论文部分内容阅读
化工园区供电网络优化设计是国家综合管理部门、电网工程项目法人及融资方进行项目决策的重要依据。化工园区内企业在生产活动中极易释放易燃易爆气体,供电危险极大。主要目的是运用程序优化方法,找到一种既经济又安全的供电线路布局方案,最终得到基于风险和成本的安全决策。基于前人在计算机智能算法优化数学模型得到的经验,运用图论Floyd算法和GA遗传算法研究了图论算法和人工智能算法在基于风险和成本的化工园区供电网络优化中的设计和应用。设计了一个化工园区,总降变电站为源点,8个可选择的化工园区内具有火灾或爆炸高危险性生产企业位置为网络节点。建立了无浓度约束范围和有浓度约束范围两种模型,并在有浓度约束范围模型的优化中运用广度优先遍历和深度优先遍历相结合的思想。给出了3套方案并且得到了理想的优化结果。首次应用计算机智能算法优化的方法研究了化工园区供电网络安全性,研究表明,图论算法和人工智能算法在化工园区供电网络优化上的应用是可行的。对于网络节点较多的实际化工园区供电布局,GA遗传算法一般能得到比Floyd算法更快更好的全局最优解。
The optimal design of the power supply network in the chemical park is an important basis for the project decision-making by the state integrated management department, the power grid project legal person and the financier. Enterprises in the chemical park can easily release flammable and explosive gases in the production activities, and the power supply is in great danger. The main purpose is to use the program optimization method to find an economical and safe layout of the power supply line, and finally get the security decision based on risk and cost. Based on the experience gained by predecessors in optimizing the mathematical model of computer intelligent algorithm, this paper studies the design and application of graph theory algorithm and artificial intelligence algorithm in optimization of power supply network based on risk and cost in chemical industry park by using the graph theory Floyd algorithm and GA genetic algorithm. Design a chemical park, the total substation as the source, the eight optional chemical park with high risk of fire or explosion production enterprise location for the network node. Two models of concentration-free and concentration-bound are established, and the combination of breadth-first traversal and depth-first traversal is used to optimize the model with concentration constraints. Three schemes are given and the optimal results are obtained. For the first time, the optimization of computer intelligent algorithm is used to study the security of power supply network in chemical park. The results show that the application of graph theory and artificial intelligence is feasible in the optimization of power supply network in chemical park. For practical chemical industry power supply layout with more network nodes, GA genetic algorithm can get global optimal solution faster and better than Floyd algorithm.