The Optimization Irrigation System Based On Intelligent Estimation Of Reference Evapotranspiration

Reference evapotranspiration(ET0)plays an important role in water resources scheduling of irrigation systems.With the development of Internet of things(IOT)technology,it is possible to collect some meteorological data and soil environment data in real time.Meanwhile,it is a key step to predict crop water requirement accurately for saving water resources continuously and effectively in the field of agricultural irrigation.At present,some neural network models are used to predict crop evapotranspiration which can be multiplied by the corresponding crop coefficient to obtain crop water demand.In this paper,we are devoted to studying bio-intelligence algorithms to improve the prediction accuracy of crop evapotranspiration.Firstly,we use biological heuristic optimization algorithm called particle swarm optimization(PSO)and select a new activation function to adjust the parameters of extreme learning machine(ELM);In addition,this paper proposes a new deep extreme learning machine network model(DELM)on the basis of studying human brain learning and memory mechanism and the characteristics of big data stream with concept drift.The DELM model improves the prediction accuracy,reduces the prediction time and computation cost,achieves real-time online data prediction of crop evapotranspiration.Also,it improves the generalization performance and prediction of the stability of the traditional extreme learning machine network.The main works of this paper are as follows:(1)Aim at the small data predicting model,this paper proposes a novel extreme learning machine optimized by particle swarm optimization named PSO-SWELM to realize more accurate estimation of evapotranspiration with limited meteorological and environmental data.The weights and thresholds of ELM are optimized by PSO and a new function which is based on the two-wave superposition is selected as the activation function of ELM,which both enhances the accuracy of estimation.The simulation results show that the proposed method has better performance in predicting the evapotranspiration than the currently prevailing methods.(2)Aim at large-scale real-time data predicting model,a new deep learning network model called DELM is proposed.The collected large-scale data is divided into multiple small data block and DELM model simulates the human brain's re-consolidating learning mechanism and adopts the strategy of "seeking common ground while reserving differences" which refers to use incremental extreme learning machine training to update model based on existing models and new block datasets when the prediction accuracy of irrigation water demand is not meet the current forecast target demand.Also,the hidden layer nodes are dynamically updated according to the prediction accuracy of each data block to improve the prediction accuracy of the model.Finally,the model not only reduces the prediction time loss and calculation cost,but also realizes large-scale real-time data online prediction and improves the traditional ELM network generalization performance and prediction of stability.(3)Realize the construction of agriculture IOT intelligent irrigation system platform;we can study the meteorological information,farmland soil information through IOT sensor devices.Then we establish a good database collection system where the real-time collected data is stored and build a bridge between the database and Matlab algorithm simulation software platform.Finally,IOT intelligent irrigation system platform is used to solve the specific crop water demand by using the proposed new intelligent predicting models,and the irrigation plan can be rationally established to control the irrigation equipment timely and appropriately.This paper improves ELM prediction model through biological heuristic algorithms and deep learning mechanism,and finally realizes more accurate evapotranspiration estimation with limited environmental and meteorological data.The simulation results of the proposed algorithms show that this paper makes a contribution in conserving the water resources and irrigating crops reasonably.