An intelligent Online monitoring and diagnostic system for manufacturing automation

Condition monitoring and fault diagnosis in modern manufacturing automation is of great practical significance. It improves quality and productivity, and prevents damage to machinery. In general, this practice consists of two parts: 1) extracting appropriate features from sensor signals and 2) recognizing possible faulty patterns from the features. Through introducing the concept of marginal energy in signal processing, a new feature representation is developed in this paper. In order to cope with the complex manufacturing operations, three approaches are proposed to develop a feasible system for online applications. This paper develops intelligent learning algorithms using hidden Markov models and the newly developed support vector techniques to model manufacturing operations. The algorithms have been coded in modular architecture and hierarchical architecture for the recognition of multiple faulty conditions. We define a novel similarity measure criterion for the comparison of signal patterns which will be incorporated into a novel condition monitoring system. The sensor-based intelligent system has been implemented in stamping operations as an example. We demonstrate that the proposed method is substantially more effective than the previous approaches. Its unique features benefit various real-world manufacturing automation engineering, and it has great potential for shop floor applications. Note to Practitioners-This paper was motivated by the problem of detecting the complicated sheet-metal processes faults in a high success rate but also applies to the real-world practices. Existing systems adopt traditional approaches fast but cannot adapt to various cases and achieve a diagnosis purpose. Using hidden Markov models and the newly developed support vector techniques, this paper proposes three approaches and develops intelligent learning algorithms to model manufacturing operations. The novel similarity measure criterion is able to implement the signal patterns comparison, overcoming the time-shifting phenomenon. The approach of modular architecture and hierarchical architecture has been presented in the software coding for the recognition of multiple faulty conditions. The sensor-based hardware design facilitates the online condition monitoring system. In this paper, we mathematically characterize the process signal features. We then develop the HMM-based modular fault diagnosis system which makes the modification expansion easy in practice. The hierarchical SVM-based fault diagnosis system uses a small-size training sample but achieves a high successful rate, and overcomes the overfitting and difficulties in structure design of the learning machine. The similarity measure approach using support vector techniques requires training sample sets produced under normal operating conditions and, thus, makes it more feasible than other techniques. Meanwhile, this approach overcomes the time-shifting. The proposed system provides important technologies and valuable reference for research areas and an innovation to manufacturing automation. The system works well in most applications; however, in some cases, certain parameters have to be tuned manually, for example, the parts of the signals to be modeled. The next step is to develop the system so it can learn human skills and set parameters automatically. In the short term, it is important to research ways to improve the system's intelligent online learning.