On the Construction of Organizational Structures in the Development of Automated Control Systems

Yu. V. Karyakin

(Presented by the scientific-technical seminar of the Department of Automation and Telemechanics)

Recently, in the field of automated control system (ACS) development, a clear tendency has emerged toward a transition from the creation of ACS subsystems to an ACS for the entire organization [1]. Such an approach requires a deeper and more complete study of the organization, the laws of its construction, functioning, and development.

On the way toward this, researchers have begun to use the method of analyzing organizational structures [2, 3, 4]. The essence of this method is that the organization under study is considered successively in various aspects, as a result of which the corresponding structures are constructed. Each of the latter describes one definite side of the organization. The identification of such structures makes it possible to describe the organization by a set of models presented in a certain formal language, which makes it possible to analyze the structures with the aid of a computer.

This article develops one of the possible approaches, set forth in [5], to the selection of a nominal set of structures that make it possible to model the principal aspects of an organization’s activity.

The organizational structure is described by a directed graph whose nodes represent operators (employees of the organization), and whose arcs represent relations of their subordination, i.e., the sequence of functioning.

The functional structure is represented in the form of a directed graph whose nodes reflect the functions performed by the elements of the organization (subdivisions, individual groups, and operators), and whose arcs represent relations of their subordination, i.e., the sequence of functioning.

For the construction of the functional structure of an organization, the definition of the concept of function is of considerable importance. The author has not encountered in the literature a definition of this concept that satisfies the requirements for constructing a formalized structure; therefore it became necessary to propose such a definition.

By a function of a system (or of its element) one should understand a property of the system (element) considered in a certain target aspect. From this general definition it is not difficult to proceed to the definition of a control function for economic systems (objects). By a control function we shall understand such an interaction of two objects in which one object controls the state of the other. In this case we shall say that the first object possesses a control function (performs a control function) with respect to the second object.

From the practice of the functioning of organizations it is known that there are two different forms of management: techno-economic and operational. Correspondingly, two kinds of management functions are distinguished: functions of the first kind manifest themselves (are performed) at certain moments after intervals of time of a definite length, whereas functions of the second kind manifest themselves (are performed) at random moments of time. A second distinction is that management functions of the first kind, in solving management problems (see below), use absolute values (levels) of parameters, by which the object is managed, as input and output information, whereas management functions of the second kind use for this purpose relative changes in the parameters or their deviations from some specified level. The division of management functions into management functions of the first kind and management functions of the second kind makes it possible, when constructing the functional structure, to divide it into two independent or weakly dependent parts, allowing them to be analyzed subsequently.

For a deeper analysis of the functional structure, it is necessary to split the functional graph in both its parts (operational and techno-economic management) into two oppositely directed graphs.

Let us consider the system of techno-economic management. Here the direct graph reflects the interrelation of planning functions, and the reverse graph reflects the interrelation of reporting functions. We shall call these functions elementary management functions, in contrast to the management functions defined above. In each functional node, there are thus elementary functions of two types. Functions of the first type assign to the managed object a certain level of its parameters, forecast for a definite moment of time in the future, while functions of the second type carry out “feedback,” i.e., they communicate to the managed object the true value of the parameters of the managed object at definite moments of time. The number of control loops formed by pairs of elementary functions corresponds to the number of independent parameters by which the objects managed by the given functional node are described.

The functional structure of the operational management system is likewise represented by two oppositely directed graphs. The elementary functions in this case are the functions of operational control and regulation, which form control loops identical in their structure to those considered.

By the economic structure we understand the totality of economic and operational indicators, together with the sequence of their formation in the process of the organization’s functioning. Economic and operational indicators are those parameters by whose system the functioning organization is described and upon which the management system and its functional elements act. The economic structure of an organization can be represented by a graph, each node of which reflects an indicator, while the arcs represent the paths of formation of reporting indicators and, at the same time, planned ones. The graph has the form of a tree with the number of vertices equal to the number of indicators planned from outside. The number of indicators is exactly equal to the number of control loops of the first type.

Information Structure

As a unit of observation of information in constructing the information structure of an organization’s management system, it is convenient

take a task. We shall consider and define the tasks solved in an organization’s management system in the informational aspect. In this connection, the following definition of a task appears satisfactory.

By a task we shall agree to understand the simplest information system, having independent significance in the system under consideration and providing for the input of information, its processing according to special algorithms, and output in a certain formalized form [6].

By the informational structure of a management system we shall understand the totality of tasks solved in the system in their interrelation. The informational structure is represented in the form of a graph, the nodes of which reflect tasks, and the arcs—the sequence of their solution.

Each task solved in the management system must be uniquely assigned to one definite management function. A task is implemented with the aid of one of four elementary functions.

According to the accepted definition of a task, the latter is characterized by three parameters: input information, output information, and an information-processing algorithm. By singling out the first two parameters of the tasks in the entire informational structure, we obtain a formalized scheme of information flows. By singling out the third parameter, we obtain the so-called algorithmic structure of the management system [2].

Analysis of the algorithmic structure makes it possible to divide the entire totality of tasks solved in the management system into tasks of machine and human management.

The isolation of the described organizational structures makes it possible to construct a comprehensive model of the management system for the purpose of analysis and the development of recommendations for improving management both at the stage of developing an ACS and in the course of its operation.

LITERATURE

1. A. A. Modin. Development of a model of management structures. In: Information and Models of Management Structures, ed. by Academician N. P. Fedorenko, Moscow, “Nauka,” 1972.

2. A. P. Kategovich, A. K. Kagramanov. Experience in developing the structure of an automated management system for a large complex of discrete-continuous production. Proceedings of the Scientific-Technical Conference “Problems in the Development and Introduction of ACS at Machine-Building Enterprises.” Novosibirsk, June 1972.

3. I. A. Malkovich. Problems of the theory of indicators of economic systems and the practice of ACS design (ibid.).

4. V. M. Ubeiko. The use of graphs in the implementation of processes of machine design of ACS (ibid.).

5. G. F. Mazurek, Yu. V. Karyakin. On one approach to designing an ACS for an organization. (The article is included in the present collection.)

6. M. S. Ilyin, T. Ya. Danielyan, A. N. Markin. Investigation of economic information systems. Statement of the problem, existing approaches, methodological guidelines for investigation. In: Theory and Practice of Designing Systems for Processing Economic Information, vol. II, 1968.