RESUMEN
As a critical support and fixed component of aero engines, electro-hydrostatic actuators, and other equipment, the operation of aviation bearings is often subject to high speed, high-temperature rise, large load, and other continuous complex fluctuation conditions, which makes their health assessment tasks more difficult. To solve this problem, an intelligent health assessment method based on a new Deep Transfer Graph Convolutional Network (DTGCN) is proposed for aviation bearings under large speed fluctuation conditions. First, a new DTGCN algorithm is designed, which mainly uses the domain adaptation mechanism to enhance the performance of Graph Convolutional Network (GCN) and the generalization performance of transfer properties. Specifically, order spectrum analysis is employed to resample the vibration signals of aviation bearings and transform them into order spectral signals. Then, the trained 1dGCN is used as the feature extractor, and the designed Dynamic Multiple Kernel Maximum Mean Discrepancy (DMKMMD) is calculated to match the difference in edge distribution. Finally, the aligned features are fed into the softmax classifier for intelligent health assessment. The effectiveness of the proposed diagnostic algorithm and method are validated by using aviation bearing fault data set under large speed fluctuation conditions.
RESUMEN
An electro-hydraulic load simulator (EHLS) is a typical case of torque systems with strong external disturbances from hydraulic motion systems. A new velocity synchronizing compensation strategy is proposed in this paper to eliminate motion disturbances, based on theoretical and experimental analysis of a structure invariance method and traditional velocity synchronizing compensation controller (TVSM). This strategy only uses the servo-valve's control signal of motion system and torque feedback of torque system, which could avoid the requirement on the velocity and acceleration signal in the structure invariance method, and effectively achieve a more accurate velocity synchronizing compensation in large loading conditions than a TVSM. In order to facilitate the implementation of this strategy in engineering cases, the selection rules for compensation parameters are proposed. It does not rely on any accurate information of structure parameters. This paper presents the comparison data of an EHLS with various typical operating conditions using three controllers, i.e., closed loop proportional integral derivative (PID) controller, TVSM, and the proposed improved velocity synchronizing controller. Experiments are conducted to confirm that the new strategy performs well against motion disturbances. It is more effective to improve the tracking accuracy and is a more appropriate choice for engineering applications.
RESUMEN
This study structures an adaptive prescribed performance output feedback controller for hydraulic systems with large uncertainties including disturbances and parametric uncertainties. Adaptive control is structured to approximate real system parameters. Based on estimated parameters, a nonlinear disturbance observer for largely mismatched disturbance and an extended state observer for matched disturbance and unmeasurable system states are integrated. Then, to guarantee prescribed tracking performance, an output feedback controller is proposed with uncertainty compensation based on the funnel function using the backstepping method. In addition, to avoid the occurrence of complex analytical calculations and "complexity explosion" in the backstepping method, a differentiator is adopted. Then, the Lyapunov method is utilized to prove the stability of the closed-loop system with estimation errors. Finally, experimental results were obtained to demonstrate that the proposed controller is valid.
RESUMEN
In this article, a novel neural network (NN)-based adaptive dynamic surface asymptotic tracking controller with guaranteed transient performance is proposed for n -degrees of freedom (DOF) hydraulic manipulators. To fulfill the work, the entire manipulator system model, including hydraulic actuator dynamics, is first established. Then, the neural adaptive dynamic surface controller is designed, in which the NN is utilized to approximate the unknown joint coupling dynamics, while the approximation error and uncertainties of the actuator dynamics are addressed by the nonlinear robust control law with adaptive gains. In addition, a modified funnel function that ensures the joint tracking errors remains within a predefined funnel boundary and is skillfully incorporated into the adaptive dynamic surface control (ADSC) design to achieve a guaranteed transient tracking performance. The theoretical analysis reveals that both the guaranteed transient tracking performance and asymptotic stability can be achieved with the proposed controller. Contrastive simulations are performed on a 2-DOF hydraulic manipulator to demonstrate the superiority of the proposed controller.
RESUMEN
This paper designs a new observer to address the issue of disturbance rejection in nonlinear plants with sense delay and measurement noise. The effect of sensor delay is characterized as a transport equation, and a predictor-based extended state observer has been designed using the PDE (partial differential equation) backstepping method. A cascaded ESO (extended state observer) architecture has been introduced to further reduce the limitation of measurement noise on observer bandwidth. This new cascade observer is capable of rapidly and accurately reconstructing system signals and compensating for output delay while avoiding the amplification of measurement noise. The numerical example has demonstrated the efficacy of the proposed observer structure, and extensive comparative experimental results have further confirmed its superiority.
RESUMEN
This article presents an active disturbance rejection adaptive precision pointing controller for all-electric tank bidirectional stability system in high speed driving environment. First, a comprehensive dynamic model for mechatronics and control integration of bidirectional stability system is modeled. Second, the backstepping idea is introduced to perfectly incorporate the parameter adaptive technique and the extended state observer to design a targeted control strategy. The matched and mismatched uncertainties of the system can be estimated separately with the construction of two extended state observers, while a novel adaptive function updated simultaneously via tracking and observation errors is synthesized to deal with the parametric uncertainties. Finally, theoretical analysis and extensive comparative results verify that the proposed controller is effective.
RESUMEN
In this article, the problem of high precision motion control for motor servo systems with modeling uncertainties and unknown actuator backlash is addressed. The combination of synthesized adaptive laws and continuous nonlinear robust term handles parameter uncertainties and system disturbances. The adaptive technique updates the unknown parameters of actuator backlash in real time and the backlash inverse function eliminates the backlash effect. Meanwhile, the designed controller without knowing the range of the disturbance upper bound but automatically estimates through the adaptive law, which improves the engineering practicability. Finally, the theoretical analysis proves the perfect asymptotic stability of the presented controller even with unmodeled disturbances and unknown actuator backlash. Extensive comparative experiments reveal the superiority of the presented method.
RESUMEN
The industrial gearboxes usually work in harsh and variable conditions, which results in partial failure of gears or bearings. Accordingly, the continuous irregular fluctuations of gearbox under variable conditions maybe increase the intraclass difference and reduce the interclass difference for the monitored samples. To this end, a new intelligent fault diagnosis method of gearbox based on adaptive intraclass and interclass convolutional neural network (AIICNN) under variable working conditions is proposed. The core of the proposed algorithm is to apply the designed intraclass and interclass constraints to improve the distribution differences of samples. Meanwhile, the adaptive activation function is added into the 1-D convolutional neural network (1dCNN) to enlarge the heterogeneous distance and narrow the homogeneous distance of samples. Specifically, the training sample subset with intraclass and interclass spacing fluctuations under variable conditions is first converted into frequency domain through the fast Fourier transform (FFT), and the designed AIICNN algorithm is employed for model training. Afterward, the testing subset is provided to the trained AIICNN algorithm for fault diagnosis. The experimental data of the planetary gearbox test rig verify the feasibility of the proposed diagnosis method and algorithm. Compared with other methods, this method can eliminate the difference of sample distribution under variable conditions and improve its diagnostic generalization.
RESUMEN
This paper addresses reinforcement learning based, direct signal tracking control with an objective of developing mathematically suitable and practically useful design approaches. Specifically, we aim to provide reliable and easy to implement designs in order to reach reproducible neural network-based solutions. Our proposed new design takes advantage of two control design frameworks: a reinforcement learning based, data-driven approach to provide the needed adaptation and (sub)optimality, and a backstepping based approach to provide closed-loop system stability framework. We develop this work based on an established direct heuristic dynamic programming (dHDP) learning paradigm to perform online learning and adaptation and a backstepping design for a class of important nonlinear dynamics described as Euler-Lagrange systems. We provide a theoretical guarantee for the stability of the overall dynamic system, weight convergence of the approximating nonlinear neural networks, and the Bellman (sub)optimality of the resulted control policy. We use simulations to demonstrate significantly improved design performance of the proposed approach over the original dHDP.
Asunto(s)
Algoritmos , Dinámicas no Lineales , Simulación por Computador , Retroalimentación , Redes Neurales de la ComputaciónRESUMEN
This article proposes an asymptotic adaptive command filtered control approach for uncertain nonlinear systems with parametric uncertainties, mismatched and matched disturbances. To accomplish the task, a disturbance observer (DO) only with one tuning parameter is firstly used to attain the disturbance compensation. The parameter uncertainties can be addressed via composite updated laws. Then, by judiciously combining DO, adaptive control and command filter technique, a novel command filtered controller with adaptive-gain auxiliary systems is developed to attain asymptotic tracking and shun "explosion of complexity". The system stability is proved by utilizing the Lyapunov function. Extensive experimental results uncover the preponderance of the exhibited strategy.
RESUMEN
Extended state observer acting as a popular tool can estimate the system states and total disturbances simultaneously. However, for extended-state-observer-based control of high-order nonlinear systems, there are still some difficult issues to solve, such as how to simultaneously reject matched and mismatched model uncertainties with strict theoretical proof, especially in the case of output feedback, "explosion of complexity" and so on. Motivated by these reasons, different control schemes in full-state feedback and output feedback conditions respectively will be integrated via the filter-based backstepping approach for saturated nonlinear systems. For the full-state feedback condition, adaptive neural network and extended state observer will be combined for each dynamic to handle the unknown nonlinear dynamics and external disturbances, respectively. For the output feedback condition, nonlinear disturbance observer design will be incorporated into the neural-network-based extended state observer scheme to handle mismatched disturbances at the same time. In particular, an auxiliary system will be constructed to compensate for the saturation influence. Moreover, the anticipate control effects of the developed controllers have been demonstrated by contrastive results for a hydraulic servo system.
Asunto(s)
Redes Neurales de la Computación , Dinámicas no Lineales , Simulación por Computador , Retroalimentación , IncertidumbreRESUMEN
Due to the nonlinearity and various uncertainties, the controller design for hydraulic servo systems with input constraint is more complicated and challenging. This paper first proposes an asymptotic tracking controller for electrohydraulic servomechanisms considering input constraint, parametric uncertainties, and unmodeled disturbances. The core innovation of this controller is to decouple the control input and the input nonlinearity while guaranteeing the nonlinear decoupling term and its derivative to be available and bounded. Meanwhile, the decoupling operation could be skillfully integrated to realize accurate adaptive model-based compensation while remaining the unique feature of asymptotic control for residual disturbances. For this purpose, based on the desired trajectory and the estimated disturbance via an extended state observer (ESO), a desired load pressure signal is constructed to replace the actual load pressure and accomplish the required decoupling operation. In this case, a desired adaptive feedforward compensation in combination with a robust integral of the sign of the error (RISE) feedback is proposed to attenuate parametric uncertainties and residual unmodeled disturbances, respectively. Subsequently, a smooth hyperbolic tangent function is integrated into the controller to handle the input constraint. Theoretical analysis proves that the developed control strategy can achieve semi-global asymptotic tracking performance. Besides, numerical simulations and experimental tests demonstrate that the proposed control scheme can ensure high-precision tracking performance and simultaneously satisfy the preset control input range when encountering the input constraint and modeling uncertainties.
RESUMEN
This paper presents an output feedback adaptive super-twisting sliding mode controller (SSMC) for hydraulic systems with unmodeled disturbances via utilizing an extended state observer (ESO). Both unmeasured system states and unmodeled disturbances are estimated by ESO based on output position signal, which avoids using noise-polluted signals and eliminates most of the disturbance effects on control performance simultaneously. Moreover, a SSMC is developed to further suppress the residual error of disturbance compensation, in which feedback gains are adapted online to further reduce the high-gain feedback. In addition, this proposed controller is continuous and chattering-free, which is beneficial to practical applications. Theoretical analysis indicates that the proposed controller ensures an asymptotic stability when existing constant disturbances, and ultimately bounded tracking performance for the time-variant disturbance case. Comparative experimental results reveal the validity of the developed approach.
RESUMEN
Comprehensive effects coming from measurement noises, matched and mismatched uncertainties make it difficult for electro-hydraulic servo systems to further attain high-accuracy tracking level. The existing control strategies often consider these control issues one-sidedly. Accordingly, we develop two different control strategies combining integral robust control and direct adaptive control for high-precision position control of double-rod electro-hydraulic systems to account for these control issues concurrently. Specially, by skillfully introducing a filtered error function, a novel desired compensation adaptive control framework will be integrated into the controller design to reduce environmental noises. Moreover, an improved noise-alleviation method is proposed to achieve high-accuracy calculation of the standard sign function in nonlinear integral robust terms. Furthermore, each of the control algorithms can guarantee asymptotic position tracking performance in general. Comparative experiments and simulation results show the evident superiorities of the developed control strategies.
RESUMEN
It is well known that hydraulic systems typically suffer from heavy disturbances including parametric uncertainties and unknown disturbances. In order to attain high performance tracking control, this paper proposes a composite design of nonlinear neural-networks (NN) and continuous robust integral of the sign of the error (RISE) feedback controller. The control development incorporates a NN feedforward component to have a compensation for unknown state-dependent disturbances and to further improve the accuracy of feedforward compensation, meanwhile input parameter is updated online. To achieve asymptotic stability, a novel RISE term with NN-based feedforward component is developed for the first time to enable the incorporation of model reference adaptive control structure where acceleration signal is not employed. The proposed controller guarantees controlled hydraulic system a semi-global asymptotic stability. For the experimental results, the prescribed transient performance is tested under rectangular trajectory and the steady state performance is tested under sinusoidal trajectory.
RESUMEN
Enhancing the robustness of output feedback control has always been an important issue in hydraulic servo systems. In this paper, an output feedback model predictive controller (MPC) with the integration of an extended state observer (ESO) is proposed for hydraulic systems. The ESO was designed to estimate not only the unmeasured system states but also the disturbances, which will be synthesized into the design of the output prediction equation. Based on the mechanism of receding horizon and repeating optimization of MPC, the output prediction equation will be updated in real time and the future behavior of the system will be accurately predicted since the disturbances are compensated effectively. Hence, the ability of the traditional MPC to suppress disturbances will be improved evidently. The experiment results show that the proposed controller has high-performance nature and strong robustness against various model uncertainties, which verifies the effectiveness of the proposed control strategy.
RESUMEN
A robust feed-forward observer is proposed for feel force control of an electro-hydraulic control loading system in flight simulators. Combining a velocity feed-forward compensator, the proposed control structure can effectively reduce a disturbance force caused by the control displacement of a pilot to further improve control performances. The electro-hydraulic control loading system is described as a force-displacement impedance system, including a mechanical system and a hydraulic control system, and its dynamic model is established as a double-loop system. Implementation of the proposed controller includes observer model design and filter optimization. Aiming at a force closed-loop which is always a non-minimum-phase system, an effective design method of an approximated inverse model is presented for obtaining an observer model, and it is an biregular transfer function. With the particular observer model, an optimization for the filter of feed-forward observer structure is carried out, by transforming the original structure into a control structure with consideration of modeling error between the obtained observer model and the actual inverse model, which is a 2-independent-controllers structure. Then, an H∞ controller is employed to optimize the transformed structure to indirectly obtain an optimized filter. In order to ensure existence of the H∞ controller for the transformed structure of the feed-forward observer with a biregular observer model, some dynamic characteristics are analyzed and proved. Based on the characteristics, a nominal model modification method is presented to obtain a well-posed H∞ control structure. For designing a performance weighting function, an optimize goal is discussed and its proof is given, and then an algorithm using a bisection method is proposed to further improve the robust performance of the proposed controller. To verify the efficiency of the proposed controller, control performances of the electro-hydraulic control loading system are evaluated in experiments and are compared with those of other controllers, including the proportional-integral-derivative controller, the velocity feed-forward compensator, and the H∞ controller. Experimental results show that the proposed controller can effectively reduce the disturbance force of the electro-hydraulic control loading system, and its impedance characteristics are closer to the ideal force feel model than the other controllers.
RESUMEN
This paper proposes a continuous adaptive integral robust control with robust integral of the sign of the error (RISE) feedback for a class of uncertain nonlinear systems, in which the RISE feedback gain is adapted online to ensure the robustness against disturbances without the prior bound knowledge of the additive disturbances. In addition, an adaptive compensation integrated with the proposed adaptive RISE feedback term is also constructed to further reduce design conservatism when the system also exists parametric uncertainties. Lyapunov analysis reveals the proposed controllers could guarantee the tracking errors are asymptotically converging to zero with continuous control efforts. To illustrate the high performance nature of the developed controllers, numerical simulations are provided. At the end, an application case of an actual electromechanical servo system driven by motor is also studied, with some specific design consideration, and comparative experimental results are obtained to verify the effectiveness of the proposed controllers.
RESUMEN
This paper addresses the high performance motion control of hydraulic actuators with parametric uncertainties, unmodeled disturbances and unknown valve dead-zone. By constructing a smooth dead-zone inverse, a robust adaptive controller is proposed via backstepping method, in which adaptive law is synthesized to deal with parametric uncertainties and a continuous nonlinear robust control law to suppress unmodeled disturbances. Since the unknown dead-zone parameters can be estimated by adaptive law and then the effect of dead-zone can be compensated effectively via inverse operation, improved tracking performance can be expected. In addition, the disturbance upper bounds can also be updated online by adaptive laws, which increases the controller operability in practice. The Lyapunov based stability analysis shows that excellent asymptotic output tracking with zero steady-state error can be achieved by the developed controller even in the presence of unmodeled disturbance and unknown valve dead-zone. Finally, the proposed control strategy is experimentally tested on a servovalve controlled hydraulic actuation system subjected to an artificial valve dead-zone. Comparative experimental results are obtained to illustrate the effectiveness of the proposed control scheme.