In-cylinder Pressure Pegging Algorithm Based on Cyclic Polytropic Coefficient Learning

Abstract

This paper presents an in-cylinder pressure pegging algorithm based on cyclic polytropic coefficient learning for combustion engines. In order to take the cycle-to-cycle variation of the polytropic coefficient into account in the in-cylinder measurement, an iterative learning algorithm is proposed to provide cyclic estimation of the polytropic coefficient and then with the estimation cyclic compensation method is proposed for the offset of in-cylinder pressure measurement. A comparative study of the proposed algorithm, the least-squares method (LSM) with a fixed polypropic coefficient and the nonlinear least-squares method (NLSM) with a variable polytropic coefficient is conducted using the simulated pressure data. Experimental validations are conducted on a six-cylinder gasoline engine at a motored condition and a steady fired operation condition.