Global Synchronous Optimization Including Intake Valves and Exhaust Valves Based on MODM Strategy

摘要

Fully variable valve actuation (FVVA) technology can improve engine performance by changing engine valve timing and valve lift in different conditions. To find the optimal valve profile parameters, engine power and emissions have to be taken into account due to the contradiction between them. Therefore, the paper presented a multi-objective decision making (MODM) strategy for valve profile parameters optimization. All profile parameters both intake valves and exhaust valves were optimized synchronously. We carried out the global synchronous MODM strategy with the Radial Basis Function (RBF) response surface method (RSM). First, a FVVA system was introduced. For the FVVA system, all valve profile parameters are variable. Second, based on GT-suite, an engine model was built and its valve sub-models were rebuilt to achieve the FVVA system. Some simulation results were used to train RSM. Third, based on the simulation results, engine valve parameters were analyzed and some response surface methods were compared. The results show that RBF method has a good agreement with simulation results. At the end, global MODM strategy which optimized both intake valves and exhaust valves synchronously for engine valve parameters was implemented at 2400 RPM based on Non-dominated Sorting Genetic Algorithm (NSGA-Ⅱ) and RBF response surfaces obtained. With the obtained response surfaces, multi-objective virtual optimization which is relied on the obtained response surfaces rather than the GT-suite model was carried out. The virtual optimization results show that engine power can be improved by 1.63%, and the engine NO _X emissions can be reduced by 2.75%. To verify the virtual optimization results, we carried out the simulation with the optimized parameters, and the virtual optimization results have a good agreement with results which were obtained from the simulation of the GT-suite model. Besides, the RSM-based virtual multi-objective optimization keeps faster optimization speed.