改进的混沌优化方法在电站机组负荷分配中的应用

利用搜索过程中的先验知识，提出了一种改进的变尺度混沌优化算法，对电厂机组负荷分配这一类高维、非线性问题进行寻优。该算法利用混沌运动的遍历性和随机性进行全局搜索，结合下降法思想在当前最优解的局部空间搜索，提高算法的收敛速度。通过罚函数，将非线性约束问题转化为非约束问题求解，在寻优操作过程中检验解的有效性。并采用目标函数中负荷取值区间判别方法，降低优化问题的规模，减少了计算次数。与传统遗传算法相比，改进的混沌优化算法获得了更好的效果。

Application of Improved Chaos Optimization Algorithm for Load Allocation Among Power Generating Units

An improved mutative scale chaos optimization algorithm, using knowledge drawn from preceding iteration processes, is being presented to solve high dimensional non-linear optimization problems, like load allocation among power generating sets. This algorithm makes use of the ergodicity and randomness of chaotic motions for global search, in combination with decreasing algorithm for local search in current space of best solutions; herewith speeding up convergence of the calculation process. For obtaining a solution, the non-linear constrained problem is converted into a non-constrained problem, by making use of a penalty function, with the validity of the solution being verified in the course of the process. Moreover, to reduce the scale of the optimization problem and cut down times of iteration, a load range discrimination method is used with the object function. Compared with traditional genetic algorithm, a more satisfactory outcome can be obtained with this improved chaos optimization algorithm.