Coal-fired unit high-temperature superheater wall temperature prediction neural network model

Abstract

The invention discloses a coal-fired unit high-temperature superheater wall temperature prediction neural network model which is formed by successive connection of a plurality of neural networks, andeach neural network is composed of an input layer, a hidden layer and an output layer. the input layer is divided into a predicted heating surface wall temperature variable input, an upstream heatingsurface wall temperature variable input and other key variable inputs, key factors influencing the wall temperature and the upstream heating surface wall temperature change condition are considered, and meanwhile the influence of predicted heating surface wall temperature historical data on the input layer is considered; an input variable structure is determined, an input parameter delay coefficient, the number of hidden layers and an activation function are corrected, the model training and generalization precision is improved, the change trend of the predicted wall temperature at different moments is obtained through successive wall temperature prediction, and the better high-temperature superheater wall temperature prediction precision is achieved.

Description

technical field [0001] The invention relates to the technical field of modeling the wall temperature characteristics of a heating surface of a coal-fired unit, in particular to a neural network model for predicting the wall temperature of a high-temperature superheater of a coal-fired unit. Background technique [0002] With the continuous improvement of coal-fired units, increasing steam temperature, pressure and other power generation parameters is an important way to improve the efficiency of ultra-supercritical units, but the increase in steam temperature puts forward higher requirements for steam pipe materials and wall temperature control. Restricted by the creep strength and durability of the material, the temperature fluctuation must be within the safety margin. Due to the deviation of the wall temperature measurement of the heating surface, the parameters cannot be adjusted in time, and the long-term over-temperature operation of the heating surface will inevitably i...

AI Technical Summary

Problems solved by technology

Among them, the wall temperature is calculated through the mechanism modeling analysis of water-cooled walls, superheaters and other components. This method is more complicated, and there are many boundary parameters. The actual measurement points of the power plant cannot give all the boundary parameters, and the model needs to be continuously revised under different conditions. Therefore, It does not meet the requirements of online calculation, and cannot participate in the closed-loop control of the power station wall temperature in real time; based on the mathematical modeling analysis method, the wall temperature prediction method based on the artificial neural network is mostly used at present, only considering the influence of external factors on the wall temperature, using BP Prediction of Boiler Tube Wall Temperature Using Static Network Structures such as Neural Network

The historical data of the current wall temperature, the historical data of the upstream wall temperature, and the rate of change of related factors are not considered, and the time series prediction neural network structure, neural network activation function, etc. are not studied. The prediction structure is relatively backward and the calculation results are poor.

[0006] To sum up, the existing countermeasures and prediction methods for wall temperature overheating only stop at displaying alarms, so that parameters can be changed based on the experience of operating personnel, and closed-loop control has not been realized.

Images