Automated psychoacoustic based method for detecting borderline spark knock

摘要

An automated method for detecting borderline spark knock in a spark ignition engine imitates borderline knock procedures in which human operators listen to engine sounds as a function of spark timing. The automated method employs a psychoacoustic model and a psychophysical process to estimate borderline spark knock. The psychoacoustic model represents the acoustic activity in the human auditory system in response to engine sounds. The psychoacoustic model is monitored as a function of spark timing to detect audible spark knocks. The psychophysical process alters the spark timing for detecting spark knocks until the spark timing converges to a value corresponding to borderline knock. The automated method eliminates human operator subjectivity in estimating borderline knock.