Skip to Main Content
Fault diagnosis in pneumatic actuators is a very difficult task due to the inherent high nonlinearity and uncertainty. Developing models of nonlinear systems with adaptive network-based fuzzy inference systems (ANFISs) has recently received attention. Models that are built upon ANFISs overcome the disadvantages of ordinary fuzzy modeling and can be very suitable for generalized modeling of nonlinear plants. We set up a group of ANFIS models which are relatively common in practice, corresponding to various situations of a pneumatic actuator, including normal, low and high supply pressure. Considering the advantage that a learning vector quantization (LVQ) neural network has a powerful ability to classification, we then utilize a LVQ neural network as a fault diagnosis scheme by abstracting the data of ANFIS models as the input vectors for nonlinear plants. The effectiveness is demonstrated via experiments on a pneumatic actuator.