Neural adaptive PID formation control of car-like mobile robots without velocity measurements

A virtual leader–follower formation control of a group of car-like mobile robots is addressed in this paper. First, the kinematic and dynamic models of car-like robots are transformed into a second-order leader–follower formation model which inherits all structural properties of the robot dynamic model. Then, a new observer-based proportional–integral-derivative formation controller is proposed to force that all robots construct a desired formation with respect to a predefined virtual leader. To improve the formation tracking and observation performance, the integral action is incorporated into the design of the observer–controller scheme. Adaptive robust and neural network techniques are also employed to compensate uncertain parameters, unmodeled dynamics, and external disturbances. Lyapunov’s direct method is utilized to show that the formation tracking and observation errors are semi-globally uniformly ultimately bounded. Then, the proposed controller is extended to the leader–follower formation of a team of tractor–trailer systems. Finally, simulation results illustrate the efficiency of the proposed controller.     

Remote secure decentralized control strategy for mobile robots

The main goal of this paper is to define, study and analyze a remote control architecture for a set of non-holonomic robotic vehicles. This project gathers three laboratories and the French Army Research Office. Each of these laboratories deals with a part of this multidisciplinary project which includes coordinated control, control architecture, control with time delay and monitoring of the wireless network. In this paper, we present the whole goal of this project including the basic experimental setup developed to validate our control algorithm. We also focus on a new decentralized control strategy that uses the Leader–Follower principle. The originality of this paper stems from the use of the signal level of wireless connection as a control vector. Indeed, each vehicle is fitted with two wireless devices. One of them is equipped with a sector antenna fitted on a DC motor to track the direction of best reception. Thus, it allows us to find the relative angular position of the Follower pointing out the Leader. Using wireless technology as a sensor, instead of vision for instance, allows a longer distance of the coordinated control loop between each vehicle (approximately 100 m) even if GPS information is not available.     

An efficient initialization approach of Q-learning for mobile robots

This article demonstrates that Q-learning can be accelerated by appropriately specifying initial Q-values using dynamic wave expansion neural network. In our method, the neural network has the same topography as robot work space. Each neuron corresponds to a certain discrete state. Every neuron of the network will reach an equilibrium state according to the initial environment information. The activity of the special neuron denotes the maximum cumulative reward by following the optimal policy from the corresponding state when the network is stable. Then the initial Q-values are defined as the immediate reward plus the maximum cumulative reward by following the optimal policy beginning at the succeeding state. In this way, we create a mapping between the known environment information and the initial values of Q-table based on neural network. The prior knowledge can be incorporated into the learning system, and give robots a better learning foundation. Results of experiments in a grid world problem show that neural network-based Q-learning enables a robot to acquire an optimal policy with better learning performance compared to conventional Q-learning and potential field-based Qlearning.     

Cyclic error correction based Q-learning for mobile robots navigation

International Journal of Control, Automation and Systems - Similar to control systems, reinforcement learning can capture notions of optimal behavior using natural interaction experience. In the...     

EP-based kinematic control and adaptive fuzzy sliding-mode dynamic control for wheeled mobile robots

This paper proposes a complete control law comprising an evolutionary programming based kinematic control (EPKC) and an adaptive fuzzy sliding-mode dynamic control (AFSMDC) for the trajectory-tracking control of nonholonomic wheeled mobile robots (WMRs). The control gains for kinematic control (KC) are trained by evolutionary programming (EP). The proposed AFSMDC not only eliminates the chattering phenomenon in the sliding-mode control, but also copes with the system uncertainties and external disturbances. Additionally, the convergence of trajectory-tracking errors is proved by the Lyapunov stability theory. Computer simulations are presented to confirm the effectiveness of the proposed complete control law. Finally, real-time experiments are done in the test field to demonstrate the feasibility of real WMR maneuvers.     

Robust adaptive neural network control for environmental boundary tracking by mobile robots

The paper addresses the problem of environmental boundary tracking for the nonholonomic mobile robot with uncertain dynamics and external disturbances. To do environmental boundary tracking, a reference velocity is designed for the nonholonomic mobile robot. In this paper, a radial basis function neural network (NN) is used to approximate a nonlinear function containing the uncertain model terms and the elements of the Hessian matrix of the environmental concentration function. Then, the NN approximator is combined with a robust control to construct a robust adaptive NN control for the mobile robot to track the desired environment boundary. It is proved that the tracking error can be guaranteed to converge to zero in the ultimate. Simulation results are presented to illustrate the stability of the robust adaptive control. Copyright © 2011 John Wiley & Sons, Ltd.     

Distributed adaptive control for consensus tracking with application to formation control of nonholonomic mobile robots

In this paper, we investigate the output consensus problem of tracking a desired trajectory for a class of systems consisting of multiple nonlinear subsystems with intrinsic mismatched unknown parameters. The subsystems are allowed to have non-identical dynamics, whereas with similar structures and the same yet arbitrary system order. And the communication status among the subsystems can be represented by a directed graph. Different from the traditional centralized tracking control problem, only a subset of the subsystems can obtain the desired trajectory information directly. A distributed adaptive control approach based on backstepping technique is proposed. By introducing the estimates to account for the parametric uncertainties of the desired trajectory and its neighbors’ dynamics into the local controller of each subsystem, information exchanges of online parameter estimates and local synchronization errors among linked subsystems can be avoided. It is proved that the boundedness of all closed-loop signals and the asymptotically consensus tracking for all the subsystems’ outputs are ensured. A numerical example is illustrated to show the effectiveness of the proposed control scheme. Moreover, the design strategy is successfully applied to solve a formation control problem for multiple nonholonomic mobile robots.     

Cooperative strategy based on adaptive Q-learning for robot soccer systems

The objective of this paper is to develop a self-learning cooperative strategy for robot soccer systems. The strategy enables robots to cooperate and coordinate with each other to achieve the objectives of offense and defense. Through the mechanism of learning, the robots can learn from experiences in either successes or failures, and utilize these experiences to improve the performance gradually. The cooperative strategy is built using a hierarchical architecture. The first layer of the structure is responsible for assigning each role, that is, how many defenders and sidekicks should be played according to the positional states. The second layer is for the role assignment related to the decision from the previous layer. We develop two algorithms for assignment of the roles, the attacker, the defenders, and the sidekicks. The last layer is the behavior layer in which robots execute their behavior commands and tasks based on their roles. The attacker is responsible for chasing the ball and attacking. The sidekicks are responsible for finding good positions, and the defenders are responsible for defending competitor scoring. The robots' roles are not fixed. They can dynamically exchange their roles with each other. In the aspect of learning, we develop an adaptive Q-learning method which is modified form the traditional Q-learning. A simple ant experiment shows that Q-learning is more effective than the traditional techniques, and it is also successfully applied to the learning of the cooperative strategy.     

Approximation-based adaptive control for a class of mobile robots with unknown skidding and slipping

An adaptive tracking control approach using function approximation technique is proposed for trajectory tracking of Type (2,0) wheeled mobile robots with unknown skidding and slipping in polar coordinates and at the dynamic level. The nonlinear disturbance observer (NDO) is used to estimate a nonlinear disturbance term including unknown skidding and slipping. The adaptive control system is designed via the function approximation technique using neural networks employed to compensate the NDO error. It is proved that all signals of the controlled closed-loop system are uniformly bounded and the point tracking errors converge to an adjustable neighborhood of the origin regardless of large initial tracking errors and unknown skidding and slipping. Simulation results are presented to validate the good tracking performance and robustness of the proposed control system against unknown skidding and slipping.     

A hybrid navigation strategy for multiple mobile robots

A hybrid navigation strategy is proposed in this paper for solving the navigation problem of multiple mobile robots. The proposed strategy integrates three algorithms that represent three different types of existing methods in a layered system. The bottom-up architecture of this system is the main contribution of this paper. This architecture pursues reliable low-level layers that can independently work in as much cases as possible, and the high-level layer is used only when it is necessary for guaranteeing convergence in complex situations. The simulation results show that the proposed strategy has well combined the algorithms of different types from the perspective of pursuing reactivity in the premise of ensuring convergence. Compared with the traditional top-down hybrid architecture, the bottom-up architecture proposed in this paper is more suitable for multi-robot navigation since it can better utilize the advantages of different algorithms to deal with different situations. The experiments on real robots have further verified the applicability of the proposed strategy.     

A reinforcement learning with evolutionary state recruitment strategy for autonomous mobile robots control

In recent robotics fields, much attention has been focused on utilizing reinforcement learning (RL) for designing robot controllers, since environments where the robots will be situated in should be unpredictable for human designers in advance. However there exist some difficulties. One of them is well known as ‘curse of dimensionality problem’. Thus, in order to adopt RL for complicated systems, not only ‘adaptability’ but also ‘computational efficiencies’ should be taken into account. The paper proposes an adaptive state recruitment strategy for NGnet-based actor-critic RL. The strategy enables the learning system to rearrange/divide its state space gradually according to the task complexity and the progress of learning. Some simulation results and real robot implementations show the validity of the method.     

Cooperative adaptive consensus tracking for multiple nonholonomic mobile robots

This paper addresses the cooperative adaptive consensus tracking for a group of multiple nonholonomic mobile robots, where the nonholonomic robot model is assumed to be a canonical vehicle having two actuated wheels and one passive wheel. By integrating a kinematic controller and a torque controller for the nonholonomic robotic system, a cooperative adaptive consensus tracking strategy is developed for the uncertain dynamic models using Lyapunov-like analysis in combination with backstepping approach and sliding mode technique. A key feature of the developed adaptive consensus tracking algorithm is the introduction of a directed network topology into the control constraints based on algebraic graph theory to characterise the communication interaction among robots, which plays an important role in realising the cooperative consensus tracking with respect to a specific common reference trajectory. Furthermore, a novel framework is proposed for developing a unified methodology for the convergence analysis of the closed-loop control systems, which can fully ensure the desired adaptive consensus tracking for multiple nonholonomic mobile robots. Subsequently, illustrative examples and numerical simulations are provided to demonstrate and visualise the theoretical results.     

Hybrid autonomous control for multi mobile robots

Reinforcement learning can be an adaptive and flexible control method for autonomous system. It does not need a priori knowledge; behaviors to accomplish given tasks are obtained automatically by repeating trial and error. However, with increasing complexity of the system, the learning costs are increased exponentially. Thus, application to complex systems, like a many redundant d.o.f. robot and multi-agent system, is very difficult. In the previous works in this field, applications were restricted to simple robots and small multi-agent systems, and because of restricted functions of the simple systems that have less redundancy, effectiveness of reinforcement learning is restricted. In our previous works, we had taken these problems into consideration and had proposed new reinforcement learning algorithm, 'Q-learning with dynamic structuring of exploration space based on GA (QDSEGA)'. Effectiveness of QDSEGA for redundant robots has been demonstrated using a 12-legged robot and a 50-link manipulator. However, previous works on QDSEGA were restricted to redundant robots and it was impossible to apply it to multi mobile robots. In this paper, we extend our previous work on QDSEGA by combining a rule-based distributed control and propose a hybrid autonomous control method for multi mobile robots. To demonstrate the effectiveness of the proposed method, simulations of a transportation task by 10 mobile robots are carried out. As a result, effective behaviors have been obtained.     

Robust control laws for wheeled mobile robots

The paper deals with feedback control of wheeled mobile robots. The proposed controller is based on a tracking scheme where the real robot tracks a fictitious reference one with equivalent kinematical properties. A solution to the parking problem is derived from the basic tracking scheme by considering a reference vehicle which converges to the desired configuration. The goal achievement is analysed by the stability of the zero equilibrium point of the tracking state error. Simulation as well as experimental results illustrate these controllers' designs. Robustness with respect to errors in the state estimation is investigated. It is defined by the existence of a compact attractive domain around the zero error. A practical computation of such a domain is a result of this paper.     

非完整移动机器人的复合编队控制

主要研究了非完整自主机器人之间的队形保持和避障问题,提出了一种新的复合编队控制方法,该方法根据机器人的期望位置在其运动约束区域内外的不同,分别以一种灵活的反馈线性化算法和最优近似目标算法来建立控制规则,并提出了编队环境中存在静态障碍物时的队形控制策略,从而实现多机器人的稳定编队控制．该方法降低了传统线性反馈控制对编队初始误差范围的要求,并且解决了非完整机器人编队的避障问题．实验结果表明了该编队控制方法的可行性和有效性．     

An adaptive technique for trajectory tracking control of a wheeled mobile robots without velocity measurements

This paper addresses an adaptive method for designing a sensorless trajectory tracking control scheme for a wheeled mobile robot. In order to reduce the cost of the robot, a new Nonlinear Observer (NOB) is used to leave out velocity sensors in the robot. Also, an adaptive model reference technique is used for designing the dynamic controller. In order to ensure the implementability of proposed controller, dynamic controller and nonlinear observer are designed in the presence of uncertainties. In addition, the Observer-based Kinematic Controller (OKC) is designed in the presence of sliding velocity. In order to improve the performance of the kinematic controller, sliding velocity is estimated and used for modification of kinematic controller. Finally, the effectiveness of the proposed method is demonstrated by simulations.     

轮式移动机器人嵌入式自适应控制器设计与仿真

在增加一阶输出信息基础上,改写了控制律算法准则函数,使用变分原理得出紧格式的改进无模型自适应控制(eMFAC)方程.以差动驱动移动机器人为研究对象,将非完整约束及传动系统误差作为外部干扰设计了具有前馈功能的Kalman改进无模型自适应控制(MFAC)运动控制器,证明了新控制方法的全局稳定性.FreeMAT仿真及ARM7系统控制实验证实了本方法的有效性.     

一种新型的移动机器人轨迹跟踪控制方法

主要是对非完整约束下移动机器人的轨迹跟踪控制进行了研究,提出了一种新型的基于移动机器人运动模型、具有全局渐近稳定性的跟踪控制方法。这种非线性控制方法主要分为前馈和反馈两个部分：前馈部分是一种滑模控制器,它是基于反演设计的思想设计了切换函数,采用指数趋近律,减少了滑模变结构控制的抖动,并使用Lyapunov第一法对控制系统进行了稳定性分析,证明了滑模跟踪控制器是稳定的;反馈部分是基于Lyapunov函数的方法设计的反馈控制器。通过前馈部分和反馈部分的相互作用,提高了移动机器人轨迹跟踪控制的精度。实验结果表明与一般的跟踪控制方法相比,控制效果明显改善,跟踪误差能在较短时间内收敛,具有很好的抗干扰性能。     

非完整移动机器人全局路径跟踪控制

根据制导路径跟踪理论,提出了一种非完整移动机器人全局路径跟踪控制方法.这一方法首先在路径坐标系上计算实际位置与期望位置的误差,利用制导的路径跟踪理论,导出消除该误差所需的姿态角和路径参数更新律,然后据此求解角速度及实际控制.文中还给出了初始路径参考点的计算方法,分析了路径跟踪方向和反转方法.稳定性分析证明该方法没有控制奇异点,受控闭环系统全局一致渐近稳定.最后通过移动机器人典型路径跟踪实验验证了所提出方法的可行性.