Prediction Method of Urine Component Concentration Based on BP Neural Network Optimized by Whale Algorithm
Aiming at the low accuracy of real-time analysis of urine component concentration,a prediction method of color sensor combined with WOA algorithm to optimize BP neural network is proposed.The RGB values of each reagent block on the urine test strip were collected by the color sensor,and the data were calibrated by the principle of white balance.The relationship between reflected light and concentration is constructed by Kubelka-Munk and Beer-Lambert laws.The objective function is minimized by parameter optimization model to eliminate system error.Based on the least square method,the mathematical model of color value and concentration of color label card is established.WOA algorithm is used to optimize the weights and thresholds of neural network,and a large number of data are used to train BP neural network for regression analysis of color value and concentration.The experimental results show that the MAE between the predicted value and the real value is 3.141 5,RMSE is 4.328,and R2 is close to 1.The WOA-BP neural network model has high precision and accuracy in predicting the concentration of urine components.
万方数据
维普
