论文部分内容阅读
故障诊断系统的性能评价是开发和验收故障诊断系统不可或缺的重要环节.针对基于贝叶斯网络(BN)故障诊断系统的性能评价需要,考虑系统诊断结果真实分布,提出采用二项分布参数估计方法来计算诊断准确度的置信区间,采用准确度期望值及其置信区间全面客观评价诊断模型的性能,形成贝叶斯网络模型诊断能力的量化评价指标,为诊断结果的可接受、可信程度以及诊断模型的训练充分性提供参考依据.最后通过燃油系统故障诊断实例验证所述性能评价的有效性.
The performance evaluation of fault diagnosis system is an indispensable and important link in the development and acceptance of fault diagnosis system.Because of the need of performance evaluation based on Bayesian Network (BN) fault diagnosis system and considering the true distribution of system diagnosis results, the binomial distribution parameter Estimation method to calculate the confidence interval of diagnostic accuracy, using the expected value of accuracy and its confidence interval to evaluate the performance of the diagnostic model in a comprehensive and objective way, forming the quantitative evaluation index of the diagnostic ability of the Bayesian network model, which is the acceptable and credible degree of the diagnostic result As well as the adequacy of the training of the diagnostic model.Finally, the effectiveness of the performance evaluation is verified by an example of fault diagnosis of the fuel system.