Genetic Programming And Its Applied Research, Structural Optimization Of The Cut Roadway In The Mining

Many problems in mining engineering relate to the dealing with abundant knowledge basedon experience and uncertainty. At present, however, there is still oo convincible formula fordecision-making mining engineering. for all mathematical and mechanical means to miningproblem solution are not satisfactory due to the complexity of mining procedure and geology conditions. Up to now, it mainly depends on experience and analogue among realistic works inorder to find out the rules in mining and solve mining problems, with the help ofHeuristic Method.Genetic algorithms (GA) and genetic programming (GP) are new technologies foroptimization, which simulate inherit and evolution in the nature and get solutions throughreproduction, crossover and mutation operations.In this paper, genetic programming is applied to optimize the opening networks ofunderground mine, which realized automation and Intelligence in mining design. Applicationsoftware is also developed, by which some parameters have been tested. Given actual geologicaland tech-economical conditions, it can find out satisfactory solutions to network design throughgenetic operations such as reproduction, crossover, mutation,etc.Intron and covergence of genetic programming are studied by means of symbolic regressionand formula discovery in GP. False convergence is defined, and its reason and the way tominimizing its effects are discussed. Intron's effects on convergence of GP are studied. By testinga great deal of examples, the effects of reproduction and crossover on intron are discovered. Then,an improved crossover, Single-parent crossover, is recommended, and two of its execution meansare recommended.