A 10,000-word article on automobile control in the intelligent era

In terms of the braking system, the mature control units currently include the anti-braking system (ABS) and the traction control system (TCS), which play an important role in the stability and safety of longitudinal driving. With the development of electronic technology, more efficient and energy-saving wire control technology has emerged. The wire control brake system replaces the traditional air pressure or brake actuators with electric drive elements. Each wheel has a separate controller to ensure the optimal distribution of braking force, realizing the organic combination of automotive electronic technology and network communication technology, and better improving the automation level of the vehicle[23]. The wire control brake system can be divided into an electro-hydraulic wire control brake system (Electronic hydraulic brake, EHB) and an electronic mechanical wire control brake system (Electronic mechanical brake, EMB) according to the different actuators[24].

The development of automatic control technology for power, transmission and braking systems provides the underlying execution foundation for longitudinal assisted driving/automatic driving. Longitudinal assisted/automatic driving mainly controls driving and braking based on the distance between the front and rear vehicles, the set expected speed and the vehicle's own state. Adaptive cruising control (ACC) integrates the anti-collision system on the basis of cruise control. Through the coordinated control of the power, transmission and braking systems, it realizes a more intelligent and safe longitudinal speed control strategy, freeing the driver from the operation of the accelerator pedal and brake pedal under specific working conditions. Reference [25] designed a PID-based adaptive cruise controller by the pole configuration method to ensure that the vehicle distance error and relative speed converge to zero. In order to comprehensively coordinate the tracking performance, following safety and fuel economy when following a vehicle, reference [26] studied the multi-objective ACC algorithm based on model predictive control and proposed an online fast calculation solution for the optimization problem. According to the comfort requirements of different drivers, references [27-28] studied the ACC control algorithm considering driving style. In addition, in recent years, the ACC system has been expanded to stop and start control for low-speed driving and frequent start and stop conditions in urban traffic. go, SG), collision avoidance control (Collision warning/avoidance, CW/CA), etc.

1.2 Review of the Development of Automobile Lateral Motion Control

The phenomenon of lateral movement of one or both axle wheels due to braking, rotational inertia, etc. when the car is driving is called skidding. The skidding of the car poses a great threat to safe driving and often causes serious traffic accidents such as tail-spinning and rollover [29-30]. The electronic stability program (ESP) is a further expansion of the ABS and TCS functions. It adds sensors related to the lateral movement of the vehicle and enhances the stability of the vehicle by controlling the driving force and braking force of the wheels [31-32]. In addition, active steering control technology has also attracted the interest of domestic and foreign research institutions in recent years. Active steering corrects the steering angle applied by the driver and automatically adjusts the steering gear ratio of the vehicle according to the different working conditions of the vehicle, thereby improving the steering ease of the vehicle at low speed and the lateral stability at high speed [33-34]. With the continuous development of technology, active steering can take human factors into consideration in the design of control algorithms, and carry out humanized design according to the driving habits of drivers with different styles, changing "people adapt to the car" to "car adapts to people", thereby better assisting the driver in controlling the driving of the vehicle [35].

In order to improve steering accuracy and response speed, steering-by-wire (SBW) systems have been widely studied. To improve the safety of the system, the SBW system may be equipped with multiple drive systems at the same time to ensure that the vehicle can continue to drive and steer in the event of a system failure [36-37]. The SBW system can change the transmission ratio in real time according to different driving conditions. Compared with the traditional mechanical steering system, the response speed and accuracy of the SBW system are greatly improved, which can further reduce the driving burden of the driver [38-39]. The SBW system is a highly electronic and intelligent product, and is also a key technology for realizing fully autonomous driving of vehicles [40].

The above-mentioned lateral motion control system provides the feasibility of autonomous lateral control of vehicles for autonomous driving technology. In the L2 stage of vehicle automation process, lateral motion control is still an auxiliary driving system for specific driving tasks and specific scenarios, such as lane keeping, automatic lane change, automatic parking control system, etc. Among them, the lane keeping assist system focuses on the safety of the vehicle. It corrects the driving posture of the vehicle through the active steering system to avoid unconscious lane deviation during driving, and does not affect the driver's normal driving while making corrections [41-42]. The automatic parking technology is to meet the convenience of car use and realize the function of helping or even replacing the driver to park. Automatic parking technology mainly adopts two research directions: path planning-based and experience-based planning [43]. In addition, the automatic lane change system makes active lane change decisions based on the driver's overall expected speed and surrounding vehicle information to help the driver change lanes.

1.3 Automobile side - Review of the development of longitudinal coupling motion control

With the increasing number of vehicle motion control subsystems, the overall performance of the vehicle cannot be optimized due to the interactive coupling and mutual constraints of the subsystem control on the vehicle motion. Therefore, how to avoid conflicts between control systems, give full play to their respective advantages, and achieve the optimal overall performance of the vehicle has become a key issue that needs to be urgently solved in the field of vehicle chassis control today [44−45].

In vehicle motion control systems, vehicle status (tire force, longitudinal velocity, lateral velocity, yaw rate, center of mass sideslip angle, etc.) and road surface information (such as road slope and adhesion coefficient) are very critical information [46]. Real-time online estimation of vehicle status and road parameters [47] is an important issue in vehicle chassis coordinated control. In terms of lateral-longitudinal motion coordinated control, references [48−49] studied the coordinated control of front-wheel steering, rear-wheel steering and direct yaw-moment control (DYC) from the perspective of optimizing tire force. Reference [50] proposed a coordinated control strategy of ARS (Active rear-wheel steering) and DYC based on feedforward and feedback control methods. Reference [51] studied the coordinated control of active steering and DYC switching under different working conditions. Reference [52] discussed the coordination of lateral motion control and continuous damping control (CDC).

At present, urban low-speed areas and highways are the two scenarios where L3 autonomous driving technology is first applied, that is, human-machine co-driving is realized in a simple traffic environment with clear structural features such as highway signs and lane lines. At the same time, the vehicle motion control system is a typical multi-input and multi-output, strongly coupled nonlinear system. In the lateral-longitudinal coupling condition, how to design a high-quality motion control strategy has become the focus and difficulty of realizing autonomous driving technology [53-55]. At present, vehicles that can achieve fully autonomous lateral and longitudinal driving are still under research, which mainly targets special vehicles and is used in specific environments, such as mines and campuses.

2 Problems in the process of automobile automation

Looking back at the development of automobile control, we can see that current automobile driving automation can basically achieve assisted driving or automatic driving under specific working conditions, including adaptive cruise control, lane keeping assist, low-speed steering following, automatic remote parking, etc. However, in the face of more complex working conditions, such as frequent changes in working conditions or emergencies, the current automatic driving control system cannot respond reliably. Recently, a series of accidents that occurred in the process of automatic driving applications have also sounded the alarm for us. The current technology development maturity is not enough to support the demand for fully automatic driving of automobiles under complex mixed traffic conditions (including motor vehicles, non-motor vehicles, pedestrians, construction obstacles, etc.). Nature's review article [56] pointed out that current automatic driving still has deficiencies in reliability and safety. Machines cannot achieve the flexibility, adaptability and creativity of humans and are still unable to cope with various emergencies. The following summarizes the level of perception, decision-making and collaborative control technology in the current process of automobile automation and analyzes some typical problems faced.

2.1 The accuracy and speed of perception and information fusion need to be improved

At present, the environmental perception technology based on a single sensor has developed relatively maturely and is widely used in assisted driving technologies with relatively low levels of automation. In the adaptive cruise control technology ACC, millimeter-wave radar can realize multi-lane target tracking and detection, and combined with curve position compensation technology, it can realize target vehicle identification and tracking [59]. In addition, the use of CMOS monocular camera can detect and identify the target vehicle and road environment and obtain the longitudinal information of the vehicle more accurately, which is also a perception technology solution in ACC control technology [60]. In terms of blind spot detection and assisted parking, perception technology based on lidar or vision can effectively detect the position of blind spots and obstacles and lock the target parking area [61−63].