基于SLPSO-WNN的铸造熔炼工序能耗预测模型

Prediction of Energy Consumption in Casting Smelting Process Based on SLPSO Initialized Wavelet Neural Network

为了实现铸造熔炼工序的能耗预测,采用小波神经网络(wavelet neural network,WNN)建立能耗预测模型。针对WNN模型初始化中存在的问题,提出一种基于SLPSO的WNN模型初始化方法。对基于随机法、RBS和SLPSO初始化的WNN模型训练效果进行对比测试,结果显示SLPSO-WNN模型具有较低的平均相对误差;采用企业实际数据进行了基于SLPSO-WNN模型的能耗预测,预测平均相对误差为2.48%:验证了模型的有效性。

In order to predict the energy consumption of casting melting procedure, an energy consumption prediction model based on wavelet neural network( WNN) is built. Aiming at the problems existing in the initialization process of WNN model,an initialization method based on SLPSO is proposed. The training results of WNN model based on random method,RBS and SLPSO are compared,the result indicate the SLPSO-WNN model has lower mean relative error( MRE). The energy consumption prediction based on SLPSO-WNN model is carried out based on the actual data of enterprise and the MRE is 2.48% which proves the validity of the model.