Abstract—Fault detection and classification is critical to the reliability of modern control systems in different industries, where detecting and classifying faults in operational processes are very important things while failure to detect and classify them, may cause irreparable damage. In this paper, fault detection and classification approach is presented. The first step, multi stage recursive least squares parameter estimation approach for controlled autoregressive autoregressive moving average systems (CARARMA) is developed with a view to estimate the parameters of the system, additionally, improve the effectiveness of the computation. By means of multi stage approach, the (CARARMA) system is decomposed into three simple identification models, and the parameters of each simple model is identified one by one. These parameters estimated by this approach are referred to as features, and not all of them have the same useful data about the system. In order to select the valuable features and improve a classification accuracy, the Linear Discriminant Analysis (LDA) approach based on scattering matrices is applied for dimension reduction. The classification between these reduced classes is done based on the Naive Bayes classifier. Finally, the obtained results explain the performance of this proposed approach.

The application of feature selection to fault diagnosis is presented. First, Filtered data identification algorithm is derived for controlled autoregressive autoregressive (CARAR) system in order to estimate the system parameters. The Proposed technique offers a high computational efficiency. These estimated parameters are referred as features and these features are not all have the same informative value. Next, features selection is carried out using principal components analysis. Finally, the simulation results demonstrate the value of the suggested procedures.

this paper presents feature selection for fault diagnosis based on recursive generalized extended least squares algorithm (RGELS). RGELS model is derived and validation of this model is tested utilizing good statistical methods, which, namely best-fit criterion. The system parameters are estimated employing the proposed algorithm. Dimension reduction of the system parameters is done to get best important features using linear discriminant analysis. Finally, the simulation results confirm the effectiveness of the algorithm.

Automobiles of today must constantly change their speeds in reaction to changing road and traffic circumstances as the pace and density of road traffic increases. In sophisticated automobiles, the Anti-lock Braking System (ABS) is a vehicle safety system that enhances the vehicle's stability and steering capabilities by varying the torque to maintain the slip ratio at a safe level. This paper analyzes the performance of classical control, model reference adaptive control (MRAC), and intelligent control for controlling the (ABS). The ABS controller's goal is to keep the wheel slip ratio, which includes nonlinearities, parametric uncertainties, and disturbances as close to an optimal slip value as possible. This will decrease the stopping distance and guarantee safe vehicle operation during braking. A Bang-bang controller, PID, PID based Model Reference Adaptive Control (PID-MRAD), Fuzzy Logic Control (FLC), and Adaptive Neuro-Fuzzy Inference System (ANFIS) controller are used to control the vehicle model. The car was tested on a dry asphalt and ice road with only straight-line braking. Based on slip ratio, vehicle speed, angular velocity, and stopping time, comparisons are performed between all control strategies. To analyze braking characteristics, the simulation changes the road surface condition, vehicle weight, and control methods. The simulation results revealed that our objectives were met. The simulation results clearly show that the ANFIS provides more flexibility and improves system-tracking precision in control action compared to the Bang-bang, PID, PID-MRAC, and FLC.

In this paper, the data filtering based Recursive Least Squares algorithm (RLS) of linear Box–Jenkins systems is proposed for fault detection. The system is decomposed into two subsystems, one containing the parameters of the system model and the other containing the parameters of the noise model to improve the computational load. In addition, the parameters of the system model and the noise model are estimated and their accuracy are developed. The model validation is tested using two statistical methods, histogram and mean square errors. The residual is generated based on the proposed algorithm to design the threshold and therefore, this designed thresholdis used for fault detection. Simulation results are performed to illustrate the algorithm performance.

this paper derives extended three-stage recursive identification algorithm of MISO for (CARARMA) systems. Based on The decomposition technique, four subsystems are obtained and the parameters of each subsystem are identified. Some model validation methods are computed to measure the model value and Akaike’s Final Prediction Error Criterion (FPE) is used to verify the selection of system order. The algorithm has a high computational efficiency because the covariance matrices dimensions become small in each subsystem. Finally, this algorithm effectiveness is demonstrated in simulation example.