Michael Ruderman

Modeling and Identification of Elastic Robot Joints With Hysteresis and Backlash

This paper presents a novel approach to the modeling and identification of elastic robot joints with hysteresis and backlash. The model captures the dynamic behavior of a rigid robotic manipulator with elastic joints. The model includes electromechanical submodels of the motor and gear from which the relationship between the applied torque and the joint torsion is identified. The friction behavior in both presliding and sliding regimes is captured by generalized Maxwell-slip model. The hysteresis is described by a Preisach operator. The distributed model parameters are identified from experimental data obtained from internal system signals and external angular encoder mounted to the second joint of a 6-DOF industrial robot. The validity of the identified model is confirmed by the agreement of its prediction with independent experimental data not previously used for model identification. The obtained models open an avenue for future advanced high-precision control of robotic manipulator dynamics.

Tracking Control of Motor Drives Using Feedforward Friction Observer

In motor drives, just as in other mechanical actuators, the friction compensation is extremely important as friction can have adverse impact on the overall control performance. In this paper, a feedforward friction observer (FFFO) is proposed as formulating an explicit analytical expression for the applied observation function. This ensures the cancelation of friction disturbances and time variances at steady state. The proposed observation scheme utilizes the two-state dynamic friction model with elastoplasticity (abbreviated as 2SEP), which is compact in parameterization and captures both the presliding and sliding phases of kinetic friction. The method to identify a motor drive plant with nonlinear friction in the frequency domain has been applied using only few frequency-response-function measurements. The feedback control design is performed with respect to the time delay detectable in the system, thus under additional constraints when determining the control gains. The optimal proportional-integral (PI) control designed this way is compared with the proportional control combined with the observer (P-FFFO). The simulation results show that P-FFFO control compensates faster for frictional disturbances at suddenly changing frictional conditions than PI control. In addition, an extensive experimental evaluation of velocity tracking control discloses P-FFFO as superior in terms of a faster steady-state convergence after various transient phases.

Use of Jiles–Atherton and Preisach Hysteresis Models for Inverse Feed-Forward Control

Hysteresis effects hinder the accurate control of electromagnetic actuators and require auxiliary sensors for properly determining the hysteretic system state. The physics-based Jiles-Atherton and the phenomenological Preisach hysteresis models provide powerful means to describe the magnetic hysteresis and its inverse. In this paper, we consider both hysteresis models in the scalar form from the control points of view, with a primary objective of the sensorless inverse feed-forward control. The identification complexity, the runtime, and the space efficiency of the control-oriented implementation are analyzed and compared for both modeling approaches. Their control performance for an inverse hysteresis compensation is experimentally evaluated on a specific force-controlled electromagnet system.

Optimal State Space Control of DC Motor

Abstract In comparison to classical cascade control architecture of DC motors, the state feedback control offers advantages in terms of design complexity, hardware realization and adaptivity. This paper presents a methodic approach to state space control of a DC motor. The state space model identified from experimental data provides the basis for a linear quadratic regulator (LQR) design. The state feedback linear control is augmented with a feedforward control for compensation of Coulomb friction. The controller is successfully applied and the closed loop behavior is evaluated on the experimental testbed under various reference signals.

Observer of Nonlinear Friction Dynamics for Motion Control

Kinetic friction in motion control systems is subject to large uncertainties due to the multiple, and often weakly known, internal and external factors such as roughness, thermal and lubricant state of contacting surfaces, varying normal loads, dwell time, wear, and others. The single modeling of friction behavior, even if comprehensive and accurate enough, appears to be insufficient for accurately compensating the friction disturbances, due to their time- and state-varying nature. In this paper, we propose a novel nonlinear friction observer aimed at the motion control. We analyze the uncertainties of viscous and Coulomb friction and derive an asymptotic observer for two-state friction dynamics without assuming a particular dynamic friction model. The linear observer gains prove to be sufficient for achieving an accurate friction estimate with a controllable eigenbehavior. Furthermore, we analyze the friction observer within the linear feedback loop and describe the required system identification and design of controller. An experimental case study accomplished on a rotary actuator system is provided for evaluating both the friction observer and its use for precise positioning.

Observer-Based Compensation of Additive Periodic Torque Disturbances in Permanent Magnet Motors

The impact of additive periodic torque disturbances on the controlled motion of permanent magnet motors can be significant. The paper shows how an observer-based drive control can efficiently reject the harmonic torque disturbances providing smooth angular velocity. The proposed control design is based on the state-space torque harmonics representation and Luenberger observer that proved to be adequate. The designed control algorithms are verified using an experimental setup with a permanent magnet synchronous motor with well-detectable torque harmonics. The rejection of additive position periodic torque disturbances is experimentally demonstrated for two first harmonics and that for different angular velocities.

Discrete dynamic Preisach model for robust inverse control of hysteresis systems

Hysteresis effects are present in diverse systems including structural mechanics, tribology, and electromagnetism. Hysteresis systems are generally known as exhibiting a memory effect which aggravates their accurate prediction and control. The classical Preisach model provides powerful means to describe arbitrary hysteresis with rate-independent behavior. In this paper, we address the problem of the robust inverse control of hysteresis systems while presenting a novel formulation of the discrete dynamic Preisach model and its inverse. The control-oriented features, among the advanced computational efficiency and the handling of hysteresis uncertainties are shown and discussed. The properly developed inverse hysteresis control is augmented by an auxiliary disturbance observer which captures uncertainties of the modeled hysteresis and its time-variant behavior. The performance of the proposed observer based control strategy is compared with a standard feedback of the controlled hysteretic value. The proposed approach is experimentally evaluated for linearizing the torsional hysteresis compliance. Providing an universal character the method is suitable for a broad class of hysteresis systems independent of the source and form of underlying hysteresis.

Control of Magnetic Shape Memory Actuators Using Observer-Based Inverse Hysteresis Approach

Magnetic shape memory (MSM) actuators belong to active material technologies with a high energy density and outstanding field-strain relation. Large-scale multivariate field-stress-strain hysteresis effects, however, antagonize their broad use because of the inherent difficulties with control. In this brief, we describe and compare experimentally several MSM control strategies, including the observer-based inverse hysteresis approach proposed in the previous works and combined here with a linear feedback controller by connecting both in parallel. For a prototypic MSM actuator with return spring, it is shown that the actuator plant can be approximated by an appropriate hysteresis operator and a linear transfer function of residual dynamics. The positioning profiles with a bandwidth 0.1-10 Hz and amplitudes between 0.01 and 0.1 mm have been evaluated by the experiments.

Modified Maxwell-slip Model of Presliding Friction

Abstract The distributed Maxwell-slip model provides a convenient way to describe the presliding friction behavior. The modified single-state Maxwell-slip (MMS) model is proposed with the main benefit to require two concentrated parameters only when describing the smooth hysteresis of the presliding friction. The model is rate-independent at both, saturated and unsaturated hysteresis, and is consistent with the generalized empirical friction model structure. Some novel perceptions on the frictional memory and drift are considered and proved in experiments. The evaluation performed on an actuator system with multiple coupled frictional surfaces reveals the proposed model as easily identifiable and accurate in prediction.

Identification of Soft Magnetic B-H Characteristics Using Discrete Dynamic Preisach Model and Single Measured Hysteresis Loop

In this paper, the identification of soft magnetic B - H characteristics when using the single measured hysteresis loop and fitting the discrete dynamic Preisach (DDP) model is addressed. The magnetic measurements with primary and secondary windings are performed to obtain the magnetostatic system response. The constrained least-squares estimation of the Preisach density function yields the global solution even for limited hysteresis data. The uncertainties detected during the measurement and computation of B and H values are discussed.