With the improvement of living standards in rural areas,the composition of pollutants contained in domestic sewage tends to be complicated,and the impact of untreated on-site discharge on the ecological environment cannot be underestimated.At present,there are various types of rural domestic sewage treatment modes,but the decision-making methods lack scientific basis.This paper taked 82 villages in Inner Mongolia as the research object.Taking the natural village as the smallest research unit,the residential density decision tree(accuracy:76％)and the terrain decision tree(accuracy:88％)were constructed by using the proportion of village homestead,the ratio of village length to width,perimeter,limit distance,the proportion of slope 0°～8 ° and the standard deviation of slope.The results of the decision tree were analyzed by the binary classification probability unit regression(Probit)analysis(accuracy rate:87％)to determine the concentration or dispersion of the villages,and then select the appropriate village sewage treatment mode(centralized mode,receiving mode,decentralized mode,pulling mode,local utilization mode).The establishment of the decision-making method of rural domestic sewage treatment model has constructed and quantified the decision-making indicators,which is of positive significance for promoting the development of rural domestic sewage treatment system and improving the effectiveness of sewage treatment.
village dispersionrural domestic sewagemachine learningdecision treegovernance model
国家科技期刊平台
NSTL
万方数据
