Abstract
In view of the randomness in the selection of kernel parameters in the traditional kernel independent component analysis (KICA) algorithm, this paper proposes a CPSO-KICA algorithm based on Chaotic Particle Swarm Optimization (CPSO) and KICA. In CPSO-KICA, the maximum entropy of the extracted independent component is first adopted as the fitness function of the PSO algorithm to determine the optimal kernel parameters, then the chaotic algorithm (CO) is used to avoid the local optimum existing in the traditional PSO algorithm. Finally, this proposed algorithm is compared with Weighted KICA (WKICA) and PSO-KICA with Tennessee Eastman Process (TEP) as the benchmark. Simulation results show that the proposed algorithm can determine the optimal kernel parameters and perform better in terms of false alarm rates (FAR), detection latency (DL) and fault detection rates (FDR).
Highlights
Fault detection and diagnosis play very important roles in the monitoring of industrial processes and have received increasing attention in recent years
Taking Tennessee Eastman Process (TEP) data as an example, this paper compares the usefulness of the two different fitness functions in Chaotic Particle Swarm Optimization (CPSO)-kernel independent component analysis (KICA)
detection latency (DL) = L − l in the calculation of the fault detection rates (FDR), N is the total number of faulty samples within the interval of analysis, and n is the number of correctly detected faulty samples
Summary
Fault detection and diagnosis play very important roles in the monitoring of industrial processes and have received increasing attention in recent years. Fault detection is the first step to realize final diagnosis of various faults and can give alarms immediately when faults happen at the early stage. Many methods, such as model-based, knowledge-based and data-driven methods, have been proposed to detect faults [1,2]. The data-driven methods have been widely used in the modern industrial processes, and models and empirical knowledge may not always be required in some cases. Among data-driven methods, PCA, ICA and PLS have received considerable attention recently
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
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
