Sinopsis
The diagnostics of industrial processes is a scientific discipline aimed at the detection of faults in industrial plants, their isolation, and finally their identification. Its main task is the diagnosis of process anomalies and faults in process components, sensors and actuators. Early diagnosis of faults that might occur in the supervised process renders it possible to perform important preventing actions. Moreover, it allows one to avoid heavy economic losses involved in stopped production, the replacement of elements and parts, etc.
Most of the methods in the fault diagnosis literature are based on linear methodology or exact models. Industrial processes are often difficult to model. They are complex and not exactly known, measurements are corrupted by noise and unreliable sensors. Therefore, a number of researchers have perceived artificial neural networks as an alternative way to represent knowledge about faults [1, 2, 3, 4, 5, 6, 7]. Neural networks can filter out noise and disturbances, they can provide stable, highly sensitive and economic diagnostics of faults without traditional types of models. Another desirable feature of neural networks is that no exact models are required to reach the decision stage [2]. In a typical operation, the process model may be only approximate and the critical measurements may be able to map internally the functional relationships that represent the process, filter out the noise, and handle correlations as well. Although there are many promising simulation examples of neural networks in fault diagnosis in the literature, real applications are still quite rare. There is a great necessity to conduct more detailed scientific investigations concerning the application of neural networks in real industrial plants, to achieve complete utilization of their attractive features.
One of the most frequently used schemes for fault diagnosis is the model based concept. The basic idea of model based fault diagnosis is to generate signals that reflect inconsistencies between nominal and faulty system operation conditions [8, 9, 10, 11]. Such signals, called residuals, are usually calculated by using analytical methods such as observers [9, 12], parameter estimation methods [13, 14] or parity equations [15, 16]. Unfortunately, the common drawback of these approaches is that an accurate mathematical model of the diagnosed plant
Content
- Introduction
- Modelling Issue in Fault Diagnosis
- Locally Recurrent Neural Networks
- Approximation Abilities of Locally Recurrent Networks
- Stability and Stabilization of Locally Recurrent Networks
- Optimum Experimental Design for Locally Recurrent Networks
- Decision Making in Fault Detection
- Industrial Applications
.jpg)
0 komentar:
Posting Komentar