| With the rapid development of microelectronics technology,more and more large-scale complex systems have emerged in industrial production,aerospace and other fields.The multi-signal model based on layering improves the deficiencies of the information flow model structure,and also takes into account the characteristics of the structured model,and has been widely used in electronic systems.Large,complex RF systems with numerous modules and test specifications make this complex block structure ideal for diagnostics using multi-signal models.In order to further improve the diagnostic efficiency,the research content of thesis is proposed.Thesis first compares the advantages and disadvantages of multiple modeling methods,analyzes the advantages of using multi-signal models for RF circuits,and introduces the modeling methods of multi-signal models in detail with a certain type of aircraft air conditioning system as an example,laying a foundation for the establishment of subsequent low-noise RF circuit models.Secondly,a low-noise amplification circuit based on the ATF54143 chip is used as a typical RF circuit as the research object of thesis.The normal specifications of the RF circuit are simulated by simulating the low noise amplifier circuit with an operating frequency of 2.45 Ghz.Select the S-parameter,stability factor,noise figure,standing wave ratio,and gain as the measurement point.According to the failure threshold of each parameter,the fault dependence matrix of the low-noise amplifier circuit is determined to determine whether the fault source device has failed under the corresponding test.It provides data support for the two following chapters to use parallel algorithms to troubleshoot the system.Third,the fault diagnosis technology of parallel RF circuit based on improved real-time diagnosis algorithm is studied.Based on the real-time diagnostic algorithm,thesis proposes a multi-core parallelization scheme.The real-time diagnostic algorithm of the algorithm is decomposed into three parts,and different parallelization schemes are adopted according to the different characteristics of each part.Applying an improved parallel algorithm to the spacecraft’s Apollo West launch system can be found to correctly diagnose the results,and the diagnosis is faster.Experiments are conducted through different scale dependency matrices,and the relationship between the acceleration ratio of the parallel algorithm and the scale of the dependency matrix isobtained,and the maximum acceleration ratio reaches 3.5 under the quad-core CPU,which has a good effect.Finally,the parallel algorithm is applied to the RF circuit of thesis and gets good results.Finally,the fault diagnosis technology of parallel RF circuit based on improved Lagrange algorithm is studied.Based on the Lagrange relaxation and Bayesian criteria,thesis proposes to parallelize and apply the algorithm to RF circuits.This optimization scheme takes full advantage of CPU usage by using modules that do not have interdependencies in multiple threading algorithms.According to the experimental results,the correctness of the parallel algorithm and the efficiency of the parallel algorithm are verified.Finally,the application of the parallel Lagrange fault diagnosis algorithm to the RF circuit of the low-noise amplifier has also achieved good results,and also laid the foundation for the fault diagnosis of large RF systems. |
