基于VEGA和DCGA的化工园区布局优化研究

Study on layout optimization for chemical industry park based on VEGA and DCGA

化工园区布局优化是一个复杂的多目标优化问题,建立了以事故风险和经济收益为两个优化指标的化工园区布局优化模型,以潜在死亡人数(PLL)表征事故风险,以销售收入表征经济收益;从算法的正确性、有效性及其搜索性能上提出了三条化工园区布局优化算法性能的评价准则;基于向量评价遗传算法(VEGA)和基于确定性拥挤机制的小生境遗传算法(DCGA)分别对化工园区布局优化进行研究,并对比分析了优化计算结果,得出:建立的化工园区双目标布局优化模型和提出的优化方法是可行的,计算所得的Pareto最优解对化工园区布局具有较好的参考价值;VEGA算法仅能满足部分准则要求,而DCGA算法能同时满足三条评价准则要求,可知DCGA算法适用于化工园区布局优化。

The layout optimization for chemical industry park is a complex multiobjective optimization problem. A layout optimization model for chemical industry park that accident risk and economic benefit were taken as two optimized targets was established, the potential accident loss of life （PLL） and sales revenue were respectively taken as characterization of accident risk and economic benefit. From the correctness, effectiveness and search performance of the algorithm, three performance evaluation criterias of optimization algorithm for the chemical industry park lay out were put forward. Research on layout optimization for the chemical industry park was made respectively accord ing to vector evaluated genetic algorithm （VEGA） and niche genetic algorithm based on the deterministic crowding mechanism （DCGA）, and optimization calculation results were comparatively analyzed. The conclusions were as follows ： the established twoobjective layout optimization model and proposed optimization methods for the chemical industry park were feasible, the calculated Pareto optimal solutions had a good reference value for the chemical in dustry park layout. VEGA could only meet some criteria requirements, but DCGA could simultaneously meet thethree evaluation criterias, which showed that DCGA is suitable for the chemical industry park layout optimization.