Model-in-the-Loop Simulation for Model Predictive Controlled High-Pressure Direct Injection Dual-Fuel Engine Combustion Control

Abstract

Given the intricate combustion process and the multitude of control parameters inherent to the high-pressure direct injection (HPDI) diesel/natural gas dual-fuel engine, achieving precise combustion control represents a significant challenge. It is imperative to develop a high-precision engine model and integrate it with advanced control algorithms to achieve an optimal combustion strategy. In this study, a system-level engine plant model with high accuracy and real-time performance was developed using a modular modeling method through the calibration of experimental data and the simplification of model calculations. In this model, the relative error of the model simulation is controlled to be less than 5%, and the real-time factor (RTF) is less than 1. The multi-stage combustion process was parameterized by performing piecewise linear fitting of the heat release rate curve, and the relationship between injection parameters and combustion parameters was established using multiple regression analysis. On this basis, a model predictive control (MPC) algorithm was designed and verified in the constructed model-in-the-loop (MiL) platform. The results demonstrate that the designed MPC algorithm can accurately track the combustion phasing CA50 and the indicated mean effective pressure (IMEP) targets with a maximum error of 0.0624° and 0.046% within 6 and 8 cycles while ensuring the stability of the control process. The MiL platform not only meets the current combustion control requirements but also provides a general basis for the development of subsequent engine multi-control strategies and cooperative control optimization.

Keywords

• dual-fuel engine
• plant model
• model predictive control
• model-in-the-loop