A new approach to fault dignosis dealing with nonlinear system Hopfieldneural networks (HNN) is presented. The model parameters of the nonlinear systemtreated as functions of measured operating points and faults are e...A new approach to fault dignosis dealing with nonlinear system Hopfieldneural networks (HNN) is presented. The model parameters of the nonlinear systemtreated as functions of measured operating points and faults are estimated by HNN. Boththe nominal model of the healthy system and HNN training models corresponding to everyoperating point are recognized. In addition, the anticipated fault models corresponding toevery kind of fault and every operating point are obtaind in advance. The real systemmodel parameters of the system estimated by generalization process of HNN are matchedwith the nominal models of the healthy system and anticipated fault models. Consequent-ly, the final result of fault detection and diagnosis is acquired. The approach to fault diag-nosis is used in an aircraft actuating poisition servo system and the simulation resu1t is re-ported.展开更多
Adaptive control of servo actuator with nonlinear friction compensation is addressed. LuGre dynamic friction model is adopted to characterize the nonlinear friction and a new kind of slid ing mode observer is designe...Adaptive control of servo actuator with nonlinear friction compensation is addressed. LuGre dynamic friction model is adopted to characterize the nonlinear friction and a new kind of slid ing mode observer is designed to estimate the internal immeasurable state of LuGre model. Based on the estimated friction state, adaptive laws are designed to identify the unknown model parameters and the external disturbances, and the system stability and asymptotic trajectory tracking perform ance are guaranteed by Lyapunov function. The position tracking performance is verified by the ex perimental results.展开更多
The hydraulic servo actuator of heavy vehicle active suspension is investigated to clarify the correlation between system parameters and the control characteristics of active suspension hydraulic servo system.Accordin...The hydraulic servo actuator of heavy vehicle active suspension is investigated to clarify the correlation between system parameters and the control characteristics of active suspension hydraulic servo system.Accordingly, a nonlinear physical model of electro-hydraulic servo active suspension system is built.Compared with the conventional nonlinear modeling, the model in this study considers the asymmetry of working areas caused by single rod hydraulic cylinder in the suspension system.In accordance with the model, a nonlinear output feedback controller based on backstepping is designed, and the effectiveness of the controller is proved based on the experimental platform.The dynamic response curve of the electro-hydraulic servo control system under the change of parameters is generated based on the simulation model.The sensitivity of electro-hydraulic servo control performance to the change of system physical parameters is investigated, and two evaluation indexes are proposed to quantify and compare the effect of all physical parameter changes on position control system.As revealed by the results, the position control characteristics of suspension actuator are more sensitive to the changes of flow gain of the servo valve, system supply oil pressure and effective working areas of cylinder, and the two evaluation indexes are over 10 times higher than other physical parameters.展开更多
The servo system actuated by oscillating pneumatic cylinder for X-Y plate is a multi-variable nonlinear control system. Its mathematical model is established, and nonlinear factors are analyzed. Due to the existence o...The servo system actuated by oscillating pneumatic cylinder for X-Y plate is a multi-variable nonlinear control system. Its mathematical model is established, and nonlinear factors are analyzed. Due to the existence of deadlock zone and the small damp of the pneumatic oscillating cylinder, it is likely to result in overshoot, and there is also certain steady-state error, so online modifying of proportion-integration-differentiation （PID） parameters is needed so as to achieve better control performance. Meanwhile considering the stability demand for long-term run, a fuzzy adaptive PID controller is designed. The result of hardware-inloop （HIL） test and real-time control experiment shows that the adaptive PID controller has desirable serfadaptability and robustness to external disturbance and to change of system parameters, and its control per- fonnance is better than that of traditional PID controllers.展开更多
This article, in order to precisely impose friction on aircraft and weapon actuation systems, presents a new friction loading method characteristic of ＂torque-zero velocity＂ switching control with an electro-hydraul...This article, in order to precisely impose friction on aircraft and weapon actuation systems, presents a new friction loading method characteristic of ＂torque-zero velocity＂ switching control with an electro-hydraulic load simulator. As the general Stribeck friction model has little related to static friction, it proposes a ＂torque-zero velocity＂ switcher, in which a zero-velocity controller is developed to load the static friction and a torque controller the kinetic friction. With the help of mathematical modeling, this article designs a ＂torque-zero velocity＂ switching controller and, correspondingly, provides a ＂dual-threshold judgment＂ algorithm. Simulation results indicate that the proposed method can be successfully used to carry out the static and kinetic friction simulation with an electro-hydraulic load simulator.展开更多
This study proposes a data-driven friction modeling and compensation method aimed at solving the problem of servo performance degradation caused by friction in rotary servo actuators.First,a data-driven friction model...This study proposes a data-driven friction modeling and compensation method aimed at solving the problem of servo performance degradation caused by friction in rotary servo actuators.First,a data-driven friction modeling method is proposed on the basis of the physics-informed neural network(PINN)and the LuGre model.The constructed friction model consists of sliding regime,static regime,and presliding regime,which extends the variables of the friction model to include velocity and position.The data-driven friction model not only retains the accuracy of the LuGre model in describing the dynamic behavior of friction at zero velocity but also improves the accuracy and convergence speed of the model through the powerful learning ability of PINN,which is verified in the two examples of constructing friction test data.Second,on the basis of the data-driven friction model,a composite compensation strategy centered on friction compensation is proposed.The friction compensator is used to compensate the internal friction of the actuator,and the extended Kalman filter is used to suppress the random disturbance to achieve the precise control of the servo actuator.Experimental validation of the proposed compensation strategy against three traditional control methods demonstrates its superiority,with average improvements of 49.5%,30.4%,and 32.7%in velocity tracking accuracy,respectively,while ensuring consistent accuracy across different positions.The proposed data-driven friction modeling and compensation method provides a new perspective and method for overcoming the effect of friction.展开更多
文摘A new approach to fault dignosis dealing with nonlinear system Hopfieldneural networks (HNN) is presented. The model parameters of the nonlinear systemtreated as functions of measured operating points and faults are estimated by HNN. Boththe nominal model of the healthy system and HNN training models corresponding to everyoperating point are recognized. In addition, the anticipated fault models corresponding toevery kind of fault and every operating point are obtaind in advance. The real systemmodel parameters of the system estimated by generalization process of HNN are matchedwith the nominal models of the healthy system and anticipated fault models. Consequent-ly, the final result of fault detection and diagnosis is acquired. The approach to fault diag-nosis is used in an aircraft actuating poisition servo system and the simulation resu1t is re-ported.
基金Supported by State Key Laboratory of Explosion Science and Technology(QNKT11-08)
文摘Adaptive control of servo actuator with nonlinear friction compensation is addressed. LuGre dynamic friction model is adopted to characterize the nonlinear friction and a new kind of slid ing mode observer is designed to estimate the internal immeasurable state of LuGre model. Based on the estimated friction state, adaptive laws are designed to identify the unknown model parameters and the external disturbances, and the system stability and asymptotic trajectory tracking perform ance are guaranteed by Lyapunov function. The position tracking performance is verified by the ex perimental results.
基金Supported by the National Natural Science Foundation of China (No. U20A20332, 52175063)Hebei Province PhD Graduate Innovation Funding Project (No. CXZZBS2021121)。
文摘The hydraulic servo actuator of heavy vehicle active suspension is investigated to clarify the correlation between system parameters and the control characteristics of active suspension hydraulic servo system.Accordingly, a nonlinear physical model of electro-hydraulic servo active suspension system is built.Compared with the conventional nonlinear modeling, the model in this study considers the asymmetry of working areas caused by single rod hydraulic cylinder in the suspension system.In accordance with the model, a nonlinear output feedback controller based on backstepping is designed, and the effectiveness of the controller is proved based on the experimental platform.The dynamic response curve of the electro-hydraulic servo control system under the change of parameters is generated based on the simulation model.The sensitivity of electro-hydraulic servo control performance to the change of system physical parameters is investigated, and two evaluation indexes are proposed to quantify and compare the effect of all physical parameter changes on position control system.As revealed by the results, the position control characteristics of suspension actuator are more sensitive to the changes of flow gain of the servo valve, system supply oil pressure and effective working areas of cylinder, and the two evaluation indexes are over 10 times higher than other physical parameters.
基金Supported by Japanese SMC Corporation with contract (No. 05-07)
文摘The servo system actuated by oscillating pneumatic cylinder for X-Y plate is a multi-variable nonlinear control system. Its mathematical model is established, and nonlinear factors are analyzed. Due to the existence of deadlock zone and the small damp of the pneumatic oscillating cylinder, it is likely to result in overshoot, and there is also certain steady-state error, so online modifying of proportion-integration-differentiation （PID） parameters is needed so as to achieve better control performance. Meanwhile considering the stability demand for long-term run, a fuzzy adaptive PID controller is designed. The result of hardware-inloop （HIL） test and real-time control experiment shows that the adaptive PID controller has desirable serfadaptability and robustness to external disturbance and to change of system parameters, and its control per- fonnance is better than that of traditional PID controllers.
基金National Natural Science Foundation of China （50825502）
文摘This article, in order to precisely impose friction on aircraft and weapon actuation systems, presents a new friction loading method characteristic of ＂torque-zero velocity＂ switching control with an electro-hydraulic load simulator. As the general Stribeck friction model has little related to static friction, it proposes a ＂torque-zero velocity＂ switcher, in which a zero-velocity controller is developed to load the static friction and a torque controller the kinetic friction. With the help of mathematical modeling, this article designs a ＂torque-zero velocity＂ switching controller and, correspondingly, provides a ＂dual-threshold judgment＂ algorithm. Simulation results indicate that the proposed method can be successfully used to carry out the static and kinetic friction simulation with an electro-hydraulic load simulator.
基金supported by the National Natural Science Foundation of China(Grant Nos.52305079 and U19A2072).
文摘This study proposes a data-driven friction modeling and compensation method aimed at solving the problem of servo performance degradation caused by friction in rotary servo actuators.First,a data-driven friction modeling method is proposed on the basis of the physics-informed neural network(PINN)and the LuGre model.The constructed friction model consists of sliding regime,static regime,and presliding regime,which extends the variables of the friction model to include velocity and position.The data-driven friction model not only retains the accuracy of the LuGre model in describing the dynamic behavior of friction at zero velocity but also improves the accuracy and convergence speed of the model through the powerful learning ability of PINN,which is verified in the two examples of constructing friction test data.Second,on the basis of the data-driven friction model,a composite compensation strategy centered on friction compensation is proposed.The friction compensator is used to compensate the internal friction of the actuator,and the extended Kalman filter is used to suppress the random disturbance to achieve the precise control of the servo actuator.Experimental validation of the proposed compensation strategy against three traditional control methods demonstrates its superiority,with average improvements of 49.5%,30.4%,and 32.7%in velocity tracking accuracy,respectively,while ensuring consistent accuracy across different positions.The proposed data-driven friction modeling and compensation method provides a new perspective and method for overcoming the effect of friction.
