一类针对带约束优化问题的进化规划算法

提出了一种适用于求解带约束优化问题的进化规划方法，其中关键的变异算子采用基于行为的架构，事先设计一系列子变异算子，如使得个体适应度函数值趋向最小方向的变异算子、逃避约束方向的变异算子、种群总体平均适应度函数值趋向最小方向的变异算子等，通过加权平均的方法决定总变异方向．结合小生境技术及最优个体保存的选择策略，该算法能在同时保证种群的多样性和个体的全局最优性的情况下快速地求得带约束条件下的最优解．仿真结果表明，该进化规划算法是可行的．

An evolutionary programming algorithm for constrained optimal planning problems

An evolutionary programming algorithm for constrained nonlinear programming problems was proposed. The organization of the most pivotal mutation sub-operator in this algorithm was based on behavioral architecture. Several mutation sub-operators were designed for the practical needs. A weighted method was used to decide the whole mutation direction. Using niche technology and the best-be held strategy to assure the diversity and the global optimum of the population, this algorithm can gain the optimal solution quickly. At the same time, the problem of path planning of the mobile robot can be classified into general nonlinear programming problems based on the district-partition modeling method. The application of the path planning of the mobile robot in dynamic environment showed that the algorithm was feasible and efficient.