Strength prediction of solid wastes filling body based on GRA--BP neural network

Abstract

In order to solve the problems of ecological environment caused by the shortage of filling materials and solid wastes in the goaf of coal mine, the paper puts forward some suggestions to solve these problems, in the paper, influencing factors of paste filling body strength were analyzed, and sample strength of filling body was determined by orthogonal test, the correlation between the influence factors and the strength of the filling body was determined with grey relational analysis method, and an improved BP strength prediction model was built, in which the factors affecting the strength of solid waste paste filling are input layer and the strength of the filling body is output layer. The strength test data obtained from orthogonal test were used as training samples and test samples of the network, the fitting ability and generalization ability of network data were tested by analog simulation. Test results show that the established prediction model has fast convergence speed and high accuracy, and its prediction accuracy reaches 93.75%, which indicates accurate prediction of the filling body strength can be realized.

Keywords

• solid waste filling body
• orthogonal test
• gra
• bp neural network
• strength prediction
• filling materials
• filling mining