为解决强制进化随机游走算法(random walk algorithm with compulsive evolution,RWCE)应用于换热网络综合时进化停滞的问题，提出了一种伴随优化策略（CO-RWCE)：对种群中各个体优化进程进行监控，当个体因接受差解陷入长期进化停滞时，将该个体历史最优解回代给个体以调整优化方向；若多次回代后个体仍未进化，则将全局最优解传给个体并摄动以实现重生。在回代或重生后，采用一种游走概率递减技术，控制游走变量个数以提升搜索精度。优化结果表明：改进策略增强了个体自身进化能力，有效提升了算法搜索能力。

To solve the deficiency that random walk algorithm with compulsive evolution (RWCE) could be in evolutionary stagnation when applied to heat exchanger network (HEN) synthesis,a company-optimization strategy was proposed.The optimization processes of all the individuals in the population were monitored.For the individuals facing the long-term evolutionary stagnation resulting from accepting bad solutions,the back substitutions of individual historical optimums were executed periodically,aiming at adjusting individual optimization direction compulsively.Then the individuals without evolving after several back substitutions were initialized with the global optimum exerted with random perturbation to achieve the renascence.After back substitution or renascence,a decreasing “walking” probability technique was developed to improve the search precision by controlling the walking variables.The optimization results proved that the proposed strategy strengthened the individual self-evolution ability and thus effectively improved the search ability.

TK124

国家自然科学基金(51176125)；上海市科委部分地方院校能力建设计划(16060502600)