This paper proposes an adaptive predictive PID control approach using recurrent polynomial-based fuzzy broad learning systems (RP-FBLS) for nonlinear discrete-time dynamic systems. The RP-FBLS combines polynomial-based fuzzy logics, broad learning systems, and recurrent networks to enhance modeling capabilities and adaptive control performance. The RP-FBLS-based adaptive predictive PID controller is evaluated through three simulation case studies on two renowned discrete-time dynamic systems and one experimental study on a heating oven in semiconductor manufacturing. The comparative studies analyze the time complexity and control performance of the proposed approach over existing methods in the aspects of setpoint tracking, disturbance rejection, and robustness. The paper also quantifies the computational effort per sampling instance by considering real-time constraints and highlighting suitability for embedded systems. Simulation results demonstrate the RP-FBLS-based PID controller's efficacy in handling nonlinear dynamics while offering favorable computational performance characteristics. Experimental validations using the heating oven are provided for illustration of the practicability of the proposed method requiring computational efficiency and memory optimization.

In this paper, a novel control architecture is presented by integrating an Output Recurrent Fuzzy Neural Long Short-Term Memory (ORFNLSTM) and a Broad Learning System (BLS) for a class of single-input-single-output (SISO) nonlinear dynamic systems. This new controller, abbreviated as the ORFNLSTM-BLS controller, is especially proposed by combining the techniques of deep learning and broad learning method to establish an adaptive intelligent controller with an online deepest gradient descent learning algorithm to online update its weights. The ORFNLSTM-BLS controller aims to improve the performance of the ORFBLS controller by incorporating the memory of LSTM to handle time-series data more effectively. A sufficient condition of the proposed controller is established to accomplish its uniformly asymptotical stability. The effectiveness and superiority of the ORFNLSTM-BLS controller are well exemplified by carrying out one comparative simulation in comparison with a fixed-gain proportional-integral-derivative (PID) controller, an adaptive predictive PID controller augmented with ORFBLS (ORFBLS-APPID), and an existing ORFBLS controller in terms of two types of control performance indexes: the overall performance indexes and transient state performance indexes. The results show that the proposed ORFNLSTM-BLS controller outperforms the three existing control methods. The developed techniques would provide useful references for professionals working in the fields of process and servomechanism control.

In this paper, an intelligent leader-following consensus formation control method using recurrent neural networks (RNNs) is presented for a team of uncertain small-size unmanned helicopters (SSUHs). After a brief description of the dynamic model of each uncertain SSUH by a set of multivariable fourth-order state equations, the leader-follower multi-SSUH system with a virtual leader is modeled by the directed graph theory. An intelligent adaptive formation control approach is proposed to fly together all the follower SSUHs in formation by using RNN to online learn the system uncertainties, consensus tracking, and the Lyapunov stability theory. The four simulations on three cooperating SSUHs are conducted to exemplify the effectiveness and merits of the proposed control method.

The paper presents a novel sliding-mode control method augmented with broad learning system (BLS), or abbreviated as BLS-SMC, for trajectory tracking and station keeping of an uncertain Inverse-Atlas ball-riding robot (IASBRR) driven by three omnidirectional wheels. After brief description of the dynamic model of the robot with frictions and gravity, a BLS-SMC controller is proposed to accomplish robust self-balancing and trajectory tracking of the IASBRR in the presence of unknown frictions, mass variations and model uncertainties. The proposed BLS-SMC controller is proven asymptotically stable using Lyapunov stability theory and Barbalta's lemma. Three comparative simulations and two experiments are conducted to show the effectiveness and merits of the proposed control method. The comparative results also indicate that the proposed controller is superior more efficient by comparing to an existing method.

In this paper, an integral terminal sliding-mode (ITSM) consensus control method is presented for a team of uncertain, networked, heterogeneous Mecanum-wheeled omnidirectional robots (MWORs) moving together in formation. After the description of the dynamic model of each kind of MWOR by a set of unified multivariable vector state equations, the interconnection structure of the multi-robots is modeled by a directed and strongly connected graph. An ITSM formation control is proposed to achieve finite-time formation control and trajectory tracking in presence of robot uncertainties. The usefulness and superiority of the proposed method are well exemplified by conducting three computer simulations on cooperative formation of four heterogeneous MWORs with a virtual leader.

This paper presents an adaptive nonsingular terminal sliding mode formation control by means of output recurrent fuzzy wavelet neural networks for a group of networked heterogeneous Mecanum-wheeled omnidirectional robots with uncertainties. The dynamic behavior of each uncertain heterogeneous omnidirectional robot is modelled by a reduced three-input-three-output second-order state equation with uncertainties and the multi-robot system is modeled by graph theory. Through the Lyapunov stability theory and online learning of the system uncertainties via output recurrent fuzzy wavelet neural networks, an adaptive nonsingular terminal sliding mode control approach is offered to attain finite-time formation control in presence of uncertainties. Conducting three simulations are to show the effectiveness and merit of the proposed method.

This paper presents a distributed consensus formation control using recurrent interval type-2 fuzzy neural networks (RIT2FNN) for a team of uncertain ballbots. The dynamic equations of each ballbot in sagittal and coronal planes can be decomposed into two identical second-order underactuated dynamic system models, and the multirobot system is modeled by graph theory. By online learning the system uncertainties using RIT2FNN and the Lyapunov stability theory, an intelligent distributed consensus formation control law is presented to carry out formation control in the presence of uncertainties. Simulations are conducted to show the effectiveness and merits of the proposed method.

This paper presents an intelligent sliding mode formation control using recurrent fuzzy wavelet neural networks (RFWNN) for a group of uncertain, networked heterogeneous Mecanum-wheeled omnidirectional platforms (MWOPs). The dynamic behavior of each uncertain MWOP is modelled by a reduced three-input-three-output second-order state equation and the multi-MWOP system is modeled by a directed graph. By using the Lyapunov stability theory and online learning the system uncertainties via RFWNN, an intelligent adaptive, sliding mode control approach is presented to carry out formation control in presence of uncertainties. Two simulations are conducted to show the effectiveness and merit of the proposed method with existing collision-free and obstacle-avoidance approaches.

In the paper, an adaptive decoupling predictive temperature control using neural networks (NN) is presented for extrusion barrels in plastic injection molding machines. Due to weakly coupling effects, the extrusion barrels are approximated by decoupling linear system models together with independent NN models. These decoupling system parameters and NN models are experimentally determined using the recursive least-squares estimation (RLSE) approach with forgetting factor. The adaptive decoupling predictive PI control laws together with NN compensating terms are developed by minimizing a generalized predictive cost function. A real-time control algorithm is then proposed to achieve temperature control of extrusion barrels. Experimental results on a laboratory-built extrusion barrel are conducted to illustrate the usefulness and applicability of the proposed method are well exemplified by conducting.

This paper presents a novel two degrees-of-freedom (DOF) digital controller using radial basis function neural network (RBFNN) for a stretch polyethylene-terephthalate (PET) blow molding machine, in order to achieve satisfactory temperature control of the PET bottle performs passing through both heating ovens. The proposed two-DOF controller is composed of a feedforward controller used to improve the transient performance and track quickly temperature setpoints, and an RBFNN self-tuning digital proportional-integral- derivative (PID) controller employed to eliminate remaining temperature errors and achieve disturbances rejection. Such a controller not only retains the practical expertise of the control practitioners working for PET blow molding machines, but also keeps automatic tuning ability of the PID controller parameters. Finally, the computer simulation and experimental result reveal disturbance rejection and good setpoint tracking performance of the proposed control method. The results clearly indicate effectiveness and merit of the proposed method.

This paper presents a direct adaptive motion controller using fuzzy wavelet neural networks (FWNN) for speed control of an electric two-wheeled robotic vehicle (ETWRV) with unknown parameters and uncertainties. With the decomposition of the overall system into two subsystems: yaw motion control and mobile inverted pendulum, two direct adaptive FWNN motion controllers are respectively proposed to achieve station keeping, speed following and yaw motion control. Asymptotic stabilities of the two controllers with their FWNN weighting updating rules are derived via the Lyapunov stability theory. Simulation results indicate that the proposed controllers are capable of providing satisfactory control actions to steer the vehicle.

This paper presents an intelligent adaptive steering control using linear quadratic regulation (LQR) approach and fuzzy cerebella model articulation control (CMAC) method for an electrical unicycle. The fuzzy CMAC is employed to on-line learn unknown frictions between the wheel and the terrain surfaces. The LQR approach is used to design a state feedback controller, in order to simultaneously achieve self-balancing and velocity control for the unicycle with different riders. The performance and merit of the proposed method are well exemplified by conducting simulations on a laboratory-built electric unicycle.

This paper presents a novel self-tuning PID control using recurrent wavelet neural networks (RWNN-PID) for a class of highly nonlinear discrete-time time-delay systems. The three-term parameters of the self-tuning PID controller are tuned based on the RWNN, in order to achieve setpoint tracking and eliminate any error caused by step disturbances. Numerical simulations for controlling two highly nonlinear process show disturbance rejection and setpoint tracking performance of the proposed control method, thus clearly indicating effectiveness and merit of the proposed method.

This paper presents an intelligent adaptive trajectory tracking controller using fuzzy basis function networks (FBFN) for an autonomous small-scale helicopter. With the on-line FBFN approximation to the vehicle mass and the coupling effect between the force and the moments, the intelligent adaptive controller is systematically synthesized using backstepping technique. This controller is shown to achieve the semi-global ultimate boundedness of the closed-loop helicopter dynamics and accommodate agile flight maneuvers in trajectory tracking. The effectiveness and merit of the proposed method are exemplified by performing one nonlinear simulation and by performance comparison with a well-known controller.

This paper presents a polar-space optimal kinematic controller design based on ant colony optimization (ACO) computing method for omnidirectional mobile robots with three independent driving wheels equally spaced at 120 degrees from one another. The optimal control parameters are obtained by minimizing the performance index using the proposed ACO computing method. These optimal parameters are used in the ACO-based polar-space kinematic controller to obtain better performance for omnidirectional mobile robots to achieve both trajectory tracking and stabilization. Simulation results are conducted to show the effectiveness and merit of the proposed ACO-based polar-space kinematic controller for omnidirectional mobile robots.

This paper presents an intelligent adaptive motion control using fuzzy basis-function networks (FBFN) for a self-balancing two-wheeled transporter (SBTWT). A mechatronic system structure driven by two DC motors is briefly described, and its nonlinear mathematical modeling incorporating the friction between the wheels and the motion surface is derived. With the decomposition of the overall system into two subsystems: yaw control and inverted pendulum, two intelligent adaptive FBFN controllers are proposed to achieve self-balancing, speed tracking and yaw motion control. Simulation results indicate that the proposed controllers are capable of providing appropriate control actions to steer the vehicle in desired manners.

This paper presents a nonsingular terminal sliding mode control using fuzzy wavelet networks for trajectory tracking and stabilization of a Mecanum wheeled omnidirectional vehicle (MWOV). A dynamic model of the vehicle including static friction is constructed; on the basis of the model, a nonsingular terminal sliding mode control is proposed to achieve trajectory tracking and stabilization. Fuzzy wavelet networks are then used to on-line approximate some uncertain nonlinear term of the controller, thereby accomplishing out better motion performance. Simulation results are conducted to show the effectiveness and the merit of the proposed approach.

This paper presents an adaptive backstepping sliding-mode motion controller using fuzzy basis function networks (FBFN) method for trajectory tracking of a self-balancing two-wheeled robot (SBTWR) with parameter variations. A decoupling method is proposed to decouple the robot's dynamic model such that the tracking controller can be synthesized using backstepping and sliding-mode control in both kinematic and dynamic levels. The FBFN is employed to on-line learn the uncertain parts of the tracking controller, thus achieving adaptive capability. Simulations results indicate that the proposed adaptive tracking controller is capable of providing satisfactory trajectory tracking performance.

This paper presents a direct adaptive motion control using fuzzy basis-function networks (FBFN) for a self-balancing two-wheeled transporter (SBTWT) with unknown parameters and uncertainties. With the decomposition of the overall system into two subsystems: yaw motion control and mobile inverted pendulum, two direct adaptive FBFN motion controllers are respectively proposed to achieve posture maintenance, speed following and yaw motion control. Asymptotic stabilities of the two controllers with their FBFN weighting updating rules are derived via the Lyapunov stability theory. Simulation results indicate that the proposed controllers are capable of providing satisfactory control actions to steer the vehicle

This paper presents an efficient parallel elite genetic algorithm (PEGA) for global path planning of an omnidirectional mobile robot moving in a static environment expressed by a grid-based map. This efficient PEGA, consisting of two parallel EGAs along with a migration operator, is proposed for global path planning of the mobile robots. The PEGA takes advantages of maintaining better population diversity, inhibiting premature convergence and keeping parallelism than conventional GAs do. The generated collision-free path is optimal in the sense of the shortest distance. The pipelined hardware implementation of IP (intellectual property) core library on a field-programmable gate array (FPGA) chip is employed to significantly speedup the processing time. Furthermore, a soft-core processor and a real-time operating system (RTOS) are embedded into the same chip to perform the global path planning using hardware/software co-design technique and SoPC (system-on-a-programmable-chip) concept. The merit and performance of the proposed SoPC-based PEGA are illustrated by conducting several simulations and experiments.

This paper presents adaptive sliding-mode control methods for self-balancing and yaw rate control of a dynamically two-wheeled human transportation vehicle (HTV) with mass variations and system uncertainties. The proposed controllers aim to provide consistent driving performance for system uncertainties and different drivers whose weights cause parameter variations of the HTV. By decomposing the overall system into the yaw subsystems and the self-balancing subsystems with parameters variations with respect to different riders, two adaptive sliding mode controls are proposed to achieve self-balancing and yaw control. Numerical simulations and experimental results on different terrains show that the proposed adaptive sliding mode controllers are capable of achieving satisfactory control actions to steer the vehicle.

This paper presents techniques for system design and human-robot interaction of an active mobile robotic assistant for the elderly people in known, cluttered and dynamic indoor environments. The RFID-based intelligent space is proposed to automatically help users to attain desired services and prevent from possible accidents. A useful human-robot interactive system (HRI) is presented which includes facial expressions, event reminder and two nursing-care functions. An operational scenario is presented for showing how the robot interacts with the user. Experimental results are conducted to show the merits and effectiveness of the proposed techniques.

This paper presents an adaptive nonlinear control using radial-basis-function neural network (RBFNN) for an electric unicycle. A mechatronic system structure of the unicycle is constructed and its simplified mathematical modeling is then established by using Newtonian mechanics and incorporating the frictions between the wheel and the terrain surface. An adaptive nonlinear control together with RBFNN is developed based on adaptive backstepping technique, in order to simultaneously achieve self-balancing and forward motion. Simulation results are conducted to illustrate feasibility and effectiveness of the proposed control method. The performance and merit of the proposed method are well exemplified by real riding test.

This paper develops a fuzzy mapping methodology for brightness improvement of color display systems using white sub-pixels. Based on human vision discrimination, three novel white sub-pixel filter structures without loss of original resolution are presented to enhance brightness of conventional stripe, Delta and PenTile color filter architectures. A new mapping fuzzy algorithm using RGB sub-pixel data around a white subpixel is proposed to improve overall image quality. Numerous simulation results are provided to show the efficacy of the proposed method for improving conventional stripe, delta and PenTile display systems. Experimental results are described which has been conducted to show that the proposed color display system performs well for electronic consumer products, such as notebooks, personal digital assistants, and mobile phone, etc.

This paper presents MIMO predictive control using recurrent neural networks for a class of nonlinear discrete-time systems. The recurrent-neural-network-based predictive control law is developed from the optimization of a generalized predictive performance criterion. A real-time adaptive control algorithm, including a neural predictor and a neural predictive controller, is proposed; the learning rates for both the neural predictor and controller are determined based on Lyapunov stability theory. Simulation results reveals that the proposed control strategy gives satisfactory tracking and disturbance rejection performance for two illustrative nonlinear multivariable systems.

This paper presents a recurrent neural network based predictive control for a class of nonlinear discrete time systems. The neural predictive control law is developed from the minimization of a generalized predictive performance criterion. A real time adaptive control algorithm, including a neural predictor and a neural predictive controller, is proposed; the adaptive learning rates for both the neural predictor and controller are determined based on Lyapunov stability theory. Simulation results reveal that the proposed control gives satisfactory tracking and disturbance rejection performance for two illustrative nonlinear systems. Experimental results for a variable frequency oil-cooling control process are performed which have shown effectiveness of the proposed method under the conditions of setpoint and load changes.

This paper develops methodologies and technologies for ultrasonic self-localization of an autonomous mobile robot (AMR) using a fuzzy adaptive extended information filtering scheme. A novel ultrasonic localization system consisting of two ultrasonic transmitters and three receivers, is presented to estimate both the static and dynamic position and orientation of the AMR. A fuzzy adaptive extended information filter (FAEIF) is presented to improve estimation accuracy and robustness for the proposed localization system, while the system lacks of sufficient information of complete models of the process and measurement noise varies with time. A static pose estimation utilizing the averaging approach is investigated as well. Six time-of-flight ultrasonic measurements together with the vehicle's dead-reckoned location information are merged to update the vehicle's dead-reckoned location information are merged to update the vehicle's pose by utilizing FAEIF sensor fusion algorithm. The proposed algorithm were implemented using an industrial personal computer with a computation speed of 800 MHz, and standard C++ programming techniques. The system prototype together with computer simulations and experimental results has been used to confirm that the system not only provides precise estimation of both the static and dynamic pose of the AMR, but also provides a simpler and more economical structure for navigation use and installation/calibration.