Feedforward Term Research Articles

Neural Network-Based Repetitive Learning Control of Euler Lagrange Systems: An Output Feedback Approach

In this letter, position tracking control problem of a class of fully actuated Euler Lagrange (EL) systems is aimed. The reference position vector is considered to be periodic with a known period. Only position measurements are available for control design while velocity measurements are not. Furthermore, the dynamic model of the EL systems has parametric and/or unstructured uncertainties which avoid it to be used as part of the control design. To address these constraints, an output feedback neural network-based repetitive learning control strategy is preferred. Via the design of a dynamic model independent velocity observer, the lack of velocity measurements is addressed. To compensate for the lack of dynamic model knowledge, universal approximation property of neural networks is utilized where an online adaptive update rule is designed for the weight matrix. The functional reconstruction error is dealt with the design of a novel repetitive learning feedforward term. The outcome is a dynamic model independent output feedback neural network-based controller with a repetitive learning feedforward component. The stability of the closed–loop system is investigated via rigorous mathematical tools with which semi-global asymptotic stability is ensured.

Robust task-space motion control of a mobile manipulator using a nonlinear control with an uncertainty estimator

In this study, a robust nonlinear control method with an uncertainty estimator is proposed and applied to a kinematically redundant mobile manipulator (MM) position tracking in its task space. The proposed MM has four degrees-of-freedom (dof) serial manipulator mounted on three of mobile bases. The proposed method incorporates a feed-forward control term to reinforce the control action with extravagance from the desired acceleration vector; a disturbance estimator to reimburse for the unknown effects such as parametric uncertainties, external disturbances, unmodeled dynamics and a decentralized PID (proportional-integral-derivative) controller as a feedback loop to strengthen the stability of the system. The main strengths of the proposed scheme are its high robustness against parameter uncertainties and external disturbances, simplicity in design and ease of implementation. The effectiveness, feasibility, and robustness of the proposed method are illustrated using computer-based simulations with and without uncertainty estimator.

Analytical expression for LVRT of BDFIG with enhanced current control to CW and reactive power support from GSC

During the low-voltage ride-through (LVRT) of the brushless doubly-fed induction generator (BDFIG), the oscillation of the power winding (PW) flux will induce high current in the control winding (CW), causing possible violation to the current constraint and weakening the power support capability of the machine-side converter (MSC). The existing PI control is designed based on the steady-state operation thus sensitive to the intense oscillation of the PW flux. To isolate the CW current control from the disturbances of the oscillating PW flux, the enhanced CW current control is newly proposed by introducing the change rates of the PW flux and the rotor flux to the inner-loop current control of the MSC as the feed-forward compensation terms. The analytical expressions of the PW current, the CW current, the rotor current, and the CW voltage are newly derived to quantify the LVRT effect of the enhanced control. With the DC-voltage constraint considered, the capacity of the grid-side converter (GSC) is newly utilized to improve the reactive power support capability. Dynamic simulation results show that the enhanced control reduces the oscillation of the CW current, the PW current and the rotor current. The accuracy of the analytical expression is validated by comparing it with the step-by-step simulation results. The selected references of the GSC current provide the additional reactive power support without violation to the constraint of the DC-voltage.

On the Design of a 4 Degrees-of-freedom Pick and Place Cable Suspended Parallel Manipulator

This paper proposes the design and control strategy for a four degrees-of-freedom spatial cable-suspended parallel robot for pick and place operations. Pick and place is a repetitive task requiring payload changes for the movement to pick-up the object, and the movement to the nal pose to release the manipulated object. In this paper, a new robust control strategy has been proposed, together with proper trajectories for the required operation. The control strategy consists on the system decoupling and linearization by means of a feedforward term and a cascade PD controller. The main advantage of the proposed solution is that its design can be scalable in size spanning from centimeters to meters with a relatively good positioning accuracy. Finally, simulations are reported to show the overall performances of the proposed con guration for pick and place operations with a medium size manipulator.

Obstacle avoidance control of a human-in-the-loop mobile robot system using harmonic potential fields

SUMMARYThis paper considers applications where a human agent is navigating a semi-autonomous mobile robot in an environment with obstacles. The human input to the robot can be based on a desired navigation objective, which may not be known to the robot. Additionally, the semi-autonomous robot can be programmed to ensure obstacle avoidance as it navigates the environment. A shared control architecture can be used to appropriately fuse the human and the autonomy inputs to obtain a net control input that drives the robot. In this paper, an adaptive, near-continuous control allocation function is included in the shared controller, which continuously varies the control effort exerted by the human and the autonomy based on the position of the robot relative to obstacles. The developed control allocation function facilitates the human to freely navigate the robot when away from obstacles, and it causes the autonomy control input to progressively dominate as the robot approaches obstacles. A harmonic potential field-based non-linear sliding mode controller is developed to obtain the autonomy control input for obstacle avoidance. In addition, a robust feed-forward term is included in the autonomy control input to maintain stability in the presence of adverse human inputs, which can be critical in applications such as to prevent collision or roll-over of smart wheelchairs due to erroneous human inputs. Lyapunov-based stability analysis is presented to guarantee finite-time stability of the developed shared controller, i.e., the autonomy guarantees obstacle avoidance as the human navigates the robot. Experimental results are provided to validate the performance of the developed shared controller.

Clamping force control based on dynamic model estimation for electromechanical brakes

The electromechanical brake (EMB) is expected to be utilized for future brake systems due to its many advantages. In this paper, keeping commercialization of the EMB in mind, a new EMB clamping force controller is proposed to overcome the limitations of the existing controller, namely, the extra cost for sensor installation and response delay. To design the controller, both mechanical parts and electrical parts in the EMB have to be mathematically analyzed. Also, dynamic models, clamping force, and friction torque are estimated to generate some feed-forward terms of the controller. With an estimation of the contact point where brake pads start to come into contact with a disk wheel, the clamping force is expressed as a polynomial curve versus the motor angle. The estimated clamping force is evaluated in comparison with measured values by a load cell. The proposed controller is based on an adaptive sliding mode control method with an adaptive law reducing errors of the friction torque model. Lastly, the performance of the entire control system is compared with that of the existing controller on a test bench.

New Adaptive Neural Network for Uncertain Nonlinear System with Disturbance

An output of RBF neural network depends linearly on the matrix weights, a training is thus a linear optimization problem. However, by adjusting the centers and the widths this type of neural network structure becomes nonlinearly parameterized. In this work, lumped disturbances consisting of both approximation errors and external disturbance are estimated by an adaptive RBF neural network structure combining with a feed-forward correction, in which the feed-forward correction term is calculated by the algebraic equation regarding the parameters of controller and the radial basis function. This estimator (using estimating the lumped disturbance) is also used both in class of SISO nonlinear and MIMO nonlinear system. In addition, an adaptive scheme for the RBF neural network (an output and n outputs) is developed to approximate unknown system functions. On the other side, the performance of closed loop system (settling time, overshoot and the static error) would be improved by using an adaptive law to update the parameters of controller instead of choosing fixed controller's parameters which are coefficients of hurwitz polynomial. Thanks to Lyapunov's theory, asymptotic stability is established with the tracking errors converging to a neighborhood of the origin. Finally, there are two examples, coupled tank liquid system and an active magnetic bearing system, are presented to illustrate the our proposed methods.

Model reference tracking control of continuous-time periodic linear systems with actuator jumping fault and its applications in orbit maneuvering

This paper deals with the model reference tracking control problem of continuous-time periodic linear systems when the actuator occurs jumping fault. The main contribution is to formulate the parametric design algorithm for the systems by utilizing the parametric solution of the generalized Sylvester matrix equations. The existence condition of the controller is deduced based on the Lyapunov stability theory. The controller consists of the additive contribution of two terms: a feedback term and a feedforward term. The feedback term is the feedback control law which can stabilize the system with finite expected cost. The feedforward term is the complete parametric feedforward tracking compensator. The simulation for flying around mission is carried out about two spacecrafts in elliptical orbit. The simulation results show the effectiveness of the proposed approach.

A Decoupled Adaptive Noise Detection Based Control Approach for a Grid Supportive SPV System

A grid supportive two-stage three-phase four-wire solar photovoltaic (SPV) system is presented in this paper, wherein a boost converter is used as a first stage to serve the function of maximum power point tracking (MPPT) and a four-leg voltage source converter (VSC) is used to feed the extracted SPV energy, along with supporting distribution system for improvement in the power quality. Unlike conventional SPV inverters, the proposed solar energy conversion system provides extra functionalities such as balancing of grid currents, reactive power compensation, mitigation of harmonics, and neutral current elimination in grid side neutral conductor. An extra feed-forward term is added in the control loop to provide fast dynamic responses. The PV array voltage is continuously adjusted using a boost converter to achieve MPPT. A control approach employing decoupled adaptive noise detection (DAND) algorithm is used for controlling the four-leg grid-tied VSC. The DAND algorithm is a simple approach using two multipliers, one integrator, and one summer per phase for detection of useful component of load current. The proposed control algorithm gives features such as simple structure, fast convergence, frequency adaptive detection, and good steady-state performance. The grid currents are found adhering to an IEEE-519 standard, even under nonlinear and unbalanced loads at common point of interconnection.

State/Model-Free Variable-Gain Discrete Sliding Mode Control for an Ultraprecision Wafer Stage

Wafer stage is an important mechatronic unit of industrial lithography tool for manufacturing integrated circuits. This paper presents a novel state/model-free variable-gain discrete sliding mode control (DSMC) to suppress the unmolded position-dependent dynamics and disturbances in the nanopositioning wafer stage. The proposed DSMC is essentially composed of feedforward control term, linear feedback control term, and nonlinear switching control term, which can be designed separately. The gain of the switching control term is meaningfully designed to be variable to balance the tradeoff between the robustness and the chattering. Data-driven parameter optimization approach is employed to achieve the optimal controller parameters of the typically nonlinear controller, where off-line parameter updating is iteratively carried out based on the input/output data to minimize a predefined objective function. This scheme facilitates a rapid implementation without either a parameter model or a state observer, and excellent tracking performance with the optimal controller parameters. Moreover, the closed-loop stability is analyzed, and the proposed DSMC is finally implemented on an ultraprecision wafer stage developed in our lab. Comparative experimental results demonstrate that it not only achieves nanoscale tracking accuracy but also possesses promising robustness to position-dependent dynamics and disturbances.

Energy Balancing Improvement of Modular Multilevel Converters Under Unbalanced Grid Conditions

This paper presents a feedback/feed-forward control strategy to improve the voltage balancing of modular multilevel converters (MMCs) under unbalanced grid conditions. The inclusion of a feed-forward compensation improves the disturbance rejection capability against asymmetrical faults and sudden voltage imbalances in the ac grid. The floating capacitor voltages of the MMC are controlled in two stages. The first one equally distributes the voltages in each arm, and the second one balances the energy among the six arms. This last control stage is also divided into the called horizontal and vertical balancing. Based on a detailed analysis of the power terms disturbing the MMC arm energies, the control strategy is tailored to improve the voltage transient response by adding feed-forward terms to both horizontal and vertical balancing controls. In addition, various approaches to manage the MMC under unbalanced grid conditions are compared regarding several aspects, such as capacitor voltage ripple, balancing performance, and negative-sequence current injection. Advantages and disadvantages of both the proposed and the conventional energy-based control schemes are also studied, showing that the voltage control performance can be improved by enhancing the existing MMC control systems.

A Sliding Mode Guidance Law with ESO for Near Space Interceptor

Considering the autopilot dynamic characteristics of the near space interceptor and the target maneuver, a novel sliding mode guidance law is proposed based on the extended state observer (ESO). Firstly, the ESO is constructed to estimate the target acceleration of the disturbances for the guidance system, and the disturbance estimated values are employed as the feed-forward compensation term to obtain the robustness of the target maneuver and system disturbances. Then, a novel sliding mode guidance law based on the ESO is given. Finally, simulation results for the near space interceptor guidance system show that the proposed guidance law can not only overcome the effect of the autopilot dynamic delay and target maneuver, but also ensure the accurate guidance results.

Control Strategy for a Grid-Connected Inverter under Unbalanced Network Conditions—A Disturbance Observer-Based Decoupled Current Approach

This paper proposes a new approach on the novel current control strategy for grid-tied voltage-source inverters (VSIs) with circumstances of asymmetrical voltage conditions. A standard grid-connected inverter (GCI) allows the degree of freedom to integrate the renewable energy system to enhance the penetration of total utility power. However, restrictive grid codes require that renewable sources connected to the grid must support stability of the grid under grid faults. Conventional synchronously rotating frame dq current controllers are insufficient under grid faults due to the low bandwidth of proportional-integral (PI) controllers. Hence, this work proposes a proportional current controller with a first-order low-pass filter disturbance observer (DOb). The proposed controller establishes independent control on positive, as well as negative, sequence current components under asymmetrical grid voltage conditions. The approach is independent of parametric component values, as it estimates nonlinear feed-forward terms with the low-pass filter DOb. A numerical simulation model of the overall power system was implemented in a MATLAB/Simulink (2014B, MathWorks, Natick, MA, USA). Further, particular results show that double-frequency active power oscillations are suppressed by injecting appropriate negative-sequence currents. Moreover, a set of simulation results provided in the article matches the developed theoretical background for its feasibility.

Frequency-Modulation Control of a DC/DC Current-Source Parallel-Resonant Converter

This paper proposes a frequency-modulation control scheme for a dc/dc current-source parallel-resonant converter with two possible configurations. The basic configuration comprises an external voltage loop, an internal current loop, and a frequency modulator: the voltage loop is responsible for regulating the output voltage, the current loop makes the system controllable and limits the input current, and the modulator provides robustness against variations in resonant component values. The enhanced configuration introduces the output inductor current as a feed-forward term and clearly improves the transient response to fast load changes. The theoretical design of these control schemes is performed systematically by first deriving their small-signal models and second using Bode diagram analysis. The actual performance of the proposed control schemes is experimentally validated by testing on a laboratory prototype.

Simultaneous Identification and Adaptive Torque Control of Permanent Magnet Synchronous Machines

This paper presents an experimentally verified, Lyapunov-based adaptive control design for permanent magnet synchronous ac machines, which exploits overactuation to achieve parameter identification and torque regulation objectives simultaneously. This is achieved by regulating the states of the system (i.e., the stator currents) to the output error-zeroing manifold, along which they are varied to provide excitation for parameter identification. The proposed control law utilizes a combination of adaptively tuned feedforward and feedback decoupling terms, in addition to proportional feedback, to achieve reference current tracking in the presence of parameter uncertainty. A switching-sigma modification to the adaptive update law is used to ensure robust stability of the closed-loop adaptive system, and excitation for parameter estimation is introduced via the direct-axis current reference input. The resulting controller achieves the simultaneous identification and control objective while providing consistent transient response characteristics with zero steady-state error over a wide range of operating points.

Echo State versus Wavelet Neural Networks: Comparison and Application to Nonlinear System Identification

This paper examines two popular neural networks that have been successfully utilized in a wide variety of applications: echo state networks (ESN) and wavelet neural networks (WNN). It introduces innovations in the structure of ESN that result in major improvements in their performance. By adding a feedforward component to the networks, and by tuning their weights using recursive least squares, we can drastically reduce the number of hidden layer neurons while reducing their error. We demonstrate the improvement in the performance of ESN and compare their performance to that of WNN by application to the identification of an un-manned vehicle. Different ESN and WNN are utilized to identify the model of the vehicle and examine the effects of adding a feedforward term to ESN and WNN. Simulation results shows that adding a feedforward term to ESN and WNN can reduce the number of hidden nodes drastically while reducing the identification error of ESN. Because the number of internal units in ESN can be very high for nonlinear systems, adding this term makes ESN more suitable for online identification and control applications. Our results show that, even though both networks yield acceptable performance, ESN outperform WNN in terms of accuracy but have a slightly higher computational cost.

A fractional-order controller for single-link flexible robots robust to sensor disturbances

This work presents a new methodology for the design of the positioning control of single-link flexible manipulators which present only one vibration mode in their dynamics. In this work, the standard phase-lead compensator will be extended to a fractional-order compensator which will be used to minimize the vibrations of the flexible-link. The control objective is the precise positioning of the link tip by combining a precise motor positioning as well as reducing large part of the flexible-link vibration. The control strategy is based on combining a feedforward term and a feedback fractional-order controller. The methodology proposed in this work deals with two perturbations which are usually present in strain gauges, which are sensor habitually used to close feedback control loops of flexible link manipulators. These perturbations are: 1) the high frequency noise and 2) the offset of the signal. Experimental results illustrate the performance of the proposed controller and the advantages that this new design methodology provides.

Robust trajectory tracking control of cable-driven parallel robots

In this paper, a robust tracking controller is designed for fully constrained cable-driven parallel robots (CDPRs). One of the main challenges of controller design for this type of robotic systems is that the cables should always be in tension, where this tension is generally generated through an actuator mechanism coupled with gearboxes. On the other hand, the presence of parametric and nonparametric modeling uncertainties is a common problem in designing a precise nonlinear tracking controller for these manipulators. To deal with these problems, in this paper two separate controllers are designed for the subsystems of the robot. First, an adaptive robust feedback controller with an adaptive feedforward term is designed for the dynamics of the CDPR, constituting the outer-loop dynamics. This controller is robust with respect to the modeling uncertainties of the system. Furthermore, the output of this controller is bounded, which guarantees a saturated desired input for the inner-loop dynamics. Next, a high-gain robust controller is developed for the inner-loop dynamics, which include the actuator–gearbox model. The stability of the overall system is analyzed through a theory of cascaded systems, and it is shown that the system is uniformly practically asymptotically stable. Finally, the effectiveness of the proposed control scheme is validated through simulations on a 4-cable planar robot in both nominal and perturbed conditions.

Active neural predictive control of seismically isolated structures

An online identification and control scheme based on a wavelet neural network (WNN) and model predictive control (MPC) are presented. The WNN comprises a backpropagation neural network with wavelet activation functions and a parallel feedforward term. The WNN is used to identify the structural system, and the model is used to provide the predictions for MPC. The backpropagation network parameters and the controller are trained by the gradient descent algorithm to minimize performance indices. The feedforward component is trained using recursive least squares. The latter is found to drastically reduce the number of hidden layer neurons and significantly reduce the computational load of the neural network. Due to the general structure of the controller, its performance is satisfactory even under the strict condition imposed by a fixed learning rate. The efficacy of the control was demonstrated through a series of computational simulations of a 5-story seismically isolated structure with conventional lead-rubber bearings. Significant reductions of all response amplitudes were achieved for both near-field (pulse) and far-field ground motions, including reduced deformations along with corresponding reduction in acceleration response. In particular, the controller effectively regulated the apparent stiffness at the isolation level.

Grid integration of single stage SPV-STATCOM using cascaded 7-level VSC

A multi-functional single-stage grid-tied solar photovoltaic (SPV) system with STATCOM (Static Compensator) capabilities using a cascaded three phase seven level voltage source converter (VSC) is presented in this paper. PS-PWM (Phase Shifted Pulse Width Modulation) technique with a low switching frequency (450Hz) is used to operate the VSC. The proposed SPV-STATCOM system works in three modes i.e. in Mode-1, only active power is supplied to the grid; in Mode-2, both active and reactive powers are supplied to the grid and in Mode-3, only reactive power is supplied to the grid thereby utilizing the proposed system to its fullest capacity in 24h of a day. To extract the maximum power from the SPV array, the incremental conductance maximum power point tracking scheme is utilized. To synchronize the SPV-STATCOM power to the grid and to maintain power factor close to unity, a decoupled current controller, feed-forward term and positive sequence detector dq phase locked loop (PSD-dqPLL) control approach are used. Lower switching losses, harmonic distortions, high output voltage and power are some of the advantages of using a single-stage 7-level cascaded H-bridges. The design and the control scheme performances in all modes are simulated in MATLAB and validated through real time hardware in loop (HIL) system.