Printer Friendly

Home energy management system for interconnecting and sensing of electric appliances.

1. Introduction

According to a research report of the Climate Change Society of the University of Oxford, if detailed power consumption data can be obtained, users can control the energy consumption situation of household appliances, and can save 5% to 15% of the power consumption in each month on average [1][2]. Therefore, HEMS have become an indispensable part of the smart grid. However, at present, most families still use traditional ammeters to record their monthly total power consumption. The recently promoted smart meters, such as Google PowerMeter [3] or Microsoft Hohm [4], can show the total household power consumption at present, but cannot show the power consumption of each household appliance, to say nothing of information about the household appliances that are consuming power. As a result, users cannot further improve their power consumption habits or avoid the use of so-called high-power electric appliances. A system that can accurately identify and detect electric appliances is a subject worthy of study. This study proposed a set of HEMS for interconnecting and sensing of electric appliances, which can measure the household power consumption through a current sensor, transmit the data back to the energy management platform, identify each electric appliance, and then determine whether it is working normally according to its staged power consumption and various effects caused by its power sine wave intervals, so as to avoid overloading problems arising from old or faulty electrical appliances. However, the use of older or large numbers of household appliances will cause power noise problems, which could result in the inaccurate identification of electric appliances and the occurrence of errors. Therefore, in this study, a set of power clustering algorithms was presented to determine the cluster value and cluster potential for measured power information. When an abnormal value arises from the system, it is identified as noise or an abnormal state according to the clustering characteristics. The main contributions of this study included:

1. Creating the ability to identify electric appliances by a single smart meter. The system achieves this function according to the power characteristics of each electric appliance. It does not require the use of identification devices such as RFID, or the matching of electric appliances with a power measurement module, and can therefore avoid additional power consumption or waste. It can also recommend appropriate electric appliances to the users through recording the power consumption of each electric appliance and comparing the operation of the same electric appliance system.

2. A high rate of accuracy. The power clustering characteristics of various electric appliances are discovered through power clustering algorithms, so as to improve the recognition rate of noise or abnormal states.

3. The establishment of an interconnection platform for electric appliances. In this study, the energy management platform was attained based on OSGi, so as to achieve the integration of mutual communication services between various electric appliances, the integration of different protocols, and the self-management and context control services of various electric appliances.

The rest of this paper is organized as follows. Section II presents related research and the technology of energy management systems. Section III introduces the framework of the proposed system, the identification of electric appliances and a detailed description of power clustering. Section IV shows the experimental results, and the conclusion is at the end.

2. Related Work and Background

This section introduces relevant recent research of HEMS and some basic knowledge about the OSGi service platform used in this study.

2.1 HEMS

A good HEMS is necessary for a smart grid, and recently there have been numerous researches about this. HEMS research mainly involves three components. The first component is the power meter of the bottom device [5][6], which is mainly used to measure the voltage and current used by a single electric appliance or the overall system of electric appliances. Ye et al. [5] designed a module with a power sensor that can automatically cut off power supply once it detected that an electric appliance was in a standby state, so as to save energy. Cho HS, et al., [7] established a position tree of electric appliances using an SMPT (Smart Multi-Power Tap) above the meter and through the startup of the power tap, so as to provide the position information of electric appliances, and provide references for context-aware and energy management services. The second component is the transmission component. How to construct a complete network system for household appliances through effective wire transmission or wireless transmission of the meter information? Information network transmission systems have evolved from Bluetooth transmission [8] in 2003 to the most frequently used ZigBee [9][10][11] or power lines [12][13] 14] at present. The last component is the system management platform [15][16][17][18][19]. Jahn M, et al. [15] measured the power consumption information of individual electric appliances using a smart socket, and established a hydra framework, enabling the power measurement system to have flexibility and expandability. Users identify household appliances with Ubilense through their mobile phones, which can also display power consumption information. Park S, et al. [20] reduced the amount of power consumption data and established a simulation display to be used in a system capable of transmitting large amounts of data used for long-term monitoring.

2.2 OSGi Platform


The OSGi [21][22][23][24][25] platform is developed by the OSGi Alliance; it supports the service providers and companies through the home gateway to provide service for devices. Through the OSGi service platform, services can be shared between bundles through the service registry. The service provider bundle registers its service objects and service items in the service registry, and the service consumer bundle searches the service objects provided by other bundles in the service registry using the service items.

[0] shows the relationship of each layer. Service layer register service in the service registry, life cycle layer manages the bundle's life cycle and defines the bundle how to start or stop as well. Module layer defines the bundle, which compost of Java class. And bundle can import or export with another bundle. OSGi platform are often employed in home security, healthcare [28], home gateway and so on.

3. Proposed System

This section introduces the overall system framework and expatiates on the various function modules.

3.1 System Architecture


[0] shows the overall system framework of HEMS proposed in the present study. The system measures the power consumption of various household appliances (including voltage and current) through a smart meter, transmits the data results to the energy management platform through a ZigBee wireless network, records the power consumption through the management platform, identifies electric devices according to the database system, and then displays relevant information through an HID or a mobile device. Users can directly transmit control commands through a remote controller a mobile device, and the system can also achieve adaptive context-aware control through a context-aware module as well as users' power consumption plans. Its energy management platform can transmit control signals through a ZigBee network, convert signals to IR signals through the ZigBee control module, or directly control the relay switches of electric appliances to achieve a complete control service platform. Its main functions include:

1. Achieving the integration of various electric appliance systems.

2. Achieving energy-saving mechanisms through an intelligent home network.

3. Providing remote operation and automation functions of various household appliances.

4. Achieving the use of a sensing controller.

3.2 Smart meter

In this study, household power consumption was measured using a smart meter, which is mainly composed of an energy metering integrated circuit (IC), voltage and current sampling circuits, and a microprocessor, to obtain the voltage and current signals.

The energy metering IC used in this paper was the ADE7763 chip produced by Analog Devices, which can be connected with a variety of power measurement circuits, including the current converter circuit and the low resistance voltage divider circuit. During current measurement, the current analog signal is sampled from the current converter circuit, amplified through a programmable gain amplifier (PGA), and then subsequently converted to a digital signal through an analog/digital converter (ADC). Current information obtained in such a manner has large amplitude, and it must be differentiated through a high pass filter so as to obtain a correct waveform level for further integration processing through an integrator. The current signal is obtained through the root mean square (RMS) operation, and its expression is shown in Eq. 1.

I (RMS) = [square root of ([[integral].sup.T.sub.0][I.sup.2](t)dt/T)] (1)

Due to the time signal sampling, Eq. 1 must be converted to Eq. 2.

I (RM S) = [square root of ([[summation].sup.N.sub.j=1] [I.sup.2](f)/N)] (2)

The process of Eq. 2 in the hardware is as follows: after the integration of the digital signal, the square of the current signal is obtained through a multiplier, and is accumulated through a low pass filter. The RMS current value can then be obtained from the square root operation.

During the voltage measurement, the sampling method is the same as that of the current signal, but the difference is that after passing through the ADC, the analog voltage signal needs to be integrated through a low pass filter, followed by the root mean square operation. Here the instantaneous power consumption can be calculated as per Eq. 3.

p(t) = v(t) x i(t) (3)

The instantaneous voltage v(t) and instantaneous current i(t) in Eq. 3 can be expressed as Eq. 4.

v(t) = [square root of (2)] x V sin ([omega]t) (4)

i(t) = [square root of (2)] x Isin ([omega]t) (5)

V and I in Eqs. 4 and 5 are respectively the RMS values of the voltage and current, so Eq. 3 can be expressed as Eq. 6.

p(t) = VI - VIcos(2 [omega]t) (6)

The actual power can be obtained through the instantaneous power, as shown in Eq. 7.


Therefore, in the actual hardware implementation process, current and voltage signals are obtained through sampling and calculation, integrated through a low pass filter, and subsequently averaged after gain adjustment through a multiplier to obtain the actual power. The two parameter, I(RMS) and power (P), are applied to appliance recognition described in Section 3.3.3.

After the power system characteristics are measured using an energy metering IC, in order to link smart meters with the household power consumption management system, the energy metering IC performs relevant measurements after receiving commands from the microprocessor. The results are then transmitted to the microprocessor, which must have a communication interface for external transmission, allowing administrators to remotely understand the measurement results, monitor household power consumption systems, or issue additional commands to the microprocessor. The structure of the communication interface is shown in [0]. After powering-on, the relays are controlled by the micro-controller unit (MCU), and sockets control whether to power on. If necessary, current values obtained from a current sensor will be converted to digital signals through a digital-to-analog converter (DAC). The digital signals are transmitted to the MCU and then to the load side. The MCU can also transmit data or commands to the central control center through ZigBee sensors, or record data in the electrically-erasable programmable read-only memory (EEPROM).


3.3 Energy management service platform

The energy management service platform is responsible for receiving smart meter data through the ZigBee network, normalizing the data and creating power clusters. After noise filtering according to clustering characteristics, the platform identifies electric appliances by comparing the power characteristics with the database, and receives surrounding context information with the aid of the context sensing module. The platform performs the corresponding context control, and transmits commands to the ZigBee control module through the ZigBee network. The energy management service platform is built on the OSGi framework. Its major function bundles include a ZigBee data bundle, a power cluster bundle, an appliance recognition bundle, a context-aware bundle and an electric appliance control bundle.

3.3.1. ZigBee Data Bundle

The ZigBee data bundle is mainly responsible for receiving information transmitted by smart meters and surrounding context sensing modules. The OSGi platform reads data from the ZigBee data bundle, and the packet format is shown in 0.


In the present study, the device type was divided into the smart meter, temperature sensors and humidity sensors. As for the smart meter, the first two bytes of its data are defined as the voltage value, while the last two bytes are the measured current values, calculated through multiplying the measured data by 0.01 V (A). As for the temperature and humidity, these are simple numerical data, which are respectively the measured data multiplied by 0.01 [degrees]C(%).

3.3.2. Power Cluster Bundle

Upon receipt of the voltage and current information, the voltage information must first be normalized to 110V. Different rooms or the number of connected electric appliances are likely to affect the voltage level, which falls within a range of 110 [+ or -] 10V. During wireless transmission, the transmission interference and noise effect will often give rise to incorrect values, as shown in [0](A), which will affect the accurate identification of electric appliances. As for an electric appliance under a variety of operating states, the conversion between its current and phase angle presents a clustering distribution, as shown in [0](B), and the final clustering distribution will present in a fixed number of regions. According to the above characteristics, whether the current is in the same state can be judged by whether the subsequent trace value falls within the clustering range through clustering operations. If a value beyond the clustering range arises, it must be observed whether abnormal clustering values or instantaneous noise distributions arise. Through the power clustering algorithm, the recognition rate of electric appliances can be effectively improved, and abnormal error detection rates caused by noise can be reduced.


In this study, power clustering characteristics were processed using the subtractive clustering method [26] in the neural algorithm. The main concept of the subtractive clustering method is to regard all data points as potential center points and select clustering standards according to the density of the surrounding data points. This method is independent of the complexity of the system dimension, but is proportional to the data amount. Here, it is supposed that M is the power group, and [r.sub.a] is the influence distance of the center point of the clustering group and is a positive constant. The potential value [P.sub.i] (Eq. 8) of the sampling point group [M.sub.i] can be calculated, which represents the potential of this point becoming the clustering center point.


After all potential values P of the sampling points are calculated, the Mc1 with the highest potential value is selected as the first clustering center point. The potential values of the other points then need to be modified, as per the following Eq. 9:


wherein [r.sub.b] is a value to be set in order to avoid getting too close to the last clustering center point [M.sub.c1]. It needs to be greater than [r.sub.a], and its value is generally recommended as 1.5 times that of [r.sub.a]. After this process is repeated, the sampling point group M can be divided into subgroups, wherein, [bar.[epsilon]] and [[epsilon].bar] are the upper and lower limit ratios of the potential value, which are defined in this study as 0.5 and 0.15, respectively.

3.3.3. Appliance Recognition Bundle

In this study, each household appliance was regarded as an RLC circuit, and each had different power characteristics under different operating states. Through the previous study of the power characteristics of household appliances, they can be identified mainly based on the four parameters of I(RMS), power (P), current and voltage phase shift angle, and the distortion power factor. I(RMS) and power (P) were introduced in the last section, and here, the definitions of current and voltage phase shift angles and distortion power factor are introduced.

In an ideal AC circuit, the voltage and current should have the same phase angle, but in actual circuits, the effects of inductors and capacitors on electric appliances have given rise to current deformations and phase difference relationships, as shown in the following [0].


The so-called distortion power factor (DPF) (Eq. 10) is defined.

DPF = 1/[square root of (1 + [THD.sup.2.sub.i])] = [l.sub.1,rms]/[l.sub.rms] (10)

[THD.sub.i] is the total harmonic distortion of the load current, [l.sub.1rms] is the fundamental component of the current, and Irms is the total current. After returning to the original waveform through the DPF, the deviation angle [omega] can be discovered according to the following Eq. 11:

[absolute value of P] = [absolute value of S] cos [omega] (11)

where S is the apparent power. When the electric appliance first joins the HEM system, the system will record and study its characteristics, and then create model data. Users need to input data for different models of this appliance, such as name, brand, power usage, and so on, so as to provide enough information for the database. Later, when the electric appliance is restarted, the system will compare its characteristics with those of the model. Comparison programs of power characteristics

The system establishes a factor queue of the various eigenvalues in sequence. After the factor queue of eigenvalues is ready, when data from new electric appliances are generated, they are input into the search system. Searched results are obtained and saved as a database. Each element within it is built with the same structure, which records the device model, device importance, video description, and power characteristics. The structure of the power characteristics is comprised of four parameters, namely the power cluster value, the angular position of the peak value of the sine wave interval, the delayed value of the sine wave interval, and the displacement value of the state power. Afterwards the retriever performs the corresponding operation of the database from the factor queue based on the factor properties. Cases of factor operations are as follows.

The Operation of power characteristics: The same power characteristics of the database are compared to eliminate elements with overlarge differences in the power characteristics of the database. Here, a set of appliance-matching algorithms is presented, which is a modified algorithm based on the Boyer-Moore algorithm. It is assumed that the system regularly captures the clustering power of every section of the electric appliance in both the first D1 seconds and the last D1 seconds as the identification standard, and that the system will capture the currently measured power clustering in the first D1 seconds as the eigenvalue of power clustering for the first identification. The system compares each phase of the power clustering in the first D1 seconds with the power clustering list obtained through the first phase of screening as a comparison target. If M1videos are successfully identified as being related, then the following electric quantity clustering data are identified with the M1 elements in the last D1 seconds as the samples and the rest may be deduced by analogy, until a complete power model is identified and the identification algorithm is completed, as shown in [0].


Suppose that there are M electric appliance models. It takes 2M * [T.sub.1](D1) to find the electric power models with a beginning and ending of D1, respectively, and [T.sub.2](D1) to identify and compare the D1. The total time consumption is as Eq. 12:

T(M) = 2M * T1(D1) + M1 * T2(D1) + M1M2 * T2(D1) + M1M2M3 * T2(D1) + ... (12)

By analogy, suppose that the same amount of models is successfully identified each time, and that M1 = M2 ... = n, it is simplified as Eq. 13:

T(M) = 2M * T1(D1) + nT2(D1) * (nm - 1)/(n - 1) (13)

When the operation corresponding to each factor is completed, an index array will be obtained. The index array records the element index in the database and continuously inputs the structural sequence of this device into the operating queue in sequence. The system shows the operating queue as the electric appliance being used, and the continuous control commands may also collect device models from this operating queue.

3.3.4. Context aware bundle

This bundle detects the regional context, such as temperature and humidity, whether there are human activities or not, and so on, and then transmits the results to the server through a network so as to be available for relevant people and other programs. It is mainly divided into four steps. Information collection

When users interact with the device, data will be collected. The system submits the collected amount of environmental parameters and related information of the interactive devices (such as UPnP, Bluetooth, infrared, ZigBee, etc.) to the learning system for information analysis, so as to produce effective information. Furthermore, useful information can be integrated with other information once again to form new information. The detailed degree of users' behavioral patterns expressed by the information depends on the integration degree; that is, users' ongoing behaviors can be more accurately described according to the information with a higher integration degree. Information analysis

Information analysis aims to determine users' behavioral patterns represented by the information. Upon receipt of the information and after analysis, the system will divide interactive information into two groups. The first group is information on further behavioral patterns after interaction with the device. The second group is information on user dissatisfaction with the environment or the response time after implementation of the operation. Information analysis integrates the synthetic information in the information collection stage, and then sorts out and strengthens the programs that can be judged by the command reasoning system. Context reasoning

[0] (A) shows the process of the basic framework of the reasoning system, which is implemented on the premise of completing data collection and data analysis. After data such as environmental parameters, interactive information of devices and behavioral patterns are obtained from the reasoning system, if the reasoning system cannot find corresponding records from previous learning experiences, it will introduce these data into the learning system; otherwise, it will begin to implement its work. The reasoning system will judge every possible context according to a standard. This standard gives priority to the information in accordance with its accuracy and integration degree. In the flowchart of context generation shown in [0] (B), if a possible context has a higher priority, then this context will be decided and implemented, while other contexts with a lower priority will not be adopted because of insufficient information.

[FIGURE 8 OMITTED] Users' feedback:

After the implementation of corresponding assistance programs, the system waits for the users' responses as to whether the assistance helps to achieve the desired functions, and then transmits these responses to the learning system. The learning system updates the regulations in response to this information, which then enhances intelligence and other subsequent work. The priority of all possible contexts forecast by the reasoning system is given in accordance with the accuracy and integration degree of the information, but user feedback will either enhance ore reduce the priority. At this stage, the priority of the implemented contexts is adjusted after prediction.

3.4 ZigBee Control Module

To integrate existing electric appliances without changing the general control model of household appliances, a ZigBee control module was responsible for receiving the commands issued by the energy management platform and for converting these commands to the command formats of the corresponding devices, where it could then control the household appliances through a ZigBee-to-IR [27] framework. The ZigBee control module is shown in the following [0]:

This module is composed of a group of IR transmission signals, a relay control and a ZigBee receiving module. An infrared remote controller transmits data using a modulated manner, that is, the data and carrier frequency at a certain frequency are taken as the AND. In this way, the transmitting efficiency can be improved and power consumption can be reduced. In this study, infrared signals could be decoded through a Linux infrared remote control kit developed for the Linux platform, and by using an infrared receiver connected to a PC via RS232, and the host computer could also be remotely controlled to do specific actions. The function of each button is recorded, and then its data are input to the ZigBee control module. When the control module receives a signal, the packet is analyzed to find the corresponding control command units, so as to issue control and achieve an integrated control service of household appliances.


4. Result and Analysis

This section introduces the results achieved in this study and experimental analysis According to this system.

4.1. Implementation


The result interface achieved in this study is shown in the following 0. This study was mainly used to measure the power consumption of general household appliances and to identify electric appliances, as well as allow users to remotely inquire about related information through the Internet. The measured information was comprised of the household temperature and humidity measured by sensors in the surrounding environment, which could simultaneously display the power, voltage and current information. The general relay controls and IR controls of a television and an electric fan were completed in the electric control section. The control method section was divided into user-controls, power-saving, content-awareness and planning. User-control was defined as all electric appliances that are self-controlled by users, and power-saving was defined as an automatic shutoff to the power supply when a standby state was detected. Content-awareness was defined as the automation of controls in electric appliances based on environmental information and historical user information. The planning section was defined as the user input of scheduled power consumption, and then allowing the system to decide the schedule the electric appliances according to the electricity price and necessity.

4.2. Experiments and analysis

In this study, a total of 40 different household appliances were used for experimental analysis. During the experiment, at most six electric appliances were randomly started for identification analysis, and there were 30 experiments in each stage.

4.2.1. Relation between recognition accuracy and recognition time

In this experiment, the recognition accuracy was studied mainly based on the sampling time and recognition time, and the accuracy rate was defined as:

F = Recognition time/Sampling time * 100% (22)


As can be seen from the experimental results, [0], when the sampling time was insufficient (15 S), the sample would be incompletely established, resulting in recognition difficulties. Even if the recognition time was extended, it was still difficult to increase the recognition rate, while too short of recognition time would also cause recognition difficulties. As can be seen from the experimental results, the best sampling time was 60 seconds. When the recognition time reached 120 seconds, the recognition accuracy was 92%, and this could reach as high as 95%.

4.2.2. Relation between recognition accuracy and the current clustering algorithm

In this experiment, [0], the effect of the current clustering algorithm on the system recognition rate was analyzed. An experiment is performed with a sampling time of 60 seconds and a recognition time of 90, 120, 150, and 180 seconds, respectively. The experimental results are shown in the following figure. As can be seen from the experimental results, the recognition accuracy rate was approximately 91.5% on average using the current clustering algorithm, and 79% on average without the current clustering algorithm, due to the effect of noise on data collection, sampling and identification.


4.2.3. Relation between recognition accuracy and recognition number

Another concern was whether increasing the number of electric appliances would affect recognition accuracy, so the detection of different numbers of electric appliances was investigated. The experimental results are listed in the following Table 1. When there was one electric appliance, the recognition rate was 100%, but when there were two or more electric appliances, the recognition rate was reduced due to the effect of capacitors and inductors in series. This will be a target of subsequent studies in the future.

5. Concluding Remarks

In this study, a set of HEMS for the interconnection and sensing of various household appliances was presented. It measured power information through a power meter and transmitted the data back to the management platform via wireless transmission. It allowed users to realize the currently used electric appliances and their power consumptions through the identification of the devices, and provided corresponding control interfaces for the users to remotely control household appliances. It established content-aware service functions with the help of context information sensors and user habits, and its recognition rate reached as high as 95% with the aid of the current clustering algorithm and the establishment of identification samples. In the future, research will be mainly engaged in establishing the planning control model matched with cloud services, so as to expand the scope of recognition and the obtainment of identification samples.


Yueh-Min Huang's research in this paper was supported by Innovation R&D and Advance Industry Research Plan (99RC13), National Science Council (NSC), Taiwan.

DOI: 10.3837/tiis.2011.07.004

Received April 1, 2011; revised June 8, 2011; accepted June 15, 2011; published July 28, 2011


[1] Wood G, Newborough M, "Dynamic energy-consumption indicators for domestic appliances: environment, behaviour and design," Energy and Buildings, vol. 35, no. 8, pp. 821-841, September, 2003. Article (CrossRef Link)

[2] Darby S, "The effectiveness of feedback on energy consumption. A review for DEFRA of the literature on metering, billing and direct displays," Environmental Change Institute, University of Oxford, pp. 1-21, April, 2006. Article (CrossRef Link)

[3] Google Inc., "Save energy. Save money. Make a difference," 2011. Article (CrossRef Link)

[4] Microsoft Corp., "How energy efficient is your home?," 2010. Article (CrossRef Link)

[5] Ye Y, Li B, Gao J and Sun Y, "A design of smart energy-saving power module," in Proc. of 5th IEEE Conf. on Industrial Electronics and Applications, Taichung, pp. 898-902, Jun 2010. Article (CrossRef Link)

[6] Serra H, Correia J, Gano AJ, de Campos AM and Teixeira I, "Domestic power consumption measurement and automatic home appliance detection," in Proc. of Int. Workshop on Intelligent Signal Processing, Faro, Portugal, pp. 128-132, September, 2005. Article (CrossRef Link)

[7] Cho HS, Yamazaki T and Hahn M, "Determining location of appliances from multi-hop tree structures of power strip type smart meters," IEEE Trans. on Consumer Electronics, vol. 55, no. 4, pp. 2314-2322, November, 2009. Article (CrossRef Link)

[8] Tajika Y, Saito T, Teramoto K, Oosaka N and Isshiki M, "Networked home appliance system using Bluetooth technology integrating appliance control/monitoring with Internet service," IEEE Trans. on Consumer Electronics, vol. 49, no. 4, pp. 1043-1048, November, 2003. Article (CrossRef Link)

[9] Han DM, Lim JH, "Design and implementation of smart home energy management systems based on ZigBee," IEEE Trans. on Consumer Electronics, vol. 56, no. 3, pp. 1417-1425, August, 2010. Article (CrossRef Link)

[10] Han DM, Lim JH, "Smart home energy management system using IEEE 802.15.4 and ZigBee," IEEE Trans. on Consumer Electronics, vol. 56, no. 3, pp. 1403-1410, August, 2010. Article (CrossRef Link)

[11] Bennett C, Highfill D, "Networking AMI Smart Meters," in Proc. of IEEE Energy 2030 Conference, pp. 1-8, November, 2008. Article (CrossRef Link)

[12] Liu J, Zhao B, Wang J, Zhu Y and Hu J, "Application of power line communication in smart power Consumption," in Proc. of IEEE Int. Symposium on Power Line Communications and Its Applications, Rio de Janeiro, pp. 303-307, March 28-31, 2010. Article (CrossRef Link)

[13] Son YS; Pulkkinen T, Moon KD and Kim C, "Home energy management system based on power line communication," IEEE Trans. on Consumer Electronics, vol. 56, no. 3, pp. 1380-1386, August, 2010. Article (CrossRef Link)

[14] Lien CH; Bai YW; Chen HC and Hung CH, "Home appliance energy monitoring and controlling based on Power Line Communication," in Proc. of Digest of Technical Papers Int. Conf. on Consumer Electronics, Las Vegas, NV, pp. 1-2, January 10-14, 2009. Article (CrossRef Link)

[15] Jahn M, Jentsch M, Prause CR, Pramudianto F, Al-Akkad A and Reiners R, "The Energy Aware Smart Home," in Proc. of 5th Int. Conf. on Future Information Technology, Busan, pp. 1-8, May 2010. Article (CrossRef Link)

[16] Capone A, Barros M, Hrasnica H, Tompros S, "A New Architecture for Reduction of Energy Consumption of Home Appliances," in Proc. of TOWARDS eENVIRONMENT, European Conf. of the Czech Presidency of the Council of the EU, March 25-27, 2009. Article (CrossRef Link)

[17] Yang CS, Liao MY and Chen CX, "Design and implementation of HEMS based on RFID and OSGi," in Proc. of Int. Conf. on Anti-counterfeiting, Security, and Identification in Communication, Hong Kong, pp. 250-253, August, 2009. Article (CrossRef Link)

[18] Heo J, Hong CS, Kang SB and Jeon SS, "Design and Implementation of Control Mechanism for Standby Power Reduction," IEEE Trans. on Consumer Electronics, vol. 54, no. 1, pp. 179-185, February, 2008. Article (CrossRef Link)

[19] Sundramoorthy V, Liu Q, Cooper G, Linge N and Cooper J, "DEHEMS: A user-driven domestic energy monitoring system," in Proc. of Internet of Things, Tokyo, pp. 1-8, November 29-December 1, 2010. Article (CrossRef Link)

[20] Park S, Kim H, Moon H, Heo J and Yoon S, "Concurrent simulation platform for energy-aware smart metering systems," IEEE Trans. on Consumer Electronics, vol. 56, no. 3, pp. 1918-1926, August, 2010. Article (CrossRef Link)

[21] Dobrev P, Famolari D and Kurzke C, "Device and service discovery in home networks with OSGi'," IEEE Communication Magazine, vol. 40, no. 8, pp. 86-92, August, 2002. Article (CrossRef Link)

[22] Jeong J, Shin D and Shin D, "An XML-based automated authentication profile for home network based on OSGi (open service gateway initiative) framework," in Proc. of Digest of Technical Papers of Int. Conf. on Consumer Electronics, Korea, pp. 99-100, January, 2006. Article (CrossRef Link)

[23] Kang, K., Lee, J. and Choi, H. "Management system for OSGi-based infrastructure," Digest of Technical Papers of International Conference on Consumer Electronics, ICCE'06, January, Korea, pp. 53-54, January, 2006. Article (CrossRef Link)

[24] Y. Lu, Y. Yuan, Y. Sun and X. Yang, "An approach to service integration in the OSGi architecture of home networks," in Proc. of IEEE Singapore Int. Conf. on Communication Systems, ICCS 2008, pp. 756-760, November 19-21, 2008. Article (CrossRef Link)

[25] C. Lee, D. Nordstedt, S. Helal, "Enabling smart spaces with OSGi," IEEE Pervasive Computing, vol.2, no.3, pp. 89-94, July-September, 2003. Article (CrossRef Link)

[26] Chiu, S.L., "Fuzzy model identification based on cluster estimation," Journal of Intelligent and Fuzzy Systems, vol. 2, pp. 267-278, September, 1994. Article (CrossRef Link)

[27] WK. Park, I. Han, and KR. Park, "ZigBee based Dynamic Control Scheme for Multiple Legacy IR Controllable Digital Consumer Devices," IEEE Trans. on Consumer Electronics, vol. 53, no. 1, pp. 172-177, February, 2007. Article (CrossRef Link)

[28] C. Min, S. Gonzalez, Leung V., Z. Qian, and L. Ming, "A 2G-RFID-based e-healthcare system," IEEE Wireless Communications, vol. 17, no. 1, pp. 37-43, February, 2010. Article (CrossRef Link)

Wei-Ting Cho received the M.S. degrees in Computer Science and Information Engineering from National Cheng Kung University in 2008. He is currently a Ph.D. student in Department of Engineering Science at National Cheng Kung University. His research interests are information retrieval, information extraction, data mining, and embedded system. Before pursuing a Ph.D., he had worked in industry for 2 years.

Chin-Feng Lai is an assistant professor in Institute of Computer Science and Information Engineering at National Ilan University (NIU). He has worked as a Post-Doctoral Fellow in Dept. of Engineering Science at National Cheng Kung University (NCKU) for one and a half year since Jan. 2008. He has published more than 50 technical papers. Dr. Lai serves as an editor of KSII Transactions on Internet and Information Systems. He has worked as a session chair in ICIMCS'09. He was the workshop Co-Chair of EMC-10, ACM Mobility Conference 2010, publicity co-chair of MMASN 2010. He serves as the corresponding guest editor for Journal of Telecommunications Management (JTM). He was selected an honorary member of the Phi Tau Phi Scholastic Honor Society of the Republic of China in 2009 and TC member of Multimedia Systems & Applications Technical Committee (MSATC), IEEE Circuits and Systems Society, 2009. He is also a project leader of several international, industrially-funded multimedia projects in NCKU. He is an IEEE member.

Yueh-Min Huang received the M.S. and Ph.D. degrees in electrical engineering from the University of Arizona, Tucson, in 1988 and 1991, respectively. He is a Distinguished Professor with the Department of Applied Geoinformatics at Chia Nan University of Pharmacy & Science and the Department of Engineering Science, National Cheng-Kung University, Taiwan. His research interests include wireless networks, artificial intelligence, and e-Learning. He has coauthored two books and has published over 140 refereed journal research papers. Dr. Huang has received many research awards, such as the Best Paper Award of 2007 IEA/AIE Conference, Best Paper Award of the Computer Society of the Republic of China in 2003, the Awards of Acer Long-Term Prize in 1996, 1998, and 1999, Excellent Research Awards of National Microcomputer and Communication Contests in 2006. He has been invited to give talks or served frequently in the program committee at national and international conferences. He is on the Editorial Board of the Journal of Wireless Communications and Mobile Computing, the Journal of Security and Communication Networks, and the International Journal of Communication Systems. He is a member of the IEEE Communication, Computer, and Circuits and Systems Societies.

Wei-Tsong Lee received B.S., M.S. and Ph.D degrees in Electrical Engineering from National Cheng Kung University, Tainan, Taiwan. In 2003, he joined the department members of Electrical Engineering of Tamkang University as associate professor, and reached professor in 2007. From 2010, he is the chairman of Electrical Engineering Department. His research interests are computer architecture, micro-processor interface and computer network.

Sing-Wei Huang received his B.S. degree in Industrial Management from National Taiwan Institute of Technology, Taipei, Taiwan, in 1984, and received his M.S. degree in Industrial Management from National Cheng Kung University, Tainan, Taiwan, in 1986. He joined Chung-Shan Institute of Science and Technology 25 years ago. He is concurrently working at Chung-Shan and pursuing his Ph.D. in Industrial and Systems Engineering at Chung Yuan Christian University, Tao-Yuan, Taiwan. His research interests are e-life digital technology, RFID applications and wireless sensor networks.

Wei-Ting Cho (1), Chin-Feng Lai (2), Yueh-Min Huang (3), Wei-Tsong Lee (4) and Sing-Wei Huang (5)

(1) Department of Engineering Science, National Cheng Kung University Tainan 701, Taiwan

(2) Institute of Computer Science & Information Engineering, National Ilan University Ilan 260, Taiwan

(3) Department of Applied Geoinformatics, Chia Nan University of Pharmacy & Science Tainan 701, Taiwan

(4) Department of Electrical Engineering, Tamkang University Taipei 151, Taiwan

(5) Department of Industrial and Systems Engineering, Chung Yuan Christian University Taoyuan 320, Taiwan

Table 1. Relation between recognition accuracy and recognition number

Number of electric      1            2            3

recognition accuracy    100%         97%          95%

Number of electric      4            5            6

recognition accuracy    88%          90%          91%
COPYRIGHT 2011 KSII, the Korean Society for Internet Information
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:Cho, Wei-Ting; Lai, Chin-Feng; Huang, Yueh-Min; Lee, Wei-Tsong; Huang, Sing-Wei
Publication:KSII Transactions on Internet and Information Systems
Article Type:Report
Date:Jul 1, 2011
Previous Article:A global graph-based approach for transaction and QoS-aware service composition.
Next Article:Code combining cooperative diversity in long-haul transmission of cluster based wireless sensor networks.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters