Feedforward Control Input Research Articles

Molten metal’s reaction force measurement for pressure estimation and control system construction in press casting

<p class="Abstract"><span lang="EN-US">This paper presents a measurement system for molten metal’s reaction force and estimation of liquid pressure during pressing to control the iron product quality. We have developed a new type of casting process. In the process, molten metal is quickly filled into casting molds by high-speed pressing. Casting defects such as physical metal penetration is often caused by excess pressure. Hence, we have constructed a pressure control system using a mathematical model-based off-line simulation to derive the ideal feedforward control input of pressing. However, it is difficult to accurately control the pressure in cases of varying conditions such as liquid volume and temperature changes. Also, pressure measurements by using contact-type sensors directly is impossible for molten metal, because of the high temperature of the liquid, over 1400 °C. So, we have proposed a new pressure estimation method with force measurement data processing. Here, the exact reaction force from the molten metal must be accurately observed by a force sensor set between the upper mold and its elevating device. The viscosity coefficient can also be calculated on a real-time basis. The proposed force measurement system will realize an improved casting quality due to the effective feed-back control system.</span></p>

Constrained optimal iterative learning control with mixed-norm cost functions

Iterative learning Control (ILC) is a widely used design technique for determining feedforward control inputs to systems that perform the same task repeatedly. This feedforward control design problem can be posed as an optimization problem in the Norm Optimal ILC (NO-ILC) framework. Although NO-ILC problems are usually formulated without constraints, they may be extended to enforce constraints by posing an analogous constrained optimization problem, termed Constrained Optimal ILC (CO-ILC). Typical NO-ILC and CO-ILC algorithms use 2-norm type cost functions (i.e., minimizing tracking error and control effort 2-norms), which are smooth and have analytical gradient expressions for design of the update law. However, in many applications, the max (∞) norm of tracking error is critical, which is a nonsmooth function and thus gradient-based ILC methods cannot be directly used. In this manuscript, we design a CO-ILC algorithm to explore the performance of using non-smooth type cost functions and compare them with the case using 2-norm cost function with actuator (or more generally state) box constraints. Specifically, CO-ILC algorithms for linear system with linear constraints are derived using (1) an ∞-norm cost function, (2) a mixed (2−∞)-norm cost function and (3) a sequential (2−∞)-norm cost function; the performance results of these are then compared with the traditional CO-ILC using 2-norm cost function on a precision motion control stage experimentally. We also provide proofs for robust monotone convergence of the proposed CO-ILC algorithms for a class of uncertainty models.

Distributed Adaptive Consensus Output Regulation of Network-Connected Heterogeneous Unknown Linear Systems on Directed Graphs

This technical note deals with output regulation of network connected linear systems under directed connection graphs. The subsystem dynamics are unknown and heterogeneous. The desired output signal, generated by a linear exosystem, is only available to some subsystems, and the others will estimate the exosystem state through the network connections for obtaining regulation errors. The exosystem is parametrised in a specific form such that it has a skew-symmetric system matrix, whose property is further explored in the estimation of exosystem state, and in the estimation of the desired feed-forward control inputs for output regulation. The unknown subsystem parameters are dealt with by the adaptive laws, driven by the estimated regulation errors. The proposed adaptive control strategy is fully distributed, in the sense that the control design only uses the information in the connected neighbourhood, without reference to the eigenstructure of the associated Laplacian matrix, as long as the connection graph is strongly connected.

Model based burnt gas fraction controller design of diesel engine with VGT/Dual loop EGR system

-Recently, regulation of NOx and PM emission in diesel engines has become stricter and the EGR system has been expanded into a dual loop EGR system to increase EGR rate as well as to utilize exhaust gas strategically. In terms of engine combustion characteristics, burnt gas fraction is becoming an important factor of solving the NOx and PM emission reduction problem more efficiently but conventional controller focused only pressure and air flow rate targets. Unlike the previous studies, this paper describes a model based burnt gas fraction control structure for a diesel engine with a dual loop EGR and a turbocharger. Feedforward control inputs based on burnt gas fraction states aids in the precise control of diesel engines, especially in transient states by considering coupled behavior within the system. For the controller validation, a control oriented reduced order model of a diesel engine air management system is established to simplify the control input computation and its stability is proved by analysing the internal dynamics stability. Then, a sliding mode controller is designed and controller robustness at certain operating points is validated using an HiLS bench.

Feedforward dynamics for the control of articulated multi-limb robots

We describe a general approach for using linearizing feedforward control inputs for large degree of freedom (dof) multi-limb robots operating in scenarios involving motion and force constraints, and under-actuated degrees of freedom arising from the task and the environment. Our solution is general and has low computational cost needed for real-time control loops. It supports the tuning of the feedforward term to meet multiple task objectives. Being structure-based, it is able to easily accommodate changes in motion and force constraints that often occur in robotics scenarios.

OPTIMAL CONTROL STRATEGY FOR LOW SPEED AND HIGH SPEED FOUR-WHEEL-ACTIVE STEERING VEHICLE

In this work, based on the optimal control theory approach, a four-wheel-active steering (4WAS) system is proposed for low speed and high speed applications. A model following the control structure is adopted consisting of a feed-forward and feedback compensation strategy that serves as correction inputs to enhance the vehicle's dynamic behavior. The velocity dependent feed-forward control inputs are based on the driver's steering intention while the feedback control inputs are based on the vehicle's state feedback errors, being the sideslip and yaw rate of the vehicle. Numerical simulations are conducted using the Matlab/Simulink platform to evaluate the control system's performance. The performance of the 4WAS controller is tested in two designated open loop tests, being the constant steer and the lane change maneuver, to evaluate its effectiveness. A comparison with conventional passive front-wheel-steering (FWS) and conventional four-wheel-steering (4WS) systems shows the preeminent result performance of the proposed control strategy in terms of the response tracking capability and versatility of the controller to adapt to the system's speed environment. In high speed maneuvers, the improvement in terms of yaw rate tracking error in rms is evaluated and the proposed active steering system considerably beat the other two structures with 0.2% normalized error compared to the desired yaw rate response. Meanwhile, in low speed, turning radius reductions of 25% and 50% with respect to the capability of normal or typical FWS vehicles are successfully achieved.

Optimal control of nonlinear continuous-time systems in strict-feedback form.

This paper proposes a novel optimal tracking control scheme for nonlinear continuous-time systems in strict-feedback form with uncertain dynamics. The optimal tracking problem is transformed into an equivalent optimal regulation problem through a feedforward adaptive control input that is generated by modifying the standard backstepping technique. Subsequently, a neural network-based optimal control scheme is introduced to estimate the cost, or value function, over an infinite horizon for the resulting nonlinear continuous-time systems in affine form when the internal dynamics are unknown. The estimated cost function is then used to obtain the optimal feedback control input; therefore, the overall optimal control input for the nonlinear continuous-time system in strict-feedback form includes the feedforward plus the optimal feedback terms. It is shown that the estimated cost function minimizes the Hamilton-Jacobi-Bellman estimation error in a forward-in-time manner without using any value or policy iterations. Finally, optimal output feedback control is introduced through the design of a suitable observer. Lyapunov theory is utilized to show the overall stability of the proposed schemes without requiring an initial admissible controller. Simulation examples are provided to validate the theoretical results.

Robust control of permanent magnet synchronous motors

In this paper, permanent magnet synchronous motors (PMSMs) are investigated. According to the feature of PMSMs, a novel state equation of PMSMs is obtained by choosing suitable state variables. Based on the state equation, robust controllers are designed via interval matrix and PI control idea. In terms of bilinear matrix inequations, sufficient conditions for the existence of the robust controller are derived. In order to reduce the conservation and the dependence on parameter, the control inputs of PMSMs are divided into two parts, a feedforward control input and a feedback control input, and relevant sufficient conditions for the existence of the controller are obtained. Because of the suitable choice of state variables, the proposed control strategies can cope with the load uncertainty and have robustness for disturbance. Finally, simulations are carried out via Matlab/Simulink soft to verify the effectiveness of the proposed control strategies. The performance of the proposed control strategies are demonstrated by the simulation results.

Control system modeling of stock management for civil infrastructure

Management of infrastructure stock is essential in sustainability of society, and its analysis and optimization are studied in the light of control system modeling in this paper. At the first part of the paper, cost of stock management is analyzed based on macroscopic statistics on infrastructure stock and economical growth. Stock management burden relative to economy is observed to become larger at low economic growth periods in developed economies. Then, control system modeling of stock management is introduced and by augmenting maintenance actions as control input, dynamic behavior of stock is simulated and compared with existing time history statistics. Assuming steady state conditions, applicability of the model to cross sectional data is also demonstrated. The proposed model is enhanced so that both preventive and corrective maintenance can be included as system inputs, i.e., feedforward and feedback control inputs. Optimal management strategy to achieve specified deteriorated stock level with minimal cost, expressed in terms of preventive and corrective maintenance actions, is derived based on estimated parameter values for corrosion of steel bridges. Relative cost effectiveness of preventive maintenance is shown when target deteriorated stock level is lower.

1P1-X02 最大筋力を考慮した筋骨格システムの位置制御

This paper proposes a set-point control method for a musculoskeletal system considering maximum muscular force. In our previous work, the set-point control method for a musculoskeletal system was proposed which combines a feedforward control with a sensory feedback control including a considerable time-delay. However, the maximum output force of each muscle have not been considered though it sometimes induces a wind-up phenomenon because of the saturation of the muscular force. The ratio function to determine the ratio of feedforward and feedback control inputs is newly introduced in this paper to prevent such a wind-up phenomenon without the saturation of muscular force, and generate more sophisticated reaction motion even if disturbances are added on the end-point during movement. Numerical simulation results are shown to demonstrate the effectiveness of the proposed method.

Experimental Verification of Adaptive Dominant Type Hybrid Adaptive and Learning Controller for Trajectory Tracking of Robot Manipulators

This paper presents an experimental study to verify an adaptive dominant type hybrid adaptive and learning controller for acquiring an accurate trajectory tracking of periodic desired trajectory of robot manipulators. The proposed controller is developed based on combining the model-based adaptive control (MBAC), repetitive learning control (RLC) and proportionalderivative (PD) control in which the MBAC input becomes dominant than other inputs. Dominance of adaptive control input gives the advantage that the proposed controller could adjust the feed-forward motion control input immediately after changing the desired motion or load of the manipulator. In motion control law, the proposed controller uses only one vector to estimate the unknown dynamical parameters. It makes the proposed controller as a simpler hybrid adaptive and learning controller which does not need much computational power and also is easily be implemented for real applications of robot manipulators. The proposed controller is verified through experiments on a four-link small robot manipulator as representation of a scale robot manipulator to ensure this controller can be applied in the real applications of robot manipulators. The experimental results show the effectiveness of the proposed controller by indicating the position tracking error approaches to zero.

An Adaptive Controller Dominant-Type Hybrid Adaptive and Learning Controller for Trajectory Tracking of Robot Manipulators

This paper proposes a new hybrid adaptive and learning control method based on combining model-based adaptive control, repetitive learning control (RLC) and proportional–derivative control to consider the periodic trajectory tracking problem of robot manipulators. The aim of this study is to obtain a high-accuracy trajectory tracking controller by developing a simpler adaptive dominant-type hybrid controller by using only one vector for estimation of the unknown dynamical parameters in the control law. The RLC input is adopted using the original learning control law, adding a forgetting factor to achieve the convergence of the learning control input to zero. We will improve and prove that the adaptive dominant-type controller could be applied for tracking a periodic desired trajectory in which adaptive control input increases and becomes dominant of the control input, whereas the other control inputs decrease close to zero. The domination of the adaptive control input gives the advantage that the proposed controller could adjust the feed-forward control input immediately and it does not spend much time relearning the learning control input when the periodic desired trajectory is switched over from the first trajectory to another trajectory. We utilize the Lyapunovlike method to prove the stability of the proposed controller and computer simulation results to validate the effectiveness of the proposed controller in achieving the accurate tracking to the periodic desired trajectory.

Control of a flexible beam actuated by macro-fiber composite patches: I. Modeling andfeedforward trajectory control

This paper considers a systematic approach for motion planning and feedforward controldesign for a flexible cantilever actuated by piezoelectric macro-fiber composite (MFC)patches. For accurate feedforward tracking control, special attention has to be paid to theinherent nonlinear hysteresis and creep behavior of these actuators. In order to account forthese effects an appropriate compensator is applied which allows us to perform the trackingcontroller design on the basis of a linear infinite-dimensional model. A detailed analysisof the nonlinear actuator behavior as well as the compensator design and theoverall experimental validation is presented in the companion paper (Schröck et al 2011 Smart Mater. Struct. 20 015016). The governing equations of motion ofthe hysteresis and creep compensated cantilever are determined by means of theextended Hamilton’s principle. This allows us to consider the influence of thebonded patch actuators on the mechanical properties of the underlying beamstructure in a straightforward manner and results in a model with spatially varyingsystem parameters. For the solution of the motion planning and feedforwardcontrol problem a flatness-based methodology is proposed. In a first step, theinfinite-dimensional system of the MFC-actuated flexible cantilever is approximated by afinite-dimensional model, where all system variables, i.e. the states, input and output, canbe parameterized in terms of a so-called flat output. In a second step, it is shownby numerical simulations that these parameterizations converge with increasingsystem order of the finite-dimensional model such that the feedforward controlinput can be directly calculated in order to realize prescribed output trajectories.

Simultaneous Control for Position and Vibration of a Planetary Rover with Flexible Structures

This paper proposes a synthesis procedure for a robust control target to be utilized in the simultaneous control of position and vibration of a planetary rover with flexible structures. The target is designed as a type of feedforward control input and its profile is derived only by solving a regulator problem with an initial velocity. In the regulator problem, the non-stationary robust control method is employed to make the target profile be robust against parameter variation. Further, the design procedure for the target is independent of the positioning distance and time schedule. The robustness of the target against parameter variation is examined through numerical calculations and the results are compared with those obtained from conventional methods. As a result, the usefulness of the proposed command shaping procedure is demonstrated and the high robustness of the target is confirmed.

Temperature control of functionally graded plates using a feedforward–feedback controller based on the inverse solution and proportional-derivative controller

As a first endeavor, an approach for the two- and three-dimensional temperature control of functionally graded (FG) plates by using the inverse solution and the proportional-differential (PD) controller is provided. For this purpose, firstly, having the desired temperatures at different locations and times, heat fluxes at the boundaries of the plates are estimated by inverse solution techniques offline. Then, the estimated heat fluxes as feedforward control inputs are combined with a PD controller to introduce a hybrid feedforward–feedback control input to the FG domain in the presence of disturbance and noise. In order to show the efficiency and accuracy of the proposed (inverse + PD) controller in two- and three-dimensional domains, different distinct examples, which include different boundary conditions, material properties and disturbance sources are presented. It is shown that the presented approach can adjust heat fluxes for control of the temperature accurately; also, the PD controller gains do not need to be re-adjusted for different problems.

Feedforward Control Using Sampled-Data Polynomial for Track Seeking in Hard Disk Drives

To decrease the seek time of hard disk drives, a feedforward control method was developed by using a sampled-data polynomial. The sampled-data polynomial satisfies the boundary conditions that include the characteristics of the zero-order hold, and compensates for the discretization error caused by the zero-order hold without the need for complicated calculations. Therefore, the feedforward control using the sampled-data polynomial enables real-time calculation and does not require lookup tables which need a large amount of memory. The parameters of the sampled-data polynomial are designed by using shock-response-spectrum analysis to minimize the settling vibrations caused by the feedforward control inputs. When the proposed method was applied on a hard disk drive, it significantly reduced the amount of tracking error in the seek control and also reduced the seek time.

Iterative Identification of Feedforward Controllers for Iterative Learning Control

One the the main disadvantages to the standard iterative learning control (ILC) problem is that the learning filter is static and does not update based on the current iteration information. This means that while the learning control is updating the feedforward control input, the learning controller itself is not updating. In this paper, we present an iteration varying learning filter which is based on identification techniques. Every iteration, this learning filter is computed to be the closed loop system inverse based on a least squares approximation. We show that the iteration varying learning filter successfully reduces the 2-norm error as fast as a static, model-based learning filter and much better than a P-Type learning filter. This results suggests the benefit of extensive preliminary system identification efforts for iterative systems is limited at best.

Lane change steering manoeuvre using model predictive control theory

This paper is a fundamental study on steering manoeuvre (driver model) when the steering angle input is obtained via Model Predictive Control (MPC). Conventional front steering angle for the lane change manoeuvre uses the sine wave input. This is very simple but this feed forward control input does not have adequate performance for mounting on real vehicle. For this reason, the steering input using MPC is developed. For example, the collision avoidance of lane-change steering manoeuvre has to deal with many constraints, so that the servo control like an LQI theory is not suitable for this problem. In addition, offline nonlinear programming calculation for optimal steering input of vehicle trajectory involves too heavy computational complexity to apply with real-time controllers. Desired control of the lane-change performance is conducted by the MPC approach with the constraint conditions. Potential advantage of the collision avoidance steering compared to the hard braking is estimated.

20815 1リンク6足歩行ロボットの姿勢センサを用いた脚リンクフィードバック入力生成法(歩行ロボット,一般講演,学術講演)

This paper dealt with a walking control method of a six-legged robot consisting of one link by the feedback control input generated from the attitude of the body. The leg of the six-legged robot is semicircle arc shape and it is suitable for walking on uneven terrain. In details, the transfer function of the leg is derived from the frequency response, and the feedforward control input of the leg is obtained from the inverse function of the transfer function. The error between the reference signal and the leg controlled by the feedforward control input is corrected by the feedback signal generated from the attitude of the body.

光ディスク装置のディフェクト補償制御

In optical-disc drives (ODDs), disturbances can be roughly classified in two groups: external disturbances and defects. The first group requires a higher closed loop bandwidth than the second group. If the closed loop has a high bandwidth the controller may have a good performance against external disturbances but it might follow the defects of the disc surface. In order to make ODDs more robust against surface defects, a new control method for compensating the defects has been developed. The developed control method can estimate the plant state. The feedforward control input, which can transfer the estimated state to zero, is systematically generated by using a final-state control (FSC) method. The simulation results show that the developed control method can significantly improve the robustness against the defects.