居民生活垃圾可燃成分的热值与RDF热值关联模拟预测研究

发布时间：2018-05-02 13:10

本文选题：居民生活垃圾 + 热值　；参考：《西南交通大学》2015年硕士论文

【摘要】：居民生活垃圾可燃成分具有较高的热值,若能充分回收利用将其能源化,对能源危的可持续利用具有非常重大的意义。.本文选择人工神经网络进行数据挖掘,以在居民小区采取的生活垃圾所测定的试验数据为基础,选择PE/PP、纸类、橡胶、含水率以及干基氢各含量作为模型的输入参数,居民生活垃圾可燃成分的低位热值为输出参数,并分别基于BP神经网络、RBF神经网络和自适应神经模糊推理系统(ANFIS),对低位热值建立了预测模型。BP神经网络模型预测准确率为93.36%,RBF神经网络模型预测准确率为96.87%,自适应神经模糊推理系统(ANFIS)的模型预测准确为91.06%。对比三种模型的预测结果可知：BP、RBF神经网络和自适应神经模糊推理系统(ANFIS)模型均可用于可燃成分热值的预测。ANFIS模型具有极高的模型拟合效果,但模型的验证误差相对较大,使该模型平均预测准确率较低。BP模型的预测准确率相对较高,能较好的对可燃成分热值进行预测,但其效果要低于RBF模型。RBF模型在BP模型的基础上加入了线性控制,使模型的预测准确率有了较大提高。由此证明：RBF模型更适用于居民生活垃圾可燃成分的低位热值预测,能获得令人更为满意的结果。当PE/PP、纸类、橡胶、含水率以及干基氢各含量分别为31.66%、59.94%、0.03%、31.37%、8.74%时,经RBF模型预测后得到的低位热值为16352.61kJ/kg,其处于RDF物料所要求热值范围,故用居民生活垃圾中的可燃成分生产RDF具有一定的理论可行性。
[Abstract]:The combustible composition of domestic refuse has a high calorific value, if it can be fully recycled and utilized, it is of great significance to the sustainable utilization of energy. In this paper, artificial neural network is selected for data mining. Based on the experimental data measured by domestic waste in residential areas, PE-P, paper, rubber, moisture content and dry base hydrogen content are selected as input parameters of the model. The low calorific value of combustible components of resident domestic waste is the output parameter. Based on the BP neural network RBF neural network and the adaptive neurofuzzy inference system, the prediction model of low calorific value is established. The prediction accuracy of BP neural network model is 93.3636. The prediction accuracy of RBF neural network model is 96.87, and the prediction accuracy of BP neural network model is 96.87 and adaptive. The model prediction of neurofuzzy inference system (ANFIS) is 91.06. Compared with the prediction results of the three models, it can be seen that the two models can be used to predict the calorific value of combustible components. ANFIS model has a very high fitting effect, but the verification error of the model is relatively large. The prediction accuracy of the model is lower than that of the BP model, and the prediction accuracy of the model is relatively high, which can better predict the calorific value of combustible composition, but its effect is lower than that of the RBF model. The BP model has linear control on the basis of the BP model. The prediction accuracy of the model has been greatly improved. It is proved that the 1: RBF model is more suitable for the prediction of low calorific value of combustible components of household solid waste and can obtain more satisfactory results. When P / P, paper, rubber, moisture content and dry base hydrogen content were 31.660.93 and 8.74, respectively, the low calorific value predicted by RBF model was 16352.61 kJ / kg, which was within the required calorific value range of RDF material. Therefore, it is theoretically feasible to produce RDF from combustible components in household garbage.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：X799.3;TP183

【参考文献】