Verification of Advanced Driver Assistance Systems (ADAS) and Autonomous Vehicles with Hardware Emulation-In-The-Loop

Abstract

Automotive Original Equipment Manufacturers (OEMs) are taking more control of Electronic Control Unit (ECU) hardware/software architecture definition and exploration mainly due to the rise of new technologies in ADAS and autonomous vehicles. The increasing number of cars with ECUs containing tens of processors, and the need to accelerate the simulation of thousands of driving scenarios and design parameters to ensure safety and reliability in autonomous driving has expanded the use of Hardware Emulation-In-The-Loop based on its capacity and functional verification merits. In this paper, an integrated heterogeneous system of systems framework to simulate and verify multiple ECUs is presented. The framework allows co-development of the automotive system application software, operating system and controller hardware together with the mechanical system actuators, sensors, and the physical world environment/traffic scenarios all in one. A comprehensive manual/automatic braking system use case consisting of four ECUs at different levels of abstraction is used to demonstrate the tools and interface technologies developed in-house and integrated with third-party tools.

Vehicle Driving Risk Prediction Model by Reverse Artificial Intelligence Neural Network

“…MATLAB software has been widely used in the design and simulation of control systems. At this stage, the field of simulation technology is a relatively mature scientific computing language, which is good at efficiently processing large amounts of data and visualizing data [8]. MATLAB is used to simulate the driving process of the car and analyze how the car engine and clutch, and other components break and operate in an emergency.…”