Method for selection of optimal composition of operating units in power plants by genetic algorithm

Abstract

The problem of selecting the optimal composition of operating units in power plants when planning short-term modes of power systems is a complex problem of nonlinear programming. Its solution usually boils down to determining the starting or stopping units in power plants for each interval of the planning period. Despite the current existence of many methods and algorithms for solving this problem, the issues of their improvement in the direction of increasing the accuracy of calculations and the reliability of convergence of iterative process, taking into account the new conditions of the functioning of power systems, as well as the capabilities of modern computing facilities, remains as an important task. In this paper, we propose a new algorithm for the selection of the composition of operating units in power plants based on a genetic algorithm, which, to a certain extent, meets modern requirements. The solution to the problem is carried out in two stages. At the first stage, generalized energy characteristics of the plants participating in the optimization are constructed. At the second stage, based on optimal coverage of the power system load schedule by all plants according to their obtained generalized energy characteristics taking into account all limiting factors, by the genetic algorithm, the optimal compositions of the operating units are determined. High accuracy of the result is ensured by the direct use of energy characteristics of plants, usually set in a tabular form, and the ability of the genetic algorithm to solve multi-extreme problems without any simplifications.The results of computational experiments, which confirm the high efficiency of the proposed algorithm, are presented.