Vehicle energy recovery method, vehicle and readable storage medium

Abstract

The embodiment of the invention provides a vehicle energy recovery method, a vehicle and a readable storage medium, a plurality of sensors are deployed in the vehicle, the types of the plurality of sensors are different, the method comprises the steps that sensing data collected by the plurality of sensors are acquired, and the sensing data are used for representing the road surface state of a road on which the vehicle is currently driven; based on the sensing data, determining a road surface adhesion coefficient of the road, the road surface adhesion coefficient being used for representing a friction force between the vehicle and the road surface of the road; based on the road adhesion coefficient, an energy recovery mode of the vehicle is determined, and the energy recovery mode is used for representing a rule adopted for energy recovery of the vehicle; and in the process of controlling the vehicle to run according to the energy recovery mode, energy generated by the vehicle is recovered into a battery of the vehicle. The technical problem that the driving safety of the vehicle is low when the energy of the vehicle is recovered is solved.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, and more specifically, to a vehicle energy recovery method, a vehicle, and a readable storage medium. Background Technology

[0002] In the field of vehicle technology, coasting energy recovery is one of the key technologies for improving vehicle energy efficiency and reducing energy consumption. By converting some of the kinetic energy during the coasting or deceleration process into electrical energy, the vehicle's battery is charged, effectively utilizing the energy of the vehicle during non-acceleration phases.

[0003] In related technologies, when recovering energy from a vehicle, fixed energy recovery parameters are usually relied upon to achieve high energy recovery. Although this method works effectively under certain stable conditions, it is prone to causing the vehicle to lose control due to the lack of real-time feedback when the road surface adhesion is weak, resulting in technical problems of low vehicle driving safety.

[0004] There is currently no good solution to the technical problem of low vehicle driving safety when energy is recovered from vehicles. Summary of the Invention

[0005] This application provides a vehicle energy recovery method, a vehicle, and a readable storage medium to at least solve the technical problem of low vehicle driving safety during energy recovery.

[0006] According to one aspect of the embodiments of this application, a vehicle energy recovery method is provided. The vehicle is equipped with multiple sensors of different types. The method includes: acquiring sensing data collected by the multiple sensors, wherein the sensing data is used to characterize the road surface state of the road currently being traveled by the vehicle; determining a road surface adhesion coefficient based on the sensing data, wherein the road surface adhesion coefficient is used to characterize the friction between the vehicle and the road surface; determining an energy recovery mode for the vehicle based on the road surface adhesion coefficient, wherein the energy recovery mode is used to characterize the rules adopted for energy recovery of the vehicle; and recovering the energy generated by the vehicle into the vehicle's battery while controlling the vehicle to travel according to the energy recovery mode.

[0007] Optionally, the road surface adhesion coefficient is determined based on the sensing data, including: determining the weighting coefficients corresponding to the sensing data collected by multiple sensors based on the road surface condition, wherein the road surface condition is a dry road surface condition or a wet and slippery road surface condition; and calculating the weighted sensing data collected by multiple sensors based on the weighting coefficients to obtain the road surface adhesion coefficient.

[0008] Optionally, the multiple sensors include at least a vision sensor, a lidar sensor, and a wheel speed sensor. Based on the road surface condition, weighting coefficients corresponding to the sensing data collected by the multiple sensors are determined, including: in response to a dry road surface condition, determining a first weighting coefficient corresponding to the first sensing data collected by the vision sensor, a second weighting coefficient corresponding to the second sensing data collected by the lidar sensor, and a third weighting coefficient corresponding to the third sensing data collected by the wheel speed sensor, wherein the first weighting coefficient is greater than the second weighting coefficient, and the second weighting coefficient is greater than the third weighting coefficient; in response to a wet road surface condition, determining a fourth weighting coefficient corresponding to the first sensing data collected by the vision sensor, a fifth weighting coefficient corresponding to the second sensing data collected by the lidar sensor, and a sixth weighting coefficient corresponding to the third sensing data collected by the wheel speed sensor, wherein the fifth weighting coefficient is greater than the fourth weighting coefficient, and the fourth weighting coefficient is greater than the sixth weighting coefficient.

[0009] Optionally, determining the vehicle's energy recovery mode based on the road surface adhesion coefficient includes: determining an energy recovery mode matching the road surface adhesion coefficient from a mapping table, wherein the mapping table includes the mapping relationship between road surface adhesion coefficient samples and energy recovery mode samples.

[0010] Optionally, controlling vehicle movement according to an energy recovery mode includes: responding to a first energy recovery mode by controlling the vehicle to move at a first deceleration, wherein the first energy recovery mode is used to characterize energy recovery using the vehicle's electric braking; responding to a second energy recovery mode by controlling the vehicle to move at a second deceleration, wherein the second energy recovery mode is used to characterize energy recovery using the vehicle's electric braking and hydraulic braking, and the second deceleration is greater than the first deceleration; and responding to a third energy recovery mode by controlling the vehicle to move at a third deceleration, wherein the third energy recovery mode is used to characterize prohibiting the vehicle from performing energy recovery, and the third deceleration is greater than the second deceleration.

[0011] Optionally, the road surface adhesion coefficient is determined based on the sensing data, including: in response to the vision sensor being in a fault mode, determining the road surface adhesion coefficient of the road currently being traveled by the vehicle based on the sensing data of the lidar sensor and the sensing data of the wheel speed sensor.

[0012] Optionally, the road surface adhesion coefficient is determined based on the sensing data, including: in response to both the visual sensor and the lidar sensor being in a fault mode, the road surface adhesion coefficient of the road currently being traveled by the vehicle is determined based on the historical road surface adhesion coefficient of the vehicle in a historical time period.

[0013] Optionally, multiple sensors can synchronously acquire sensing data through a time synchronization protocol.

[0014] According to another aspect of the embodiments of this application, a vehicle energy recovery device is also provided. The vehicle is equipped with multiple sensors of different types. The device includes: an acquisition unit for acquiring sensing data collected by the multiple sensors, wherein the sensing data characterizes the road surface condition of the road currently being traveled by the vehicle; a first determination unit for determining the road surface adhesion coefficient based on the sensing data, wherein the road surface adhesion coefficient characterizes the friction between the vehicle and the road surface; a second determination unit for determining the vehicle's energy recovery mode based on the road surface adhesion coefficient, wherein the energy recovery mode characterizes the rules used for energy recovery of the vehicle; and a control unit for recovering the energy generated by the vehicle into the vehicle's battery while controlling the vehicle to travel according to the energy recovery mode.

[0015] According to another aspect of the embodiments of this application, a vehicle is also provided, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods in various embodiments of this application when it runs.

[0016] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of this application.

[0017] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0018] According to another aspect of the embodiments of this application, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the methods in various embodiments of this application.

[0019] According to another aspect of the embodiments of this application, a computer program is also provided, which, when executed by a processor, implements the methods of the various embodiments of this application.

[0020] In this embodiment, the sensing data collected by multiple different types of sensors of the vehicle determines the road surface adhesion coefficient of the road on which the vehicle is currently traveling. Then, based on the road surface adhesion coefficient, a matching energy recovery mode is adopted to recover the vehicle's energy. This can improve the energy recovery efficiency and ensure the driving safety of the vehicle when recovering energy, thereby solving the technical problem of low vehicle driving safety when recovering energy in related technologies. Attached Figure Description

[0021] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0022] Figure 1 This is a flowchart of a vehicle energy recovery method according to an embodiment of this application;

[0023] Figure 2 This is a flowchart of a fault degradation operation method according to an embodiment of this application;

[0024] Figure 3 This is a flowchart of a method for determining the road surface adhesion coefficient according to an embodiment of this application;

[0025] Figure 4 This is a flowchart of a multi-source data fusion method according to an embodiment of this application;

[0026] Figure 5 This is a schematic diagram of a vehicle energy recovery device according to an embodiment of this application. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0028] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0029] According to an embodiment of this application, an embodiment of a vehicle energy recovery method is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0030] This embodiment provides a vehicle energy recovery method, in which multiple sensors of different types are deployed. Figure 1 This is a flowchart of a vehicle energy recovery method according to an embodiment of this application, such as... Figure 1 As shown, the process includes the following steps.

[0031] Step S101: Acquire sensing data collected by multiple sensors.

[0032] In the technical solution provided by step S101 of this application, the aforementioned sensing data is used to characterize the road surface condition of the road currently being traveled by the vehicle.

[0033] In this embodiment, multiple sensors are deployed in the vehicle. These multiple sensors are of different types. For example, the multiple sensors may include, but are not limited to, vision sensors, lidar, and wheel speed sensor groups. Based on this, sensing data collected by the multiple sensors can be acquired.

[0034] Optionally, the visual sensor can be equipped with a high-resolution (e.g., 2 megapixels) and high-frame-rate (e.g., 30fps) RGB camera, capable of capturing visual features such as the color and texture of the road surface, as well as detecting the coverage of ice or water film through an infrared sensor.

[0035] Optionally, the lidar can use a 905nm wavelength and scan at a frequency of 10Hz to construct a high-density three-dimensional point cloud image of the road surface, thereby obtaining microstructural information of the road surface, including cracks, unevenness, and obstacles.

[0036] Optionally, the wheel speed sensor group consists of wheel speed sensors for each tire of the vehicle, with an accuracy of up to ±0.1 km / h. It can monitor the rotation speed of the tires in real time. By analyzing the changes in wheel speed, it can determine whether the vehicle is slipping and indirectly assess the road surface adhesion conditions.

[0037] Optionally, acquiring sensing data collected by multiple sensors can provide a comprehensive understanding of road conditions from multiple angles and levels, overcoming the limitations that may exist with a single sensor. For example, visual sensors perform well in well-lit conditions, but their effectiveness decreases at night or in adverse weather conditions (such as rain or snow); lidar can work in all weather conditions, but it is sensitive to the reflective properties of certain special road materials (such as snow and ice); wheel speed sensors play a role when a vehicle slips, but may lack insight into the microscopic features of the road surface.

[0038] Alternatively, since different types of sensors acquire data at different speeds and in different ways, without a strict time synchronization mechanism, this data will be difficult to use for real-time analysis. Therefore, a Precision Time Protocol (PTP) can be used to achieve hardware-level time synchronization of multiple sensors. Precise time synchronization ensures that all sensors operate on the same time reference, allowing for consistent and reliable road condition information even at high speeds or when road conditions change rapidly.

[0039] In this step, real-time road condition information is acquired through various sensors deployed on the vehicle, and this data is then synchronized in time, providing high-quality raw data for subsequent data fusion.

[0040] Step S102: Determine the road surface adhesion coefficient based on the sensing data.

[0041] In the technical solution provided by step S102 of this application, the road surface adhesion coefficient is used to characterize the friction between the vehicle and the road surface of the road on which the vehicle is currently traveling.

[0042] In this embodiment, after acquiring sensing data from multiple sensors, effective information can be extracted from the sensing data collected by multiple sensors to dynamically calculate the road surface adhesion coefficient of the current road surface, denoted as μ. This road surface adhesion coefficient can accurately assess the friction level between the vehicle and the road surface.

[0043] Optionally, the coefficient of friction (COP) is an important indicator of the friction between a vehicle's tires and the road surface. A higher COP means greater friction between the tires and the road, allowing the vehicle to withstand greater deceleration force during braking without slipping. Conversely, a lower COP indicates slippery surfaces or low-friction conditions such as ice or snow. In these situations, excessive regenerative braking force can easily lead to wheel lock-up or slippage, affecting the vehicle's stability and safety.

[0044] Optionally, when determining the road surface adhesion coefficient of the road currently being traveled by the vehicle based on sensing data, feature extraction can be performed on sensing data from different sensors (e.g., vision sensors, LiDAR sensors, wheel speed sensors). For example, scale-invariant feature transform (SIN) technology can be used to extract the road surface texture features of the road currently being traveled by the vehicle from the sensing data collected by the vision sensor. The road surface smoothness of the road currently being traveled by the vehicle can be calculated from the sensing data collected by the LiDAR sensor. The vehicle's slip state can be determined based on the sensing data obtained by the wheel speed sensor. After obtaining the road surface texture features, road surface smoothness, and vehicle slip state of the road currently being traveled by the vehicle, the road surface texture features, road surface smoothness, and vehicle slip state can be weighted and calculated according to the weight coefficients corresponding to the different sensors of the vehicle to determine the road surface adhesion coefficient of the road currently being traveled by the vehicle. This road surface adhesion coefficient can be the road surface adhesion coefficient within 50m of the road ahead of the vehicle.

[0045] Optionally, when fusion processing is performed on sensing data from multiple different sensors, the outputs of different sensors are assigned different weights. These weights are adjusted according to changes in road conditions. For example, in snowy conditions, the weight of lidar may be increased to 70% to rely more on high-precision point cloud data for road adhesion coefficient estimation.

[0046] Alternatively, an improved Kalman filter algorithm can be used to dynamically calculate the road surface adhesion coefficient of the road currently being driven by the vehicle. This algorithm can adjust the road surface adhesion coefficient in real time at a high update frequency of 100Hz.

[0047] In this step, by accurately estimating the road surface adhesion coefficient, the problem of inaccurate control in traditional energy recovery systems under complex road conditions is solved. This enables the vehicle to intelligently adjust the intensity of energy recovery based on the real-time road surface friction, avoiding skidding instability caused by energy recovery on low-adhesion surfaces, while maximizing energy recovery efficiency under high-friction conditions.

[0048] Step S103: Determine the vehicle's energy recovery mode based on the road surface adhesion coefficient.

[0049] In the technical solution provided by step S103 of this application, the above-mentioned energy recovery mode is used to characterize the rules adopted for energy recovery of the vehicle.

[0050] In this embodiment, after calculating the road surface adhesion coefficient of the road currently being driven by the vehicle through multi-source data fusion of vision, lidar and wheel speed sensors, and dynamic weight allocation algorithm, the energy recovery mode of the vehicle can be determined based on the road surface adhesion coefficient.

[0051] Optionally, the road surface adhesion coefficient is an important indicator for evaluating whether a vehicle can safely and efficiently recover coasting energy. A higher road surface adhesion coefficient means that the road surface is dry and the friction is high, which is suitable for using a more efficient energy recovery strategy; while a lower road surface adhesion coefficient may indicate that the road surface is wet or snowy. In this case, if an overly aggressive energy recovery strategy is adopted, it is easy to cause the vehicle to skid or become unstable. Therefore, it is necessary to adjust or even disable energy recovery in a timely manner to ensure driving safety.

[0052] Optionally, when determining the vehicle's energy recovery mode based on the road surface adhesion coefficient, the following rules can be followed. When the road surface adhesion coefficient > 0.7, it indicates that the road surface the vehicle is currently traveling on is dry, meaning the road conditions are good and the energy recovery mode can be fully activated. In this case, the vehicle can be controlled to recover energy according to the maximum energy recovery mode. For example, controlling the vehicle to apply electric braking with a maximum recovery capacity of -0.2g means that the vehicle can fully utilize the kinetic energy generated during coasting, converting it into electrical energy stored in the battery through the electric braking system without negatively impacting vehicle stability.

[0053] Optionally, if 0.5 ≤ road surface adhesion coefficient ≤ 0.7, it indicates that the road surface may be gradually changing from a dry state to a wet and slippery state, or is in a medium adhesion condition. In this case, the vehicle can be controlled to perform energy recovery in a limited recovery mode. For example, the maximum recovery capacity of electric braking can be adjusted by linear interpolation to adapt to changes in the road surface adhesion coefficient. Specifically, the maximum recovery capacity is smoothly transitioned from 0g to -0.2g to ensure that the vehicle achieves optimal energy recovery under different road surface conditions, while maintaining good driving stability and comfort.

[0054] Optionally, if the road surface adhesion coefficient is less than 0.5, the road surface on which the vehicle is currently traveling is determined to be slippery. In this case, the vehicle's energy recovery mode will be limited or completely disabled. That is, in this mode, the maximum recovery capacity of electric braking is 0g, and the system will rely entirely on hydraulic braking to maintain vehicle stability, avoiding the risk of vehicle instability due to excessive energy recovery on low-adhesion surfaces.

[0055] Optionally, to further optimize the control strategy, an execution time T for the coasting recovery capability limit is introduced, which is interpolated based on the vehicle speed. This is to ensure that when the vehicle speed changes, the adjustment of the energy recovery mode can respond quickly to changes in road conditions, while avoiding the impact of excessively rapid adjustments to the energy recovery strategy on vehicle ride stability.

[0056] In this step, the energy recovery mode of the vehicle is determined based on the road surface adhesion coefficient, which can realize the optimal energy recovery mode under different road surface conditions, thereby improving energy recovery efficiency while ensuring vehicle driving safety and comfort.

[0057] Step S104: During the process of controlling the vehicle to drive in the energy recovery mode, the energy generated by the vehicle is recovered into the vehicle's battery.

[0058] In the technical solution provided by step S104 of this application, during the process of controlling the vehicle to drive in the energy recovery mode, the energy generated by the vehicle can be recovered into the vehicle's battery to provide energy for subsequent electric drive or other vehicle electronic devices.

[0059] In this embodiment, after determining the energy recovery mode that matches the road surface conditions of the road surface where the vehicle is currently traveling based on the road surface adhesion coefficient, the vehicle's driving process can be controlled according to the energy recovery mode, and energy can be recovered during the vehicle's driving process.

[0060] Optionally, when controlling the vehicle to drive in the "maximum energy recovery mode", the maximum energy recovery capacity of the vehicle's electric braking can be set to -0.2g. This means that the vehicle will use electric braking to recover energy as much as possible, while hydraulic braking will hardly intervene unless the electric braking reaches its limit.

[0061] Optionally, when controlling the vehicle in the "limited energy recovery mode", the maximum recovery capacity of the vehicle's electric braking will be dynamically adjusted according to the road surface adhesion coefficient, ranging from 0 to -0.2g. The insufficient energy recovery demand will be supplemented by hydraulic braking to balance the recovery efficiency and the stability of vehicle control.

[0062] Optionally, when controlling the vehicle in "stop energy recovery" mode, the vehicle's electric braking energy recovery capability is completely disabled, and the necessary braking force is borne by the hydraulic braking system to prevent wheel slippage and ensure vehicle driving safety.

[0063] Optionally, depending on the energy recovery mode, the vehicle's motor can be controlled to operate as a generator, converting the kinetic energy of the vehicle during coasting or downhill driving into electrical energy. This conversion process is achieved by reversing the motor's rotation, using the rotation of the wheels to drive the motor to generate electricity.

[0064] Optionally, the converted electrical energy is then fed into the vehicle's Battery Management System (BMS) and stored in the vehicle's battery to power subsequent electric drives or other onboard electronic devices. The BMS monitors the battery status to ensure efficient energy storage and safe use.

[0065] In this step, the energy recovery process is intelligently controlled, and the energy recovery strategy can be dynamically optimized according to the actual road conditions. This maximizes the recovery of usable energy while ensuring vehicle stability, reducing the number of battery charge-discharge cycles and extending battery life.

[0066] In steps S101 to S104 above, the sensing data collected by multiple different types of sensors of the vehicle determines the road surface adhesion coefficient of the road on which the vehicle is currently traveling. Then, based on the road surface adhesion coefficient, a matching energy recovery mode is adopted to recover the vehicle's energy. This can improve the energy recovery efficiency and ensure the driving safety of the vehicle when recovering energy, thereby solving the technical problem of low vehicle driving safety when recovering energy in related technologies.

[0067] The vehicle energy recovery method described in this application will be further described below.

[0068] As an optional implementation, step S102, determining the road surface adhesion coefficient based on sensing data, includes: determining weighting coefficients corresponding to sensing data collected by multiple sensors based on the road surface condition, wherein the road surface condition is either a dry road surface condition or a wet and slippery road surface condition; and performing weighted calculations on the sensing data collected by multiple sensors based on the weighting coefficients to obtain the road surface adhesion coefficient.

[0069] In this embodiment, when determining the road surface adhesion coefficient of the road on which the vehicle is traveling based on sensing data, since different types of sensors perform differently under different road conditions, the degree of dependence on the sensing data collected by different sensors varies when determining the road surface adhesion coefficient of the vehicle. Based on this, the weighting coefficients corresponding to the sensing data collected by multiple sensors can be determined according to the road surface condition of the road currently being traveled by the vehicle.

[0070] Optionally, on dry roads, visual sensors (e.g., RGB cameras + infrared sensors) can more effectively capture road texture and the presence of ice / water film, thus their weighting coefficients can be relatively high. On wet roads, however, LiDAR (especially in snow conditions) performs better because point cloud data better reflects the three-dimensional features and subtle structural changes of the road surface; in this case, the weighting coefficient of LiDAR can be relatively increased. By dynamically adjusting the weighting coefficients of the sensing data from each sensor based on the road conditions of the vehicle's current route, the advantages of various sensors can be effectively utilized under different road conditions, improving the accuracy and reliability of road adhesion coefficient estimation.

[0071] Optionally, after determining the weighting coefficients corresponding to multiple sensors, the sensing data collected by each sensor can be weighted and calculated to obtain the final road surface adhesion coefficient.

[0072] In this step, by dynamically adjusting the weighting coefficients of the sensing data from multiple sensors and performing weighted calculations on the data from various sensors, an accurate estimation of the road surface adhesion coefficient is achieved. This overcomes the limitations of a single sensor in a specific environment, improves the responsiveness and adaptability of the intelligent coasting energy recovery control system, and thus maximizes energy recovery efficiency while ensuring driving safety.

[0073] As an optional implementation, the multiple sensors include at least a vision sensor, a lidar sensor, and a wheel speed sensor. Based on the road surface condition, weighting coefficients corresponding to the sensing data collected by the multiple sensors are determined, including: in response to the road surface condition being a dry road surface, determining a first weighting coefficient corresponding to the first sensing data collected by the vision sensor, a second weighting coefficient corresponding to the second sensing data collected by the lidar sensor, and a third weighting coefficient corresponding to the third sensing data collected by the wheel speed sensor, wherein the first weighting coefficient is greater than the second weighting coefficient, and the second weighting coefficient is greater than the third weighting coefficient; in response to the road surface condition being a wet road surface, determining a fourth weighting coefficient corresponding to the first sensing data collected by the vision sensor, a fifth weighting coefficient corresponding to the second sensing data collected by the lidar sensor, and a sixth weighting coefficient corresponding to the third sensing data collected by the wheel speed sensor, wherein the fifth weighting coefficient is greater than the fourth weighting coefficient, and the fourth weighting coefficient is greater than the sixth weighting coefficient.

[0074] In this embodiment, the multiple sensors include at least a vision sensor, a lidar sensor, and a wheel speed sensor. When determining the weighting coefficients corresponding to the sensing data collected by the multiple sensors based on the road surface condition, if the road surface condition is dry, the vision sensor can effectively identify the texture and color of the road surface under dry conditions. Therefore, the weighting coefficient corresponding to the first sensing data collected by the vision sensor can be set as the first weighting coefficient. Although lidar can provide valuable point cloud data under various road conditions, its effectiveness may be slightly less than that of the vision sensor in a dry road environment. Therefore, the weighting coefficient corresponding to the second sensing data collected by the lidar sensor can be set as the second weighting coefficient, and the weighting coefficient corresponding to the third sensing data of the wheel speed sensor can be set as the third weighting coefficient. The first weighting coefficient is greater than the second weighting coefficient, and the second weighting coefficient is greater than the third weighting coefficient.

[0075] Optionally, if the road surface is wet and slippery, the visual sensor may be affected by rain and snow, leading to a decrease in image quality and affecting the accurate estimation of the road adhesion coefficient. Therefore, the weighting coefficient of the visual sensor can be set as the fourth weighting coefficient. Since the LiDAR sensor can better penetrate the rain and snow film on wet or snowy roads to detect the true condition of the road surface, the 3D point cloud data it provides is more critical in assessing adhesion. Therefore, the weighting coefficient of the LiDAR sensor can be set as the fifth weighting coefficient. The weighting coefficient of the wheel speed sensor is set as the sixth weighting coefficient, where the fifth weighting coefficient is greater than the fourth weighting coefficient, and the fourth weighting coefficient is greater than the sixth weighting coefficient.

[0076] In this step, the weighting coefficients are dynamically adjusted according to different road surface conditions, enabling intelligent responses to changes in the external environment. On dry roads, visual information is more reliable, while on wet or complex road conditions, the roles of lidar and wheel speed sensors become more prominent. This flexible weighting mechanism ensures that, regardless of road conditions, the system can make accurate road condition assessments based on the most reliable sensor data, thereby effectively guiding the formulation of dynamic coasting energy recovery strategies. This ensures both the efficiency of energy recovery and the stability and safety of vehicle operation.

[0077] As an optional implementation, step S103, determining the vehicle's energy recovery mode based on the road surface adhesion coefficient, includes: determining the energy recovery mode matching the road surface adhesion coefficient from a mapping table, wherein the mapping table includes the mapping relationship between road surface adhesion coefficient samples and energy recovery mode samples.

[0078] In this embodiment, the mapping table provides a series of road surface adhesion coefficient samples and corresponding energy recovery mode samples. Based on this, after determining the road surface adhesion coefficient of the road currently being driven by the vehicle, the energy recovery mode matching the road surface adhesion coefficient can be determined from the mapping table.

[0079] Optionally, if the road surface adhesion coefficient is >0.7, it indicates that the road surface the vehicle is currently driving on is a dry road surface. The energy recovery mode that matches the road surface adhesion coefficient can be determined as the maximum coasting energy recovery mode, that is, the maximum recovery capacity of electric braking is set to -0.2g, which allows the vehicle to recover as much coasting energy as possible while ensuring driving safety, so as to reduce energy consumption and improve energy efficiency.

[0080] Optionally, if 0.5 ≤ road surface adhesion coefficient ≤ 0.7, it indicates that the vehicle is currently traveling on a slightly wet or semi-dry road surface. In this case, the vehicle's energy recovery mode is limited, meaning the primary energy recovery mode is determined. In this situation, the maximum recovery capacity of the vehicle's electric braking will be dynamically adjusted based on the precise value of the road surface adhesion coefficient. For example, a linear interpolation method can be used to ensure that the energy recovery intensity matches the road surface adhesion, preventing the vehicle from losing control due to excessive energy recovery when road conditions are poor.

[0081] Optionally, if the road surface adhesion coefficient is less than 0.5, it indicates that the vehicle is currently traveling on a wet or snowy road surface. In this case, the electric braking energy recovery capability will be stopped or significantly reduced to avoid wheel slippage or vehicle instability. That is, in this situation, the vehicle's energy recovery mode can be determined as the stop recovery mode. At this time, the maximum regenerative braking capability of the vehicle is set to 0g, and any braking demand will be entirely handled by the hydraulic braking system to ensure safe driving of the vehicle under complex road conditions.

[0082] Optionally, by selecting an energy recovery mode that matches the current road surface adhesion coefficient from a mapping table, the energy recovery strategy can be adjusted in real time during vehicle operation. Under conditions of high adhesion coefficient and good road surface conditions, gliding energy can be recovered to the maximum extent, improving the vehicle's overall energy efficiency. When the vehicle is traveling on low-traction surfaces such as wet, slippery, or icy / snow-covered surfaces, electric braking regeneration is automatically reduced or even disabled, ensuring vehicle safety through the hydraulic braking system and avoiding potential skidding risks caused by energy recovery.

[0083] In this step, by flexibly using the mapping relationship table, the energy recovery mode and the road surface adhesion coefficient are accurately matched. This enables intelligent, efficient and safe management of vehicle coasting energy recovery, overcoming the limitations of traditional fixed energy recovery strategies. By dynamically adjusting the energy recovery parameters, stable driving and efficient energy recovery of the vehicle under various road conditions are ensured.

[0084] As an optional implementation, controlling vehicle movement according to an energy recovery mode includes: responding to a first energy recovery mode by controlling the vehicle to move at a first deceleration, wherein the first energy recovery mode is used to characterize energy recovery using the vehicle's electric braking; responding to a second energy recovery mode by controlling the vehicle to move at a second deceleration, wherein the second energy recovery mode is used to characterize energy recovery using the vehicle's electric braking and hydraulic braking, and the second deceleration is greater than the first deceleration; and responding to a third energy recovery mode by controlling the vehicle to move at a third deceleration, wherein the third energy recovery mode is used to characterize prohibiting the vehicle from performing energy recovery, and the third deceleration is greater than the second deceleration.

[0085] In this embodiment, the vehicle deceleration process can be dynamically controlled according to the determined energy recovery mode to achieve efficient and safe energy recovery.

[0086] Optionally, if the vehicle's energy recovery mode is the first energy recovery mode, i.e., the maximum energy recovery mode, the vehicle will utilize the electric motor's electric braking function to recover as much energy as possible without the intervention of hydraulic brakes. The first deceleration is set to -0.2g, meaning the vehicle will decelerate at a relatively high rate while converting most of the kinetic energy into electrical energy, which is stored in the battery. In this mode, the system prioritizes energy recovery efficiency and is suitable for road conditions with sufficient friction.

[0087] Optionally, if the vehicle's energy recovery mode is the second energy recovery mode, i.e., the limited energy recovery mode, the energy recovery force of the vehicle's electric braking is limited to adapt to the current low road adhesion coefficient. The second deceleration is higher than the first deceleration because, in this second energy recovery mode, in addition to electric braking, hydraulic braking will also intervene moderately to compensate for insufficient electric braking recovery force to cope with sudden deceleration demands. This measure helps maintain the vehicle's driving stability and control on slippery roads while still enabling a certain degree of energy recovery.

[0088] Optionally, if the vehicle's energy recovery mode is the third energy recovery mode, i.e., the stop recovery mode, then the vehicle's electric braking energy recovery is completely disabled to avoid the risk of wheel lock-up or slippage caused by using electric braking on low-friction surfaces. All necessary deceleration is performed by the hydraulic braking system. The third deceleration is the greatest of the three modes because hydraulic braking can provide high braking force in a very short time, ensuring that the vehicle can safely decelerate to a stop even under the most adverse road conditions.

[0089] In this step, by dynamically adjusting the vehicle's deceleration behavior, energy recovery control adapted to road surface adhesion conditions is achieved. A balance is found between energy recovery efficiency and driving safety, ensuring that the vehicle can safely decelerate and stop under any conditions, while also making the most of the kinetic energy during coasting.

[0090] As an optional implementation, step S102, determining the road surface adhesion coefficient based on sensing data, further includes: in response to the vision sensor being in a fault mode, determining the road surface adhesion coefficient of the road currently being traveled by the vehicle based on sensing data from the lidar sensor and sensing data from the wheel speed sensor.

[0091] In this embodiment, assuming the vision sensor, lidar, and wheel speed sensor are all functioning normally, the data collected by these sensors are combined and an improved Kalman filter algorithm is used to estimate the road surface adhesion coefficient in real time through a dynamic weight allocation mechanism. This process fully utilizes the advantages of different sensors; for example, the vision sensor can identify the texture features of the road surface, the lidar can provide three-dimensional structural information of the road surface, and the wheel speed sensor can monitor the rotation state of the wheels in real time. The combination of the three can more accurately reflect the actual friction condition of the road surface, providing a key basis for energy recovery control.

[0092] Optionally, when a vision sensor malfunctions, the system will automatically switch to backup mode and readjust the sensor data fusion strategy. For example, the weight of the LiDAR sensor will be significantly increased, and the wheel speed sensor will be given more consideration to compensate for the lack of vision data.

[0093] For example, because lidar provides stable road surface information under various weather and lighting conditions, its 3D point cloud data can indirectly reflect the smoothness of the road surface and the condition of obstacles. Smoothness is a crucial factor in estimating the road adhesion coefficient. In the event of visual sensor failure, increasing the weight of lidar means the system relies more heavily on the road features captured by lidar for road adhesion coefficient estimation, thus maintaining the continuity and reliability of the control strategy. Wheel speed sensors can monitor the vehicle's slip state in real time, which is particularly helpful in determining road adhesion conditions. When visual information is lacking, the analysis of wheel speed data, such as changes in slip ratio, can serve as supplementary information, helping the system more accurately determine whether the vehicle is traveling on a low-μ surface, further optimizing energy recovery control.

[0094] In this step, by fusing multimodal sensor data, not only is the road surface adhesion coefficient accurately estimated under normal operating conditions, but also, when the visual sensor fails, the weight of the lidar and wheel speed sensor is increased to ensure that the system can continue to operate under various adverse conditions, providing a robust decision-making basis for the vehicle's intelligent coasting energy recovery.

[0095] As an optional implementation, step S102, determining the road surface adhesion coefficient based on sensing data, includes: in response to both the visual sensor and the lidar sensor being in fault mode, determining the road surface adhesion coefficient of the road currently being traveled by the vehicle based on the historical road surface adhesion coefficient of the vehicle within a historical time period.

[0096] In this embodiment, when determining the road surface adhesion coefficient based on sensing data, if both the visual sensor and the lidar sensor malfunction and lose the ability to collect sensing data in real time, the road surface adhesion coefficient of the road currently being driven by the vehicle can be determined based on the historical road surface adhesion coefficient of the vehicle within a historical time period.

[0097] For example, historical road surface adhesion coefficient values ​​recorded within a recent period (e.g., the past few seconds or tens of seconds) can be used to determine the road surface adhesion coefficient on the vehicle's current driving path. This data is typically collected during normal sensor operation and reflects trends in road conditions along the vehicle's path. For instance, the road surface adhesion coefficient on the vehicle's current driving path can be determined using the following formula.

[0098]

[0099] in, It can be used to represent the road surface adhesion coefficient on the road where the vehicle is currently traveling. It can be used to represent the historical road surface adhesion coefficient determined at the previous time step. It can be used to represent the historical road surface adhesion coefficient determined at the penultimate time of the current moment.

[0100] Optionally, after determining the road surface adhesion coefficient of the road the vehicle is currently traveling on based on the historical road surface adhesion coefficient within a historical time period, the instrument panel will illuminate a corresponding warning light to remind the driver that the system is operating in degraded mode. The driver can use the instrument panel to understand the current operating status of the vehicle and whether manual intervention is required, such as adjusting the vehicle speed or paying closer attention to changes in the road surface.

[0101] In this step, under sensor failure mode, the road adhesion coefficient is estimated using historical data and predictive models, ensuring that the core functions of the system can continue to some extent. Although this strategy is not as accurate as real-time multimodal sensor fusion under ideal conditions, it provides a basic operating framework for the system until the sensor is repaired or replaced. Furthermore, the implementation of the failure degradation operation strategy enhances the system's robustness in emergency situations, providing necessary safety assurances and operational guidance for the vehicle driver even under adverse conditions.

[0102] As an optional implementation, multiple sensors synchronously acquire sensing data through a time synchronization protocol.

[0103] In this embodiment, in intelligent coasting energy recovery control based on multi-sensor fusion, multiple sensors can be hardware synchronized via a time synchronization protocol to ensure that they collect data synchronously in time, thereby achieving accurate road condition analysis and effective energy recovery control. This time synchronization protocol can be the PTP protocol, used to calibrate and coordinate the sampling times of multiple sensors, enabling the processing and analysis of sensor data based on a consistent time reference.

[0104] Optionally, road conditions may change rapidly within a very short time during vehicle operation. If the time bases of the sensors are not synchronized, the data collected from different sensors may reflect different road conditions at the same instant, leading to biased analysis results. For example, a vision sensor may capture a momentary dry road surface, while a wheel speed sensor may detect signs of wheel slippage at the same instant, which is clearly caused by inconsistencies in data acquisition time.

[0105] Optionally, multimodal sensor fusion algorithms rely on synchronized data input. Only when all sensor data are temporally aligned can the accuracy of the road adhesion coefficient estimate output by the algorithm be ensured. For example, when using an improved Kalman filter algorithm for real-time estimation of the μ value, even small temporal differences in the sensor data can amplify algorithm errors and affect the final road adhesion coefficient calculation result.

[0106] Optionally, coasting energy recovery control requires rapid decisions based on instantaneous road surface adhesion conditions. The use of asynchronous data may lead to system response delays, making it impossible to accurately capture road surface changes, thus affecting the real-time performance and effectiveness of braking control. Especially when the vehicle is traveling at high speed or road conditions are changing rapidly, the lack of time synchronization will significantly reduce the system's control accuracy and safety.

[0107] In this step, time synchronization is fundamental to ensuring effective fusion of multimodal sensor data and accurate road condition analysis in vehicle coasting energy recovery control. By using time synchronization technologies such as the PTP protocol, time discrepancies between sensors can be overcome, enabling real-time and consistent data acquisition, thereby ensuring the accuracy of dynamic road adhesion coefficient estimation and the efficient execution of the coasting energy recovery strategy.

[0108] The above technical solutions of the present application embodiments will be further illustrated below with reference to preferred embodiments of the present invention.

[0109] The following section will provide a further introduction to the vehicle's multimodal perception system.

[0110] The multimodal perception system aims to achieve comprehensive and accurate perception of the vehicle's driving environment by integrating various types of sensors. In intelligent coasting energy recovery control based on multimodal sensor fusion, this multimodal perception system ensures that the vehicle can safely and efficiently recover energy under different road conditions. The main components and operating mechanism of this system are described in detail below.

[0111] Visual sensors (such as RGB cameras and infrared sensors) are used to capture visual features of the road, including but not limited to the texture, color, and presence of ice or water film. In this system, the RGB camera used has a 2-megapixel resolution and a 30fps frame rate, enabling it to quickly acquire clear road images. Combined with an infrared sensor (e.g., with a wavelength range of 8-14μm), the system can identify the slippery condition of the road surface even in low light or special environmental conditions, which is crucial for accurately estimating the road adhesion coefficient.

[0112] LiDAR (Light Detection and Ranging) is a sensor that measures distance by emitting laser pulses and receiving the echoes, used to construct three-dimensional maps of the vehicle's surroundings. In the intelligent coasting energy recovery control system, the LiDAR used operates at a wavelength of 905nm, a scanning frequency of 10Hz, and has 16 laser emission lines. This configuration allows the LiDAR to quickly and accurately acquire three-dimensional point cloud data of the road surface, reflecting not only the location of obstacles but also revealing the road surface's microstructure, such as potholes and cracks—key factors affecting the road surface's adhesion coefficient.

[0113] Wheel speed sensors monitor the rotational speed of a vehicle's tires, providing the system with fundamental information about the vehicle's motion. This data is crucial for understanding the immediate contact between the wheels and the road surface, as changes in wheel speed indirectly reflect changes in road adhesion conditions. By monitoring the independent speeds of the four wheels, the system can gain a more detailed understanding of the vehicle's coasting and braking states, which is extremely important for dynamically adjusting energy recovery strategies.

[0114] This application employs a triple data fusion architecture combining visual sensors, LiDAR, and wheel speed sensors. In this way, data from different sensors are comprehensively analyzed in a central processing unit, forming a complete understanding of the current road conditions. This fusion not only improves the integrity and reliability of the data but also reduces false alarms or missed alarms that might arise from a single sensor through mutual verification.

[0115] Optionally, the improved Kalman filter algorithm plays the role of the core computing engine in the multimodal perception system, with an update frequency of up to 100Hz. This high frequency ensures that the estimation of the road adhesion coefficient (μ value) can keep up with the rapid pace of vehicle movement and reflect changes in road conditions in a timely manner. The Kalman filter can effectively handle noise in sensor data, continuously updating the estimate of the μ value through prediction and correction steps, making it closer to the true value, and providing stable estimation results even under complex and changing road conditions.

[0116] To adapt to different driving conditions, a dynamic confidence weighting mechanism is introduced, adjusting the data weights based on sensor performance in specific environments. For example, the weight of LiDAR increases to 70% in snowy conditions because the reflective properties of snow make LiDAR superior to other sensors in acquiring three-dimensional road surface information, while the weights of infrared and vision sensors are correspondingly reduced. This mechanism ensures that the system always relies on the most reliable sensor data under various conditions, providing the most accurate support for subsequent adjustments to the energy recovery mode.

[0117] This multimodal perception system, working in conjunction with improved control algorithms, significantly enhances the accuracy and efficiency of intelligent coasting energy recovery control. By comprehensively analyzing visual features, 3D point cloud data, and motion states, it can assess the road surface adhesion coefficient in real time and intelligently adjust the energy recovery strategy accordingly, thereby maximizing energy recovery benefits while ensuring vehicle safety.

[0118] Optionally, in intelligent coasting energy recovery control based on multimodal sensor fusion, a dynamic fusion algorithm plays a core role. By comprehensively analyzing data from multiple sensors, it accurately assesses the road surface adhesion coefficient (μ) within 50 meters in front of the vehicle. The μ value is an important parameter measuring the magnitude of friction between the wheels and the road surface, directly affecting the vehicle's energy recovery strategy and driving safety. The dynamic fusion algorithm will be introduced below.

[0119] Optionally, the algorithm input parameters include texture entropy (H) and reflection intensity variation coefficient (CV). Texture entropy (H), provided by a visual sensor (e.g., an RGB camera, an infrared sensor), reflects the complexity and randomness of the texture on the road surface and is generally proportional to the road's friction. On dry roads, the texture is detailed and the entropy is high; while on wet or icy roads, the texture becomes smoother and the entropy decreases. The reflection intensity variation coefficient (CV), obtained by LiDAR and infrared sensors, reflects the fluctuation of the road surface's reflection intensity. A higher CV value indicates greater inconsistency in reflection intensity, potentially suggesting the presence of low-adhesion areas such as wet, slippery areas, water accumulation, or frost.

[0120] Alternatively, the dynamic fusion algorithm can be expressed by the following formula.

[0121] μ=0.6 sigmoid(2H-3)+0.4 tanh(CV / 0.1)

[0122] Where H is the texture entropy value and CV is the reflection intensity variation coefficient.

[0123] The fault degradation operation strategy in the embodiments of this application will be further described below.

[0124] Figure 2 This is a flowchart of a fault degradation operation method according to an embodiment of this application, such as... Figure 2 As shown, if a sensor malfunction is detected in the multimodal perception system, such as a visual sensor failure or a LiDAR failure, a corresponding backup mode can be adopted to determine the road surface adhesion coefficient of the road being traveled. For example, when the visual sensor fails, the system can switch to a LiDAR + wheel speed sensor mode, that is, the road surface adhesion coefficient of the road being traveled can be determined using the sensing data from the LiDAR sensor and the wheel speed sensor. When the LiDAR sensor fails, a texture feature compensation algorithm can be enabled to determine the road surface adhesion coefficient of the road being traveled, wherein the error compensation for the road surface adhesion coefficient is ≤ ±0.1.

[0125] Optionally, Table 1 is a mapping relationship between road surface adhesion coefficient and vehicle energy recovery mode according to an embodiment of this application.

[0126] Table 1. Mapping Relationship between Road Surface Adhesion Coefficient and Vehicle Energy Recovery Mode

[0127]

[0128] Optionally, the execution time T for limiting the coasting recovery capability can be interpolated and calibrated according to vehicle speed. This execution time T refers to the time interval from detecting the current road conditions requiring limitation of the coasting recovery capability to actually implementing the limitation (e.g., reducing the electric braking recovery torque). The calibration of this time interval needs to take vehicle speed into account, as vehicle speed directly affects braking distance, and thus also affects the safe execution of the energy recovery strategy.

[0129] Optionally, interpolation calibration is based on a series of preset speed-execution time T correspondences to quickly find or calculate the most suitable execution time T for the current vehicle speed. For example, a table is pre-defined with speeds ranging from low to high, corresponding to different execution times T. For instance, when the vehicle speed is 30 km / h, the execution time T is 0.5 seconds; when the vehicle speed is 60 km / h, the execution time T is 1 second; and when the vehicle speed is 90 km / h, the execution time T is 1.5 seconds. When the actual vehicle speed is between these preset values, such as 45 km / h or 75 km / h, the above preset time values ​​will not be directly applied. Instead, interpolation will be used to calculate a more accurate execution time T. The interpolation method can be linear interpolation or a higher-order interpolation method, such as quadratic interpolation or cubic spline interpolation, to more smoothly transition changes in execution time T. No specific limitation is made here.

[0130] The process of determining the road surface adhesion coefficient of the road on which the vehicle travels will be described in more detail below.

[0131] Figure 3 This is a flowchart of a method for determining the road surface adhesion coefficient according to an embodiment of this application, as shown below. Figure 3 As shown, the method includes the following steps.

[0132] Step S301, data synchronization.

[0133] In this embodiment, the PTP protocol is first used to ensure time synchronization of all sensors (vision, lidar, wheel speed, etc.) at the hardware level. Then, a deviation compensation formula is used to compensate for synchronization errors caused by sensor spacing (d) and the speed of light (c), which can be expressed as follows.

[0134]

[0135] Where d is the sensor spacing and c is the speed of light.

[0136] Step S302, Feature extraction.

[0137] In this embodiment, an improved Scale-Invariant Feature Transform (SIFT) algorithm can be used to extract gradient features of road surface texture from images acquired by a visual sensor. Geometric features of the road surface are extracted from point cloud data acquired by LiDAR by calculating the point cloud curvature, where the formula for calculating the point cloud curvature c can be expressed as follows.

[0138]

[0139] Where λ is the eigenvalue of the covariance matrix.

[0140] Step S303: Multi-source data fusion.

[0141] In this embodiment, during the data fusion process, the weights of each sensor can be dynamically adjusted according to the dryness or wetness of the road surface.

[0142] For example, Table 2 is a dynamic weight allocation table according to an embodiment of this application.

[0143] Table 2 Dynamic Weight Allocation Table

[0144]

[0145] Optionally, as shown in Table 2, on dry surfaces, the visual sensor has a higher weight (0.5), while the lidar and wheel speed sensors have relatively lower weights (0.3 and 0.2, respectively). On wet surfaces, the lidar weight increases to 0.6, the visual sensor weight decreases to 0.3, and the wheel speed sensor weight remains at 0.1. This weighting strategy considers the reliability and effectiveness of different sensors under different environments, ensuring the accuracy and robustness of the μ value estimation.

[0146] Figure 4 This is a flowchart of a multi-source data fusion method according to an embodiment of this application, such as... Figure 4 As shown, before data fusion, the data collected from visual sensors, LiDAR sensors, and infrared sensors are aligned and spatiotemporally registered. This is a crucial step to ensure data consistency in both time and space. Only when the data is synchronized and within the same reference frame can subsequent effective fusion and analysis be performed. Next, the confidence level of each sensor's data is tested. Confidence level is a quantitative indicator reflecting the accuracy and reliability of sensor data. If a type of data is affected by environmental interference (e.g., lighting conditions, atmospheric conditions) or sensor malfunction, its confidence level will decrease accordingly.

[0147] Optionally, when the confidence level of all sensor data exceeds a preset threshold (e.g., 90%), the weighted fusion calculation stage begins. A specific weighting formula is used here: μ = 0.6H + 0.3CV + 0.1IR, where H is the texture entropy value (i.e., the complexity of the road surface texture features extracted by the visual sensor), CV is the reflection intensity variation coefficient (i.e., the surface characteristics of the lidar point cloud data), and IR is the thermal radiation value (i.e., the road surface temperature or humidity-related attributes detected by the infrared sensor). The above three feature values ​​are weighted and averaged according to preset weights to obtain a comprehensive road adhesion coefficient μ value. This μ value can more comprehensively and accurately reflect the current road surface state, providing crucial information input for subsequent vehicle control strategies. Based on the calculated μ value, the specific state of the current road surface (e.g., dry, slippery, or icy) can be determined, and optimal coasting energy recovery control decisions can be made accordingly, such as adjusting the electric braking recovery torque or activating hydraulic braking assist, to ensure that the vehicle maximizes energy recovery efficiency while maintaining stability.

[0148] Step S304, output μ value.

[0149] In this embodiment, a second-order Kalman filter algorithm is used to further process the fused data to output the final road adhesion coefficient μ value. Kalman filtering is a highly efficient dynamic data fusion algorithm that combines sensor measurements and system predictions. Through recursive prediction-correction steps, it can provide continuous and smooth μ value estimates, maintaining high estimation accuracy even when sensor data is noisy or incomplete.

[0150] In steps S301 to S304 above, starting with data synchronization, the time base of multiple sensors is ensured to be consistent, avoiding data misalignment caused by time differences. Next, feature extraction technology is used to obtain the texture and geometric features of the road surface from visual images and LiDAR point clouds, respectively. Then, in the multi-source data fusion stage, sensor weights are dynamically adjusted to adapt to data reliability under different road conditions. Finally, the fused data is smoothed using a Kalman filter algorithm, outputting a precise μ value to guide the intelligent coasting energy recovery control strategy, which can significantly improve the driving safety and energy efficiency of new energy vehicles in complex environments.

[0151] The energy recovery control process of the vehicle will be described in more detail next.

[0152] The coasting energy recovery control strategy is subdivided into three stages based on changes in the road surface adhesion coefficient (μ value): a coasting energy recovery fully released stage, a coasting energy recovery limited stage, and a coasting energy recovery disabled stage. Adopting different energy recovery modes at different times ensures effective energy recovery under various road conditions while improving vehicle stability and safety.

[0153] Triggering conditions for the full release of coasting energy: When the road surface adhesion coefficient μ reaches or exceeds 0.7, this generally indicates excellent road conditions and sufficient friction, suitable for efficient energy recovery. Control parameters for the full release of coasting energy: Maximum electric braking recovery capacity, i.e., -0.2g (g is the acceleration due to gravity, approximately 9.8 m / s²), which means that during braking, the vehicle will experience a deceleration equivalent to 0.2g. In this stage, the vehicle can fully utilize electric braking for efficient energy recovery without the need for additional hydraulic braking intervention.

[0154] Triggering conditions for the coasting energy recovery limiting phase: When the road surface adhesion coefficient μ is between 0.5 and 0.7, it indicates that road conditions are becoming more complex, and friction may be insufficient to support the full-range energy recovery mode, requiring limitation to prevent vehicle instability. Control parameters for the coasting energy recovery limiting phase: Maximum electric braking recovery capacity = -0.2g (μ-0.5) / 0.2, with the insufficient portion compensated by hydraulic pressure. This means that as the μ value increases from 0.5 to 0.7, the electric braking regenerative braking capacity will smoothly increase from 0 to -0.2g. In this stage, when the electric braking cannot provide sufficient braking force, the system will automatically activate hydraulic braking to compensate for the insufficient portion, ensuring the total braking force of the vehicle while recovering as much energy as possible.

[0155] Triggering conditions for the coasting energy recovery disabled phase: When the road surface adhesion coefficient μ value is below 0.5, which is usually a low-adhesion road surface covered with wet, slippery, or icy and snowy conditions, the use of electric braking may cause the vehicle to skid or lose control. In this case, the use of electric braking to recover energy should be avoided. Control parameters for the coasting energy recovery disabled phase: Maximum electric braking recovery capacity = 0, with the insufficient portion compensated by hydraulic pressure. That is, the electric braking energy recovery function is disabled, and the hydraulic braking system completely takes over the vehicle's braking needs to ensure driving safety on low-adhesion road surfaces. In this phase, although energy recovery is not possible, safety and stability take precedence over energy efficiency.

[0156] The above control strategy fully considers the impact of road surface adhesion coefficient on vehicle braking force. By adjusting the maximum recovery capacity of electric braking in stages and dynamically, a balance between energy recovery efficiency and driving safety is achieved. On high-adhesion roads, electric braking is fully utilized for energy recovery; in complex road conditions, the use of electric braking is flexibly limited, and hydraulic braking compensation is used to ensure vehicle driving stability; while on low-adhesion roads, electric braking is completely disabled to avoid any risk that may lead to vehicle instability.

[0157] The fault handling method for the sensor in the embodiments of this application will be further described below.

[0158] In intelligent coasting energy recovery control based on multimodal sensor fusion, considering the possibility of sensor malfunction or failure, corresponding fault handling strategies are designed to ensure that basic functionality and safety are maintained even when some sensors fail to function properly. The following describes the countermeasures for visual sensor failure and dual-sensor failure.

[0159] When visual sensors (including RGB cameras and infrared sensors) malfunction, the system automatically switches to backup mode to minimize the impact of information loss on μ value estimation and energy recovery control. In backup mode, it relies more heavily on data from LiDAR and wheel speed sensors. For example, the weight of LiDAR is increased to 0.8 in backup mode because LiDAR provides high-precision 3D point cloud data, which is significantly advantageous for detecting minute road surface structures and changes (such as potholes, cracks, and water accumulation). In the absence of visual information, increasing the weight of LiDAR helps to more accurately estimate the road adhesion coefficient μ, thereby guiding adjustments to the energy recovery strategy.

[0160] Optionally, wheel speed sensors assist the lidar in correcting the slip ratio (i.e., the difference between wheel slip and the actual vehicle motion). The correction amount ΔS is calculated using the formula ΔS = 0.05·a_y, where a_y represents the lateral acceleration. When the visual sensor fails, by monitoring changes in lateral acceleration and wheel speed, the wheel speed sensor can help detect potential lateral slip of the vehicle, thereby adjusting the control strategy and preventing the vehicle from losing control during coasting energy recovery.

[0161] When two sensors are detected to fail simultaneously (e.g., a vision sensor and a lidar sensor fail at the same time), a prediction mode based on historical data will be activated. This mode uses sensor data from before the vehicle failed to perform trend analysis and prediction, and estimates the current road adhesion coefficient μ and vehicle status.

[0162] Optionally, the prediction mode has a time limit, with a maximum prediction duration of 3 seconds. This is because after a certain period, the relevance and accuracy of historical data may significantly decrease, especially when road conditions change rapidly. The 3-second time window ensures that the system can still maintain a certain level of normal operation even if the sensor fails in the short term.

[0163] Optionally, in the event of dual sensor failure, a warning light on the instrument panel can be automatically illuminated to alert the driver. This not only indicates that the energy recovery control system has entered a degraded operation mode, but also alerts the driver to potential vehicle performance degradation and suggests appropriate driving measures to ensure driving safety.

[0164] In this embodiment, the fault handling strategy is an important component of the multimodal sensor fusion design, aiming to improve the robustness and safety of the system. By dynamically adjusting sensor weights, using wheel speed sensors for auxiliary correction, and enabling prediction modes based on historical data and system degradation prompts when dual sensors fail, the challenges posed by sensor failures can be effectively addressed, ensuring that the vehicle's coasting energy recovery control maintains basic stability and safety under complex road conditions.

[0165] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0166] According to an embodiment of this application, an embodiment of a vehicle off-line detection device is provided. It should be noted that the device can be used to perform the above-described vehicle energy recovery method.

[0167] Figure 5 This is a schematic diagram of a vehicle energy recovery device according to an embodiment of this application, such as... Figure 5 As shown, the energy recovery device 500 of the vehicle includes: an acquisition unit 501, a first determination unit 502, a second determination unit 503, and a control unit 504.

[0168] The acquisition unit 501 is used to acquire sensing data collected by multiple sensors, wherein the sensing data is used to characterize the road surface condition of the road currently being traveled by the vehicle.

[0169] The first determining unit 502 is used to determine the road surface adhesion coefficient based on sensing data, wherein the road surface adhesion coefficient is used to characterize the friction between the vehicle and the road surface.

[0170] The second determining unit 503 is used to determine the energy recovery mode of the vehicle based on the road surface adhesion coefficient, wherein the energy recovery mode is used to characterize the rules used for energy recovery of the vehicle.

[0171] Control unit 504 is used to recover the energy generated by the vehicle into the vehicle's battery during the process of controlling the vehicle to drive in energy recovery mode.

[0172] Optionally, the first determining unit 502 is further configured to: determine the weighting coefficients corresponding to the sensing data collected by multiple sensors based on the road surface condition, wherein the road surface condition is a dry road surface condition or a wet and slippery road surface condition; and calculate the weighted sensing data collected by multiple sensors based on the weighting coefficients to obtain the road surface adhesion coefficient.

[0173] Optionally, the first determining unit 502 is further configured to: in response to a dry road surface condition, determine a first weighting coefficient corresponding to the first sensing data collected by the vision sensor, a second weighting coefficient corresponding to the second sensing data collected by the lidar sensor, and a third weighting coefficient corresponding to the third sensing data collected by the wheel speed sensor, wherein the first weighting coefficient is greater than the second weighting coefficient, and the second weighting coefficient is greater than the third weighting coefficient; in response to a wet road surface condition, determine a fourth weighting coefficient corresponding to the first sensing data collected by the vision sensor, a fifth weighting coefficient corresponding to the second sensing data collected by the lidar sensor, and a sixth weighting coefficient corresponding to the third sensing data collected by the wheel speed sensor, wherein the fifth weighting coefficient is greater than the fourth weighting coefficient, and the fourth weighting coefficient is greater than the sixth weighting coefficient.

[0174] Optionally, the second determining unit 503 is further configured to: determine an energy recovery mode matching the road surface adhesion coefficient from a mapping table based on the road surface adhesion coefficient, wherein the mapping table includes a mapping relationship between road surface adhesion coefficient samples and energy recovery mode samples.

[0175] Optionally, the control unit 504 is further configured to: control the vehicle to travel at a first deceleration in response to the energy recovery mode being a first energy recovery mode, wherein the first energy recovery mode is used to characterize energy recovery using the vehicle's electric braking; control the vehicle to travel at a second deceleration in response to the energy recovery mode being a second energy recovery mode, wherein the second energy recovery mode is used to characterize energy recovery using the vehicle's electric braking and the vehicle's hydraulic braking, and the second deceleration is greater than the first deceleration; and control the vehicle to travel at a third deceleration in response to the energy recovery mode being a third energy recovery mode, wherein the third energy recovery mode is used to characterize prohibiting the vehicle from performing energy recovery, and the third deceleration is greater than the second deceleration.

[0176] Optionally, the first determining unit 502 is further configured to: in response to the vision sensor being in a fault mode, determine the road surface adhesion coefficient of the road currently being traveled by the vehicle based on the sensing data of the lidar sensor and the sensing data of the wheel speed sensor.

[0177] Optionally, the first determining unit 502 is further configured to: in response to both the visual sensor and the lidar sensor being in a fault mode, determine the road surface adhesion coefficient of the road currently being traveled by the vehicle based on the historical road surface adhesion coefficient of the vehicle within a historical time period.

[0178] In the vehicle energy recovery device described in this application, the sensing data collected by multiple different types of sensors of the vehicle determines the road surface adhesion coefficient of the road on which the vehicle is currently traveling. Then, based on the road surface adhesion coefficient, a matching energy recovery mode is adopted to recover the vehicle's energy. This can improve the energy recovery efficiency and ensure the driving safety of the vehicle when recovering energy, thereby solving the technical problem of low vehicle driving safety when recovering energy in related technologies.

[0179] Embodiments of this application also provide a vehicle, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods described in various embodiments of this application when it runs.

[0180] Embodiments of this application also provide a computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of this application.

[0181] Embodiments of this application also provide a computer program product, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0182] Embodiments of this application also provide a computer program product, including a non-volatile computer-readable storage medium for storing a computer program that, when executed by a processor, implements the methods in various embodiments of this application.

[0183] Embodiments of this application also provide a computer program that, when executed by a processor, implements the methods described in the various embodiments of this application.

[0184] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0185] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0186] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0187] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0188] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0189] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for recovering vehicle energy, characterized in that, The vehicle is equipped with multiple sensors of different types, and the method includes: Acquire sensing data collected by the multiple sensors, wherein the sensing data is used to characterize the road surface condition of the road currently being traveled by the vehicle; Based on the sensed data, the road surface adhesion coefficient is determined, wherein the road surface adhesion coefficient is used to characterize the friction between the vehicle and the road surface. Based on the road surface adhesion coefficient, the energy recovery mode of the vehicle is determined, wherein the energy recovery mode is used to characterize the rules used for energy recovery of the vehicle; During the operation of the vehicle in accordance with the energy recovery mode, the energy generated by the vehicle is recovered into the vehicle's battery.

2. The method according to claim 1, characterized in that, Based on the sensed data, the road surface adhesion coefficient of the road is determined, including: Based on the road surface condition, weighting coefficients are determined corresponding to the sensing data collected by multiple sensors, wherein the road surface condition is either a dry road surface condition or a wet and slippery road surface condition. Based on the weighting coefficients, the sensing data collected by the multiple sensors are weighted and calculated to obtain the road surface adhesion coefficient.

3. The method according to claim 2, characterized in that, The plurality of sensors include at least a vision sensor, a lidar sensor, and a wheel speed sensor. Based on the road surface condition, weighting coefficients are determined corresponding to the sensing data collected by the plurality of sensors, including: In response to the road surface condition being the dry road surface condition, a first weighting coefficient corresponding to the first sensing data collected by the vision sensor, a second weighting coefficient corresponding to the second sensing data collected by the lidar sensor, and a third weighting coefficient corresponding to the third sensing data collected by the wheel speed sensor are determined, wherein the first weighting coefficient is greater than the second weighting coefficient, and the second weighting coefficient is greater than the third weighting coefficient. In response to the road surface condition being the slippery road surface condition, a fourth weighting coefficient is determined for the first sensing data collected by the vision sensor, a fifth weighting coefficient is determined for the second sensing data collected by the lidar sensor, and a sixth weighting coefficient is determined for the third sensing data collected by the wheel speed sensor, wherein the fifth weighting coefficient is greater than the fourth weighting coefficient, and the fourth weighting coefficient is greater than the sixth weighting coefficient.

4. The method according to claim 1, characterized in that, Based on the road surface adhesion coefficient, the energy recovery mode of the vehicle is determined, including: Based on the road surface adhesion coefficient, the energy recovery mode that matches the road surface adhesion coefficient is determined from the mapping table, wherein the mapping table includes the mapping relationship between road surface adhesion coefficient samples and energy recovery mode samples.

5. The method according to claim 3, characterized in that, Controlling the vehicle's movement according to the energy recovery mode includes: In response to the energy recovery mode being a first energy recovery mode, the vehicle is controlled to travel at a first deceleration, wherein the first energy recovery mode is used to characterize energy recovery using the vehicle's electric braking. In response to the energy recovery mode being a second energy recovery mode, the vehicle is controlled to travel at a second deceleration, wherein the second energy recovery mode is used to characterize energy recovery using the vehicle's electric braking and hydraulic braking, and the second deceleration is greater than the first deceleration; In response to the energy recovery mode being a third energy recovery mode, the vehicle is controlled to travel at a third deceleration, wherein the third energy recovery mode is used to indicate that the vehicle is prohibited from performing energy recovery, and the third deceleration is greater than the second deceleration.

6. The method according to claim 1, characterized in that, Based on the sensed data, the road surface adhesion coefficient of the road is determined, including: In response to a failure mode of the vision sensor, the road surface adhesion coefficient of the road currently being traveled by the vehicle is determined based on sensing data from the lidar sensor and wheel speed sensor.

7. The method according to claim 1, characterized in that, Based on the sensed data, the road surface adhesion coefficient of the road is determined, including: In response to the fact that both the vision sensor and the lidar sensor are in fault mode, the road surface adhesion coefficient of the road on which the vehicle is currently traveling is determined based on the historical road surface adhesion coefficient of the vehicle within a historical time period.

8. The method according to any one of claims 1 to 7, characterized in that, The multiple sensors synchronously acquire the sensing data through a time synchronization protocol.

9. A vehicle, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the method according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the storage medium is located to perform the method according to any one of claims 1 to 8.

Images