A relevance vector machine (RVM) based fault diagnosis method was presented for non-linear circuits. In order to simplify RVM classifier, parameters selection based on particle swarm optimization (PSO) and preprocessi...A relevance vector machine (RVM) based fault diagnosis method was presented for non-linear circuits. In order to simplify RVM classifier, parameters selection based on particle swarm optimization (PSO) and preprocessing technique based on the kurtosis and entropy of signals were used. Firstly, sinusoidal inputs with different frequencies were applied to the circuit under test (CUT). Then, the resulting frequency responses were sampled to generate features. The frequency response was sampled to compute its kurtosis and entropy, which can show the information capacity of signal. By analyzing the output signals, the proposed method can detect and identify faulty components in circuits. The results indicate that the fault classes can be classified correctly for at least 99% of the test data in example circuit. And the proposed method can diagnose hard and soft faults.展开更多
To detect the bias fault in stochastic non-linear dynamic systems, a new Unscented Kalman Filtering（UKF） based real-time recursion detection method is brought forward with the consideration of the flaws of tradition...To detect the bias fault in stochastic non-linear dynamic systems, a new Unscented Kalman Filtering（UKF） based real-time recursion detection method is brought forward with the consideration of the flaws of traditional Extended Kalman Filtering（ EKF）. It uses the UKF as the residual generation method and the Weighted-Sum Squared Residual （WSSR） as the fault detection strategy. The simulation results are provided which demonstrate better effectiveness and a higher detection ratio of the developed methods.展开更多
A fault diagnosis method of bearing based on integration of non-linear geometric invariables was presented for the non-linearity exiting in bearing system but ignored in traditional fault diagnosis.The meanings of non...A fault diagnosis method of bearing based on integration of non-linear geometric invariables was presented for the non-linearity exiting in bearing system but ignored in traditional fault diagnosis.The meanings of non-linear geometric invariables,such as fractal dimension,Lyapunov exponent,Kolmogorov entropy,correlation distance entropy and their calculation method were analyzed.Grey theory is applied to integrate these parameters and the correlation values as fault characteristic value was input into the support vector machines for diagnosis.The experimental results show that this method can distinguish the bearing fault effectively,it provides a new approach for the fault diagnosis of rotating machinery.展开更多
Although lots of valuable results for fault diagnosis based on model have been achieved in linear system, it is difficult to apply these results to non-linear system due to the difficulty of modeling the non-linear sy...Although lots of valuable results for fault diagnosis based on model have been achieved in linear system, it is difficult to apply these results to non-linear system due to the difficulty of modeling the non-linear system by analysis. Adaptive Fuzzy system provides a way for solving this problem because it can approximate any non-linear system at any accuracy. The key for adaptive Fuzzy system to solve problem is its learning ability, so the authors present a learning algorithm for Adaptive fuzzy system, which can build the system's model by learning from the measurement data as well as experience knowledge with high accuracy. Furthermore, the experiment using the learning algorithm to model a servo-mechanism and to construct the fault diagnosis system based on the model is carried out, the results is very good.展开更多
基金Project(Z132012)supported by the Second Five Technology-based in Science and Industry Bureau of ChinaProject(YWF1103Q062)supported by the Fundemental Research Funds for the Central Universities in China
文摘A relevance vector machine (RVM) based fault diagnosis method was presented for non-linear circuits. In order to simplify RVM classifier, parameters selection based on particle swarm optimization (PSO) and preprocessing technique based on the kurtosis and entropy of signals were used. Firstly, sinusoidal inputs with different frequencies were applied to the circuit under test (CUT). Then, the resulting frequency responses were sampled to generate features. The frequency response was sampled to compute its kurtosis and entropy, which can show the information capacity of signal. By analyzing the output signals, the proposed method can detect and identify faulty components in circuits. The results indicate that the fault classes can be classified correctly for at least 99% of the test data in example circuit. And the proposed method can diagnose hard and soft faults.
文摘To detect the bias fault in stochastic non-linear dynamic systems, a new Unscented Kalman Filtering（UKF） based real-time recursion detection method is brought forward with the consideration of the flaws of traditional Extended Kalman Filtering（ EKF）. It uses the UKF as the residual generation method and the Weighted-Sum Squared Residual （WSSR） as the fault detection strategy. The simulation results are provided which demonstrate better effectiveness and a higher detection ratio of the developed methods.
文摘A fault diagnosis method of bearing based on integration of non-linear geometric invariables was presented for the non-linearity exiting in bearing system but ignored in traditional fault diagnosis.The meanings of non-linear geometric invariables,such as fractal dimension,Lyapunov exponent,Kolmogorov entropy,correlation distance entropy and their calculation method were analyzed.Grey theory is applied to integrate these parameters and the correlation values as fault characteristic value was input into the support vector machines for diagnosis.The experimental results show that this method can distinguish the bearing fault effectively,it provides a new approach for the fault diagnosis of rotating machinery.
文摘Although lots of valuable results for fault diagnosis based on model have been achieved in linear system, it is difficult to apply these results to non-linear system due to the difficulty of modeling the non-linear system by analysis. Adaptive Fuzzy system provides a way for solving this problem because it can approximate any non-linear system at any accuracy. The key for adaptive Fuzzy system to solve problem is its learning ability, so the authors present a learning algorithm for Adaptive fuzzy system, which can build the system's model by learning from the measurement data as well as experience knowledge with high accuracy. Furthermore, the experiment using the learning algorithm to model a servo-mechanism and to construct the fault diagnosis system based on the model is carried out, the results is very good.
