Based BP Neural Network Prediction Of Rice Pests

The main pests do serious harm to crops perennial and damage the economy.To give the pests effective governance, an dynamic way of preventing pests israised. The key to prevent pests is that the preliminary forecasting work about pestscan be carried out smoothly. To make the right and reasonable plans to prevent andcontrol pests, we should take the precise and seasonable forecasting scheme.So wecan adopt the right measures to reduce the number of pests and the pest damage,ensuring the crops harvest.The forecast quantity for the pests decides the area the frequency and thescope of crops we take preventive measures. But the whole progress of the researchis behind the prediction of the occurrence time. It is because that there are too manyeffective factors, which are uncertain. So we propose a there-layer BP neuralnetwork that can better depict the model’s feature of complex nonlinear, manyinput-output and indefinite.Artificial neural network is a class of simulated human nervous systemstructure, hereveals the data sample contains non-linear relationship, a lot ofnon-linear adaptive processing unit dynamic system, with good adaptability,self-organization and strong Learning, associative, fault tolerance and anti-jammingcapability. BP network in particular is widely used in recent years, patternrecognition, prediction assessment areas, and achieved good results. BP network tothe current error back propagation training neural network learning algorithm, thealgorithm is based on the network error function gradient descent.The MATLAB software which has ready-made network toolbox is the bestcondition to develop and use neural network. An available Neural Toolbox inMATLAB. However, solves the problem, we will introduce the ways of networkmodel training, network model building and network model stimulating based onthe toolbox of The MATLAB neural network much more.