关于动力环境监控贝叶斯预测系统中参数获取的研究

针对目前动力环境监控系统中存在的一些问题,采用贝叶斯推理的方式处理一些不确定的信息,用于指导系统维护和决策。简单叙述了目前常见的一些抽样方法和预测方法,从统计学的角度讨论了如何对监控数据进行抽样才能更好的获取贝叶斯网络的先验概率和后验概率,并结合一些其他预测方法来完善推理系统,以实现对系统故障的有效预测。

The Study on the Parameters of Bayesian Forecasting Obtained in Power and Environment Monitoring System

Against some problems in the current power and environment monitoring system,this paper uses Bayesian Inference to approach to some uncertain information for guiding system maintenance and decision-making.It briefly describes the current common sampling methods and some forecasting methods,and discusses how to sampling the monitoring data can access to priori probability and posterior probability of the Bayesian network from the statistical point of view,and improves the forecasting system by combining some of the other forecasting methods to achieve an effective forecasting for the failure of the System.