BP-PSO在电加热炉中的温度智能预测

Intelligent Temperature Prediction Based on BP-PSO for Electric Furnace

针对电加热炉难以建立精确模型的问题,提出采用BP神经网络与粒子群优化(PSO)相结合的算法对电加热炉的温度变化进行辨识,并建立系统模型。在建立系统模型的基础上,对温度变化趋势进行了预测。试验结果显示,与BP神经网络算法相比,粒子群优化BP神经网络算法所得到的预测值有效时间范围延长了60%;在相同有效的预测时间内,预测值精度提高了43%。

Aiming at the features of electric furnace, e. g. , multiple variables, non-uniform distribution, and slow real time performance, that bring difficulty for establishing accurate model, the algorithm that combining BP neural network and particle swarm optimization （ PSO ） is proposed to establish the system model through recognizing the temperature variation of electric furnace, and to implement prediction of temperature varying trend based on the model. The experimental results show that comparing the BP neural network algorithm, through PSO and BP neural network algorithm, the effective time range of the predictive value is extended by 60%, and the accuracy of the predictive value is enhanced by 44% within the same effective prediction time.