Accelerating the pace of engineering and science

Development of Integrated Vehicle Safety Applications Through Model-Based Design

Siddharth D'Silva, Autoliv
Richard Rakes, Autoliv

Autoliv is an industry leader in the development and manufacture of automotive safety systems with an enterprise-wide mission of saving lives. For this reason, safety is viewed as a unified domain and not just a collection of individual applications. While production groups work on next-generation environmental sensor technologies like radar and vision, development teams are focused on integrating these functions for enhanced vehicle state estimation and coordinated motion control. The long-term vision is a safety-domain controller for enhanced motion control that is capable of interpreting driver intention and, in conjunction with inertial (e.g., yaw, lat, roll, pitch), environmental (e.g., radar, camera), and peripheral (e.g. high-g impact) sensing, coordinating the control of restraints and vehicle motion through brakes, steering, suspension, and powertrain.

Model-Based Design ensures a disciplined approach to integration from routine tasks such as sensor signal processing and reference model tuning/calibration to more advanced concepts such as sensor fusion, vehicle state estimation, and the generation, arbitration, and distribution of vehicle or specific actuator control. For example, Autoliv’s enhanced Restraints Control Module combines inertial sensing capabilities for stability control with traditional airbag control in a single integrated controller. The improved vehicle state estimates provide invaluable information to the airbag controller, enhancing passenger safety while also enabling auxiliary features such as vehicle stabilization in response to light to moderate impact crash forces.

MATLAB® and Simulink® provided a starting point for concept evaluation. Those concepts were extended and improved through cosimulation with CarSim®, an advanced vehicle dynamics simulation tool. The algorithms were then evaluated in-vehicle using a dSPACE® rapid prototyping ECU. Finally, code was automatically generated from Real-Time Workshop Embedded Coder™ and implemented on a production enhanced Restraints Control Module. Additionally, the production code was subjected to and passed the standard automotive verification and validation procedures before customer release.

Product Focus

  • Simulink

Recorded: 4 May 2011