Abstract: The scientific prediction of municipal solid waste is the basis of the government department of sanitation safe assessment, planning and decision-making. This paper takes Xi’an city, the biggest central city in Northwest China, as an example. By qualitative and quantitative analysis of the factors which affect the waste output, three significant factors which include city’s built-up area, urban per capita disposable income and the urban population are used as predictive indexes. The GM（1，1） grey forecasting model and the polynomial fitting model based on least square method forecast the three indexes, and then a multivariate linear regression model is set up to forecast the emission of MSW of Xi’an 2010 to 2020 on the basis of the early prediction. Significant F test indicates that multiple integrated prediction model is reasonable, and prediction results are meaningful.

Key words: municipal solid waste, GM（1, 1） grey forecasting model, Multivariate Linear Regression（MLR） model

摘要： 大城市生活垃圾预测是政府环卫部门安全评价、规划和决策的基础。以西北最大城市西安为例，在分析现状基础上得出目前影响西安市生活垃圾产量的三个显著的驱动因子：西安市辖区建成区土地面积、城镇居民人均可支配收入和市区户籍人口。先建立GM（1，1）灰色预测模型和最小二乘法多项式拟合人均可支配收入增长模型，对三个影响因素指标分别进行预测；后建立多元线性回归模型，结合前期三个指标预测值，预测2011—2020年西安市生活垃圾产生量。显著性F检验表明多模型拟合预测合理，预测结果有意义。

关键词: 城市固体垃圾, GM（1, 1）灰度预测模型, 多元线性回归模型