International Journal of Control Automation and Systems最新文献

Programmable logic controllers (PLCs) are usually equipped with only basic direct control algorithms like proportional-integral-derivative (PID). Modules included in engineering software running on a personal computer (PC) are usually used to tune controllers. In this article, an alternative approach is considered, i.e. the development of a stochastic optimizer based on the (μ,λ) evolution strategy (ES) in a PLC. For this purpose, a pseudorandom number generator (pRNG) was implemented, which is not normally available in most PLCs. The properties of popular random number generation methods were analyzed in terms of distribution uniformity and possibility of implementation in a PLC. The Wichmann-Hill (WH) algorithm was chosen for implementation. The developed generator with a uniform distribution was the basis for the implementation of a generator with a normal distribution. Both generators are the engines of the stochastic optimization algorithm in the form of the (μ, λ) strategy. For verification purposes, a modular servomechanism laboratory set was used as a test object for PID and linear-quadratic regulator (LQR) control. Moreover, the possibility of using the developed optimizer was shown in an application of model predictive control (MPC). Comprehensive tests confirmed the correctness of the implementation and high functionality of the developed software. Calculation time issues are also investigated.

The running gear system provides the safety guarantee for the normal operation of high-speed trains. The massive historical data in the system can be used for fault detection and diagnosis. This data inevitably exists redundancy, which makes the valuable data not fully utilized in the process of extracting latent variables. In this paper, to make full and effective use of historical data, a dynamic weighted slow feature analysis (DWSFA) method is proposed, which can detect slow-change faults in the running gear system of high-speed trains. The proposed method based on basis functions can reduce the amount of time lags required for the process of extracting latent variables, and it obtains the better fault detection (FD) performance. The effectiveness of the proposed method is verified via a running gear system of high-speed train.

The present paper proposes a model-free extended state observer (MFESO) based model and model-free predictive control (MPC-MFESO-MFPAC) approach for achieving trajectory tracking and obstacle avoidance of unmanned surface vehicle (USV) in complex environments. MFPAC is investigated for addressing the uncertainty in the kinetics modeling of USV system, eliminating the need for an accurate mathematical model of the USV. Additionally, the backstepping method is employed to eliminate rotational characteristics, enabling direct application of MFPAC in USV control. In the computation of the virtual control law, MPC is utilized for kinematics which can be easily modeled, while incorporating obstacle avoidance performance. By utilizing the model-free ESO for estimating additional unknown perturbations, this approach obviates the need for any prior knowledge of the system’s dynamics. The stability analysis demonstrates that the proposed control strategy is bounded-input and bounded-output (BIBO) stable. The simulation results validate the effectiveness and advancement of the algorithm.

This paper is concerned with the transformed parameter-dependent H_∞ controller design for semi-Markovian jump linear parameter varying (S-MJLPV) systems under actuator saturation and faults. In the S-MJLPV system, the semi-Markov process transition rate is time-varying during the semi-Markov process and a plant includes time-varying parameters which are bounded and measurable in magnitude. For more practical analysis and synthesis of the S-MJLPV systems, a time-varying actuator fault model and actuator saturation of the controller are considered into account simultaneously. The primary goal of this paper is to develop a transformed parameter-dependent control that makes the closed-loop system stochastically stable with H_∞ performance index γ and provides less conservative results against actuator saturation and faults. Based on the mode-dependent Lyapunov function, new sufficient conditions are obtained to ensure that the stochastic stability of S-MJLPV systems. Eventually, an example based on the turbofan-engine model is presented to demonstrate the efficacy of our proposed methods.

Iterative learning control (ILC) is a popular scheme in the trajectory tracking of manipulators, greatly improving tracking accuracy despite often requiring multiple iterations over identical trajectories. This research introduces an optimization technique for ILC parameters, enhanced with proportional-derivative (PD) feedback control, which aims to significantly reduce tracking errors within a single iteration. In the proposed approach, a PD feedback controller is utilized in the first run, collecting error data. An ILC controller is then incorporated in the second run to minimize the tracking error. Utilizing the dynamic model of the system, the transcription method transforms the continuous-form optimization problem concerning the ILC parameters into a discrete form, enabling its solution via standard numerical optimization algorithms. To demonstrate the effectiveness of the proposed approach in reducing tracking errors, we compared the tracking errors for the first and second runs of the system using frequency-domain analysis and conducted simulations and experiments on two different trajectory types.

The well known back-propagation algorithm has revolutionized machine learning and artificial intelligence, particularly in neural network applications. Although gradient descent-based algorithms are utilized in control applications, they are not as prevalent as in neural network applications. This discrepancy can be attributed to the successful development of various adaptation laws which ensure system stability while meeting the required design criteria. Many of these laws can be found in model reference adaptive control (MRAC) and adaptive sliding mode control (ASMC). This paper investigates the applicability of the Brandt-Lin (B-L) learning algorithm, mathematically equivalent to the back-propagation algorithm, in adaptive control applications. We find that combining the B-L learning algorithm with SMC yields a robust controller suitable for model reference adaptive sliding mode control (MRA-SMC). The controller is applicable to linear and a class of nonlinear dynamic systems and is suitable for efficient implementation. We derive the stability criteria for this controller and conduct simulations to study the adaptation’s impact on chattering. Our work exemplifies one approach to adopt the back-propagation algorithm in control applications.

It is always problematic that the initial value of the trajectory tracking error must be inside the area included in the prescribed performance constraint function. To overcome this problem, a novel fault-tolerant control strategy is designed for a second-order multi-input and multi-output nonlinear system (MIMO-NLS) with unknown initial states, actuator faults, and control saturation. Firstly, a predefined time convergence (PTC) stability criterion is theoretically proven. Then, an error conversion function is introduced to convert the trajectory tracking error to a new error variable with an initial value of zero, and an adaptive fuzzy system is designed to approximate the compound interference composed of actuator fault, parameter perturbation, control saturated overamplitude, and external disturbance. Based on the backstepping control method, prescribed performance control method, and predefined time convergence stability theory, an adaptive fuzzy fault-tolerant controller for the new error variable is designed and theoretically proven for the predefined time convergence of the closed-loop system. The numerical simulation results of the guaranteed performance trajectory tracking control for industrial robots with actuator faults demonstrate that the adaptive fuzzy fault-tolerant control algorithm has strong fault tolerance to actuator faults and anti-interference capabilities. The convergence time and performance of trajectory tracking errors can be preset in advance, and the parameter settings of the prescribed performance constraint function are not affected by the initial state values.

Hand tracking is relevant to such a variety of applications including human-robot interaction (HRI), human-computer interaction (HCI), virtual reality (VR), and augmented reality (AR). Accurate and robust hand tracking however is challenging due to the intricacies of dynamic motion within small space and the complex interactions with nearby objects, coupled with the hurdles in real-time hand mesh reconstruction. In this paper, we conduct a comprehensive examination and analysis of existing hand tracking technologies. Through the review of major works in the literature, we have discovered numerous studies employing a diverse array of sensors, leading us to propose their categorization into seven types: vision, soft wearable, encoder, magnetic, inertial measurement unit (IMU), electromyography (EMG), and the fusion of sensor modalities. Our findings indicate that no singular solution surpasses all others, attributing to the inherent limitations of using a single sensor modality. As a result, we assert that integrating multiple sensor modalities presents a viable path toward devising a superior hand tracking solution. Ultimately, this survey paper aims to bolster interdisciplinary research efforts across the spectrum of hand tracking technologies, thereby contributing to the advancement of the field.

The simultaneous presence of uncertain data delays and data loss in a network control system complicates the state estimation problem and its solution. This paper redesigns the Kalman filter (KF) algorithm for systems with k-step random delayed data and data loss to improve estimation accuracy. A binary Bernoulli distribution is employed in the modified KF algorithm to model the received data with the knowledge of data delay and loss probabilities. Besides, the distribution of the non-Gaussian noise in the measurement system will degrade the performance of the conventional KF algorithm based on the minimum mean square error. Therefore, the modified KF algorithm is extended to the maximum correntropy Kalman filter (MCKF) algorithm to overcome the effect of non-Gaussian noise. The estimation accuracy of the modified KF and MCKF algorithms are experimentally compared under Gaussian and non-Gaussian noises, respectively. The simulation results demonstrate the excellent estimation performance of the proposed modified KF and MCKF algorithms under Gaussian and non-Gaussian noises, respectively.

The vertical tire force can be utilized to obtain information on the longitudinal and lateral force of the tire through the tire friction circle. This means that ride safety can be improved by using the longitudinal force that affects the vehicle’s driving performance and the lateral force that allows for stable cornering without slips. In this paper, we propose a vertical tire force estimation method using sensors that can be implemented in the vehicle. First, the issue of the observability of the tire force is investigated, then we introduce a tire force observer that utilizes the acceleration of the sprung and the displacement between the sprung and unsprung mass. In the proposed observer design, the change in the road surface is taken into consideration as a Gaussian random variable. In addition, a 1/5 scaled quarter car model is developed as an experimental apparatus to evaluate the proposed method, and the proposed method is validated through simulation and experiment.