Study On Intelligent Decision Model Of Tilapia Pond Culture

With the rapid development of aquaculture, informational aquaculture is getting into the process of aquaculture, by the using of water quality sensors, the promotion of internet of things, the farmer can find out the aquaculture process data in a variety of environmental factors. However, because the lower level aquaculture management, the imperfect information technology and other reasons, we can not be timely and accurately on the aquaculture process management and control intelligently, Therefore it is reason of the research of tilapia intelligent decision models, to explore the relationship between the growth.diseases of tilapia with a variety of environmental factors and other environmental factors. We can provide intelligent decision for the future of aquaculture precision farming, but also provide a reference for other aquaculture animals intelligent farming.The intelligence neural network is an important branch of artificial intelligence, it is an important way. In recent years, it is frequently applied to the prediction of nonlinear system. It is a network which is connected by artificial neurons, it simply describes the human brain by a microscopic a point of view, it realizes the human brain’s learning, memory and other functions.In this thesis, we build the tilapia growth model, predictive model of disease, pond dissolved oxygen prediction model by using the technology of neural network. The fish pond tilapia growth models, predictive models of disease is built by BP neural network to explore the key technologies to establishment BP neural network model, including the selection of the samples and pre-treatment, the selection of the input and output variables, the number of nodes in the hidden layer and so on. Tilapia pond growth model discussed the relationship among body weight weekly during a tilapia growth period and dissolved oxygen, water temperature, air temperature, pH. The predictive models of disease explores the relationship among aquaculture diseases and environmental measured factors. The least squares parameter optimization vector machine is used on the tilapia ponds DO prediction model to process data with radial basis function as a core function, we study the relationship between dissolved oxygen and other pond biological and environmental factors. The fittable error in three models are derived through research within the allowable range, the neural network model constructed in this paper is precise and accurate, it proves that the neural network is feasible and applicable in the field of aquaculture.