Control method, device, vehicle, equipment and product of autonomous vehicle

Abstract

The application discloses a control method and device of an automatic driving vehicle, a vehicle, equipment and products, and belongs to the technical field of automatic driving. The control method of the automatic driving vehicle comprises the following steps: obtaining a target operation amount of a cloud cabin driver on a pedal; determining a target longitudinal motion amount corresponding to the target operation amount based on a preset mapping relationship between the operation amount and the motion amount of a controlled vehicle; loading longitudinal characteristic parameters of a current control vehicle and vehicle end control interface parameters; converting the target longitudinal motion amount into an initial control amount based on the longitudinal characteristic parameters of the current control vehicle and the target longitudinal motion amount; converting the initial control amount into vehicle end control amount corresponding to the current control vehicle based on the vehicle end control interface parameters of the current control vehicle; and sending the vehicle end control amount to the current control vehicle to control the operation of the current control vehicle. The control method of the automatic driving vehicle reduces the risk of misoperation when switching vehicle models.

Description

Technical Field

[0001] This application belongs to the field of autonomous driving technology, and in particular relates to a control method, device, vehicle, equipment and product for an autonomous vehicle. Background Technology

[0002] Parallel driving cloud cabins are used to provide remote monitoring, remote takeover, remote driving, and remote obstacle avoidance support for L4 autonomous vehicles under complex conditions. In actual operation, the same cloud cabin often needs to connect to and control multiple vehicle models. However, there are significant differences in the power response of vehicles from different manufacturers and models, resulting in inconsistent operating feel for cloud cabin drivers. The speed output obtained by cloud cabin drivers for the same pedal opening is completely different on different models, making it difficult to establish a stable operating intuition. The training and adaptation costs are high, as drivers need to familiarize themselves with the acceleration characteristics of each model and build muscle memory, and need to readjust when switching models. The risk of misoperation is high when switching models. After switching models, drivers are prone to situations where the vehicle suddenly surges when it should be slowing down or sluggishly accelerates when it should be accelerating, especially in scenarios such as low-speed obstacle avoidance, which require extremely precise control, and can easily cause the vehicle to overshoot, collide, or fail to escape. Summary of the Invention

[0003] This invention aims to at least solve one of the technical problems existing in the prior art. To this end, this invention proposes a control method, device, vehicle, equipment, and product for autonomous vehicles, which eliminates the impact of differences in pedal and acceleration response between different vehicle models on the driver's operating feel, reduces the risk of misoperation when switching vehicle models, and especially avoids the risk of vehicle overshooting or failure to escape from difficulties in fine control scenarios such as low-speed traction.

[0004] In a first aspect, this application provides a control method for an autonomous vehicle, including: Obtain the target amount of pedal operation by the cloud cabin driver; Based on the preset mapping relationship between the operation amount and the motion amount of the controlled vehicle, the target longitudinal motion amount corresponding to the target operation amount is determined. The mapping relationship is the same for different controlled vehicles. Load the longitudinal characteristic parameters and vehicle-side control interface parameters of the currently controlled vehicle; Based on the longitudinal characteristic parameters of the currently controlled vehicle and the target longitudinal motion, the target longitudinal motion is converted into an initial control quantity; Based on the vehicle-side control interface parameters of the currently controlled vehicle, the initial control quantity is converted into the vehicle-side control quantity corresponding to the currently controlled vehicle. Send the vehicle-side control quantity to the currently controlled vehicle to control the operation of the currently controlled vehicle.

[0005] According to the control method for autonomous vehicles provided in this application, the unified mapping mechanism with the same mapping relationship for different controlled vehicles eliminates the impact of differences in pedal and acceleration response between different vehicle models on the driver's operating feel. This allows the cloud-based driver to become familiar with only one set of unified pedal operation mappings without having to rebuild operating intuition and muscle memory for each vehicle model. On this basis, by adaptively converting the vehicle-side control quantity according to the longitudinal characteristic parameters of the vehicle model, it ensures that the unified goal can be accurately executed on different vehicle models, so that the same pedal opening corresponds to the same speed output, reducing the risk of misoperation when switching vehicle models. Especially in fine control scenarios such as low-speed traction, it avoids the risk of vehicle overshoot or traction failure.

[0006] An embodiment of the control method for an autonomous vehicle according to this application includes converting the target longitudinal motion into an initial control quantity based on the longitudinal characteristic parameters of the currently controlled vehicle and the target longitudinal motion quantity, comprising: Obtain the actual longitudinal motion state of the currently controlled vehicle; Calculate the deviation between the target longitudinal motion and the actual longitudinal motion state; Based on the deviation and the longitudinal characteristic parameters, the initial control quantity is calculated using a feedforward method.

[0007] An embodiment of the autonomous vehicle control method of this application includes a vehicle-side control interface parameter comprising the dimension type and value range of the vehicle-side control interface of the currently controlled vehicle, wherein the step of converting the initial control quantity into a vehicle-side control quantity corresponding to the currently controlled vehicle based on the vehicle-side control interface parameter of the currently controlled vehicle comprises: Based on the unit type, the initial control quantity is converted into a control command with at least one of the following units: target acceleration, target vehicle speed, percentage of drive or brake request, and torque request. Based on the range of values, the value of the control command is limited.

[0008] An embodiment of the present application provides a method for controlling an autonomous vehicle, wherein controlling the operation of the currently controlled vehicle includes: Obtain the actual longitudinal motion state of the currently controlled vehicle; Based on the deviation between the actual longitudinal motion state and the target longitudinal motion amount, the vehicle-end control quantity is corrected, so as to control the operation of the currently controlled vehicle based on the corrected vehicle-end control quantity.

[0009] One embodiment of the control method for an autonomous vehicle in this application includes obtaining the target input amount of the driver's pedal input in the cloud cabin, comprising: The target operation amount of the cloud cabin driver on at least one of the accelerator pedal and the brake pedal is obtained, and the target operation amount includes at least one of pedal travel, pedal opening and pedal force.

[0010] One embodiment of the control method for an autonomous vehicle in this application further includes: The same upper limit of acceleration and upper limit of jerk are applied to the change process of the target longitudinal motion corresponding to different vehicle models.

[0011] An embodiment of the control method for an autonomous vehicle according to this application further includes, before sending the vehicle-side control quantity to the currently controlling vehicle: A unified speed limit is determined based on the current control scenario; The longitudinal motion of the target is limited based on the unified speed limit.

[0012] An embodiment of the control method for an autonomous vehicle according to this application further includes, before sending the vehicle-side control quantity to the currently controlling vehicle: The vehicle-side control quantity is limited based on the current maximum acceleration and maximum deceleration of the controlled vehicle.

[0013] An embodiment of the control method for an autonomous vehicle according to this application includes an accelerator pedal and a brake pedal. Before sending the vehicle-side control quantity to the currently controlling vehicle, the method further includes: In response to both the accelerator pedal and the brake pedal being operated, the operation amount of the brake pedal is responded to first.

[0014] One embodiment of the control method for an autonomous vehicle in this application further includes: The vehicle speed feedback status, communication link delay time, and vehicle model parameter loading status of the currently controlled vehicle are obtained. In response to at least one of the following abnormal conditions, the currently controlled vehicle is controlled to enter a degraded mode or execute a safe parking strategy: The vehicle speed feedback status is abnormal; The delay time of the communication link exceeds a preset threshold; The vehicle model parameters are missing.

[0015] One embodiment of the control method for an autonomous vehicle in this application further includes: In response to a switching event of the controlled vehicle, the mapping relationship remains unchanged, and the longitudinal characteristic parameters and vehicle-side control interface parameters of the vehicle after the switching are reloaded.

[0016] Secondly, this application provides a control device for an autonomous vehicle, comprising: The first processing module is used to obtain the target amount of pedal operation by the cloud cabin driver. The second processing module is used to determine the target longitudinal motion amount corresponding to the target operation amount based on the preset mapping relationship between the operation amount and the motion amount of the controlled vehicle. The mapping relationship is the same for different controlled vehicles. The third processing module is used to load the longitudinal characteristic parameters and vehicle-side control interface parameters of the currently controlled vehicle. The fourth processing module is used to convert the target longitudinal motion into an initial control quantity based on the longitudinal characteristic parameters of the currently controlled vehicle and the target longitudinal motion. The fifth processing module is used to convert the initial control quantity into the vehicle-side control quantity corresponding to the current controlled vehicle based on the vehicle-side control interface parameters of the currently controlled vehicle. The sixth processing module is used to send the vehicle-side control quantity to the currently controlled vehicle in order to control the operation of the currently controlled vehicle.

[0017] The control device for autonomous vehicles provided in this application eliminates the impact of differences in pedal and acceleration response between different vehicle models on the driver's operating feel through a unified mapping mechanism that makes the mapping relationship the same for different controlled vehicles. This allows the driver in the cloud cabin to become familiar with a unified set of pedal operation mappings without having to rebuild operating intuition and muscle memory for each vehicle model. On this basis, by adaptively converting the vehicle-side control quantity according to the longitudinal characteristic parameters of the vehicle model, it ensures that the unified goal can be accurately executed on different vehicle models, so that the same pedal opening corresponds to the same speed output, reducing the risk of misoperation when switching vehicle models. Especially in fine control scenarios such as low-speed traction, it avoids the risk of vehicle overshoot or traction failure.

[0018] Thirdly, this application provides an autonomous vehicle that operates based on the autonomous vehicle control method described in the first aspect.

[0019] Fourthly, this application provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the control method for an autonomous vehicle as described in the first aspect above.

[0020] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the control method for an autonomous vehicle as described in the first aspect above.

[0021] The above-described one or more technical solutions in the embodiments of this application have at least one of the following technical effects: By using a unified mapping mechanism that ensures the same mapping relationship for different controlled vehicles, the impact of differences in pedal and acceleration response between different vehicle models on the driver's operating feel is eliminated. This allows cloud-based drivers to become familiar with only one set of unified pedal operation mappings, without having to rebuild their operating intuition and muscle memory for each vehicle model. Furthermore, by adaptively converting vehicle-side control quantities based on the longitudinal characteristic parameters of the vehicle model, it ensures that the unified objective can be accurately executed on different vehicle models, making the same pedal opening correspond to consistent speed output. This reduces the risk of misoperation when switching vehicle models, especially in fine control scenarios such as low-speed traction, avoiding the risk of vehicle overshoot or failure to escape.

[0022] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0023] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a schematic diagram of the system architecture of the control method for an autonomous vehicle provided in the embodiments of this application; Figure 2 This is one of the flowcharts illustrating the control method for an autonomous vehicle provided in this application embodiment; Figure 3 This is a second schematic flowchart of the control method for an autonomous vehicle provided in the embodiments of this application; Figure 4 This is a schematic diagram of the structure of the control device for an autonomous vehicle provided in an embodiment of this application; Figure 5 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0024] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0025] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0026] The following description, in conjunction with the accompanying drawings, details the control method, control device, electronic equipment, and readable storage medium for autonomous vehicles provided in this application, through specific embodiments and application scenarios.

[0027] The control method for autonomous vehicles can be applied to the terminal, and can be executed by the hardware or software in the terminal.

[0028] The autonomous vehicle control method provided in this application can be executed by a control system deployed on the cloud-based control module, or by a functional module or entity within that control system capable of implementing the control method. This control system can be an independent computing device or a control unit integrated into the remote driving cloud module. The following description uses the cloud-based control system as an example to illustrate the autonomous vehicle control method provided in this application.

[0029] like Figure 1 As shown, the cloud-based remote driving terminal acts as the control center, establishing connections with vehicles A, B, C, D, and more via the vehicle-to-cloud communication network. The driver operates the pedals on the cloud-based remote driving terminal, and the control system issues unified control commands to the currently connected target vehicles for execution. The cloud-based remote driving terminal can switch between controlling different vehicle models at different times, enabling a single cloud-based terminal to support remote driving of multiple vehicle models.

[0030] The control system includes the following modules: The unified pedal input acquisition module is used to collect the input quantities of the driver's input to the unified pedals (accelerator pedal and / or brake pedal) in the cloud-based cockpit. These input quantities may include pedal travel, pedal opening, pedal force, or their normalized values ​​(e.g., normalized to 0 to 1, or 0% to 100%). The cloud-based cockpit provides the driver with the same set of physical pedals and the same set of pedal semantics, independent of the target vehicle model.

[0031] The vehicle longitudinal characteristic parameter library module stores longitudinal dynamic characteristic parameters and control interface information for different vehicle models. Each vehicle model corresponds to a set of calibration parameters, including but not limited to: the response characteristics of the pedal or control quantity with acceleration or vehicle speed (e.g., the relationship curve between pedal opening or control request and steady-state acceleration and steady-state vehicle speed); maximum acceleration, maximum deceleration, and maximum permissible vehicle speed; actuator response delay and acceleration / braking settling time; parameters affecting longitudinal response such as vehicle mass and powertrain type; and the type and dimensions of the vehicle-side control interface, such as the received target acceleration, target vehicle speed, percentage of drive / braking request, or torque request, as well as the corresponding value range and unit. Calibration parameters can be obtained through offline calibration, real-vehicle testing, parameters provided by the manufacturer, or online identification.

[0032] The unified mapping module establishes a standardized mapping relationship between pedal opening and target longitudinal motion across all vehicle models. This mapping relationship is consistent for all connected vehicle models and is independent of the target model. In some embodiments, pedal opening is mapped to a standardized target vehicle speed. For example, in a remote escape scenario, the maximum standardized vehicle speed is set to 10 km / h, and a standardized mapping function from accelerator pedal opening to target vehicle speed is defined: Target vehicle speed = Maximum standardized vehicle speed × Pedal opening. Pedal opening can also be mapped to a standardized target acceleration, or simultaneously to a standardized target vehicle speed and a standardized target acceleration limit. Brake pedal opening is mapped to a standardized target deceleration or a standardized target deceleration rate. Thus, the driver's pedal input in the cloud-based control system is transformed into a standardized target longitudinal motion (standardized target vehicle speed and / or standardized target acceleration / deceleration) independent of the vehicle model.

[0033] The vehicle model adaptive conversion module is used to convert the unified target longitudinal motion quantity output by the unified mapping module into a vehicle-side control quantity executable by the vehicle model based on the longitudinal characteristic parameters of the currently controlled vehicle model. In some embodiments, this module calculates the initial control quantity using a feedforward method based on the response characteristics of the vehicle model in the vehicle model longitudinal characteristic parameter library: according to the unified target vehicle speed or target acceleration, combined with the inverse model of the vehicle model's pedal or control quantity-acceleration (or vehicle speed) response characteristics, it calculates the vehicle-side control quantity required for the vehicle model to achieve the unified target, and converts the control quantity into the dimensions and value range corresponding to the vehicle-side control interface of the vehicle model. For example, for a vehicle model receiving target acceleration, it is directly converted into a target acceleration request; for a vehicle model receiving a percentage of drive / brake request, it is converted into the corresponding percentage based on its response curve; for a vehicle model receiving target vehicle speed, it is converted into a target vehicle speed request.

[0034] The closed-loop feedback correction module is used to perform closed-loop correction of the control quantity based on the actual longitudinal motion state of the vehicle, so that the actual output of the vehicle converges to a unified target and eliminates the influence of vehicle model differences, calibration errors, and external disturbances. In some embodiments, this module receives the actual vehicle speed and / or actual acceleration fed back from the vehicle end, calculates the deviation between the actual value and the unified target value, and performs closed-loop control (e.g., proportional-integral-derivative control or other feedback control) based on the deviation to correct the control quantity output by the vehicle model adaptive conversion module, so that the actual steady-state vehicle speed of different vehicle models at the same pedal opening converges consistently to the unified target speed. By combining feedforward and feedback, even if the calibration parameters of a certain vehicle model deviate or the dynamic characteristics change, the system can still keep its actual speed output consistent with the unified target.

[0035] A unified dynamic response shaping module is used to shape the change process of longitudinal motion of a unified target, so that the dynamic feel of different models during acceleration and deceleration is consistent and smooth. In some embodiments, this module applies a unified upper limit of acceleration and a jerk upper limit to the change of the unified target speed or target acceleration, and smoothly transitions according to a unified response curve. Even if the driver presses the pedal to the maximum position instantly, the unified target speed will not undergo a step change, but will rise smoothly according to the unified acceleration and jerk limits. This makes the acceleration and deceleration process obtained from the same pedal action of different models feel consistent, avoiding the phenomenon of some models overshooting and others sluggish.

[0036] The safety constraint module is used to apply safety restrictions during the unified mapping and control quantity conversion process. In some embodiments, the safety constraints include: setting a unified speed limit based on the current scenario, such as limiting the maximum target vehicle speed in a low-speed escape scenario; tailoring the unified target based on the maximum acceleration and deceleration capabilities of the controlled vehicle model to avoid exceeding the vehicle's physical capabilities; prioritizing the brake pedal over the accelerator pedal, allowing emergency braking commands to be directly transmitted to the vehicle; and implementing a downgrade or safe parking strategy when there is abnormal vehicle speed feedback, excessive communication delay, or missing vehicle model parameters.

[0037] like Figure 2 As shown, the control method for the autonomous vehicle includes steps S1, S2, S3, S4, S5, and S6.

[0038] Step S1: Obtain the target input amount of the cloud cabin driver to the pedals; In this step, the cloud cabin serves as a remote driving control seat, i.e., a remote driver control console or a remote assistance processing terminal.

[0039] The pedals are physical input devices used by the cloud cabin driver for longitudinal control, including the accelerator pedal and the brake pedal.

[0040] The target input quantity is the operative input applied by the driver to the pedals in the cloud-based control system. It can include pedal travel (the amount of displacement of the pedal when it is pressed down), pedal opening (the ratio of pedal travel to maximum travel, usually expressed as a percentage or a value between 0 and 1), pedal force (the amount of force applied by the driver to the pedal), or their normalized values. The input quantity reflects the driver's desired longitudinal control strength.

[0041] Step S2: Based on the preset mapping relationship between the operation amount and the motion amount of the controlled vehicle, determine the target longitudinal motion amount corresponding to the target operation amount. The mapping relationship is the same for different controlled vehicles. In this step, the mapping relationship between the operation amount and the motion amount of the controlled vehicle is a corresponding rule that converts the pedal operation amount into the vehicle motion target. This mapping relationship is preset and consistent across different vehicle models; that is, regardless of the vehicle model currently being controlled, the same pedal opening corresponds to the same target longitudinal motion amount.

[0042] The target longitudinal motion is the target value of the vehicle's longitudinal motion obtained after mapping, corresponding to the pedal operation amount. Longitudinal motion refers to the movement of the vehicle along the direction of travel (forward or backward). The target longitudinal motion can include the target vehicle speed (the speed value that the vehicle is expected to reach) and / or the target acceleration or deceleration (the acceleration or deceleration value that the vehicle is expected to reach).

[0043] Step S3: Load the longitudinal characteristic parameters and vehicle-side control interface parameters of the currently controlled vehicle; In this step, the vehicle currently under control is the autonomous vehicle that the cloud-based control system is currently remotely operating. In the cloud-based remote driving scenario, the cloud-based control system can control different vehicles at different times.

[0044] Longitudinal characteristic parameters are a set of parameters reflecting the inherent properties of a vehicle's longitudinal dynamics. These parameters differ between different vehicle models. Longitudinal characteristic parameters may include at least one of the following: the response characteristics of the pedal or control input to acceleration or vehicle speed (i.e., the change in the vehicle's actual acceleration or speed after applying a certain control input); maximum acceleration (the maximum positive acceleration the vehicle can achieve); maximum deceleration (the maximum braking deceleration the vehicle can achieve); actuator response delay (the delay from the issuance of the control input to the vehicle's actual response); 0-100 km / h acceleration time (the time required for the vehicle to accelerate from a standstill to 100 km / h, reflecting the vehicle's power performance); vehicle model coefficient (a normalized coefficient determined based on the vehicle's power performance, used for differentiated power allocation for different vehicle models); and dimensions of the vehicle-side control interface (the type and unit of the control signals received by the vehicle).

[0045] The vehicle-side control interface refers to the communication interface through which a vehicle receives control commands from the cloud. Different vehicle models may have different physical quantity definitions (dimensions) and effective numerical ranges (value ranges) for their vehicle-side control interfaces.

[0046] Dimensions refer to the type of physical quantity of the control commands received by the vehicle-side control interface, such as acceleration (m / s²), vehicle speed (m / s or km / h), percentage of drive or braking request (%), or torque request (N·m).

[0047] The value range refers to the valid numerical range of the control command under this dimension, for example, a drive request percentage of 0% to 100%, or a torque request of -500 N·m to 500 N·m, etc.

[0048] Because different car models have different power responses (some accelerate to 100 km / h in 5 seconds, some in 10 seconds), and different models receive different control command formats (some receive acceleration commands, some receive percentage commands), the system needs to obtain these two types of parameters for the current car model before it can perform subsequent adaptation.

[0049] Step S4: Based on the current longitudinal characteristic parameters of the controlled vehicle and the target longitudinal motion, convert the target longitudinal motion into the initial control quantity; In this step, the power characteristics of the current vehicle model (such as acceleration capability and response curve) can be used to calculate how much control intensity needs to be applied to the vehicle model in order to achieve a unified target value (target longitudinal motion), i.e., the initial control quantity.

[0050] Step S5: Based on the vehicle-side control interface parameters of the currently controlled vehicle, convert the initial control quantity into the vehicle-side control quantity corresponding to the currently controlled vehicle. In this step, the initial control quantity is a preliminary value of the control quantity, which may include the power request value required to make the vehicle model achieve a uniform target (e.g., a power request percentage of 26%).

[0051] The vehicle-side control quantity is the final control command obtained by converting the initial control quantity through the vehicle-side control interface parameters, which can be directly recognized and executed by the vehicle-side control interface of the currently controlled vehicle.

[0052] In actual execution, after determining the longitudinal motion of the unified target, the system needs to convert it into specific control commands that the currently controlled vehicle can execute.

[0053] Step S6: Send vehicle-side control signals to the currently controlled vehicle to control the operation of the currently controlled vehicle.

[0054] In this application, the driver's pedal operation in the cloud cabin is decoupled from the original longitudinal dynamic characteristics of each vehicle model, so that the pedal input is completely independent of the vehicle's dynamic characteristics.

[0055] On the cloud-based control module, a unified pedal operation semantics independent of vehicle model can be established. The unified pedal input acquisition module can acquire the driver's pedal input. Then, the unified mapping module, based on a preset mapping relationship between the input and the controlled vehicle's motion, determines the target longitudinal motion corresponding to that input. This mapping relationship is consistent across all vehicle models; that is, different types of controlled vehicles have the same mapping relationship (for example, in a scrambling scenario with a maximum unified speed of 10 km / h, there is a unified linear mapping relationship between the accelerator pedal opening and the target speed; any vehicle model with an accelerator pedal opening of one-third corresponds to a unified target speed of approximately 3.3 km / h, regardless of the vehicle model's native pedal-acceleration response curve). Then, the vehicle model adaptive conversion module determines the executable vehicle-side control quantity based on the longitudinal characteristic parameters of the currently controlled vehicle and this unified target longitudinal motion quantity. Finally, the vehicle is controlled based on this vehicle-side control quantity. Through this method, the cloud-based control module ensures consistent speed output and acceleration / deceleration feel regardless of the vehicle model being controlled, even with the same pedal travel.

[0056] This application first obtains the target pedal input from the cloud-based driver, then determines the target longitudinal motion corresponding to the target input based on a consistent mapping relationship across different vehicle models. This ensures that different vehicle models correspond to the same target value at the same pedal opening, eliminating the impact of pedal-acceleration response differences on driver feel. Drivers only need to familiarize themselves with a unified pedal input mapping to establish stable operating intuition, eliminating the need to relearn acceleration characteristics for each vehicle model, thus reducing training and vehicle adaptation costs. Furthermore, based on the longitudinal characteristic parameters of the currently controlled vehicle, the target longitudinal motion is converted into the vehicle-side control quantity that the vehicle can execute. This ensures the unified target can be accurately executed across different vehicle models. When the cloud-based driver switches controlled vehicles, the mapping relationship remains unchanged, and the system automatically converts the input based on the new vehicle's longitudinal characteristic parameters, maintaining a consistent pedal input feel for the driver. This avoids sudden acceleration during slow driving or sluggish acceleration during acceleration, significantly improving safety in delicate control scenarios such as low-speed escape from difficult situations. Moreover, when adding a new vehicle model, only the longitudinal characteristic parameters and control interface conversion strategy for that model need to be added; the unified mapping relationship can be reused without modification, demonstrating good scalability.

[0057] According to the control method for autonomous vehicles provided in this application, the unified mapping mechanism with the same mapping relationship for different controlled vehicles eliminates the impact of differences in pedal and acceleration response between different vehicle models on the driver's operating feel. This allows the cloud-based driver to become familiar with only one set of unified pedal operation mappings without having to rebuild operating intuition and muscle memory for each vehicle model. On this basis, by adaptively converting the vehicle-side control quantity according to the longitudinal characteristic parameters of the vehicle model, it ensures that the unified goal can be accurately executed on different vehicle models, so that the same pedal opening corresponds to the same speed output, reducing the risk of misoperation when switching vehicle models. Especially in fine control scenarios such as low-speed traction, it avoids the risk of vehicle overshoot or traction failure.

[0058] In some embodiments, based on the longitudinal characteristic parameters of the currently controlled vehicle and the target longitudinal motion, the target longitudinal motion is converted into an initial control quantity, including: Obtain the actual longitudinal motion state of the currently controlled vehicle; Calculate the deviation between the target longitudinal motion and the actual longitudinal motion state; The initial control quantity is calculated using a feedforward method based on the deviation and longitudinal characteristic parameters.

[0059] In this embodiment, the actual longitudinal motion state refers to the longitudinal motion physical quantity fed back by the currently controlled vehicle during actual operation, which may include the actual vehicle speed (the vehicle's current real speed) and / or the actual acceleration (the vehicle's current real acceleration). The actual longitudinal motion state is acquired through vehicle-mounted sensors or a positioning system and reported to the cloud in real time via vehicle-to-cloud communication.

[0060] Deviation refers to the difference between the target's longitudinal motion and the actual longitudinal motion state, reflecting the gap between the current actual motion state of the vehicle and the target.

[0061] The feedforward method is a control quantity calculation method that does not rely on feedback signals. Based on the mathematical model of the system or known characteristics, it directly calculates the required control quantity according to the input target.

[0062] The feedforward method can utilize the longitudinal characteristic parameters of the controlled vehicle model (i.e., the response characteristics of the vehicle model's pedal or control quantity with acceleration or speed) to directly calculate the vehicle-end control quantity required to achieve the unified target based on the longitudinal motion quantity of the unified target.

[0063] In actual operation, after the driver presses the pedal, the system determines the target longitudinal motion (e.g., target speed 3.3 km / h) based on a unified mapping relationship across all vehicle models. Based on this, the system obtains the actual longitudinal motion state of the currently controlled vehicle (e.g., actual speed 2 km / h). Then, it calculates the deviation between the target longitudinal motion and the actual longitudinal motion state (i.e., speed difference of 1.3 km / h). Next, based on this deviation and longitudinal characteristic parameters, the initial control input is calculated using a feedforward approach.

[0064] For example, the system calculates the basic power request based on the speed difference between the target vehicle speed and the actual vehicle speed, combined with the control coefficient. Then, it corrects this basic power request based on the longitudinal characteristic parameters of the currently controlled vehicle (such as the vehicle model coefficient determined by the vehicle's 0-100 km / h acceleration time) to obtain the initial control quantity. For instance, if the control coefficient K is 20 and the speed difference is 1.3 km / h, the basic power request is: speed difference × K = 1.3 × 20 = 26%. If the vehicle model coefficient of the currently controlled vehicle is 1 (such as the base model B), the initial control quantity is 26% × 1 = 26%; if it is the faster-accelerating A vehicle (model model coefficient 0.5), the initial control quantity is 26% × 0.5 = 13%; and if it is the slower-accelerating C vehicle (model model coefficient 0.8), the initial control quantity is 26% × 0.8 = 20.8%.

[0065] According to the control method for autonomous vehicles provided in the embodiments of this application, the feedforward calculation is based on the current actual longitudinal motion state, thereby calculating a control quantity that is more in line with the current working conditions. By combining the deviation with the feedforward calculation of longitudinal characteristic parameters, less power is given to models with stronger acceleration capabilities, and more power is given to models with weaker acceleration capabilities, so that different models eventually converge to the same target speed under the same pedal opening, realizing the rapid response and accurate adaptation of feedforward control to the differences in physical characteristics of each model.

[0066] In some embodiments, the vehicle-side control interface parameters include the dimension type and value range of the vehicle-side control interface of the currently controlled vehicle. Based on the vehicle-side control interface parameters of the currently controlled vehicle, converting the initial control quantity into the vehicle-side control quantity corresponding to the currently controlled vehicle may include: Based on the type of dimension, the initial control quantity is converted into a control command with at least one of the following dimensions: target acceleration, target vehicle speed, percentage of drive or brake request, and torque request. The value of the control command is limited based on the range of values.

[0067] In this embodiment, the dimension type refers to the type of physical quantity in the vehicle-side control command. Specific dimension types may include: target acceleration (i.e., the acceleration value the vehicle is expected to achieve, in m / s²), target vehicle speed (i.e., the speed value the vehicle is expected to achieve, in km / h or m / s), drive or braking request percentage (i.e., the percentage control request applied to the vehicle's drive or braking system, dimensionless, typically expressed as 0%-100%), and torque request (i.e., the torque command applied to the vehicle's motor or engine, in N·m).

[0068] The value range is the valid numerical range under this dimension type, such as a drive request percentage of 0% to 100%, or a torque request of -500 N·m to 500 N·m, etc.

[0069] Control commands convert the initial control quantity (such as a 26% power request) into a command value that conforms to the current vehicle dimension type (such as a target acceleration value of 0.5 m / s² or a drive percentage of 26%).

[0070] After calculating the initial control quantity (such as a 26% power request) using a feedforward method based on deviation and longitudinal characteristic parameters, this initial control quantity needs to be converted into a format that the currently controlled vehicle can directly recognize and execute. The system processes this information according to the vehicle-side control interface parameters of the currently controlled vehicle.

[0071] First, the initial control quantity is converted into the corresponding control command according to the dimension type. Specifically, if the dimension type of the currently controlled vehicle is target acceleration, the system converts the initial control quantity into the corresponding target acceleration value (e.g., 0.5 m / s²) based on the response characteristics of that vehicle model; if the dimension type is a percentage of drive or braking request, the initial control quantity (26%) is directly used as the percentage of drive request; if the dimension type is target vehicle speed, it is converted into the target vehicle speed value (e.g., 3.3 km / h); if the dimension type is torque request, it is converted into the corresponding torque value (e.g., 50 N·m).

[0072] Then, the value of the control command is limited based on the value range. For example, if the current driving request percentage of the controlled vehicle is in the range of 0%-100%, then 26% is within the effective range and does not need to be adjusted; however, if the maximum allowed driving request of a certain vehicle is only 60%, while the initial control quantity is converted to 80%, then the system limits the value of the command to 60% to avoid sending values ​​to the vehicle that exceed the effective range of its control interface.

[0073] In this application, by explicitly listing multiple control interface dimension types, it is possible to adapt to the diverse control interfaces of different manufacturers and different vehicle models, thereby enhancing scalability and universality. When adding a new vehicle model, it is only necessary to supplement the control interface dimension type and longitudinal characteristic parameters of that vehicle model, without changing the unified mapping relationship for reuse.

[0074] By separating dimensional conversion and amplitude limiting into two sub-steps, the format adaptation of control commands is made more complete. The combination of these two steps ensures that the vehicle-side control quantities can be correctly received and executed by the currently controlled vehicle in both format and numerical dimensions, adapting to diverse control interfaces from different manufacturers and vehicle models. Compared to related technologies that calibrate fixed pedal mappings only for a single vehicle model, this application achieves adaptation of the same control method to vehicles with different interface types through the two steps of dimensional conversion and amplitude limiting.

[0075] In actual implementation, the initial control quantity can first be calculated using a feedforward approach based on the longitudinal characteristic parameters of the currently controlled vehicle. The system then calculates the required power demand value to achieve the target speed or acceleration based on a unified target vehicle speed or acceleration, combined with the vehicle model's response characteristics (i.e., the response curve between the pedal or control quantity and the acceleration or speed). This response characteristic's inverse model is used to calculate the power demand value required for the vehicle model to reach the unified target. For example, when the unified target speed is 3.3 km / h, for vehicle B, which accelerates from 0 to 100 km / h in 10 seconds, the system calculates an initial control quantity of 26% power demand. This initial control quantity is a purely power value and does not yet consider the format requirements of the current vehicle control interface.

[0076] Then, based on the dimensions and value range of the vehicle-side control interface currently controlling the vehicle, the initial control quantity is converted into the final vehicle-side control quantity. Different vehicle models have different communication protocols for their chassis drive-by-wire systems. Some receive target acceleration (in m / s²), some receive drive / brake request percentage (in %), some receive target vehicle speed (in km / h), and some receive torque requests (in N·m). Furthermore, even when receiving the same dimension, the effective value range varies between vehicle models. For example, some vehicles have a drive request percentage range of 0%-100%, while others have a range of 0-1 (normalized value), and some vehicles have a maximum acceleration limit of only 2 m / s². The system converts the initial control quantity (e.g., 26%) into the physical quantity and numerical range actually received by the vehicle's control interface based on the control interface information stored in the vehicle's longitudinal characteristic parameter library. If the vehicle receives a drive percentage with a value range of 0%-100%, then 26% is sent directly; if the value range is 0-1, then 0.26% is sent; if the vehicle receives a target acceleration, then 26% is converted into the corresponding acceleration value (e.g., 0.5m / s²) based on the vehicle's response characteristics before being sent.

[0077] According to the control method for autonomous vehicles provided in the embodiments of this application, by performing differentiated calculations of quantities based on the differences in power performance of different vehicle models, and then performing differentiated format adaptation based on the differences in control interfaces of different vehicle models, it is ensured that the unified target can be accurately calculated, correctly issued and executed on different vehicle models. When adding a new vehicle model, it is only necessary to supplement the response characteristic parameters and interface dimension information of the vehicle model in the longitudinal characteristic parameter library, and the unified mapping relationship can be reused without changing it.

[0078] In some embodiments, controlling the operation of the currently controlled vehicle may include: Obtain the actual longitudinal motion state of the currently controlled vehicle; Based on the deviation between the actual longitudinal motion state and the target longitudinal motion amount, the vehicle-side control quantity is corrected, and the current vehicle operation is controlled based on the corrected vehicle-side control quantity.

[0079] In this embodiment, the actual longitudinal motion state refers to the longitudinal motion physical quantity fed back by the currently controlled vehicle during actual operation, which may include the actual vehicle speed (the vehicle's current real speed) and / or the actual acceleration (the vehicle's current real acceleration). The actual longitudinal motion state is acquired through vehicle-mounted sensors or a positioning system and reported to the cloud in real time via vehicle-to-cloud communication.

[0080] In this application, while using a feedforward method to calculate the initial control quantity, a closed-loop feedback correction mechanism is also introduced. The closed-loop feedback correction module obtains the actual longitudinal motion state of the currently controlled vehicle (such as actual vehicle speed and / or actual acceleration), calculates the deviation between the actual value and the unified target value, and then performs closed-loop control (e.g., proportional-integral-derivative control or other feedback control) on the vehicle-side control quantity based on this deviation, correcting the control quantity output by the vehicle model adaptive conversion module, so that the actual steady-state vehicle speed of different vehicle models at the same pedal opening consistently converges to the unified target vehicle speed.

[0081] like Figure 3As shown, taking three car models with different acceleration capabilities as an example, car A accelerates the fastest (0-100 km / h in 5 seconds), car B the slowest (0-100 km / h in 10 seconds), and car C in the middle (0-100 km / h in 8 seconds). The system sets the model coefficients based on car B (0.5 for car A, 1 for car B, and 0.8 for car C). When the maximum unified speed is 10 km / h and the pedal opening is one-third, the unified target speed is approximately 3.3 km / h. If the current actual speeds of cars A, B, and C are all 2 km / h, then the speed difference is: 3.3 - 2 = 1.3 km / h. Assuming the control coefficient K is 20, the basic power request is: speed difference × K = 1.3 × 20 = 26%. Subsequently, based on the model coefficients, the actual output of each model is obtained: car A is 26% × 0.5 = 13%, car B is 26% × 1 = 26%, and car C is 26% × 0.8 = 20.8%. Although the three vehicles issued different power requests (less power to the vehicle with stronger acceleration and more power to the vehicle with weaker acceleration), their control target was the same: 3.3 km / h.

[0082] Building upon the feedforward conversion, closed-loop feedback correction further ensures that even if the calibration parameters of a particular vehicle model deviate or its dynamic characteristics change (such as external disturbances like vehicle aging or load variations), the system can still dynamically adjust the control input based on the deviation between the actual vehicle speed and the target value, ensuring that the actual output of different vehicle models converges consistently to a unified target. By combining feedforward and feedback, the driver in the cloud-based control system receives consistent speed output and acceleration / deceleration feel regardless of the vehicle model being controlled, even with the same pedal travel.

[0083] By combining feedforward and feedback, not only is the rapid response and accuracy of the initial control quantity guaranteed, but the influence of external disturbances such as vehicle model calibration error, vehicle aging, and load changes can also be eliminated. This allows the actual steady-state vehicle speed of different models to converge to a unified target speed under the same pedal opening, thus achieving stable and accurate unified speed control for multiple vehicle models.

[0084] In some embodiments, obtaining the target amount of pedal operation by the cloud cabin driver may include: The target input quantity of the cloud cabin driver to at least one of the accelerator pedal and brake pedal is obtained. The target input quantity includes at least one of pedal travel, pedal opening and pedal force.

[0085] In this embodiment, the accelerator pedal is the pedal used by the cloud cabin driver to control the acceleration of the vehicle. When the accelerator pedal is pressed, the vehicle receives positive driving force and the vehicle speed increases.

[0086] The brake pedal is used by the driver in the cloud cabin to control the vehicle's deceleration. When the brake pedal is pressed, the vehicle receives braking force and its speed decreases.

[0087] Pedal travel is the amount of displacement of the pedal when it is pressed down, usually expressed in millimeters or as a percentage of the maximum travel.

[0088] Pedal opening is the ratio of pedal travel to maximum travel, usually expressed as a percentage between 0% and 100% or a value normalized to 0 to 1. 0% means the pedal is not depressed, and 100% means the pedal is depressed to its maximum position.

[0089] Pedal force refers to the amount of force applied by the driver when pressing the pedal.

[0090] In actual operation, the unified pedal input acquisition module is used to collect the target pedal input quantities of the driver in the cloud cabin. The cloud cabin provides the driver with the same set of physical pedals (accelerator and brake pedals) and the same set of pedal semantics, independent of the target vehicle model. The driver in the cloud cabin expresses longitudinal control intentions by operating these pedals: when the driver presses the accelerator pedal, the unified pedal input acquisition module obtains the target input quantity of the accelerator pedal (including at least one of pedal travel, pedal opening, and pedal force, or its normalized value), and obtains the target acceleration or target vehicle speed after unified mapping; when the driver presses the brake pedal, the unified pedal input acquisition module obtains the target input quantity of the brake pedal, and obtains the target deceleration or target deceleration rate after unified mapping. The normalized value of the input quantity can be normalized to 0 to 1 or 0% to 100%.

[0091] The control method for autonomous vehicles provided in the embodiments of this application, by comprehensively covering the accelerator pedal and brake pedal as well as various types of operation quantities, enables the support of complete longitudinal control scenarios such as acceleration and deceleration, and the multiple expression methods of operation quantities (stroke, opening degree, force) enhance the universality under different cloud cabin hardware configurations.

[0092] In some embodiments, the method may further include: The same upper limit of acceleration and upper limit of jerk are applied to the change process of the target longitudinal motion corresponding to different vehicle models.

[0093] In this embodiment, different vehicle models refer to various vehicle models that the cloud-based remote driving system may need to access and control, such as Robotaxi (autonomous taxis) from different OEMs, low-speed unmanned vehicles and park vehicles with different drive-by-wire chassis, etc. These vehicle models have different longitudinal dynamic characteristics and response calibrations.

[0094] The process of change in the longitudinal motion of the target is the time history of the longitudinal motion of the target (such as the target vehicle speed) changing from the current value to the target value when the driver of the cloud cabin operates the pedal, for example, the process of changing from 0 km / h to 3.3 km / h.

[0095] The upper limit of acceleration refers to the maximum rate at which the longitudinal motion of a target increases or decreases per unit time, and is used to limit how fast a vehicle accelerates or decelerates.

[0096] The upper limit of acceleration refers to the maximum rate at which acceleration changes per unit time, that is, how fast acceleration changes.

[0097] The maximum acceleration and maximum jerk applied to different vehicle models are numerically consistent and are independent of the vehicle model.

[0098] In actual operation, when the driver operates the pedals in the cloud cabin, the system determines the target longitudinal motion based on a unified mapping relationship across all vehicle models. However, if the target value changes abruptly (for example, the driver suddenly presses the accelerator from 0 to the maximum), the target vehicle speed jumps from 0 km / h to 10 km / h instantly. The actuator response speed varies across different vehicle models, with some vehicles surging forward (high acceleration) and others slowing down (low acceleration), resulting in significant differences in perceived speed.

[0099] A uniform upper limit on acceleration and a uniform upper limit on jerk are applied to the change in the longitudinal motion of the target vehicle. Specifically, the target vehicle speed will not experience abrupt changes, but will rise smoothly according to uniform acceleration limits (e.g., the maximum acceleration does not exceed a certain value) and jerk limits (e.g., the rate of change of acceleration does not exceed a certain value). Since all vehicle models follow the same upper limits on acceleration and jerk, the acceleration and deceleration processes perceived by the driver tend to be consistent.

[0100] In this application, by applying a uniform upper limit for acceleration and a uniform upper limit for jerk, the dynamic feel of different models during acceleration and deceleration is kept consistent and smooth, avoiding the phenomenon of some models overshooting and others laging, thus improving the driver's operating comfort and control precision. Especially in fine control scenarios such as low-speed escape from difficulties, the smooth acceleration and deceleration process helps the driver to make precise control.

[0101] In some embodiments, the pedals include an accelerator pedal and a brake pedal, and may further include: [The text abruptly ends here, so the translation stops as well.] A uniform speed limit is determined based on the current control scenario, and the longitudinal motion of the target is limited based on the uniform speed limit; And / or, The vehicle-side control parameters are limited based on the current maximum acceleration and maximum deceleration of the controlled vehicle. And / or, When both the accelerator and brake pedals are operated, the brake pedal operation is responded to first.

[0102] In this embodiment, the current control scenario is the type of operating scenario in which the vehicle is currently located, such as a low-speed escape scenario (a scenario in which the vehicle needs to be slowly and carefully controlled in complex terrain), a normal driving scenario, a parking scenario, etc.

[0103] The uniform speed limit is the maximum permissible speed set according to the current control scenario. For example, in a low-speed obstacle-avoidance scenario, the maximum uniform speed is set to 10 km / h. The uniform speed limit applies to all vehicle models and is independent of the vehicle type.

[0104] Maximum acceleration is the maximum positive acceleration value that the currently controlled vehicle can achieve; it represents the vehicle's physical limits.

[0105] Maximum deceleration is the maximum braking deceleration value that the vehicle can currently control, and it represents the vehicle's physical limits.

[0106] The longitudinal motion of the target or the control quantity of the vehicle can be limited within the safety boundary to avoid exceeding the vehicle's physical capabilities or the safety requirements of the scenario.

[0107] When the accelerator pedal and brake pedal are operated simultaneously, the system responds to the amount of operation of the brake pedal, while ignoring or delaying the response to the amount of operation of the accelerator pedal.

[0108] In actual execution, the safety constraint module is used to impose safety restrictions during the unified mapping and control quantity conversion process. Before the control quantity is sent to the vehicle for execution, any one or more combinations of the following safety constraints can be applied: First, the system determines a uniform speed limit based on the current control scenario. For example, in a low-speed escape scenario, the maximum uniform speed is limited to 10km / h. Based on this uniform speed limit, the longitudinal movement of the target is restricted. Even if the driver presses the accelerator pedal to its maximum opening, the target speed will not exceed the uniform speed limit in this scenario.

[0109] Secondly, the system limits the vehicle-side control quantity based on the current maximum acceleration and maximum deceleration of the controlled vehicle. If the vehicle-side control quantity calculated based on the unified target exceeds the maximum acceleration or maximum deceleration that the vehicle model can physically execute, the control quantity will be limited to the capability range of the vehicle model to avoid issuing control commands that the vehicle cannot execute.

[0110] Third, in response to the simultaneous operation of both the accelerator and brake pedals (i.e., the driver simultaneously presses both the accelerator and brake pedals), the system prioritizes the operation of the brake pedal, enabling the emergency braking command to be executed immediately.

[0111] Any one or more combinations of the above safety constraints can be executed before the control quantity is issued. Each of the three is independent and can be selected for execution according to actual needs. Furthermore, when there is abnormal vehicle speed feedback, excessive communication delay, or missing vehicle model parameters, the system enters a degraded or safe stopping strategy.

[0112] In this application, a unified speed limit ensures that the vehicle's operating speed remains within a safe range under different scenarios; maximum acceleration and deceleration capacity limits prevent the issuance of control commands exceeding the vehicle's physical limits; and a brake priority mechanism ensures that braking commands can be executed immediately in emergency situations. These three elements together constitute a safety barrier between the unified objective and vehicle-side execution, guaranteeing the safety of remote driving. Furthermore, any one or more combinations of the aforementioned safety constraints can be executed before the control input is issued, allowing the system to flexibly select the appropriate option based on actual needs.

[0113] In some embodiments, the method may further include: Obtain the vehicle speed feedback status, communication link latency, and vehicle model parameter loading status of the currently controlled vehicle; In response to at least one of the following abnormal conditions, control the currently controlled vehicle to enter degraded mode or execute a safe parking policy: The vehicle speed feedback status is abnormal; The communication link delay exceeds a preset threshold; Vehicle model parameters are missing.

[0114] In this embodiment, the vehicle speed feedback status refers to the availability and validity of the vehicle speed signal reported by the currently controlled vehicle to the cloud through vehicle-cloud communication, including two states: normal (vehicle speed data is valid and updated in real time) and abnormal (vehicle speed data is missing, jumps, or timed out and not updated).

[0115] The latency of the communication link refers to the time difference between when the data is reported from the vehicle and when the data is received by the cloud, reflecting the real-time performance of the vehicle-cloud communication network.

[0116] Vehicle parameter loading status refers to the completion status of the system loading the longitudinal characteristic parameters of the currently controlled vehicle and the vehicle-side control interface parameters from the vehicle longitudinal characteristic parameter library, including two statuses: loading success and loading failure (parameter missing).

[0117] Degraded mode refers to a restricted operating state that the system enters when it detects abnormal conditions. In this mode, the system limits the vehicle's maximum speed or only allows braking operations to reduce operational risks.

[0118] A safe parking strategy refers to a preset control scheme that controls the vehicle to decelerate safely to a complete stop when the system detects a serious anomaly.

[0119] Abnormal conditions refer to abnormal situations that affect the safety of remote driving control, including at least one of the following three situations: abnormal vehicle speed feedback, communication link delay exceeding a preset threshold, and failure to load vehicle parameters.

[0120] During remote driving control, the system continuously monitors the vehicle speed feedback status, vehicle-to-cloud communication link latency, and vehicle model parameter loading status of the currently controlled vehicle. When any of the following abnormal conditions are detected, such as abnormal vehicle speed feedback (e.g., missing or outdated speed data), communication link latency exceeding a preset threshold (e.g., latency exceeding 500ms causing control commands to fail to reach the vehicle in real time), or missing vehicle model parameter loading status (e.g., the system cannot obtain the longitudinal characteristic parameters of the current vehicle from the parameter library), the system triggers a safety response mechanism, controlling the currently controlled vehicle to enter a degraded mode or execute a safe parking strategy.

[0121] In degraded mode, the system limits the vehicle's maximum speed or allows only braking. Under the safe stopping strategy, the system controls the vehicle to decelerate to a complete stop at a preset safe deceleration rate.

[0122] By continuously monitoring three abnormal conditions—vehicle speed feedback, communication latency, and parameter loading—and automatically triggering downgrade or parking strategies when any of these abnormalities occur, a complete remote driving safety assurance mechanism has been formed. This mechanism effectively addresses emergencies such as vehicle-side communication anomalies and data loss, thereby improving the safety of remote driving.

[0123] In some embodiments, it also includes: In response to the switching event of the controlled vehicle, the mapping relationship remains unchanged, and the longitudinal characteristic parameters and vehicle-side control interface parameters of the vehicle after the switch are reloaded.

[0124] In some embodiments, a controlled vehicle switching event refers to an event in which the cloud cabin switches from controlling the current vehicle to controlling another vehicle of a different model, such as a driver switching from controlling vehicle A to controlling vehicle B.

[0125] The mapping relationship between pedal opening and target longitudinal motion on the cloud-based platform remains unchanged across all vehicle models, so drivers do not need to change their pedal operation habits due to vehicle model changes.

[0126] Based on the new vehicle model after the switch, the system reads the corresponding longitudinal characteristic parameters and vehicle-side control interface parameters from the vehicle model longitudinal characteristic parameter library.

[0127] When the cloud-based control system switches between controlled vehicles (e.g., from vehicle A to vehicle B), the unified mapping relationship remains unchanged, meaning the driver does not need to change their operating habits. The system automatically loads the longitudinal characteristic parameters and control interface information of the new vehicle model from the vehicle model longitudinal characteristic parameter library module, only switching the longitudinal characteristic parameters and control quantity conversion strategy of the controlled vehicle. Specifically, after the switch, the pedal operation amount collected by the system is still mapped to a unified target longitudinal motion amount through the same unified mapping relationship, and then adaptively converted and closed-loop corrected according to the longitudinal characteristic parameters of the new vehicle model. Therefore, the pedal operation feel remains consistent when the driver switches between different vehicle models, eliminating the risk of misoperation due to vehicle model switching.

[0128] The following is a detailed description of the control method flow for autonomous vehicles provided in this application embodiment: The cloud cabin accesses and identifies the currently controlled vehicle model, and loads the longitudinal characteristic parameters and control interface information of the vehicle model from the vehicle model longitudinal characteristic parameter library; the unified pedal input acquisition module acquires the pedal opening of the driver in the cloud cabin; the unified mapping module maps the pedal opening to a unified target longitudinal motion amount according to the unified mapping relationship of all vehicle models (e.g., at a maximum unified vehicle speed of 10km / h, one-third of the pedal opening is mapped to a target vehicle speed of approximately 3.3km / h); the system calculates the speed difference based on the target vehicle speed and the current actual vehicle speed, and calculates the basic power request based on the control coefficient; vehicle model adaptive switching... The calculation module converts the basic power request into a vehicle-side control quantity that can be executed for the vehicle model based on the longitudinal characteristic parameters or vehicle model coefficient of the controlled vehicle model; the unified dynamic response shaping module applies unified acceleration and jerk limits to the change process of the unified target; the closed-loop feedback correction module performs closed-loop correction on the control quantity based on the actual vehicle speed and / or acceleration feedback, so that the actual output converges to the unified target; the safety constraint module performs speed limit, amplitude limit and safety verification on the control quantity before sending it to the vehicle for execution; when the cloud cabin switches the controlled vehicle model, the unified mapping relationship remains unchanged, only the longitudinal characteristic parameters of the vehicle model and the control quantity conversion strategy are switched to keep the driver's pedal operation feel consistent.

[0129] In practical applications involving unified operation of multiple vehicle models, such as a driver controlling three different vehicle types (A, B, and C) for remote escape from a difficult situation, the cloud-based control system demonstrates significant differences in performance: vehicle A accelerates from 0 to 100 km / h in 5 seconds, vehicle B in 10 seconds, and vehicle C in 8 seconds. With the same pedal opening, vehicle A naturally accelerates rapidly, vehicle B naturally accelerates slowly, and vehicle C is somewhere in between. The cloud-based control system can establish a unified mapping between pedal opening and target speed, with a maximum unified speed of 10 km / h. When the driver depresses the pedal one-third of its travel, regardless of which vehicle type is being controlled, the unified target speed is approximately 3.3 km / h. The system converts a unified target into different vehicle-side control quantities based on the longitudinal characteristic parameters of each vehicle model (model coefficients: A 0.5, B 1, C 0.8). Vehicle A receives a lower power demand (13%), vehicle B receives a higher power demand (26%), and vehicle C receives a moderate power demand (20.8%). Closed-loop feedback correction ensures that the actual vehicle speed for each model converges to 3.3 km / h. When the driver switches between different vehicle models, the same pedal operation provides consistent speed output and acceleration / deceleration feel.

[0130] In low-speed obstacle avoidance scenarios, when a vehicle is stuck in complex terrain, the driver in the cloud-based control system needs to perform precise maneuvers at extremely low speeds to escape. In this scenario, a safety constraint module sets a uniform speed limit of 10 km / h to prevent the driver from pressing the pedals too hard due to nervousness, which could cause the vehicle to suddenly lurch forward. Simultaneously, a unified dynamic response shaping module applies uniform acceleration and jerk limits to changes in the target vehicle speed, ensuring smooth and controllable acceleration and deceleration. The driver only needs to perform precise control according to standardized pedal operation habits, and the system automatically adapts to the longitudinal characteristics of the current vehicle model, ensuring smooth and controllable acceleration and deceleration.

[0131] The autonomous vehicle control method provided in this application can be executed by an autonomous vehicle control device. This application uses the autonomous vehicle control device executing the autonomous vehicle control method as an example to illustrate the autonomous vehicle control device provided in this application.

[0132] This application also provides a control device for an autonomous vehicle.

[0133] like Figure 4 As shown, the control device for the autonomous vehicle includes: a first processing module 410, a second processing module 420, a third processing module 430, a fourth processing module 440, a fifth processing module 450, and a sixth processing module 460.

[0134] The first processing module 410 is used to obtain the target amount of operation of the cloud cabin driver on the pedal; The second processing module 420 is used to determine the target longitudinal motion amount corresponding to the target operation amount based on the preset mapping relationship between the operation amount and the motion amount of the controlled vehicle. The mapping relationship is the same for different controlled vehicles. The third processing module 430 is used to load the longitudinal characteristic parameters and vehicle-end control interface parameters of the currently controlled vehicle. The fourth processing module 440 is used to convert the target longitudinal motion into an initial control quantity based on the longitudinal characteristic parameters of the currently controlled vehicle and the target longitudinal motion. The fifth processing module 450 is used to convert the initial control quantity into the vehicle-side control quantity corresponding to the current controlled vehicle based on the vehicle-side control interface parameters of the currently controlled vehicle. The sixth processing module 460 is used to send vehicle-side control quantities to the currently controlled vehicle in order to control the operation of the currently controlled vehicle.

[0135] The control device for autonomous vehicles provided in this application eliminates the impact of differences in pedal and acceleration response between different vehicle models on the driver's operating feel through a unified mapping mechanism that makes the mapping relationship the same for different controlled vehicles. This allows the driver in the cloud cabin to become familiar with a unified set of pedal operation mappings without having to rebuild operating intuition and muscle memory for each vehicle model. On this basis, by adaptively converting the vehicle-side control quantity according to the longitudinal characteristic parameters of the vehicle model, it ensures that the unified goal can be accurately executed on different vehicle models, so that the same pedal opening corresponds to the same speed output, reducing the risk of misoperation when switching vehicle models. Especially in fine control scenarios such as low-speed traction, it avoids the risk of vehicle overshoot or traction failure.

[0136] In some embodiments, the fourth processing module 440 can also be used for: Obtain the actual longitudinal motion state of the currently controlled vehicle; Calculate the deviation between the target longitudinal motion and the actual longitudinal motion state; The initial control quantity is calculated using a feedforward method based on the deviation and longitudinal characteristic parameters.

[0137] In some embodiments, the vehicle-side control interface parameters include the dimension type and value range of the vehicle-side control interface currently controlling the vehicle. The fifth processing module 450 can also be used for: Based on the type of dimension, the initial control quantity is converted into a control command with at least one of the following dimensions: target acceleration, target vehicle speed, percentage of drive or brake request, and torque request. The value of the control command is limited based on the range of values.

[0138] In some embodiments, the sixth processing module 460 can also be used for: Obtain the actual longitudinal motion state of the currently controlled vehicle; Based on the deviation between the actual longitudinal motion state and the target longitudinal motion amount, the vehicle-side control quantity is corrected, and the current vehicle operation is controlled based on the corrected vehicle-side control quantity.

[0139] In some embodiments, the first processing module 410 may also be used for: The target input quantity of the cloud cabin driver to at least one of the accelerator pedal and brake pedal is obtained. The target input quantity includes at least one of pedal travel, pedal opening and pedal force.

[0140] In some embodiments, the control device of the autonomous vehicle may further include a seventh processing module for: The same upper limit of acceleration and upper limit of jerk are applied to the change process of the target longitudinal motion corresponding to different vehicle models.

[0141] In some embodiments, the control device of the autonomous vehicle may further include an eighth processing module, which is used to determine a uniform speed limit based on the current control scenario before sending the vehicle-side control quantity to the currently controlling vehicle. The longitudinal motion of the target is limited based on a uniform speed limit.

[0142] In some embodiments, the control device of the autonomous vehicle may further include a ninth processing module for limiting the vehicle-end control quantity based on the maximum acceleration and maximum deceleration of the currently controlled vehicle before sending the vehicle-end control quantity to the currently controlled vehicle.

[0143] In some embodiments, the pedals include an accelerator pedal and a brake pedal. The control device of the autonomous vehicle may further include a tenth processing module, which, before sending a vehicle-side control quantity to the currently controlling vehicle, prioritizes responding to the operation quantity of the brake pedal in response to the operation of both the accelerator pedal and the brake pedal.

[0144] In some embodiments, the control unit of the autonomous vehicle may further include an eleventh processing module, for: Obtain the vehicle speed feedback status, communication link latency, and vehicle model parameter loading status of the currently controlled vehicle; In response to at least one of the following abnormal conditions, control the currently controlled vehicle to enter degraded mode or execute a safe parking policy: The vehicle speed feedback status is abnormal; The communication link delay exceeds a preset threshold; Vehicle model parameters are missing.

[0145] In some embodiments, the control unit of the autonomous vehicle may further include a twelfth processing module for: In response to the switching event of the controlled vehicle, the mapping relationship remains unchanged, and the longitudinal characteristic parameters and vehicle-side control interface parameters of the vehicle after the switch are reloaded.

[0146] The control device for the autonomous vehicle in this application embodiment can be an electronic device or a component within an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or other devices besides a terminal. For example, the electronic device can be a mobile phone, tablet computer, laptop computer, PDA, in-vehicle electronic device, mobile internet device (MID), augmented reality (AR) / virtual reality (VR) device, robot, wearable device, ultra-mobile personal computer (UMPC), netbook, or personal digital assistant (PDA), etc. It can also be a server, network attached storage (NAS), personal computer (PC), television (TV), ATM, or self-service machine, etc. This application embodiment does not specifically limit the scope of the device.

[0147] The control device for the autonomous vehicle in this application embodiment can be a device with an operating system. This operating system can be a Microsoft (Windows) operating system, an Android operating system, an iOS operating system, or other possible operating systems; this application embodiment does not specifically limit the specific operating system.

[0148] The control device for autonomous vehicles provided in this application embodiment can achieve… Figures 1 to 3 The various processes implemented in the method implementation examples will not be described again here to avoid repetition.

[0149] In some embodiments, this application also provides an autonomous vehicle that operates based on the control method for autonomous vehicles described in any of the above embodiments.

[0150] In some embodiments, such as Figure 5 As shown, this application embodiment also provides an electronic device 500, including a processor 501, a memory 502, and a computer program stored in the memory 502 and executable on the processor 501. When the program is executed by the processor 501, it implements the various processes of the above-described autonomous vehicle control method embodiment and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0151] It should be noted that the electronic devices in the embodiments of this application include the mobile electronic devices and non-mobile electronic devices described above.

[0152] This application also provides a non-transitory computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the above-described autonomous vehicle control method embodiments and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0153] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0154] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described control method for an autonomous vehicle.

[0155] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0156] This application embodiment also provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement the various processes of the above-described autonomous vehicle control method embodiment and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0157] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0158] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0159] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the related technology, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0160] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

[0161] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0162] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.

Claims

1. A control method for an autonomous vehicle, characterized in that, include: Obtain the target amount of pedal operation by the cloud cabin driver; Based on the preset mapping relationship between the operation amount and the motion amount of the controlled vehicle, the target longitudinal motion amount corresponding to the target operation amount is determined. The mapping relationship is the same for different controlled vehicles. Load the longitudinal characteristic parameters and vehicle-side control interface parameters of the currently controlled vehicle; Based on the longitudinal characteristic parameters of the currently controlled vehicle and the target longitudinal motion, the target longitudinal motion is converted into an initial control quantity; Based on the vehicle-side control interface parameters of the currently controlled vehicle, the initial control quantity is converted into the vehicle-side control quantity corresponding to the currently controlled vehicle. Send the vehicle-side control quantity to the currently controlled vehicle to control the operation of the currently controlled vehicle.

2. The control method for an autonomous vehicle according to claim 1, characterized in that, The step of converting the target longitudinal motion into an initial control quantity based on the longitudinal characteristic parameters of the currently controlled vehicle and the target longitudinal motion includes: Obtain the actual longitudinal motion state of the currently controlled vehicle; Calculate the deviation between the target longitudinal motion and the actual longitudinal motion state; Based on the deviation and the longitudinal characteristic parameters, the initial control quantity is calculated using a feedforward method.

3. The control method for an autonomous vehicle according to claim 1, characterized in that, The vehicle-side control interface parameters include the dimension type and value range of the vehicle-side control interface of the currently controlled vehicle. The step of converting the initial control quantity into a vehicle-side control quantity corresponding to the currently controlled vehicle based on the vehicle-side control interface parameters includes: Based on the unit type, the initial control quantity is converted into a control command with at least one of the following units: target acceleration, target vehicle speed, percentage of drive or brake request, and torque request. Based on the range of values, the value of the control command is limited.

4. The control method for an autonomous vehicle according to any one of claims 1-3, characterized in that, The control of the currently controlled vehicle includes: Obtain the actual longitudinal motion state of the currently controlled vehicle; Based on the deviation between the actual longitudinal motion state and the target longitudinal motion amount, the vehicle-end control quantity is corrected, so as to control the operation of the currently controlled vehicle based on the corrected vehicle-end control quantity.

5. The control method for an autonomous vehicle according to any one of claims 1-3, characterized in that, The acquisition of the target input amount of the cloud cabin driver to the pedals includes: The target operation amount of the cloud cabin driver on at least one of the accelerator pedal and the brake pedal is obtained, and the target operation amount includes at least one of pedal travel, pedal opening and pedal force.

6. The control method for an autonomous vehicle according to any one of claims 1-3, characterized in that, Also includes: The same upper limit of acceleration and upper limit of jerk are applied to the change process of the target longitudinal motion corresponding to different vehicle models.

7. The control method for an autonomous vehicle according to any one of claims 1-3, characterized in that, Before sending the vehicle-side control quantity to the currently controlled vehicle, the method further includes: A unified speed limit is determined based on the current control scenario; The longitudinal motion of the target is limited based on the unified speed limit.

8. The control method for an autonomous vehicle according to any one of claims 1-3, characterized in that, Before sending the vehicle-side control quantity to the currently controlled vehicle, the method further includes: The vehicle-side control quantity is limited based on the current maximum acceleration and maximum deceleration of the controlled vehicle.

9. The control method for an automated driving vehicle according to any one of claims 1-3, characterized in that, The pedals include an accelerator pedal and a brake pedal, and before sending the vehicle-side control quantity to the currently controlled vehicle, they also include: In response to both the accelerator pedal and the brake pedal being operated, the operation amount of the brake pedal is responded to first.

10. The control method for an automated driving vehicle according to any one of claims 1-3, characterized in that, Also includes: The vehicle speed feedback status, communication link delay time, and vehicle model parameter loading status of the currently controlled vehicle are obtained. In response to at least one of the following abnormal conditions, the currently controlled vehicle is controlled to enter a degraded mode or execute a safe parking strategy: The vehicle speed feedback status is abnormal; The delay time of the communication link exceeds a preset threshold; The vehicle model parameters are missing.

11. The control method for an autonomous vehicle according to any one of claims 1-3, characterized in that, Also includes: In response to a switching event of the controlled vehicle, the mapping relationship remains unchanged, and the longitudinal characteristic parameters and vehicle-side control interface parameters of the vehicle after the switching are reloaded.

12. A control device for an autonomous vehicle, characterized in that, include: The first processing module is used to obtain the target amount of pedal operation by the cloud cabin driver. The second processing module is used to determine the target longitudinal motion amount corresponding to the target operation amount based on the preset mapping relationship between the operation amount and the motion amount of the controlled vehicle. The mapping relationship is the same for different controlled vehicles. The third processing module is used to load the longitudinal characteristic parameters and vehicle-side control interface parameters of the currently controlled vehicle. The fourth processing module is used to convert the target longitudinal motion into an initial control quantity based on the longitudinal characteristic parameters of the currently controlled vehicle and the target longitudinal motion. The fifth processing module is used to convert the initial control quantity into the vehicle-side control quantity corresponding to the current controlled vehicle based on the vehicle-side control interface parameters of the currently controlled vehicle. The sixth processing module is used to send the vehicle-side control quantity to the currently controlled vehicle in order to control the operation of the currently controlled vehicle.

13. An autonomous vehicle, characterized in that, It operates based on the control method for autonomous vehicles as described in any one of claims 1-11.

14. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the control method for an autonomous vehicle as described in any one of claims 1-11.

15. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the control method for an autonomous vehicle as described in any one of claims 1-11.

Images