A new sensor fault diagnosis method based on structured kernel principal component analysis （KPCA） is proposed for nonlinear processes. By performing KPCA on subsets of variables, a set of structured residuals, i.e....A new sensor fault diagnosis method based on structured kernel principal component analysis （KPCA） is proposed for nonlinear processes. By performing KPCA on subsets of variables, a set of structured residuals, i.e., scaled powers of KPCA, can be obtained in the same way as partial PCA. The structured residuals are utilized in composing an isolation scheme for sensor fault diagnosis, according to a properly designed incidence matrix. Sensor fault sensitivity and critical sensitivity are defined, based on which an incidence matrix optimization algorithm is proposed to improve the performance of the structured KPCA. The effectiveness of the proposed method is demonstrated on the simulated continuous stirred tank reactor （CSTR） process.展开更多
In this paper, a robust sensor fault diagnosis observer with non-singular structure is proposed for a class of linear sampled-data descriptor system with state time-vary delay. Firstly, a sampled-data descriptor model...In this paper, a robust sensor fault diagnosis observer with non-singular structure is proposed for a class of linear sampled-data descriptor system with state time-vary delay. Firstly, a sampled-data descriptor model with time-vary delay is proposed and transformed into a discrete-time non-singular one. Then, a robust sensor fault diagnosis observer is proposed based on the state estimation error and the measurement residual, this observer can guarantee the robustness of the residual against the augmented disturbance and the sensor fault, which means the H∞ performance index is satisfied. As the confining matrix of the designed observer parameters does not meet the Linear Matrix Inequality (LMI), a cone complementary linearization (CCL) algorithm is proposed to solve this problem. The decision logic of the residual is obtained by the residual evaluation function. Simulation results show the effectiveness of the method.展开更多
As the energy crisis intensifies,the organic Rankine cycle(ORC)is increasingly employed for efficient recovery of low-temperature waste heat.The operation of the ORC system necessitates the use of numerous sensors to ...As the energy crisis intensifies,the organic Rankine cycle(ORC)is increasingly employed for efficient recovery of low-temperature waste heat.The operation of the ORC system necessitates the use of numerous sensors to monitor its status.Over time,these sensors may become faulty,rendering accurate and timely diagnosis is critical for proper and safe functioning of the ORC system.Currently,there is a lack of rapid diagnostic methods for sensor faults in ORC systems.This study establishes an ORC test bench utilizing cyclopentane as the working fluid.Experimental data incorporating induced faults from the ORC test bench is employed to train machine learning-based models for sensor fault diagnosis.The test results indicate that the diagnostic model developed in this study can accurately diagnose various sensor faults in the ORC system,thereby ensuring its safe operation.Notably,the method based on Bayesian-optimized long short-term memory network(BO-LSTM)achieved the highest diagnostic accuracy,reaching up to 95.92%.展开更多
This paper addresses the gas path component and sensor fault diagnosis and isolation(FDI) for the auxiliary power unit(APU). A nonlinear dynamic model and a distributed state estimator are combined for the distributed...This paper addresses the gas path component and sensor fault diagnosis and isolation(FDI) for the auxiliary power unit(APU). A nonlinear dynamic model and a distributed state estimator are combined for the distributed control system. The distributed extended Kalman filter(DEKF)is served as a state estimator,which is utilized to estimate the gas path components’ flow capacity. The DEKF includes one main filter and five sub-filter groups related to five sensors of APU and each sub-filter yields local state flow capacity. The main filter collects and fuses the local state information,and then the state estimations are feedback to the sub-filters. The packet loss model is introduced in the DEKF algorithm in the APU distributed control architecture. FDI strategy with a performance index named weight sum of squared residuals(WSSR) is designed and used to identify the APU sensor fault by removing one sub-filter each time. The very sensor fault occurs as its performance index WSSR is different from the remaining sub-filter combinations. And the estimated value of the soft redundancy replaces the fault sensor measurement to isolate the fault measurement. It is worth noting that the proposed approach serves for not only the sensor failure but also the hybrid fault issue of APU gas path components and sensors. The simulation and comparison are systematically carried out by using the APU test data,and the superiority of the proposed methodology is verified.展开更多
Batterymanagement systems(BMSs)are essential in ensuring the safe and stable operation of lithium-ion batteries(LIBs)in electric vehicles(EVs).Accurate sensor signals,particularly voltage,current,and temperature senso...Batterymanagement systems(BMSs)are essential in ensuring the safe and stable operation of lithium-ion batteries(LIBs)in electric vehicles(EVs).Accurate sensor signals,particularly voltage,current,and temperature sensor signals,are essential for a BMS to performfunctions such as state estimation,balance control,and fault diagnosis.The smooth operation of a BMS depends primarily on sensor signals,which provide current,voltage,and temperature information to maintain the battery pack in a safe running state.However,sensor failures and inaccurate measurement data can easily occur because of external interference and complex operating conditions.Therefore,an investigation into the faultdiagnosis of battery sensors and fault-tolerant control(FTC)is necessary to ensure the normal operation of a BMS.This paper analyzes themodes of sensor faults,fault diagnosismethods,and fault-tolerant control techniques.First,the differentmodes of sensor faults are analyzed,andmathematical expressions for these faults are provided.Second,diagnostic methods for sensor faults based on models,signal processing,and data-drivenmethods are analyzed in detail.Finally,FTC techniques are introduced to ensure stable sensor operation.Based on an analysis of the research status of sensor fault diagnosis,a new development direction for sensor fault diagnosis is proposed.展开更多
According to fault type diversity and fault information uncertainty problem of the hydraulic driven rocket launcher servo system（HDRLSS） , the fault diagnosis method based on the evidence theory and neural network e...According to fault type diversity and fault information uncertainty problem of the hydraulic driven rocket launcher servo system（HDRLSS） , the fault diagnosis method based on the evidence theory and neural network ensemble is proposed. In order to overcome the shortcomings of the single neural network, two improved neural network models are set up at the com-mon nodes to simplify the network structure. The initial fault diagnosis is based on the iron spectrum data and the pressure, flow and temperature（PFT） characteristic parameters as the input vectors of the two improved neural network models, and the diagnosis result is taken as the basic probability distribution of the evidence theory. Then the objectivity of assignment is real-ized. The initial diagnosis results of two improved neural networks are fused by D-S evidence theory. The experimental results show that this method can avoid the misdiagnosis of neural network recognition and improve the accuracy of the fault diagnosis of HDRLSS.展开更多
基金supported by Scientific Reserch Fund of SiChuan Provincial Education Department (No.07ZB013)by the Scientific ResearchFoundation of CUIT (No.CSRF200704)
文摘A new sensor fault diagnosis method based on structured kernel principal component analysis （KPCA） is proposed for nonlinear processes. By performing KPCA on subsets of variables, a set of structured residuals, i.e., scaled powers of KPCA, can be obtained in the same way as partial PCA. The structured residuals are utilized in composing an isolation scheme for sensor fault diagnosis, according to a properly designed incidence matrix. Sensor fault sensitivity and critical sensitivity are defined, based on which an incidence matrix optimization algorithm is proposed to improve the performance of the structured KPCA. The effectiveness of the proposed method is demonstrated on the simulated continuous stirred tank reactor （CSTR） process.
基金Sponsored by the National Natural Science Foundation of China(Grant No.61021002)
文摘In this paper, a robust sensor fault diagnosis observer with non-singular structure is proposed for a class of linear sampled-data descriptor system with state time-vary delay. Firstly, a sampled-data descriptor model with time-vary delay is proposed and transformed into a discrete-time non-singular one. Then, a robust sensor fault diagnosis observer is proposed based on the state estimation error and the measurement residual, this observer can guarantee the robustness of the residual against the augmented disturbance and the sensor fault, which means the H∞ performance index is satisfied. As the confining matrix of the designed observer parameters does not meet the Linear Matrix Inequality (LMI), a cone complementary linearization (CCL) algorithm is proposed to solve this problem. The decision logic of the residual is obtained by the residual evaluation function. Simulation results show the effectiveness of the method.
基金supported by the National Key R&D Program of China(2024YFE0213200)supported by the Tianjin Municipal Science and Technology Bureau(23jCjQjC00260).
文摘As the energy crisis intensifies,the organic Rankine cycle(ORC)is increasingly employed for efficient recovery of low-temperature waste heat.The operation of the ORC system necessitates the use of numerous sensors to monitor its status.Over time,these sensors may become faulty,rendering accurate and timely diagnosis is critical for proper and safe functioning of the ORC system.Currently,there is a lack of rapid diagnostic methods for sensor faults in ORC systems.This study establishes an ORC test bench utilizing cyclopentane as the working fluid.Experimental data incorporating induced faults from the ORC test bench is employed to train machine learning-based models for sensor fault diagnosis.The test results indicate that the diagnostic model developed in this study can accurately diagnose various sensor faults in the ORC system,thereby ensuring its safe operation.Notably,the method based on Bayesian-optimized long short-term memory network(BO-LSTM)achieved the highest diagnostic accuracy,reaching up to 95.92%.
基金supported by the National Natural Science Foundation of China(No.91960110)the National Science and Technology Major Project(No. 2017-I0006-0007)the Fundamental Research Funds for the Central Universities(NP2022418)。
文摘This paper addresses the gas path component and sensor fault diagnosis and isolation(FDI) for the auxiliary power unit(APU). A nonlinear dynamic model and a distributed state estimator are combined for the distributed control system. The distributed extended Kalman filter(DEKF)is served as a state estimator,which is utilized to estimate the gas path components’ flow capacity. The DEKF includes one main filter and five sub-filter groups related to five sensors of APU and each sub-filter yields local state flow capacity. The main filter collects and fuses the local state information,and then the state estimations are feedback to the sub-filters. The packet loss model is introduced in the DEKF algorithm in the APU distributed control architecture. FDI strategy with a performance index named weight sum of squared residuals(WSSR) is designed and used to identify the APU sensor fault by removing one sub-filter each time. The very sensor fault occurs as its performance index WSSR is different from the remaining sub-filter combinations. And the estimated value of the soft redundancy replaces the fault sensor measurement to isolate the fault measurement. It is worth noting that the proposed approach serves for not only the sensor failure but also the hybrid fault issue of APU gas path components and sensors. The simulation and comparison are systematically carried out by using the APU test data,and the superiority of the proposed methodology is verified.
基金funded by the Key Industrial Chain Technology Research Program of Xi’an(24ZDCYJSGG0048)the Key Research and Development Program of Xianyang(L2023-ZDYF-SF-077)the Fundamental Research Funds for the Central Universities in Chang'an University(300102384201).
文摘Batterymanagement systems(BMSs)are essential in ensuring the safe and stable operation of lithium-ion batteries(LIBs)in electric vehicles(EVs).Accurate sensor signals,particularly voltage,current,and temperature sensor signals,are essential for a BMS to performfunctions such as state estimation,balance control,and fault diagnosis.The smooth operation of a BMS depends primarily on sensor signals,which provide current,voltage,and temperature information to maintain the battery pack in a safe running state.However,sensor failures and inaccurate measurement data can easily occur because of external interference and complex operating conditions.Therefore,an investigation into the faultdiagnosis of battery sensors and fault-tolerant control(FTC)is necessary to ensure the normal operation of a BMS.This paper analyzes themodes of sensor faults,fault diagnosismethods,and fault-tolerant control techniques.First,the differentmodes of sensor faults are analyzed,andmathematical expressions for these faults are provided.Second,diagnostic methods for sensor faults based on models,signal processing,and data-drivenmethods are analyzed in detail.Finally,FTC techniques are introduced to ensure stable sensor operation.Based on an analysis of the research status of sensor fault diagnosis,a new development direction for sensor fault diagnosis is proposed.
文摘According to fault type diversity and fault information uncertainty problem of the hydraulic driven rocket launcher servo system（HDRLSS） , the fault diagnosis method based on the evidence theory and neural network ensemble is proposed. In order to overcome the shortcomings of the single neural network, two improved neural network models are set up at the com-mon nodes to simplify the network structure. The initial fault diagnosis is based on the iron spectrum data and the pressure, flow and temperature（PFT） characteristic parameters as the input vectors of the two improved neural network models, and the diagnosis result is taken as the basic probability distribution of the evidence theory. Then the objectivity of assignment is real-ized. The initial diagnosis results of two improved neural networks are fused by D-S evidence theory. The experimental results show that this method can avoid the misdiagnosis of neural network recognition and improve the accuracy of the fault diagnosis of HDRLSS.
