Abstract
Anomaly detection for discrete manufacturing systems is important in intelligent manufacturing. In this paper, we address the problem of anomaly detection for the discrete manufacturing systems with complicated processes, including parallel processes, loop processes, and/or parallel with nested loop sub-processes. Such systems can generate a series of discrete event data during normal operations. Existing methods that deal with the discrete sequence data may not be efficient for the discrete manufacturing systems or methods that are dealing with manufacturing systems only focus on some specific systems. In this paper, we take the middle way and seek to propose an efficient algorithm by applying only the system structure information. Motivated by the system structure information that the loop processes may result in repeated events, we propose two algorithms—centralized pattern relation table algorithm and parallel pattern relation table algorithm—to build one or multiple relation tables between loop pattern elements and individual events. The effectiveness of the proposed algorithms is tested by two artificial data sets that are generated by Timed Petri Nets. The experimental results show that the proposed algorithms can achieve higher AUC and F1-score, even with smaller sized data set compared to the other algorithms and that the parallel algorithm achieves the highest performance with the smallest data set.
Highlights
Detection is a hot topic in various domains, such as manufacturing anomaly detection [1,2], cyber attack detection [3,4,5,6], and crowded scene video anomaly detection [7,8].Cyber attacks detection typically detects three types of external attacks, i.e., false data injection attack, denial-of-service attack, and confidentiality attack
This paper seeks to deal with anomaly detection for discrete event data that result from discrete manufacturing systems
The parallel pattern relation table algorithm in Algorithm 2 has the same work flow as in in this algorithm, we take a different point of view of the loops in the data set, i.e., when the loop process happens, there must exist some events that will be repeated and the combinations of repeated events show the patterns of the loops
Summary
Detection is a hot topic in various domains, such as manufacturing anomaly detection [1,2], cyber attack detection [3,4,5,6], and crowded scene video anomaly detection [7,8]. This paper seeks to deal with anomaly detection for discrete event data that result from discrete manufacturing systems. This paper will take these two features into account when designing the anomaly detection algorithms, and such feature-based algorithms facilitate us to reach good performance at low costs: required small-sized data. Kernel-based methods are one of the similarity-based algorithms, which detect anomalies by comparing the similarity between the testing sequences and the training sequences These algorithms are tolerant of false negatives, which may result in huge loss when considering the fact that customs may lose their money once ATM has a problem and manufacturing companies may suffer great losses once severe abnormal behaviors occur. We proposed a centralized pattern relation-based algorithm (Algorithm 1) and a parallel pattern relation-based algorithm (Algorithm 2) to detect anomalies in manufacturing systems, especially for parallel processes and loop processes.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
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
