Studies of joint work of genetic algorithm and coordinate search algorithm to optimize temperature of heated indoor areas

The article is focused on studying optimization algorithms that are relevant both for solving applied problems and for studying the artificial intelligence in general. Optimization methods are used to solve environmental problems including the issues of energy saving. It is important to solve problems of optimizing the thermo-hydraulic modes of buildings (as a part of the “Smart City” project), in particular, problems of eliminating temperature imbalance in terms of saving thermal energy and improving the microclimate in apartments. There is shown a mathematical formulation of the problem of optimizing the temperature modes of the indoor areas using adjustable devices. A hybrid algorithm applied to solve the problem has been described. The considered algorithm combines two optimization methods: a coordinate search method and a genetic algorithm. Thus, the stochastic component (element of the genetic algorithm) and the gradient component (element of the coordinate search method) are used in the hybrid algorithm. A description of the above algorithms is given including the mathematical apparatus used and the design formulas. The results of the numerical experiment for the suggested algorithm are presented. These results are compared with the results of applying the genetic algorithm and the method of coordinates search separately. There has been confirmed the hypothesis that in order to increase the efficiency of solving the considered class of problems, it is necessary to combine the genetic algorithm and gradient methods. At the same time, it has been inferred that in cases of low thermal power of radiators, optimization of the hydraulic resistance of valves is not sufficient, thermal insulation measures and replacement of radiators are also required. The practical value of the work lies in the possibility of solving the problem of saving thermal energy in the housing and communal services system.