Suitability Analysis Of The Tropical Fruit Trees Planting Based On GA-BP Neural Network

With the improvement of people's living standards,the demand for fruits is also growing,and the scale of China's tropical fruit industry is also growing.In recent years,extreme weather has often brought disasters to the tropical fruit industry,causing economic losses in the tropical fruit industry.If computer science methods can be used to assess the suitability of tropical fruit trees in a region in advance and rationally allocate the planting range,many unnecessary losses can be avoided.This paper firstly studies the methods for the analysis of the suitability of tropical fruit trees.It is found that the current domestic analysis methods of tropical fruit trees mostly are based on suitable zoning of planting based on mathematical statistics,aiming at the problem of low adaptive ability of traditional mathematical statistics methods.BP neural network was introduced to establish the suitability analysis model of tropical fruit tree planting.Secondly,aiming at the situation that BP neural network often falls into local optimum,the method of increasing network depth is introduced to train and improve the model,and the reliability of the model is verified based on the dataset of the KEEL.Then through analysis of predecessors' research and discussion with experts,12 representative evaluation indicators were selected,use Matlab established multiple hidden layer BP neural networks as The tropical fruit tree planting suitability model,The simulation experiment was carried out based on the meteorological data of 32 stations such as guilin and bobai.The results show that the BP neural network with increased network depth has a significant improvement in classification accuracy compared with the ordinary BP neural network.Finally,in order to improve the shortcomings of BP neural network which is easy to fall into local minimum value,it is found that genetic algorithm has outstanding advantages in global optimization.Therefore,BP neural network improved by genetic algorithm is used to establish a tropical fruit tree planting suitability analysis model.The initial threshold and weight of the BP neural network are optimized by genetic algorithm(GA),and the reliability verification work is performed after using the same KEEL data set.The simulation results of the same meteorological data set are selected.The results show that the BP neural network optimized by genetic algorithm has higher convergence rate and classification accuracy index than the original BP neural network classification.This GA-BP neural network model can be used to analyze the suitability of tropical fruit tree planting...