城市生活垃圾热值的特征变量选择方法及预测建模  

Characteristic Variable Selection Method and Predictive Modeling for Municipal Solid Waste Heat Value

作　　者：丁晨曦 严爱军 DING Chenxi;YAN Aijun(College of Artificial Intelligence and Automation,Faculty of Information Technology,Beijing University of Technology,Beijing 100124,China;Engineering Research Center of Digital Community,Ministry of Education,Beijing 100124,China;Beijing Laboratory for Urban Mass Transit,Beijing 100124,China)

机构地区：[1]北京工业大学信息学部人工智能与自动化学院,北京100124 [2]数字社区教育部工程研究中心,北京100124 [3]城市轨道交通北京实验室,北京100124

出　　处：《北京工业大学学报》2021年第8期874-885,共12页Journal of Beijing University of Technology

基　　金：国家自然科学基金资助项目(61873009);北京市自然科学基金资助项目(4192009)。

摘　　要：在垃圾焚烧的过程中,垃圾热值的波动会影响垃圾焚烧的稳定性.为了实现城市生活垃圾热值的实时在线预测以及变化趋势预测,采用模糊神经网络软测量方法,利用焚烧发电厂在线运行数据作为输入,实现垃圾热值的实时预测功能.首先采用互信息方法从若干特征变量中剔除部分无关变量;然后将模糊神经网络和粒子群优化算法结合起来从上述选择出的特征变量中进一步剔除冗余变量,从而确定预测垃圾热值的输入变量,并从中训练出垃圾热值的模糊神经网络预测模型;最后通过采集的样本数据进行性能测试.结果表明该方法有较好的预测准确率和实时性,适用于垃圾热值的在线预测.In the process of municipal solid waste incineration,the fluctuation of waste heat value affects the stability of waste incineration.To make the real-time online prediction and change the trend of the waste heat value,the fuzzy neural network soft sensing method was adopted,and the on-line operation data of the incineration power plant was used as the input to accomplish the real-time prediction function of the waste heat value.First,the mutual information method was used to eliminate irrelevant variables from characteristic variables.Then,the fuzzy neural network and particle swarm optimization algorithm were combined to further eliminate redundant variables from the selected characteristic variables,so as to determine the input variables for predicting the waste heat value,and the fuzzy neural network prediction model for waste heat value was trained.Finally,the performance test was carried out through the collected sample data.Results show that this method has good prediction accuracy and real-time performance,and is suitable for online prediction of waste heat value.

关 键 词：城市生活垃圾 热值 特征选择 互信息 模糊神经网络 实时预测 

分 类 号：U461[机械工程—车辆工程] TP308[交通运输工程—载运工具运用工程]