DBN-GA model-based thermal power prediction method for solar heat collection system

Abstract

The invention discloses a DBN-GA model-based thermal power prediction method for a solar heat collection system, and the method comprises the steps: 1) employing a solar radiation online monitoring system to obtain solar radiation data, recording real-time meteorological measurement data, and obtaining historical thermal power data through calculation; 2) dividing the data in the step 1) into twotypes according to the acquired characteristic information parameters, namely a training set and a test set; 3) establishing a thermal power prediction model based on a DBN-GA algorithm, determining arestricted Boltzmann machine model, and inputting the training set characteristic parameter sample obtained in the step 2) into the prediction model for training learning to obtain an output result,namely the thermal power of the solar heat collection system; and 4) inputting the test set sample into the trained thermal power prediction model, and completing thermal power prediction of the solarheat collection system by the thermal power prediction model. The method provided by the invention can predict the thermal power of the solar heat collection system more accurately and effectively.

Description

technical field [0001] The invention belongs to the technical field of thermal power prediction of a solar heat collection system, and relates to a method for predicting thermal power of a solar heat collection system based on a DBN-GA model. Background technique [0002] With the emergence of global problems such as "energy crisis" and "environmental pollution", solar energy has been used more and more widely as a clean energy source. Today, the solar thermal collection system is advocating the construction of a green and harmonious earth. It has the advantages of energy saving, environmental protection and practicality. The thermal power of the solar thermal collection system is one of the very important factors in the solar thermal collection system, which is directly related to the performance of the entire heating system. Safe and stable operation, resource utilization optimization, and benefit maximization. Real-time prediction of thermal power is of great significanc...

AI Technical Summary

Problems solved by technology

The BP neural network can establish a feed-forward prediction model for this problem, but the BP neural network has the phenomenon of "overfitting". Due to the learning of too many sample details, the learned model can no longer reflect the information contained in the sample. law

The radial basis neural network provides a better structural system for the prediction of thermal power, but it has the disadvantage that it cannot explain its own reasoning process and reasoning basis, and the neural network cannot work normally when the data is insufficient.

Images