This article addresses the problem of resilient$H_\infty$control for direct-drive-wheel (DDW) systems by developing an adaptive hybrid-triggered (AHT) scheme that improves security and communication efficiency against hybrid cyberattacks. First, the attacked DDW system is modeled as a switched cyber-physical system (CPS) to effectively capture the impact of such attacks on signal transmission. Then, an attack-information-dependent AHT scheme is proposed, where triggering instants are determined by system states and attack information, and the adaptive threshold is tuned via a self-adaptive differential evolution (DE) algorithm to balance dynamic performance and communication frequency. By constructing a Lyapunov functional, sufficient conditions are derived to co-design the control gain and AHT parameters, ensuring global exponential stability and guaranteed$H_\infty$performance. Finally, the proposed method is validated on examples, showing its efficacy in maintaining system stability while considerably reducing communication overhead.

This article is concerned with the stochastic finite-timeL1control problem of multi-loop networked control systems (NCSs) with network-induced delay, random packet loss, and external interference. Firstly, considering the data processing mode jumping, data transmission channel switching, and positive total amount of data, the multi-loop NCSs with network-induced delay, random packet loss, and external interference are modeled as a more general class of variable dual switching positive time-delay systems (VDSPTDSs) for the first time. Secondly, a new scheduling strategy that fully considers the random packet loss and the total amount of data, named positive minimum state expectation (PMSE), is proposed. Under this scheduling strategy, the channel with the smaller expectation of the total amount of data is selected to reduce the communication overhead. Subsequently, a stochastic multiple co-positive Lyapunov-Krasovskii functional (SMCPLKF) is constructed to establish the criteria of stochastic finite-time bounded (SFTB) and finite-timeL1-gain performance. A mode-dependent finite-timeL1-gain state feedback controller is further designed such that the closed-loop VDSPTDSs are positive and SFTB withL1-gain characterization. Finally, a multi-loop data communication NCS model is given to demonstrate the validity and generality of the proposed methods.

In a continuous annealing process (CAP), an accurate and timely temperature prediction is crucial, particularly during transitions between different types of strips. Predicting transition states is challenging due to process nonlinearities and time-varying characteristics induced by these transitions. In addition, extensive historical datasets hold valuable insights that can be harnessed to detect consistent and similar changes within transition groups. In this study, we introduce a data-driven framework known as the “progressive search and parallel model” to forecast transition states rapidly. The “progressive search” is executed by categorizing subclasses based on parameter properties and spatial locations, each with varying priority levels. Specifically, parameters are subdivided into segments along a piecewise linear scale, and a just-in-time learning framework is employed to establish local models at these levels concurrently. We apply this framework in both stable and transitional CAP conditions, demonstrating its robust performance in effectively tracking shifting trends during state transitions.

In practice, repetitive control (RC) is a type of learning control that exhibits good tracking performance. However, existing nonlinear RC methods lack analysis and design of the learning property, which results in limited performance. This study addresses the learning-enhanced issue. First, a new fuzzy Lyapunov candidate is constructed for stability analysis, which contains an integral term associated with the membership function and a double integral term. The design of the learning-dependent term integrates the nonlinear membership function information, which enhances the learning ability. Second, an additional first-order low-pass filter is incorporated into the conventional equivalent input disturbance estimator. The new filter acts as an integrator that adjusts the bandwidth of the disturbance compensation and gradually eliminates the disturbances in the output error. Third, a recursive optimization algorithm is used to design the controllers. Experimental comparisons on motor drive systems demonstrate the effectiveness and superiority of the method.

The stability issues of linear systems with time-varying delays are tackled in this article. Several positive augmented Lyapunov-Krasovskii (L-K) functionals are proposed by introducing integral quadratic functions based on the L-K stability theorem. To further reduce the estimation gap caused by the existing integral inequalities, which were applied for dealing with the derived augmented-type integral term from augmented functional, some refined matrix-separation-based inequalities are introduced. With fewer decision variables, the proposed method considers the information among the system state, its derivative, and their related terms. After these, some cubic functions in the time-varying delay appear and show nonconvexity. Then, negative definite conditions are imposed on such cubic terms to obtain the stability criterion in the form of the linear matrix inequality (LMI). Inspired by the Taylor’s expansion methodology and the delay-partitioning techniques, we improve the existing negative definite conditions on the cubic function without introducing any decision variable. For linear systems with a time-varying delay, this relaxed condition, the refined matrix-separation-based inequalities, and the constructed L-K functionals, are combined to produce a less conservatism stability criterion. Two numerical examples illustrate the effect of the offered methods and the stability condition.

Weight learning forms a basis for the machine learning and numerous algorithms have been adopted up to date. Most of the algorithms were either developed in the stochastic framework or aimed at minimization of loss or regret functions. Asymptotic convergence of weight learning, vital for good output prediction, was seldom guaranteed for online applications. Since linear regression is the most fundamental component in machine learning, we focus on this model in this paper. Aiming at online applications, a deterministic analysis method is developed based on LaSalle's invariance principle. Convergence conditions are derived for both the first-order and the second-order learning algorithms, without resorting to any stochastic argument. Moreover, the deterministic approach makes it easy to analyze the noise influence. Specifically, adaptive hyperparameters are derived in this framework and their tuning rules disclosed for the compensation of measurement noise. Comparison with four most popular algorithms validates that this approach has a higher learning capability and is quite promising in enhancing the weight learning performance.

This study presents a two-dimensional (2-D) repetitive control method to address the issues of periodic tracking and disturbance suppression in uncertain Takagi–Sugeno systems. The disturbance and uncertainty are treated as an equivalent-input-disturbance (EID). However, the conventional EID estimators typically suppress the EID through high gain. Meanwhile, the low-pass filter associated with EID causes a certain degree of phase lag. A proportional–integral (PI) filter is integrated with an EID estimator to develop a PI-EID structure to improve the estimation accuracy. Based on the self-learning mechanism of repetitive control, the 2-D repetitive controller is used to achieve a high level of tracking. Unlike the conventional nonlinear repetitive control methods, the state observer and the PI-EID estimator are membership function dependent. The gains of both controllers switch in line with the signs of the time derivative of the normalized premise variables, and this framework takes full account of the information of the nonlinear membership functions. The controller design procedures and the stability conditions are detailedly presented. Finally, a rotation speed control experiment is conducted to validate the developed PI-EID method.

Optimization control of the rate of penetration (ROP) is crucial due to its vital role in maximizing the drilling efficiency. In this article, a dynamic optimization-based intelligent control system for drilling ROP is proposed considering the drilling characteristics. First of all, the framework of the proposed system is described, which has two layers (intelligent optimization layer and basic automation layer). In the former layer, two stages (ROP modeling stage and ROP optimization/implementation stage) are executed alternatively by using the moving window strategy. Moving window-Extreme learning machine and tenfold cross validation are used to establish the dynamic model between the rotation speed (RPM), weight on bit (WOB), depth, and ROP. In addition, hybrid bat algorithm is introduced to search the controllable parameters while the inputs for ROP optimization are RPM, WOB, depth, constraints, and ROP model, and the outputs are the optimized RPM and optimized WOB. After that, the outputs are recommended to the driller to be used as the setpoints of the basic automation layer. Comparison results of simulation with seven well-known methods demonstrate the effectiveness of the proposed system. In addition, in an industrial application to a real-world drilling process in the Xiangyang area, Central China, the ROP was improved by 15.9%.

Rate of penetration (ROP) optimization is crucial for drilling processes due to its vital role in increasing efficiency. This article proposes a new hybrid bat algorithm (HBA) to achieve the maximum ROP accurately. A data-driven ROP model is established by combining the wavelet filtering and optimized support vector regression according to the drilling characteristics. After that, five modifications, namely, iterative local search, stochastic inertia weight, pulse rate and loudness improvement, directional echolocation, and modified local random walk are combined to further improve the global optimization performance of the bat algorithm. Extensive experiment results on the IEEE Congress on Evolutionary Computation 2005 standard benchmark problems show that the proposed algorithm has successful outcomes compared with ten conventional algorithms. Additionally, in the application of the HBA to a real-world drilling process in Shennongjia area, Central China, the ROP has been improved by 34.84%, which is the largest compared with the conventional popular ROP optimization methods.

This paper proposes a cruise control for a two-wheeled mobile vehicle on a two-dimensional plane. The cruise control is executed using model predictive control (MPC), such that input and state constraints can be addressed explicitly. For real-time execution of MPC, we model the dynamics of the vehicle as a mixed logical dynamical system rather than an underlying nonlinear model. The cruise control is then reduced to a mixed integer quadratic programming problem, for which efficient solvers using the convexities are available. Simulations and experiments demonstrate the effectiveness of the proposed cruise control through a computation time comparison with the nonlinear MPC method.

This paper proposes a novel adaptive flux observer for the stator resistance identification of speed sensorless controlled induction motors (IM). Generally, to realize the stable parameter identification, the adaptive controller requires strictly positive-realness (SPR) property for the transfer function of the error system of observer. Since this error system does not have the SPR property in the low-speed region when operated in the regenerative mode, this paper proposes a new compensation method to realize the stable stator resistance identification. Moreover, this paper provides a stability analysis for the stator resistance identification system. This paper also demonstrates the feasibility of the proposed method through simulations.

This paper addresses the real-time control of transmission power for small cell base stations (SBSs) exploiting energy- harvesting sources. We employ model predictive control and optimize an objective function that contains the number of users with a given quality of experience and the average state of charge. We first determine the number of active SBSs in the viewpoint of energy efficiency and then approximate the objective function. Finally, we illustrate the proposed method through a numerical example, comparing it with a static method based on statistical information.

An essential issue in the hybrid energy storage system (HESS) control for smoothing wind farm output is designing a controller that can decompose the frequency component of wind power quickly and allocate the power to the respective device in online processing. In this paper, a HESS consists of conventional battery-Li-Ion Battery and Ultra Capacitor (UC) is adopted to meet the electric grid technical requirements for smoothing wind power. We propose a HESS power allocation control consisting of two stages: online wavelet filter and a novel power allocation optimization algorithm. An online wavelet filtering method is established to minimize the phase lag in decomposing frequency components. The results show a significant reduction in UCs' utilization as compared with conventional method. The optimization algorithm also provides an optimal power allocation between storage devices with fast enough computation time to satisfy the need of online processing.

This paper addresses an optimal event generator synthesis for networked control systems (NCSs). The event generator is installed in a sensor, and sampled data packets are chosen for transmission through a communication network while desired performance is guaranteed to use network resources efficiently. We propose an optimal event generator synthesis method using a model predictive-based scheme. The effectiveness of the proposed method is illustrated with an example of a rotary inverted pendulum system.

In networked control systems, the data needs to be quantized for transmission over a limited bandwidth communication channel. We consider a dynamic quantizer in a quantized feedback control system. In this system, the quantizer parameters and input are optimized on-line by using the model predictive control to achieve the optimal control performance. In our method, the system constraints explicitly considered. The effectiveness of the proposed method is verified through simulation and experimental results.

Recently, as an effective means of dealing with global warming and fossil fuel depletion, the introduction of wind power generation is promoted all over the world. In the wind power generation, a very important issue is the smoothing of its output power which usually fluctuates with the wind speed. In this paper, a new method is proposed for the output power smoothing of wind farms. This method is based on an optimization approach and makes use of the kinetic energy of windmills. Detailed algorithm is provided and validated through numerical simulations with a quite complete simulator. The results are quite promising which show that the sharp fluctuation of wind farm power can be smoothed without much sacrifice of generation efficiency.

Networked control systems (NCSs) have been receiving much attention in the field of remote robot operation, surgery and some operations. In the NCSs, data needs to be quantized since it is transmitted over a limited network channel. In earlier works, we considered an NCS with a variable discrete quantizer. In the system, both input and a parameter of the quantizer are optimized online with the help of model predictive control (MPC) so that the system is stabilized. However, the method is not applied to the tracking control, directly. This paper addresses an extension of the quantized feedback control system in order to consider the tracking control performance as well as the stabilization of the system. In the system, the center of the quantization is considered as a variable to be optimized. The optimization problem is reduced to a mixed integer quadratic programming so that the tracking control can be realized online. Experimental results are demonstrated to verify the effectiveness of the proposed method.

Data dropout in networked control systems (NCSs) is unavoidable. The desired performance may not be achieved if the data dropout is not considered. Therefore, a controller in the NCS needs to be designed which takes the data dropout into account. In this paper, the data dropout is assumed to follow a stochastic process. For the NCS, a model predictive control (MPC) method is proposed for the design of an optimal input sequence. The effectiveness of the proposed method is verified through simulations and experiments.

In this paper, a method for automatic parking systems is proposed based on the approach of Kang-Zhi Liu, (2002) and the corresponding system package is developed. This method is composed of control laws for the steering angle and the driving velocity, algorithms for various kinds of parking tasks under steering angle arid parking space limitations. The system works pretty well in experiments subject to various parking situations.