Research On Neural Network Quality Prediction Model Based On Genetic Algorithm

Abstract

In industrial production process, it is difficult to predict product quality in advance. Traditional prediction methods are mostly based on complex mechanism models, and the prediction accuracy is not high. This paper uses the historical data of industrial production to forecast, constructs the quality prediction model of neural network, and uses genetic algorithm to optimize the network parameters, so as to avoid the neural network falling into local optimum. This paper takes the data of hot rolling production line as an example and adopts the method to predict the quality index of steel plate. The simulation result shows that the quality prediction model of neural network based on genetic algorithm has better prediction ability. This method has important theoretical value and practical significance for production workers to improve product quality.