Taguchi-based combined grey relational and principal component analyses for multi-response optimization of diesel engines

Abstract

Purpose Diesel engine can produce power more efficiently with lower exhaust emissions when operated at optimum input parameter settings. To achieve this goal, the purpose of this paper is to optimize the input parameters of diesel engine which will lead to optimum performance and exhaust emissions. Design/methodology/approach To achieve the goal of improving diesel engine performance and exhaust emissions, four input parameters were considered in the study. Five different levels of each input parameter were taken. Four response variables under no load, half load and full load conditions were recorded. Experiments were performed in random manner according to selected Taguchi L25 orthogonal array. The data were analyzed using grey relational analysis coupled with principal component analysis. Analysis of S/N ratio was performed to obtain the optimum combination of input parameters. The grey relational grade at optimum setting of the input parameters was obtained by regression analysis. Findings Results of the current research work give the optimum input parameter settings for no load, half load and full load conditions of diesel engine. Engine produces power more efficiently with low exhaust emissions when operated at these optimum settings. Practical implications In view of the compliance to the stringent air pollution norms of the nations and fast depleting fossil fuels, it is of the utmost importance to design and operate the engine in the optimum range of its input parameters so that it produces more power with low exhaust emissions. This paper aims at optimizing input parameters of diesel engine to improve performance and exhaust emissions. Results of the study presented in this paper are significantly useful for diesel engine-related researchers and professionals. Originality/value From the literature review, it appears that only few researchers have conducted studies pertaining to the optimization of the input parameters of diesel engine to improve performance or exhaust emissions. Although few studies related to the optimization of compression ratio, fuel injection timing, fuel injection pressure and air pressure have been reported, no work related to optimization of temperature and pressure of turbocharged air has been reported. Therefore, the main focus of the current research work is on optimizing the charge air temperature and pressure with respect to performance and exhaust emissions.

Simulation-based lean six sigma for Industry 4.0: an action research in the process industry

“…Taguchi's Robust Design methodology is a popular optimization technique to ensure the robustness of the product by accommodating both noise and control parameters (Almansoori et al , 2020). However, it has the limitation of optimizing a single output (Muqeem et al , 2017), while the GRA assists in drawing a robust inference by transforming the multi-output into a single output optimization problem (Naresh et al , 2014; Neeli et al , 2018). This approach is used to establish optimal solutions to uncertain problems having incomplete information and discrete data (Wu, 2002).…”

Multi-Response Optimization of Nanofluid-Based I. C. Engine Cooling System Using Fuzzy PIV Method

“…MADM techniques are best suited when selection problem involves attributes/criteria that are conflicting in nature. Literature reveals application of these techniques for solving decision problems pertaining to different knowledge domain [15][16][17][18][19]. For instance, Mufazzal and Muzakkir [20] proposed a novel MADM technique called Proximity Index Value (PIV) method to minimize the rank reversal problems arising due to either addition or deletion of alternatives.…”

Multi objective optimization of diesel engine performance and emission characteristics using Taguchi-grey relational analysis

“…The measured output values from the Taguchi design OA are accepted as input and normalized as the initial step of data pre-processing. The normalized values now fall inside the interval of 0 to 1 [37]. The normalization equation for the smaller the better characteristic is written as shown in Equation 3.…”