Artificial Intelligence Enhanced Resolver System for Automotive Traction Inverter Applications Based on AURIX TC4x

Abstract:
Continuous performance increase due to scaling of semiconductor technology allows to utilize more artificial intelligence (AI) approaches on embedded devices. This paper focuses on an AI algorithm running on the automotive microcontroller of the new 28nm AURIX TC4x family to implement a novel resolver to digital converter solution. The proposed approach demonstrates how the AI is enhancing the commonly used hardware solution in the automotive industry by utilizing the new computation hardware inside the AURIX TC4x family. By doing so the safety compliance can be still guaranteed due to a redundant path approach and specific hardware features.