污泥干燥特性实测及预测模型研究  

Study on the M Teasurement and Prediction Model of Sludge Drying Characteristics

作　　者：王振宇 王强 刘东[2] 王令[2] 陈永灿[1] WANG Zhenyu;WANG Qiang;LIU Dong;WANG Ling;CHEN Yongcan(School of Environment and Resource,Southwest University of Science and Technology,Mianyang 621010,Sichuan,China;School of Civil Engineering and Architecture,Southwest University of Science and Technology,Mianyang 621010,Sichuan,China)

机构地区：[1]西南科技大学环境与资源学院,四川绵阳621010 [2]西南科技大学土木工程与建筑学院,四川绵阳621010

出　　处：《西南科技大学学报》2024年第1期66-74,110,共10页Journal of Southwest University of Science and Technology

基　　金：国家自然科学基金面上项目(5207082284)。

摘　　要：污泥脱水干化是污泥资源化利用的重要环节,目前缺乏通识性的污泥干燥模型。实验分析了干燥温度、相对湿度、污泥厚度、干燥时间等因素对污泥干燥过程的影响,比较了常用的5种薄层干燥模型对污泥干燥过程的拟合效果,建立了一种BP神经网络污泥干燥预测模型,并与传统的拟合效果较优的Midilli模型进行了预测精度比较。结果表明:当污泥低温干燥时,温度、相对湿度对污泥干燥有显著影响,相对湿度越高、温度越低,污泥干燥速率越慢;Midilli模型决定系数高、卡方系数和均方根误差均较小,是5种常用薄层干燥模型中拟合效果最好的模型,其与实验结果误差在15%以内;BP神经网络污泥干燥预测模型能很好预测污泥的干燥过程,预测结果与实验测试结果误差在5%以内,具有比Midilli模型更高的预测精度。BP神经网络污泥干燥预测模型为污泥干燥过程模拟提供了一种新的方法。Sludge dewatering and drying are crucial steps in the resource utilization of sludge.There is currently a lack of a comprehensive drying model of sludge.The effects of drying temperature,relative humidity,sludge thickness,and drying time on drying sludge were analyzed in the experiment.The fitting effects of five commonly used thin-layer drying models on the sludge drying process were compared.A BP neural network model was established for predicting sludge drying.This was compared with the Midilli model,which has traditionally shown better fitting results for prediction accuracy.The results indicate that the drying of sludge is significantly affected by temperature and relative humidity when it is dried at low temperatures.The higher the relative humidity and the lower the temperature,the slower the rate at which the sludge dries.The Midilli model has a high coefficient of determination,and its chi-square and RMSE values are relatively low.It is the best-fitting model among the five commonly used thin-layer drying mod-els.The error compared to the experimental results is within 15%.The BP neural network sludge drying prediction model can predict the sludge drying process very well.The prediction results have less than 5%errors compared to the experimental results.The model has a higher predictive accuracy than that of the Midilli model.The BP neural network model for sludge drying prediction provides a new method to simulate the sludge drying process.

关 键 词：污泥 薄层干燥模型 BP神经网络污泥干燥预测模型 

分 类 号：X703[环境科学与工程—环境工程]