基于单亲遗传算法的区域供冷系统管网的布置优化

Optimization of district cooling pipe network layout with single parent genetic algorithm

区域供冷树状管网布置优化属于典型的组合优化问题.针对区域供冷管网系统特点,建立了以管网年度费用最小为优化目标的数学模型和约束条件.以图论和遗传算法为理论基础,应用改进的单亲遗传算法对区域供冷树状管网进行了优化布置,并对编码方案,遗传算子设计,适应度函数进行了设计.应用实例表明,单亲遗传算法直接以管网年度费用最小为优化目标,能够获得年度费用最小的布置方案,与Dijkstra算法对比,算法的寻优效率较高,收敛性和稳定性较好.

Optimization of district cooling pipe network layout is a typical combination optimal problem. The mathematical model with constrained conditions is built with the minimal annual cost of pipe network as the optimal layout objective function by the characteristics of the district cooling pipe network. Based on the graphic theory and genetic algorithm, an improved genetic algorithm, namely single parent genetic algorithm （SPGA）,is introduced to optimize a district cooling tree pipe network layout,and the code technology , transposition operators,as well as the fitness function are designed. A practical application demonstrated that SPGA uses the minimal annual cost of pipe network as the optimal layout objeetive,a set of district cooling pipe network layout with the minimal annual cost can be acquired. In comparison with the Dijkstra algorithm, the SPGA can save invest ment and running-cost which possesses the advantages of high searching efficiency, quick convergence and good stability.