Time-frequency signal analysis of the acoustic emission of Formula 1 engines

Presents the application of frequency domain signal analysis methods to estimation of engine performance parameters from measurements of sound intensity. The method is applied to measurements obtained from Formula 1 (F1) race cars during practice and race conditions. The paper begins by summarizing the issues of interest in the analysis of engine noise from high-performance engines, the information content of the signal, and presents the analysis of some experimental data. The results of this study indicate that joint time-frequency signal processing may permit the estimation of various performance-related engine parameters in the absence of telemetry data. The analysis methods are applied to acoustic emission data obtained from F1 vehicles through in-car cameras during 1998 Formula 1 practice, qualifying and race sessions. The data presented in this paper were acquired during the 1998 F1 season, and pertain to the performance of the Ferrari and McLaren-Mercedes vehicles during the Grand Prix of San Marino (Imola) and of France (Magny-Cours). The results of the analysis demonstrate that it is possible to estimate a number of useful engine performance variables from simple sound measurements. These estimates are related both to engine performance (e.g., engine speed of rotation, acceleration, power, top speed) and to driving strategy (e.g., gear ratios, shifting strategy).