基于NSGA-Ⅱ的工业区急救点优化布局研究

Research on the optimization of the emergency point layout for industrial zone base on NSGA-Ⅱ

针对工业区急救点布局存在的问题,综合考虑多种影响因素,建立多目标多约束数学模型。将成本、急救时间和急救点数量作为3个优化目标,并根据实际情况构建相应的约束条件。运用NSGA-Ⅱ(带精英策略的快速非支配排序遗传算法)解决多目标优化问题,最终得到一系列Pareto最优解,其解空间分布均匀,并且具有良好的收敛性和鲁棒性。文章结合某工业区急救点优化问题对算法进行验证,结果与实际情况匹配良好,此研究具有重要的现实意义。

This paper takes it as its main objectives to fully consider the various influential factors of the given area in the locality optimization in establishing an emergency point layout, which is expected to be popular for industrial authorities. Actually, it is both realistic and desirable to make such a kind of emergency point layout optimization for industrial zones. For the given purpose, we have put forward a multiple-objective and multiple-constraint mathematical model by taking into full account all the influential factors. Ⅲ doing so, we have converted the traditional single-objective optimization of emergency point layout into a multiple-objective one. In our proposed model, cost, emergency time and number of emergeney points can be taken as the optimal targets and given constraints simultaneously. The problem can thus be solved by NSGA-Ⅱ, a new multiple-objective generic algorithm for the purpose. By means of the algorithm, it is possible to produce a series of Pareto-optimal solutions as a result of filtering out so-called dominant solutions to restrict the number of expected solution candidates to the minimum. As one of the advan- tages, the given solution sets enjoy a uniform distribution as well as nice property of convergence and robustness. The final satisfactory. solution can be gained by the decision-maker in accordance with the system demands, which can provide an effective tool for measuring the performance of different objective functions. In addition, alternative techniques both from the GA field and from other meta-heuristic areas have been investigated and compared to NSGA-Ⅱ. The model can thus be used for practical application, enabling the decision-makers to verify their decisions by comparing it with the output of the model. Combining the proposed model with the emergency point layout optimization of an industrial zone with six operation zones, this paper has succeeded in validating the algorithm, whose results agree well with the actual situation. Thus it can be seen that the model is appropriate as well as available for multiple-objective emergency point optimization in industrial zone planning.