模拟退火遗传算法的泵站优化运行

针对泵站在不同流量扬程要求时，水泵运行组合不合理，导致泵站运行效率低的问题，提出以泵站总功率最小为目标函数的优化运行数学模型.此模型以水泵扬程、总供水量和调速泵的调速率为约束条件，利用模拟退火遗传算法确定并联运行泵的台数、调速泵的调速率及泵的流量分配，实现泵站的优化运行.该算法引入了不可行度对每代种群做初始选择，并利用不可行度取代传统的惩罚函数，平衡目标函数最优化和满足约束条件这两方面的要求.最后通过实例验证了本算法在求解泵站优化调度问题方面的可行性.

Optimization of Pumping Station with Simulated Annealing and Genetic Algorithm

To cope with the problem of low efficiency of pump station due to unreasonable combination of the pumps, a mathematic model for optimizing the combination is proposed leading to minimal power consumption. The constraining conditions in the model are the pump head, total water supply and rotation speed of the pumps. A simulated annealing genetic algorithm is used to determine the number of pumps in operation, the rotation speed and the flow rate of each pump when the pump station is in an optimal state. The infeasibility degree is introduced into the algorithm to choose a suitable result from every population and to balance the demand between the target function and the constraining conditions. The result of an example shows feasibility of the present algorithm for pump station optimization.