基于DRBF-EKF算法的车辆质心侧偏角与路面附着系数动态联合估计

车辆质心侧偏角和路面附着系数是实现车辆底盘智能化所需要的关键参数. 车辆质心侧偏角对于提高车辆安全性和操控性至关重要, 轮胎-路面附着系数决定轮胎力的峰值, 进而确定汽车的动力学稳定性边界. 本文针对四轮独立驱动电动汽车提出了一种基于惯性测量单元、轮毂电机内置转速/转角传感器的车辆质心侧偏角和路面附着系数动态联合估计方法. 对四轮独立驱动电动汽车进行车辆动力学分析, 结合Dugoff轮胎计算模型得到车辆质心侧偏角估计器; 利用机器学习中高维数据降维PCA多元分析方法, 提取主元特征参数, 建立路面附着系数估计器. 采用可自适应调节网络结构的双径向基神经网络和扩展卡尔曼滤波DRBF-EKF方法, 通过K-means算法改进RBF神经网络结构, 扩展卡尔曼滤波进行噪声滤波提高估计精度, 实现车辆质心侧偏角和路面附着系数的动态联合估计. 通过仿真和实车实验表明, 所设计的DRBF-EKF动态联合估计器实时性和估计精度均优于扩展卡尔曼滤波算法, 可以适应车辆行驶过程中路面附着特性与车速的变化, 表现出较强的鲁棒性; 与DRBF方法相比, 显著提高了估计精度; 并且分析了可以同时满足估计精度和实时性要求的最佳隐含层神经元个数.      

基于虚拟力概念的柔性臂空间机器人模糊退步自适应控制算法设计

讨论了漂浮基柔性臂空间机器人系统的动力学模拟、运动轨迹跟踪控制算法设计及柔性振动主动抑制。采用多体动力学建模方法并结合假设模态法,建立了漂浮基柔性臂空间机器人的系统动力学模型。基于该模型,针对系统惯性参数未知情况,提出了刚性运动基于模糊基函数网络自适应调节的退步控制算法,以完成柔性臂空间机器人载体姿态及机械臂各关节铰的协调运动。然后,为了主动抑制系统柔性振动,运用虚拟力的概念,构造了同时反映柔性模态和刚性运动轨迹的混合期望轨迹,通过改造原有的控制算法,提出了基于虚拟力概念的模糊退步自适应控制算法;这样不但保证了之前刚性运动控制方案对模型不确定的鲁棒性,而且能主动抑制柔性振动,从而提高了轨迹跟踪性能。理论分析及数值仿真算例均表明了控制方法的可行性。     

柔性空间机械臂协调运动的鲁棒神经网络控制及柔性振动PD反馈控制

针对具有未知惯性参数柔性空间机械臂系统的动力学和柔性振动控制问题,设计了基于奇异摄动理论的载体、关节铰轨迹跟踪的鲁棒神经网络控制算法和柔性振动反馈PD控制算法.首先,构建神经网络函数逼近慢变子系统的综合建模误差,设计载体、关节铰协调运动鲁棒控制算法,同时,通过稳定性分析选择神经网络自适应律;应用PD反馈控制来主动控制并消除快变子系统的柔性振动模态.仿真结果表明,所设计的控制算法解决了系统参数未知等因素带来的影响,能快速准确地进行轨迹追踪,同时,柔性杆的振动模态得到明显抑制,在2秒后基本消除.     

履带车辆路面激励响应的简化计算

为了快速计算履带车辆在路面不平度激励下的动力学响应,基于合理假设采用理论力学方法建立了某履带车辆的简化动力学模型,用ADAMS 软件建立了同种工况下的履带车辆动力学模型,并把两种模型的求解结果进行了对比,验证了简化模型的合理性. 研究结果为基于简化模型的动力学方程对车辆悬挂系统进行优化和逆动力学分析奠定了基础.     

柔性机械臂运动轨迹变结构双模控制

本文针对柔性机械臂运动轨迹的控制问题，采用假设模态法，利用Ｋａｎｅ春动力学模型，采用变结构控制理论跟踪关节转角的运动轨迹，当关切转角到达其终点的领域时，利用线性稳态器来控制末端点的弹性振动。提出了一种弯以模控制算法，并将其应用于单杆柔性机械臂的控制中，文中最后给出仿真研究结果，验证了该算法的有效性。     

基于非线性级联系统的弹性关节空间机器人新型滑模控制

研究了双臂弹性关节空间机器人的改进型非线性干扰观测器(nonlinear disturbance observer,NDO)设计、新型自适应动态终端滑模控制和弹性振动抑制问题。首先,考虑空间机器人的关节弹性,基于非线性级联系统的结构建立弹性关节空间机器人模型,分为外环机械臂动力学和内环关节动力学,具有渐近稳定性。针对外环机械臂动力学模型,设计基于改进型NDO的新型自适应动态终端滑模控制算法。针对内环关节动力学模型,设计力矩反馈控制算法来抑制弹性关节振动。本文提出的基于非线性级联系统的自适应动态终端滑模控制算法具有良好的动态特性及较强的鲁棒性,可在关节柔性刚度较小情况下,快速完成弹性关节振动抑制,实现空间机器人轨迹的精确跟踪。系统仿真试验证明了本文控制算法的正确性。     

存在关节力矩输出死区及外部干扰的漂浮基空间机械臂积分滑模神经网络自适应控制

探讨了载体位置和姿态都不受控时,漂浮基空间机械臂在带有关节力矩输出死区及外部干扰情况下轨迹跟踪的控制算法设计问题。死区与外部干扰影响系统的跟踪精度与稳定性。为此引入积分型切换函数,减少外部干扰引起的稳态误差,并利用径向基函数神经网络逼近动力学方程的未知部分,设计了一种积分滑模神经网络控制方案。控制算法的优点是,在死区斜率与边界参数不确定及最优逼近误差上确界未知的条件下,可以利用最优逼近误差、死区及干扰的补偿项来消除影响。李亚普诺夫稳定性分析证明了闭环系统的稳定性,且轨迹跟踪误差将收敛到0的某个小邻域内。仿真算例证实了该控制算法的有效性,实现了空间机械臂的轨迹跟踪控制。     

考虑轮胎力耦合约束的智能汽车轨迹跟踪控制算法

针对在低附着变速工况下,忽略纵、侧向轮胎力耦合约束可能导致轨迹跟踪时车辆失稳问题,提出了一种模型预测控制框架下的考虑轮胎力耦合约束的车辆轨迹跟踪控制方法。首先,通过摩擦圆假设建立纵、侧向轮胎力的耦合关系,并推导与之等效的输入量边界约束,将该问题转化为约束二次规划问题;其次,提出了一种基于交叉方向乘子法的数值求解构型,降低了求解约束优化问题时Karush-Kuhn-Tucker方程的维数,实现了求解加速。仿真结果表明,在低附着变速工况下,所提出的算法能够实现最大0.166 m误差的稳定跟踪;同时数值求解过程最多仅需8次迭代,增强了控制过程的实时性。     

基于轮胎非线性特性的汽车动力学问题

长期以来,人们对轮胎的非线性进行了大量的理论与试验研究,总结出各种理论模型与经验模型。利用这些非线性轮胎模型建立汽车动力学的非线性常微分方程组,通过数值积分,可以获得汽车在各种工况条件下的稳态与瞬态转向特性。但这些模型的普遍缺点是不能用于对汽车行驶的稳定性作定性分析。本文提出了一种轮胎非线性侧特性的摄动模型,利用近似解析方法,讨论了轮胎非线性特性对汽车的转向特性、动态响应和汽车行驶稳定性的影响,导     

EXPERIMENTAL ANALYSIS OF ACTIVE STEERING HARDWARE-IN-LOOP BASED ON SLIDING MODE CONTROL 1)

为了提高重型车辆在转向过程中的稳定性和安全性,本文提出了一种基于滑模变结构控制的主动前轮转向控制策略,基于这种策略设计了主动转向控制器,建立了三轴商用车的二自由度车辆动力学简化模型及整车模型,利用TruckSim--Simulink建立联合仿真平台以及进行硬件在环实验。在不同工况、不同车速下,分别对有无主动转向控制器的车辆进行了操纵稳定性分析,并在此基础上进行了滑模变结构控制的主动转向影响因素敏感性分析。实验结果表明,这种控制器策略在不同工况下具有较强的适应性。     

Design of neural network-based control systems for active steering system

Nowadays, safety of road vehicles is an important issue due to the increasing road vehicle accidents. Passive safety system of the passenger vehicle is to minimize the damage to the driver and passenger of a road vehicle during an accident. Whereas an active steering system is to improve the response of the vehicle to the driver inputs even in adverse situations and thus avoid accidents. This paper presents a neural network-based robust control system design for the active steering system. Primarily, double-pinion steering system used modeling of the active steering system. Then four control structures are used to control prescribed random trajectories of the active steering system. These control structures are as classical PID Controller, Model-Based Neural Network Controller, Neural Network Predictive Controller and Robust Neural Network Predictive Control System. The results of the simulation showed that the proposed neural network-based robust control system had superior performance in adapting to large random disturbances.     

Distributed coordinated tracking of multiple autonomous underwater vehicles

This paper considers the distributed coordinated tracking problem of multiple autonomous underwater vehicles with a time-varying reference trajectory. Each vehicle is subject to model uncertainty and time-varying ocean disturbances. A novel predictor-based neural dynamic surface control design approach is proposed to develop the node controllers, under which synchronization between vehicles can be reached on condition that the augmented graph induced by the vehicles and the reference trajectory contains a spanning tree. The prediction errors are used to update the neural adaptive laws, which enable fast identifying the vehicle dynamics without excessive knowledge of their dynamical models. Further, this result is extended to the output-feedback case where only position-yaw information can be measured. A local predictor, based on its own position-yaw information, is constructed, not only to recover the unmeasured velocity information, but also to identify the unknown dynamics for each vehicle. A linear matrix inequality-based analysis is performed for the stability of the predictor. Then, distributed output-feedback tracking controllers are developed to achieve synchronization between vehicles in the presence of unknown dynamics and unmeasured velocities. For both cases, the stability properties of the closed-loop network are established via Lyapunov analysis. Simulation results demonstrate the effectiveness of the proposed methods.     

A feedback linearization design for the control of vehicle’s lateral dynamics

In this paper, the feedback linearization scheme is applied to the control of vehicle’s lateral dynamics. Based on the assumption of constant driving speed, a second-order nonlinear lateral dynamical model is adopted for controller design. It was observed in (Liaw, D.C., Chung, W.-C. in 2006 IEEE International Conference on Systems, Man, and Cybernetics, 2006) that the saddle-node bifurcation would appear in vehicle dynamics with respect to the variation of the front wheel steering angle, which might result in spin and/or system instability. The vehicle dynamics at the saddle node bifurcation point is derived and then decomposed as an affine nominal model plus the remaining term of the overall system dynamics. Feedback linearization scheme is employed to construct the stabilizing control laws for the nominal model. The stability of the overall vehicle dynamics at the saddle-node bifurcation is then guaranteed by applying Lyapunov stability criteria. Since the remaining term of the vehicle dynamics contains the steering control input, which might change system equilibrium except the designed one. Parametric analysis of system equilibrium for an example vehicle model is also obtained to classify the regime of control gains for potential behavior of vehicle’s dynamical behavior.     

多体系统轨迹跟踪的瞬时最优控制保辛方法

随着近年来机器人在各行业领域的广泛应用,对机器人的动力学与控制性能不断提出新的要求,特别是对设计越来越复杂、操作越来越灵巧的智能机器人,要求其能够对目标轨迹实现高精度跟踪以满足实际工作需求. 因此,针对机器人多体系统对目标轨迹跟踪的任务需求,基于微分代数方程提出瞬时最优控制保辛方法. 首先,采用多体动力学绝对坐标建模方法建立机器人系统的普适动力学方程,即微分代数方程;然后,采用保辛方法将连续时间域内的微分代数方程进行离散化,进而得到以当前位置、速度和拉式乘子为未知量的非线性代数方程组;其次,通过引入对目标轨迹跟踪以及对控制加权的瞬时最优性能指标,根据瞬时最优控制理论获得当前最优控制输入;最后,通过离散时间步的更新完成对目标轨迹的跟踪任务. 为了验证本文方法的有效性,以双摆轨迹跟踪控制为例进行了数值仿真,结果表明:针对机器人轨迹跟踪任务所提出的瞬时最优控制保辛方法能够实现对目标轨迹的高精度跟踪,且瞬时最优控制由受控微分代数方程推导获得,更具一般性,能够适应其他复杂多体系统的轨迹跟踪控制问题.      

A new contact & slip model for tracked vehicle transient dynamics on hard ground

This study presents a new general transient contact and slip model for tracked vehicles on hard ground which is simple, accurate, and in agreement with the test results to a satisfactory level. Simulating zero track speed instances become possible with the new contact/shear model which is the major proposed improvement in addition to more accurate results for transient steering and tractive inputs. The model represents a general tracked vehicle having rear or front sprockets, with parameters for center of gravity, wheel positions, number of wheels, and track-pretention. To calculate longitudinal and lateral forces, a transient shear model is used. Shear stress under each track pad is assumed to be a function of shear displacement. The contact time formulation used in shear displacement calculation is improved to gain accuracy for transient and zero track speed conditions.The model is implemented on the Matlab/Simulink platform and verified with a comprehensive program of road tests composed of transient steering and tractive/braking scenarios. The results of the simulations and the road tests are satisfactorily similar for both constant and transient input maneuvers. Moreover, sensitivity simulations for vehicle parameters are conducted to show that the model responses are inline with the expected vehicle dynamics behaviours.     

Algorithms for a Vehicle Dynamics Monitoring System,Based on Model Reference Structure

Vehicle control depends heavily on the knowledge of the vehicle operatingconditions. One of the most important parameters for its control is thetyre–road friction coefficient (µ). An appropriate way toestimate the vehicle operating conditions is the Model Reference Approach.This technique requires a model that provides estimated states which can becompared with the measured states, the difference is used to determine thereal operating conditions. This paper presents two different applications ofthe Model Reference Techniques to estimate tyre–road frictioncoefficient; these are based on the relation between tyre forces and slip,and on the vehicle lateral behaviour using an extended Kalman filter.Experimental data from the test vehicle confirms the good results obtainedin the friction estimation based on the tyre slip–force relation. Theestimation using an extended Kalman filter on lateral behaviour showsaccurate tracking. The next step to be taken is to integrate all thealgorithms in the test vehicle and to validate them for a wide range ofoperating conditions, in order to have reliable information for the activesystem control.     

An instrumented vehicle for offroad dynamics testing

The paper presents an instrumented vehicle that was equipped with measuring systems to perform complete dynamics tests, especially in off-road conditions. The equipment consists of four wheel dynamometers, a steering robot, and a differential GPS system together with an inertial platform, a non-contact vehicle speed sensor, and an on-board computer with software to control the devices and collect experimental data. The four wheel dynamometers measure six elements; based on strain gage force transducers, it measures three orthogonal forces and three moments. The steering robot can control the steering wheel of the vehicle at a variety of excitation modes; it can carry out typical vehicle dynamics tests (ISO 7401, ISO 4138, ISO/TR3888, etc.) as well as custom engineered tests at a wide range of setting parameters (steer angle rate up to 1600 deg/s). The differential GPS system gives true time vehicle kinematics data (velocities, accelerations, angles, etc.) at 10-ns sample rate and 20-mm accuracy. The base vehicle, a Suzuki Vitara 4 × 4, required no special modifications or changes to install the measuring equipment. The paper also describes typical tests performed with the use of the instrumented vehicle together with sample results.     

Velocity-sensorless proportional–derivative trajectory tracking control with active damping for quadcopters

The proposed observer-based control mechanism solves the trajectory tracking problem in the presence of external disturbances with the reduction in sensor numbers. This systematically considers the quadcopter nonlinear dynamics and parameter and load variations by adopting the standard controller design approach based on a disturbance observer (DOB). The first feature is designing first-order observers for estimating the velocity and angular velocity error, with their parameter independence obtained from the DOB design technique. As the second feature, the resultant velocity observer-based control action including active damping and DOBs secures first-order tracking behavior for the position and attitude (angle) loops through pole zero cancellation, thereby forming a proportional–derivative control structure. Closed-loop analysis results reveal the performance recovery and steady-state error removal properties in the absence of tracking error integrators. The numerical verification confirms the effectiveness of the proposed mechanism using MATLAB/Simulink.