Neural networks applied for the identification and fault diagnosis of process valves and actuators

Conditioning Monitoring and Fault Diagnosis for a Servo-Pneumatic System with Artificial Neural Network Algorithms

“…The BP estimates the weights of the neurons by updating them after the forward and backward propagation of error. The learning rate and the momentum are two important parameters of the BP for training the network successfully (Chen & Mo, 2004;McGhee et al, 1997). Levenberg-Marquardt algorithm (Beale et al, 2010) generally estimates the parameters of the FFNNs.…”

“…Acoustic emission is an excellent example of high frequency monitoring signal (Yang, 2006;Chena et al, 2007). The frequency of the pressure, flow, and timing signals are low (Sepasi & Sassani, 2010;Nogami et al, 1995;Bouamama et al, 2005;Nakutis & Kaškonas, 2005Wang et al, 2004;Karpenko et al, 2003;Li & Kao, 2005;McGhee et al, 1997). The gathered signals are encoded to obtain their most descriptive features.…”

“…The gathered signals are encoded to obtain their most descriptive features. The encoded signals were classified by using various classification techniques such as ANNs (Karpenko et al, 2003;Nakutis & Kaškonas, 2003;Sepasi & Sassani, 2010;Nogami et al, 1995;McGhee et al, 1997), fuzzy method (Mendonca et al, 2009;, neuro fuzzy method (Shi & Sepehri, 2004;, statistical technique (Song et al, 2003), bond graphs (Bouamama et al, 2005), genetic programming (Wang et al, 2004;Yang, 2006), and expert/intelligent systems (Chen & Mo, 2004). It is not difficult to develop programs for classification of the sensory signals of pneumatic systems.…”

Fault diagnosis of pneumatic systems with artificial neural network algorithms

“…Otherwise, a new category is created. The most important task is the selection of the vigilance of the neural network to create the minimum number of categories without classifying normal and faulty cases in the same category (McGhee, Henderson, & Baird, 1997;Yang & Han, 2004).…”

Artificial Neural Network Approach for Fault Detection in Pneumatic Valve in Cooler Water Spray System

“…Another trend is towards intelligent FDD. McGhee et al [3] used a backward propagation artificial neural network to model a process valve actuator. The estimate of the torque by the neural network was compared to the actual measured torque from a torque-measuring device for the FDD.…”