TeraGlobus的动态

Transcript

The Automotive Intelligence for Connected Shared Mobility AI 4 CSM project is dedicated to developing Europe's intelligent electronic components and systems for the next generation of mass market electric, connected, autonomous and shared vehicles to accelerate the transition towards a sustainable ecosystem. AI4 CSM follows the trends in electrification, standardization, automation, and digitalization. In line with the Green Deal principles, supply chain 2 envisions an EV 5.0 vehicle by redesigning a Mercedes-Benz EQC. This electric passenger car features an upgraded highly efficient powertrain controlled by newly developed methods for real time AI based fault detection analysis and mitigation. An 800 Volt silicone carbide inverter along with an E axle that includes a high performance 800. Electric machine and gearbox and a revised oil cooling system were integrated to meet the increased fast charging requirements. An 800 Volt battery replacing the original 400 Volt version has been developed in the 1000 Kilometer Plus project. The battery is equipped with an AI enhanced battery management system based on the Fox BMS platform. The highly reliable 800 Volt sick based power inverter from ZF has been integrated. Into the vehicle, This inverter has been developed within the activities of supply Chain 4, which dealt with robust propulsion systems. The redesigned Control Board integrates the next generation or XPPU platform enabling real time control and diagnostics. AI algorithms exploiting PPU capabilities. 3D time of flight sensors were integrated into a mobile 360?? perception system. Leveraging a hybrid AI or conventional calibration procedure for data registration. The time off flight camera grid was mounted on the demonstrator vehicle and through a calibration routine it is possible to align the camera outputs to the new positions. A GUI assists the user by providing critical information about the environment surrounding the vehicle. Using processed information from the time of flight sensors, a comprehensive vehicle simulation model developed within the AVL Model CONNECT framework was implemented to support the development of subsystems and components for the EV 5.0 demonstrator vehicle. A prediction model for energy consumption was trained using data gathered from over 13, 000 kilometers of test drives with multiple drivers in various environmental conditions. The model leverages reinforcement learning to provide an accurate prediction of the energy required for the trip based on fleet knowledge. The vehicle interface can query a routing service in the cloud and receive a route for on-screen display. Supply Chain 2 demonstrates a new generation of highly automated electric passenger vehicles, integrating an improved intelligent electric propulsion system for continuous use with lower power consumption. Utilizing predictive diagnosis methods. And real time driving assistance systems.