Abstract
The task of fault detection is crucial in modern chemical industries for improved product quality and process safety. In this regard, data-driven fault detection (FD) strategy based on independent component analysis (ICA) has gained attention since it improves monitoring by capturing non-gaussian features in the process data. However, presence of measurement noise in the process data degrades performance of the FD strategy since the noise masks important information. To enhance the monitoring under noisy environment, wavelet-based multi-scale filtering is integrated with the ICA model to yield a novel multi-scale Independent component analysis (MSICA) FD strategy. One of the challenges in multi-scale ICA modeling is to choose the optimum decomposition depth. A novel scheme based on ICA model parameter estimation at each depth is proposed in this paper to achieve this. The effectiveness of the proposed MSICA-based FD strategy is illustrated through three case studies, namely: dynamic multi-variate process, quadruple tank process and distillation column process. In each case study, the performance of the MSICA FD strategy is assessed for different noise levels by comparing it with the conventional FD strategies. The results indicate that the proposed MSICA FD strategy can enhance performance for higher levels of noise in the data since multi-scale wavelet-based filtering is able to de-noise and capture efficient information from noisy process data.
Highlights
In modern petrochemical industries, process abnormalities or faults hamper smooth functioning and this needs to be addressed before they pose a further threat
An multi-variate statistical process monitoring (MSPM) technique based on Independent component analysis (ICA) has been investigated in the last few years to handle the problem of monitoring non-gaussian processes
The fault detection rate (FDR) and false alarm rate (FAR) indices are used to evaluate the performance of each fault detection strategy and they are calculated in percentage using the below representation: FDR = Total samples (J > Jth)| abnormal condition Total samples in faulty region (20)
Summary
Process abnormalities or faults hamper smooth functioning and this needs to be addressed before they pose a further threat. In the early years when multi-scale filtering was proposed for fault detection domain, it was integrated with PCA to have an MSPCA strategy [26]. In the work by [26], the PCA strategy was developed at each scale followed by combining the results at individual scales where significant events were present This enabled wavelets’ ability to detect deterministic changes and capture those features where abnormal operations were evident. This work aims to develop an FD strategy by demonstrating the ability of wavelets in de-noising and capturing essential non-gaussian features from process data through ICA modeling technique. 2 of this paper, the PCA- and ICA-based fault detection strategies are presented This is followed by discussion on wavelets, multi-scale representation of data and proposed MSICA fault detection strategy in sect.
Sign-in/Register to view highlights & summary 
Paper version not known (
Free)
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call