APPLICATION OF THE GENETIC ALGORITHM (GA) IN THE SELECTION OF GENERATORS FOR PRODUCTION AT PERUCHICA HPP 93

In order to make the best possible use of the water potential and ensure the balance reserve for the needs of energy system regulation, it is necessary to optimally plan the operation of the hydroelectric power plant. The values of hydropower production are set in defined time intervals based on short- term planning obtained by optimizing the production mix from the balance responsible party, and data on water resources and availability of generators are also known. Under these conditions, optimal use of generators and the available amount of water from reservoirs should be foreseen. In HPPs with multiple generators, the selection of generators depends on pipeline losses, mechanical losses on turbines, electrical losses on generators, etc. The goal of planning should be the optimal use of water potential and the use of generators, and the result of the optimization is flows through generators and pipelines, such that the given production of the hydroelectric power plant is ensured. This objective is described by an objective function, which is defined by a non-linear relationship of several variables, in this case the degree of utilization and the flow of each unit individually. The obtained results for flows represent input data for further planning, mainly for planning the provision of a sufficient amount of water in the input building, for the production of the power plant, which practically means maintaining the height of the input building in the given range. The paper presents the procedure for obtaining and solving the objective function using a genetic algorithm (GA), with the simplest restrictions on the input variables. The principle of the genetic algorithm consists of evaluation, selection, crossover and mutation within a population. Through the iterative procedure of these four principles, through several generations, the solution of the objective function is obtained.