In actual production processes, the occurrence probability of multiple faults is much higher than that of a single fault, which will affect the process industry operating performance and final products quality. This paper is concerned with industrial practices and theoretical approaches for detection and location of key performance indicator (KPI)-related multiple faults in process industries. Firstly, a new KPI-related multiple fault monitoring scheme is addressed from the subprocess level based on the developed correlation-based canonical variable analysis (CCVA) model. Then, Bayesian fusion is implemented to form the final monitoring decisions from the plant-wide level. After that, a tensor subspace analysis-based discriminant analysis (TSADA) method is proposed for locating the root causes, which will help field engineers to take correction actions and recover the process operations. Finally, the application to a typical industry process, i.e. hot strip mill process (HSMP), is given to demonstrate the performance and effectiveness of the proposed methods with real industrial data.
Large-scale manufacturing processes are usually made up of multiple interrelated and distributed continuously subprocesses, which are transmitted and connected by information and quality flow. The characteristics of long processes, quality heritability between subprocesses, and dynamic-coupled variables bring severe challenges to conventional quality-related fault diagnosis. Against this background, a novel distributed diagnosis framework for quality-related faults is proposed in this article. First, the sequential manufacturing process is decomposed into multiple subprocesses based on mechanism knowledge. Second, a novel dual-attention quality-driven autoencoder method is designed as the model for local fault diagnosis. Deep nonlinear features are extracted under quality supervision; meanwhile, the dynamic information and the different correlations among variables are also considered. Then, based on the tandem structure of the manufacturing process, multiple dual-attention quality-driven autoencoder models corresponding to each subprocess are constructed and stacked into a distributed model. Bayesian inference is used to build global monitoring statistics. Moreover, after faults occur, the intervariable attention weights are achieved to identify faulty variables. Finally, the effectiveness and advantages of the proposed framework are demonstrated via a practical large-scale sequential manufacturing process, the hot strip mill process.