Applied heuristic information technologies for automating power system management in mining and metallurgical complexes

Abstract

This paper addresses the problem of constructing a knowledge base for an expert system to ensure the quality of the post-contingency state of an electric power system based on the use of heuristic methods of dispatch control. The aim of the article is to study a potential approach to constructing the knowledge base of an expert system to support the decisions of dispatch personnel during the management of the post-contingency state of an electric power system. This study proposes developing the knowledge base using the sensitivity matrix of the controlled operating parameters. The research methodology relies on the principles of the theories of electrical networks and power systems, automatic control, experimental design, artificial intelligence systems, and mathematical statistics. As a result of experimental design, dependencies linking the parameters of the normal pre-contingency state and the network structure with the optimal magnitudes of control actions were obtained. The scientific novelty lies in the integration into a single knowledge base of dispatch operational guidelines during the elimination of emergency states, as well as the coefficients of the sensitivity matrix obtained from computational experiments, taking into account the configuration of the electrical network. The results of the work are structural models of the expert system knowledge base, a method of experimental design to construct the sensitivity matrix of regression polynomials of the efficiency of dispatch actions, and algorithms for the functioning of the expert automation system for heuristic control of the post-contingency state of the power system. The heuristic control system provides static stability support, power flow regulation, real-time state adjustment, display of current state parameters, and monitoring of the actions of operating personnel. The practical significance of the obtained results lies in the intelligent expert automation of post-contingency power system state management. The expert system can be integrated into the control subsystem of a general automated dispatch control system, thereby increasing management efficiency and the quality of the post-contingency state. Furthermore, the system can be used to train and verify the professional knowledge and skills of operational dispatch and technical personnel.