Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Driver model optimization method based on GRU

A technology of driver model and optimization method, applied in the field of driver model optimization based on GRU, can solve problems such as deviation and inability to apply driver model, and achieve the effect of reducing deviation, reducing traffic accidents and avoiding fatigue

Active Publication Date: 2021-09-07
HEFEI UNIV OF TECH
View PDF9 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, the driver model used in the prior art belongs to an older technical solution, which is more accurate for the following path formed by the straight path, but for the route with many curves, it has a larger Therefore, the traditional driver model cannot be directly applied to the actual vehicle control or driving assistance, but some upgrade and optimization are needed to improve its fit in following the road

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Driver model optimization method based on GRU
  • Driver model optimization method based on GRU
  • Driver model optimization method based on GRU

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0031] figure 1 and figure 2 A GRU-based driver model optimization method is provided, which uses pre-set road trajectory and expected vehicle speed information to establish a driver's preview following model, uses the model to determine the expected value of the driver's steering wheel angle applied to the car, and uses The expected path of the driver is obtained by fitting the two-degree-of-freedom vehicle model. Using the Gated Recurrent Unit (GRU) network, the desired path decision-making process is trained with a multi-factor weight-adjusted deep neural network, thereby reducing the deviation between the expected path and the actual path, and obtaining a good preview path following driving member model. The present invention can be applied to the field of smart cars in practical applications, and provides technical solutions for some technical problems in the fields of automatic driving, driving assistance, and airline supervision. The optimal path with high matching d...

Embodiment 2

[0143] The general steps of this embodiment adopt the content described in Embodiment 1, wherein the values ​​of the vehicle parameters used for simulation and the actual path parameters are as follows, and the driver model based on GRU optimization is obtained through the method described in Embodiment 1 The preview path of the decision, here the optimized preview path is called the comparison result of the GRU path and the actual path simulated by the driving simulator. However, it should be noted that the vehicle parameters and the actual path of the simulation can be changed arbitrarily, and this embodiment only shows the simulation experiment results under one set of parameters.

[0144] Vehicle parameter determination:

[0145] Vehicle mass m = 1301kg;

[0146] Wheelbase L=2.537m;

[0147] Front wheelbase a=1.074m;

[0148] Rear wheelbase b=1.463m;

[0149] Vehicle yaw moment of inertia I Z =1600kg·m 2 ;

[0150] Front and rear axle cornering stiffness k 1 ,k 2 ...

Embodiment 3

[0162] A GRU-based driver model optimization system is provided, which includes an expected path generation module, a driver model generation module, a two-degree-of-freedom vehicle model generation module, a GRU module and an output module.

[0163] The expected path generation module generates the actual path through simulation, or can also retrieve the actual collected path data in the training library, and transmit the expected path to the driver model generation module.

[0164] The preset program in the driver model generation module establishes the driver model based on the preview follow according to the actual path, and analyzes the lateral displacement process of the vehicle and corrects the analysis results according to the driver's hysteresis effect transfer function and lateral acceleration error feedback. Get the steering wheel angle as δ sw And output to the two-degree-of-freedom vehicle model generation module.

[0165] The two-degree-of-freedom vehicle model ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

PropertyMeasurementUnit
Wheelbaseaaaaaaaaaa
Login to View More

Abstract

A driver model optimization method based on GRU is used for performing deep neural network training of multi-factor weight adjustment on a preview path process of driver model decision so as to reduce the deviation between a preview path and an actual path, and the GRU refers to a gated cyclic unit network; the optimization method comprises the following total steps: training is carried out through the process of obtaining the side slip angle, the yaw velocity, the lateral acceleration and the like through a vehicle two-degree-of-freedom model by taking a steering wheel steering angle [delta]sw obtained on the basis of preview path data output by a gated cyclic unit network GRU to a driver preview model and a preset vehicle speed and taking the steering wheel steering angle [delta]sw optimized by the driver preview model as input to obtain various optimized training data, the training data and actual path data obtained by a driving simulator are fitted together, and finally a comparison image between the optimized GRU path curve and an actual path is obtained.

Description

technical field [0001] The invention relates to the field of calculation and simulation of the driving process, in particular to a GRU-based driver model optimization method. Background technique [0002] With the development of computers and the emergence of artificial intelligence technology, traditional car driving methods are facing tremendous changes, especially the emergence of smart driving cars and intelligent assisted driving has gradually begun to change people's driving experience. Intelligent driving is an interdisciplinary comprehensive technology, and its problems are complex. One of the problems is how to use intelligent algorithms to improve the fit between the running track and the expected track when the car follows the road. [0003] CN108829110A A driver model modeling method with a unified framework for horizontal / longitudinal motion, comprising the following steps: using a hyperbolic tangent function to establish a vehicle lateral single lane change tra...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G05D1/02G06N3/04G06N3/08
CPCG05D1/0253G05D1/0223G05D1/0276G06N3/04G06N3/08Y02T10/40
Inventor 张良祁永芳饶泉泉续秋锦李鑫
Owner HEFEI UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com