A Research Of Water Temperature Warning System For Aquaculture Based On Improved BP Neural Network Algorithm

As people’s demand for aquatic products continues to increase,the intensive aquaculture industry has achieved vigorous development by virtue of its advanced and efficient advantages.However,the high density of intensive aquaculture can easily lead to deterioration of water quality,and the scientific management of aquaculture water quality has become the main bottleneck restricting its development.Considering that the aquaculture water temperature is the most important water quality parameter that affects the output and quality of aquatic products,and the application of the existing water temperature prediction model in the field of aquaculture is still shallow,farmers mostly manage and control the aquaculture water temperature based on aquaculture experience,and the error rate is high.Low efficiency,this research puts forward a complete solution around pond aquaculture water temperature warning,that is,based on the improved BP neural network algorithm to construct a water temperature prediction model for hairy crab culture,and to develop aquaculture water temperature warning system.The main work of this research includes:(1)Combing the status quo based on literature analysis and integrating the necessary theories: analyzing domestic and foreign water quality prediction and early warning,aquaculture water temperature early warning,and BP neural network algorithm improvement status,analyzing existing deficiencies in the research field and drawing lessons for reference.At the same time,around the aquaculture water temperature measurement and prediction,the theories of water temperature parameter sensing technology,Zigbee technology,genetic algorithm and BP neural network algorithm are introduced to lay the foundation of research theory.(2)Improve the traditional BP neural network algorithm: Considering that the water temperature prediction model based solely on the traditional BP neural network is difficult to meet the high requirements of intensive aquaculture for water temperature prediction,based on the limitations of the traditional BP neural network algorithm,genetic algorithm(GA)is introduced.To optimize it,the GABP algorithm is obtained.The algorithm first assigns the optimal network weights and thresholds obtained by the genetic algorithm to the BP neural network,and then trains the BP network.Through the fitting simulation experiment,it is concluded that the GABP algorithm has a faster convergence speed than the traditional BP algorithm,and has a higher-precision predictive return,and the overall performance of the algorithm is effectively improved.Therefore,it is feasible to apply the improved BP neural network algorithm to the construction of aquaculture water temperature prediction model.(3)Construction of aquaculture water temperature prediction model based on GABP algorithm: Taking hairy crab pond culture as an example,first find out the key factors affecting pond water temperature based on system dynamics theory,and then select the water temperature prediction collected by Changsha Qingyang Lake hairy crab breeding base based on these factors The model has 600 training samples and 120 test samples.The MATLAB2019 a software is used to construct a 5-5-1 structure aquaculture water temperature prediction model according to the input and output of the model,and with the help of empirical trial and error method.The GABP algorithm is used to train the model,and finally a water temperature prediction model based on the five major meteorological data of temperature,solar radiation,rainfall,wind speed and air humidity to output the prediction results of the water temperature of the breeding pond is obtained.Model testing shows that the model prediction process is stable and the prediction results are accurate.(4)Development of aquaculture water temperature early warning system:with the built aquaculture water temperature prediction model as technical support,the development of aquaculture water temperature early warning system was completed in accordance with the software development ideas of system demand analysis,system design,implementation and testing.The hardware of the system is deployed with water temperature sensor module,weather station parameter module,data acquisition node,wireless transmission module,on-site monitoring module,remote monitoring module;and the software has built data management,intelligent prediction,water temperature warning,water temperature control and other functions to support Breeders use the dual-end access of mobile terminals and Web terminals.(5)Application description and prospect exploration of the aquaculture water temperature early warning system: discuss the main functions and applications of the aquaculture water temperature early warning system to fully reflect the availability and practicality of the system in the aquaculture water temperature intelligent early warning,automatic control and statistical analysis.At the same time,combined with the comprehensive requirements of aquaculture water quality control and the current system’s shortcomings in cross-regional management,the system has been integrated with other key aquaculture water quality factors for early warning and the introduction of GIS technology in the two dimensions of the future feasible application prospects or development directions of the system.Preliminary exploration.The research in this article is of practical significance for improving the scientificity and effectiveness of aquaculture water temperature management and control,promoting the long-term development of my country’s intensive aquaculture industry,and helping my country’s aquaculture intelligence and information process.