Research On Control Strategy And System Of Water And Fertilizer Integrated Machine In Greenhouse Based On Multivariate Data Fusion

Agriculture is China's basic industry,which is related to national economic development and people's living standards.In recent years,The state attaches importance to the development of rural agriculture informatization.Facility agriculture can adjust the industrial structure of agriculture,change the traditional planting structure,and improve environmental pollution caused by the abuse of pesticides and fertilizers.At the same time,China's agricultural water consumption and fertilizer application are huge,and the waste of resources is serious.Although the country has made certain efforts to regulate and control water resources,pesticides,fertilizers,etc.,the pressure on resources and the environment is still great.Therefore,the development direction of modern facility agriculture requires increase the level of intelligence in facility agriculture,reasonable management of water and fertilizer resources,and the agricultural management method should be changed from the extensive type of large water and fertilizer to the intensive type of fine control.In this context,based on integrated water and fertilizer irrigation technology and Internet technology,this paper focuses on the following aspects:(1)To study the relationship of multi-sensor data processing in the greenhouse,according to the environmental characteristics of the greenhouse,an adaptive weighted fusion method based on t criterion is proposed to process the environmental data collected in greenhouses,and the fusion results are more accurate.On the basis of data fusion,the heterogeneous sensor data is analyzed and synthesized,and a device control scheme that is convenient for the unified control of intelligent equipment in the greenhouse is proposed,which improves the control efficiency.(2)According to the water and fertilizer mixing process of the intelligent water and fertilizer integrated machine in the greenhouse,a process control model of soil conductivity and soil acidity and alkalinity concentration was constructed.According to the model analysis,it can be known that there are nonlinearity,time lag and time variability in the process of water and fertilizer mixing.Combining traditional PID control,fuzzy control theory and BP neural network to build a BP neural network fuzzy PID controller.The controller uses neural network to adjust fuzzy control rules,and online self-tuning output variables(35)K _p?(35)K _i?(35)K_d,thereby adjusting K _p?K _i?K _d.Through simulation analysis,BP neural network fuzzy PID control has fast control time,small overshoot,and high control accuracy.(3)Based on the Shanke intelligent controller,in response to the functional requirements of the greenhouse water and fertilizer integrated machine,the underlying driver is optimized under the programming software Keil.According to the control requirements of the intelligent water and fertilizer integrated machine,design the irrigation and fertilization program and write the man-machine interactive interface to meet the actual operation requirements.(4)Based on the idea of layering,a multi-data integrated control platform for water and fertilizer in the greenhouse was constructed.The platform includes a decision analysis layer,a platform management layer,an application control layer and a resource layer,which are used for information interaction,analysis decision,and remote control of smart devices.