Multivariable Iterative Learning Control Design for Precision Control of Flexible Feed Drives.

Advancements in machining technology demand higher speeds and precision, necessitating improved control systems in equipment like CNC machine tools. Due to lead errors, structural vibrations, and thermal deformation, commercial CNC controllers commonly use rotary encoders in the motor side to close the position loop, aiming to prevent insufficient stability and premature wear and damage of components. This paper introduces a multivariable iterative learning control (MILC) method tailored for flexible feed drive systems, focusing on enhancing dynamic positioning accuracy. The MILC employs error data from both the motor and table sides, enhancing precision by injecting compensation commands into both the reference trajectory and control command through a norm-optimization process. This method effectively mitigates conflicts between feedback control (FBC) and traditional iterative learning control (ILC) in flexible structures, achieving smaller tracking errors in the table side. The performance and efficacy of the MILC system are experimentally validated on an industrial biaxial CNC machine tool, demonstrating its potential for precision control in modern machining equipment.

Robust model-free adaptive iterative learning control for an autonomous bus trajectory tracking system.

A robust model-free adaptive iterative learning control (R-MFAILC) algorithm is proposed in this work to address the issue of laterally controlling an autonomous bus. First, according to the periodic repetitive working characteristics of autonomous buses, a novel dynamic linearized method used in the iterative domain is utilized, and a time-varying data model with a pseudo gradient (PG) is given. Then, the R-MFAILC controller is designed with a proposed adaptive attenuation factor. The proposed algorithm's advantage lies in the R-MFAILC controller, which solely utilizes the input and output data of the regulated entity. Moreover, the R-MFAILC controller has strong robustness and can handle the nonlinear measurement disturbances of the system. In simulations based on the Truck-Sim simulation platform, the effectiveness of the proposed algorithm is verified. A rigorous mathematical analysis is employed to demonstrate the stability and convergence of the proposed algorithm.

Precise tracking control via iterative learning for one-sided Lipschitz Caputo fractional-order systems.

This paper investigates iterative learning control for Caputo fractional-order systems with one-sided Lipschitz nonlinearity. Both open- and closed-loop P-type learning algorithms are proposed to achieve perfect tracking for the desired trajectory, and their convergence conditions are established. It is shown that the algorithms can make the output tracking error converge to zero along the iteration axis. A simulation example illustrates the application of the theoretical findings, and shows the effectiveness of the proposed approach.

Event-triggered based predictive iterative learning control with random packet loss compensation for nonlinear networked systems.

In this paper, a novel event-triggered predictive iterative learning control (ET-PILC) method with random packet loss compensation (RPLC) mechanism is proposed for unknown nonlinear networked systems with random packet loss (RPL). First, a new RPLC mechanism is designed by utilizing both the historical and predictive data information to avoid the deterioration of control performance due to RPL. Then, a new event-triggered condition is designed based on the proposed RPLC mechanism to save communication resources and reduce computational burden. Moreover, the convergence of the modeling error and tracking control error are analyzed theoretically, and simulation results are given to demonstrate the effectiveness of the proposed method further.

Feedforward-feedback-enhanced model-free adaptive iterative learning control with measurement disturbance and data dropout for an autonomous bus trajectory tracking system.

This article presents an innovative enhanced model-free adaptive iterative learning control approach suited for autonomous bus trajectory tracking systems that may experience measurement disruptions and random data dropouts. Data loss can occur independently and randomly at different times and in different iterations with varying probabilities, leading to successive data dropouts on both the time and iteration axes. The proposed enhanced model-free adaptive iterative learning control controller incorporates a data compensation mechanism to compensate for missing data, ensuring excellent control performance. This data-driven control strategy requires only input/output data for controller design. The convergence and effectiveness of the proposed approach are verified through rigorous mathematical analysis and simulation outcomes.

Accelerating P300-based neurofeedback training for attention enhancement using iterative learning control: a randomised controlled trial.

Objective. Neurofeedback (NFB) training through brain-computer interfacing has demonstrated efficacy in treating neurological deficits and diseases, and enhancing cognitive abilities in healthy individuals. It was previously shown that event-related potential (ERP)-based NFB training using a P300 speller can improve attention in healthy adults by incrementally increasing the difficulty of the spelling task. This study aims to assess the impact of task difficulty adaptation on ERP-based attention training in healthy adults. To achieve this, we introduce a novel adaptation employing iterative learning control (ILC) and compare it against an existing method and a control group with random task difficulty variation.Approach. The study involved 45 healthy participants in a single-blind, three-arm randomised controlled trial. Each group underwent one NFB training session, using different methods to adapt task difficulty in a P300 spelling task: two groups with personalised difficulty adjustments (our proposed ILC and an existing approach) and one group with random difficulty. Cognitive performance was evaluated before and after the training session using a visual spatial attention task and we gathered participant feedback through questionnaires.Main results. All groups demonstrated a significant performance improvement in the spatial attention task post-training, with an average increase of 12.63%. Notably, the group using the proposed iterative learning controller achieved a 22% increase in P300 amplitude during training and a 17% reduction in post-training alpha power, all while significantly accelerating the training process compared to other groups.Significance. Our results suggest that ERP-based NFB training using a P300 speller effectively enhances attention in healthy adults, with significant improvements observed after a single session. Personalised task difficulty adaptation using ILC not only accelerates the training but also enhances ERPs during the training. Accelerating NFB training, while maintaining its effectiveness, is vital for its acceptability by both end-users and clinicians.

Latent Representation-Based Learning Controller for Pneumatic and Hydraulic Dual Actuation of Pressure-Driven Soft Actuators.

The pneumatic and hydraulic dual actuation of pressure-driven soft actuators (PSAs) is promising because of their potential to develop novel practical soft robots and expand the range of soft robot applications. However, the physical characteristics of air and water are largely different, which makes it challenging to quickly adapt to a selected actuation method and achieve method-independent accurate control performance. Herein, we propose a novel LAtent Representation-based Feedforward Neural Network (LAR-FNN) for dual actuation. The LAR-FNN consists of an autoencoder (AE) and a feedforward neural network (FNN). The AE generates a latent representation of a PSA from a 30-s stairstep response. Subsequently, the FNN provides an individual inverse model of the target PSA and calculates feedforward control input by using the latent representation. The experimental results with PSAs demonstrate that the LAR-FNN can meet the requirements of dual actuation control (i.e., accurate control performance regardless of the actuation method with a short adaptation time) with a single neural network. The results suggest that a LAR-FNN can contribute to soft dual-actuation robot development and the field of soft robotics.

A survey of methods for handling initial state shifts in iterative learning control.

This paper introduces three types of controllers: a PID-type iterative learning controller, an adaptive iterative learning controller, and an optimal iterative learning controller, and reviews the history and research status of initial shifts rectifying algorithms. Initial state shifts have attracted research attention because they affect both the tracking performance and system stability. This study focuses on the current common initial shifts rectifying methods and analyzes the underlying mechanism in detail. To verify the effectiveness of the presented initial shifts rectifying algorithms, we simulated those using ideal first- and second-order systems. Finally, directions for the future development of iterative learning control (ILC) and some challenging topics related to initial shifts rectifying for ILC are presented. This article aims to introduce recent developments and advances in initial shifts rectifying algorithms and discuss the directions for their further exploration.

Design of robust fuzzy iterative learning control for nonlinear batch processes.

In this paper, a two-dimensional (2D) composite fuzzy iterative learning control (ILC) scheme for nonlinear batch processes is proposed. By employing the local-sector nonlinearity method, the nonlinear batch process is represented by a 2D uncertain T-S fuzzy model with non-repetitive disturbances. Then, the feedback control is integrated with the ILC scheme to be investigated under the constructed model. Sufficient conditions for robust asymptotic stability and 2D $ H_\infty $ performance requirements of the resulting closed-loop fuzzy system are established based on Lyapunov functions and some matrix transformation techniques. Furthermore, the corresponding controller gains can be derived from a set of linear matrix inequalities (LMIs). Finally, simulations on the three-tank system and the highly nonlinear continuous stirred tank reactor (CSTR) are carried out to prove the feasibility and efficiency of the proposed approach.

PMSM Torque Ripple Suppression Method Based on SMA-Optimized ILC.

Periodic torque ripple often occurs in permanent magnet synchronous motors due to cogging torque and flux harmonic distortion, leading to motor speed fluctuations and further causing mechanical vibration and noise, which seriously affects the performance of the motor vector control system. In response to the above problems, a PMSM torque ripple suppression method based on SMA-optimized ILC is proposed, which does not rely on prior knowledge of the system and motor parameters. That is, an SMA is used to determine the optimal values of the key parameters of the ILC in the target motor control system, and then the real-time torque deviation value calculated by iterative learning is compensated to the system control current set end. By reducing the influence of higher harmonics in the control current, the torque ripple is suppressed. Research results show that this method has high efficiency and accuracy in parameter optimization, further improving the ILC performance, effectively reducing the impact of higher harmonics, and suppressing the torque ripple amplitude.

Fractional-order iterative learning control for fractional-order systems with initialization non-repeatability.

The effect of initialization non-repeatability on iterative learning control performance for fractional-order systems has not been sufficiently investigated. It is a hidden deficiency that leads directly to the breaking of perfect tracking conditions in both theoretical analysis and real-world applications. Therefore, under the framework of general fractional-order nonlinear systems, this paper proposes an open-close loop Dα-type iterative learning control scheme based on system preconditioning and strictly derives two convergence conditions. By applying the preconditioning optimization strategy based on the short-memory principle, the tracking error due to initialization nonrepetition can converge to any desired range. Compared with the existing results, the proposed iteration scheme fully considers the complexity of the initialization and initial conditions of fractional-order systems, and provides several practical preconditioning methods to improve the tracking efficiency. Two numerical examples are presented to validate the above conclusions.

Extended state observer based dynamic iterative learning for trajectory tracking control of a six-degrees-of-freedom manipulator.

With the development of industrial automation comes an ever, broadening number of application scenarios for manipulators along with increasing demands for their precise control. However, manipulator trajectory tracking control schemes often exhibit problems such as those related to high levels of coupling, complex calculations, and in various difficulties in application for industrial environments. For the problems of low accuracy in control and poor robustness of multiple-jointed robotic trajectory tracking, iterative learning control (ILC) with model compensation (MC) based on extended state observer (ESO) has been proposed for the trajectory tracking control of six-degrees-of-freedom (six-DOF) manipulators. The scheme has excellent features to overcome uncertainties in repetitive tasks, including unknown bounded perturbations that are external to the model or dynamic perturbations that are internal to the model. The proposed control strategy combines ESO, iterative learning, and MC, for precise control of trajectory tracking. Here, ESO is used to estimate disturbances, iterative learning allows fast and accurate control in repeated tasks, and the model-compensated control algorithm alleviates the necessary for many inverse operations. The convergence of our proposed control scheme is proved through Lyapunov function and time-varying approximation theory. Simulation and experimental results verify the validity of the proposed scheme.

Safe deep reinforcement learning in diesel engine emission control.

A deep reinforcement learning application is investigated to control the emissions of a compression ignition diesel engine. The main purpose of this study is to reduce the engine-out nitrogen oxide (NOx) emissions and to minimize fuel consumption while tracking a reference engine load. First, a physics-based engine simulation model is developed in GT-Power and calibrated using experimental data. Using this model and a GT-Power/Simulink co-simulation, a deep deterministic policy gradient is developed. To reduce the risk of an unwanted output, a safety filter is added to the deep reinforcement learning. Based on the simulation results, this filter has no effect on the final trained deep reinforcement learning; however, during the training process, it is crucial to enforce constraints on the controller output. The developed safe reinforcement learning is then compared with an iterative learning controller and a deep neural network-based nonlinear model predictive controller. This comparison shows that the safe reinforcement learning is capable of accurately tracking an arbitrary reference input while the iterative learning controller is limited to a repetitive reference. The comparison between the nonlinear model predictive control and reinforcement learning indicates that for this case reinforcement learning is able to learn the optimal control output directly from the experiment without the need for a model. However, to enforce output constraint for safe learning reinforcement learning, a simple model of system is required. In this work, reinforcement learning was able to reduce NOx emissions more than the nonlinear model predictive control; however, it suffered from slightly higher error in load tracking and a higher fuel consumption.

Hybrid and adaptive control of functional electrical stimulation to correct hemiplegic gait for patients after stroke.

Introduction: Gait, as a fundamental human movement, necessitates the coordination of muscles across swing and stance phases. Functional electrical stimulation (FES) of the tibialis anterior (TA) has been widely applied to foot drop correction for patients with post-stroke during the swing phase. Although the gastrocnemius (GAS) during the stance phase is also affected, the Functional electrical stimulation of the gastrocnemius received less attention. Methods: To address this limitation, a timing- and intensity-adaptive Functional electrical stimulation control strategy was developed for both the TA and GAS. Each channel incorporates a speed-adaptive (SA) module to control stimulation timing and an iterative learning control (ILC) module to regulate the stimulation intensity. These modules rely on real-time kinematic or kinetic data during the swing or stance phase, respectively. The orthotic effects of the system were evaluated on eight patients with post-stroke foot drop. Gait kinematics and kinetics were assessed under three conditions: no stimulation (NS), Functional electrical stimulation to the ankle dorsiflexor tibialis anterior (SA-ILC DS) and FES to the tibialis anterior and the ankle plantarflexor gastrocnemius (SA-ILC DPS). Results: The ankle plantarflexion angle, the knee flexion angle, and the anterior ground reaction force (AGRF) in the SA-ILC DPS condition were significantly larger than those in the NS and SA-ILC DS conditions (p < 0.05). The maximum dorsiflexion angle during the swing phase in the SA-ILC DPS condition was similar to that in the SA-ILC DS condition, with both being significantly larger than the angle observed in the NS condition (p < 0.05). Furthermore, the angle error and force error relative to the set targets were minimized in the SA-ILC DPS condition. Discussion: The observed improvements can be ascribed to the appropriate stimulation timing and intensity provided by the SA-ILC DPS strategy. This study demonstrates that the hybrid and adaptive control strategy of functional electrical stimulation system offers a significant orthotic effect, and has considerable potential for future clinical application.

Q-learning based fault estimation and fault tolerant iterative learning control for MIMO systems.

This paper proposes a Q-learning based fault estimation (FE) and fault tolerant control (FTC) scheme under iterative learning control (ILC) framework. Due to the repetitive demands on control actuators for repetitive tasks, ILC is sensitive to actuator faults. Moreover, unknown faults varying with both time and trial axes pose a challenge to the control performance of ILC. This paper introduces Q-learning algorithm for FE to continuously adjust the estimator and adapt the changing faults. Then, FTC is designed by adopting the norm-optimal iterative learning control (NOILC) framework, where the controller is adjusted based on the FE results from Q-learning to counteract the influence of faults. Finally, the simulation on the plant of a mobile robot verifies the effectiveness of the proposed algorithm.

A 2-dimensional model framework for blood glucose prediction based on iterative learning control architecture.

The accurate, timely, and personalized prediction for future blood glucose (BG) levels is undoubtedly needed for further advancement of diabetes management technologies. Human inherent circadian rhythm and regular lifestyle resulting in similarity of daily glycemic dynamics play a positive role in the prediction of blood glucose. Inspired by the iterative learning control (ILC) method in the field of automatic control, a 2-dimensional (2-D) model framework is constructed to predict the future blood glucose levels by taking both the short-range information within a day (intra-day) and long-range information between days (inter-day) into account. In this framework, the radial basis function neural network was applied to capture nonlinear relationships in glycemic metabolism, that is, short-range temporal dependence and long-range contemporaneous dependence on previous days. We build models for each patient, and the models were tested on the in silico datasets at various prediction horizons (PHs). The learning model developed in the 2-D framework successfully increases the accuracy and reduces the delay of predictions. This modeling framework provides a new point of view for BG level prediction and contributes to the development of personalized glucose management, such as hypoglycemia warning and glycemic control.

Waveform Design Method for Piezoelectric Print-Head Based on Iterative Learning and Equivalent Circuit Model.

Piezoelectric print-heads (PPHs) are used with a variety of fluid materials with specific functions. Thus, the volume flow rate of the fluid at the nozzle determines the formation process of droplets, which is used to design the drive waveform of the PPH, control the volume flow rate at the nozzle, and effectively improve droplet deposition quality. In this study, based on the iterative learning and the equivalent circuit model of the PPHs, we proposed a waveform design method to control the volume flow rate at the nozzle. Experimental results show that the proposed method can accurately control the volume flow of the fluid at the nozzle. To verify the practical application value of the proposed method, we designed two drive waveforms to suppress residual vibration and produce smaller droplets. The results are exceptional, indicating that the proposed method has good practical application value.

Enhancing projection based iterative learning control: A set-membership approach.

Lithographic machine tools require both high motion accuracy and high motion flexibility. Projection based iterative learning control (P-ILC) is appealing for wafer stages to achieve two goals, simultaneously. P-ILC contains a nonparametric feedforward controller based on ILC, and a parametric feedforward controller with a projection step for feedforward tuning. In this paper, a set-membership based frequency-domain ILC algorithm (SM-F-ILC) is employed in the enhancing P-ILC scheme to improve the performance in the nonparametric feedforward control mode. SM-F-ILC can effectively compensate for repetitive errors, attenuate the nonrepetitive error accumulation and achieve fast convergence speed with model uncertainties. These superiorities also facilitate to improve the performance of P-ILC in the parametric feedforward control mode. The validity of the enhancing P-ILC scheme is demonstrated by experimental results.

Large-range high-speed dynamic-mode atomic force microscope imaging: adaptive tapping towards minimal force.

In this paper, a software-hardware integrated approach is proposed for high-speed, large-range tapping mode imaging of atomic force microscope (AFM). High speed AFM imaging is needed in various applications, particularly in interrogating dynamic processes at nanoscale such as polymer crystallization process. Achieving high speed in tapping-mode AFM imaging is challenging as the probe-sample interaction during the imaging process is highly nonlinear, making the tapping motion highly sensitive to the probe sample spacing, and thereby, difficult to maintain at high speed. Increasing the speed via hardware bandwidth enlargement, however, leads to a substantially reduction of the imaging area. Contrarily, the imaging speed can be increased without loss of the scan size through control (algorithm)-based approach. For example, the recently-developed adaptive multiloop mode (AMLM) technique has demonstrated its efficacy in increasing the tapping-mode imaging speed without loss of scan size. Further improvement, however, has been limited by the hardware bandwidth and the online signal processing speed and computation complexity involved. Thus, in this paper, the AMLM technique is further enhanced to optimize the probe tapping regulation, and integrated with a field programmable gate array platform to further increase the imaging speed without loss of quality and scan range. Experimental implementation of the proposed approach demonstrates that high-quality imaging can be achieved at a high-speed scanning rate of 100 Hz and higher, and over a large imaging area of over 20µm.

Iterative neural network adaptive robust control of a maglev planar motor with uncertainty compensation ability.

In this paper, an iterative neural network adaptive robust control (INNARC) strategy is proposed for the maglev planar motor (MLPM) to achieve good tracking performance and uncertainty compensation. The INNARC scheme consists of adaptive robust control (ARC) term and iterative neural network (INN) compensator in a parallel structure. The ARC term founded on the system model realizes the parametric adaptation and promises the closed-loop stability. The INN compensator based on the radial basis function (RBF) neural network is employed to handle the uncertainties resulted from the unmodeled non-linear dynamics in the MLPM. Additionally, the iterative learning update laws are introduced to tune the network parameters and weights of the INN compensator simultaneously, so the approximation accuracy is improved along the system repetition. The stability of the INNARC method is proved via the Lyapunov theory, and the experiments are conducted on an home-made MLPM. The results consistently demonstrate that the INNARC strategy possesses the satisfactory tracking performance and uncertainty compensation, and the proposed INNARC is an effective and systematic intelligent control method for MLPM.