Design of electric vehicle warning sound systems to minimise drive-by noise

Abstract

Electric and hybrid-electric vehicles are required to emit artificial warning sounds at low speeds, as their quiet operation poses a potential hazard for vulnerable road users. At the same time, concerns have been expressed over the increase in environmental noise levels this measure might bring. This thesis aims to design a warning sound system that can both fulfil its role as a safety measure while minimising its noise contribution. In addition, the proposed system must be viable for a wide scale implementation by keeping the manufacturing and maintenance costs low, and ensuring that it is physically robust for long-term operation within a vehicle. For this purpose, a directional sound system based on structural vibration has been designed, utilising an array of inertial actuators forcing the structure upon which they are attached to vibrate and radiate a sound field of controllable directivity. An analytical model was formulated to describe the physical system, combining existing models of structural vibration and sound radiation. A simulations-based parametric study performed using this model provided insight into the design parameters, indicating that a larger number of actuators distributed evenly along the entire radiating surface ensures the greatest possible bandwidth and steerability for the system. The construction and measurement of a simple physical prototype system allowed for the experimental validation of the analytical model. The proposed system was evaluated in its intended implementation by installing the array in a test vehicle and measuring the directivity of the radiated sound field for different arrangements of the actuators on the vehicle. Experimental results indicated that an actuator array installed in the bumper of the vehicle can achieve directivity of at least 10 dB in terms of acoustic contrast level, for a bandwidth from 300 Hz to 5 kHz, while minimising interior and drive-by noise. An additional approach to reducing the impact of warning sounds was also investigated, in the form of an environmentally adaptive warning sound system. The proposed system employs an algorithm that estimates the auditory thresholds due to a changing sonic environment, and uses this information to adapt the warning sound. This aims to render the vehicle detectable in all environments without unnecessarily increasing its level, therefore limiting noise pollution. The adaptation algorithm was tested in a simulation-based application study for a variety of environmental noise scenarios, showing that it is capable of selectively adjusting the output of the warning sound at specific frequency bands to match the audibility threshold.

More information

Identifiers

ORCID for Jordan Cheer: orcid.org/0000-0002-0552-5506

Catalogue record

Export record