用遗传算法的兼容度极大化模型选优水污染治理方案

Choosing optimal plan of water pollution treatment with compatibilitydegree maximum model based on RAGA

采用不同评价方法对水污染治理的16种初选方案进行排序选优,得到不同评价结果,为克服各单个评价方案的片面性,充分挖掘各评价方案包含的信息,提出了基于实数编码加速遗传算法(RAGA)的兼容度极大化模型(RAGA-CDMM)对各单个评价方案进行组合,得到一个最优评价方案。结果表明:RAGA-CDMM得到的最优评价方案,结果更为合理可靠,为决策者进行"多中选好"提供了科学的依据;利用RAGA-CDMM可直接根据各单个评价方案的评价结果数据来确定各评价方案的权重,避免了主观赋权的不足;RAGA-CDMM简便可行,适用性和可操作性强,具有推广应用价值。

Choosing the optimal plan from 16 initial plans using different methods, in order to overcome unilateralism of each single evaluation method and to mine sufficiently the evaluation information of each single evaluation method, a real coding accelerating genetic (RAGA) based compatibility degree maximum model (RAGA-CDMM) is presented to combine each single evaluation plan, then a optimal evaluation plan is get. The result shows that the optimal plan get by using RAGA is more rational and reliable, provide scientific basis for makers to choose the best one from many results. RAGA makes the determine weights of evaluation methods in objective function and avoid the shortcoming of subjective weighting method. RAGA-CDMM simple, feasible, and general, so it can be applied to different multiple attribute decision making problems.