Search Results

An analytical methodology to efficiently evaluate design alternatives in the conversion of a Common Rail Diesel engine to either CNG dedicated or dual fuel engine has been presented in a previous investigation. The simulation of the dual fuel combustion was performed with a modified version of the KIVA3V code including a modified version of the Shell model and a modified Characteristic Time Combustion model. In the present investigation, this methodology has been validated at two levels. The capability of the simulation code in predicting the emissions trends when changing pilot specification, like injected amount, injection pressure and start of injection, and engine configuration parameters, like compression ratio and axial position of the diesel injector has been verified. The second validation was related to the capability of the proposed computer-aided procedure in finding optimal solutions in a reduced computational time.

An adaptative energy management strategy for series hybrid electric vehicles based on optimized maps and the SUMO (Simulation of Urban MObility) predictor is presented here. The first step of the investigation is the off line optimization of the control strategy parameters (already developed by the authors) over a series of reference mini driving cycles (duration of 60s) obtained from standard driving cycles (UDDS, EUDC, etc) and realistic driving cycles acquired on the ITAN500 HEV. The optimal variables related to each mini driving cycle are stored in maps that are then implemented on the ITAN500 vehicles. When the vehicle moves, a wireless card is used to exchange information with surrounding vehicle and infrastructure. These information are used by a local instance of the SUMO traffic prediction tool (run on board) to predict the driving conditions of the HEV in the future period of time T=60s.

The optimization procedure adopted in the present investigation is based on Genetic Algorithms (GA) and allows different fitness functions to be simultaneously maximized. The parameters to be optimized are related to the geometric features of the combustion chamber, which ranges of variation are very wide. For all the investigated configurations, bowl volume and squish-to-bowl volume ratio were kept constant so that the compression ratio was the same for all investigated chambers. This condition assures that changes in the emissions were caused by geometric variations only. The spray injection angle was also considered as a variable parameter. The optimization was simultaneously performed for different engine operating conditions, i.e. load and speed, and the corresponding fitness values were weighted according to their occurrence in the European Driving Test.

The aim of the present investigation is the implementation of an innovative procedure to optimise the design of a high pressure common rail electro-injector. The optimization method is based on the use of genetic programming, a search procedure developed by John Holland at the University of Michigan. A genetic algorithm (GA) creates a random population which evolves combining the genetic code of the most capable individual of the previous generation. For the present investigation an algorithm which includes the operators of crossover, mutation and elitist reproduction has been developed. This genetic algorithm allows the optimization of both single and multicriteria problems. For the determination of the multi-objective fitness function, the concept of Pareto optimality has been implemented. The performance of the multiobjective genetic algorithm was examined by using appropriate mathematical functions and was compared with the single objective one.