Image Processing-Based Methodology for Optimizing Automotive Ignition Timing

Modern fuel-injected automobiles use an electronic engine control unit (ECU) that is programmed to ignite the air-fuel mixture in a way that maximizes the torque (TQ) and power output of the engine without compromising its reliability. The ECU achieves this by reading a base value from the primary ignition map (PIM) and then manipulating various other parameters to come up with the final solution. Each input into the final solution comes from a so-called <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">map</i> , which is a 3-D image that represents a set of all the possible values that each parameter can take on. In this paper, we develop a robust methodology using image processing techniques to optimize the PIM for automobiles. The techniques are used to compare, combine, manipulate, and reconstruct various maps. Experiments show that the process yields a gain of about 7 hp and a TQ of 15 ft middotlb measured at the wheels.