The Study Of The Yield Estimation Method About Processing Tomatoes Based On The Growth And Development Model

Crop models are foundation and premises to achieve crop planning, crop breeding,cultivation and management of modern agricultural technology. Crop growth model, yield formation model and fertilization models are the most important crop models in the planting industry of processing tomato. Through the study of the growth model and crop physiology to achieve the prediction of growth stages and production has significance for processing tomato planting management and production planning.This paper mainly according to the planting environment and climate characteristics in Xinjiang, collecting extensive production management data and combining field experiments,using mathematical modeling techniques and control algorithms to construct the growth models, dry matter accumulation and yield formation model and fertilization model.Achieving yield prediction and field management, provide technical guidance to crop breeding and planting management at the same time. Besides, using wavelet neural to build processing tomato production prediction model under construction situations, achieving macro forecast of the production to provide basis for decisions for the formulation of industrial chain and construction planing.The results show that, the prediction of the growth period of processing tomato is similar with the actual situation, the difference between the prediction value of yield forecast model and the actual production is small, showing the prediction results are better and it can provide decision support for management. Using statistical test to the fertilization model,indicates that the return relationship between the production of processing tomatoes with inputs of three kinds of fertilizer is significant, so this model can provide scientific guidance for rational application of fertilizer. Processing tomato production prediction model based on wavelet neural network prediction results are better than BP neural network, indicates that it has overcame the shortages of the model and improved the model further combing theadvantages of wavelet analysis and BP neural network,therefore the model can provide basis for the processing planing of the tomato products.