технические науки / 6.Электротехника и радиоэлектроника

O. S. Bordiug 6 - year student
Government Nautical Technological University of Kerch.

APPLICATION of Precedential Structures for Electromechanical Systems

Precedent (Case-Based Reasoning) is a specification of the sequence of actions performed by a system or subsystem while interacting with external actants. The precedent includes a problematic situation and its solution (fig. 1). If you find a problematic situation, the decision should be based on previous precedents. The information is packed into a container called a precedent and is stored in the repository of precedents. Traditional means of database management, specialized knowledge base servers, multidimensional database, etc. may be used as repository of precedents. 

Fig.1. - Decision-making based on the precedents.

The state of S is defined as a sum of values parameters of its factors .

Let denote the sum of values of a system (finite or infinite) as . The state of a system may be characterized as : where - set of committed actions (procedures) - the set of objects - the set of object properties - set of relations between objects.

In order to set a situation in a such space of states, one need to mark the properties of objects and relations between them and also to point out the set of performed actions that took, take or will take place at a certain period of time .

Intelligent systems are an integral part in the field of electrical and electromechanical systems today. Discrete modes of transmission and transformation of signals using the principle of signals’ discretization in time or level or in time and level simultaneously are widely used in modern automatic control systems.

Autoregressive moving-average model (ARMA) is one of the mathematical models used for analysis and prediction of stationary time series in statistics.

ARMA-model generalizes two more simple time series models - model autoregressive (AR) and model moving average (MA).

It is obvious that while modeling the dynamics of control objects, ARMA-model should have adaptive properties [1], which may be realized while using lagged operator (B):

. (1)

When you apply a polynomial of log

, (2)

to variable x, then we get:

The operative of the disparity (absolute growth) Δ, defined as (1-B), the second disparity can be determined by the following pattern:

If we apply operative logs (1) to obtain the model ADL-ADL (p, q)-model in operative form: , where f (B) and g (B) - polynomials.

Disparity equation (3) describes the autoregressive process (AR (p)-model with distributed log criterion parameter) of the moving average (MA (q)-model with a distributed lag of regressors).

Using this model we can get more precise indicators of the value of the largest log, based on the study of finite disparity transition function of the object forming the output signal ARMABIS-structure.

References:

1. Щекин В. П. AARMA-эмуляторы динамического состояния нелинейных нестационарных систем / В. П. Щекин, О. В. Щекина // Вісник КДПУ ім. М. Остроградського. – Кременчук, КДПУ, 2009. – Ч. 1, № 3(56) – С.127 – 130.

2. A.Dubois D.Case-Based Reasoning: A Fuzzy Approach.- Lecture Notes in Computer Science / Dubois D., Esteva F., Garcia P., Godo L., Lopez de Mantaras R., Prade H.- Springer.- 1999.- Vol.1566.- Pp.79-91. (ru)