An Improved Evolutionary Algorithm without Penalty Function for Multi-objective Optimization Problems
this paper presents an improved genetic algorithm without penalty function for multi-objective and multiconstraint optimization problems. A conventional genetic algorithm with penalty function probably falls into local convergence or gets infeasible optimal solution, if the penalty function is improper. In order to avoid above situations, some improved strategies such as dynamically dividing populations, adaptive probabilities of cross and mutation and multi- selection strategies are contributed to enhance the optimization performance of genetic algorithm. Simulation result and engineering application show that the new algorithm can effectively handle multi-constraint and multi-objective optimization problems and has good optimization performance.
genetic algorithm multi-objective optimization adaptive cross and mutation probability
Dongmei Cheng Changhua Qiu
College of Mechanical and Electrical EngineeringHarbin Engineering UniversityHarbin, China College of Mechanical and Electrical Engineering Harbin Engineering University Harbin, China
国际会议
哈尔滨
英文
407-411
2011-01-18（万方平台首次上网日期，不代表论文的发表时间）