EP-4737253-A1 - DATA PROCESSING METHOD AND APPARATUS, AND VEHICLE

EP4737253A1EP 4737253 A1EP4737253 A1EP 4737253A1EP-4737253-A1

Abstract

This application provides a data processing method and apparatus, and a vehicle, and may be applied to the field of intelligent driving. The data processing method includes: obtaining first environment information around a vehicle and first status information of the vehicle; sending the first environment information and the first status information to a cloud server; receiving first latency optimization configuration information sent by the cloud server, where the cloud server stores a mapping relationship between the first environment information, the first status information, and the first latency optimization configuration information, and the first latency optimization configuration information includes a first latency optimization parameter for a processing unit in the vehicle to process data collected by a sensor; and processing, based on the first latency optimization parameter, first data collected by the sensor. Embodiments of this application may be applied to an intelligent vehicle or a new energy vehicle, to help reduce an end-to-end latency, thereby helping ensure driving safety of users.

Inventors

GUAN, Gaoyang
WANG, Qiangdong
CHEN, RUINING

Assignees

Shenzhen Yinwang Intelligent Technologies Co., Ltd.

Dates

Publication Date: 20260506
Application Date: 20240605

Claims (20)

A data processing method, comprising: obtaining first environment information around a vehicle and first status information of the vehicle; sending the first environment information and the first status information to a cloud server; receiving first latency optimization configuration information sent by the cloud server, wherein the cloud server stores a mapping relationship between the first environment information, the first status information, and the first latency optimization configuration information, and the first latency optimization configuration information comprises a first latency optimization parameter for a processing unit in the vehicle to process data collected by a sensor; and processing, based on the first latency optimization parameter, first data collected by the sensor.
The method according to claim 1, wherein the method further comprises: obtaining second environment information around the vehicle and second status information of the vehicle when a latency of processing, by the processing unit based on the first latency optimization parameter, the data collected by the sensor is greater than or equal to a latency threshold; sending the second environment information and the second status information to the cloud server; receiving second latency optimization configuration information sent by the cloud server, wherein the cloud server stores a mapping relationship between the second environment information, the second status information, and the second latency optimization configuration information, and the second latency optimization configuration information comprises a second latency optimization parameter for the processing unit to process data collected by the sensor; and processing, based on the second latency optimization parameter, second data collected by the sensor.
The method according to claim 1 or 2, wherein the processing unit comprises a primary processing unit and a secondary processing unit; and processing, based on the first latency optimization parameter, the first data collected by the sensor comprises: deploying the first latency optimization parameter on the secondary processing unit; and deploying the first latency optimization parameter on the primary processing unit when, in a first time period, a latency of processing the first data by the secondary processing unit based on the first latency optimization parameter is less than a latency of processing the first data by the primary processing unit.
The method according to claim 3, wherein before deploying the first latency optimization parameter on the primary processing unit, the method further comprises: determining that vehicle control instructions output by the secondary processing unit and the primary processing unit are the same in the first time period.
The method according to any one of claims 1 to 4, wherein the first environment information comprises environment information around the vehicle at a plurality of future moments, and the first status information comprises status information of the vehicle at the plurality of future moments; and the first latency configuration information comprises a latency optimization parameter for the processing unit to process, at the plurality of future moments, data collected by the sensor.
The method according to any one of claims 1 to 5, wherein the first environment information comprises at least one of time, a geographical location, climate, and a road condition; and/or the first status information comprises at least one of a hardware specification, a system status, and an autonomous driving status of the vehicle.
A data processing method, comprising: obtaining first environment information around a vehicle and first status information of the vehicle; and sending first latency optimization configuration information to the vehicle based on the first environment information, the first status information, and a mapping relationship, wherein the mapping relationship comprises a mapping relationship between environment information, status information, and latency optimization configuration information, and the first latency optimization configuration information comprises a first latency optimization parameter for a processing unit in the vehicle to process data collected by a sensor.
The method according to claim 7, wherein the first environment information comprises environment information around the vehicle at a plurality of future moments, and the first status information comprises status information of the vehicle at the plurality of future moments; and the first latency optimization configuration information comprises a latency optimization parameter for processing, at the plurality of future moments, data collected by the sensor.
The method according to claim 7 or 8, wherein the method further comprises: retrieving the mapping relationship based on the first environment information and the first status information and by using a hierarchical navigable small world HNSW algorithm, to obtain the first latency optimization configuration information.
The method according to any one of claims 7 to 9, wherein before obtaining the first environment information around the vehicle and the first status information of the vehicle, the method further comprises: obtaining third environment information around an autonomous vehicle and third status information of the autonomous vehicle; determining, based on a third latency optimization parameter set by an offline simulation system, the third environment information, and the third status information, a first latency of processing data collected by a sensor; updating the third latency optimization parameter based on the first latency, to obtain a fourth latency optimization parameter; and storing, in the mapping relationship, a correspondence between the third environment information, the third status information, and the fourth latency optimization parameter.
The method according to any one of claims 7 to 9, wherein before obtaining the first environment information around the vehicle and the first status information of the vehicle, the method further comprises: obtaining third environment information around an autonomous vehicle and third status information of the autonomous vehicle; extracting latency model data based on the third environment information and the third status information, wherein the latency model data comprises one or more of a quantity of threads of an application, periodic data of a thread, temporal probability distribution of thread execution, and a dependency relationship between thread data; inputting the latency model data into a latency simulator to obtain a second latency; determining, based on the second latency, a fourth latency optimization parameter that corresponds to the third environment information and the third status information; and storing, in the mapping relationship, a correspondence between the third environment information, the third status information, and the fourth latency optimization parameter.
A data processing apparatus, comprising: an obtaining unit, configured to obtain first environment information around a vehicle and first status information of the vehicle; a sending unit, configured to send the first environment information and the first status information to a cloud server; and a receiving unit, configured to receive first latency optimization configuration information sent by the cloud server, wherein the cloud server stores a mapping relationship between the first environment information, the first status information, and the first latency optimization configuration information, and the first latency optimization configuration information comprises a first latency optimization parameter for a processing unit in the vehicle to process data collected by a sensor; and the processing unit is configured to process, based on the first latency optimization parameter, first data collected by the sensor.
The apparatus according to claim 12, wherein the obtaining unit is further configured to obtain second environment information around the vehicle and second status information of the vehicle when a latency of processing, by the processing unit based on the first latency optimization parameter, the data collected by the sensor is greater than or equal to a latency threshold; the sending unit is further configured to send the second environment information and the second status information to the cloud server; the receiving unit is further configured to receive second latency optimization configuration information sent by the cloud server, wherein the cloud server stores a mapping relationship between the second environment information, the second status information, and the second latency optimization configuration information, and the second latency optimization configuration information comprises a second latency optimization parameter for the processing unit to process data collected by the sensor; and the processing unit is further configured to process, based on the second latency optimization parameter, second data collected by the sensor.
The apparatus according to claim 12 or 13, wherein the processing unit comprises a primary processing unit and a secondary processing unit, and the processing unit is configured to: deploy the first latency optimization parameter on the secondary processing unit; and deploy the first latency optimization parameter on the primary processing unit when, in a first time period, a latency of processing the first data by the secondary processing unit based on the first latency optimization parameter is less than a latency of processing the first data by the primary processing unit.
The apparatus according to claim 14, wherein the apparatus further comprises: a determining unit, configured to: before the processing unit deploys the first latency optimization parameter on the primary processing unit, determine that vehicle control instructions output by the secondary processing unit and the primary processing unit are the same in the first time period.
The apparatus according to any one of claims 12 to 15, wherein the first environment information comprises environment information around the vehicle at a plurality of future moments, and the first status information comprises status information of the vehicle at the plurality of future moments; and the first latency configuration information comprises a latency optimization parameter for the processing unit to process, at the plurality of future moments, data collected by the sensor.
The apparatus according to any one of claims 12 to 16, wherein the first environment information comprises at least one of time, a geographical location, climate, and a road condition; and/or the first status information comprises at least one of a hardware specification, a system status, and an autonomous driving status of the vehicle.
A data processing apparatus, comprising: an obtaining unit, configured to obtain first environment information around a vehicle and first status information of the vehicle; a determining unit, configured to determine first latency optimization configuration information based on the first environment information, the first status information, and a mapping relationship, wherein the mapping relationship comprises a mapping relationship between environment information, status information, and latency optimization configuration information, and the first latency optimization configuration information comprises a first latency optimization parameter for a processing unit in the vehicle to process data collected by a sensor; and a sending unit, configured to send the first latency optimization configuration information to the vehicle.
The apparatus according to claim 18, wherein the first environment information comprises environment information around the vehicle at a plurality of future moments, and the first status information comprises status information of the vehicle at the plurality of future moments; and the first latency optimization configuration information comprises a latency optimization parameter for processing, at the plurality of future moments, data collected by the sensor.
The apparatus according to claim 18 or 19, wherein the apparatus further comprises: an information retrieval unit, configured to retrieve the mapping relationship based on the first environment information and the first status information and by using a hierarchical navigable small world HNSW algorithm, to obtain the first latency optimization configuration information.

Description

This application claims priority to Chinese Patent Application No. 202310792509.3, filed with the China National Intellectual Property Administration on June 29, 2023 and entitled "DATA PROCESSING METHOD AND APPARATUS, AND VEHICLE", which is incorporated herein by reference in its entirety. TECHNICAL FIELD This application relates to the field of intelligent driving, and more specifically, to a data processing method and apparatus, and a vehicle. BACKGROUND In recent years, with rapid and vigorous development of autonomous driving artificial intelligence (artificial intelligence, AI) algorithms and technologies, numerous traditional automakers and emerging autonomous driving players have embarked on developing advanced L4 fully automated driving systems and applications. However, people are concerned about implementation of the autonomous driving technology, and one of the most concerned problems is whether the autonomous driving technology can ensure safety of other vehicles and pedestrians. A very important indicator for measuring safety is an end-to-end latency of a data flow link of the automated driving system, for example, a latency from a moment at which a sensor like a lidar or a camera collects an original signal to a moment at which a brake instruction is sent to a chassis domain controller. A low end-to-end latency (for example, less than 100 ms) can greatly improve a safety coefficient of vehicles and pedestrians, and is one of key objectives for subsequent optimization of systems and applications. Therefore, how to reduce the end-to-end latency becomes an urgent problem to be resolved. SUMMARY This application provides a data processing method and apparatus, and a vehicle, to help reduce an end-to-end latency, thereby helping ensure driving safety of users. According to a first aspect, this application provides a data processing method, where the method includes: obtaining first environment information around a vehicle and first status information of the vehicle; sending the first environment information and the first status information to a cloud server; receiving first latency optimization configuration information sent by the cloud server, where the cloud server stores a mapping relationship between the first environment information, the first status information, and the first latency optimization configuration information, and the first latency optimization configuration information includes a first latency optimization parameter for a processing unit in the vehicle to process data collected by a sensor; and processing, based on the first latency optimization parameter, first data collected by the sensor. Based on the foregoing technical solution, the cloud server can send corresponding latency optimization configuration information to the vehicle based on environment information around the vehicle, the status information of the vehicle, and a mapping relationship. In this way, the vehicle can obtain appropriate latency optimization configuration information in a current scenario from the cloud server. This helps reduce a latency of processing, by the processing unit in the vehicle, the data collected by the sensor, thereby helping reduce an end-to-end latency and helping ensure driving safety of users. The mapping relationship stored in the cloud server may be a mapping relationship between environment information, status information, and latency optimization configuration information in different scenarios. Currently, an end-to-end latency optimization configuration is basically performed offline based on engineers' experience for some common environments, and is difficult to perform adaptive dynamic adjustment for different scenarios during traveling of the vehicle. In addition, an end-to-end latency optimization configuration adjustment process is currently time-consuming. In this embodiment of this application, latency optimization configuration information for different scenarios is deployed in advance in the cloud server, so that the vehicle can obtain appropriate latency optimization configuration information in the different scenarios. In addition, a manner in which the cloud server determines appropriate latency optimization configuration information in a current scenario of the vehicle based on a mapping relationship can reduce waiting duration for the vehicle to obtain the latency optimization configuration information. In some possible implementations, the first latency optimization parameter includes one or more of a priority parameter of a task in the processing unit, a core binding parameter, and a scheduling parameter of a scheduler in an operating system. In some possible implementations, the first latency optimization configuration information includes an end-to-end latency optimization parameter, and the end-to-end latency is an end-to-end latency of a data flow link of an automated driving system in the vehicle, for example, a latency from a moment at which the sensor (