基于多维高斯贝叶斯算法在雷达故障信息诊断系统应用

为提升雷达故障信息的诊断效率,节约执行故障诊断指令所需消耗的数据成本,设计基于多维高斯贝叶斯算法的雷达故障信息诊断系统;利用完整的框架结构设计客户机/服务器模式,联合服务器与信息查询模块,实现雷达故障信息诊断系统的体系结构搭建;在此基础上,计算故障信息的松弛度数值,通过判定迭代门限的方式,实现对多维系数信号的处理,完成基于多维高斯贝叶斯算法的雷达故障信息感知;联合已嵌入的雷达故障信息,连接Access数据库,实现诊断系统的接口技术研究;综合上述所有理论依据,设计基于多维高斯贝叶斯算法的雷达故障信息诊断系统;借助Windows平台,分别连接上述系统与远程故障诊断原型系统,在雷达故障信息诊断系统的应用过程中得出,基于多维高斯贝叶斯算法雷达故障信息诊断系统的诊断效率最大值为45％,RSR指标最大值为16.7％,指令执行消耗的数据成本较低,表明所设计系统具有一定的实际应用性.

Design of radar fault information diagnosis system based on multi-dimensional Gaussian Bayesian algorithm

In order to greatly improve the diagnostic efficiency of radar fault information and save the data cost required for the execution of fault diagnosis instructions, a radar fault information diagnosis system based on multi-dimensional Gaussian Bayes algorithm is designed. Utilize the complete frame structure, design the client/server model, and combine the fault diagnosis server and the information query module to realize the architecture of the radar fault information diagnosis system. On this basis, the convex slack value of the fault information is calculated, and the multi-dimensional coefficient signal is processed by determining the iteration threshold, and the radar fault information perception based on the multi-dimensional Gaussian Bayes algorithm is completed. Combine the embedded radar fault information and connect to the Access database to realize the interface technology research of the diagnosis system. Based on all the above theoretical basis, the design of radar fault information diagnosis system based on multi-dimensional Gaussian Bayes algorithm is completed. With the help of the Windows platform, the above-mentioned system and the remote fault diagnosis prototype system are respectively connected. Through comparative experiments, it is concluded that the diagnosis efficiency of the radar fault information diagnosis system based on the multi-dimensional Gaussian Bayes algorithm is high, and the data cost of the instruction execution is low, and it has a certain degree of Practical applicability.