资源描述
Artificial Intelligence Applications for 4WD Electric Vehicle Control SystemKeywords4WD; PI; Adaptive Fuzzy PI; Fuzzy Controller; Direct Torque Control.SOCABSTRACTA novel speed control design of 4WD electric vehicle (EV) to improve the comportment and stability under different road constraints condition is presented in this paper. The control circuit using intelligent adaptive fuzzy PI controller is proposed. Parameters which guide the functioning of PI controller are dynamically adjusted with the assistance of fuzzy control. The 4WD is powered by four motors of 15 kilowatts each one, delivering a 384 N.m total torque. Its high torque (338 N.m) is instantly available to ensure responsive acceleration performance in built-up areas. The electric drive canister of tow directing wheels and tow rear propulsion wheels equipped with tow induction motors thanks to their light weight simplicity and their height performance. Acceleration and steering are ensure by electronic differential, the latter control separately deriving wheels to turn at any curve. Electric vehicle are submitted different constraint of road using direct torque control. Electric vehicle are simulated in Matlab Simulink. The simulation results have proved that the intelligent fuzzy PI control method decreases the transient oscillations and assure efficiency comportment in all topologies road constraints, straight, curved road, descent.1. IntroductionThe principal constraints in vehicle design for transportation are the development of a non-polluting high safety and comfortable vehicle. Taking into account these constraints, our interest has been focused on the 4WD electrical vehicle, with independent driving wheel-motor at the front and with classical motors on the rear drive shaft 1-4. This configuration is a conceivable solution, the pollution of this vehicle is strongly decreased and electric traction gives the possibility to achieve accurate and quick control of the distribution torque. Torque control can be ensured by the inverter, so this vehicle does not require a mechanical differential gear or gearbox. One of the main issues in the design of this vehicle (without mechanical differential) is to assume the car stability. During normal driving condition, all drive wheel system requires a symmetrical distribution of torque in the both sides. In recent years, due to problems like the energy crisis and environmental pollution, the Electric Vehicle (EV) has been researched and developed more and more extensively 1,2. Currently, most EVs are driven by two front wheels or two rear wheels. Considering some efficiency and space restrictions on the vehicle, people have paid more and more attention in recent years to fourwheel drive vehicles employing the IM in-wheel motor.Research has shown that EV control methods such as, PI control are able to perform optimally over the full range of operation conditions and disturbances and it is very effective with constant vehicle torque, Moreover these non-linear vehicle torque are not fixed and change randomly. However EV with conventional PI control may not have satisfactory performance in such fast varying conditions, the system performance deteriorates. In addition to this, it is difficult to select suitable control parameters Kp and Ki in order to achieve satisfactory compensation results while maintaining the stability of EV traction, due to the highly complex, non-linear nature of controlled systems. These are two of the major drawbacks of the PI control. In order to overcome these difficulties, adaptive PI controller by fuzzy control has been applied both in stationary and under roads constraints, and is shown to improve the overall performance of 4WD electric vehicle.The aim of this paper is to understand the impact of intelligent fuzzy speed controller using lithium-ion battery controlled by DC-DC converter, each wheels is controlled independently by via direct torque control based space vector modulation under several topologies. Modelling and simulation are approved out using the Matlab/Simulink tool to study the performance of 4WD proposed system.2. Electric Vehicle DescriptionAccording to Figure 1 the opposition forces acting to the vehicle motion are: the rolling resistance force due to the friction of the vehicle tires on the road; the aerodynamic drag force caused by the friction on the body moving through the air; and the climbing force that depends on the road slope.The total resistive force is equal to and is the sum of the resistance forces, as in (1).(1)The rolling resistance force is defined by:(2)The aerodynamic resistance torque is defined as follows:(3)The rolling resistance force is usually modeled as:(4)where r is the tire radius, m is the vehicle total mass, is the rolling resistance force constant, g the gravity acceleration, is Air density, is the aerodynamic drag coefficient, is the frontal surface area of the vehicle, v is the vehicle speed, is the road slope angle. Values for these parameters are shown in Table 1.3. Direct Torque Control StrategyThe basic DTC strategy is developed in 1986 by Takahashi. It is based on the determination of instantaneous space vectors in each sampling period regarding desired flux and torque references. The block diagram of the original DTC strategy is shown in Figure 2. The reference speed is compared to the measured one. The obtained error is applied to the speed regulator PI whose output provides the reference torque. The estimated stator flux and torque are compared to the corresponding references. The errors are applied to the stator flux and torque hysteresis regulators, respectively. The estimation value of flux and its phase angle is calculated in expression(5)(6)And the torque is controlled by three-level Hysteresis. Its estimation value is calculated in expression (7).(7)Figure 1. The forces acting on a vehicle moving along a slope.Figure 2. PI gains online tuning by fuzzy logic controller.Table 1. Parameters of the electric vehicle model.4. Intelligent Fuzzy PI ControllerFuzzy controllers have been widely applied to industrialprocess. Especially, fuzzy controllers are effective techniques when either the mathematical model of the system is nonlinear or no the mathematical model exists. In this paper, the fuzzy control system adjusts the parameter of the PI control by the fuzzy rule. Dependent on the state of the system. The adaptive PI realized is no more a linear regulator according to this principle. In most of these studies, the Fuzzy controller used to drive the PI is defined by the authors from a series of experiments 5-8.The expression of the PI is given in the Equation (2).(8)where: Output of the control; : Input of the control. The error of the reference current and the injected speed ; Parameter of the scale; Parameter of the integrator.The discrete equation:(9)where: Output on the time of k th sampling; : Error on the time of k sampling; Cycle of the samplingOn-line Tuning:The on-line tuning equation for kp and ki are shown below:(10)(11)The frame of the fuzzy adaptive PI controller is illustrated in Figure 2.The linguistic variables are defines as NL, NM, NS, Z, PS, PM, PB meaning negative large, negative medium, negative small, zero, positive small, positive medium, positive big.The Membership function is illustrated in the Figures 3-6.The view plot surface of fuzzy controller for kp and ki are shown in Figures 7 and 8 respectively.Table 2 shows the fuzzy tuning rules.Table 2. Fuzzy tuning rules.Figure 3. The membership function of input e(k).Figure 4. The membership function of input e(k).kpFigure 5. The membership function of output kp.kiFigure 6. The membership function of output ki.Figure 7. View plot surface of fuzzy controller for kp.Figure 8. View plot surface of fuzzy controller for ki.5. Simulation ResultsIn order to analyze the driving wheel system behavior, Simulations were carried using the model of Figure 9. The following results were simulated in MATLAB and its divided in two phases. The first one deal with the test of the EV performances controlled with DTC strategy under several topology variation in the other hand we show the impact of this controller on vehicle power electronics performances. Only the right motor simulations are shown. The assumption that the initialized lithiumion battery SOC is equal to 70% during trajectories.5.1. Intelligent Fuzzy PI Controller for Direct Torque Control SchemeThe topology studied in this present work consists of three phases: the first one is the beginning phases with speed of 80 Km/h in straight road topology, the second phase present the curved road with the same speed, finally the 4WD moving up the descent road of 10% under 80 Km/h, the specified road topology is shown in Figure 10, when the speed road constraints are described in the Table 3.Refereed to Figure 11at time of 2 s the vehicle driver turns the steering wheel on a curved road at the right side with speed of 80 Km/h, the assumption is that the four motors are not disturbed. In this case the front and rear driving wheels follow different paths, and they turn in the same direction but with different speeds. The electronic differential acts on the four motor speeds by decreasing the speed of the driving wheel on the right side situated inside the curve, and on the other hand by increasing the wheel motor speed in the external side of the curve. The behaviors of these speeds are given in Figure 11. At t = 3 s the vehicle situated in the second curve but in the left side, the electronic differential compute the novel steering wheels speeds references in order to stabilize the vehicle inside the curve. The battery initial SOC of 70 % is respected. In this case the driving wheels follow the same path with no overshoot and without error which can be justified with the good electronic differential act coupled with DTC performances.Figures 12-15 show the variation of kp and ki of the four intelligent speed controller.Figure 16 describes the variation of current for the front motor right in different phases. In the first step and to reach 80 Km/h, the EV demands a current of 48.75 A for each motors which explained with electromagnetic torque of 138.20 N.m. In the curved road the current and electromagnetic torque demand are computed using the electronic differential process according to the driver decision by means that the speed reference of each wheels is given by the electronic differential computations witch convert the braking angle in the curve on linear speeds. Figure 17 shows the electromagnetic torque of the front motor right. The third phase explains the effect of the descent slopped road the electromagnetic torque decrease and the current demand undergo half of the current braking phases. The presence of descent causes a great decrease in the phase current of each motor by means that the sloped force became an motor force. They develops approximately 96.17 N.m each one. The linear speeds of the four induction motors stay the same and the descent sloped road does not influence the torque control of eachFigure 9. The driving wheels control system.Figure 10. The chosen road topology of tests.Figure 11. Variation of vehicle speeds in different phases.Figure 12. Variation gain kp of intelligent fuzzy PI for the front right and left speed controller.Figure 13. Variation gain ki of intelligent fuzzy PI for the front right and left speed controller.Figure 14. Variation gain kp of intelligent fuzzy PI for the rear right and left speed controller.Figure 15. Variation gain ki of intelligent fuzzy PI for the rear left and left speed controller.Figure 16. Variation of phase current of the front motor right in different phases.Figure 17. Variation of electromagnetic torque of the front motor right in different phases.Table 3. The driving road topology description.wheels. The results are listed in Table 4.According to the formulas (1), (2), (3) and (4) and Table 4, the vehicle resistive torque was 127.60 N.m in the first case (beginning phase) when the power propulsion system resistive one is 127.60 N.m in the curved road. The driving wheels develop more and more efforts to satisfy the traction chain demand which justify a resistive torque equal to 127.60 N.m in the third descent slopped phase. The result prove that the traction chain under descent demand develop less effort comparing with the breaking phase cases by means that the vehicle needs the half of its energy in the descent sloped phases compared with the sloped ones as it specified in Table 5 and Figure 17.5.2. Power ElectronicsThe Lithium-ion battery must be able to supply sufficient power to the EV in accelerating and decelerating phase, which means that the peak power of the batteries supply must be greater than or at least equal to the peak power of the both electric motors. The battery must store sufficient energy to maintain their SOC at a reasonable level during driving, the Figure 18, describe the changes in the battery storage power in different speed references.Table 4. Values of phase current driving force of the right motor in different phases.Table 5. Variation of vehicle torque in different phases.Figure 18. Variation of Lithium-ion battery power in different phases.It is interesting to describe the power distribution in the electrical traction under several speed references as it described in Table 6. The battery provides about 20.73 Kw in the first phase in order to reach the electronic differential reference speed of 80 Km/h. In the second phase (phase 2: curved phases) the demanded power battery stay the same which present amount of 66.87% of the globally nominal power battery (31 Kw). In third phase the battery produced power equal to 13.73 Kw under descent slopped road state. The battery produced power depend only on the electronic differential consign by means the courved and descente sloped road driver state which can be explained by the battery SOC of Figure 19.Figure 19 explains how SOC in the Lithium-ion battery changes during the driving cycle; it seems that the SOC decreases rapidly at acceleration, by means that the SOC ranges between 68.44% to 70% during all cycles phases from beginning at the end cycles.At t = 4 s, the battery SOC becomes lower than 68.44% (it was initialized to 70% at the beginning of the simulation).Table 7 reflects the variation of SOC in different simulations phases. The relationship between SOC and left time in three phases are defined by the flowing linear fitting formula:(12)Table 6. Variation of battery power in different trajectory phases.Table 7. Evaluation of SOC % in the different phases.Figure 19. Battery efficiency versus state-of-charge.Moreover the simulation results specified by Figure 20, we can define the relationship between the sate of charge and the traveled distance in each cases as it shown in Table 8 and the relationship between power consumed and state of charge during each phase as it shown in Table 9, the first one (beginning phase) is defined by the linear fitting formula:(13)This power is controlled by the Buck Boost DC-DC converter current and distribute accurately for three phases. Figure 21 shows the buck boost DC-DC converter robustness under several speed cycles. The buck boost converter is not only a robust converter which ensures the power voltage transmission but also a good battery recharger in deceleration state that help to perfect the vehicle autonomous with no voltage ripple.6. ConclusionThe research outlined in this paper has demonstrated theFigure 20. Evaluation traveled distance en function the SOC.Figure 21. Buck boost DC-DC converter behavior under several speed variations.Table 8. Evaluation of distance traveled and SOC.Table 9. The relationship between the traction chain power electronics characteristics and the distance traveled in differ-ent phases.feasibility of improved vehicle stability for 4WD electric vehicle using DTC controls. DTC with intelligent fuzzy speed controller is able to adapt itself the suitable control parameters which are the proportional and integral gains and ki to the variations of vehicle torque. This method was improved proposed traction system steering and stability during different trajectory this. The advantage DTC controller is robustness and performance, there capacity to maintain ideal trajectories for four wheels control independently and ensure good disturbances rejections with no overshoot and stability of vehicle perfected ensured with the speed variation and less error speed. The 4WD electric vehicle was proved best comportment and stability during different road path by maintaining the motorization error speed equal zeros and gives a good distribution for electromagnetic torque. The electric vehicle was proved efficiency comportment under different road topologies.四驱越野车控制系统人工智能应用关键词：四驱，PI，自适应模糊 PI（比例积分）,模糊控制器，直接转矩控制。摘要本文提出了在不同道路约束条件下，一种新型四驱电动车的速度控制设计如何提高其准确性和稳定性。提出了使用智能自适应模糊 PI 控制器的控制电路。引导 PI 控制器功能的参数在模糊控制器帮助下的动态调整。4WD 由每台 15KW 的 4 台电机提供动力，总扭矩为384Nm。其高扭矩（384Nm）的立即可用确保在组合区域相应加速性能。由于质轻简单和高性能，牵引转向轮和后牵引运动轮的电力驱动罐装备了牵引感应发动机。通过电子差速器确保加速和转向，后者控制单独导出以任何角度转动的车轮。电动车辆使用直接转矩控制提供不同的道路约束。电动汽车在 Matlab Simulink 中模拟。模拟结果证明，智能模糊 PI 控制方法减少了瞬时震荡，并确保了所有拓扑中道路约束，直线，弯曲，下降的效率。1.介绍运输车辆设计的主要制约因素是无污染的高安全性和舒适性车辆的开发。考虑到这些限制，我们的兴趣集中在 4WD 电动车辆，前置独立的驱动轮电机和后置驱动轴上的传统电机上。这种配置是可以想到的解决方案，车辆的污染被大大减少，电牵引给出了实现对分配扭矩的准确和快速控制的可能性。变频器可以确保扭矩控制，因此该车辆不需要机械差速齿轮或变速箱。这种车辆（无机械差速）设计的主要问题是保证汽车的稳定性。在正常行驶状态下，所有的驱动轮系统在两侧都需要扭矩对称分布。近年来，由于能源危机和环境污染等问题，电动汽车（EV）的研究和开发越来越广泛。目前，大多数电动汽车是前轮驱动或后轮驱动。考虑到车辆的效率和空间限制，近年来人们越来越重视 IM 轮内动力四轮驱动车辆的使用。研究表明，诸如 PI 控制的 EV 控制方法能够在整个操作条件和干扰范围内进行最佳动作，并且在车辆转矩恒定方面非常有效，而且这些非线性车辆转矩不是固定的，而是随机变化的。然而具有常规 PI 控制的 EV 可能在这种快速变化，系统性能恶化的条件下不具有令人满意的效果。除此之外，由于控制系统非常复杂，非线性的特性，想要达到满意的补偿结果，同时维持 EV 牵引器的稳定性，所以选择稳定控制参数 Kp 和 Ki 很困难。这就是 PI控制的两个主要缺点。为了克服这些缺点，通过模糊控制的自适应 PI 控制器已经在静止和道路限制下应用，并且显示出 4WD 电动车整体性能的提高。本文的目的是了解在使用 DC-DC 转换器控制锂离子电池对智能模糊速度控制器的影响，在多个拓扑下，每个车轮基于空间向量调制通过直接转矩独立控制控制。使用Matlab/Simulink 工具批准的建模和仿真，以研究 4WD 提出的系统性能。2.电动汽车概述根据图 1，作用于车辆运动的反作用力的是：由于汽车轮胎与地面的摩擦所产生的滚动阻力 Ftire，车身移动过程中与空气摩擦所产生的空气动力学阻力 Faero 和由于道路坡度产生的攀爬力 Fslope。总阻力等于 Fr，并且是各阻力的总和，如（1）所示。（1）滚动阻力由下式定义：（2 ）气动阻力矩定义如下：（3）滚动阻力通常建模为：（4 ）其中 r 是轮胎半径，m 是车辆总质量，f r 是滚动阻力常数，g 是重力加速度， air 是空气密度，c d 是气动阻力系数， Af 是车辆的正面面积，v 是车速， 是道路坡度角。这些参数的值如表 1 所示。3.直接转矩控制策略基本直接转矩控制策略是 Takahashi 在 1986 年开发的。它基于每个采样周期中关于期望的通量和转矩参考的瞬时空间矢量的确定。原始 DTC 策略的框图如图 2 所示。将参考速度与测量速度进行对比。所获得的误差被应用到输出提供参考扭矩的速度调节器 PI。将估计的定子磁通和转矩与相应的参考值相比较。误差分别应用于定子磁链和转矩磁质调节器。在表达式中计算磁链和相位角的估计值，且转矩由三电平滞环控制。其估计值由表达式（7）计算。（5）（6）（7）图 1：沿斜坡行驶的车辆上的作用力图 2：PI 控制器通过模糊逻辑控制器的在线调整表 1：电动汽车模型的参数4.智能模糊 PI 控制器模糊控制器已经广泛地应用到了工业过程中了。特别地，当系统的数学模型是非线性的或者没有数学模型存在时，模糊控制器是有效的技术手段。在本文中，模糊控制系统通过模糊规则来调整 PI 控制参数。由于提高系统的稳定性这一原则，自适应 PI 不再是一个线性调节器。在大量研究中，驱动 PI 的模糊控制器是作者从一系列的试验中确定下来的。PI 的表达式由等式 2 给出。（8 ）其中：y（t）：控制输出；e（t）：控制输入；参考电流 w（t）和注入速度 w（t ）的误差；K p：刻度参数；T i：积分器参数。离散方程：（9）其中：y（k）：第 k 次采样时间的输出；e（k ）：k 的采样时间误差；T：采样周期。在线调整：kp 和 ki 的在线调整方程如下所示：（10）（11）模糊自适应 PI 控制器的框架如图 2 所示。语言变量的定义是NL, NM, NS, Z, PS, PM, PB，其意义为负大，负中，负小，零，正小，正中，正大。参数功能如图 3-6 所示。kp 和 ki 模糊控制器的视图曲面分别如图 7 和图 8 所示。图 2 表明了模糊调整规则。表 2:模糊调整规则图 3：输入 e（k）的函数图像图 4：输入 e（k）的函数图像图 5：输入 kp 的函数图像图 6：输入 ki 的函数图像图 7：k p 模糊控制器的曲面视图图 8：k i 模糊控制器的曲面视图5.仿真结果为了分析驱动轮系统的运动，采用图 9 的模型进行仿真。在 MATLAB 中模拟了以下结果，且分为两个阶段。第一个处理在几个拓扑变化下用 DTC 策略控制的 EV 性能测试，另一方面我们显示了该控制器对车辆电力电子性能的影响。只显示正确的电机模拟。假设在轨迹中初始化的锂离子电池 SOC 等于 70%。5.1.用于直接转矩控制的智能模糊 PI 控制器当前工作研究的拓扑结构由三个阶段组成:第一阶段是直路拓扑中四驱车以速度为 80 Km/h 行驶的起始阶段，第二阶段是相同速度的上下坡阶段，最后是速度低于 80 Km/h，且坡度为 1:10 的下坡阶段。道路中速度约束由表 3 所示，指定的道路拓扑如图 10 所示。根据图 11，假设四个电机不受干扰，在第二秒的时候，汽车驾驶员以 80 Km/h 的速度在弯道右侧转动方向盘。在这种情况下，前后驱动轮行驶路径不同，虽然朝着相同的方向，但转速不同。电子差速器通过降低位于曲线内侧的右侧驱动轮速度以及提高曲线外侧的右侧驱动轮速度来控制四个电机的速度。它们的速度变化如图 11 所示。在 t=3s 时，位于第二曲线左侧的车辆，电子差速器计算新的方向盘速度参考值，以使车辆在曲线内部趋于稳定。电池初始 SOC 为 70%。在这种情况下，通过与 DTC 性能相结合的良好电子差动证明，驱动轮路径相同没有过冲和错误。图 12-15 表明了四个智能速度控制器参数 kp 和 ki 的变化。图 16 描述了不同阶段前置电机电流的变化。在第一步车速达到 80 Km/h，电动机要求每个电机的电流为 48.75A，电磁转矩为 138.20Nm。在弯道中，根据驾驶员决定，使用电子差速过程来计算电流和电磁转矩需求，这意味着每个车轮的速度基准由电子差动计算开关给出，以线性速度转换曲线中的制动角度。图 17 显示了前置电机的电磁转矩。第三阶段说明了斜坡下坡的影响，电磁转矩
展开阅读全文