摘要
采用遗传算法对约束函数进行优化也得到了广泛关注与认同,但在解决高维、多约束优化时存在"早熟"和发散现象。采用整合数列编码和实数编码遗传算法,设计出序列-实数混合编码遗传算法,并利用次算法对多个典型约束函数进行优化测试,对算法性能进行测试评估,结果表明,混合编码遗传算法在求解高维复杂问题的性能较实数编码遗传算法有较大改善。
Using genetic algorithm to optimize the constraint function is concerned and identified. The phenomena of ' precocious' and divergence are existed when dealing with the high-dimensionalmulti-constrained optimization. This paper designs the sequence-real genetic algorithm based on hybrid encoding by integrating the sequence encoding and real encoding genetic algorithm. It utilizes its optimization results of several typical constraint functions to test and evaluate the algorithm. The results show that in the capability of solving high-dimensional problems,genetic algorithm based on hybrid encoding has bigger rise than real encoding.
出处
《仪表技术》
2015年第8期33-35,51,共4页
Instrumentation Technology
关键词
遗传算法
混合编码
约束函数
genetic algorithm
hybrid encoding
constrained function
