Systems and methods for estimating an engine event location are disclosed herein. In one embodiment, a controller is configured to receive a signal from at least one knock sensor coupled to a reciprocating engine, transform the signal, using a multivariate transformation algorithm, into a power spectral density, transform the power spectral density into a plurality of feature vectors using predictive frequency bands, predict the engine event location using at least the plurality of feature vectors and a predictive model, and adjust operation of the reciprocating engine based on the engine event location.

A system includes an engine control system configured to receive a first vibration signal from a first knock sensor disposed about a reciprocating engine, apply a binaural model to the first vibration signal, derive an engine health condition based on the application of the binaural model to the first vibration signal, and communicate the engine health condition.

A system includes a cylinder, a piston, a sensor configured to detect vibrations of the cylinder, piston, or both that correspond with varying pressures within the cylinder, and a controller coupled to the sensor. The controller is configured to receive a first signal from the sensor corresponding with first vibrations of the cylinder and to deduce from the first signal a first operating value of a parameter indicative of peak firing pressure at a first time, to compare the first operating value with a baseline value of the parameter indicative of peak firing pressure to detect a change in peak firing pressure, to receive a second signal from the sensor corresponding with second vibrations of the cylinder and to deduce from the second signal a second operating value of the parameter indicative of peak firing pressure at a second time, and to compare the second operating value with the baseline value to confirm the change in peak firing pressure.