Research And Implementation Of GA-BP Based Forecast Algorithm For Greenhouse Irrigation Water Conservation

In the context of water shortage in China,61.4% of the total water resources are used for agricultural irrigation,and irrigation is the only way to meet the water demand of greenhouses crops.At present,the irrigation management methods of many greenhouses are extensive,and there are some shortcomings:(1)Artificial control of irrigation amount,and no reasonable irrigation system for reference to the water demand of greenhouses crops,so the irrigation is extremely inaccurate.Even most of the time it is flooded irrigation,such an extreme waste of water resources,and the labor intensity is too large for a large-scale greenhouse.(2)Most greenhouse irrigation systems mainly through Zig Bee,WIFI to realize the collection of environmental parameters.However,Zig Bee and WIFI have short transmission distance and fast attenuation,and are susceptible to interference from other signal frequency bands.Their networks are relatively complex,and cannot satisfy the requirements of low power consumption and long-distance transmission at the same time,and cannot well satisfy the actual environment of large-scale greenhouses.To solve the above problems,establish a crop water demand prediction model,and choose Lo Ra technology to build a network with the characteristics of long transmission distance,large capacity and low power consumption.Through scientific and accurate prediction,the optimized irrigation scheme effectively changes the extensive greenhouse irrigation method and the blindness and randomness in the irrigation process.Finally,the efficient utilization of water resources can be used for achieving the purpose of saving water.The main work and innovation points of this paper are as follows:(1)Establishing a GA-BP crop water demand prediction model that optimized the BP neural network through GA,establishing a multi-dimensional target model using factors such as air temperature and humidity,soil humidity and light intensity that have a greater impact on crops,forecasting the water demand of crops,the results show that the GA-BP prediction results are highly accurate.(2)A demonstration and verification system based on water-saving irrigation in greenhouses was built with Lo Ra technology.The Lo Ra terminal node is designed to realize the automatic collection of environmental data and upload it to the Lo Ra gateway,which can upload data to the server through WIFI or 4G.Finally,the Android APP monitoring software remotely monitors the greenhouse environmental data andcontrols the greenhouse irrigation.(3)The GA-BP crop water demand forecast model is applied to Lo Ra irrigation system,through the irrigation system collects and processes the data as the input of the GA-BP crop water demand prediction model,and the output of the model is the crop water demand.The output result is fed back to the irrigation system,and the irrigation system performs intelligent and accurate irrigation according to the predicted water demand.Research and implementation of GA-BP based forecast algorithm for greenhouse irrigation water conservation provides a reasonable and effective irrigation schedule for crops,and realizes the automatic collection and irrigation control of long-distance greenhouse data under low power consumption,realizes accurate irrigation to achieve the purpose of water saving and has certain process application value.