Unmanned Vehicle Sensor Layout Using Simulation-Based Perception Correction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing sensor solutions for unmanned vehicles have weak perception capability and low perception accuracy, primarily due to reliance on physical parameters alone, which affects safe operation.

Innovation Solution

A method that establishes a simulated unmanned vehicle and simulation scene for simulation driving, determining a sensor solution based on initialization parameters, including sensor installation parameters, and correcting the solution using simulation data and perception parameters to enhance perception capability and accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If sensor solution is determined only based on physical parameters of sensor, then device complexity is reduced, but perception capability and perception accuracy deteriorate

Engineering Contradiction:
Improvesensor solution design complexityVSAvoidperception accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent creates a virtual copy of the unmanned vehicle and its sensor system in a simulation environment. This virtual model replicates the physical sensor characteristics, vehicle geometry, and operating conditions to evaluate sensor solutions before physical deployment, thereby improving perception accuracy without increasing physical device complexity

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent performs sensor solution evaluation and optimization in advance through simulation testing. By conducting virtual experiments before actual deployment, the system pre-determines optimal sensor configurations, reducing the need for complex iterative adjustments and improving perception accuracy from the outset

Inventive Principle:
Principle #10Preliminary action

2Device complexity

If detection distance is estimated only based on obstacle size and vehicle model size, then measurement process is simplified, but detection distance accuracy deteriorates

Engineering Contradiction:
Improvedetection estimation complexityVSAvoiddetection distance accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent uses a virtual copy of the detection system in simulation to measure detection distance. The simulation replicates sensor beam patterns, obstacle properties, and environmental conditions to provide accurate detection distance measurements without requiring complex physical measurement processes

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent replaces physical measurement methods with computational simulation. Instead of using complex physical measurement equipment and procedures, the system uses virtual sensor models and computational algorithms to accurately determine detection distance based on simulated sensor data

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Measurement precision

If simulation-driven sensor solution correction is implemented, then perception capability and accuracy are improved, but device complexity and computation requirements increase

Engineering Contradiction:
Improveperception accuracyVSAvoidsensor solution design complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent uses a virtual simulation model to evaluate and correct sensor solutions. The virtual copy allows comprehensive testing of different sensor configurations without physical prototyping, improving perception accuracy while managing complexity through software-based iteration

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent implements a feedback loop where simulation results are used to evaluate sensor solution performance and automatically adjust configurations. The simulation provides performance feedback on perception accuracy, which is then used to refine sensor parameters, creating an iterative optimization process

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3816663B1Method, device, equipment, and storage medium for determining sensor solution
Publication Date: 2024.08.07 BEIJING BAIDU NETCOM SCI & TECH CO LTD
  • EP3816663B1 patent drawingFigure 1~1b
  • EP3816663B1 patent drawingFigure 2~3
  • EP3816663B1 patent drawingFigure 4~5

AI summary

The present application discloses a method, a device, equipment, and storage medium for determining a sensor solution, where the method includes: establishing (101) a simulated unmanned vehicle and a simulation scene, where the simulation scene is used for the simulated unmanned vehicle to perform simulation driving; determining (102) a first sensor solution according to a initialization parameter, and determining, according to the first sensor solution, simulation data generated by the simulated unmanned vehicle during the simulation driving in the simulation scene; and; and determining (103) a first perception parameter of the first sensor solution according to the simulation data, and correcting the first sensor solution according to the first perception parameter to obtain a sensor solution applied to an unmanned vehicle. This method fully considers and combines the physical parameters of the sensor, the shape and movement characteristics of obstacles, the size and appearance of the unmanned vehicle, as well as the traffic environment in which the vehicle is driving, and the determined sensor solution applied to the unmanned vehicle has strong perception capability and high perception accuracy, which is beneficial to guarantee the safe operation of the unmanned vehicle.