Study on Parameter Inversion of Displacement Based on Homotopy Theory Genetic Algorithms

摘要

The principle and the main characteristics of Genetic Algorithm (GA) were stated. Crossover was a main process producing new generation through genetic genes recombination at the whole exploration space in GA. But some traditional crossover mechanism made the extent and profundity of cross among operators limited. There were some thorny issues such as crossover efficiency slowly, some operator perdition no chance crossing, the population diversity playing down quickly etc. In order to improve genetic cross-strategy, Homotopy Theory Genetic Algorithm (HTGA) was presented by combining the improved homotopy theory and GA. Homotopy genetic curve directed GA genes operating which enhanced genes' interior exchange of population. Then the HTGA was applied to the multi-parameters inversion of cavern displacement in geotechnical engineering. Numerical example shows that HTGA is of strong ability to obtain global optimization solution, and its performances are superior to those of single GA.