Research And Design Of Greenhouse Monitoring System Based On Wireless Sensor Network

Greenhouse cultivation,as a technology of agricultural production and management under relatively controllable environment,has the unique advantages of not being affected by seasonal changes.It has gradually become an important agricultural production means to reduce the cost of planting,avoid the influence of extreme climate,and improve the yield of crops in northern China.In recent years,the modern greenhouse,which combines microprocessor control,wireless sensor network,cloud computing platform and other technologies,has achieved a great leap forward development,and it is gradually moving towards the direction of intelligence.However,with the expansion of the area of a single greenhouse and the scale of greenhouse group control,the early scheme of wireless sensor network applied in greenhouse control is no longer suitable,and the original control system gradually presents problems such as the increase of node energy consumption,the difficulty of long-distance communication,and the high cost of technology upgrading.In addition,the greenhouse environment has the characteristics of lag,strong coupling and time-varying,and it also leads to the difficulties of designing greenhouse control strategies.In view of the above problems,according to the control characteristics of greenhouse environment,the methods of sensor data processing and greenhouse temperature control are studied in this thesis.First,the data collected by sensors are preprocessed by the method based on support degree and compatibility matrix.Aiming at the problem of data loss in wireless communication caused by greenhouse environment interference and network communication,Kalman filtering is used to estimate the transmission data of sensor nodes.Secondly,support vector machine(SVM)algorithm is used to predict the greenhouse temperature under the influence of multiple factors,and the target value of greenhouse temperature in the next stage is obtained.Finally,in order to meet the effect of greenhouse temperature regulation,a fuzzy PID greenhouse temperature control algorithm based on Smith predictor is designed by using greenhouse mechanism model.On the basis of the above research,the greenhouse control system based on low-power long-distance communication wireless sensor network is designed by using narrowband IOT Lora technology.In the hardware circuit design,to achieve the greenhouse environment data collection and actuator control,the sensor node uses STM32F051K8U6 low-power microcontrollers as the controller to build the circuit of each module.The STM32F103C8T6 microcontrollers is used as the controller to design the concentrator circuit,coordinate the successful networking of nodes,receive node data,and it can also send and receive information with the upper computer.The software program of the control system includes the data acquisition program of the node,the networking program of the node and the concentrator.This thesis pay special attention to the anti-collision problem of network channel,and design a Mac layer communication association based on CSMA/CA and TDMA.The communication between the concentrator and the master computer is based on the GPRS module,and the TCP communication protocol is used to realize the low-cost and large capacity data transmission.According to the design requirements,the master computer interface is designed with Labview software,which realizes the real-time display and control of greenhouse environmental information.In the process of system debugging,the thesis successively completes the self-organizing network debugging of node and concentrator,GPRS module communication debugging,the master computer data acquisition and actuator control debugging.Display the current greenhouse parameters and equipment status by PC and Wechat T-Link applet,and record the alarm information.The experimental results are good,and meet the requirements of greenhouse system design scheme.