Unknown External Forces Research Articles

Adaptive neural control for an uncertain robotic manipulator with joint space constraints

ABSTRACTIn this paper, adaptive neural tracking control is proposed for a robotic manipulator with uncertainties in both manipulator dynamics and joint actuator dynamics. The manipulator joints are subject to inequality constraints, i.e., the joint angles are required to remain in some compact sets. Integral barrier Lyapunov functionals (iBLFs) are employed to address the joint space constraints directly without performing an additional mapping to the error space. Neural networks (NNs) are utilised to compensate for the unknown robot dynamics and external force. Adapting parameters are developed to estimate the unknown bounds on NN approximations. By the Lyapunov synthesis, the proposed control can guarantee the semi-global uniform ultimate boundedness of the closed-loop system, and the practical tracking of joint reference trajectory is achieved without the violation of predefined joint space constraints. Simulation results are given to validate the effectiveness of the proposed control scheme.

A real-time Lie-group differential algebraic equations method to solve the inverse nonlinear vibration problems

Desiring a real-time recovery of an unknown external force in the nonlinear inverse vibration problem, we transform the nonlinear ordinary differential equation (ODE) of motion into a nonlinear parabolic type partial differential equation (PDE), which can raise the robustness against large noise. Then, we come to a heat source identification problem, of which the numerical method of lines is used to discretize the PDE into a system of differential algebraic equations (DAEs). Consequently, we can develop an implicit Lie-group scheme and a Newton algorithm to stably solve the DAEs for finding the unknown force, damping function, or stiffness function, which is well recovered even under a large noise. Superficially, we seem to transform a simple ODE to a more complex PDE; however, the advantages gained in this transformation will be seen when we test some nonlinear inverse vibration problems with large time span and under large noise.

On the rejection of internal and external disturbances in a wind energy conversion system with direct-driven PMSG

This paper deals with the critical issue in a wind energy conversion system (WECS) based on a direct-driven permanent magnet synchronous generator (PMSG): the rejection of lumped disturbance, including the system uncertainties in the internal dynamics and unknown external forces. To simultaneously track the motor speed in real time and capture the maximum power, a maximum power point tracking strategy is proposed based on active disturbance rejection control (ADRC) theory. In real application, system inertia, drive torque and some other parameters change in a wide range with the variations of disturbances and wind speeds, which substantially degrade the performance of WECS. The ADRC design must incorporate the available model information into an extended state observer (ESO) to compensate the lumped disturbance efficiently. Based on this principle, a model-compensation ADRC is proposed in this paper. Simulation study is conducted to evaluate the performance of the proposed control strategy. It is shown that the effect of lumped disturbance is compensated in a more effective way compared with the traditional ADRC approach.

Closed-Loop Neuromorphic Benchmarks.

Evaluating the effectiveness and performance of neuromorphic hardware is difficult. It is even more difficult when the task of interest is a closed-loop task; that is, a task where the output from the neuromorphic hardware affects some environment, which then in turn affects the hardware's future input. However, closed-loop situations are one of the primary potential uses of neuromorphic hardware. To address this, we present a methodology for generating closed-loop benchmarks that makes use of a hybrid of real physical embodiment and a type of “minimal” simulation. Minimal simulation has been shown to lead to robust real-world performance, while still maintaining the practical advantages of simulation, such as making it easy for the same benchmark to be used by many researchers. This method is flexible enough to allow researchers to explicitly modify the benchmarks to identify specific task domains where particular hardware excels. To demonstrate the method, we present a set of novel benchmarks that focus on motor control for an arbitrary system with unknown external forces. Using these benchmarks, we show that an error-driven learning rule can consistently improve motor control performance across a randomly generated family of closed-loop simulations, even when there are up to 15 interacting joints to be controlled.

A Balance Control Strategy for a Walking Biped Robot under Unknown Lateral External Force using a Genetic Algorithm

In this paper, we propose a force-resisting balance control strategy for a walking biped robot subject to an unknown continuous external force. We assume that the biped robot has 12 degrees of freedom (DOFs) with position-controlled joint motors, and that the unknown continuous external force is applied to the pelvis of the biped robot in the single support phase (SSP) walking gait. The suggested balance control strategy has three phases. Phase 1 is to recognize the application of an unknown external force using only zero moment point (ZMP) sensors. Phase 2 is to control the joint motors according to a method that uses a genetic algorithm and the linear interpolation technique. Against an external continuous force, the robot retrieves the pre-calculated solutions and executes the desired torques with interpolation performed in real time. Phase 3 is to make the biped robot move from the SSP to the double support phase (DSP), rejecting external disturbances using the sliding mode controller. The strategy is verified by numerical simulations and experiments.

An Endorectal Ultrasound Probe Comanipulator With Hybrid Actuation Combining Brakes and Motors

—A robotic device, aimed at assisting a urologist who positions an endorectal ultrasound probe to perform prostate biopsies, is presented. The proposed system is a comanipulator that holds the probe simultaneously with the urologist. This robot supports two modes of operation: the free mode, where the entire movement control is left to the urologist when he/she positions the probe with respect to the prostate thanks to the feedback provided by the ultrasound images; the locked mode, where the robot's role is to precisely maintain the targeted biopsy site at a given location, while the urologist can insert a needle through a guide mounted to the probe and proceed to biopsy. The device combines three brakes and three motors. This allows both transparent comanipulation in the free mode with 6 Degrees of Freedom (DoFs) liberated and stabilization of the probe in the locked mode. At the control level, a main challenge in the locked mode raises from antagonistic constraints: the needle placement shall be precise in spite of unknown external forces due to the contact between the probe and the rectum; the robot apparent impedance shall be low due to security constraints. This is solved by an inner low stiffness controller and an outer slow integration for canceling steady state errors. Both in vitro and in cadavero experimental results show the efficiency of the system in the two modes of operation.

An Improved eXtended Classifier System for the Real-time-input Real-time-output (XCSRR) Stability Control of a Biped Robot

Abstract In this paper a revised reinforcement learning method is presented for stability control problems with real-value inputs and outputs. The revised eXtended Classifier System for Real-input and Real-output (XCSRR) controller is designed, which is capable of working at fully real-value environment such as stability control of robots. XCSRR is a novel approach to enhance the performance of classifier systems for more practical problems than systems with merely binary behaviour. As a case study, we use XCSRR to control the stability of a biped robot, which is subjected to unknown external forces that would disturb the robot equilibrium. The external forces and the dynamics of the upper body of the biped robot are modelled in MATLAB software to train the XCSRR controller. Theoretical and experimental results of the learning behaviour and the performance of stability control on the robot demonstrate the strength and efficiency of the proposed new approach.

Stabilizing saddles

A synergetic control technique for stabilizing a priori unknown saddle steady states of dynamical systems is described. The method involves an unstable filter technique combined with a derivative feedback. The cut-off frequency of the filter is not limited by the damping of the system, and therefore can be set relatively high. This essentially increases the rate of convergence to the steady state. The synergetic technique is robust to the influence of unknown external forces, which change the coordinates of the steady state in the phase space.

Enhanced control of saddle steady states of dynamical systems

An adaptive feedback technique for stabilizing a priori unknown saddle steady states of dynamical systems is described. The method is based on an unstable low-pass filter combined with a stable low-pass filter. The cutoff frequencies of both filters can be set relatively high. This allows considerable increase in the rate of convergence to the steady state. We demonstrate numerically and experimentally that the technique is robust to the influence of unknown external forces, which change the position of the steady state in the phase space. Experiments have been performed using electrical circuits imitating the damped Duffing-Holmes and chaotic Lindberg systems.

Online state and input force estimation for multibody models employing extended Kalman filtering

This paper discusses the use of Subsystem Global Modal Parameterization (SS-GMP) reduced multibody models in an augmented discrete extended Kalman filter (A-DEKF) to generate a general formalism for online coupled state/input estimation in mechanisms. The SS-GMP approach is proposed to reduce a general multibody model of a mechanical system into a real-time capable model without considerable loss in accuracy. In order to use these reduced models with an extended Kalman filter, the necessary derivatives of this model are provided. An exponential integration scheme is used to discretize the model in order to be compatible with discrete time filters. Finally, the augmented approach is used for the estimation, in which the unknown external forces are considered as additional states to be estimated. The proposed approach is validated numerically and compared to three other filtering approaches. The validation demonstrates that the proposed approach provides accurate results while still maintaining real-time performance.

Improved neural network control of inverted pendulums

Nowadays, neural network controllers (NNCs) are getting more and more prevalent because they are able to handle unknown systems by learning them and adapt to their changing behaviour. The family of robust fixed point transformations (RFPT) has been partly developed to solve control tasks without knowing the exact parameters of a controlled system. When disturbances effect a plant or the neural network controller is not trained properly RFPT integrated to the controller is suitable to reduce the problems caused by the model approximation and make the controller robust to the unknown external forces. In this paper, a novel combination of neural networks and robust fixed point transformations is introduced to balance an inverted pendulum on the top of a cart of changing nominal position. The results show that the inaccuracies caused by the disturbances can be reduced significantly when RFPT is used in the control process.

Adaptive Regulation of Tracking Servo for Next Generation Optical Data Storage Systems

The tracking servo control has played an important role in the data storage servo systems. In the next generation optical data storage systems, i.e. the near-field recording system, the tracking error of the servo system should be below 10 nanometers under various unknown situations. However, higher data transfer rate and higher data density make it difficult to maintain the desired tracking precision during normal disk operation. It is proposed in this paper to use an adaptive regulation approach to maintain the tracking error below its desired value, despite unknown track eccentricity and external force disturbance. The performance of the proposed control approach is analyzed and simulation results are presented to illustrate the capability of the proposed adaptive regulator to achieve and maintain the desired tracking precision.

Investigation of a Stochastic Inverse Method to Estimate an External Force: Applications to a Wave‐Structure Interaction

The determination of an external force is a very important task for the purpose of control, monitoring, and analysis of damages on structural system. This paper studies a stochastic inverse method that can be used for determining external forces acting on a nonlinear vibrating system. For the purpose of estimation, a stochastic inverse function is formulated to link an unknown external force to an observable quantity. The external force is then estimated from measurements of dynamic responses through the formulated stochastic inverse model. The applicability of the proposed method was verified with numerical examples and laboratory tests concerning the wave‐structure interaction problem. The results showed that the proposed method is reliable to estimate the external force acting on a nonlinear system.

Story‐by‐story estimation of the stiffness parameters of laterally‐torsionally coupled buildings using forced or ambient vibration data: I. Formulation and verification

SUMMARYA new parameter estimation algorithm is described for identifying the stiffness properties of torsionally coupled shear buildings from their linear response due to ambient excitations or during low‐amplitude forced‐vibration tests. The algorithm is based on the time‐domain equations of motion, and yields estimates of the stiffness properties using a measure of the equilibrium of forces acting on each floor over a time interval. The banded structure of the stiffness matrix — a property intrinsic to torsion‐shear buildings — is exploited to decompose the initial inverse problem into several problems of reduced size. This decomposition allows the identification of lateral and torsional stiffnesses of individual stories, independent of the others. The algorithm utilizes vibration data where input excitation is known/measured, which is typical for forced‐vibration tests and earthquakes. If the ambient vibrations of the structure are adequately uncorrelated to the (unknown) external forces that induce such vibrations, then the algorithm can also be modified for output‐only system identification. The proposed algorithm is verified — and its various attributes are investigated — using simulation data from the ‘Analytical Phase I’ of the IASC (International Association for Structural Control)‐ASCE (American Society of Civil Engineers) benchmark studies. The companion article is devoted to the algorithm's application to experimental data, using data from the ‘Experimental Phase’ of the same benchmark studies. Copyright © 2011 John Wiley & Sons, Ltd.

Combining a feedback linearization controller with a disturbance observer to control a chaotic system under external excitation

This paper proposes a robust controller which combines a feedback linearization controller with a disturbance observer. This controller can suppress the chaotic motion of an unknown nonlinear system even though it receives an unknown external force. Two numerical simulations are performed to demonstrate the feasibility of the proposed method.

Disturbance Compensation Control for a Biped Vehicle

This paper describes an avoidance behavior against unknown forces acting from outside the system on a biped vehicle. External forces that act on a biped vehicle can be divided into two categories: disturbances from the rider of the vehicle and disturbances from outside the system. Disturbances from the rider are measured by a force sensor placed between the rider's seat and the pelvis, while disturbances from outside the system are estimated from zero moment point (ZMP) errors. To guarantee walking stability, the waist position is adjusted to match the measured ZMP to the reference ZMP and the position of the landing foot is adjusted such that the waist trajectory does not diverge. On implementing the developed method on the human-carrying biped robot in an experimental setup, the robot could realize a stable walk even while subjected to unknown external forces in the environment. On applying a pushing force to the walking robot, the robot moved in the pushed direction and away from the source of the externally generated forces. On pushing the robot that was walking forward, the robot stopped moving forward and avoided getting closer to the source of the externally generated forces. We confirmed the effectiveness of the proposed control through these experiments.

A biped static balance control and torque pattern learning under unknown periodic external forces

This paper addresses a biped balancing task in which an unknown external force is exerted, using the so-called ‘ankle strategy’ model. When an external force is periodic, a human adaptively maintains the balance, next learns how much force should be produced at the ankle joint from its repeatability, and finally memorized it as a motion pattern. To acquire motion patterns with balancing, we propose a control and learning method: as the control method, we adopt ground reaction force feedback to cope with an uncertain external force, while, as the learning method, we introduce a motion pattern generator that memorizes the torque pattern of the ankle joint by use of Fourier series expansion. In this learning process, the period estimation of the external force is crucial; this estimation is achieved based on local autocorrelation of joint trajectories. Computer simulations and robot experiments show effective control and learning results with respect to unknown periodic external forces.

双边缘测风激光雷达法布里-珀罗鉴频器控制方法

The paper demonstrates a nonlinear design of the stabilization system of Fabry-Perot(F-P) interferometer which is a frequency discriminator used in double-edge wind lidar. First, a nonlinear model is set up to describe the F-P interferometer transmission peak locking problem. The disturbance caused by air pressure fluctuation, temperature change or mechanical vibration is characterized by the unknown external force acting on piezoelectric ceramics. For such a disturbance rejection problem, the nonlinear proportional-integral-derivative(PID) control is developed by utilizing arctan function to improve the feedback gain of the typical PID control design. The results show that, with the new controller, the discriminator can be stabilized much faster, and the steady-state error is approximatel

Three-dimensional elasticity solution of simply supported functionally graded rectangular plates with internal elastic line supports

This paper studies the stress and displacement distributions of simply supported functionally graded rectangular plates with internal elastic line supports. The Young's modulus is graded through the thickness following the exponential law and the Poisson's ratio is kept constant. On the basis of three-dimensional elasticity theory, the solutions of displacements and stresses of the plate under static loads, which exactly satisfy the governing differential equations and the simply supported boundary conditions at four edges of the plate, are analytically derived. The reaction forces of the internal elastic line supports are regarded as the unknown external forces acting on the lower surface of the plate. The unknown coefficients in the solutions are then determined by the boundary conditions on the upper and lower surfaces of the plate. Convergence and comparison studies demonstrate the correctness and effectiveness of the proposed method. The effect of variations in Young's modulus on the displacements and stresses of rectangular plates and the effect of internal elastic line supports on the mechanical properties of plates are investigated.

Three-dimensional elasticity solution of simple-supported rectangular plate on point supports, line supports and elastic foundation

This paper studies the bending of simple-supported rectangular plate on point supports, line supports and elastic foundation. On the basis of three-dimensional elasticity theory, the exact expressions of the displacement functions, which satisfy the governing differential equations and the simply supported boundary conditions at four edges of the plate, are analytically derived. The reaction forces of the intermediate supports are regarded as the unknown external forces acting on the lower surface of the plate. The unknown coefficients are then determined by the boundary conditions on the upper and lower surfaces of the plate. Comparing the numerical results obtained from the proposed method to those obtained from Kirchhoff plate theory, Mindlin plate theory and those obtained from the commercial finite element software ANSYS, the high accuracy of the present method has been demonstrated.