Abstract
Advanced engine control systems require accurate models of the thermodynamic-mechanical process, which are substantially nonlinear and often time-variant. After briefly introducing the identification of nonlinear processes with grid-based look-up tables and a special local linear Radial Basis Function network (LOLIMOT), a comparison is made with regard to computation effort, storage requirements and convergence speed. A new training algorithm for online adaptation of look-up tables is introduced which reduces the convergence time considerably. Application examples and experimental results are shown for a multidimensional nonlinear model of NOx emissions of a Diesel engine, and for the adaptive feedforward control of the ignition angle of a SI engine.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call