چكيده به لاتين
Improvement of fuel consumption, emissions and performance of vehicles have been the main subject of many researches in recent decades. These parameters along with a proper driving experiment can be considered as objective functions in an optimization problem. It has been proven that gear ratios of gearbox and final drive affect mentioned functions. The objective of this thesis is to find the optimized gear ratios in order to optimize 0-60 and 50-75 mph acceleration times, fuel consumption, CO, HC and NOx emissions and the number of gearshifts in a vehicle. For this purpose, multi-objective genetic algorithm is used as the optimization tool and MATLAB and ADVISOR are used for optimization and simulation processes. The optimization process is implemented on five different drive cycles with various patterns (FTP, UDDS, NEDC, INRETS and ARB02). After optimization process, Pareto curves are built to analyze the optimization results and the best optimized point is then chosen between optimal Pareto front points using Nearest to Ideal Point (NIP) method.
In order to find cycle-independent answers two innovative methods are suggested in this thesis; Combined-Cycle and Multi-Cycle analyses. Additionally, another approach has been implemented to obviate dependency of the answers from driving and shifting patterns which is the use of a universal gearshift strategy based on driver’s Sportivity Index instead of conventional shifting maps.
Pareto curves showed same results as similar documents and results were as predicted. After applying NIP on the pareto points, results indicated that generally all methods improve fuel consumption, emissions, and number of gearshifts. The mean values of changes in objective functions for all methods are calculated as 1.7%, 2.2% and 2.6% reduction in fuel consumption, emissions and shift count respectively. Conversely, accelerating times have been weakened by 3.7%. Note that changing the vehicle model or the setting of objective functions might cause alteration in all of the results.