High-Precision Positioning Based on Neural Network Iterative Learning Control

This work focuses on the model-free iterative learning control (ILC) design with good transient performance and fast convergence rate that is employed for precision control of a wafter stage. First, the experimental setup, the mathematical model, and the desired reference are introduced. Notice that the desired reference and its internal model (IM) are known in advance. Then, based on the internal model principle (IMP), a new ILC scheme, named as 1M-based ILC, is proposed by incorporating the 1M of the reference. The proposed 1M-based ILC is model-free, that is, independent on the system model of the controlled plant. The 1M -based ILC is carefully constructed to track the reference along the time axis while tracking/rejecting repetitive variations along the iteration axis. Precisely because of the tracking capability along the time axis in every iteration, the 1M-based ILC can achieve satisfactory performance even in the 0th iteration and improve the transient performance and the convergence speed. In addition, the 2-D $H_{\infty}$ theory is used to establish the design criterion for the 1M-based ILC. Finally, in the experiments, the efficiency of the 1M-based ILC is illustrated by comparing with a P-type ILC and a model-based one.

Load simulator is a kind of hardware in the loop (HIL) simulation equipment, which is used to simulate the real load of the actuator in the laboratory environment. The main goal of the load simulator is to eliminate the surplus torque caused by the movement of the tested actuator and track the load spectrum accurately. When the load simulator tracks the high-frequency load signal, the feedback based control method has a serious lag, which can be avoided by using iterative learning control. However, iterative learning control is suitable for the case of constant reference trajectory and cannot deal with non-repetitive reference trajectory. B-spline decomposition based iterative learning control method is proposed. The reference load spectrum signal is decomposed into the basic spline signal, and the corresponding basic input is obtained by iterative learning control, the final control input is synthesized for tracking reference load spectrum. The simulation and experimental results show that compared with the traditional control method, this method can effectively improve the loading accuracy. The main contribution of this paper is to apply iterative learning control to non repetitive trajectory tracking.

While iterative learning control (ILC) and repetitive control (RC) have much ground in common, they fundamentally differ in the initial conditions at each repetition. This difference has lead to distinct analysis techniques, hereby clouding the interrelations between both control strategies. To facilitate the transfer of results, this paper presents a unified approach to ILC and RC. Both control problems are formulated in the trial domain using so-called system lifting. For a given system, the corresponding ILC and RC trial-domain models differ, and a thorough system theoretic analysis and comparison of these models is performed. To illustrate the value of a unified formulation of ILC and RC, the analysis of the most commonly used ILC and RC structures is harmonized. This analysis reveals central differences and similarities between various stability, monotonic convergence and steady-state performance conditions.

Iteration dependent interval based open‐closed‐loop iterative learning control for time varying systems with vector relative degree

A Multi-Input Single-Output iterative learning control for improved material placement in extrusion-based additive manufacturing

In this paper, an adaptive iterative learning control (AILC) method combined with sliding mode technique is proposed to improve the force control performance for repeating tasks of fluidic muscle (FM) driven parallel manipulators. Different from the traditional iterative learning control method, the proposed AILC is to learn the controller time-varying parameters rather than to learn the control signals. Since the AILC is sensitive to non-repetitive disturbances, the sliding mode technique is introduced to enhance the robustness. Since no model information involved in the controller design, the proposed method is a complete data-driven method. Hence, the difficulty of obtaining accurate model is avoided. Simulation studies are performed on a two degrees of freedom FM driven parallel manipulator. Simulation results demonstrate that the proposed method can achieve high force tracking performance and robustness.

Control engineering approaches may be applied to irrigation management to make better use of available irrigation water. These methods of irrigation decision-making are being developed to deal with spatial and temporal variability in field properties, data availability and hardware constraints. One type of control system is advanced process control which, in an irrigation context, refers to the incorporation of multiple aspects of optimisation and control. Hence, advanced process control is particularly suited to the management of site-specific irrigation. This paper reviews applications of advanced process control in irrigation: mathematical programming, linear quadratic control, artificial intelligence, iterative learning control and model predictive control. From the literature review, it is argued that model-based control strategies are more realistic in the soil–plant–atmosphere system using process simulation models rather than using ‘black-box’ crop production models. It is also argued that model-based control strategies can aim for specific end of season characteristics and hence may achieve optimality. Three control systems are identified that are robust to data gaps and deficiencies and account for spatial and temporal variability in field characteristics, namely iterative learning control, iterative hill climbing control and model predictive control: from consideration of these three systems it is concluded that the most appropriate control strategy depends on factors including sensor data availability and grower’s specific performance requirements. It is further argued that control strategy development will be driven by the available sensor technology and irrigation hardware, but also that control strategy options should also drive future plant and soil moisture sensor development.

The control strategies for batch processes in the past are usually categorized into two levels. The higher level is economic optimization running at a low frequency and the lower one tracks the reference given at the higher level using MPC or PID. The lower level regards all of the disturbances as something to reject using a quadratics-based optimization objective. However, not all of the disturbances are unfavorable to batch processes; some of them would be helpful. In this paper, an economic optimization for batch processes is directly applied at the lower level. It replaces the conventional tracking strategy. With the collection of the information on disturbances in the previous batches, the iterative learning control strategy (ILC) can determine better operation profiles. ILC has the advantage of continuously improving the economic performance of the current batch with enriched information on disturbances from batch to batch. The convergence of the proposed ILC-based economic optimization is proved. To demonstrate the potential applications of the proposed design method, a typical dynamic batch reactor is applied.

In order to improve the convergence speed of iterative learning control and reduce the fluctuation of the system error, a class of linear steady-state systems is considered. The convergence of the algorithm and error fluctuations are studied by introducing the variable-gain idea into the D-type iterative learning control algorithm with variable forgetting factor. According to the related properties of the λ norm theory, the convergence of the improved iterative learning algorithm is proved. Compared with iterative learning control with forgetting factor and iterative learning control with variable gain, MATLAB simulation analysis is performed. The simulation results show that the algorithm is effective. The improved iterative learning law not only makes the iterative error smoother, but also improves the convergence speed.

Significant advances in smart material actuators have taken place in the past decade. The holy grail of actuator research is an architecture that can generate high displacement and force throughout a broad frequency range while not consuming a significant amount of electrical power. The large appeal of using smart material actuators stems from their high mechanical energy density. However, all smart material actuators generally have at least one shortcoming involving either mechanical stroke, force, or frequency capability. Whenever speed is a consideration, piezoelectric actuation is the most commonly employed. The purpose of this paper is to review the most current trends in piezoelectric actuation architectures. The paper does not present the theoretical details of each actuator, but instead strives to highlight the novel concepts used in each design to overcome the stroke limitation of the material.

Optimal Iterative Learning Control for Batch Processes Based on Linear Time-varying Perturbation Model

Collective point-to-point iterative learning control of multi-agent system with switched reference

This paper presents RoFaLT, a freely-available, model-based iterative learning control (ILC) tool for nonlinear systems, that strives to close the gap between the theory of nonlinear ILC and successful applications. By providing a simple yet powerful syntax, all phases of the design of a nonlinear ILC - from modeling, tuning and execution, to analysis - are supported. RoFaLT implements an optimization-based two-step approach that yields fast convergence rates for nonlinear systems and is easily tunable to trade convergence speed off for robustness. To demonstrate the efficiency of RoFaLT, an application to a tracking problem of a race car is considered, where a simple kinematic bicycle model is used to iteratively learn the control of complex vehicle dynamics. The results show fast convergence and the effect of different tunings is shown.

Hydraulic press as a kind of very important press machining equipment has been widely used in national defense and civilian industry. This paper developed a new kind of hydraulic press that synthesizes the advantages of both hydraulic and SRM (Switched Reluctance Motor) driving technology. Aiming at the serious nonlinearity and time-variability existing in electro-hydraulic servo system of pump-controlled cylinder, the ILC (Iterative Learning Control) method was applied to track the shift curve. The fuzzy strategy was utilized to adjust the iterative learning gain adaptively in order to improve the convergence speed and precision of ILC. The simulation and experimental results show that the iterative learning speed is raised enormously by this method, and tracking of the slider shift curve with high response speed and control accuracy is realized.

This work investigates robust monotonic convergent iterative learning control (ILC) for uncertain linear systems in both time and frequency domains, and the ILC algorithm optimizing the convergence speed in terms ofl2norm of error signals is derived. First, it is shown that the robust monotonic convergence of the ILC system can be established equivalently by the positive definiteness of a matrix polynomial over some set. Then, a necessary and sufficient condition in the form of sum of squares for the positive definiteness is proposed, which is amendable to the feasibility of linear matrix inequalities. Based on such a condition, the optimal ILC algorithm that maximizes the convergence speed is obtained by solving a set of convex optimization problems. Moreover, the order of the learning function can be chosen arbitrarily so that the designers have the flexibility to decide the complexity of the learning algorithm.