基于神经网络的锅炉过热器和再热器管壁温度预测研究

STUDY ON PREDICITION OF TUBE-WALL TEMPERATURE IN SUPERHEATER AND REHEATER OF BOILER BASED ON NEURAL NETWORK

提出了采用BP神经网络和RBF神经网络对锅炉过热器和再热器壁温进行预测的方法,经过网络训练和测试,使预测的管壁温度有一定的准确度。RBF神经网络较BP神经网络误差更小,更稳定,更适合于预测锅炉过热器和再热器的管壁温度。

A prediction method for tube- wall temperature in superheater and reheater of boilers, which adopts BP neural network and RBF neural network,has been put forward,through training and testing of the neural networks, making the above - mentioned prediction method to have certain accuracy. Comparatively speaking,the error of RBF neural network is more smaller than that of BP neural net- work,the work of RBF neural network is more stable, and is mroe suitable to predict the tube - wall temperature in superheater and reheater of boilers.