基于遗传算法的供水系统一级优化调度研究

根据多水源多水塔供水系统的网络特性，以测压点压力宏观模型、水源供水量和水头之间的关系模型，替代管网水力平衡方程，并以泵站的供水压力、水塔初始水头为决策变量，建立了供水系统一级优化调度模型．采用罚函数法将各种约束转化为目标函数，然后利用乘除法将多目标优化转化为单目标优化问题，最后利用遗传算法求解．针对遗传算法容易早熟等不足，将遗传算法和模拟退火技术相结合，并采用自适应交叉和变异率的解决方法．算例计算表明，该算法具有较强的适用性．

Optimal Operation of Water Supply Systems Based on Genetic Algorithm

According to network properties of muhisource- multitank water supply system, the hydraulic equations are replaced by macroscopic nodal pressure model and the model of relationship between supply flow and head of water source. By using pump - station pressure heads and initial tank water levels as decision variables, the optimal allocation model of water supply between pump - sources is developed. Penalty function methods are used for converting constraints into the objective functions, and multiply method is adopted to transform multiple objectives into single objective. Then genetic algorithm is introduced to solve the single objective optimal problem, and in allusion to the shortages such as premature of genetic algorithm, the solving method is brought forward in which genetic algorithm is combined with simulated annealing technology and self- adaptive crossover and mutation probabilities. The application example indicates that this algorithm is feasible.