Comprehensive Performance Detection System For Intelligent Agriculture Soil Media

Wisdom agriculture is an advanced stage of agricultural production.It integrates the Internet of Things & the Internet,and depends on the modules of soil media data collection to realize the intelligent analysis of agriculture.Ultimately,it can provide intelligent decision-making for agricultural production.This paper provides a better solution for the comprehensive performance testing of soil media by improving related algorithms,which focuses on completing the following works:1)The improved Kalman filtering algorithm is based on BP(back propagation)neural network.The Kalman filter is improved by BP neural network algorithm,which solves the disadvantages that it is difficult to apply to nonlinear equations,by applying this algorithm to the real state of the soil media.The data of Gaussian white noise is verified.The results indicate that:(1)the filtered value obtained by the optimized algorithm is approximate to the real state value.(2)The optimized filtering algorithm is superior to the unmodified Kalman filter algorithm in absolute error,root mean square error and RMAD(Root Mean Absolute Deviation).2)The BP neural network algorithm is based on whale hybrid genetic optimization.The hybrid algorithm of whale algorithm will solve the shortcomings of BP neural network with local minimum value and slow convergence speed of the algorithm.The algorithm is applied to the real state data of soil medium for verification.The results show that:(1)the predicted value obtained by the optimized algorithm is close to the actual value.(2)the optimized prediction algorithm is R2(Root squre),MSE(mean-square error),RMSE(Root mean-square error),MAPE(mean absolute percent error)and MAD(Mean Absolute Deviation)are superior to the unoptimized BP neural network algorithm.3)Based on the support vector machine of particle swarm optimization algorithm.The particle swarm optimization algorithm is used to solve the shortcomings of the support vector machine kernel parameters and the disciplinary parameter model selection.The algorithm is applied to the real state data of soil media for verification.The results show that:(1)the classification value obtained by optimized algorithm is basically consistent with the actual value.(2)The optimized classification algorithm is superior to the unoptimized support vector machine algorithm in the classification accuracy.4)By applying the optimized algorithm to the soil medium comprehensive performance detection module to verify the authenticity and effectiveness of the system.