基于改进神经网络的垃圾焚烧锅炉设备床温预测方法  

A Bed Temperature Prediction Method for Waste Incineration Boiler Equipment Based on Improved Neural Network

作　　者：景玉博 JING Yubo(Zhengzhou Xingze Environmental Protection Energy Co.Ltd.,Zhengzhou 450100,China)

机构地区：[1]郑州荥泽环保能源有限公司,河南郑州450100

出　　处：《工业加热》2025年第3期35-40,共6页Industrial Heating

摘　　要：垃圾成分的不确定性给垃圾焚烧锅炉设备床温预测带来了极大的影响,一旦预测不准确,会导致垃圾处理不当,对生态环境造成破坏。由于垃圾焚烧锅炉的燃烧过程受到多种因素的影响,因素的变化会导致燃烧过程的不稳定,从而影响床温的预测,因此研究一种基于改进神经网络的垃圾焚烧锅炉设备床温预测方法。该方法利用相关系数法筛选出对垃圾焚烧锅炉设备床温产生影响的主要因素,以此作为神经网络的输入,采集床温数据,以此作为神经网络的输出,以有效避免燃烧过程受到多种因素的影响,确保对动态变化的处理能力。然后通过输入与输出对神经网络进行训练,以并采用水波优化算法来寻优神经网络参数,实现神经网络的改进,最终利用改进后的神经网络实现床温预测。结果表明:所研究方法应用下,可有效实现对测试样本的床温预测,其误差在±5℃以下,具有较强的稳定性,且其熵值结果在0.9以上,适应能力较强。The uncertainty of garbage composition has a great impact on the bed temperature prediction of garbage incineration boiler equipment.Once the prediction is inaccurate,it can lead to improper garbage treatment and damage to the ecological environment.Due to the influence of various factors on the combustion process of garbage incineration boilers,changes in these factors can lead to instability in the combustion process,thereby affecting the prediction of bed temperature.Therefore,a bed temperature prediction method for garbage incineration boiler equipment based on an improved neural network is studied.This method uses the correlation coefficient method to screen out the main factors that affect the bed temperature of waste incineration boiler equipment.It used as the input of the neural network to collect bed temperature data,which is used as the output of the neural network to effectively avoid the influence of multiple factors on the combustion process and ensure the ability to handle dynamic changes.Then,the neural network is trained through input and output,and the water wave optimization algorithm is used to optimize the neural network parameters,achieving improvement of the neural network.Finally,the improved neural network is used to achieve bed temperature prediction.The results show that under the application of the research method,the bed temperature prediction of the test samples can be effectively achieved,with an error below+5℃,strong stability,and an entropy value result above 0.9,indicating strong adaptability.

关 键 词：改进神经网络 水波算法 垃圾焚烧锅炉设备 床温影响因素 床温预测方法 

分 类 号：TP41.62[自动化与计算机技术]