Piston motion control for a dual free piston linear generator: predictive-fuzzy logic control approach

Abstract

The free-piston engine has a lot of potential to alter the conventional combustion engine due to mechanical structural simplicity, high engine efficiency, reducing hazardous gas emission and fuel flexible combustion. However, accuracy of piston motion control is a major challenge in all types of free-piston engine as well a dual free piston engine generator (DFPG) because of not having a crankshaft mechanism. This article describes a simulation investigation of a predictive-fuzzy logic control method conducted on the DFPG prototype in University of Ulsan. First, the simulation model of the DFPG consisted of a free-piston engine and a linear alternator is established and running in Matlab/Simulink environment. Then simulation result is validated against experimental data to prove validity of the simulation model. Second, a predictive-fuzzy logic (PFL) control method is proposed to apply to the DFPG and its performance is carried out using the full-cycle simulation model. The PFL strategy is demonstrated to improve the DFPG performance and more effective than the PID one in terms of settling time, TDC error, and cycle-to-cycle variation. According to results, the PFL controller reduces settling time 30 % in comparison to PID controller. The PFL controller has also 0.52 mm better TDC position error than the PID controller is 0.76 mm. Coefficient of variation values is found to be 1.0 %, and 1.9 % for the PFL controller and PID controller, respectively. This means that the PFL controller is a powerful method to dampen cycle to cycle variations in the DFPG.