Research On Hybrid Circuit Fault Diagnosis Method For Radio Frequency Transceiver Chip Of Wireless Sensor Networks

In the wireless sensor networks, the faults' of sensor nodes are inevitable in the process of operation. These are the main sources of hardware failure, link failure and the influence from the environment, in which there is a hardware failure often cause relatively serious consequences, sometimes even the failure of a component will cause the entire system paralyzed, so the research of fault detection and fault diagnosis method which is about circuit system of wireless sensor network has become a hot topic in the research of wireless sensor network.This thesis mainly studies the fault diagnosis method of wireless sensor node related electronic circuit system, the main work is as follows:(1) take a wireless sensor node as an example to analyse the RF chip and the control board by using quantization FMEA, extract the relevant fault sensitive parameters, based on which, and design an online fault diagnosis platform, including the core processing unit, RF :currentdetection, signal processing, condition monitoring module; (2) according to the fault waveform data extracted from the online fault platform,design a mixed circuit fault diagnosis algorithm, focus on the research offault diagnosis algorithm based on BP neural network, the simulation tests found that the fault recognition rate need to be further improved; (3)the introduction of the thought and method of decision-making, and design a fault diagnosis algorithm based on decision-making of distance measure, through theoretical calculation and practical test to prove that the fault recognition rate compared to BP neural network fault diagnosis algorithm has been greatly improved.The hybrid circuit fault diagnosis method can be used in other typesof circuit fault diagnosis.The following research work is to research in-depth about the extraction method of the fault waveform feature of the under test system,in order to extract appropriate feature parameters, at the same time, an in-depth study about the calculation method of establishing weights and decision-making operator will be carried out, in order to further improve the recognition rate of the online fault diagnosis system.