活性污泥工艺的多目标优化分析

Analysis of activated sludge process by multi-objective optimization

研究了多目标优化方法在活性污泥工艺优化中应用的可行性,以国际水协(IWA)发布的基准仿真模型BSM1的工艺流程为对象,以出水水质、过程能耗和反应池体积的指标为优化目标,以反应池池容(5个)、混合液回流量、污泥回流量、污泥排放量、反应池氧气传质系数(3个)等11个变量为决策变量,构建了基于BSM1的多目标优化模型,并利用MatLab非支配排序遗传算法NSGA-II进行求解,获得了BSM1多目标优化模型的Pareto解.研究结果表明,在Dry进水条件下,通过将体积指标增大4.1%,可以明显改善出水水质指标(降低81.5%),能耗指标也能得到一定改善(降低13.3%);在3种进水条件(Dry,Rain,Storm)下,相对于BSM1工艺参数的缺省值,利用本研究的优化解作为工艺参数,均能不同程度改善出水水质、降低能耗.研究结果显示,在活性污泥过程不同目标之间进行权衡可以通过多目标优化方法来实现.

The feasibility of applying the method of multi-objective optimization in analysis of the activated sludge process was studied. A benchmark process in BSM1 issued by IWA was taken as a target process. Three indexes of effluent quality, energy consumption, and total volume of reactors were the objectives. The volumes （ for each of 5 reactors）, mixed liquor return rate, sludge return rate, excess sludge wasting rate, oxygen transfer coefficients （ in each of 3 reactors） were the decision variables. The model of multi-objective optimization was solved by the non-dominated sorting genetic algorithm Ⅱ（NSGA-Ⅱ） in MatLab, and the Pareto solutions were obtained. Resuhs show that under the condition of Dry weather influent, increasing the volume index by 4.1% may enhance the effluent quality index by 81.5% and the energy consumption index by 13.3%. Under all influent conditions （ Dry,Rain, Storm）, taking Pareto solutions as process parameters, all performance indexes are better than with the default process parameters of BSM1. The results indicate that the trade-offs among different objectives of the activated sludge process may be achieved by using the method of the multi-objective optimization.）