基于模糊神经网络的混联式混合动力汽车驱动系统工作状态控制研究

通过分析混联式混合动力汽车理想工作模式的控制,提出了一种新型的汽车驱动系统工作模式的控制方法——模糊神经网络控制.采用模糊神经网络能够自适应的控制汽车驱动模式工作状态的切换,能够节能减排,提高发动机的动力.利用Matlab/Simulink建立模糊神经网络控制模型,仿真实验结果表明,该控制方法达到了预期的目的,具有较好的效果.     

基于扩张状态观测器估计的混合动力汽车协调控制

为了提高功率分流式混合动力汽车模式切换的稳定性,提出干扰补偿的转矩协调控制策略. 针对发动机动态响应振荡及车辆行驶工况多变的问题,设计多变量线性扩张状态观测器,从频域角度验证了观测器对于上述2种干扰的估计精确性. 研究不同干扰对基础电机补偿控制稳定性的影响,指出负载干扰对车辆模式切换响应的影响最大,引起的切换冲击最大可至24.5 m/s³. 提出基于干扰补偿的动力源转矩再分配算法,开展仿真验证. 结果表明,该协调控制策略在受到明显的外界干扰时能够保证系统的稳定性及模式切换的平顺性.     

高速无人驾驶车辆轨迹跟踪和稳定性控制

针对高速无人驾驶车辆运动控制过程中轨迹跟踪精度和稳定性难以同时保障的问题,提出综合前馈-反馈及自抗扰控制(ADRC)补偿相结合的横向控制算法. 通过车速和道路曲率信息计算前馈稳态前轮转向角,将质心侧偏角引入航向偏差,以车辆航向角偏差和侧向偏差作为参考量进行反馈控制,通过前馈-反馈控制提升瞬态轨迹跟踪性能. 设计自抗扰控制器,通过扩张状态观测器对未建模动态和内外界干扰进行估计,通过将后轮侧偏角控制在参考值附近来补偿前轮转角,提升无人驾驶车辆的转向稳定性和控制器的鲁棒性. 不同工况下的仿真结果表明,利用该方法可以保证高速无人驾驶车辆稳定地跟踪期望路径行驶,轨迹跟踪偏差较小,对车辆参数变化和外界干扰具有较强的鲁棒性.     

基于理想转向传动比的汽车线控转向控制算法

以29自由度汽车动力学模型为基础,提出了保证汽车转向增益不变的理想传动比稳态控制策略,使线控转向汽车转向特性不受车速和方向盘转角变化的影响;提出了基于状态反馈的动态校正稳定性控制算法。仿真和驾驶模拟器实验表明,基于理想转向传动比的稳态控制策略保证了汽车转向增益不变,减轻了驾驶员的负担,适合于更多的驾驶人群;基于状态反馈的动态校正稳定性控制算法有效提高了汽车的稳定性。     

The improvement of Clausius entropy and its application in entropy analysis

The defects of Clausius entropy which include a premise of reversible process and a process quantity of heat in its definition are discussed in this paper. Moreover, the heat temperature quotient under reversible conditions, i.e. (δQ/T)_rev, is essentially a process quantity although it is numerically equal to the entropy change. The sum of internal energy temperature quotient and work temperature quotient is defined as the improved form of Clausius entropy and it can be further proved to be a state function. Unlike Clausius entropy, the improved definition consists of system properties without premise just like other state functions, for example, pressure p and enthalpy h, etc. It is unnecessary to invent reversible paths when calculating entropy change for irreversible processes based on the improved form of entropy since it is independent of process. Furthermore, entropy balance equations for internally and externally irreversible processes are deduced respectively based on the concepts of thermal reservoir entropy transfer and system entropy transfer. Finally, some examples are presented to show that the improved definition of Clausius entropy provides a clear concept as well as a convenient method for entropy change calculation. Supported by the National Basic Research and Development Program of China (Grant No. 2007CB206901)     

用于爆胎车辆仿真的驾驶员模型

分析了驾驶员模型在汽车爆胎仿真应用中存在的问题及爆胎后车辆响应特性的变化,指出爆胎后车辆响应特性发生改变是导致驾驶员模型不能控制爆胎后汽车回到原路径的根本原因。建立了考虑爆胎后车辆左、右转向特性不对称的驾驶员模型,并进行了仿真。最后在原驾驶员模型中加入状态参考器,并对不同车速下的状态参考器参数进行辨识。解决了驾驶员模型在汽车爆胎仿真中的应用问题,为进一步研究汽车爆胎后驾驶员的操作规律及爆胎后汽车的稳定性控制提供了参考。     

基于状态反馈精确线性化Buck变换器的微分平坦控制

针对Buck变换器由于输入电压和输出电流等外部扰动及电路参数摄动等不确定因素引起系统输出电压变化的问题,提出一种基于状态反馈精确线性化的微分平坦控制方法。根据状态空间平均方程建立了变换器的仿射非线性模型,通过坐标变换推导出状态反馈控制率。在状态反馈精确线性化模型基础上,设计了微分平坦控制器,同时将扰动观测器嵌入控制器中,进一步提高了系统对电路参数摄动和外部扰动等内外干扰的处理能力。最后应用灵敏度理论分析了变换器的稳定性,可知减小电感量、增大电容值有利于提高系统的稳定性。实验结果表明,与传统PI控制方法相比,本文所提控制方法对输入电压和输出电流扰动具有更强鲁棒性,输出电压纹波更小,且系统响应速度能提升近50%。本文研究对DC-DC变换器控制器的设计具有参考意义。     

Slow driving control of tracked vehicles with automated mechanical transmission based on fuzzy logic

In order to move tracked vehicles at an extremely slow speed with automated mechanical transmission (AMT), slow driving function was added in the original system. The principle and requirement of slow driving function were analyzed. Based on analysis of slow driving characteristic, identification of slow driving condition and fuzzy control algorithm, a control strategy of the clutch was designed. In order to realize slow driving, the clutch was controlled in a slipping mode as manual driving. The vehicle speed was increased to a required speed and kept in a small range by engaging or disengaging the clutch to the approximate half engagement point. Based on the control strategy, a control software was designed and tested on a tracked vehicle with AMT. The test results show that the control of the clutch with the slow driving function was smoother than that with original system and the vehicle speed was slower and steadier.     

车辆驾驶性评估方法在蠕行工况中的应用

为解决评估体系构建方法不科学,评估指标权重计算及主客观权重分配不合理的问题,提出构建驾驶性评估体系和计算指标权重的新方法.考虑车辆驾驶性在蠕行工况的内涵,基于SMART原则对驾驶性评估体系进行设计,从纵向响应特性、纵向平顺特性、纵向稳态特性3个维度构建了蠕行工况驾驶性评估体系;提出用网络层次分析和组合熵分别计算主客观权重,并设计结合网络层次和组合熵的综合权重优化模型,用于计算评估体系中各评估指标权重,使其兼顾主观性和客观性.应用该方法进行多辆车的蠕行工况驾驶性评估,对比分析不同权重计算方法的评估结果.结果表明:本文提出的方法得到的评估结果较层次分析、熵权等方法更准确,该评估体系构建方法更具科学性,权重计算方法更有效.     

Combined Control Allocation and Sliding Mode Control in the Dynamic Control of a Vehicle with Eight In-Wheel Motors

An eight wheel independently driving steering (8WIDBS) electric vehicle is studied in this paper. The vehicle is equipped with eight in-wheel motors and a steer-by-wire system. A hierarchically coordinated vehicle dynamic control (HCVDC) system, including a high-level vehicle motion controller, a control allocation, an inverse tire model and a lower-level slip/slip angle controller, is proposed for the over-actuated vehicle system. The high-level sliding mode vehicle motion controller is designed to produce desired total forces and yaw moment, distributed to longitudinal and lateral forces of each tire by an advanced control allocation method. And the slip controller is designed to use a sliding mode control method to follow the desired slip ratios by manipulating the corresponding in-wheel motor torques. Evaluation of the overall system is accomplished by sine maneuver simulation. Simulation results confirm that the proposed control system can coordinate among the redundant and constrained actuators to achieve the vehicle dynamic control task and improve the vehicle stability.     

车辆路面附着条件自辨识模型研究

车辆的主动安全性问题和操纵稳定性问题受到广泛的关注,自主辨识道路附着条件可有效提高车辆在行驶过程中的安全稳定性.实时识别道路的状况,有利于车辆稳定性控制系统、防滑控制系统充分利用极限驱动力或制动力,保证车辆行驶安全性.以冰雪及湿滑等路面的附着条件辨识为重点研究内容,通过采集相关参数,利用模糊控制器对车辆动、稳态操控输入条件下车辆的响应辨识路面附着条件.     

State Estimation of the Electric Drive Articulated Dump Truck Based on UKF

The requirements of vehicle dynamic stability control are higher than ever as the significant increase of electric drive articulated vehicle speed. According to the construction features of articulated dumping truck and nonlinear characteristics of moving vehicles, nonlinear observer of vehicle status is designed to strength robustness of dynamic control system in this paper. A 4-degree-of-freedom nonlinear dynamic model of articulated electric drive vehicle is built as reference model to estimate the state of the articulated vehicle. And by adopting Unscented Kalman Filter (UKF) algorithm, a series of state parameters such as longitudinal velocities of front and rear frames, yaw rate and side-slip angle are estimated. During the test of 60 t articulated electric drive vehicle, 2 inertial navigation modules are installed in the front frame and rear frame respectively and the speed of each electric drive wheel is obtained simultaneously. As the test results suggest, in various working conditions, the algorithm based on UKF is able to accurately estimate the state parameters of articulated vehicle with the estimated error less than 5%. The proposed method is justified to be the theoretical basis and application guidance for articulated vehicle stability control.     

高超声速飞行器有限时间LPV滑模控制器设计

高超声速飞行器在机动飞行时易受到外界扰动,若采用传统的状态反馈控制方法,闭环控制系统极易引起振荡,无法满足机动飞行指令信号跟踪的精度要求;若采用传统的滑模控制方法,由于系统存在奇异值的问题,且计算过程较为复杂,控制系统不易于实现.针对上述问题,考虑高速机动飞行控制实际要求,提出了一种基于有限时间时变滑模的线性变参数(LPV)控制器设计方法并应用于高超声速飞行控制.首先不考虑外界扰动,通过传统的状态反馈控制方法使系统保持稳定.然后,在扰动存在的情况下,通过选取一个特殊的滑动函数,设计有限时间时变滑模控制律.为减小系统的抖振现象,引入饱和函数来替换控制律中的符号函数.经理论推导证明了闭环系统中的所有信号都是有界的,并且可以在预定的时间内将跟踪误差控制在零点的一个很小的邻域范围内.仿真验证结果表明,高超声速飞行器机动飞行条件下的状态量可在有限时间内稳定跟踪参考指令信号,且有效地抑制了闭环系统的振荡现象,验证了本方法所设计控制器的有效性.     

基于线性矩阵不等式的智能车轨迹跟踪控制

针对传统的基于精确数学模型的智能车轨迹跟踪控制器跟踪精度低,鲁棒性弱,很难适应复杂多变的驾驶环境等问题,结合线性矩阵不等式(LMI)鲁棒控制具有易于求解、抗干扰能力强等优点,提出基于LMI的智能车轨迹跟踪控制方法. 将车辆侧向动力学状态空间模型进行坐标变换,得到基于跟踪误差的车辆侧向动力学状态空间模型,采用饱和线性轮胎得到车辆侧向动力学多胞型模型;设计LMI反馈控制器,在控制器中引入前馈控制量,以消除侧向位置稳态误差. Carsim和Matlab/Simulink的联合仿真表明,该控制器在保证车辆稳定性的基础上具有较高的跟踪精度,对车速和路面附着系数具有较强的鲁棒性. 与模型预测控制器(MPC)和预瞄驾驶员模型(PDM)控制器进行对比,结果表明,设计的该控制器轨迹跟踪精度更优.     

Control allocation algorithm for over-actuated electric vehicles

A control allocation algorithm based on pseudo-inverse method was proposed for the over-actuated system of four in-wheel motors independently driving and four-wheel steering-by-wire electric vehicles in order to improve the vehicle stability. The control algorithm was developed using a two-degree-of-freedom(DOF) vehicle model. A pseudo control vector was calculated by a sliding mode controller to minimize the difference between the desired and actual vehicle motions. A pseudo-inverse controller then allocated the control inputs which included driving torques and steering angles of the four wheels according to the pseudo control vector. If one or more actuators were saturated or in a failure state, the control inputs are re-allocated by the algorithm. The algorithm was evaluated in Matlab/Simulink by using an 8-DOF nonlinear vehicle model. Simulations of sinusoidal input maneuver and double lane change maneuver were executed and the results were compared with those for a sliding mode control. The simulation results show that the vehicle controlled by the control allocation algorithm has better stability and trajectory-tracking performance than the vehicle controlled by the sliding mode control. The vehicle controlled by the control allocation algorithm still has good handling and stability when one or more actuators are saturated or in a failure situation.     

Anomaly detection method based on kinematics model and nonholonomic constraint of vehicle

A method used to detect anomaly and estimate the state of vehicle in driving was proposed.The kinematics model of the vehicle was constructed and nonholonomic constraint conditions were added,which refer to that once the vehicle encounters the faults that could not be controlled,the constraint conditions are violated.Estimation equations of the velocity errors of the vehicle were given out to estimate the velocity errors of side and forward.So the stability of the whole vehicle could be judged by the veloci...     

基于熵矩阵的多目标非线性0-1规划近似算法

为了简化多目标二元匹配问题的求解,将该问题建模为多目标非线性0-1规划模型,该模型将变量约束转移到目标函数中,从而降低了问题求解难度.针对该模型,设计了基于熵矩阵计算的贪心近似算法,该算法通过熵矩阵的熵值计算确定多目标二元匹配度,并根据熵值的大小预先优化匹配顺序,从而使近似解更快速地接近最优解.仿真实验结果证明,对于单目标非线性0-1规划问题,本算法优于已有的近似算法,对于多目标非线性0-1规划问题,本算法在计算时间以问题规模的指数级减少的情况下,近似解能够很好地逼近最优解.因此,本算法与其它近似算法相比,在不增加时间复杂度的前提下,结果更优,近似度更高.     

基于2级控制系统的量化状态观测器设计

研究2级控制系统中量化状态观测器的设计方法.对设计的控制系统的输入、输出信号进行量化,使得这些数据可以在网络环境中传输.研究对量化后的数据经过非完美的频道传输时造成的数据损失的处理方法.设计状态观测器,对控制系统的各个信息进行估计.设计控制器,能够根据状态观测器输出的估计值,控制2个致动器协调工作,成功地实现了2级控制.在实验平台dSPACE-DS1103上对所设计的2级控制系统进行验证,实验数据验证了本文提出的2级控制器设计方法的有效性和可行性.     

基于视觉的自动寻迹智能车系统设计与实现

以全国大学生智能汽车比赛为背景,设计一种在专门赛道上自动行驶的智能车控制系统,能够在遵守大赛的一系列规则下,最短时间内跑完全程并不脱离跑道。本系统以飞思卡尔公司的16位单片机MC9S12XS128为控制核心,提出一种基于模糊预瞄跟踪控制算法。以模拟人类驾车的方式,根据不同赛道情况,通过模糊预瞄器在线调节预瞄距离,并通过判断预瞄点位置来调节速度和转向角度。该方法使智能车系统达到要求的稳定性及快速性。实际运行结果表明,该系统具有较好的稳定性和鲁棒性.     

基于多级鲁棒PID控制的汽车稳定性控制策略

利用AMESim软件搭建了整车和液压制动系统模型。将基于H∞控制理论的PID控制算法应用于汽车稳定性控制的研究。根据车辆行驶状态的变化调整鲁棒PID控制器的参数,构建以横摆角速度和质心侧偏角为控制目标的汽车稳定性控制算法。进而利用PID控制算法得出制动轮缸压力,实现了整车的稳定性控制。利用Matlab/Simulink和AMESim建立联合仿真平台,对控制算法进行验证。结果表明,该控制算法具有很好的实时性和控制效果,能够满足车辆稳定性控制的要求。