| There is a shortage of water resources in China for irrigation in agricultural production.Effective use of water resources and reasonable control of irrigation water consumption play a vital role in promoting agricultural production and improving the utilization rate of water resources.Up to now,the planting area of tomato crops in our country ranks among the top countries in the world.In order to ensure the normal growth of tomatoes while saving irrigation water,it is very necessary to predict the water demand of the key growth stages of tomatoes.At present,there are the following shortcomings in the process of crop production management: environmental data collection depends on manual observation and recording,which cause that data collection is not real-time and massive data is difficult to store;the key growth stages of tomato are not irrigated in time,which affects tomato growth,and over-irrigation cause water waste;lack of integrated intelligent agriculture platform that supports data collection and timely guidance of irrigation and other functions.In order to solve the above problems,the main contributions of this thesis are as follows:(1)The whole growth cycle of tomato crops requires adequate water,especially during the critical period of tomato growth.Therefore,it is necessary to ensure sufficient soil moisture.During tomato planting,there are many factors that affect the moisture content of soil,such as air humidity,soil humidity,and atmospheric pressure.To facilitate the collection of these environmental data,this article uses environmental sensors to collect the data,customize the communication protocol according to the needs,and transmit the data through the Lo Ra wireless communication technology.So as to solve the problem of difficult data collection.Customizing the communication protocol according to requirements,use Lo Ra wireless communication technology to transmit data,thereby solving the problem of difficult data collection.(2)Aiming at the problem of numerous environmental factors affecting tomato water requirement,information overlap among the factors was reduced.Principal component analysis(PCA)algorithm was used to reduce the dimension of the environmental factors that affected the prediction results of water demand of tomato.There is an extremely complex nonlinear relationship between environmental factors that affect the water demand during the tomato fruiting period.Radial Basis Function Neural Network(RBF)neural network is good at solving nonlinear problems,so it is suitable for water demand forecasting.In order to solve the problem of unstable initialization of RBF neural network parameters,this thesis proposes a method for water demand prediction in tomato fruit period based on Beetle Antennae Search(BAS)algorithm to optimize RBF neural network parameters.In this study,BAS algorithm is used to optimize the initialization parameters of the RBF neural network,and the tomato fruiting period prediction model of the BAS-RBF neural network is constructed.The results show that: Compared with the results of the RBF neural network model,the root mean square error of the BAS-RBF neural network model is reduced by 0.1168.The prediction performance is better than the traditional RBF neural network prediction model,which has better applicability in the prediction of water demand of the tomato fruit period.(3)In this research,a smart irrigation prototype system was built,which uses environmental parameter sensors to collect environmental data.The collected data is packaged through a custom communication protocol,and the data packets are uploaded to the Lo Ra gateway through Lo Ra wireless communication technology.The Lo Ra gateway uses 4G signals to upload the data packets to the server.Finally,the PC terminal displays the collected environmental data.You can also operate the PC terminal interface to issue instructions to control irrigation.In summary,the smart irrigation system platform can collect and record crop environmental data in real time.Users can use the BAS-RBF neural network prediction model to predict tomato water demand.Set the environmental parameter domain value on the system,and the system can automatically control the switch of the water pump.The platform automatically controls the irrigation equipment for timely water supply function.This article realizes the information management of crops and saves human resources by building a smart irrigation system platform.Therefore,the research has certain practical significance and application value. |
