Promoting excellence in mobility engineering

  1. FISITA Store
  2. Technical Papers

Ray Optics Approach to Model Light Detection And Ranging (LIDAR) Technology for Selfdriving Cars Under Different Surrounding Obstructions and Atmospheric Conditions
F2018/F2018-ACV-037

Authors

Uttam Pal
COMSOL Multiphysics Pvt. Ltd

Prajakta Sabnis, Swapnil Badjugar

Christopher Boucher
COMSOL Inc, USA

Abstract

Research and/or Engineering Questions/Objective: Light Detection And Ranging (LIDAR) has been used extensively while prototyping selfdriving cars, especially because of its ability to create a 2D / 3D map of nearby surroundings. The LIDAR subsystem involves a laser source which gets spatially broadened in horizontal and vertical directions, also known as field of view (FOV). This broadened source of light, while hitting a nearby object, scatters some of the energy back to the photodetectors embedded in the LIDAR subsystem. This simulation model evaluates the distance between the LIDAR subsystem and nearby obstructions under different atmospheric conditions such as rain, snow, or fog. The obstructions can consist of trees, pedestrians, walls of buildings in the periphery, and other vehicles. The objective is to investigate the positioning, field of view (FOV) and arrangement of the array of laser source and photodetectors to give full coverage of obstacle detection and ranging. Methodology: In this study, we create a model of the LIDAR subsystem with ray tracing approach, using the COMSOL Multiphysics® simulation software (1). To begin with, we introduce the laser source angle using a cone release feature with a certain amplitude of power, launching in the order of ten thousand rays in different directions. The scattering of the rays from different boundaries such as trees, building walls, metallic boundaries of vehicles, and a human body are modeled, while taking into consideration the specific mixed specular and diffuse scattering coefficient. The particle density and the visibility of a particular atmospheric condition can be used to evaluate the imaginary component of the refractive index of the atmosphere, from which the absorption and scattering losses are estimated. Results: We were able to detect nearby building walls, pedestrians, and vehicles with good accuracy, predicting the difference in power detected under different atmospheric conditions. The positioning and arrangement of the array of photodetectors was investigated to provide full coverage of the surrounding area with an error of less than 5%. Limitations of this study The model doesn’t support any Monte Carlo scattering analysis, and so the optical scattering is instead just treated as attenuation with no change of ray direction. What does the paper offer that is new in the field including in comparison to other work by the authors? The paper brings a new approach to model the LIDAR ranging systems using the ray optics technique, which assumes the laser light source as photons with a particular operating wavelength.

Add to basket

Back to search results