【作者】 张锐；

【导师】 谢志江；

【作者基本信息】 重庆大学 ， 机械（专业学位）， 2024， 硕士

【摘要】 相较于传统的移动机器人,四轮独立驱动转向的移动机器人具备更强的机动性和适应能力。然而其强耦合、冗余驱动的复杂特性给运动控制方法提出了挑战。尤其在转向时,移动机器人会产生车轮难以协调控制、偏离期望轨迹、车体摆动等现象,给自动化转运带来了安全性问题。本文以项目背景下的四轮转向移动机器人为研究对象,提出并设计了一种分层的横向控制方法,顶层为状态观测器和路径跟踪控制器,中层为力矩决策,底层为力矩优化分配。目的在于提升移动机器人转向时的跟踪精度、横摆稳定性和工作效率。本文具体研究内容如下:针对四轮转向移动机器人的底盘特性与工作需求,建立包含横向、纵向和横摆的三自由度整车动力学模型。通过建立电机、车轮动力学以及轮胎模型,完成了驱动单元参数计算。为便于控制器设计,基于阿克曼转向定理得出了车轮转向角关系,并建立了单轨动力学模型。通过Matlab/Simulink与Carsim联合仿真,验证了动力学模型的准确性。设计了移动机器人状态观测器。利用扩展卡尔曼滤波器算法,将横摆角速度、质心侧偏角和纵向速度作为估计参数,侧向加速度作为观测参数,建立移动机器人状态空间方程和测量方程,进行线性化和离散化处理,并转化为扩展卡尔曼滤波器的可实现形式。仿真结果验证了该状态观测器的鲁棒性和实时性。基于模型预测控制设计了移动机器人路径跟踪控制器。基于非线性动力学方程建立预测模型,设定包含控制量、输出量和轮胎侧偏角的约束条件,创建系统目标函数并将其转化为二次规划标准形式求解。不同速度工况下的仿真结果显示,路径跟踪控制器可以对车轮转角进行实时最优控制,有效地跟踪期望轨迹。设计了移动机器人横摆稳定控制器,并对分层控制器进行了仿真验证。经过稳定性参数分析,结合顶层控制器,对横摆稳定控制器进行设计。中层力矩决策模块一方面将移动机器人的横摆角速度、质心侧偏角与期望值的偏差和偏差率作为控制量,作用于滑模控制器,输出期望的附加横摆力矩;另一方面将纵向速度和期望速度作用于PID车速跟踪控制器,输出广义纵向力矩;底层力矩优化分配模块以轮胎负荷率加权平方和最低为目标函数,并设置约束条件,对决策出的力矩进行合理分配,保证四个车轮协调运转。搭建仿真平台,设计对比试验,结果证明本文搭建的横向运动控制系统可以达到转角调控和横摆力矩协同控制的效果。更多还原

【Abstract】 Compared with the traditional mobile robot,the mobile robot with four-wheel independent drive steering has stronger mobility and adaptability.However,its complex characteristics of strong coupling and redundancy drive pose challenges to motion control methods.Especially in the steering,the mobile robot will produce wheel coordination control,deviation from the expected trajectory,body swing and other phenomena,which brings safety problems to the automatic transfer.Taking the four-wheel steering mobile robot under the project background as the research object,this paper proposes and designs a hierarchical horizontal control method,with the top layer being the state observer and path tracking controller,the middle layer being the moment decision making,and the bottom layer being the moment optimization distribution.The aim is to improve the tracking accuracy,yaw stability and working efficiency of the mobile robot when it turns.The specific research content of this paper is as follows:According to the chassis characteristics and working requirements of four-wheel steering mobile robot,a three-degree-of-freedom vehicle dynamics model including transverse,longitudinal and yaw is established.Through the establishment of motor,wheel dynamics and tire models,the driving unit parameters are calculated.In order to facilitate the design of the controller,the wheel steering Angle relation is obtained based on Ackermann steering theorem,and the monorail dynamics model is established.The accuracy of the dynamic model was verified by Matlab/Simulink and Carsim co-simulation.A mobile robot state observer is designed.Using the extended Kalman filter algorithm,the yaw velocity,side deflection Angle of the center of mass and longitudinal velocity are taken as the estimation parameters,and the lateral acceleration is taken as the observation parameters.The state space equation and measurement equation of the mobile robot are established,linearized and discretized,and converted into the realizable form of the extended Kalman filter.Simulation results verify the robustness and real-time performance of the state observer.The path tracking controller of mobile robot is designed based on model predictive control.Based on the nonlinear dynamic equation,the prediction model is established,the constraint conditions including control quantity,output quantity and tire side deflection Angle are set up,and the system objective function is created and converted into quadratic programming standard form.The simulation results under different speed conditions show that the path tracking controller can control the wheel Angle optimally in real time and track the desired trajectory effectively.A mobile robot yaw stabilization controller is designed,and the layered controller is verified by simulation.After analyzing the stability parameters and combining the top-level controller,the yaw stabilization controller is designed.On the one hand,the middle layer moment decision module takes the yaw velocity of the mobile robot,the deviation of the side yaw Angle of the center of mass from the expected value and the deviation rate as the control quantity,and acts on the sliding mode controller to output the expected additional yaw moment.On the other hand,the longitudinal speed and the expected speed are applied to the PID speed tracking controller,and the generalized longitudinal torque is output.The bottom torque optimization allocation module takes the lowest weighted sum of square of tire load rate as the objective function,and sets constraint conditions to distribute the decided torque reasonably,so as to ensure the coordinated operation of four wheels.The simulation platform is built and the comparative test is designed.The results show that the lateral motion control system built in this paper can achieve the effect of Angle control and yaw moment collaborative control.更多还原