Research And Implementation Of Herb Watering Model And System

Recently,more and more people have been raising herbs at home as the gradual improvement of the quality of life.However,due to many reasons such as being away from home or busy work,the plants can't get properly water sometimes.In order to solve this problem,the goal of this thesis is to design a watering model for herbs,which can decide water time and calculate water volume according to the environment and the herbaceous variety.Besides,this thesis designs and implements a watering system for herbs,which can automatically water herbs.To implement the watering system for herbs,several technologies are used including Arduino platform and related equipment,SpringMVC framework,Flask framework,Mosquitto software,MySQL database and Android development technology.To study the watering model for herbs,the time series prediction model based on fuzzy theory and the machine learning regression model are used.First,the requirement analysis of the watering system for herbs is carried out,including user control,device control,herb control and watering control.Base on the requirement analysis,the summary design is carried out including system architecture,deployment structure,4 function modules and the database ER diagram.Then,the detailed design is obtained including process for function modules,design for classes,interface design for HTTP and MQTT protocol and design for database tables.Based on the design work,the initial system implementation is completed,including using Arduino to implement automatic watering electronic equipment,using MySQL as database to store data,using SpringMVC framework to implement background server.Then,the watering model for herbs is studied.First,dataset is obtained by means of experiment.Two kinds of prediction models are trained according to the dataset,one is the soil moisture prediction based on fuzzy theory and the time series prediction model,another model is the soil moisture prediction model based on evapotranspiration.After comparing the predictive effect,the soil moisture prediction model based on evapotranspiration is chosen to become the prediction model for watering system for herbs.Then,watering strategy is obtained considering water preference for herbs,pot volume and soil quality.The watering model for herbs is obtained according to the moisture prediction model and the watering strategy,which will water herbs without soil moisture sensor.Finally,on the basis of system design,with the watering model for herbs,the watering system for herbs is completed,including implementation of Android App,Web client,the application logic control server based on Flask framework,etc.29 test cases is designed according to 4 function modules and all of them pass.