ON ONE APPROACH TO SOLVING THE PROBLEM OF SYNTHESIZING UNIVERSAL AUTOMATED CONTROL SYSTEMS

G. P. Pavlenko, V. M. Razin

(Presented by the scientific seminar of the Department of Computer Engineering)

Despite the tendency toward universalization, automated control systems (ACS) for complex objects remain, for the most part, specialized systems that provide control of the equipment of objects of certain types.

In this connection, questions of synthesizing universal ACS remain highly relevant.

Universal ACS may be assigned to the class of adaptive systems. In [1] it was shown that, in order to ensure high flexibility, such systems must possess the capability of promptly replacing some system components with others having new characteristics. Here the latter are functions of changes in the control objects (CO). An ACS may also increase its adaptive capabilities by increasing the number of its variables (components), i.e., in practice the system must have a very extensive arsenal of constituents possessing various technical characteristics adequate to the parameters of the CO.

In addition, ACS can be combined, which will make it possible to increase the adaptive capabilities of the systems. The creation of such combinations may be identified with the development of a range (series) of modifications of universal ACS capable of adapting to certain groups (classes) of control objects.

Therefore, in what follows the term “universal ACS” will be used with reference to an ACS that is universal for a given class of CO.

Obviously, the above-mentioned qualities of an ACS can be achieved through a certain redundancy of the technical components of the system, which, naturally, must be limited by their cost.

Thus, the solution of the problem of synthesizing universal ACS may be divided into the following stages:

— typification of the system structure;

— development of a complex of unified technical components (UTC);

— synthesis of a system on the basis of the UTC;

— evaluation of the effectiveness of applying the universal ACS according to the selected criterion.

For typifying the structure, the method of decomposition may be used, the essence of which consists in dividing the system under consideration into parts, and solving the problems with respect to these individual parts, rather than to the entire system as a whole.

As a result of using this technique, in [2] a variant of the structure shown in Fig. 1 was obtained.

The main task of the stage of developing the UTC is the transformation of the functional diagrams of devices obtained as a result of typifying the ACS structure into a set of standard structurally and functionally complete components.

Fig. 1. System structure: CU — control device; IDD — information distribution device; RAM — random-access memory; IPD — information-processing device; OID — information-output device; ID — input device; ALD — long-term storage device; CD — communication device; PPID — primary-processing device; SFA — device for forming stimulating actions.

The solution of such a problem can be reduced to the development of methods of structural analysis of a representative sample of functional diagrams in order to identify in them frequently recurring, logically connected sets of elements (standard structures) and to distribute the standard structures obtained among blocks (subblocks) in such combinations that would subsequently make it possible to cover the functional diagrams of devices of universal ACSs with components of the set, making maximum use of the elements contained in them [3].

For such methods, corresponding algorithms and working programs have been developed and tested [3, 4], which can serve as a basis in developing a UTC complex for universal ACSs.

The problem of synthesizing a universal ACS consists in ensuring the required characteristics of the devices and the specified throughput capacity of the system by varying the composition and number of components in the devices.

The synthesis of devices of a universal ACS whose parameters have no direct dependence on the characteristics of the control object (input, storage, and information-processing devices, etc.) does not present serious difficulties and can be carried out using existing methods [5, 6] based on estimating device performance from the specified volume of information in the system. As for devices connected with the object (OUSPО, SFA, CD), whose characteristics are determined mainly by the parameters of the CO of a given class, another approach has been proposed for their synthesis [7], taking into account the initial uncertainty of the object of control.

In [7] a universal ACS is interpreted as a multichannel mass-service system whose task consists in processing the flow of demands (requests) entering it for the control of systems (devices) of equipment of the entire diversity of complex objects of a given class.

One of the main features of such a system is that each request at the system input is characterized not only by the arrival time \(t_j\), but also by the parameters \(a_0,\ a_1,\ a_2,\ldots,\ a_n\), i.e., each request is an \((n + 1)\)-dimensional vector of the form

\[ V_j = V(t_j,\ a_{0j},\ a_{1j},\ a_{2j},\ldots,\ a_{nj}) \]

in the space of the parameters \(a_0,\ a_1,\ a_2,\ldots,a_n\).

The purpose of the behavior of such a system is to solve the following problems:

— determining the number of identical servicing instruments (i.e., the number of channels, blocks, or devices of the ACS) depending on the types of requests \((\alpha_0)\), the intensity of their arrival \((\lambda)\), the service time \((t \text{ serv})\), and the waiting time \((t \text{ wait})\), for a given probability of the number of unserviced requests \(P(N)\);

— optimizing the characteristics of the servicing instruments \((\beta_i^k)\) depending on changes in the parameters of the requests \((\alpha_j^k)\) entering the system, for a given probability of the number of unserviced requests \(P(N)\).

The indicated problems are solved by the method of statistical modeling on the “Minsk-22” computer, for which the corresponding modeling algorithm and working programs have been developed.

To evaluate the effectiveness of using a universal ACS, we introduce the following criterion:

\[ K_{\mathrm{e}}=\frac{K_{\mathrm{p}}}{K_{\mathrm{c}}}, \]

where

\[ K_{\mathrm{p}}=\frac{n}{N} \]

is the coefficient of adaptability of the system;

\[ K_{\mathrm{c}}=\frac{C_{\mathrm{u}}}{C_{\mathrm{s}}} \]

is the cost coefficient;

\(n\) is the number of types of COs whose control is provided by the given ACS;

\(N\) is the total number of types of objects of the given class;

\(C_{\mathrm{s}}\) is the cost of using specialized ACSs;

\(C_{\mathrm{u}}\) is the cost of using a universal ACS.

From the expression given it is evident that the effectiveness of using a universal ACS is increased mainly by increasing the number of types of COs whose control is provided by the given system, when the cost of using specialized control systems for these COs is high; i.e., the expediency of developing universal ACSs for a given class of COs, on the assumption that the system provides control of all COs of the given class, or for a specified \(K_{\mathrm{p}}\), is determined by the ratio of the cost of using specialized and universal control systems.

The approach described above to solving the problem of synthesizing universal ACSs was used in designing an automated control system for a given class of objects on the basis of the UMShN “Dnepr-1.”

REFERENCES

1. G. P. Pavlenko. On the question of the synthesis of universal automated control systems for complex objects. In: Automated Control Systems and Technical Means. Omsk, SKB PA, 1972.

2. G. P. Pavlenko. On the similarity of the structures of a universal automated control system and a control computer. Abstracts of reports of the All-Union Scientific and Technical Seminar “Problems of the Transmission, Conversion, and Processing of Information by Automatic Means.” Rostov-on-Don, 1973.

3. V. K. Pogrebnoi. Development and investigation of algorithms for automating the technical stage of designing computing devices. Dissertation. Tomsk, TPI, 1970.

4. M. Ya. Kulikov. Questions of the investigation and development of a rational set of blocks in designing local discrete devices of control systems. Dissertation. Omsk, SKB PA, 1973.

5. I. V. Kuzmin. Evaluation of the Effectiveness and Optimization of Automatic Control and Management Systems. Moscow, “Sovetskoe radio,” 1971.

6. Principles of Constructing Automated Control Systems. Edited by P. I. Kuznetsov. Moscow, “Energiya,” 1969.

7. G. P. Pavlenko. On one model of an automated control system universal for a given class of complex objects. In: Automated Systems and Technical Means. Omsk, SKB PA, 1972.