The Robust Data Communication And Data Processing Study Of Wireless Equipment Condition Monitoring System In Nuclear Power Plants

The monitoring of Equipment operation and health condition is an important method to ensure the safety in manufacture of nuclear power plants(NPPs), but the wired system of currently used has large limits; for example, it is high cost and is hardly extended at a finish building NPPs. It is an innovation of monitoring technology of NPPs to deploy wireless sensor networks in equipment condition monitoring of NPPs,for it is not only as an effective supplementary, but also it can extend range of monitoring, improve monitoring ability, and decease the cost of monitoring. But owing to the severe demand of safety and reliability in NPPs, and owing to the harsh environment of NPPs(nuclear irradiation, electromagnetic interference, high temperature, high humidity), there are some challenges of the application of WSN technology, for example, they may increase the rate of faults of monitoring system and may degrade its performance of communication and measurement, meanwhile it make parameter perturbation of system. Thereby it is very important to study the robustness of wireless condition monitoring system. In the other hand, wireless monitoring system is a loose coupling wireless sensor network, and it has application constraint of limited resource, limited calculating ability, limited energy, and these make the research of robustness of wireless monitoring system complicate.To maintain the normal performance of equipment condition monitoring system under the harsh environment stressor and parameter perturbation, the paper studies the robustness of wireless monitoring system from the view of data communication and data processing, and mainly solve following problem: 1) How to decrease the abnormal change of test data under parameter perturbation of measurement device; 2) How to decrease the abnormal change of missing data under parameter perturbation of wireless communication of sensor nodes. 3) How to decrease the abnormal change of the performance of wireless equipment condition monitoring system under parameter perturbation of WSN.To solve these problems, firstly, the paper studies the data communication and processing method in the field test layer, the wireless communication layer, and the network monitoring layer; and focuses on researching maintain system robustnessmethods of decreasing the uncertainty of test data, decreasing the number of missing data, processing missing data, simplifying the data stream, normalizing the network structure, and so on. Secondly, the paper studies the method of improving robustness of monitoring system that has large number of data and complex network topology, mainly studies robust compressing sense arithmetic for equipment condition sound signal, the chaos phenomenon in complex wireless network, and the relationship between chaos and monitoring parameter estimation.The main contributions of paper are:1. The contributions in the field test layer include:At first, the paper uses outlier detection as filter, uses the chance measurement as decouple, combined with short time state space equation with disturbance, establishes measurement signal and its feature value filter. And the paper solves problem of how to build a resource limited robust filter under environment stressing and measurement structure changing. Results of theory and experiment show that the robust measurement filter based on differential manifold is stable and effective.Secondly, the paper uses cluster analysis method and feature ARMA method to judge the faults of sensors, combined with distributed redundancy measurement short time adaptive filter arithmetic, establishes self-judge redundancy measurement filter.And the paper solves the problem that test results are influenced by nodes dismayed or nodes performance degraded under nuclear irradiation environment. The theory and experiment prove that the arithmetic decrease measurement uncertainty of test, improve the robustness, and meanwhile it decrease the influence of input and measure structure change.2. The contributions in the wireless communication layer include:The paper uses Krylov subspace method, designs dynamic channel selection arithmetic; the paper solves the problem that lots of communication data is missed by the change of channel parameter.The paper uses decreasing dimension and PCA method, based on Fisher information distance and CRLB, designs a robust analysis method of random missing data.The results of theory and experiment show that the dynamic channel selection arithmetic can decrease the communication missing data, the robustness analysis ofmissing data can decrease that the test results is influenced by missing data.3. The contributions in the network monitoring layer include:The paper uses methods of simplifying data stream and decreasing uncertainty of network, proposes regulations of the standard deterministic deployment; and designs the information routing wireless access arithmetic based on independent relay nodes; the paper solves the problem of the uncertainty of network data communication and data process when there is unmoral changing of network data in the wireless sensor network.The results of theory and experiment show the valid of information routing wireless access arithmetic and the standard deterministic deployment regulation of WSN.4. The contributions in the equipment acoustic signal and chaos of network include:The paper proposes the HRS compressed sensing arithmetic based on hardware threshold value; and analyze the robustness of the arithmetic. The paper builds nonlinear dynamic model of WSN under the complex status of nodes losing function and being attacked, studies the phenomenon of chaos in complex network, and proposes the local bidirectional circle relay topology to improve robustness and decrease number of missing data. The simulation and experiments prove the valid of HRS compressed sensing arithmetic and local bidirectional circle topology.