Fault Diagnosis Of The WSN Of Measuring And Controlling System In Greenhouse Based On The Time-Series And Space-Series Analysis

Reliability is an important index of the Data Acquisition System,especially to the WSN control system in greenhouse which works in the High Temperature and High Humidity environment,high reliability is more important.In general,the power supply system of sensor node in WSN is battery and the sensor node receives and transmits data through wireless communication,so it has such weaknesses as limited energy and weak anti-jamming ability, resulting in a high probability of occurring fault.Fault diagnosis is an effective method in ensuring high reliability of the WSN of Controlling System in Greenhouse.Fault diagnosis for the WSN of Controlling System in Greenhouse should try to reduce communication cost,energy consumption, the computational complexity of fault diagnosis algorithm and the false alarm rate during the fault diagnosis.However,the current researches on the fault diagnosis method of WSN mainly focus on the theory,and can not meet the above demands.Therefore,it is necessary to conduct studies on more practical fault diagnosis methods for the WSN of Controlling System in Greenhouse.The environmental information collected by the WSN of Controlling System in Greenhouse changes slowly,so the time-series and the space-series formed by the information has an obvious stationary.When fault occurs,the data got by the WSN of Controlling System in Greenhouse will change abruptly,so the series will become unstable at the fault point accordingly.Therefore, through analyzing the time-series and the space-series,it can be known whether fault occurs. Furthermore,the time series analysis method is characteristic of easy modeling,small computational amount,and strong practicality,which can meet the demands of the fault analysis for the WSN of Controlling System in Greenhouse.From the perspective of increasing the practicality of fault analysis method,according to the basic theory of time-series analysis,on the basis of theoretical analysis,simulation and experiments,an effective fault diagnosis method for the WSN of Controlling System in Greenhouse is set up,and this method does not need more communication costs and energy consumption of sensor node.The collected real time environmental data is pre-treated as the diagnosis sample series,and when the time series analysis model becomes abnormal,the space series diagnosis model is used to determine the fault.The results show that generally the above method can monitor the running state of the nodes of the WSN of Controlling System in Greenhouse,find out the anomaly of the WSN of Controlling System in Greenhouse and judge the fault node,thus enhancing the reliability of WSN.