Search Results

A computer code oriented to S.I. engine control and powertrain simulation is presented. The model, developed in Matlab-Simulink® environment, predicts engine and driveline states, taking into account the dynamics of air and fuel flows into the intake manifold and the transient response of crankshaft, transmission gearing and vehicle. The model, derived from the code O.D.E.C.S. for the optimal design of engine control strategies now in use at Magneti Marelli, is suitable both for simulation analysis and to achieve optimal engine control strategies for minimum consumption with constraints on exhaust emissions and driveability via mathematical programming techniques. The model is structured as an object oriented modular framework and has been tested for simulating powertrain system and control performance with respect to any given transient and control strategy.

A computer code oriented to S.I. engine control and powertrain simulation is presented. The model predicts engine and driveline states, taking into account the dynamics of air and fuel flows into the intake manifold and the transient response of crankshaft, clutch, transmission gearing and vehicle. The whole model is integrated in the code O.D.E.C.S., now in use at Magneti Marelli, and is based on a hierarchical structure composed of different classes of models, ranging from black-box Neural Network to grey-box mean value models. By adopting the proposed approach, a satisfactory accuracy is achieved with limited computational demand, which makes the model suitable for the optimization of engine control strategies. Furthermore, in order to simulate the driver behavior during the assigned vehicle mission profile, two drive controllers have been implemented for throttle and brakes actuation, based on classical PID and fuzzy-logic theory.

The paper describes suited methodologies for developing Recurrent Neural Networks (RNN) aimed at estimating NOx emissions at the exhaust of automotive Diesel engines. The proposed methodologies particularly aim at meeting the conflicting needs of feasible on-board implementation of advanced virtual sensors, such as neural network, and satisfactory prediction accuracy. Suited identification procedures and experimental tests were developed to improve RNN precision and generalization in predicting engine NOx emissions during transient operation. NOx measurements were accomplished by a fast response analyzer on a production automotive Diesel engine at the test bench. Proper post-processing of available experiments was performed to provide the identification procedure with the most exhaustive information content. The comparison between experimental results and predicted NOx values on several engine transients, exhibits high level of accuracy.

An integrated system of phenomenological models is applied in conjunction with identification techniques to simulate SI engine performance and emissions. In the framework of a hierarchical model architecture, the model structure provides the steady state engine data required for the design and validation of synthetic engine models. This approach allows limiting the recourse to the experimental data and speeds up the engine control strategies prototyping. The model structure is composed of a multi-zone thermodynamic engine model linked to a 1-D commercial fluid-dynamic model for intake and exhaust gas flow and to a physical model for NOx exhaust emissions. In order to improve model accuracy and generalization, an identification methodology is applied to estimate the optimal parameters for the turbulent combustion model. Due to the built-in physical content, the proposed methodology requires a relatively limited amount of experimental data for characterizing the under-study engine.

The paper focuses on the experimental identification and validation of recurrent neural networks (RNN) for real-time prediction and control of air-fuel ratio (AFR) in spark-ignited engines. Suited training procedures and experimental tests are proposed to improve RNN precision and generalization in predicting both forward and inverse AFR dynamics for a wide range of operating scenarios. The reference engine has been tested by means of an integrated system of hardware and software tools for engine test automation and control strategies prototyping. The comparison between RNNs simulation and experimental trajectories showed the high accuracy and generalization capabilities guaranteed by RNNs in reproducing forward and inverse AFR dynamics. Then, a fast and easy-to-handle procedure was set-up to verify the potentialities of the inverse RNN to perform feed-forward control of AFR.

A set of models for the prediction of mechanical efficiency as function of the operating conditions for an automotive spark ignition engine is presented. The models are embedded in an integrated system of models with hierarchical structure for the analysis and the optimal design of engine control strategies. The validation analysis has been performed over a set of more than 400 steady-state operating conditions, where classical engine variables and pressure cycles were measured. Models with different functional structures have been tested; parameter values and indices of statistical significance have been determined via non-linear and step-wise regression techniques. The Neural Network approach (Multi Layer Perceptrons with Back-Propagation) has been also used to evaluate the feasibility of using such an approach for fast black-box modelization.

The paper focuses on the experimental identification and validation of recurrent neural networks (RNN) for virtual sensing of NO emissions in internal combustion engines (ICE). Suited training procedures and experimental tests are proposed to improve RNN precision and generalization in predicting NO formation dynamics. The reference Spark Ignition (SI) engine was tested by means of an integrated system of hardware and software tools for engine test automation and control strategies prototyping. A fast response analyzer was used to measure NO emissions at the exhaust valve. The accuracy of the developed RNN model is assessed by comparing simulated and experimental trajectories for a wide range of operating scenarios. The results evidence that RNN-based virtual NO sensor will offer significant opportunities for implementing on-board feedforward and feedback control strategies aimed at improving the performance of after-treatment devices.

The growing concern about Greenhouse Gas (GHG) emissions led institutions to further reduce the limits on vehicle-related CO2 emissions. Therefore, car manufacturers are developing vehicles with low environmental impact, like Hybrid-Electric Vehicles (HEVs), which in the series architecture employ an Internal Combustion Engine (ICE) coupled with an electric generator for battery recharging, thus extending the range of a Battery Electric Vehicle (BEV). For this kind of application, small four-stroke Spark Ignition (SI) engines are preferred, as they are a proven and reliable solution to increase the driving range with very low environmental impact. In series hybrid-electric powertrains, the ICE is decoupled from the drive wheels, then it can operate in a steady-state high-efficiency working point, regardless of the power required by the mission profile. The benefits of lean combustion can be exploited to increase efficiency and reduce CO2 and NOx emissions.

International regulations continuously restrict the standards for the exhaust emissions from automotive engines. In order to comply with these requirements, innovative control and diagnosis systems are needed. In this scenario the application of methodologies based on the in-cylinder pressure measurement finds widespread applications. Indeed, almost all engine thermodynamic variables useful for either control or diagnosis can be derived from the in-cylinder pressure. Apart for improving the control accuracy, the availability of the in-cylinder pressure signal might also allow reducing the number of existing sensors on-board, thus lowering the equipment costs and the engine wiring complexity. The paper focuses on the detection of the engine thermal state, which is fundamental to achieve suitable control of engine combustion and after-treatment devices.

In the recent years, growing attention has been focused on internal combustion engines, considered as the main sources of Particulate Matter (PM) in urban air. Small particles are associated to fine dust formation in the atmosphere and to pulmonary diseases. The legislation proposes a stronger restriction in terms of particulate mass concentrations for both Diesel and gasoline engines and a limitation on number concentration. Unfortunately, the experimental evaluation of particles number and size is a hard task as they are strongly affected by the dilution conditions, due to condensation and nucleation phenomena, which may occur during the sampling. Even if a considerable amount of basic research on particulate matter emitted by engines has been carried out, the mechanisms governing particle formation are still not fully understood, neither for Diesel nor for gasoline engines.

The paper addresses the potentialities of Recurrent Neural Networks (RNN) for modeling and controlling Air-Fuel Ratio (AFR) excursions in Spark Ignited (SI) engines. Based on the indications provided by previous studies devoted to the definition of optimal training procedures, an RNN forward model has been identified and tested on a real system. The experiments have been conducted by altering the mapped injection time randomly, thus making the effect of fuel injection on AFR dynamics independent of the other operating variables, namely manifold pressure and engine speed. The reference engine has been tested by means of an integrated system of hardware and software tools for engine test automation and control strategies prototyping. The developed forward model has been used to generate a reference AFR signal to train another RNN model aimed at simulating the inverse AFR dynamics by evaluating the fuel injection time as function of AFR, manifold pressure and engine speed.

In the last few years, concern about the environmental impact of vehicles has increased, considering the growth of the dangerous effects on health of noxious exhaust emissions. For this reason, car manufacturers are moving towards more efficient combustion systems for Spark Ignition (SI) engines, aiming to comply with the increasingly stringent regulation imposed by EU and other legislators. Engine operation with very lean air/fuel ratios has demonstrated to be a viable solution to this problem. Stable ultra-lean combustion can be obtained with a Pre-Chamber (PC) ignition system, installed in place of the conventional spark plug. The efficiency of this configuration in terms of performance and emissions is due to its combustion process, that starts in the PC and propagates in the main chamber in the form of multiple hot turbulent jets.

The paper deals with the application of two techniques for the selection of the training data set used for the identification of Neural Network black-box engine models; the research starts from previous studies on Sequential Experimental Design for regression based engine models. The implemented methodologies rely on the Active Learning approach (i.e. active selection of training data) and are oriented to drive the experiments for the Neural Network training. The methods allow to select the most significant examples leading to an improvement of model generalization with respect to a heuristic choice of the training data. The data selection is performed making use of two different formulation, originally proposed by MacKay and Cohn, based on the Shannon's Statistic Entropy and on the Mean Error Variance respectively.

In the last years, the research effort of the automotive industry has been mainly focused on the reduction of CO2 and pollutants emissions. In this scenario, concepts such as the engines downsizing, stop/start systems as well as more costly full hybrid solutions and, more recently, Waste Heat Recovery technologies have been proposed. These latter include Thermo-Electric Generator (TEG), Organic Rankine Cycle (ORC) and Electric Turbo-Compound (ETC) that have been practically implemented on few heavy-duty applications but have not been proved yet as effective and affordable solutions for passenger cars. The paper deals with modeling of ORC power plant for simulation analyses aimed at evaluating the opportunities and challenges of its application for the waste heat recovery in a compact car, powered by a turbocharged SI engine.

A thermodynamic model for the simulation of performance and emissions in a spark ignition engine is presented. The model is part of an integrated system of models with a hierarchical structure developed for the study and the optimal design of engine control strategies. In order to reduce the uncertainty due to the mutual interference during the validation phase, the model has been developed accordingly with a hierarchical and sequential structure. The main thermodynamic model is based on the classical two zone approach. A multi-zone model is then derived form the two zone calculation, for a proper evaluation of temperature gradients in the burned gas region. The emissions of HC, CO and NOx are then predicted by three sub-models. In order to make the precision of emission models suitable for engine control design, an identification technique based on decomposition approach has been developed, for the definition of optimal model structure with a minimum number of parameters.

The paper deals with the use of Recurrent Neural Networks (RNNs) for the Air-Fuel Ratio (AFR) control in Spark Ignition (SI) engines. Because of their features, Neural Networks can perform an adaptive control more efficiently than classical techniques. In the paper, a review of the most useful control schemes based on Neural Networks is presented and the potential use in the field of engine control is analyzed. A preliminary controller has been implemented making use of a Direct Inverse Modeling approach. The controller compensates for the wall wetting dynamics and estimates the right amount of fuel to be injected to meet the target AFR during engine transients. The Direct Inverse Controller has been tested within an engine/vehicle simulator. The simulation tests have been performed by imposing a set of throttle transients at different engine speeds. The results show that the Inverse Model can satisfactorily bound the AFR excursions around the target value.

The computer code ODECS (Optimal Design of Engine Control Strategies) for the design of Spark Ignition engine control strategies is presented. This code has been developed starting from the author's activity in this field, availing of some original contributions about engine stochastic optimization and dynamical models. This code has a modular structure and is composed of a user interface for the definition, the execution and the analysis of different computations performed with 4 independent modules. These modules allow the following calculations: (i) definition of the engine mathematical model from steady-state experimental data; (ii) engine cycle test trajectory corresponding, to a vehicle transient simulation test such as ECE15 or FTP drive test schedule; (iii) evaluation of the optimal engine control maps with a steady-state approach.

In the paper, a rule-based (RB) control strategy is proposed to optimize on-board energy management on a Hybrid Solar Vehicle (HSV) with series structure. Previous studies have shown the promising benefits of such vehicles in urban driving in terms of fuel economy and carbon dioxide reduction, and that economic feasibility could be achieved in a near future. The control architecture consists of two main loops: one external, which determines final battery state of charge (SOC) as function of expected solar contribution during next parking phase, and the second internal, whose aim is to define optimal ICE- EG power trajectory and SOC oscillation around the final value, as addressed by the first loop. In order to maximize the fuel savings achievable by a series architecture, an intermittent ICE scheduling is adopted for HSV. Therefore, the second loop yields the average power at which the ICE is operated as function of the average values of traction power demand and solar power.