fault detection

Gears are a very important part of mechanical equipment in industry. Due to the fact that in mechanical processes, the teeth are subjected to long-term load, the surface of their teeth is usually rusty, worn out and even broken. Timely fault detection cannot only increase the life cycle of the gears, however it can even prevent property losses and losses due to breakdowns. Therefore, it is necessary to monitor and diagnose the health of the gears to ensure the normal operation of the invaluable machines in industry. In this research, fault detection in polymer gears using audio signal is considered as a non-contact inspection method. Sound signals were recorded from 50 pairs of gears in normal condition, worn teeth and broken teeth at two speeds of 66 and 99 rpm. In the following, using wavelet packet transformation (WPT), the sound signal is analyzed in the time-frequency domain and 12 statistical features are extracted from the 16 coefficients of the fourth level of WPT. In order to study the performance of the fault detection algorithm, four classifications of linear discriminant analysis, K-nearest neighbor, decision tree and support vector machine have been used. The values of accuracy, true positive rate, true negative rate, positive predictive value, negative predictive value, geometric-mean, F1 score, and Matthews correlation coefficient have shown that by using WPT, a significant distinction can be found between normal and faulty gears. Therefore, the proposed method is a suitable approach for timely error detection of polymer gears used in mechanical equipment.

System faults, usually lead to changes in critical system parameters or even system dynamics, may lead to reduced performance and unsafe operating conditions. Fault detection plays an important role in ensuring system safety and reliability for unmanned aerial vehicles. Artificial neural networks have a good potential to detect and isolate errors in complex processes. In this paper, an observer based on adaptive neural network is presented. In this study, the adaptive neural network is designed as an intelligent learning system to detect and isolate sensor and actuator error in a nonlinear dynamic model of an unmanned aerial vehicle. Due to the nonlinearity of the system, the weighting parameters of the neural network are updated using the extended Kalman filter, which increases the convergence rate of the neural network. A set of sudden and intermittent faults is applied to a nonlinear dynamic model of a tilting multirotor to evaluate the method. Due to the high rate of updating the neural network weightings, the proposed method is able to detect sudden and intermittent faults with appropriate accuracy and speed. Numerical simulation results are also given to show the performance of the proposed method. which shows the proper performance of this design.

In this article, a method for designing a fault-tolerant optimal attitude tracking control (FTOATC) for a quadrotor UAV subject to component and actuator faults is presented. The proposed fault-tolerant method is based on safe reinforcement learning (SRL) and is capable of ensuring input and state constraints without the need for prior knowledge of the quadrotor dynamics. To this end, the proposed optimal method is presented with a dual neural network (NN) structure consisting of identifier-critic neural networks. In the identifier NN update law, in addition to considering the variable forgetting factor dependent on measurement noise, the experience response method is used, which increases convergence speed and robustness to measurement noise and reduces estimation error. In this method, solving the constrained FTOATC problem is equivalent to solving an unconstrained optimal stabilization problem for an augmented system, where control input constraints and states are guaranteed by selecting suitable cost functions on the input signal and appropriate control barrier functions (CBF)on the states, respectively. Furthermore, fault detection is performed without the need for any model or filter bank, simply by comparing the residual value of the Hamilton-Jacobi-Bellman (HJB) equation with a predetermined threshold. The Uniformly Ultimately Boundedness (UUB) of identifier and critic NN weight errors and, as a result, the convergence of the control input to the neighborhood of the optimal solution are all proved by Lyapunov theory and the performance of the method is validated through simulation results.

In this research, parameter estimation and fault detection in hydraulic pump which is one of the most common systems used in sensitive industries such as aerospace and it has nonlinear behavior is investigated. The detection criterion is the residual signal calculated from the difference between the system output and an unknown input parameter. The reason for using the nonlinear observer is the timely detection of faults and its resistance to model disturbances and uncertainties. In this research, the leakage in the pump and the change of swashplate angle of the hydraulic pump are considered as faults and they could be detected by a nonlinear observer in time. Finally, it is proved by Matlab simulation software that the proposed method is able to estimate the system states well and reveal the faults.

A fault tolerant control is designed for a four-legged bird. This control consists of two parts: status control and position control. Position control is performed by a fuzzy controller and position control is performed by a PD controller. In the next step, the parity space method is implemented in order to diagnose and estimate the defect. In addition to finding the analytical relationship of the defect, in this article, position control will also be performed. Operator saturation and its effects on system performance is another issue not seen in reviewed articles. In this paper, operator saturation is considered and consequently the maximum fault tolerable by the system will be estimated. Using fault signals, control redesign operations have been performed to tolerate faults and correct control inputs; So using the PID controller, the annoying outputs created by the fault are zeroed. The defect is assumed to be of the operative type and two scenarios have been assumed for its occurrence.

Inertial navigation systems are one of the main components in advanced equipment navigation systems such as drones and satellites. Therefore, their reliability is very important for system mission success. For increasing the reliability while developing advanced sensors, the issue of sensor placement and arrangement, the use of redundancy, and error detection should also be considered. This paper evaluates different sensor configurations and reliability analyses of fault detection and isolation of an inertial navigation system. First, the design of the navigation system is analyzed in terms of the location of the sensors, then the detection and fault detection unit based on the excess sensors. The system's reliability is calculated based on exponential distribution and reliability block diagram, and then the reliability fault detection unit is calculated using the Monte Carlo method. The sensitivity analysis has been performed, and the results show that the reliability depends on the noise value. Because the reliability of this system is a function of the fault detection and its threshold values, the optimal values for fault detection threshold are obtained using two iterative methods and estimating the minimum nonlinear squares.

Intelligent fault detection diagnosis methods are one of the common topics in recently investigations. In this paper, a new hybrid technique is presented based on discrete wavelet transform (DWT) and multi – class support vector machine (Multi-SVM). The considered vibrational signals are collected in three conditions: normal, chipped tooth and worn teeth. These signals are decomposed using DWT methods with different wavelet base functions and the most appropriate level of decomposition are selected by the cross - correlation concept. The feature vector for each sample is extracted using different time domain statistical functions. «One – against - one» support vector machine (SVM-OAO) is utilized for detecting the gearbox conditions. The condition recognition of a gearbox is depended on the extracted features type and setting the SVM parameters. Therefore, in this study, particle swarm optimization (PSO) is used for identifying the most sensitive features to the defect and its type and determining the optimal parameters of SVM method. The obtained results show that the identification accuracy of the gearbox conditions is significantly increased with improving the feature matrix and the SVM classifier method.

Using mathematical and databased modeling for designing navigation subsystems has some disadvantages such as the presence of uncertainty in the parameter's identification. In this paper, the dynamic Mathematical model of the attitude channels of the satellite is investigated.For this purpose, second-order sliding mode theory is employed to estimate external perturbation torques acting on an Earth-orbiting satellite. Then, due to the repetition of external perturbation torques in each orbital period, grey box modeling is proposed by applying estimated external perturbation torques to attitude mathematical modeling, attitude rate can be derived simultaneously. Next, to design fault detection and isolation subsystems, a proposed soft sensor is employed to generate the estimation residual as an indicator of predefined navigation faults. A nonlinear model of the earth-orbiting satellite is simulated using specific navigation failures. The results verified the feasibility of the proposed system. The simulation results show an improvement in the accuracy of the navigation data instead of the addition of a new hardware sensor.

This paper deals with the problem of fault detection and isolation for discrete time-varying systems with stochastic and bounded uncertainties, and in presence of noises in the plant and sensors. Faults can occur simultaneously or sequentially, so the designed filter has the ability to detect and isolate these faults, and handle the challenges posed by uncertainty and the effects of noises. In solving the problem of fault diagnosis, fault detection and isolation filter based on the robust Kalman filter are presented. For this purpose, a time-varying threshold is defined based on the upper bound of covariance of the residuals. This threshold helps in better performance and prevents misdiagnosis. In the design of the fault detector, due to the number of outputs, fault detectors are designed. Moreover, by examining the residuals of the system, some conditions are obtained, which, by applying these conditions, a robust fault isolator is achieved. Finally, using three examples, the efficiency and performance of the proposed method are shown. In the first example, the performance of the proposed method is studied in the presence of uncertainty and noise, and in the second and third examples, the performance of the method is compared with other methods and the superiority of the proposed approach in the presence of uncertainties is shown.

To minimize the cost of maintenance and repair of rotating industrial equipment, one of the methods used is condition monitoring by sound analysis. This study was performed to diagnose the fault of a single-phase electric motor through machine learning method aiming to monitor its situation by sound analysis. Test conditions included healthy state, bearing failure, shaft imbalance and shaft wear at two speeds of 500 and 1400 rpm. A microphone was installed on the electric motor to record data. After data acquisition, signal processing and statistical analysis, the best characteristics were selected by PCA method and then the data were clustered by machine learning method and K mean algorithm. These features used in the ANFIS modeling process were common features selected in both electromotor speed situations. After evaluating the models, the best model had the highest accuracy value of 96.82%. The average accuracy was 96.71% for overall fault classification. The results showed that the analysis of acoustic signals and modeling process can be used to diagnose electromotor defects by machine learning method. Based on the obtained results, condition monitoring of the electromotor through acoustic analysis reduces its stop and continues its work process in the industry. The repair costs of the electromotor are reduced by its proper condition monitoring.