Research On Fault Diagnosis Technology Based On Agricultural Internet Of Things

Agriculture is the foundation of our national economy. With the rising level of informatization, much more technology related with agricultural internet of things is appearing. Besides, the government is paying more attention to agricultural internet of things. However, the nodes of wireless sensor network in agricultural internet of things are more likely to malfunction because the environment of the agricultural green house is harsh, and the agricultural internet of things in the market is not able to solve the technology of fault diagnosis. Therefore, this paper does research on fault diagnosis technology based on Agricultural Internet of Things.This paper chooses an agricultural IOT based on ZigBee as the platform for research on Fault Diagnosis Technology and introduces the agriculture IOT using wireless sensor network technology and ZigBee technology. Besides, the paper illustrates the framework of agricultural IOT based on ZigBee, and makes a brief description of the hardware.The fault of agricultural IOT is divided for hard fault and soft fault. Hard fault includes the communication fault and processor fault. Soft fault includes impact-faul, constant-fault, deviation-fault and multiple-fault. Hard fault diagnosis algorithm was developed by using group algorithm for fault diagnosis of cluster head, to ensure that the node of cluster head is normal node. Under the condition that the head cluster is normal node, the parper designs a fault diagnosis algorithm of the nodes in the cluster, which realize the fault diagnosis of collector nodes and controller nodes in the field of communication fault and processor fault. For soft fault, this paper uses the method of time series to set up time series by subtracting the preset value, and uses the auto regressive model as the mathematical model of time series. The least square method is chosen as the calculation method of the auto regressive parameters. The fault diagnosis algorithm based on time series is developed by using the parameter calculation and the order calculation.Two kinds of fault diagnosis algorithms are tested by using the Agricultural IOT cloud platform, which can realize the real-time display and fault detection of the collector and controller nodes. The user can observe the fault status of the nodes in the greenhouse by the Agricultural IOT cloud platform, and gives an example of the application of fault diagnosis technology in agricultural IOT.At the end of the paper, this paper gives some suggestions of the future research direction on the fault diagnosis of agricultural internet of things, including reducing network traffic to carry out fault diagnosis, improving the accuracy of fault diagnosis and remotely promotes the updating of the fault diagnosis module.