A Study on Powertrain Noise Reduction Control Vehicles for Improving Speech Recognition Accuracy in Hybrid Electric Vehicle

Abstract

Voice recognition technological development has prompted the creation of many convenience functions using a driver's voice recognition in a vehicle. However, the driver's voice recognition in a vehicle has some difficulties in the recognition process due to various noises generated by the vehicle. This paper proposed a powertrain control strategy to improve voice recognition rate by reducing noise in electrified vehicles. In an electrified vehicle equipped with various powertrain modes, it is possible to obtain the effect of reducing the noise inside the vehicle through appropriate mode transition during voice recognition. The suggested control strategy was implemented in the Hybrid Control Unit of a hybrid powertrain, and vehicle tests were conducted using the Kia Niro Hybrid vehicle. The results demonstrated that transitioning the powertrain mode from HEV mode to EV mode during voice control achieved noise reduction effects of approximately 12 dB during idling and 3.5 dB during driving. The noise reduction of the vehicle using the proposed algorithm helps improve the voice recognition rate and is expected to increase the utilization of various voice recognition applications.