Research On Fault Diagnosis Method For Sensor Network Node

Wireless Sensor Network(WSN) is a new information acquisition and processing technology, its application has been extended to many fields, caused great attention in the world military, industrial, environmental and academic circles. At present the structure of WSN system becomes more and more complex, the function more perfect, increasingly high degree of automation, prolong the service life of the topic is more popular in recent years as much as possible. The fault node of the wireless sensor network system will reduce the quality of service, node fault diagnosis plays an important role in the WSN system, so the need for timely and accurately through the fault diagnosis of various fault state to make the diagnosis, in order to improve the system operation reliability, safety and effectiveness.This paper studied the detection of transient and noise, constant failure. Detection method for the least squares method for detection and estimation method for first order rules, two yuan AR detection method based on time sequence based on, and gives the application of these methods in the practical data in the detection of. By injecting method of fault data, to analyze the detection performance of three kinds of methods, not only embodies the difference between three methods in detection accuracy and robustness, but also can be used for sensor fault detection effectiveness to provide reference for other methods.This paper further on the wireless sensor network fault detection technology comparison method has significant weaknesses and put forward a new kind of self-diagnosis based on Petri net model, the model can be dependent on the adjacent node data is reduced to a minimum, when the sensor of all parts and components are the link between the intact, sensor was diagnosed no fault. The model by analyzing the sensor component behavior can be accurately determined sensor state, and the application of related images, using Pearson correlation coefficient was used to assess the association state of internal sensor, if the above two parts are in line with the requirements of the sensor is fault free. The simulation results show that with the help of HPSIM can detect the permanent fault.