"Everything should be made as simple as possible, but not simpler." - Albert Einstein

Seminar Announcement

Latin Hypercube Sampling and Thin Plate Splines Interpolation Approach for Developing Optimal Fuel Injection Strategies for Diesel Engines

Zonglin He, M.S. Candidate, Department of Statistics, Colorado State University

Monday, May 10, 2010

3:00 p.m., room 006, Statistics Bldg


In this paper, we consider the problem of optimizing diesel engine performance subject to emission and fuel efficiency constraints. This involves determining an optimal “engine map” for a given engine type. We examine how Latin Hypercube Sampling and Thin Plate Splines Interpolation can be used to make the optimization problem computationally feasible. We propose a simulation based approach for evaluating the suitability of different sampling and surface fitting strategies. Our findings are applied to a case study using a real engine map. We find that the proposed strategy does offer a reasonable approach for simplifying the engine performance constrained optimization problem.

Advisory Committee:

Dr. Hari Iyer, Advisor
Dr. Haonan Wang, Committee Member
Dr. Jie Luo, Electrical & Computer Engineering, Outside Member