Unlock instant, AI-driven research and patent intelligence for your innovation.

Unmanned vehicle intelligent training method based on simulation learning in virtual environment

A technology of simulation learning and virtual environment, applied in the field of intelligent training of unmanned vehicles, can solve the problems of low efficiency and slow learning speed of unmanned vehicles, and achieve the effect of improving learning efficiency and shortening the process of learning tasks

Active Publication Date: 2020-08-25
ZHENGZHOU UNIVERSITY OF LIGHT INDUSTRY
View PDF8 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0010] The main technical problem to be solved by the present invention is to provide an intelligent training method for unmanned vehicles based on simulation learning in a virtual environment, which solves the problem that the learning speed of unmanned vehicles is relatively slow and the efficiency is relatively low

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
  • Unmanned vehicle intelligent training method based on simulation learning in virtual environment
  • Unmanned vehicle intelligent training method based on simulation learning in virtual environment
  • Unmanned vehicle intelligent training method based on simulation learning in virtual environment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0034] Example 1, such as Figure 1-3 As shown, an intelligent training method for unmanned vehicles based on simulation learning in a virtual environment includes the following steps:

[0035] Step 1. Build a virtual scene; the virtual scene includes natural environment, road environment, vehicle model and road network semantics.

[0036] Step 2. Establish an unmanned vehicle motion model;

[0037] Step 3. Carry out simulation in the virtual scene built in step 1 according to the vehicle motion model in step 2, and obtain the motion information of the unmanned vehicle as the simulation data set; the motion information includes acceleration, constant speed, deceleration, left turn, right turn and stop. The simulation data set includes environmental information and unmanned vehicle driving actions.

[0038] The simulation data set includes Λ(t) and a(t), (t ∈ (0, T), where Λ(t) is the environmental information at time t, which is the high-dimensional environment intercepted ...

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

No PUM Login to View More

Abstract

The invention discloses an unmanned vehicle intelligent training method based on simulation learning in a virtual environment. According to the invention, the motion behavior of the unmanned vehicle is simulated; a simulation data set is obtained, wherein the simulation data set comprises environment information in the current environment and unmanned vehicle driving actions; then compressed encoding preprocessing is carried out on the environment information; a data set is grouped by using a cross grouping method, a fitting value in a relational expression is solved by using least square fitting, finally, a CMA-ES algorithm is evolved and optimized on the basis of the fitting value, and an obtained optimal solution is added into an intelligent training process of the unmanned vehicle to guide the vehicle to learn good behaviors.

Description

technical field [0001] The invention relates to the field of virtual reality, in particular to an intelligent training method for unmanned vehicles based on simulation learning in a virtual environment. Background technique [0002] Unmanned driving is in line with the trend of automobile intelligence and Internetization, and it is a major opportunity for the transformation of the automobile industry under the wave of the Internet. "Made in China 2025" clearly proposes to accelerate the intelligent transformation of automobile and other industries. 2020 will be the first year of commercialization of unmanned vehicles, and it will enter a period of explosive growth. [0003] Specifically, the goal of unmanned vehicle driving research is to completely replace or partially replace the driver's function. In daily life, the driving control of vehicles is not only boring but also dangerous. Drivers need to pay attention to changes in the environment at all times, and at the same ...

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): G06F30/27G06N3/04G06N3/08G06N3/00G09B9/04
CPCG06F30/27G06N3/08G06N3/006G09B9/04G06N3/045Y02T10/40
Inventor 王华孟颍辉张旭马亚丹韩丽张静张伟伟徐盛金保华
Owner ZHENGZHOU UNIVERSITY OF LIGHT INDUSTRY