A vehicle energy-saving method

Abstract

The present invention relates to a vehicle energy-saving method, which includes the following steps: The vehicle establishes communication with the external environment; The vehicle sends its relevant information to the roadside unit; The vehicle performs perception operations independently; The roadside unit sends the perception results to the vehicle in real time; The vehicle and the roadside unit perform information fusion; The vehicle and the roadside information system respectively evaluate whether the current environment meets the takeover conditions. If so, the vehicle and the roadside unit obtain the satisfaction degree and conclusion of the takeover conditions and request to perform computing power takeover; The vehicle requests to perform computing power takeover; The roadside device evaluates the computing power resources and determines whether to pass the request; The vehicle reduces its computing power and is in a low computing power mode; Determine whether the vehicle currently meets the requirements for autonomous driving. If so, maintain the low computing power mode. If not, the vehicle restores its computing power; The vehicle restores its computing power and confirms with the roadside information system. Compared with the prior art, the present invention has the advantages of reducing vehicle energy consumption and energy conservation and emission reduction, etc.

Description

Technical Field

[0001] The present invention relates to the field of energy consumption in autonomous driving, and particularly to a vehicle energy-saving method. Background Art

[0002] With the development of autonomous driving technology, more and more tasks that originally needed to be completed by humans in driving behaviors have gradually been completed by machines. Among them, the vehicle perception technology in the autonomous driving system takes over the work of observing the road by the human eye. The driver's perception is the cognition of the environment or things formed through observation by various organs and processed and judged by the brain. Correspondingly, in the field of autonomous driving, the process in which the central controller receives the information from each sensor and processes it through a series of fusion algorithms is the perception technology. The sensor is the perception organ of the vehicle. Therefore, in recent years, the perception technology has continuously undergone technological innovation following the development of sensors.

[0003] The key technologies of autonomous driving are mainly in three aspects: perception, planning, and control. Among them, the perception system takes the data captured by multiple sensors and the information of the high-definition map as input, and through a series of calculations and processes, estimates the state of the vehicle and realizes the precise perception of the vehicle's surrounding environment, and can provide rich information for the downstream decision-making system module.

[0004] Currently, traditional chip manufacturers such as NVIDIA, Intel, and Qualcomm, relying on their own integration and R & D capabilities, are committed to establishing autonomous driving chip production capabilities through self-research or acquisition, and have successively launched in-vehicle computing platforms that can achieve high-level autonomous driving above L3. The use of these intelligent sensors and deep learning high-performance computing platforms in vehicles consumes a large amount of electricity and needs to work in real time. Currently, this kind of vehicle intelligent driving is mainly single-vehicle intelligence, relying on the vehicle's own battery for power supply, with relatively high power consumption.

[0005] While high computing power brings high performance, it also causes the vehicle's power consumption to increase rapidly. And due to reasons such as safety monitoring, all sensors and the central computer are working at full load at all times when the vehicle is working. For example, when waiting for a red light at an intersection or during a parking wait, traditional fuel vehicles can reduce operation by turning off the engine through the automatic start-stop function. Although intelligent vehicles can turn off the electric drive, there is currently no good way to reduce the computing power and save power consumption in specific scenarios for the intelligent part. Therefore, the present invention proposes a method for reducing vehicle energy consumption by reducing the computing power of intelligent sensors and intelligent controllers in autonomous driving and assisted driving vehicles. Summary of the Invention

[0006] The purpose of the present invention is to provide a vehicle energy-saving method to overcome the defects existing in the above-mentioned prior art.

[0007] The object of the present invention can be achieved by the following technical solutions:

[0008] A vehicle energy-saving method, the method comprising the following steps:

[0009] Step 1: The vehicle establishes communication with the external environment;

[0010] Step 2: The vehicle sends its relevant information to the roadside unit in the external environment;

[0011] Step 3: The vehicle performs perception operations autonomously;

[0012] Step 4: The roadside unit perceives the surrounding environment in real time and sends the perception result to the vehicle;

[0013] Step 5: The vehicle and the roadside perception device of the roadside unit confirm the information and fuse the information;

[0014] Step 6: The roadside information systems of the vehicle and the roadside unit respectively evaluate whether the current environment meets the takeover condition for the roadside perception to take over the vehicle perception. If so, the vehicle and the roadside obtain the satisfaction degree and conclusion of the takeover condition and execute Step 7. If not, return to Step 2;

[0015] Step 7: The vehicle requests to take over the computing power;

[0016] Step 8: The roadside unit evaluates the computing power resources and determines whether to pass the request. If so, execute Step 9. If not, return to Step 2;

[0017] Step 9: The vehicle reduces the computing power and is in the low computing power mode;

[0018] Step 10: Determine whether the current low computing power state of the vehicle meets the autonomous driving requirements. If so, maintain the low computing power mode. If not, execute Step 11;

[0019] Step 11: The vehicle restores the computing power, is in the stage of using the autonomous computing power, confirms with the roadside information system, and returns to Step 2.

[0020] In the said Step 1, the roadside unit includes a roadside perception device and a roadside information system.

[0021] The said roadside perception device includes a lidar, a camera and a millimeter wave radar.

[0022] The said roadside information system is a roadside edge computing unit, that is, a computer on the roadside, used to receive the data transmitted by the roadside perception device.

[0023] The said relevant information includes the current state of the vehicle, the autonomous driving level of the vehicle, and whether the autonomous driving is activated.

[0024] In step 6 described above, the takeover conditions include that the vehicle has no current faults, the roadside perception device has no faults, the communication signal strength is greater than the set strength threshold, the handshake between the vehicle and the roadside is normal, and the reserved computing power of the roadside perception device is greater than the set computing power threshold.

[0025] In step 9 described above, the low computing power mode is specifically as follows:

[0026] When the vehicle enters a specific area, the roadside perception device within the specific area receives the perception results of the vehicle's surrounding environment based on V2X technology, fuses them with the relevant information sent by the vehicle to obtain the fused perception results. The roadside information system and the information integration system in the cloud send the fused perception results and the planned path within this area to the vehicle. The vehicle controls its computing power by receiving information that is not perceived and calculated by itself, that is, the roadside and the cloud inform the vehicle of the results of perception and calculation to reduce the vehicle's power consumption.

[0027] The specific areas described above include intersections, underground garages, and highways.

[0028] The information that is not perceived and calculated by the vehicle itself includes path planning, temperature control, and future congestion conditions.

[0029] In step 9 described above, the vehicle's actions to reduce computing power include turning off some cameras, turning off all cameras, turning off some millimeter-wave radars, turning off all millimeter-wave radars, turning off some lidars, turning off all lidars, turning off infrared sensors, turning off some perception and calculation algorithms, turning off all perception and calculation algorithms, turning off some perception fusion algorithms, turning off all perception fusion algorithms, turning off high-precision maps, turning off planning and decision-making algorithms, reducing the camera sampling rate, reducing the millimeter-wave radar sampling rate, reducing the lidar sampling rate, turning off some AI chips in the vehicle domain controller, and reducing the working frequency of the AI chips.

[0030] Compared with the prior art, the present invention has the following advantages:

[0031] 1. The present invention forms a distributed computing system by integrating the vehicle, the cloud, and the roadside, forming a computing architecture of the cloud, the edge, and the vehicle end, and optimizing the computing power allocation of the cloud, the edge, and the vehicle end based on an optimization algorithm to achieve the optimal energy consumption of the entire distributed computing system;

[0032] 2. The vehicle of the present invention has an LTE-V2X communication function to achieve remote driving. When the vehicle is taken over by a remote driver, it issues a requirement to reduce the automatic driving computing power of the vehicle itself. After obtaining a handshake, it reduces the computing power of the vehicle or turns off some automatic driving functions to achieve power consumption reduction of the vehicle;

[0033] 3. The present invention can reduce the computing power of a vehicle in a specific scenario, thereby saving power consumption and achieving the technical effects of energy conservation and emission reduction. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a flowchart of the method of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0035] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0036] Embodiment

[0037] The present invention provides a vehicle energy-saving method, which includes the following steps:

[0038] Step 1: The vehicle establishes communication with the external environment;

[0039] Step 2: The vehicle sends its relevant information to the roadside unit (road end) in the external environment;

[0040] Step 3: The vehicle performs perception operations independently;

[0041] Step 4: The roadside unit perceives the surrounding environment in real time and sends the perception result to the vehicle;

[0042] Step 5: The vehicle and the roadside perception device confirm the information, confirm whether the information is credible according to the information sent and received by both parties, and fuse the information;

[0043] Step 6: The vehicle and the roadside information system respectively evaluate whether the current environment meets the takeover condition for the roadside perception to take over the vehicle perception. If so, the vehicle and the roadside obtain the satisfaction degree and conclusion of the takeover condition and execute Step 7. If not, return to Step 2;

[0044] Step 7: One party in the vehicle or the roadside information system requests to perform computing power takeover;

[0045] Step 8: The roadside device evaluates the computing power resources and determines whether the request is passed. If so, execute Step 9. If not, return to Step 2;

[0046] Step 9: The vehicle reduces its computing power and is in a low computing power mode;

[0047] Step 10: Determine whether the current low computing power state of the vehicle meets the requirements of autonomous driving. If so, maintain the low computing power mode. If not, execute Step 11;

[0048] Step 11: The vehicle restores its computing power, enters the stage of using its own computing power, confirms with the roadside information system, and returns to Step 2.

[0049] In step 1, the relevant information of the vehicle includes the current state of the vehicle, the vehicle's autonomous driving level, and whether autonomous driving is activated.

[0050] In step 9, the vehicle controls its computing power by receiving information that is not self-perceived and calculated by the vehicle itself, that is, the roadside and the cloud tell the vehicle the results of their perception and calculation to save the power consumption of the vehicle itself. The information that is not self-perceived and calculated by the vehicle itself is the result of the roadside and the cloud's perception and calculation, including path planning, temperature control, and future traffic congestion.

[0051] The roadside perception devices include lidar, cameras, millimeter-wave radars, etc. The roadside information system is the roadside edge computing unit, that is, the computer on the roadside. The roadside perception devices transmit data to the roadside information system.

[0052] The control system of the vehicle is the vehicle's computer, which includes a CPU, memory, and GPU. TOPS is used to represent the computing power.

[0053] The specific way for the vehicle and external devices (roadside and cloud) to coordinate and communicate about the computing power is as follows:

[0054] When the vehicle enters a specific area, the roadside perception system and the cloud information integration system in the specific area provide the vehicle with all the perception results and information such as the planned path in the area. The specific area includes intersections, underground garages, and highways. The vehicle sends a request to reduce the computing power. The roadside perception device receives the perception results around the vehicle based on V2X technology and generates a fused perception result. The fused perception result includes the result of decision-making and planning to replace the in-depth calculation of the vehicle itself. The roadside and the cloud define the tasks to be taken over according to the state of the vehicle, the available state of its own resources, and the level of computing power requested by the vehicle. According to the task definition and the interaction information of the vehicle, the tasks include the external device replacing the vehicle for visual perception, radar perception, or replacing the vehicle for path planning. The interaction information is the process of information passing back and forth, and reaches an agreement with the vehicle's requirements for the roadside and the cloud, and conducts the transfer of computing power, transferring the calculation of the vehicle's controller to the roadside or the cloud for calculation to complete the reduction of the vehicle's computing power.

[0055] The behaviors of the vehicle to reduce the computing power include turning off some cameras, turning off all cameras, turning off some millimeter-wave radars, turning off all millimeter-wave radars, turning off some lidars, turning off all lidars, turning off infrared sensors, turning off some perception and calculation algorithms, turning off all perception and calculation algorithms, turning off some perception fusion algorithms, turning off all perception fusion algorithms, turning off high-precision maps, turning off planning and decision-making algorithms, reducing the camera sampling rate, reducing the millimeter-wave radar sampling rate, reducing the lidar sampling rate, turning off some AI chips of the vehicle domain controller, and reducing the working frequency of the AI chips.

[0056] The present invention forms a distributed computing system by integrating vehicles, the cloud, and the roadside, establishing a computing architecture for the cloud, the edge, and the vehicle. By optimizing algorithms, the computing power allocation among the cloud, the edge, and the vehicle is optimized to achieve the optimal energy consumption of the entire distributed computing system. That is, the vehicle, the roadside, and the cloud are regarded as a large computer system, and the computing power of each party is a part of the entire computer system. Through the allocation mechanism, the computing power is redistributed. Moreover, the vehicle has the LTE-V2X communication function to enable remote driving. When the vehicle is taken over by a remote driver, a request to reduce the autonomous driving computing power of the vehicle is sent. After obtaining a handshake (establishing a correct connection), the computing ability of the vehicle is reduced or some autonomous driving functions are turned off to achieve power consumption reduction of the vehicle.

[0057] As described above, the above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any staff member familiar with the technical field of the present invention can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A vehicle energy-saving method, characterized in that, the method comprises the following steps: Step 1: The vehicle establishes communication with the external environment; Step 2: The vehicle sends its relevant information to the roadside unit in the external environment; Step 3: The vehicle performs perception operations independently; Step 4: The roadside unit perceives the surrounding environment in real time and sends the perception result to the vehicle; Step 5: The vehicle and the roadside perception devices of the roadside unit confirm the information and fuse the information; Step 6: The roadside information systems of the vehicle and the roadside unit respectively evaluate whether the current environment meets the takeover conditions for the roadside perception to take over the vehicle perception. If so, the vehicle and the roadside obtain the satisfaction degree and conclusion of the takeover conditions and execute Step 7. If not, return to Step 2; Step 7: The vehicle requests to take over the computing power; Step 8: The roadside unit evaluates the computing power resources and determines whether to pass the request. If so, execute Step 9. If not, return to Step 2; Step 9: The vehicle reduces the computing power and is in the low computing power mode; Step 10: Determine whether the current low computing power state of the vehicle meets the autonomous driving requirements. If so, maintain the low computing power mode. If not, execute Step 11; Step 11: The vehicle resumes the computing power, is in the stage of using the autonomous computing power, confirms with the roadside information system, and returns to Step 2; In the said Step 9, the low computing power mode is specifically: When the vehicle enters a specific area, the roadside perception devices located in the specific area receive the perception results of the vehicle's surrounding environment based on the V2X technology, fuse them with the relevant information sent by the vehicle, obtain the fused perception results, and the roadside information system and the information integration system in the cloud send the fused perception results and the planned path in this area to the vehicle. The vehicle controls the computing power by receiving the information that is not perceived and calculated by the vehicle itself, that is, the roadside and the cloud tell the vehicle the results of the perception and calculation to reduce the power consumption of the vehicle.

2. A vehicle energy-saving method according to claim 1, characterized in that, in the said Step 1, the roadside unit includes roadside perception devices and a roadside information system.

3. A vehicle energy-saving method according to claim 2, characterized in that, the said roadside perception devices include lidar, cameras and millimeter wave radars.

4. A vehicle energy-saving method according to claim 2, characterized in that, the said roadside information system is a roadside edge computing unit, that is, a computer on the roadside, used to receive the data transmitted by the roadside perception devices.

5. A vehicle energy-saving method according to claim 1, characterized in that, the said relevant information includes the current state of the vehicle, the autonomous driving level of the vehicle, and whether the autonomous driving is activated.

6. A vehicle energy-saving method according to claim 1, characterized in that, in the said Step 6, the takeover conditions include that the vehicle has no current faults, the roadside perception devices have no faults, the communication signal strength is greater than the set strength threshold, the handshake between the vehicle and the roadside is normal, and the reserved computing power of the roadside perception devices is greater than the set computing power threshold.

7. A vehicle energy-saving method according to claim 1, characterized in that, the said specific areas include intersections, underground garages and highways.

8. A vehicle energy-saving method according to claim 1, characterized in that, the information that is not perceived and calculated by the vehicle itself includes path planning, temperature control, and future traffic congestion conditions.

9. A vehicle energy-saving method according to claim 1, characterized in that, in step 9, the behaviors of the vehicle to reduce computing power include turning off some cameras, turning off all cameras, turning off some millimeter-wave radars, turning off all millimeter-wave radars, turning off some lidars, turning off all lidars, turning off infrared sensors, turning off some perception and calculation algorithms, turning off all perception and calculation algorithms, turning off some perception fusion algorithms, turning off all perception fusion algorithms, turning off high-precision maps, turning off planning and decision-making algorithms, reducing the camera sampling rate, reducing the millimeter-wave radar sampling rate, reducing the lidar sampling rate, turning off some AI chips of the vehicle domain controller, and reducing the working frequency of the AI chips.

Images