贝叶斯推理在动力环境监控系统中的应用研究.pdfVIP

摘 要 作为电信运营商维护管理系统的一个重要手段，动力环境监控系统其 功能的完善对于提高服务质量具有很高的经济价值。本文分析了非智能化 动力环境监控系统中存在的一些问题，针对系统中存在大量的不确定性信 息，讨论了如何采用贝叶斯推理的方式来完善其预警以及智能化分析方面 的功能，重点研究贝叶斯分类器在监控技术中，如何利用自适应滤波算法 完成参数的学习实现推理。 通过对监控系统的告警产生过程进行仿真，测试了该模型的非线性处 理能力和稳定性，并验证系统通过推理模型能够实现预警功能和告警过滤 功能。同时，根据贝叶斯推理的特点提出了一种新的告警产生和过滤流程， 为进一步提高监控系统的智能化拓宽了思路。 关键词 ：监控系统，粒子滤波，自适应滤波，贝叶斯推理,后验概率 1 Abstract As an important tool that the telecom operators use to protect and manage their system, the power and environment monitoring system to improve its function is of high economic value in improving the quality of services. This paper analyzed some problems existed in the field of early warning in the non-intelligent power and environment m onitoring system. Against there is a lot of uncertain information in the system. It discussed how to use the way of Bayesian Inference to improve its early warning and intelligence analysis function. It focused on how the Bayesian Classifier achieved inference through using the adaptive particle filter algorithm to learn the parameter in the monitoring technology. Using the simulation of a warning process it tested the non-linear processing capacity and stability of the model, and confirmed that the monitoring system had the early warning and filtering function by using Bayesian inference model. At the same time, according to the characteristics of Bayesian Inference, it proposed a new process to report and filter the warning. This paper broadened a thinking to further improve the intelligent of the monitoring system. Keywords : Monitoring System,