Development of a traction controller for drivability improvement of electric vehicles

本博士项目旨在开发用于改善全电动卡车驾驶性能的抗抽搐和牵引控制器。目标车辆将用于城市和城市外的交付，并将面临各种气候和道路条件：从干燥平坦的柏油路面到下雪的上坡道路。控制器将在两个不同的演示器上进行测试。第一种是两轮驱动车辆，配备中央电动机，单齿轮传动，通过半轴和等速接头与车轮连接的开放式差速器。第二种是四轮驱动汽车，在前轴和后轴上都使用了以前的动力传动系统配置。该抗突控制器将基于非线性显式模型预测控制技术，并将与文献中五种不同的控制结构进行比较：i)纯前馈控制器；Ii)基于电机转速高频分量的反馈控制器；Iii)前馈控制器与反馈扰动观测器相结合的系统；Iv)基于传动系扭率的反馈控制器；v)基于估计的传动系统转矩的反馈控制器。在牵引控制器的开发中，将采用基于PID滑动的算法设定基准，并将其与一种新的非线性显式模型预测牵引控制器进行比较。萨里大学汽车工程中心在模型预测控制领域有着丰富的经验，因为它是为数不多的开发了显式非线性模型预测控制器设计工具箱的研究小组之一。该工具箱已被用于带轮内电机的电动汽车牵引控制器的初步开发，这在文献中代表了一种新奇的东西。然而，这种控制器仍然是一个概念的证明，因为它对于轮胎-道路参数的变化不是鲁棒的。该博士项目旨在提高控制器的鲁棒性和技术成熟度，以实现工业上可实现的系统。工业合作伙伴提供的创新控制器及其对电动汽车演示的综合评估相结合，将使结果能够在机械和汽车工程主题领域的四分之一期刊上发表。此外，该活动将包括对抗抽搐控制器的广泛调查，因为文献缺少对该主题的全面回顾。

This Ph.D. project aims at developing anti-jerk and traction controllers for the drivability improvement of fully electric lorries. The target vehicles will be employed for urban and extra-urban deliveries, and will face a wide range of climate and road conditions: from dry and flat tarmac to snowy uphill roads. The controllers will be tested on two different demonstrators. The first one is a two-wheel-drive vehicle with a central electric motor, a single gear transmission, an open differential connected to the wheels through half-shafts and constant velocity joints. The second one is a four-wheel-drive vehicle that uses the previous drivetrain configuration on both the front and rear axles.The anti-jerk controller will be based on nonlinear explicit model predictive control technology, and will be compared with five different control structures from the literature: i) a pure feedforward controller; ii) a feedback controller based on the high frequency component of the motor speed; iii) a system combining a feedforward controller with a feedback disturbance observer; iv) a feedback controller based on the drivetrain torsion rate; and v) a feedback controller based on the estimated drivetrain torque.Within the traction controller development, the benchmark will be set by a PID slip based algorithm, which will be compared with a novel nonlinear explicit model predictive traction controller. The Centre for Automotive Engineering of the University of Surrey has experience in the field of model predictive control, as it is one of the few research groups to have developed a toolbox for the design of explicit nonlinear model predictive controllers. This toolbox has been exploited for the preliminary development of a traction controller for an electric vehicle with in-wheels motors, which represents a novelty in the literature. Nevertheless, such controller is still a proof of concept, as it is not robust with respect to tyre-road parameter variations. This Ph.D. project targets the enhancement of the level of robustness and technology readiness of the controller, to achieve an industrially implementable system.The combination of innovative controllers and their comprehensive assessment on the electric vehicle demonstrators provided by the industrial partners will allow the publication of the results in quartile 1 journals in the subject areas of mechanical and automotive engineering. Furthermore, the activity will include an extensive survey on anti-jerk controllers, as the literature misses a thorough review of this topic.