FIELD: acoustics.
SUBSTANCE: invention relates to processing and conversion of acoustic signals into electrical signals in a robot. Technical result of the proposed technical solution is achieved by calibrating the robot working microphones in two steps, at the first step, transmitting a pulsed audio signal which is translated by the robot dynamics speakers and received by the working microphones, determining time delay of passage of pulse audio signal, transmitting obtained values of time delays to memory of control device, at second stage of calibration sound signal is transmitted successively in different frequency bands, translating it with robot speakers, determining the sound level value at each frequency for each working microphone, calculating the audio signal attenuation coefficient for each operating microphone, transmitting obtained values of attenuation coefficients to memory of control device, after which audio signals of robot are delayed in buffer, value of external speech signal of each working microphone is corrected by subtraction of time delay value.
EFFECT: technical result is reduced noise and reduced level of audio signal from robot loudspeakers.
1 cl, 3 dwg
