Technical note: Optimum predictive search method for automatic acoustical instrument calibration

Acoustical measuring instruments such as sound level meters, require period ic calibration to comply with noise control regulations. For laboratories t hat are involved with calibrations of acoustical instruments, it is desirab le to develop automatic or semiautomatic calibration systems that are econo mically attractive. The purpose of this note is to discuss a new optimum pr edictive search method developed for acoustical instrument calibrations. In this method, the instrument setting-reading response is modeled as a time- invariant unknown system response. Previous measurement data (settings and readings taken from the instrument) are used to estimate the system model p arameters. The desired setting is then predicted according to the estimated system model, The search for the optimum becomes solving a linear/nonlinea r system equation. This fast optimum search algorithm can be used in other automatic measurement systems with substantial reduction in calibration tim e. The automatic calibration system developed has a precision that is bette r than a manual calibration system, and the uncertainty of the measured rel ative sound pressure level is within +/- 0.1 dB, (C) 1998 Institute of Nois e Control Engineering. [S0736-2501(98)00406-8].