Development of a Hybrid Algorithm for Creating Escape Routes Based on Fuzzy Control

In the article, the task of developing a hybrid algorithm for constructing escape routes based on a genetic search algorithm and a fuzzy control unit is considered. The task of increasing the effectiveness of existing methods of solving the problem of evacuation of the population in emergency situations is urgent. To ensure mass evacuation of the population, evacuation plans are needed that take into account the ways of transporting the evacuated population to places of temporary deployment. Having data on the coordinates of the most potentially dangerous objects, you can write an expression for the function of estimating the time of evacuation of the population. It is proposed to use a genetic algorithm as the main tool for searching and laying alternative escape routes and routes. To maintain a balance between improving existing solutions and finding new solutions, it is proposed to introduce a fuzzy logic controller into the genetic algorithm. To increase the efficiency of work, it is proposed to supplement the fuzzy logic controller with a learning unit based on the fuzzy Wang-Mendel production model. The proposed hybrid method was implemented as a software application based on the Qt 5.6 framework. The QCustomPlot widget was used for plotting and visualizing data. During the testing of the developed software application, a series of tests were conducted. The results show that using a fuzzy controller can improve the quality of the solution by an average of 20%. Analysis of graphs of changes in the average and minimum values of the objective function, as well as graphs of changes in the probability values of genetic operators, allows us to conclude that the fuzzy control unit effectively regulates the values of the algorithm parameters depending on the assessment of the evolutionary process.