Improved Parallel Genetic Algorithm Based on Resetting Strategy and Hash Table

Abstract

Abstract In the process of parallel genetic algorithm (PGA) convergence, the probability of generating repeated individuals in a new generation is gradually increasing, which leads to repetitive computations of fitness values. In order to improve the time-efficiency and quality of solutions, an improved algorithm based on resetting strategy and hash table (RHPGA) is proposed. On the one hand, RHPGA takes advantage of the low time complexity of hash table lookup to reduce fitness values’ calculation of repeated individuals. On the other hand, RHPGA uses a resetting strategy to improve the optimal solution searching ability. By solving a set covering problem, PGA and RHPGA are compared. The experiment proves that RHPGA can almost increase the time-efficiency by 300% and quality of solutions by 15%.