Giuseppe Conte

HRR and MFB50 Estimation in a Euro 6 Diesel Engine by Means of Control-Oriented Predictive Models

The paper has the aim of assessing and applying control-oriented models capable of predicting HRR (Heat Release Rate) and MFB50 in DI diesel engines. To accomplish this, an existing combustion model, previously developed by the authors and based on the accumulated fuel mass approach, has been modified to enhance its physical background, and then calibrated and validated on a GM 1.6L Euro 6 DI diesel engine. It has been verified that the accumulated fuel mass approach is capable of accurately simulating medium-low load operating conditions characterized by a dominant premixed combustion phase, while it resulted to be less accurate at higher loads. In the latter case, the prediction of the heat release has been enhanced by including an additional term, proportional to the fuel injection rate, in the model. The already existing and the enhanced combustion models have been calibrated on the basis of experimental tests carried out on a dynamic test bench at GMPT-E. A comparison has been made between the models, in terms of accuracy in the prediction of HRR and MFB50, as well as of the required computational time and calibration effort, at several steady-state operating conditions as well as over NEDC and WLTP cycles. The values of MFB50 predicted by means of the two approaches, for the same steady-state tests and driving cycles, have been compared with those obtained from a low-throughput invertible MFB50 predictive model that has recently been developed by the authors, which is characterized by an extremely low computational time

Feedback linearization control for the air & charging system in a diesel engine

The air and charge subsystem of a diesel engine is a quite complex system with strong couplings, actuator constraints, and fast dynamics. MIMO model based control strategies are essential to deal with such a process. Furthermore, the air-path process is highly nonlinear, based on thermodynamical phenomena. Therefore, classical linear MIMO approaches are difficult to be applied and require system linearization over equilibrium operating points. Conversely, feedback linearization control can be applied smoothly. In this paper, an air and charging model-based multivariable control approach is described, which has been developed for a single-stage turbocharged diesel architecture, enabling an improved trade-off of NOx and PM (particulate matter) emissions and better transient performances. The developed approach uses a nonlinear physics-based model of the air and charging system, and the control architecture is based on the feedback linearization technique that decouples interactions and compensates nonlinearities. The nonlinear control is then coupled with PI controllers in order to guarantee the transient performances and robustness. Test-bench results show the effectiveness of the proposed approach.

Nonlinear MIMO Data-Driven Control Design for the Air and Charging Systems of Diesel Engines

Emission requirements for diesel engines are becoming increasingly strict, leading to the increase of engine architecture complexity. This evolution requires a more systematic approach in the development of control systems than presently adopted, in order to achieve improved performances and reduction of times and costs in design, implementation and calibration. To this end, large efforts have been devoted in recent years to the application of advanced Model-Based MIMO control systems. In the present paper a new MIMO nonlinear feedback control is proposed, based on an innovative data-driven method, which allows to design the control directly from the experimental data acquired on the plant to be controlled. Thus, the proposed control design does not need the intermediate step of a reliable plant model identification, as required by Model-Based methods. In this way, significant advantages over Model-Based methods can be achieved in terms of times and costs in design and deployment as well as in terms of control performances. The method is applied to the control design for the air and charging systems, using experimental data measured on a four cylinder diesel engine with single stage turbocharger. The performances of the designed controller are evaluated on an accurate nonlinear engine model, showing significant reductions of up to 2.7 times for the intake manifold pressure, up to 2.7 times for the oxygen concentration tracking errors and about 4 times in controller design and calibration efforts with respect to a decoupled-gain-scheduled PID controller typically applied for the air charging system control of diesel engines

Neural-Network Based Approach for Real-Time Control of BMEP and MFB50 in a Euro 6 Diesel Engine

A real-time approach has been developed and assessed to control BMEP (brake mean effective pressure) and MFB50 (crank angle at which 50% of fuel mass has burnt) in a Euro 6 1.6L GM diesel engine. The approach is based on the use of feed-forward ANNs (artificial neural networks), which have been trained using virtual tests simulated by a previously developed low-throughput physical engine model. The latter is capable of predicting the heat release and the in-cylinder pressure, as well as the related metrics (MFB50, IMEP - indicated mean effective pressure) on the basis of an improved version of the accumulated fuel mass approach. BMEP is obtained from IMEP taking into account friction losses. The low-throughput physical model does not require high calibration effort and is also suitable for control-oriented applications. However, control tasks characterized by stricter demands in terms of computational time may require a modeling approach characterized by a further lower throughput. To this aim, feed-forward NNs have been trained to predict MFB50 and BMEP using a large dataset of virtual tests generated by the well-calibrated low-throughput physical engine model. The real-time approach has also been applied to derive the start of injection of the main pulse and the injected fuel quantity to achieve specific targets of MFB50 and BMEP. The accuracy of the real-time approach has been assessed based on experimental data taken at GM-GPS (General Motors - Global Propulsion Systems) facilities and its computational time has been compared to that of the low-throughput physical engine model, at steady-state and transient conditions over the WLTP cycle. Copyright