Font Size: a A A

Study On Prediction Method Of Soil Moisture Content In Tea Garden Based On Extended Autoregressive Model

Posted on:2018-07-29Degree:MasterType:Thesis
Country:ChinaCandidate:X Q ZhuFull Text:PDF
GTID:2393330518977780Subject:Agricultural informatization
Abstract/Summary:PDF Full Text Request
Intelligent agriculture based on Internet of things relies on national agricultural development policy,has gradually become the main force of scientific and technological progress in the future.As a new trend in the development of modern agriculture,precision irrigation technology is developed on the basis of organic combination of agricultural production measures and high-tech.The main purpose of precision irrigation technology is to avoid blindly farmland irrigation induced water resources waste and agricultural production through science and technology,rational utilization of water resources,so as to achieve the purpose of improving the economic benefit.The application of drip irrigation system in tea garden realizes the fine management of precision irrigation and tea plantation,which is of great significance to improve the modernization level of tea plantation,save water resources and improve tea quality.Soil water content is one of the main basis of determining whether tea garden needs irrigation.Therefore,the accurate grasp of soil moisture content is the premise of precision irrigation,soil moisture content in different depths is an important indicator of soil moisture.The tea plant is densely planted,mostly in mountainous area,slope,and there are obstacles such as utility poles and other trees in tea garden.It is very important to find the reasonable layout of the sensor to Obtain the soil moisture temperature and nutrient information effectively.The main contents and research results of this paper are as follows:(1)The remote monitoring system of tea garden information is introduced,the remote monitoring of tea garden information consists of two parts: information collection method and remote transmission.Based on the ZigBee wireless sensor network node module,the convergence point module and the routing module,a multi-hop and self-organizing wireless sensor network can be constructed to cover the tea garden target information.The information acquisition mainly completes the real-time acquisition and remote transmission of tea garden information through high-precision environment-aware equipment,wireless sensor network(WSN)network node,outdoor PH automatic weather station and real-time terminal control equipment.Remote transmission is used with the Welleplus WR200 series controller.The central processing unit supporting the GPRS module for the RS485 communication,the use of Modbus as a communication protocol.(2)The error data that appears in the monitoring data can not be directly used for analysis and modeling.Pauta is used to eliminate gross errors,and then through the wavelet denoising to reduce noise.As the uneven distribution of parameters in the tea garden environment,the data time or space of single sensor may be deficient in locality.The data fusion technology based on batch estimation is used to combine the time and space of the same sensor data respectively to form the description of the uniformity of the environmental parameters of the tea garden.(3)The prediction of soil moisture content in different depths has important practical significance for the formulation of reasonable scientific irrigation plan.This paper introduces a model of prediction of soil moisture content of different depth based on ARX model,the model can predict the soil moisture of other levels according to the measurement of soil water content in a few levels.Using correlation analysis to determine model input and output,using a period of soil moisture data to train the model,determine model structure and model parameters.In the prediction stage,the trained model can predict soil moisture content of other depths according to available data.The effectiveness of the model is evaluated by relative prediction error,root mean squared error and model Fit.(4)Construct the tea garden information platform,and early warning after comparison between prediction value of soil moisture prediction and water index of tea tree.The tea garden information database is constructed,and the information platform is connected with the information database to obtain tea garden information.The tea garden information platform realizes real-time data display,equipment status display,implant warning,user management,remote control and other functions.Interface of the information platform is simple,and easy to operate,easy to expand.The tea garden drip irrigation system automation and convenient control is achieved.
Keywords/Search Tags:Tea garden, Monitoring, Data preprocessing, Modeling, Control
PDF Full Text Request
Related items