Printer Friendly

Advanced control strategies for HVAC&R systems--an overview: part II: soft and fusion control.


In large commercial and residential buildings, energy management control systems (EMCS) play a major role in maintaining good control of temperature, human comfort, and overall operational and energy efficiency. In a typical situation, HVAC&R systems provide a central air supply at a controlled temperature and flow rate for heating or cooling a particular unit, zone, or entire building complex. The two main requirements of any HVAC&R system are, first, to provide satisfactory indoor comfort (temperature and relative humidity) conditions to the building housing both humans and equipment and, at the same time, minimize overall energy consumption. Another important requirement is to prevent the spread of any chemical or biological species from any point where these species are released to the rest of the building.

In this second part of a two-part series, the focus is on soft-control techniques applied to the HVAC&R field and the techniques using the synergy of hard- and soft-control strategies. The article concludes with some comments on future directions for this field.

Soft computing or control (SC)

Control techniques, in general, are categorized into hard control and soft control. The hard-computing or control techniques use proportional-integral-derivative (PID), optimal, nonlinear, adaptive, and/or robust control. On the other hand, SC is an emerging field based on the application of neural networks (NNs), fuzzy logic (FL), and genetic algorithms (GAs) and other evolutionary techniques to any field such as HVAC&R (see Jang et al. 1997; Tsoukalas and Uhrig 1997; Nguyen et al. 2003; Tettamanzi and Tomassini 2001; Karray and De Silva 2004; Konar 2005; Kasabov 2007; Sumathi et al. 2008).


NNs are engineering or mathematical representations of biological neurons that relate input and output actions and take the form of a massively connected and parallel distributed network. These are also called artificial NNs (ANNs) to distinguish them from the biological NNs. The NNs have long been recognized for the identification and control of physical systems, especially when the system models are not analytically known fully (Hunt et al. 1992; Poznyak et al. 2001). See Curtiss et al. (1994,1996) for a comparison of local (decentralized) and global (centralized) control of an HVAC&R plant using ANNs with that of PID controllers. To take care of the nonlinearities of HVAC&R plants, a weighted radial basis function (RBF) NN was used for control of the plant (Hepworth and Dexter 1994), and it was found that the performance of the plant was better than with conventional PI controllers.

For an air-handling unit (AHU) with seven variables (room temperature, room humidity, pressure behind supply air damper, supply air temperature, supply airflow rate, return airflow rate, and supply air moisture content) and five actuator variables (re-heater power output, humidifier output, supply air fan speed, chilled water flow rate, and supply air damper resistance), a NN-based system identifier and controller was designed by So et al. (1994) with 12 input layer nodes and 7 output layer nodes, which acted first as an identifier and then as a controller with the control algorithm to minimize set-point error deviation and energy consumption. To improve the operation of a multi-zone space heating (MZSH) system in terms of disturbance rejection and occupied/unoccupied set-point control, an NN-based, decentralized controller (NNDC) was designed by Saboksayr et al. (1995) to accomplish this task. The occupied (day) and unoccupied (night) set-point controls refer to the overall energy demand for buildings where the comfort of people is of importance without any reference to the equipment. In Chow and Teeter (1997) and Teeter and Chow (1998), identification and control of a single-zone thermal system were developed using a functional-link NN, where the functional links mainly expand the original input space for training with the idea of reducing the burden of the training phase of NNs.

Robustness and stability issues relating to a variable airflow volume (VAV) system in HVAC&R was studied by Song et al. (2003), who used the motor-fan model to describe the dynamics of VAV system given by Zaheer-Uddin and Zheng (1994) in terms of motor voltage, motor current, fan speed, airflow volume rate, and pressure rise. A normalized learning algorithm is used to train an NN, and an adaptive dead-zone scheme was employed to provide adaptability and robustness due to disturbance and uncertain parameters in the model with stability and convergence proofing.

For controlling the AHU in a typical HVAC&R system, Guo et al. (2007), instead of using typical PID controllers in both outer and inner loops, used an NN controller in the outer loop to make the whole system more adaptive and robust by training the multi-layer NN by a unique simultaneous perturbation stochastic approximation (SPSA) based training algorithm, which provided a guaranteed stability of the overall control system.


FL is basically a methodology to represent human knowledge and reasoning in the form of membership functions and rules to make useful inference actions for the modeling and control of uncertain physical systems (see Zadeh 1965; Mamdani 1974; Tagaki and Sugeno 1985; Abonyi 2003; Zhang and Liu 2006).

Problems, such as thermal regulation and maintenance of temperature set-point, in HVAC&R systems were addressed using a FL controller (FLC) (Huang and Nelson 1994a, 1994bb; Arima et al. 1995; Lea et al. 1996). An interesting application of the use of FL for a thermal control system, referred to as a thermal enclosure system (TES), and a commercial refrigerator/incubator module (CRIM) was developed by Lembeck (1992) and flown in a space shuttle flight in August 1992 to control the temperature in protein crystal growth experiments. The investigations in Dounis et al. (1996), Dounis and Manolakis (2001), and Dounis and Caraiscos (2007) basically focused on FLC applied to improve the thermal comfort level and indoor air quality (IAQ) in the buildings. A comparison of traditional or conventional control techniques, such as ON-OFF, PID, PI with dead-beat (PIdb), with the FLC method was done in Dounis et al. (1996), showing that the fuzzy and PIdb controllers were better than the PID or ON-OFF controllers. Two types of fuzzy controllers--the Mamdani fuzzy controller (MFC) and the Gupta fuzzy controller (GFC)--were used in the comparison.

The IAQ index was chosen based on the carbon dioxide concentration. In another work by Dounis and Manolakis (2001), by considering thermal comfort as a fuzzy concept, using mathematical models for indoor temperature in terms of heat flux through the window, auxiliary heating, heat transfer due to ventilation, heat transfer due to infiltration, and relative humidity of indoor air; wall temperature in terms of relative heat exchange, heat flow through walls including conduction, and mean-radiant temperature; thermal comfort index; and outdoor climate in terms of temperature, wind speed, and relative humidity, a simulator was developed using the Turbo-C programming language to test the FLC to regulate thermal comfort in terms of the predicted mean vote (PMV) along with the development of general guidelines for the design of an FL system. However, the problems associated with FLCs are that it is very difficult and time consuming to come up with (or tune for optimization) the FLCs to get the right set of membership functions associated with the data base (DB) and the set of pre-defined rule base; however, these FLCs are not optimized, especially if one has multiple objective functions, such as minimizing energy consumption, maximizing interior comfort, peak electrical load demand, etc. Thus, there is a real need for bringing GAs with a multi-objective feature to complement FLCs, as shown in Alcalaet al. (2003).

Without increasing thermal dissatisfaction of the occupants (in an office building with many cubicles) as measured by Fanger's predicted percentage dissatisfied-PMV (PPD-PMV) (Fanger 1972),a distributed control system was designed to minimize the energy consumption using an FL approximation approach with results that mimic the equivalent obtained by gradient-based optimization techniques, leading to high computational complexity (Ari et al. 2005). The building model used had three types of zones: corner, perimeter, and interior offices, leading to three different types of fuzzy inference systems (FISs); four neighboring office temperatures along with their own desired temperatures were used as inputs and the optimized temperature set-point as the output of the FIS.

A two-level hierarchical structure was developed in He et al. (2005) for multiple-model predictive control (MMPC) using Takagi-Sugeno (T-S) fuzzy models leading to linear time varying parameter (LPV) controllers and testing with real process simulations. A biologically motivated algorithm using a modified version of brain emotional learning (BEL) was proposed by Sheikholeslami et al. (2006) to a multi-variable, nonlinear and nonminimum phase HVAC control system to provide a robust and desired performance. An intelligent coordinator for the operation of a local PI-like FLC was proposed in Dounis and Caraiscos (2007) to provide better working conditions based on three factors: thermal comfort, visual comfort (VC), and IAQ. Further, the coordination has been achieved using two hierarchical subsystems--the master and slave agent. A fuzzy controller was designed by Chiou et al. (2009) for controlling the temperature of air-conditioners in a unitary system to realize energy savings and stable operation.

In a two-part investigation, the authors Zhang and Zhang (2009) first developed a mathematical model for a chiller with a screw compressor with both economized and non-economized modes with a provision for automatic switching between the two modes and validated the model with experiments and simulations, and then two control algorithms based on PID and FL are compared in Zhang et al. (2009a) for electronic expansion valve (EXV) controlling the suction superheat, the compressor controlling the leaving water temperature and the sub EXV regulating the injection superheat showing that the fuzzy controller providing higher reliability and better performance.

Two advanced fuzzy tuning techniques--lateral and amplitude (LA) tuning--along with fuzzy rule selection were used in Alcala et al. (2009) to improve (optimize) the FLCs with application to the control of HVAC systems.

A self-extracting rules fuzzy control (SERFC) strategy was proposed by Lu et al. (2010) for a multi-variable, large time-delay, nonlinear hot water radiator system to maintain a stable temperature using transfer function identification schemes and simulations results leading to satisfying performance.


GAs are derivate-free optimization techniques based on biological evolution theory involving crossover and mutation and survival of the fittest (Goldberg 1989; Gen and Cheng 1997; Haupt and Haupt 1998), including optimization of multi-objective functions (Deb 2001).

For a single-zone "all outside air" HVAC&R system consisting of a regenerative heat exchanger, cooling and heating coils, and supply fan, a control scheme was designed by Wright et al. (2002) to determine supply air temperature and flow rate to realize three optimization criteria--operating cost of HVAC&R, maximum thermal zone comfort, and "in-feasibility objective" (the aggregated value of constraint violations), using a multi-objective GA (MOGA) (Van Velhhuizen and Lamont 1970). The MOGA had the potential to find Pareto optimal solutions given in Engwerda (2005) for building design problems.

For global optimization of an HVAC&R system characterized by mathematical models of various components and constraints, a modified GA was used with better performance compared to other traditional control methods (Lu et al. 2005a, 2005b). In the first part by Lu et al. (2005a), for global optimization, the objective function to be minimized was formed as the sum of energy consumption of the cooling coil fans, chilled water pumps, chillers, condenser water pumps, and cooling tower fans under the constraints due to physical limitations of the components and the interaction between the components and units. The resulting mixed-integer, constraint, nonlinear optimization (MICNO) problem was transformed and simplified for solving using an adaptive neuro-FIS (ANFIS) (Jang et al. 1997; Karray and De Silva 2004), which is basically a multi-layer network with the functions of fuzzification, FIS, and defuzzification. The ANFIS had as inputs mass flow rate of chilled water to each cooling coil and as output optimal set-point of the chilled water pump head. In the second part by Lu et al. (2005b), the MICNO problem was solved using a modified GA consisting of four parts: encoding, construction of fitness function, evolution, and termination. Compared to traditional methods, this investigation showed substantial reduction in energy demand by the HVAC&R system. Also, see a closely related work in Lu et al. (2005c).

Using GAs for a multi-objective optimization problem involving supervisory control, the set-points (such as supply air temperature and chilled water supply temperature) were obtained by Nassif et al. (2005) to realize minimum energy use and maximum comfort validated with a VAV system with high-load distributed multiple zones. Using an optimal control model, parameter identification, obtained by Yan et al. (2008), resulting in a 7% reduction in energy consumption. For optimization of HVAC systems, a robust evolutionary algorithm (REA, different from the traditional GA) was developed in Fong et al. (2008), where the method emphasized both mutation and recombination, as well as selection, and used a constraint-handling operator instead of the penalty-based approach, which successfully demonstrated through an application example claiming the capability of this REA for handling different HVAC optimization problems that possess multi-modal, multi-dimensional, non-linear, continuous-discrete, and highly constrained characteristics. GA-based optimization was used in Xu et al. (2009) for a model-based optimal ventilation control scheme for a multi-zone VAV air-conditioning system with a cost function comprising thermal comfort, IAQ, and total energy consumption.

An adaptive FLC (AFLC) to control a water valve for a cooling coil in an AHU of an HVAC system was developed by Navale and Nelson (2010) using GAs for fuzzy rule matrix and membership functions along with experimental validation.

Using simplified linear models of major components along with recursive least squares (RLS) on a simulated virtual system representing the central chiller plant in a super high-rise building, a supervisory and optimal control strategy for energy efficiency using system and subsystem level characteristics and interactions was developed in Ma and Wang (2011) using GAs.

It was demonstrated that this strategy can save about 0.73%-2.55% daily energy compared to the traditional settings method.

Fusion of hard and soft controls

This study focuses on methods synergizing more than one of the methods involving NNs, FL, and GAs to capture the best features of the individual methods (Jang et al. 1997; Tsoukalas and Uhrig 1997; Nguyen et al. 2003; Karray and De Silva 2004; Konar 2005; Kasabov 2007; Sumathi et al. 2008). It further looks at the direction of developing an integrated structure by blending (see Ovaska et al. 2002) SC techniques and conventional hard-control or computing (HC) techniques comprising optimal control (Naidu 2003), model predictive control (Camacho and Bordons 2004), robust control (Sinha 2007), reconfigurable control (Benitez-Perez and Garcia-Nocetti 2005; Isermann 2006), adaptive control (Ustrom and Wittenmark 1995), networked control (Wang and Liu 2008), and resilient control systems (Mitchell and Mannan 2006; Amin and Horowitz 2007; Hollnagel et al. 2008; Weiss 2010) with specific applications to HVAC&R systems. The integration of SC and HC methodologies, shown in Figure 1, has the following attractive features (see Ovaska et al. 2002; Tettamanzi and Tomassini 2001):

1. The methodology based on SC can be used, in particular with FL, at upper levels of the overall mission, where human involvement and decision making is of primary importance, whereas the HC can be used at lower levels for accuracy, precision, stability, and robustness.

2. In another situation using a hybrid scheme, an NN of the SC is used to supplement the control provided by a linear, fixed-gain controller for a missile autopilot.

3. Further, the SC-based GA can be used for the parameter tuning of a PID controller to achieve good performance and robustness for a wide range of operating conditions.

4. SC and HC are potentially complementary methodologies.

5. The fusion could solve problems that cannot be solved satisfactorily by using either methodology alone.

6. Novel synergetic combinations of SC and HC lead to high performance, robust, autonomous, and cost-effective missions.


The fusion of GA and FL, resulting in the genetic fuzzy optimization method, was investigated in Parameswaran et al. (2010) for a combined variable refrigerant volume (VRV) and VAV air-conditioning system to address enhanced thermal comfort and IAQ. The experimentation involving three distinct supply air temperature categories yielded daily energy savings of the order of 54% in summer and 61% in winter. Fusion among the soft control methods of NN, FL, and GA, resulting in a hybrid GA hierarchical adaptive network based FIS (GA-HANFIS), was applied in Li and Su (2010) to predict the daily air conditioning consumption for a hotel, where the objective function in the GA is to minimize the coefficient of variation (CV) of predictions made by the HANFIS. It was further demonstrated that this fusion methodology gives a better performance than that using NN alone.

A simple fuzzy PID (FPID) design strategy was proposed by Jantzen (1998) consisting of the following steps:

1. tune a PID controller using Ziegler-Nichols (Z-N) method of Ziegler and Nichols (1942),

2. replace the summation in PID control with an equivalent linear fuzzy controller acting like a summation,

3. transfer the PID gains to the linear FPID controller,

4. make the linear FPID controller nonlinear by introducing fuzzy rule base, and

5. fine- (hand-) tune the resulting FPID controller based on intuition and experience.

An NN for predicting the building electric demand and GA-based supervisory control scheme was developed in Gibson (1997) with application to a high school in California. Automatic tuning of PID controllers for HVAC&R systems was achieved using an adaptive learning algorithm based on GAs by Huang and Lam (1997) with performance indicators as overshoot, settling time, and mean squared error for evaluating controller performance using the modular dynamic optimization software package HVACSIM+. A simple and practical hybrid fuzzy sliding mode adaptive control method for HVAC&R was presented in Ying-Guo et al. (1998), and the superiority of the method was demonstrated by comparing the results of fuzzy adaptive with a simple PID and FPID methods. In an HVAC&R system, the supply air pressure is controlled by the supply air fan driven by a variable-speed drive modeled as a second-order dynamics with dead time.

Here, these methods are combined in such a way to retain the most attractive features of each method. For example, one can use the GA to optimize the weights of NNs and/or fuzzy rules and membership functions; a hybrid synergy of FL and GA to optimally combine NNs is proposed in Choi (2002). In order to optimize energy consumption and user comfort in an HVAC&R system, an integrated (hybrid) control architecture was designed by optimizing fuzzy controller parameters using GAs (see Pargfrieder and Jorgl 2002). In addition, a generalized predictive controller (GPC) is used to identify the room model and designing an adaptive controller. In Chow et al. (2002), a study was done based on integrating an NN (for modeling system characteristics) and a GA (for global optimization) for the optimal control of fuel and electricity in a direct-fired absorption chiller system.

Using a third-order, nonlinear model composed of the thermal space temperature, the thermal space humidity ratio and the supply air temperature and as three state variables, the first two states as two output variables, and the volumetric airflow rate and chilled water flow rate as two control (input) variables, a hybrid method was developed by Rahmati et al. (2003) consisting of the traditional PID controller and an FPID controller, which showed that the hybrid method gave a more robust performance. In Alcala et al. (2003), GAs are used to tune intelligently the FLC to control an HVAC&R system satisfying the requirements of energy saving and indoor comfort. On the other hand, weighted linguistic fuzzy rules and the corresponding rules selection were addressed using an evolutionary optimization technique, such as the GA, in Alcala et al. (2005) based on the previous results for weighted fuzzy rules (Cho and Park 2000) the and use of GAs for selecting FL rules, as shown in Ishibuchi et al. (1995). Further, it is worth noting from Alcala etal. (2005) that the testing of the control algorithms at a real test site under the generic embedded system (GENESYS) project was within the framework of the JOULE-THERMIE program.

Using FL to characterize the PMV as in Fanger (1972) and ISO-1995 (1995) for assessing the thermal comfort conditions, the work by Calvino et al. (2004) developed a PID-based adaptive fuzzy controller with an experimental application for a wintertime period in an office room belonging to the University of Palermo, Palermo, Italy. Recognizing the two limitations of requiring significant storage space for fuzzy rules and the need for switching fuzzy rules based on changes in outside temperature in their earlier work (see Ari et al. 2005) to minimize the energy consumption without increasing thermal discomfort, the researchers modified their work in Ari et al. (2006) using FL and NNs to investigate the performance of FL approximation.

Using a simple fuzzy rule base for errors between the set-point and output, and training the parameters of an NN, an adaptive neuro-fuzzy (ANF) method was developed by Jian and Wenjian (2000) to be used along with a secondary loop with a PID controller to regulate the supply air pressure, which was found to offer a better performance compared to a pure PID controller alone.

An FPID controller was developed for a complex chilling unit in the overall building energy management system (BEMS) by Talebi-Daryani (2001) and was found to be superior to the distributed digital controller (DDC). Here the chilling system supplies chilled water to the air-conditioning systems installed in different research labs and computer rooms at the Max Plant Institute for Radio Astronomy in Bonn. The investigation in Zaheer-Uddin and Zheng (2001) focused on determining room temperature set-point scheduling to minimize the total energy consumption and demand costs subject to the constraints (dynamics of the system, room temperature comfort range, heat exchanger temperature range, and load limit), uncertainties (uncontrolled and thermal loads), and objective function. Here, the dynamics of the system was predicted by using NNs, and the multi-stage stochastic optimization problem was solved using Lagrangian relaxation (LR; a decomposition and relaxation approach) and a stochastic dynamic programming (SDP) approach. Four examples were presented with two single-zone HVAC&R units focusing on load shutting function, computational efficiency, quality of the LR method, and implementation to an industrial building.

Using the fuzzy-GA (FGA) technique, a method was proposed in Lo et al. (2007) for automatic fault detection (AFD) for an HVAC system verified by simulations. A model-based supervisory control strategy was developed and customized, as shown in Ma et al. (2008), for the currently existing high-rise commercial office building in Kowloon, Hong Kong, for online control and operation of building central cooling water systems using simplified models for the chiller and two cooling towers (CTA and CTB) and the performance index to be minimized, consisting of power consumption for chiller and the cooling towers. Here, a hybrid optimization technique, called the performance map and exhaustive search (PMES) based method, was developed to seek (steady-state) optimal solutions and compared with that searched by using GAs. It was found that the developed strategy was more energy efficient and computationally cost effective, making it simple and easy for implementation for online applications.

A related work with a simple HVAC&R experimental setup is shown (see Rieger 2008) in Figure 2. A graphical user interface or human machine interface (HMI) was used to perform all experiments as shown in Figure 3. As a major goal of this HVAC&R control system is to optimize the pressure gradients and associated flows for the facility, a linear quadratic tracker (LQT) method provides a time-based approach to guiding facility interactions. However, LQT methods are susceptible to modeling and measurement errors, and therefore, the additional use of soft-computing methods is proposed for implementation to account for errors and non-linearities.

The resulting hybrid (soft and hard) controller design provides a framework for supervisory control of an HVAC&R system and shows that significant improvements can be achieved when the overall plant dynamics of the ventilation system are considered in the control system design. The HVAC&R work performed at Idaho State is summarized below (Rieger 2008).

1. Advanced control methodologies and a resulting hybrid controller was proposed and tested for differential pressure control of Department of Energy (DOE) facilities, which was unique compared to other HVAC&R control research applications in design and focus (see Rieger and Naidu 2005).

2. The analytical solution to the Riccati equation for the linear quadratic regulator (LQR) nonsteady-state design was extended to the LQT design. This extension allowed the development of an automated mechanism for finding a controller system of any order and tracking reference, which was performed and tested using MATLAB[C] (see Rieger and Naidu 2004, 2005).

3. An NN was proposed as a vehicle for the capturing of LQT data in off-the-shelf control systems, providing a straightforward framework for implementation, as shown in Rieger and Naidu (2005).

4. A hybrid controller design using soft-computing and hard-computing techniques was proposed, tested, and implemented. The hybrid controller provides a pathway for allowing both faster performance and inclusion of temperature controls in the differential pressure makeup of DOE facilities (see Rieger and Naidu 2005, 2008).



A T-S-based neuro-fuzzy network was used in Huang and Dexter (2008) for nonlinear robust model-based predictive control (MPC) of the temperature of an experimental AHU.

A review was conducted by Dounis and Caraiscos (2009) on the work on the state-of-the art intelligent (neural, fuzzy, neuro-fuzzy, PI-fuzzy, adaptive fuzzy PD and PID) and conventional (classical, optimal, predictive and adaptive) control systems for improving the efficiency and indoor environment in buildings with particular emphasis on multi-agent control systems (MACS) with simulations using TRNSYS/MATLAB software.

For controlling the temperature and humidity of an HVAC system using PID controllers, an intelligent approach for modeling and the control of the system was achieved by Soyguder and Alli (2009b), using ANFIS, leading, in particular, to a more accurate prediction of damper gap rate and faster and simplified solutions. A related work by Soyguder and Alli (2009a) on the application of self-tuning PID-fuzzy adaptive controller to an HVAC system showed improvement over the classical PID controller. A programmable logic controller (PLC) based PID controller with optimal tuning of the controller parameters using fuzzy sets was designed by Soyguder and Alli (2010) for controlling the two different damper gap rates of an HVAC system leading to a fuzzy adaptive control (FAC) method.

A flexible software framework using GenOpt1 along with a GA was developed in Coffey et al. (2010) for model predictive control of a zone temperature ramping in an office space.

GenOpt R is an optimization program for the minimization of a cost function that is evaluated by an external simulation program, such as EnergyPlus, TRNSYS, SPARK, IDA-ICE, or DOE-2

Internet-based HVAC&R

Here the focus is on control methods related to the issues of fault-detection, reliability, Internet (cyber) security, etc. In particular, the Internet-based HVAC&R refers to the situation where the information between the various elements, such as sensors and actuators, is channeled via Internet and/or the complete systems being networked using Internet.

Fault detection

A two-part review on automated fault detection and diagnostics (FDD) along with prognostics was by Katipamula and Brambley (2005a, 2005b) was provided with specific applications to the HVAC&R field. The first part of the review focuses on generic FDD and prognostics, providing various methods and their primary strengths and weaknesses, whereas the second part reviews research and specific applications to HVAC&R fields, such as refrigerators, air conditioners and heat pumps, chillers, and AHUs.

For modeling HVAC systems using measured data, principle component analysis (PCA) was successfully demonstrated in Hao et al. (2005), which splits the measurement space into two subspaces of principle component subspace (PCS) and residual subspace (RS), resulting in the recovery of faulty or missed data to provide a reliable and optimal control structure for the HVAC system.

A fault tolerant control method was presented by Jin and Du (2006) to control the outdoor air ventilation and AHU supply air temperature (which affect IAQ and humidity) to satisfy the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) standard for VAV systems using PCA, the joint angle method, and compensatory reconstruction for detection, isolation, and reconstruction of the fault with testing and evaluation. Using a sequencing strategy for AHUs based on finite-state machine logic and a model-based fault detection system, a new integrated control and fault detection system was described in Seem and House (2009) through simulations of 16 faults relating to temperature sensor offset, stuck and leaking damper, and stuck and leaking valves with the ultimate goal of achieving AFD strategy.

Using a dynamic model in state space form (Talukdar and Patra 2010), stuck faults of the damper vanes of the AHU were detected, isolated, and estimated, and the fault identification scheme was implemented via an interactive multiple observe method using the control distribution concept, thus preventing the convergence of two faults at two different locations in the air volume box to the same value.

Reliability and Internet (cyber) security

A general framework was presented by Huang et al. (2008) to enhance the reliability of the chilled sequencing control using the fused measurement involving the combination of two approaches: one direct approach measurement using differential water temperature and water flow and the other indirect measurement of building cooling load based on chiller electrical power input. It was demonstrated by a case study that the proposed framework effectively eliminates measurement noises, outliers, systematic errors, and model errors present in the individual direct and indirect approaches.

An experimental system was constructed in Lin and Brogerg (2002) to investigate monitoring and control of the proposed Internet-based HVAC&R systems consisting of the major modules: sensors and actuators, DB, data acquisition, processing and signal conditioning, input/output (I/O) links and devices, network interface, common gateway interface (CGI)-accessed control programs, and a Web server using the transmission control protocol/Internet protocol (TCP/IP) and Web technologies for achieving the goals of combining the functions of control, programming, operator interface, data collection, analysis and management and provide real-time monitoring and control capability. It was found that the proposed Internet-based scheme is flexible and scalable, and easy integration of multi-vendor equipment.

In a typical building automation and control system (BACS), shown in Novak and Gerstinger (2009), with services such as HVAC&R, lighting, and shading, the safety- and security-critical services (SSCS) are integrated with the building automation technology with an SSCS package consisting of a fire alarm, access control, and standard HVAC&R service.

In Zhang et al. (2009b), a personal digital assistant (PDA) based intelligent building service system was developed to control HVAC&R components, proposing that the system can be connected to the Ethernet network and communicate with PDAs or computers using the BACnet protocol for the communication process between a system server and other connected smart devices.

As industrial control systems (ICS), such as HVAC&R systems, are being continuously upgraded with advanced communication capabilities and networked to improve process efficiency, productivity, regulatory compliance, and safety, there is a real need for some kind of cyber security (different from physical security and information technology [IT] security) to avoid any minor to catastrophic damages to the whole ICS (Weiss 2010). This is more critical in the wake of planned "Smart Grid" connection with the Internet of the electric grid and all the devices and equipment that work with electricity at home, office, and industry (see SMART-GRID 2008).

In summary, it is noted that fault detection and reliability and Internet security are closely related important topics for HVAC&R. Recognizing that a fault is different from a disturbance, one can design a controller that can accommodate or is tolerant of the fault. If a fault-tolerant or faculty-accommodation controller is no longer able to manage the system under these faulty conditions, the supervisory-level control design switches to a new control configuration, leading to reconfiguration control (Blanke et al. 2006). Internet (cyber) attack and security are important issues for infrastructure facilities, such as HVAC&R, and there is urgent need to probe further to identify, isolate, and eliminate these attacks.

Conclusions and future directions


In this article, an overview of a topical survey on the control strategies for HVAC&R systems was presented. In particular, this overview focused on

1. hard-control techniques, such as PID, optimal, robust and adaptive;

2. soft-control techniques, such as NNs, FL, and GAs;

3. the fusion of hard- and soft-control techniques; and

4. a brief description of HVAC&R experimental setup at Idaho State University.

The major outputs of the overview in terms of contributions of the hard- and soft-control techniques to the HVAC&R field are summarized below.

1. The main applications to the HVAC&R field come from hard-control techniques, such as PID, optimal, and adaptive. However, the optimal control method, due to its attractive features of energy saving and thermal comfort, dominated the applications to the field.

2. On the other hand, the application of soft-control techniques to the HVAC&R field has accelerated in the recent years due to their attractive features of nonlinear identification and control (NNs), human knowledge, and reasoning in the form of membership functions and rules (FL) and global nonderivative-based optimization (GAs). In particular, it appears that the NN, although useful in cases where there is no mathematical model, suffers from the enormous time taken for off-line training. On the other hand, fuzzy and neuro-fuzzy techniques capable of tuning fuzzy PI controllers are adequate for most HVAC applications. GAs are attractive for optimization purposes without involving the mathematical theory, such as calculus or calculus of variations; however, they are not always suitable for real-time HVAC applications. See Dounis and Caraiscos (2009) for further views on this topic.

3. The major contribution of this overview is to identify the application of the technique fusion of hard and soft control to the HVAC&R field, where the best features of hard and soft control techniques are captured.

Future directions

Some of the areas identified for future investigations in the area of HVAC&R are (see Sane et al. 2006):

1. More accurate, physics-based dynamic (both lumped- and distributed-parameter) models for both existing and future building systems.

2. Advanced algorithms for optimal, model predictive, robust, reconfigurable, adaptive, networked, and resilient control systems with industry-standard embedded platforms for the integrated building systems.

3. Advanced algorithms based on soft computing techniques, including NNs, FL, genetic logic, genetic programming, swarm intelligence, probabilistic reasoning, and others.

4. More sophisticated HVAC&R software packages for modeling, analysis, design, development, testing, and validation with capabilities of modularity and integration with Internet connectivity.

5. All the research and development work stated above needs to focus on the critical issues of thermal comfort for both people and equipment, energy usage, feasibility, compliance with industry standardization, physical and cyber security, and cost.


AFD = automatic fault detection

AFLC = adaptive fuzzy logic controller

AHU = air-handling unit

ANF = adaptive neuro-fuzzy

ANFIS = adaptive neuro-fuzzy inference system

ANN = artificial neural network

ASHRAE = American Society of Heating, Refrigerating and Air-Conditioning Engineers

BACS = building automation and control systems

BEL = brain emotional learning

BEMS = building energy management system

CGI = common gateway interface

CRIM = commercial refrigerator/incubator module

CV = coefficient of variation

DB = data base

DDC = distributed digital controller

DOE = Department of Energy

EMCS = energy management control systems

EXV = electronic expansion valve

FAC = fuzzy adaptive control

FDD = fault detection and diagnostics

FGA = fuzzy genetic algorithm

FIS = fuzzy inference system

FL = fuzzy logic

FLC = fuzzy logic controller (or fuzzy logic control)

FPID = fuzzy proportional-integral derivative

GA = genetic algorithms

GA-HANFIS = genetic algorithm hierarchical adaptive network based fuzzy inference system

GENESYS = generic embedded system

GFC = Gupta fuzzy controller

GPC = generalized predictive controller

GUI = Graphical User Interface

HANFIS = hierarchical adaptive network based fuzzy inference system

HC = hard computing or control

I/O = input/output

IAQ = indoor air quality

ICS = industrial control systems

IT = information technology

LA = lateral and amplitude

LPV = linear time varying parameter

LQR = linear quadratic regulator

LQT = linear quadratic tracker

LR = Lagrangian relaxation

MACS = multi-agent control systems

MFC = Mamdani fuzzy controller

MICNO = mixed-integer, constraint, nonlinear optimization

MMPC = multiple-model predictive control

MOGA = multi-objective genetic algorithm

MPC = model-based predictive control

MZSH = multi-zone space heating

NLP = nonlinear programming

NN = neural network

NNDC = neural network based, decentralized controller

PCA = principle component analysis

PCS = principle component subspace

PDA = personal digital assistant

PID = proportional-integral-derivative

PIdb = proportion integral with dead-beat

PLC = programmable logic controller

PMES = performance map and exhaustive search

PMV = predicted mean vote

PPD = predicted percentage dissatisfied

RBF = radial basis function

REA = robust evolutionary algorithm

RLS = recursive least squares

RS = residual subspace

SC = soft computing or control

SDP = stochastic dynamic programming

SERFC = self-extracting rules fuzzy control

SPSA = simultaneous perturbation stochastic approximation

SSCS = safety- and security-critical services

TCP/IP = transmission control protocol/Internet protocol

TES = thermal enclosure system

T-S = Takagi-Sugeno

VAV = variable airflow volume

VC = visual comfort

VRV = variable refrigerant volume

Z-N = Ziegler-Nichols

DOI: 10.1080/10789669.2011.555650


Abonyi, J. 2003. Fuzzy Model Identification for Control. Boston, MA: Birkhauser.

Alcala, R., J. Alcala-Fdez, M. Gacto, and F. Herrera. 2009. Improving fuzzy logic controllers obtained by experts: A case study in HVAC systems. Applied Intelligence 31:15-30.

Alcala, R., J. Benitez, J. Casillas, O. Cordon, and R. Perez. 2003. Fuzzy control of HVAC systems optimized by genetic algorithms. Applied Intelligence18:155-77.

Alcala, R., J. Casillas, O. Cordon, A. Gonzalez, and F. Herrera. 2005. A genetic rule weighting and selection process for fuzzy control of heating, ventilating and air conditioning systems. Engineering Applications of Artificial Intelligence 18:279-96.

Amin, S., and B. Horowitz. 2007. Flexibility with Business Excellence in the Knowledge Economy, Toward agile and resilient large-scale systems: Adaptive robust national/international infrastructures, pp. 1-22. Delhi, India: Global Institute of Flexible Systems Management.

Ari, S., I. Cosden, H. Khalifa, J. Dannenhoffer, P. Wilcoxen, and C. Isik. June 2005. Constrained fuzzy logic approximation for indoor comfort and energy optimization. Proceedings of the 2005 Annual meeting of the North American Fuzzy Information Processing Society (NAFIPS), Detroit, MI, pp. 500-4.

Ari, S., H. Khalifa, J. Dannenhoffer, P. Wilcoxen, and C. Isik. June 2006. Fuzzy logic and neural network approximation to indoor comfort and energy optimization. Proceedings of the Annual meeting of the North American Fuzzy Information Processing Society (NAFIPS), Montreal, Canada, pp. 692-5.

Arima, M., E. Hara, and J. Katzberg. May 1995. A fuzzy logic and rough sets controller for HVAC systems. Proceedings of the IEEE WESCANEX Conference Communications, Power, and Computing, Winnipeg, Canada, pp. 133-8.

Ustrom, K., and B. Wittenmark. 1995. Adaptive Control, 2nd ed., Reading, MA: Addison-Wesley Publishing Company, Inc.

Benitez-Perez, H., and F. Garcia-Nocetti. 2005. Reconfigurable Distributed Control. London, UK: Springer-Verlag.

Blanke, M., M. Kinnaert, J. Lunze, and M. Staroswiecki. 2006. Diagnosis and Fault-Tolerant Control, 2nd ed. Berlin, Germany: Springer-Verlag.

Calvino, F., M. Gennusa, G. Rizzo, and G. Scaccianoce. February 2004. The control of indoor thermal comfort conditions: Introducing a fuzzy adaptive controller. Energy and Buildings 36(2):97-102.

Camacho, E., and C. Bordons. 2004. Model Predictive Control, 2nd ed. London, UK: Springer-Verlag.

Chiou, C., C. Chiou, C. Chu, and S. Lin. February 2009. The application of fuzzy control on energy saving for multiunit room air-conditioners. Applied Thermal Engineering 29(2-3):310-6.

Cho, J.-S., and D.-J. Park. 2000. Novel fuzzy logic control based on weighting of partially inconsistent rules using neural network. Journal of Intelligent & Fuzzy Systems 8(2):99-110.

Choi, S.-B. 2002. Fusion of neural networks with fuzzy logic and genetic algorithm. Integrated-Computer-Aided Engineering 9:363-72.

Chow, M.-Y., and J Teeter. June 9-12 1997. Reduced-order functional link neural network for HVAC thermal system identification and modeling. Proceedings of the International Conference on Neural Networks, Houston, TX, pp. 5-9.

Chow, T., G. Zhang, Z. Lin, and C. Song. September 2002. Global optimization of absorption chiller system by genetic algorithm and neural network. Energy and Buildings 34(1):103-9.

Coffey, B., F. Haghighat, E. Morofsky, and E. Kutrowski. July 2010. A software framework for model predictive control with Gen Opt. Energy and Buildings 42(7):1084-92.

Curtiss, P., J. Kreider, and M. Brandemuehl. June 29-July 1,1994. Local and global control of commercial building HVAC systems using artificial neural networks. Proceedings of the 1994 American Control Conference, Baltimore, MD, pp. 3029-44.

Curtiss, P., G. Shavit, and M. Brandemuehl. 1996. Neural networks applied to buildings--a tutorial and case studies in prediction and adaptive control. ASHRAE Transactions 102(7):1141-6.

Deb, K. 2001. Prentice Hall Series on Environmental and Intelligent Manufacturing Systems, Vol. 5, Multi-objective optimization using evolutionary algorithms. Chichester, UK: John Wiley & Sons.

Dounis, A., M. Bruant, M. Santamouris, G. Guaraccino, and P. Michel. 1996. Comparison of conventional and fuzzy control of indoor air quality in buildings. Journal of Intelligent and Fuzzy Systems 4:131-40.

Dounis, A., and C. Caraiscos. July 23-26, 2007. Intelligent coordinator of fuzzy controller-agents for indoor environment control in buildings using 3-d fuzzy comfort set. Proceedings of the IEEE International Fuzzy Systems Conference, London, UK, pp. 1-6.

Dounis, A., and C. Caraiscos. August-September 2009. Advanced control systems engineering for energy and comfort management in a building environment: A review. Renewable and Sustainable Energy Reviews 13(6-7):1246-61.

Dounis, A., and D. Manolakis. June 2001. Design of a fuzzy system for living space thermal-comfort regulation. Energy and Buildings 69(2):119-44.

Engwerda, J. 2005. LQ Dynamic Optimization and Differential Games. New York: John Wiley & Sons.

Fanger, P. 1972. Thermal Comfort: Analysis and Applications in Environmental Engineering. New York: McGraw-Hill.

Fong, K., V Hanby, and T. Chow. September 2008. A robust evolutionary algorithm for HVAC engineering optimization. HVAC&R Research 14(5):683-705.

Gen, M., and R. Cheng. 1997. Genetic Algorithms and Engineering Design. New York: John Wiley & Sons.

Gibson, G. 1997. A supervisory controller for optimization of building central cooling systems. ASHRAE Transactions 103(1):486-493.

Goldberg, D. 1989. Genetic Algorithms in Search, Optimization and Machine Learning. Reading, MA: Addison-Wesley.

Guo, C., Q. Song, and W. Cai. February 2007. A neural network assisted cascade control system for air handling unit. IEEE Transactions on Industrial Electronics 54(1):620-8.

Hao, X., G. Zhang, and Y. Chen. February 2005. Fault-tolerant control and data recovery in HVAC monitoring system. Energy and Buildings 37(2):175-80.

Haupt, R., and S. Haupt. 1998. Practical Genetic Algorithms. New York: John Wiley & Sons, Inc.

He, M., W.-J. Cai, and S.-Y. Li. January 2005. Multiple fuzzy model-based temperature predictive control for HVAC systems. Information Sciences 169:155-74.

Hepworth, S., and A. Dexter. August 24-26, 1994. Neural control of non-linear HVAC plant. Proceedings of the Third IEEE Conference on Control Applications, Glasgow, UK, pp. 1849-54.

Hollnagel, E., C. Nemeth, and S. Dekker. 2008. Resilience Engineering Perspectives: Volume 1: Remaining Sensitive to the Possibility of Failure. Williston, VT: Ashgate.

Huang, G., and A. Dexter. June 2008. Realization of robust non-linear model predictive control by off-line optimization. Journal of Process Control 18(5):431-8.

Huang, G., S. Wang, and Y. Sun. November 2008. Enhancing the reliability of chiller control using fused measurement of building cooling load. HVAC&R Research 14(6):941-58.

Huang, S., and R. Nelson. 1994a. Rule development and adjustment strategies of a fuzzy logic controller for an HVAC system: Part one--analysis. ASHRAE Transactions 100(1):841-50.

Huang, S., and R. Nelson. 1994b. Rule development and adjustment strategies of a fuzzy logic controller for an HVAC system: Part two--experiment. ASHRAE Transactions 100(1):851-6.

Huang, W., and H. Lam. 1997. Using genetic algorithms to optimize controller parameters for HVAC systems. Energy and Buildings 26(3):277-82.

Hunt, K., D. Sbarbaro, R. Zbikowski, and P. Gawthrop. 1992. Neural networks for control systems--a survey. Automatica 28:1083-112.

Isermann, R. 2006. Fault-Diagnosis Systems: An Introduction from Fault Detection to Fault Tolerance. Berlin: Springer-Verlag.

Ishibuchi, H., K. Nozaki, N. Yamamoto, and H. Tanaka. August 1995. Selecting fuzzy if-then rules for classification problems using genetic algorithms. IEEE Transactions on Fuzzy Systems 3(2):260-70.

ISO. 1995. ISO-1995, Moderate thermal environments-determination of the PMV and PPD indices and specification of the conditions for thermal comfort. Geneva, Switzerland: International Standardization Organization.

Jang, J.-S., C.-T. Sun, and E. Mizutani. 1997. Neuro-Fuzzy and Soft Computing: A Computational Approach to Learning and Machine Intelligence. Upper Saddle River, NJ: Prentice Hall PTR.

Jantzen, J. September, 30, 1998. Tuning of fuzzy PID controllers. Technical report, Department of Automation Department of Automation, Technical University of Denmark, Lyngby, Denmark.

Jian, W., and C. Wenjian. October 8-11,2000. Development of an adaptive neuro-fuzzy method for supply air pressure control in HVAC system. Proceedings of the2000 IEEE International Conference on Systems, Man, and Cybernetics, Nashville, TN, pp. 3806-9.

Jin, X., and Z. Du. August 2006. Fault tolerant control of outdoor air and AHU supply air temperature in VAV air conditioning systems using PCA method. Applied Thermal Engineering 26(11-2):1226-37.

Karray, F., and C. De Silva. 2004. Soft Computing and Intelligent Systems Design: Theory, Tools and Applications. Harlow, England, UK: Pearson Educational Limited.

Kasabov, N. 2007. Evolving Connectionist Systems: The Knowledge Engineering Approach, 2nd ed. London, UK: Springer-Verlag.

Katipamula, S., and M. Brambley. January 2005a. Methods for fault detection, diagnostics, and prognostics for building systems: A review, Part I. HVAC&R Research 11(1):3-25.

Katipamula, S., and M. Brambley. April 2005b. Methods for fault detection, diagnostics, and prognostics for building systems: A review, Part II. HVAC&R Research 11(2):169-87.

Konar, A. 2005. Computational Intelligence: Principles, Techniques and Applications. Berlin, Germany: Springer-Verlag.

Lea, R., E. Dohmann, W. Prebilsky, and Y. Jani. September 8-11, 1996. An HVAC fuzzy logic zone control system and performance results. Proceedings of the Fifth IEEE International Conference on Fuzzy Systems, New Orleans, LA, pp. 2175-80.

Lembeck, M. March 1992. TES/CRIM Fuzzy Logic Thermal Control System, Presented to NASA/JSC Software Technology Branch.

Li, K., and H. Su., 2010. Forecasting building energy consumption with hybrid genetic algorithm-hierarchical adaptive network-based fuzzy inference system. Energy and Buildings 42(11):2070-6.

Lin, P.-H., and H. Brogerg. January-February 2002. Internet-based monitoring and controls for HVAC applications. IEEE Industry Applications Magazine 8(1):49-54.

Lo, C., P. Chan, Y. Wong, A. Rad, and K. Cheung. March 2007. Fuzzy-genetic algorithm for automatic fault detection in HVAC systems. Applied Soft Computing 7(2):554-60.

Lu, L., W. Cai, Y. Chai, and L. Xie. May 2005a. Global optimization for overall HVAC systems. Part I: Problem formulation and analysis. Energy Conversion and Management 46:999-1014.

Lu, L., W. Cai, Y. Chai, and L. Xie. May 2005b. Global optimization for overall HVAC systems. Part II: Problem solution and simulations. Energy Conversion and Management 46:1015-28.

Lu, L., W. Cai, L. Xie, S. Li, and Y. Soh. 2005c. HVAC system optimization--in building section. Energy and Buildings 37:11-22.

Lu, Z., J. Zhang, Y. Chen, T. Zhao, and H. Liu. March 2010. Fuzzy control model and simulation of supply airsystem in a test rig of low-temperature hot-water radiator system. Energy and Buildings 42:386-92.

Ma, Z., and S. Wang. 2011. Supervisory and optimal control of central chiller plants using simplified adaptive models and genetic algorithm. Applied Energy88:198-211.

Ma, Z., S. Wang, X. Xu, and F. Xiao. 2008. A supervisory control strategy for building cooling water systems for practical and real time applications. Energy Conversion & Management 49:2324-36.

Mamdani, E. December 1974. Applications of fuzzy algorithms for control of a simple dynamic plant. IEE Proceedings 121(12):1585-8.

Mitchell, S., and M. Mannan. April 2006. Designing resilient engineered systems. Chemical Engineering Progress 102(4):39-45.

Naidu, D. 2003. OptimalControlSystems. BocaRaton, FL: CRC Press.

Nassif, N., S. Kajl, and R. Sabourin. July 2005. Optimization of HVAC control system strategy using two-objective genetic algorithm. HVAC&R Research 11(3):459-86.

Navale, R.L., and R.M. Nelson. May 2010. Use of genetic algorithms to develop an adaptive fuzzy logic controller for a cooling coil. EnergyandBuildings 42(5):708-16.

Nguyen, H., N. Prasad, C. Walker, and E. Walker. 2003. A First Course in Fuzzy and Neural Control. Boca Raton, FL: Chapman & Hall/CRC.

Novak, T., and A. Gerstinger. 2009. Safety- and security-critical services in building automation and control systems. IEEE Transactions on Industrial Electronics 57:3614.

Ovaska, S., H. Van Landingham, and A. Kamiya. May 2002. Fusion of soft computing and hard computing in industrial applications: An overview. IEEE Transactions on Systems, Man, and Cybernetics--Part C: Applications and Reviews 32(2):72-9.

Parameswaran, R.P., R.K.C.V.R. Kumar, and S. Iniyan. 2010. Energy conservative building air conditioning system controlled and optimized using fuzzy-genetic algorithm. Energy and Buildings 42:745-62.

Pargfrieder, J., and H. Jorgl. 2002. An integrated control system for optimizing the energy consumption and user comfort in buildings. Proceedings of the 2002 IEEE International Symposium on Computer Aided Control System Design, Glasgow, Scotland, pp. 127-32.

Poznyak, A., E. Sanchez, and W. Yu. 2001. Differential Neural Networks for Robust Nonlinear Control: Identification, State Estimation and Trajectory Tracking. Singapore: World Scientific Publishing Co.

Rahmati, A., F. Rashidi, and M. Rashidi. October 5-8, 2003. A hybrid fuzzy logic and PID controller for control of nonlinear HVAC systems. Proceedings of the 2003 IEEE International Conference on Systems, Man and Cybernetics, Washington, DC, pp. 2249-54.

Rieger, C., August 2008. Advanced control strategies for heating, ventilation and air conditioning (HVAC) systems in critical building structures. Ph.D. thesis, Measurement and Control Engineering, Idaho State University.

Rieger, C., and D. Naidu. 2004. Linear quadratic regulator and tracking control algorithms implemented in MATLAB. Technical report, Measurements and Controls Engineering Research Center (MCERC), Idaho State University, Pocatello, ID.

Rieger, C., and D. Naidu. June 2005. Implementation of a hybrid controller for ventilation control using soft computing. Proceedings of the2005 American Control Conference, Portland, OR, pp. 2245-50.

Rieger, C., and D. Naidu. October 2008. Implementation of a hybrid controller for critical building HVAC systems using soft computing. Proceedings of the Sixth IASTED International Conference on Intelligent Systems and Control, Orlando, FL, pp. 126-33.

Saboksayr, S., R. Patel, and M. Zaheer-uddin. June 1995. Energy-efficient operation of HVAC systems using neural network based decentralized controllers. Proceedings of the American Control Conference, Seattle, WA, pp. 4321-5.

Sane, H., C. Haugstetter, and S. Bortoff, June 14-16,2006. Building HVAC control systems--role of controls and optimization. Proceedings of the 2006 American Control Conference, Minneapolis MN, pp. 1121-6.

Seem, J., and J House., January 2009. Integrated control and fault detection of air-handling units. HVAC&R Research 15(1):25-55.

Sheikholeslami, N., D. Shahmirzadi, E. Semsar, C. Lucas, and M. Yazdanpanah. January 2006. Applying brain emotional learning algorithm for multivariable control of HVAC systems. Journal of Intelligent and Fuzzy Systems 17(1):35-46.

Sinha, A. 2007. Linear Systems: Optimal and Robust Control. Boca Raton, FL: Taylor & Francis Group.

SMARTGRID. 2008. The Smart Grid: An Introduction. Washington, DC: U.S. Department of Energy.

So, A., W. Chan, T. Chow, and W. Tse. 1994. A neural-network-based identifier/controller for modern HVAC control. ASHRAE Transactions Research 101(2):14-31.

Song, Q., W. Hu, and T. Zhao. March 2003. Robust neural network controller for variable airflow volume system. IEE Proceedings Control Theory and Applications 150:112-8.

Soyguder, S., and H. Alli. April 2009a. Design and simulation of self-tuning PID-type fuzzy adaptive control for an expert HVAC system. Expert Systems with Applications 36(3):4566-73.

Soyguder, S., and H. Alli. August 2009b. An expert system for the humidity and temperature control in HVAC systems using ANFIS and optimization with fuzzy modeling approach. Energy and Buildings 41(8):814-22.

Soyguder, S., and H. Alli. March 2010. Fuzzy adaptive control for the actuators position control and modeling of an expert system. Expert Systems with Applications 37(3):2072-80.

Sumathi, S., T. Hamsapriya, and P. Surekha. 2008. Evolutionary Intelligence: An Introduction to Theory and Applications with MATLAB[C]. Berlin, Germany: Springer-Verlag.

Tagaki, T., and M. Sugeno. 1985. Fuzzy identification of systems and its applications to modeling and control. IEEE Transactions on Systems, Man, and Cybernetics 15(1):116-32.

Talebi-Daryani, R. 2001. Intelligent Control Systems Using Soft Computing Methodologies, Application of fuzzy logic for control of heating, chilling, and air-conditioning systems. A. Zilouchian and M. Jamshidi, eds., pp. 261-290. Boca Raton, FL: CRC Press.

Talukdar, A., and A. Patra. March 2010. Dynamic model-based fault tolerant control of variable air volume air conditioning system. HVAC&R Research 16(2):233-54.

Teeter, J., and M.-Y. Chow. February 1998. Application of functional link neural network to HVAC thermal dynamic system identification. IEEE Transactions on Industrial Electronics 45(1):170-6.

Tettamanzi, A., and M. Tomassini. 2001. Soft Computing: Integrating Evolutionary, Neural, and Fuzzy Systems. Berlin, Germany: Springer-Verlag.

Tsoukalas, L., and R. Uhrig. 1997. Fuzzy and Neural Approaches in Engineering. New York: John Wiley & Sons, Inc.

Van Velhhuizen, D., and G. Lamont. 1970. Multi-objective evolutionary algorithms: Analyzing the state-of-the art. Evolutionary Computation 8(2):125-47.

Wang, F.-Y., and D. Liu. (Eds.). 2008. Networked Control Systems: Theory and Applications. London, UK: Springer-Verlag.

Weiss, J., 2010. Protecting Industrial Control Systems from Electronic Threats. New York: Momentum Press.

Wright, J., H. Loosemore, and R. Farmani. 2002. Optimization of building thermal design and control by multi-criterion genetic algorithm. Energy and Buildings 34(9):959-72.

Xu, X., S. Wang, Z. Sun, and F. Xiao. January 2009. A model based optimal ventilation control strategy of multi-zone VAV air-conditioning systems. Applied Thermal Engineering 29(1):91-104.

Yan, Y., J. Zhou, Y. Lin, W. Yang, P. Wang, and G. Zhang. August 2008. Adaptive optimal control model for building cooling and heating sources. Energy and Buildings 40(8):1394-401.

Ying-Guo, P., Z. Hua-Guang, and Z. Bien. May 4-9,1998. A simple fuzzy adaptive control method and application in HVAC. Proceedings of The 1998 IEEE International Conference on FuzzySystems, Anchorage, AK, pp. 528-32.

Zadeh, L. 1965. Fuzzy sets. Information and Control 8:338-53.

Zaheer-Uddin, M., and G. Zheng. 1994. A dynamic model of multi-zone VAV system for control analysis. ASHRAE Transactions Research 100:219-29.

Zaheer-Uddin, M., and G. Zheng. 2001. Multistage optimal operating strategies for HVAC systems. ASHRAE Transactions Research 107:346-52.

Zhang, H., and D. Liu. 2006. Fuzzy Modeling and Fuzzy Control. Control Engineering. Boston, MA: Birkhauser.

Zhang, W.-J., S.-F. Ding, and C.-L. Zhang. August 2009a. Transient modeling of an air-cooled chiller with economized compressor. Part II: Application to control design. Applied Thermal Engineering 29(11-2):2403-7.

Zhang, W.-J., and C.-L. Zhang. August 2009. Transient modeling of an air-cooled chiller with economized compressor. Part I: Model development and validation. Applied Thermal Engineering 29(11-2):2396-402.

Zhang, X., M. Chang, M.Y. Man, W.K. Mo, and Y.C. Hang. April 2009b. A PDA-based intelligent building service system. IEEE Circuits and Systems International Conference on Testing and Diagnosis (ICTD 2009), Chengdu, China, pp. 1-3.

Ziegler, J., and N. Nichols. November 1942. Optimum settings for automatic controllers. Transactions of the ASME 65:433-44.

D. Subbaram Naidu (1) * and Craig G. Rieger (2)

(1) School of Engineering, Department of Electrical Engineering and Computer Science, Idaho State University, 921 S. 8th Avenue, Stop 8060, Pocatello, ID, 83209-8060, USA

(2) Idaho National Laboratory, P.O. Box 1625, Idaho Falls, ID 83415-3779, USA

* Corresponding author e-mail:

Received May 27, 2010; accepted November 29, 2010

D. Subbaram Naidu, PhD, is Director and Professor. Craig G. Rieger, PhD, ICIS Distinctive Signature Lead.
COPYRIGHT 2011 Taylor & Francis Ltd.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Naidu, D. Subbaram; Rieger, Craig G.
Publication:HVAC & R Research
Date:Mar 1, 2011
Previous Article:Evaluation of refrigerator/freezer gaskets thermal loads.
Next Article:Infiltration effects on residential pollutant concentrations for continuous and intermittent mechanical ventilation approaches.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters