It is the policy of OMNI Engineering Services Inc. not to disclose customer names and/or proprietary details of projects without the client’s release to do so.

Identify and implement a method for optimizing power train efficiency and determine the feasibility of optimizing efficiency "on the fly" in an embedded controller.

Matlab was used as the analysis engine. Several Matlab scripts (m-files) were written to simulate the hydraulic pump, hydraulic motors, and the vehicle engine. Efficiency maps for each component in the system were included. Scripts were written to perform the optimization routines, with the ability to optimize in various ways, I.e., maximal power, minimal fuel consumption, maximal traction, etc. And, scripts were written to compute optimal conditions along various operational trajectories, and to visualize the results in various useful ways.

Algorithms were developed and implemented. With the simulations, it became possible to verify, analytically, empirically known data about system efficiency, as well as to identify previously unknown optimal system configurations and behaviors.
Analysis of the simulations themselves showed that it was economically infeasible to create a real-time optimizing transmission control. The primary bottleneck was the very large number of floating point math operations required to find just a single optimal operating point, resulting in extremely long processing times. Even on the Matlab simulation, running on a 233MHz Pentium, it took nearly one minute to compute one such point, approximately 5,000 times too slow for a real-time system.

Matlab