文献类型：专利

英文题名：Vine described copula relevance-based multi-model processing fault detection method, involves calculating generalized local probability index value, and calculating Bayesian inference generalized BIP index value

作者：LI S;REN X;ZHENG W;YANG Y;XU W

机构：[1]UNIV EAST CHINA SCI & TECHNOLOGY

申请号：CN104914775-B

公开日：2017-05-31

语种：英文

收录：DERWENT

摘要：NOVELTY - The method involves obtaining training sample normal information (S1). Unite probability density model establishing process is performed (S2). A unite probability density function value is calculated (S3). Density static state determining process is performed (S4). Time monitoring data is obtained. A generalized local probability index value is calculated (S5). A Bayesian inference generalized BIP index value is calculated (S6). A random vector value is calculated. Cumulative distribution information is obtained. A discrete fractile value is calculated. USE - Vine described copula relevance-based multi-model processing fault detection method. ADVANTAGE - The method enables realizing multi-model processing fault detection process in a simple manner. DETAILED DESCRIPTION - An INDEPENDENT CLAIM is also included for a vine described copula relevance-based multi-modal process fault detection system. DESCRIPTION OF DRAWING(S) - The drawing shows a flow diagram illustrating a vine described copula relevance-based multi-model processing fault detection method. '(Drawing includes non-English language text)' Step for obtaining training sample normal information (S1) Step for performing unite probability density model establishing process (S2) Step for calculating unite probability density function value (S3) Step for performing density static state determining process (S4) Step for calculating generalized local probability index value (S5) Step for calculating Bayesian inference generalized BIP index value (S6)