Stages in developing methodological assets and software to be used in planning armed forces employment.

Planning the use of forces and weapon systems is seen as one of the main measures in preparing forces, troops and systems of armaments for combat employment in would-be wars. Planning is essentially about command and control agencies developing variants of a plan of combat employment of forces and weapon systems that are in conformity with the forecasted (emerging) conditions of the military-political and military-strategic situation (MPS and MSS) and provide for the assured execution of tasks set to troops (forces) in a war. Tasks to use separate components of forces and weapon systems are defined within the context of the general concept of employment of the Armed Forces. For this reason, planning the use of weapons is part of a more general operations (forms of actions) planning process and is made to form a system with measures to develop the concept and plans of the employment of general-purpose forces.

Planning is based on three important principles that require that it should be continuous, done in good time, and multi-variant. The latter is aimed to deal with uncertainties in forecasts for the changing MPS and MSS, and, accordingly, to strengthen the guarantee that the combat assignments set within a broad spectrum of emerging real situational conditions should be performed after all.

The combat task to be assigned to a weapon system and each asset is determined by an invariantly interpreted set of quantitatively concretized parameters of target designation, time, and combat employment method. Referred to the existing structural-organizational formations, weapon employment assignments determine the combat task to be addressed respectively by components of forces and armaments force in each planned variant of strikes. Thus, in the military-technical aspect the aim of each variant of the plan and of their totality is to determine such characteristics of target designations and methods of employment of a separate asset, complexes of assets, and their organizational formations and forces as will secure, in forecasted (existing) situational conditions, the assured execution of weapon employment assignments, given the optimal use of the combat capabilities of weapons in coordination with other weapon systems, troops and forces.

As is common knowledge, making a plan of combat employment of forces (weapons) is a complicated, integrated military-technical job. There is an organizational-technical automated planning system to address it. The planning process within the system is a sequence of parts or sub-processes that have information links with each other, a technique designed to overcome the difficulties involved in determining the entire totality of characteristics of the plan in each variant of the strike and as a whole. The main instrument obtaining quantitative evaluations and recommendations as produced in the process of development of the variants of the plan is the existing planning-adjusted specialized mathematical support and software (SMSS). From the start its development and stage-by-stage improvement were tackled by research teams of the 27 Central Scientific Research Institute of the RF Defense Ministry.

By tradition the SMSS work was based on operational and organizational requirements, principles and provisions that reflected the content and peculiarities of the multistage planning process itself, on the one hand, and the specifics of its separate stages, on the other. The characteristic properties of operational and computation tasks addressed in the process of planning were what explained the problems that arose in the course of work. We will single out the following essential properties determining the methodological and methods approaches to formulas and solution procedures.

First, the multi-aspect (multi-factor) nature of tasks. The objective conditions of the situation and operational requirements determine the necessity of analyzing the processes of employment of forces (assets) and weapons from different sides and within a broad range of their heterogeneous properties and connections (within different system sections). To be able to study the processes in this way, we have to introduce and use a vector scale for measuring the system's properties, and to design quite a set of indicators and evaluations of those properties, as well as criteria for choosing the optimal (rational) controls to handle the analyzed processes. If the purpose is to find rational solutions, the multi-variant analysis of complex processes of warfare and weapon employment holds out no prospects in the general case should only qualitative methods of comparison and selection of alternatives be in use. Those solutions may be obtained only on the basis of quantitative analysis methods realized in informational-computational problems (ICP), models and methods of SMSS, both developed and improved.

Second, weak quantitative definiteness of a part of conditions of operational problems addressed at the input of the designed computation process and the complete determinateness of the plan's characteristics (parameters) sought at the output of the process. The weak formal definiteness of conditions of the operational problems of planning is displayed in the non-removable subjectivity of suppositions and assumptions, which the problem tackler accepts at the stage where he formulates computation conditions and in the course of computations. Along with quantitatively expressed constraints, some of the decision-maker's (DM) requirements take the form of qualitative judgments that determine various preferences. More often than not, attempting to formalize those preferences leads to new subjective suppositions.

The DM would put forward a number of informal judgments and requirements directly in the course of computations as an analysis is being made of intermediate and final results. All of that conveys to the search for solutions to operational problems some characteristics of the man-machine process, and to the planning problem, some traits of an under-structured quantitative-qualitative problem. The latter circumstance is pushing to the fore the necessity of combining and making mutually complementary, in a man-machine system, the capabilities of the DM and the computer. It is also necessary to construct an effective computation control process in order to determine interactively the rational values of the desired characteristics (parameters) of a force (weapons) employment plan.

Third, the big dimension of problems. The conditions of planning problems are based on data determining the properties and states of many targets and weapons. The possible methods of employment of each weapon may be characterized by a considerable number of alternatives. Solution results may contain, along with the desired plan characteristics, a plethora of detailed and aggregated evaluations and indicators of the efficiency of weapons used by different methods against an entire totality of targets as well as against separate targets.

Fourth, the complexity and comprehensiveness of problems. The planning problems analyze indicators and criteria determining the dependence of the desired plan characteristics on evaluations of the state of different targets that possess different properties and a heterogeneous structure. Weapon employment processes a problem investigates are characterized by complicated regularities and linkages determined, among other things, by the necessity of coordinating employment methods of many, more often different-type weapons as to where, when and why they are used. This coordination requires a comprehensive consideration of the entire totality of conditions determining the alternative variants of weapon employment. The values of the desired variables at the output of information-computational problems are, as a rule, the functions of the characteristics, which have to be defined as well. Both, in turn, depend on a large number of different factors, indefinite and chance ones included.

Fifth, multi-variant nature of solutions and the hierarchic structure of the computation system. The multi-variant nature of computations holds important reserves for finding rational methods of employment of forces (weapons) in operations, given the uncertainty of data on possible conditions of their pursuance. The multi-variant nature of the analysis of the subject area--a plethora of variants of employment of forces and assets with a parallel evaluation of their efficiency--makes it possible to enhance the definiteness of the a priori data on possible outcomes of planned operations.

If planning is undertaken in good time, the multi-variant principle reflects in the development of a totality of variants of a plan of employment of forces (weapons). Preparing a totality (plurality) of rational variants of a plan is due to secure an effective use of the potential of force groupings and weapon systems under war conditions as they shape in reality. Please note that the necessity of this approach generates the problem of how to overcome the discrepancy between the detailed nature of the formalized representation in models of the analyzed warfare (weapon employment) processes and the labor-consuming nature of computations.

A way to deal with the discrepancy is to decompose the input planning problem, to wit, to represent it through a hierarchic totality of smaller and less complicated problems coordinated in terms of conditions. The top level of the computation system is reserved for models (ICP, methodologies) with aggregated, generalized structure that contain a typified description of the principal properties of the modeled processes.

This kind of typifying and generalization enables investigation of the most important regularities of processes within a broad range of conditions of current operational missions, that is, within the scheme of the multi-variant analysis of the problem. Selecting, on the basis of generalized criteria, results of multi-variant solutions using aggregated models of operations makes it possible to narrow down the range of analyzed computation conditions to their certain rational totality. The latter is the information input in detailed-level models.

Detailed models usually represent the investigated warfare (weapon employment) processes through more detailed formalized schemes (system procedures). But the possibilities of making multi-variant computations with the use of detailed models are limited by computer performance powers and the ability of DM to work with large amounts of information.

Generally, multi-level systems determine the need, as planning SMSS is developed, to deal with a number of difficult structural unification problems. They also require that models (ICP, methodologies) assuring high efficiency and precision of current computations be coordinated in terms of information and criteria.

The difficulties involved in developing the planning SMSS we have characterized in brief as well as many others were tackled at stages where the methodology and methodological assets applied to problems in the subject area under consideration came into being and made strides.

The initial period that saw the emergence of trends in investigations and developments in the area of methodological and program assets directed at coping with planning problems was associated with the growth of technical capabilities and greater practical role of computers.

A first trend in investigations was oriented toward creating mathematical models of group and later massed strikes to evaluate efficiency of combat employment of weapons against separate targets and groups of targets. A second trend embraced a broad spectrum of problems involved in the development of models of combat use of weapons with account taken of the spatial-temporal structure of the processes of functioning of forces and assets in operations. Successes achieved in the context of this trend were due to the available theoretical and practical results in the area of operations investigation methodology as well as to the development and use of mathematical programming methods. The available results served as the methodological basis for a serious progress in investigations within a third important trend associated with the development of algorithms for a broad class of problems of so-called optimal distribution of resources.

Tendencies in the problems within the first trend at the stage under review may be characterized by the following fundamental peculiarities:

* the transition from the simplified model of a homogeneous engaged target to a formalized scheme that took into account the heterogeneous composition of vulnerable elements in a facility, their uneven distribution, the intricate configuration of disposition area, and other properties;

* the introduction, along with the concept of engaged target (facility), of the category of material resources of the military and the military-economic potentials (MP and MEP) of a region or a country to coordinate particular indicators of the forecasted state of separate targets with integral evaluations of expected results of group (massed) strikes against a totality of facilities;

* the inclusion in models of the complex spatial-temporal structure of strikes, the multiple impact several different-effect explosions bring to bear on the vulnerable elements of targets, and, in some problems, also the main parameters of a simplified process of the functioning of facilities.

The consistent expansion of the possibilities of inclusion in the computer algorithms of an increasingly representative totality of conditions, factors and constraints of the formulated problems made their results more trustworthy. Successes in creating analytical and statistical methods to model group (massed) strikes against heterogeneous targets secured the acceptable precision of computations. The said circumstances explain the rather rapid diffusion of computer methods evaluating the effectiveness of group (massed) strikes for use in dealing with many practical tasks. A considerable contribution to problems within the first trend came in the form of theoretical and applied works done by many scientists and authoritative specialists of the 27th Central Scientific Research Institute of the RF Defense Ministry, primarily N.P. Buslenko, L.N. Kutsev, V.P. Luzyanin, K.K. Yefremov and others.

The development of the methodology and methods within the second trend was due to the necessity of analyzing the efficiency of weapons systems, weapons and their separate components with account taken of the potential enemy's counteraction. Given the limited memory and performance characteristics of the computers in the 1960s and the 1970s, the practical needs that dictated this kind of analysis accelerated the process whereby models with the designation we have just characterized were made to form independent classes of problems. Some of these are problems, important planning-wise, of modeling combat operations by opposing force groupings, reconnaissance models, problems of modeling the defense systems (antimissile defenses, air defenses, anti-submarine defenses) fighting against incoming weapons of the potential enemy, and others.

A special class of problems includes models intended to locate basing areas of weapon systems, which secure, given the necessary level of operational requirements, an effective realization of the combat capabilities of weapons, their high survivability, mobility, and reliable control.

Of much theoretical and practical importance were works to create a class of models for military-economic analysis, designed to identify generalized and detailed functional characteristics of material production systems of countries, regions and coalitions of states at different stages in the military-political situation, including in wartime and during subsequent economic rehabilitation period.

The said and other classes of mathematical models in the second trend served as an instrument that helped to analyze and identify quantitative recommendations and indicators to be used in formulating the conditions of problems in the first and the third trends. Some of these recommendations are proposals on the allocation of strength levels to stages of combat operations, on formulations of objectives of a strike against material resources and facilities, on the location of weapon basing areas, on expedient functioning regimes for weapon systems, etc. A list of the most important indicators at the output of mathematical models in the second trend includes evaluations of technical reliability and survivability of basing facilities and weapon delivery vehicles, evaluations of effectiveness of weapons penetrating enemy air defenses, antisubmarine defenses and antimissile defenses, as well as other characteristics. Some major theoretical and practical results in this area of investigations were obtained by the scientists of the 27th Central Scientific Research Institute of the RF Defense Ministry A.V. Zhuravlyov, Ya.D. Lekanov, Yu.M. Likhterov, Yu.S. Solnyshkov, B.F. Finkelshtein, I.V. Khudyakov, and others.

Practical supplements on the planning of weapon effectiveness evaluation often list problems in the third trend under the general heading of target allocation problems. Despite a great variety of formulas the said problems are characterized by common stable components of conditions and solution results. These components at the input of each problem are the data on facilities (targets), weapons, and planning conditions; at the output these are numerical characteristics of the weapon employment plan. The procedures for reducing input data to sought-for results are essentially dependent on the forms of expression of planning goals, the structure of chosen criteria, as well as many additional criteria that often generate within a concrete problem a unique system of constraints and functional ties between variables. The latter remark refers primarily to the consideration of different factors characterizing the minuteness and precision of a problem's formalized description of physical facilities and real processes: targets, weapons, and peculiarities of their functioning and employment. Many of these factors are of chance or indefinite nature. The necessity of considering the stochastic and quantitative-qualitative elements of a modeled process leads to additional difficulties in the formulation of problems and the search for their solution methods.

The research stage under consideration has evolved a tendency toward differentiating a broad class of target allocation problems by two characteristic signs: methods of formalized description of problems' conditions and their solution methods in use. In accordance with these signs, a first major group is formed by allocation problems with a simplified structure of conditions. The basis of simplification was the said supposition about the homogeneity of facilities and their engagement independence. The logically strict definiteness of conditions and functional ties in problems in the first group was one of the prerequisites for successful application of formal methods of optimal resource allocation to solve many of these problems. Some important results in developing the said problems were obtained by the scientists of the 27th Central Scientific Research Institute of the RF Defense Ministry G.M. Arkhipova, V.M. Kisler, A.A. Pozanenko, N.S. Krasilnikov, R.R. Naumova, A.I. Titova, M.N. Tyuleneva, and A.I. Cherkashchenko.

Practical approaches to the formulations and methods of dealing with allocation problems in the second group took shape when it became necessary to give a satisfactory consideration, in problems, to not easily formalized realities of the modeled physical processes and characteristics of facilities to be engaged: their complex structure, system connections between elements to be engaged, their changed characteristics in the course of strikes, etc.

Describing damage and casualty effect of munitions became important in second-group problems, as did the making of acceptable models of facilities and their systems and the creation of usually approximate target allocation procedures. If in formal allocation problems the descriptions of weapons' damage and casualty effect were detailed to the extent that they defined so-called effectiveness matrix coefficients, then the second-group problems take into account a much greater amount of properties of the engagement process where both one and several munitions explode. Created in that period, the algorithms of rational aiming points and munitions detonation altitudes at a separate complex facility illustrate one of the techniques specifying models of damage as inflicted on a group of engagement-dependent targets by several explosions. Another technique was specifying methods of determining indicators of effectiveness of strikes on the basis of available information on the regularities of the damage effect explosions have on the vulnerable elements of targets.

Methodological recommendations and practical solutions for algorithms of effectiveness criteria, required weapon details, aiming points and detonation altitudes when planning a group strike at separate complex facilities found a reflection in works by scientists of the 27th Central Scientific Research Institute of the RF Defense Ministry, including the works by V.M. Bagno, I.A. Isayev, A.N. Kuchin, as well as in some early works by the present writers.

A major theoretical generalization of the results of investigations in the period under consideration was in the works by O.V. Sosyura. These clearly reflected the then level of practical realization of the results obtained.

The decisive surge in the effectiveness of weapons, the expansion of boundaries and possibilities of their combat employment led to a serious complication of planning conditions. The scale and comprehensiveness of problems tackled increased. Clearly it became necessary to take into account the consequences of strikes directed not only at planned facilities but also "incidental engagement" targets. The fact that general and particular combat missions were mutually conditional, as were weapon employment methods, became clearly manifest. A situation took shape that determined the evident expedience of finding the best method of combat employment of the entire totality of components of weapon systems in a strike planned against all engagement targets. The difficulties of a formally strict formulation of this problem on the basis of the classical methods that are used to deal with tasks of optimal resource allocation seemed insurmountable from the start. They are hardly likely to be removed in future too.

The first practical results of the quantitative methods of analysis and those that sought for rational solutions where the planning of massed strikes was concerned were due to the development and application of a methodology based on the systems concept that was designed to study major complex problems.

S.A. Yeryomin should be mentioned in the first place as the premier contributor to this area of investigations. His works formulated these important designing principles that interpreted planning SMSS as an integral system:

* SMSS structure ought to conform to the organizational system of command and control and be hierarchically organized;

* the quantitative characteristics of the force (weapons) employment plan ought to be defined at several coordinated levels;

* the formalized description of processes in combat operations and conditions of strikes ought to be made consistently more detailed as a transition is performed from defining the plan's generalized characteristics in upper-level subsystems to finding weapon employment characteristics in lower-level subsystems;

* feedback relations ought to be used to coordinate models and methodologies at different SMSS levels in accordance with optimization criteria and formal constraints.

The said works demonstrated the constructive importance of the SMSS system design principles, including the use of feedback relations, by the example of computations intended to determine the plan's so-called generalized characteristics in SMSS upper-level models: strength level and strike objectives at stages in combat operations, approximate weapon details per target, parameters to control the search for rational aiming points. The present writers took determining the expedient structure of complexes of models that covered planning and coordination of solutions at different levels of the computation process a step further.

When accomplished, theoretical justifications and recommendations stimulated investigations into how to develop complexes of mathematical models and methodologies due to support all the main stages in the assault planning process, given system coordination of separate problems by their designation, minuteness of descriptions of the formalized combat employment methods, information connections, criteria, constraints and precision of calculations.

The theoretical investigations and methodological works the present writers as well as V.M. Tkachenko, A.V. Ivanov, V.G. Prudnikov, A.A. Sennitsky, E.M. Khamaganov and others performed at the turn of the 1980s assured a comprehensive solution to the problem of mathematical support of planning.

Works by N.A. Morozov developed the theoretical tenets of a methodology of mathematical models designed to provide a multifold evaluation of the effectiveness of massed strikes at different stages in planning and operations. The models formed a crucial part of the specialized mathematical support of preplanning activities envisaging combat force (weapons) employment. SMS also was due to adjust weapon employment methods in a period of threat, early in combat operations and in subsequent actions. A major contribution to models evaluating effectiveness of massed strikes against targets and their systems came also from Yu.N. Tronin, S.S. Bashirov, V.A. Krasilnikov, B.A. Lozhkin, and G.G. Pelikh.

Results of investigations into mathematical support of planning and effectiveness evaluation were realized in a system of mathematical models and methodologies for automated preparation of force (weapons) employment plans, which was developed, in the early 1980s, by a team of specialists at the 27th Central Scientific Research Institute of the RF Defense Ministry and by the research institutes of the AF services. The system was handed down to staffs for practical use. Its structural and methodological solutions were further developed in works by a number of staff members of the 27th Central Scientific Research Institute of the RF Defense Ministry, specifically S.A. Brusnitsyn, V.P. Gudyno, A.I. Zayikin, V.V. Zizyuk, I.N. Ivanov, N.A. Moiseyev, A.E. Monakhov, A.A. Muravyov, Yu.V. Pankov, I.V. Repin, A.V. Sinelnikov, and Yu.M. Smirnov. Their results, in turn, found a reflection in the improved planning SMSS developed for large computers and accepted in the early 1990s.

Its practical use by command and control agencies revealed that the realized rigid structural scheme of planning computations proved a very labor-consuming proposition under the new conditions characterized by the grown proportion of solutions validating possible variants of the combat force (weapons) employment concept and developing variants to plan limited strikes. New information technology--artificial intellect and expert systems (ES)--contributed to a higher intellectualization of the entire preplanning process. Now that staffs possessed numerous PCs, it became possible to further develop and improve operations SMSS on the basis of local PC networks (LN).

During the last few years the present writers supervised system, methodological and program research pursued by A.A. Protasov, A.B. Zelvin, V.V. Okuntsov, S.G. Kirilov, K.V. Nazarov and others, which led to structural coordination of mathematical, informational and program support of staff-based work to prepare operations at stages ranging from situation evaluation to timely plan-making to current adjustment of the plan in the course of the command and control process following the start of combat operations. Their works expanded the possibilities of determining rational variants of strikes at each of the potential enemy's targets with account taken not only of quantitative but also of weakly formalized planning factors in the process of man-machine dialogue with the use of PC LN and ES.

The new works also remarkably enhanced the reliability of the computation process and made it less labor-consuming due to the decomposition of the most sizeable part of the planning conditions related to the informational description of target situation considered. Realized in SMSS, the structural and methodological solutions approximated the system of computations to the conditions of planning of not only a massed strike but also selective (group, single) strikes at groups of targets and separate targets and made calculations involved in preparing variants of the plan less labor-consuming. The latter circumstance is important for ensuring the multi-variant nature and flexibility of planning where time is in short supply, including in a period of threat and during military operations in local conflicts that might conceivably escalate into large-scale ones.

The ways of further improvement of planning SMSS are associated with an increased promptness in developing a multitude of variants of the plan, considering the vagueness of the forecasted situation, and with changed views on forms and methods of strikes in the course of a large-scale military conflict.

Increasing precision and restricting labor intensity of the computation procedures is still a topical task where development of planning methodology is concerned and for the future. The following techniques that might be used to address these problems are theoretically grounded and tested in practice:

* using formal optimization methods at most important stages in the computation process;

* employing in practice operational assets of the systems analysis when constructing design procedures;

* combining in a single process that deals with difficult complex problems heuristic techniques and approximated methods with elements of formal methodology;

* direct participation of DM in evaluation of the quality of results and in control of problems as they get solved.

The formal methods are aimed to introduce in heuristic computation procedures some logically conditioned operations that make it possible to trace quantitative connections between the main structural components of a problem: information input, constraints, criteria and characteristics of the desired plan. Using the formal methods to control solutions is typical primarily of the "upper-level" models or the "generalized planning." The upper-level models enable establishing dependence between evaluations of the state of separate targets and indicators of the objective of a strike against a totality of targets (resources) under different conditions of employment of allocated weapons. The objectively existing dependence between the said characteristics finds a reflection in values of so-called plan search control parameters (target damage levels, importance evaluations), which are used in constructions of criteria (constraints) in the computation process that deals with the lower-level problems. Adjusting the controlling parameters in the design system's feedback contour provides the basis for a consistent improvement of the plan's sought-for characteristics in terms of the global criterion expressed in the minimized amount of assets expended.

The possibilities of improving the quality of solutions with the help of feedback are restricted by the increased complexity of problems under consideration. As mentioned above, the system operation involving the decomposition of a problem's conditions is used to overcome hitches generated by the difficulty and complexity of calculations. Yet decomposition of problems is often of informal nature and a source of danger that precision of calculations may deteriorate.

It is, as before, the planner's prerogative to estimate how much of a risk there is that informal decomposition of a problem's conditions will reduce the precision of solutions. In the system sense, problem precision evaluation may be interpreted as an operation to check the conformity of computation process output to expected output in terms of chosen criterion. The planner's inference as regards the error in solving a decomposed problem on the global criterion determines the adjustment of its results (formation of a controlling impact).

The process controlling solutions to problems in a system is of man-machine nature. Control is effected in the feedback contour embracing the decomposed procedure for solving an input problem as a whole. Aimed to boost the effectiveness of the solution of separate problems, adjustment of results may be performed by the DM, the computer or their system team.

The above-mentioned tenets of the methodology of a complex of planning ICP (models, methods) have been concretized in the formulations of separate problems and their solution methods and are realized in improved SMSS.

Col. V.A. SOBOLEVSKIY (Ret.)

Leading research fellow, the 27th Central Scientific Research Institute of the RF Defense Ministry

Doctor of Technical Sciences

Col. V.V. SUKHORUTCHENKO

Deputy chief, the 27th Central Scientific Research Institute of the RF Defense Ministry

Doctor of Technical Sciences