Search

The stability of dual active bridge converter (DAB) is threatened when feeding the constant power loads (CPLs). This article proposes a deep reinforcement learning-based backstepping control strategy to solve this problem. First, a nonlinear disturbance observer is adopted to estimate the large-signal nonlinear disturbance. Then, a backstepping controller is used to stabilize the voltage response of the DAB under the large-signal disturbance. Finally, a compensation method based on deep reinforcement learning is developed to intelligently minimize output voltage tracking error and improve the operating efficiency of the system. The proposed controller can guarantee system stability under the large-signal disturbance of the CPL and achieve a fast dynamic response with accurate voltage tracking; it is more adaptive by using the deep reinforcement learning technique through the learning of its neural networks. The effectiveness of the proposed controller is verified by experiments. [ABSTRACT FROM AUTHOR]

For high-precision contouring tasks in a typical Cartesian motion system, multiaxis cooperation is a long-standing challenging issue. Inevitably, various factors pose substantial difficulty in the multiaxis cooperation leading to degraded contouring performance, such as the strong coupling effect between different axes, nonlinearity, the unknown dynamics due to the friction, and the difficulties in accurate system identification. To enhance the contouring performance of a flexure-joint dual-drive gantry system against the aforementioned issues, this article presents a hybrid active–passive robust control framework leveraging a model-free architecture. In this control scheme, all the coupling effects, nonlinearity, disturbance, and unknown dynamics are considered as “lumped uncertainty”. Then, a super-twisting sliding mode control method with a signum-type iterative learning law is proposed to passively suppress the lumped uncertainty during iterations; and an extended state observer is deployed to actively compensate the lumped uncertainty and ensure the establishment of sliding motion in the time domain. As supported by theoretical analysis, the proposed controller is shown to exhibit several important properties. First, the establishment of the sliding motion is guaranteed globally, in both the time domain and the iteration domain. Second, the properties of short establishment time of the sliding motion, fast convergence during the iterations, and low chattering are achieved. Moreover, a series of comparative experiments are conducted, and the proposed method is shown to be rather effective in achieving excellent contouring performance in the high-speed and complex-curvature contouring tasks, without relying on the system model. [ABSTRACT FROM AUTHOR]

In this article, we propose the dynamic coverage control method based on K-means. In the traditional coverage control, Voronoi partition method is used to assign the coverage positions for intelligent units. However, the Voronoi partition method requires that the space to be covered is compact and convex, and it is difficult to realize the coverage control of high-dimensional space. Therefore, in this article, we propose a dynamic coverage planning algorithm based on K-means, which relaxes the requirements on the coverage objects and can calculate the optimal coverage positions of the intelligent units. In addition, we also design a dynamic coverage control law based on discrete sliding mode control to drive the intelligent units to the optimal coverage positions. Through the combination of planning algorithm and control law, the optimal coverage control performance can be achieved for the specified targets and the specified area, which can be discrete, nonconvex, and high dimensional. The stability of the planning algorithm and the control law are proved by theoretical deduction. The effectiveness and superiority of the dynamic coverage control method are verified by two examples. [ABSTRACT FROM AUTHOR]

The heating furnace is a sizeable energy-consuming device in iron and steel industry. In compact strip production (CSP), the size of the heating furnace is large, and the working conditions are complicated. Furthermore, the large fluctuation of furnace temperature leads to large loss of billet combustion and energy consumption. This article deals with the design of hybrid intelligent control based on condition identification for the combustion process of the CSP heating furnace. By analyzing the process of the heating furnace and existing problems, the structure of a hybrid intelligent control system is proposed, and the working conditions are divided into two categories: stable and fluctuating. A fuzzy controller is designed to improve the control accuracy of furnace temperature under the stable working condition, and an expert controller is used to adjust the temperature rapidly under the fluctuating working condition. Besides, calorific value compensation is introduced to reduce the influence of calorific value of gas fluctuation on temperature. After verifying the effectiveness, the proposed method has been used in a steel plant and achieved a good control result. The intelligent control system improves the control precision of the furnace temperature and reduces energy consumption compared with the traditional control system. [ABSTRACT FROM AUTHOR]

This article suggests an intelligent position-tracking control algorithm for rover vehicles considering actuator (dc motor) dynamics. The parameter and load uncertainties in the vehicle and dc motor dynamics are explicitly handled by modifying the original open-loop system dynamics. The proposed controller forms the conventional multiloop structure including disturbance observers for each loop. The features of this article fall into the following three parts: first, the learning part from the real-time feedback gain mechanism (named the self-tuner) in the closed-form for outer loop; second, the adaptation part from the online wheel radius estimation securing the outer-loop control accuracy; and third, the parameter-independent angular acceleration observer-based pole-zero cancellation dc motor speed controller without current feedback considering the inner- and outer-loop vehicle control algorithms. Experimental evidence is also provided to demonstrate the practical merits of the proposed technique with the use of the TETRIX, myRIO-1900, and LabVIEW. [ABSTRACT FROM AUTHOR]

Considering the increasingly strict motion precision requirements and repetitive task characteristics of magnetic levitation systems (MLSs) in the sophisticated industry, this article proposes a novel learning adaptive sliding mode control (LASMC) strategy for the MLS to achieve an excellent tracking performance. The LASMC scheme is obtained by combining adaptive sliding mode control (ASMC) and iterative learning control (ILC) terms in a parallel structure. ASMC can guarantee system stability and strong robustness, which employs an adaptive switching gain and parameter adaption algorithm. Even if the accurate disturbance bounds of MLS are not known, the ASMC item can also adjust the switching control gain online to ensure the stability and robustness of the system. Additionally, the ILC term can further improve the MLS performance for repetitive motion tasks without the accurate dynamics model. The asymptotic stability of the LASMC strategy is verified based on the Lyapunov theorem in the presence of unmodeled dynamics, disturbances, and saturation. Comparative experiments carried out on a maglev rotary table demonstrate that the proposed control strategy achieves excellent tracking accuracy and disturbance robustness in the practical MLS. The LASMC scheme provides a decent idea for the application of intelligent control technology in the controller design of MLSs. [ABSTRACT FROM AUTHOR]

A novel bidirectional fuzzy brain emotional learning (BFBEL) controller is proposed to control a class of uncertain nonlinear systems such as the quadcopter unmanned aerial vehicle (QUAV). The proposed BFBEL controller is nonmodel-based and has a simplified fuzzy neural network structure and adapts with a novel bidirectional brain emotional learning algorithm. It is applied to control all six degrees-of-freedom of a QUAV for accurate trajectory tracking and to handle the payload uncertainties and disturbances in real-time. The trajectory tracking performance and the ability to handle the payload uncertainties are experimentally demonstrated on a QUAV. The experimental results show a superior performance and rapid adaptation capability of the proposed BFBEL controller. The proposed BFBEL controller can be used for the commercial drone applications. [ABSTRACT FROM AUTHOR]

Tray indexing driven by a timing-belt actuator is commonly used in the manufacturing industry, in situations where precise and fast reference tracking is required to ensure product quality and manufacturing efficiency. However, as a commonly encountered challenge, several factors make accurate system identification here to be rather difficult, such as the existence of flexible modes and other unmodeled dynamics in the system model, including the deadzone and backlash caused by the friction. Due to these unknown dynamics, the system performance (when system identification is not suitably accurate) could be diminished when traditional model-based controllers are used. To overcome the above-mentioned problems arising from these uncertainties/perturbations and further improve the tracking performance, this work investigates the design of an intelligent model-free controller for such a tray indexing system. Specifically, a 2-degree-of-freedom control structure is developed, which consists of a generalized super-twisting sliding mode control and a unique iterative learning law, and appropriately combined. The proposed method solely makes use of the input–output data obtained during iterative experiments such that the system is driven toward proper satisfactory performance without any prior knowledge of the system parameters. Also, in our work here, rigorous proof is furnished on the convergence properties of the method, and real-time experiments are conducted on the timing-belt stage to validate the effectiveness and the practical appeal of the proposed approach. It is pertinent to note also that applications of the proposed methodology are certainly not constrained to such an industrial timing-belt application only, and its potential can be further deployed advantageously to other appropriate mechatronic systems with similar difficulties involving such existence of uncertainties/perturbations. [ABSTRACT FROM AUTHOR]

High accuracy of terrain identification is essential for intelligent control of tracked mobile robots. In this article, a learning-based identification framework is proposed to achieve precise driving torque prediction. Experiments under straight-line and steady-state turning maneuvers were conducted to develop terrain identification in three similar terrains. A multiple deep belief networks is applied as the identification layer with three kinds of signal sources. An equivalent weight algorithm with training experience effectively integrates the results from different signal sources to improve the identification accuracy. With the experiment and identification results, a method combining numerical approximation and Gaussian process (GP) is presented to predict driving torque. A combined Gaussian kernel with long- and short-term characteristics is selected to enhance prediction performance. The results from combined signal sources under straight-line maneuvers yield over 98% accuracy, which exceeds that from other sources. The integration algorithm obviously improves the identification accuracy and stability compared with a single signal source. The influence of window length in GP is explored with hyperparameters and results. The performance of torque prediction in different terrains is analyzed with inner and outer tracks. Compared with existing methods, the results validate the effectiveness and superior performance of the proposed framework. [ABSTRACT FROM AUTHOR]

For the ongoing revolution in developing intelligent and connected vehicles (ICVs), there is a lack of research for powertrain control systems using the latest artificial intelligence and vehicle-to-everything technology that have already been widely adopted in the autonomous driving systems. In this context, recent development of deep reinforcement learning (DRL) and one of the latest computing frameworks are coupled to facilitate an onboard-based intelligent powertrain control. Taking the boost control of a diesel engine equipped with variable geometry turbocharger as an example, the results show that the final control behavior indicated by the cumulated rewards is improved by 50.43% and the learning efficiency is improved by 74.29% for the proposed curiosity-driven DRL algorithm, compared with the same structure DRL algorithm with classic random exploration policy. In addition, unlike most of the DRL-based powertrain optimization algorithms, which have only been applied to single-machine architecture, this work manages the proposed DRL algorithm in parallel and, more importantly, from an edge computing perspective. This, in addition to greatly speeding up the algorithm training, can also realize a good balance of control accuracy and generality depending upon the selected training scenario. Moreover, unlike most of the cloud computing frameworks, which require low network latency, the proposed architecture can achieve a similar final control performance even if good network communication is not allowed. Compared with other existing powertrain control methods, the proposed algorithm is able to approximate a global powertrain control optimization autonomously in a connected manner, making it attractive to current ICVs with advanced automated driving and traditional powertrain control. [ABSTRACT FROM AUTHOR]

The presented research activities deals with the idea of electromagnetically actuated switching devices switching-on process improvement by means of an electronic controller application. Based on the analysis of the results of the electromagnetic contactors dynamic at switching-on selection, the most efficient condition was determined. Making process improvement methods based on the principle of controlling the supply voltage phase were presented in this article: selected dynamic parameters at switching-on optimization in view of its parameters values minimization and intelligent compensation of the movable armature inertia using predictive control of the driving coil supply voltage. Experimental verification of the developed controller confirmed its applicability for various contactor types. Tested objects increased its ability for “near to zero,” switching-on operation realization and selected parameters optimization without the necessity of device construction modification. [ABSTRACT FROM AUTHOR]

Wireless power transfer (WPT), in a general scene, is the transmission of electrical energy for various technologies by means of magnetic couple resonance. Its potential applications change from low power electric equipment (like smartphones) to high-power devices over short distances of centimeters to tens of centimeters. In this letter, a pulsewidth-modulation-based adaptive data-driven strategy is newly developed for the output voltage regulation of the series–series (SS) compensated WPT. In an attempt to achieve a fast and adaptive voltage regulation, a deep deterministic policy gradient (DDPG) based on an ultralocal model (ULM) control scheme is applied to the dc–dc buck–boost converter feeding constant power loads installed in the receiver side of the WPT. In the suggested scheme, an intelligent feedback controller based on the ULM is used to stabilize the voltage response of the buck–boost converter while the unknown dynamics of the converter are estimated by a sliding mode observer (SMO). An auxiliary deep DDPG algorithm with Actor-Critic architecture is implemented to employ its adaptive capability to eliminate the SMO estimation error and improve the voltage regulation of the WPT plant. Some experimental outcomes on a laboratory testbed of the WPT plant are presented to assess the merits and real-time implementation of the adaptive data-driven scheme from a systematic perspective. [ABSTRACT FROM AUTHOR]

The sintering ignition is an important step in the sintering process and is the starting point of the entire combustion process. It is of great economic value and scientific value to stabilize the sintering ignition temperature in the case of the unstable gas pressure. An intelligent control strategy is proposed for the iron ore sintering ignition process based on the prediction of ignition temperature in this paper. First, the ignition temperature prediction model is established by the combination of the mechanism analysis method and the data-driven method. Then, the control mode is determined by the actual operation experience, and the intelligent controller for the ignition temperature is designed to obtain the required gas flow, which is stabilized by the gas flow controller. Finally, with a case-matching method applied, the experiment of the intelligent control strategy is carried out based on the data collected from practical processes. The experimental results show that the proposed control strategy is good at controlling ignition temperature under the situation of the unstable gas pressure and has good application perspective. [ABSTRACT FROM AUTHOR]

Interconnected dc grids are studied in this paper, which comprise resistive and constant-power loads (CPLs) fed by photovoltaic (PV) units. All the sources and CPLs are connected to the grid via dc–dc converters. Nonlinear behavior of PV units in addition to the effect of negative-resistance CPLs can destabilize the dc grid. Thus, the decentralized nonlinear model and intelligent control are proposed using adaptive output-feedback controller to stabilize the grid. The use of the output-feedback control makes possible the utilization of other available signals, in case of loss of main signal, at the converter location and creates measurement redundancy that improves reliability of the dc network. The switching between measured signals of different types are performed through using the neural network (NN) controllers without the need to further tuning. The stability of the entire network is assured through Lyapunov stability method while each converter employs only local measurements. The adaptive NNs are utilized to overcome the unknown dynamics of the dc–dc converters at distributed energy resources and CPLs and those of the interconnected network imposed on the converters. Simulation and experimental results are provided on a small-scale dc grid to show the effectiveness of the developed model and the proposed controller. [ABSTRACT FROM AUTHOR]

This paper presents a green controller for a governor-less diesel engine driven two-winding single phase standalone self-excited induction generator (SEIG) based on search-then-converge network algorithm. Diesel engines used for driving standalone electric generators are designed with a rated power output. Normally such engines exhibit maximum efficiency at about 80% of their maximum power rating. This point of operation is called as maximum efficiency point of operation. The proposed control operates the single-phase SEIG to generate a fixed electric power to achieve a maximum efficiency point of operation of the diesel engine irrespective of quantum and nature of load. The proposed controller diverts the surplus generated electric power to the battery energy storage system after satisfying the power demand of the load in order to conserve electrical energy and to improve the overall system efficiency. The operation of a diesel engine at maximum efficiency point of operation with constant mechanical loading irrespective of amount of electrical loading on the generator reduces the brake specific CO, smoke, and unwanted hydro carbon emission from the engine. Thus this scheme is named as a green controller. [ABSTRACT FROM PUBLISHER]

An intelligent controller based on the Takagi–Sugeno–Kang-type probabilistic fuzzy neural network with an asymmetric membership function (TSKPFNN-AMF) is developed in this paper for the reactive and active power control of a three-phase grid-connected photovoltaic (PV) system during grid faults. The inverter of the three-phase grid-connected PV system should provide a proper ratio of reactive power to meet the low-voltage ride through (LVRT) regulations and control the output current without exceeding the maximum current limit simultaneously during grid faults. Therefore, the proposed intelligent controller regulates the value of reactive power to a new reference value, which complies with the regulations of LVRT under grid faults. Moreover, a dual-mode operation control method of the converter and inverter of the three-phase grid-connected PV system is designed to eliminate the fluctuation of dc-link bus voltage under grid faults. Furthermore, the network structure, the online learning algorithm, and the convergence analysis of the TSKPFNN-AMF are described in detail. Finally, some experimental results are illustrated to show the effectiveness of the proposed control for the three-phase grid-connected PV system. [ABSTRACT FROM PUBLISHER]

This paper proposes a more efficient intelligent control algorithm that can be applied to unknown nonlinear systems. First, a more general neural network (NN) referred to as a function-link cerebellar model articulation controller (FLCMAC) is proposed. In some cases, this FLCMAC can be reduced to an NN and a function-link NN. An FLCMAC-based control system is then developed. The proposed control system comprises an FLCMAC and a robust controller. The FLCMAC is the principal tracking controller used to mimic an ideal controller, and the parameters of FLCMAC are online tuned using derived adaptation laws that use the Lyapunov function. The robust controller can eliminate the approximation error so that the asymptotic stability of the system is achieved. The proposed control system is then applied to a magnetic ball levitation system (MBLS), which is an intricate and highly nonlinear system. For practical experiment, the proposed control scheme is implemented using a system-on-programmable-chip technique. Finally, the simulations and experiments are performed for an MBLS in order to illustrate the effectiveness of the proposed control system for achieving precise trajectory tracking. [ABSTRACT FROM PUBLISHER]

This paper presents the design method of a permanent-magnet (PM) actuator and an intelligent control module for the associated PM contactor. The original design parameters of the PM actuator are obtained using magnetic circuit method. The optimization model for the PM actuator is solved by genetic algorithm and simulated annealing algorithm. A monostable cylindrical 140-A PM contactor is designed, and the corresponding intelligent control module of the actuator is developed. A dedicated control technique based on voltage feedback is proposed to realize the intelligent dynamic adjustment of the making process for the PM contactor. The electromagnetic force calculation model, the dynamic making model of the iron core and contact, the buffering model for the iron core, and the contact bounce model are all built using MATLAB. Experimental results are reported to showcase the correctness of the simulation method and the effectiveness of the feedback control technique to decrease the closing velocities of both the movable contact and the moving iron core that, in turn, will reduce the first and the second contact bounces significantly to lengthen the life span of the device. [ABSTRACT FROM PUBLISHER]

In this paper, we consider the friction compensation problem for a class of mechanical systems. The friction behavior is described by a nonlinear dynamical model. Since it is difficult to know the nonlinear parts in the frictional model accurately, two neural networks (NNs) are employed in the proposed intelligent controller. Due to the learning capability of the NNs, the designed NN controller can compensate the effects of the nonlinear friction. Stability of the thus proposed learning control system is guaranteed by a rigid proof. Simulation and experimental results are provided to verify the effectiveness of the proposed intelligent scheme. [ABSTRACT FROM PUBLISHER]

An increasing number of carlike mobile robot (CLMR) studies have addressed the issues of autonomous parking and obstacle avoidance. An autonomous parking controller can provide convenience to a novice driver. However, if the controller is not designed adequately, it may endanger the car and the driver. Therefore, this paper presents a novel multifunctional intelligent autonomous parking controller that is capable of effectively parking the CLMR in an appropriate parking space, by integrating sensor data capable of obtaining the surrounding data of the robot. An ultrasonic sensor array system has been developed with group-sensor firing intervals. A binaural approach to the CLMR has been adopted for providing complete contactless sensory coverage of the entire workspace. The proposed heuristic controller can obtain the posture of a mobile robot in a parking space. In addition, the controller can ensure the ability of the CLMR to withstand collision to guarantee safe parking. Moreover, the CLMR can recognize the parking space and the obstacle position in dynamic environment. Therefore, the proposed controller installed in a car could ensure safe driving. Finally, practical experiments demonstrate that the proposed multifunctional intelligent autonomous parking controllers are feasible and effective. [ABSTRACT FROM AUTHOR]

This paper addresses the application of an intelligent optimal control system (IOCS) to control an indirect field-oriented induction servo motor drive for tracking periodic commands via a wavelet neural network. With the field orientation mechanism, the dynamic behavior of an induction motor is rather similar to a linear system. However, the uncertainties, such as mechanical parametric variation, external load disturbance and unmodeled dynamics in practical applications, influence the designed control performance seriously. Therefore, an IOCS is proposed to confront these uncertainties existed in the control of the induction servo motor drive. The control laws for the IOCS are derived in the sense of the optimal control technique and Lyapunov stability theorem, so that system-tracking stability can be guaranteed in the closed-loop system. With the proposed IOCS, the controlled induction servo motor drive possesses the advantages of good tracking control performance and robustness to uncertainties under wide operating ranges. The effectiveness of the proposed control scheme is verified by both simulated and experimental results. Moreover, the advantages of the proposed control system are indicated in comparison with the sliding-mode control system. [ABSTRACT FROM AUTHOR]

Presents a supervisory controller for discrete-event systems that uses a novel matrix formulation. Flexible manufacturing systems; Intelligent control; Petri nets; Reentrant flow lines; Intelligent material handling.

Focuses on dealing with the drawbacks of piezomechanics using an intelligent variable-structure control. Hysteresis feature and expansion dependence on temperature; Use of a neural network to learn the dynamics of piezomechanism; Synthesis of a discrete-time variable-structure to improve performance.

Proposes an intelligent controller for a robot manipulator, that consists of an intelligent planner and an adaptive fuzzy neural position/force controller. Application of the proposed controller to the task of deburring with an unknown object; Evaluation of the effectiveness of the proposed controller by computer simulations.

Editorial. Introduces several articles discussing intelligent systems. Knowledge-rules-based method for solving multicriteria optimization problems; Soft computing; Identification of an unknown polynomial structure through genetic algorithms; Verification of a stable grasp mechanism of a robotic hand; Robust controller for feeback systems control; Feature extraction analysis.