基于组分比例的垃圾热解神经网络预测模型.pdfVIP

第 25 卷 第 5 期 中 国 电 机 工 程 学 报 Vol.25 No.5 Mar. 2005 2005 年 3 月 Proceedings of the CSEE ©2005 Chin.Soc.for Elec.Eng. 0258-8013 (2005) 05-0154-05 X705 A 610 30 文章编号： 中图分类号： 文献标识码： 学科分类号： ⋅ 垃圾热解实验研究及其神经网络预测模型 温俊明，池 涌，金余其，严建华，倪明江，岑可法 (浙江大学热能工程研究所，能源洁净利用与环境工程教育部重点实验室，浙江省 杭州市 310027) EXPERIMENTAL STUDY ON MSW PYROLYSIS AND ITS NEURAL NETWORKS PREDICTION MODEL WEN Jun-ming, CHI Yong, JIN Yu-qi, YAN Jian-hua, NI Ming-jiang, CEN Ke-fa (Clean Energy Environment Engineering Key Laboratory of MOE, Institute for Thermal Power Engineering of Zhejiang University, Hangzhou 310027, Zhejiang Province, China) ABSTRCT: A thermogravimetry analysis system was designed, 义。然而由于垃圾的成分比较复杂，各组分的热解 in which more massive sample (about 10g) can be analyzed. 特性有很大差别，所以即使得到垃圾的工业及元素 Experimental studies were performed on the mixed pyrolysis 分析也很难指导焚烧炉的设计和运行。我国的情况 characteristics of typical organic components of MSW. Based 是，一般各地环保部门都有当地垃圾组成的统计数 on these experimental results, a prediction model was established for MSW pyrolysis by using the principle of 据，因此，提出了利用环保部门提供的垃圾组分比 chemical reaction kinetics and the theory of BP neural networks, 例预测垃圾热解特性的新思路。 which consists of a pyrolysis kinetics submodel and seven BP 以前一些学者对垃圾的热解特性研究都是在热 neural networks submodels. This model is able to predict the 天平上进行，一般物料的量仅为几毫克[6-9]，虽然有 overall pyrolysis characteristics of MSW by inputting the 学者研究了部分垃圾组分的混合热解特性[8-9]，但由 component proportions of MSW. Accordingly, it is able to guide the design and operation of MSW incinerator. 于数量