CN-121980946-A - Method for measuring and calculating human driving task load by using calculation power consumption of automatic driving vehicle

CN121980946ACN 121980946 ACN121980946 ACN 121980946ACN-121980946-A

Abstract

The invention relates to a method for measuring and calculating human driving task load by utilizing calculation power consumption of an automatic driving vehicle, which comprises the following steps of S1, discretizing and defining a scene, S2, generating a human driver comprehensive driving task load quantized value corresponding to each test road section unit, S3, reading and recording hardware performance data of a domain controller in the same divided test road section unit in real time when the road section driving task is processed, cleaning and averaging the acquired data, S4, constructing a statistical mapping model for describing the relation between the human driving task load and the automatic driving calculation power consumption, S5, obtaining road traffic scene parameters of any road section which is not determined with the human driving load, and outputting a measuring and calculating value of the human driver driving task load of the road section. According to the invention, the driving load level of the human driver under the corresponding scene is reversely deduced through objective and easily available automatic driving calculation force data.

Inventors

JIANG ZEHAO
YE MENG
CHENG YAOJIE

Assignees

华中科技大学

Dates

Publication Date: 20260505
Application Date: 20260126

Claims (10)

1. A method for measuring and calculating the driving task load of a human driver by using the calculation power consumption of an automatic driving vehicle, comprising the following steps: S1, discretizing and defining a scene, dividing a target city road network into a plurality of representative test road section units according to fixed distance intervals or based on key traffic characteristic points, taking each unit as an independent road traffic scene sample, and recording corresponding scene parameters; S2, in the divided test road section units, organizing a plurality of human drivers to drive a vehicle equipped with a data acquisition system to drive an actual road, synchronously acquiring objective indexes capable of reflecting the driving task load of the vehicle, fusing and standardizing the acquired multi-source data, and generating a comprehensive driving task load quantification value of the human drivers corresponding to each test road section unit; S3, in the same divided test road section units, repeatedly running the automatic driving vehicle on the ring test platform for a plurality of times, reading and recording hardware performance data of the domain controller when the road section driving task is processed in real time, and cleaning and averaging the acquired data to obtain a comprehensive power consumption quantized value of the automatic driving vehicle corresponding to each test road section unit; S4, based on the corresponding data obtained in the steps S2-S3, combining the road traffic scene parameter vector of the corresponding test road section unit recorded in the step S1 as S, and constructing a statistical mapping model for describing the relation between the human driving task load and the automatic driving calculation power consumption; and S5, for any new road scene of which the human driving load is not determined, acquiring road traffic scene parameters S' of the road section, and outputting a measuring and calculating value of the human driver driving task load of the road section.
2. The method for measuring and calculating the driving task load of the human driver by using the calculation power consumption of the automatic driving vehicle according to claim 1, wherein in the step S1, the scene parameters include road type, number of lanes, curvature, gradient, traffic sign line density, and dynamically collected traffic flow, average speed, weather condition and illumination condition.
3. The method for measuring and calculating the driving task load of the human driver by using the calculation power consumption of the automatic driving vehicle according to claim 2, wherein the fixed distance interval is 1km, the road type comprises a expressway/main road, and the key traffic feature point is an intersection, a ramp and a curve starting and ending point.
4. The method for measuring and calculating the driving task load of the human driver by using the calculation power consumption of the automatic driving vehicle according to claim 1, wherein in the step S2, the objective index comprises a gaze point distribution, a saccade speed, a pupil diameter change rate collected based on an eye movement meter, a heart rate variability collected based on a physiological meter, and skin electric activity, and a standardized steering entropy and an accelerator pedal entropy collected based on a vehicle bus.
5. The method for measuring and calculating the driving task load of the human driver by using the calculation power consumption of the automatic driving vehicle according to claim 1, wherein in the step S3, the hardware performance data includes average occupancy rate of the central processing unit and the graphics processing unit, power consumption, time consumption of processing single frame data by the perception module, time consumption of calculating single period by the planning decision module and total power consumption of the system.
6. The method for measuring and calculating the driving task load of the human driver by using the calculated power consumption of the automatic driving vehicle according to claim 5, wherein the high-simulation automatic driving system can be adopted to carry out repeated driving tests on the ring test platform in the same divided test road section units.
7. The method for measuring and calculating the load of the driving task of the human driver by using the calculated power consumption of the automatic driving vehicle according to claim 1, wherein in the step S4, the statistical mapping model is used for representing the indicating effect of the change of the calculated power consumption of the automatic driving on the load level of the driving task under the given traffic scene condition, and the expression is as follows: L_human = f(C_av, S) + ε Wherein f is a mapping function to be determined, and ε is a model error.
8. The method for measuring and calculating the task load of the human driver by using the calculation power consumption of the automatic driving vehicle according to claim 1, wherein in the step S5, the method for outputting the measured value of the task load of the human driver on the road section is to obtain the road traffic scene parameter S 'of the road section, collect the corresponding calculation power consumption data c_av' by the automatic driving vehicle or the simulation system driving on the road section, and input the c_av 'and S' into the correlation model f constructed and verified in the step S4 to obtain the measured value.
9. An apparatus for measuring and calculating a driving task load of a human driver using a calculation power consumption of an automatically driven vehicle, comprising: The scene discretization and definition module is used for discretizing and defining a scene, dividing a target city road network into a plurality of representative test road section units according to fixed distance intervals or based on key traffic characteristic points, taking each unit as an independent road traffic scene sample, and recording corresponding scene parameters; The human driving task load data acquisition module is used for organizing a plurality of human drivers to drive vehicles equipped with the data acquisition system to drive actual roads in the divided test road section units, synchronously acquiring objective indexes capable of reflecting the driving task load of the vehicles, fusing and standardizing the acquired multi-source data, and generating a comprehensive driving task load quantification value of the human drivers corresponding to each test road section unit; The automatic driving calculation power consumption data acquisition module is used for repeatedly performing running tests on the loop test platform for a plurality of times by using an automatic driving vehicle or a highly-simulated automatic driving system in the same divided test road section units, reading and recording hardware performance data of a domain controller when the domain controller processes a driving task of the road section in real time, and cleaning and averaging the acquired data to obtain a comprehensive calculation power consumption quantized value of the automatic driving vehicle corresponding to each test road section unit; The method comprises the steps of constructing a scene-calculation force-load association model module, wherein the scene-calculation force-load association model module is used for constructing a statistical mapping model for describing the relation between the human driving task load and the automatic driving calculation force consumption by combining the recorded road traffic scene parameter vector of the corresponding test road section unit as S based on the obtained corresponding data; The model application and human load measuring and calculating module is used for obtaining road traffic scene parameters S' of any section of a new road scene of which the human driving load is not determined and outputting measuring and calculating values of the human driver driving task load of the section of the road.
10. The electronic equipment is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory are communicated with each other through the communication bus; the memory has stored therein a computer program which, when executed by the processor, causes the processor to perform the steps of the method of any of claims 1-8.

Description

Method for measuring and calculating human driving task load by using calculation power consumption of automatic driving vehicle Technical Field The invention relates to the crossing fields of intelligent traffic, human engineering and automatic driving technologies, in particular to a method for measuring and calculating the driving task load of a human driver by using the calculated power consumption of an automatic driving vehicle. Background The driving task load is a core human factor parameter for evaluating driving safety, road design and optimization and human-computer interaction interface. The accurate measurement of the driving task load has important significance for preventing traffic accidents, improving road traffic efficiency, designing an advanced driving auxiliary system and realizing smooth man-machine co-driving. Existing assessment of driving scene complexity typically relies on the driver wearing touch sensors (e.g., brain electrode caps, eye-tracker) or post-hoc subjective scoring. However, the contact device has the problems of complex wearing, easy interference of signals by limb actions and incapability of being normally deployed in mass production vehicles in actual road tests, and meanwhile, the subjective score lacks of real-time performance, and is difficult to correspond to specific instantaneous road conditions. There is a need for a detection means that uses the onboard hardware of the vehicle for non-contact, continuous quantification. With development of automatic driving technology, in the running process of an automatic driving vehicle, in order to cope with different traffic scenes, calculation tasks such as environment perception, behavior decision, track planning, vehicle control and the like need to be continuously completed. The core computing unit is used for processing driving scenes with different complexity, and the generated computing power consumption is an internal objective variable which can be continuously monitored. The computational effort expended by an autopilot system to cope with a particular traffic scenario is conceptually similar to the cognitive and operational resources that human drivers need to invest to accomplish the same driving task, i.e., driving task load. However, a method for quantitatively estimating the load of a human driving task by using the calculation power consumption data of an automatic driving vehicle has not been established in the prior art. Disclosure of Invention The technical problem to be solved by the invention is to provide a method for measuring and calculating the driving task load of a human driver by using the calculation power consumption of an automatic driving vehicle, which realizes objective, continuous and non-invasive measurement of the driving task load. The technical scheme adopted by the invention for solving the technical problems is that a method for measuring and calculating the driving task load of a human driver by utilizing the calculation power consumption of an automatic driving vehicle is constructed, and comprises the following steps: S1, discretizing and defining a scene, dividing a target city road network into a plurality of representative test road section units according to fixed distance intervals or based on key traffic characteristic points, taking each unit as an independent road traffic scene sample, and recording corresponding scene parameters; S2, in the divided test road section units, organizing a plurality of human drivers to drive a vehicle equipped with a data acquisition system to drive an actual road, synchronously acquiring objective indexes capable of reflecting the driving task load of the vehicle, fusing and standardizing the acquired multi-source data, and generating a comprehensive driving task load quantification value of the human drivers corresponding to each test road section unit; S3, in the same divided test road section units, repeatedly running the automatic driving vehicle on the ring test platform for a plurality of times, reading and recording hardware performance data of the domain controller when the road section driving task is processed in real time, and cleaning and averaging the acquired data to obtain a comprehensive power consumption quantized value of the automatic driving vehicle corresponding to each test road section unit; S4, based on the corresponding data obtained in the steps S2-S3, combining the road traffic scene parameter vector of the corresponding test road section unit recorded in the step S1 as S, and constructing a statistical mapping model for describing the relation between the human driving task load and the automatic driving calculation power consumption; and S5, for any new road scene of which the human driving load is not determined, acquiring road traffic scene parameters S' of the road section, and outputting a measuring and calculating value of the human driver driving task load of the road section. According to the above sch