Neural network models for virtual sensing of NOx emissions in automotive diesel engines with least square-based adaptation

To meet current Diesel engine pollutant legislation, it is important to manage after-treatment devices. The paper describes the development of Neural Network based virtual sensors used to estimate NO_x emissions at the exhaust of automotive Diesel engines. Suitable identification methodologies and experimental tests were developed with the aim of meeting the conflicting needs of feasible on-board implementation and satisfactory prediction accuracy. In addition, since the prediction of control-oriented models is typically affected by engine aging and production spread as well as components drift, least square technique features were exploited in order to overcome these issues by adapting the virtual sensor output. The NO_x adaptive virtual sensor was tested via comparison with experimental data, measured at the engine test bench on a turbocharged common-rail automotive Diesel engine. Furthermore, besides model validation, the experimental measurements were modified to simulate a sensor drift in order to enable full assessment of the proposed LS-based algorithm adaptation capabilities.     

Development and experimental studies of a control-oriented SCR model for a two-catalyst urea-SCR system

This paper presents the development and experimental studies of a complete selective catalytic reduction (SCR) system control-oriented model of a two-catalyst SCR system with onboard NO_x and ammonia sensors. SCR catalysts have been popularly regarded as effective means for NO_x emission control in medium- and heavy-duty vehicles in recent years. However, control of urea dosing upstream of the SCR systems still remains a challenge in the field mainly due to the complicated SCR dynamics and limited/inaccurate feedback information. A control-oriented SCR model is thus indispensable for SCR control systems. A variety of experimental tests were examined using a Diesel engine-aftertreatment system consisting of a diesel oxidation catalyst (DOC)/diesel particulate filter (DPF), two-SCR catalysts (Fe-Zeolite type) in series, three NO_x sensors, and two NH₃ sensors. By utilizing multiple emission sensors and the two-catalyst SCR setup, the sensor properties and SCR system dynamics were studied. Grounded in the experimental investigations and the physical insights, a control-oriented model for a complete SCR system was developed and validated with experimental data.     

Model-based fault detection and isolation for a diesel lean NOx trap aftertreatment system

The lean ${\mathrm{NO}}_x$ trap (LNT) is an aftertreatment device used to reduce nitrogen oxides emissions on Diesel engines. To operate the LNT with high conversion efficiency, an optimized regeneration schedule is required, together with closed-loop control of the air/fuel ratio during regeneration. Furthermore, to comply with emissions regulations, diagnostic schemes are needed to detect and isolate faults, typically related to aging, sulfur poisoning and thermal deactivation.The paper describes a step-by-step methodology for the design and validation of model-based fault diagnosis for a LNT aftertreatment system. The approach is based on a control-oriented model of the LNT validated with experimental data.The proposed diagnostic approach is based on the generation of residuals using system models, through the comparison of the predicted and measured values of selected output variables. The paper focuses on the detection and isolation of sensor faults and LNT parametric faults. Different diagnostic methodologies are presented in relation to the detection of specific faults.Starting from sulfur poisoning detection in a laboratory environment which represents a preliminary validation of the approach, the diagnostic scheme is extended to detect various faults under different plant configurations and operating conditions, with a final application to on-board fault detection and isolation.     

Estimation and adaptive nonlinear model predictive control of selective catalytic reduction systems in automotive applications

The demands of high NO_x conversion efficiency and low tailpipe ammonia slip for urea-based selective catalytic reduction (SCR) systems have been substantially increased in the past decade, as NO_x emission legislations for Diesel engines are becoming more stringent than ever before. Since catalyst aging has a significant impact on SCR performance, robust and adaptive SCR control has been preferred for degraded SCR systems to realize emission control objectives. The purpose of this paper is twofold. Firstly, a robust ammonia coverage ratio observer was designed for estimating the ammonia coverage ratio reference for catalysts with different aging levels. An ammonia storage capacity observer was developed for estimating the actual ammonia storage capacity which can be reduced due to catalyst aging. An adaptive ammonia coverage ratio reference design was then developed to estimate the desired ammonia coverage ratio ranges at each instantaneous engine operating point for both single-cell and two-cell SCR systems at different aging levels based on a singular perturbation method. Secondly, to ensure the estimated ammonia coverage ratio falls in the desired ranges for most of engine operating conditions, robust nonlinear model predictive control (NMPC) algorithms were designed for both single-cell and two-cell SCR systems. Experimental data over US06 cycle were collected from a Diesel engine and aftertreatment system platform for controller verification. Simulation results under US06 test cycle demonstrate that the proposed NMPC algorithms were capable of consistently achieving high NO_x conversion efficiency (>95.6%) and constrained tailpipe ammonia slip (<10 ppm on average and <12 ppm on the peak) for both fresh catalyst and aged catalyst with 30% loss of ammonia storage capacity.     

Pressure-based transient intake manifold temperature reconstruction in Diesel engines

Temperature measurements by the typical thermocouples contain some first-order dynamics with varying time-constants and need to be reconstructed in transient conditions for improving the accuracy of the temperature information. Particularly, for Diesel engine advanced combustion mode control, the accurate acquisitions of the rapidly varying transient temperatures, such as the intake manifold gas temperature, are of importance. In this paper, a temperature reconstruction method, without using additional sensors, is proposed by utilizing the counterpart pressure signal. Through investigating the thermocouple dynamics in terms of the intake manifold pressure and temperature, an intake manifold temperature model was derived. According to this proposed temperature model, the transient temperature reconstruction can be formulated as a thermocouple time-constant estimation problem. Within this framework, an extended Kalman filter (EKF) based method was devised for the parameter estimations. The proposed method was validated through high-fidelity GT-Power engine model simulations as well as experimental results obtained on a multi-cylinder medium-duty Diesel engine.     

Design and experimental validation of an extended Kalman filter-based NOx concentration estimator in selective catalytic reduction system applications

This paper presents an extended Kalman filter (EKF) based approach of integrating NO_x and NH₃ sensors to estimate the NO_x concentrations in Diesel engine selective catalytic reduction (SCR) aftertreatment systems. NO_x sensors have been commonly used by vehicles for aftertreatment system control and onboard diagnostics (OBD) purposes. However, most currently available NO_x sensors are cross-sensitive to ammonia. Based on the experimental observations and physical inferences, the cross-sensitivity characteristics may change with temperature and is hard to be predicted by a model. This feature limits the applications of NO_x sensors on urea-SCR systems where ammonia is the reductant for NO_x conversions. Grounded in the insight into SCR dynamics and NO_x sensor properties, a novel approach of using an extended Kalman filter to estimate the actual exhaust gas NO_x concentration was proposed. The estimator was examined by NO_x measurements from a Horiba gas analyzer under different engine operating conditions. The experimental results show that the EKF-based approach can significantly improve the accuracy of NO_x concentration measurements from the original NO_x sensor readings.     

Rail pressure controller design of GDI basing on predictive functional control

Gasoline direct injection (GDI) is a pivotal technique for a highly efficient engine. However, how to maintain a stable rail pressure which offers good fuel economy and low emissions, is still a challengeable work. In this paper, a rail pressure controller is designed basing on predictive functional control (PFC), a model predictive control (MPC) method, to surmount the nonlinearity and discontinuity brought by the common rail pressure system (CRPS). A control-oriented piecewise linear model is presented to simplify the CRPS. The simulation results on a benchmark show that rail pressure tracks the setpoint accurately even with some perturbations. Profiting from the conciseness of PFC algorithm, the controller can compute the online solution in a short time, which makes it possible to realize the strategy on a fast response system.     

A computationally efficient Kalman filter based estimator for updating look-up tables applied to NOx estimation in diesel engines

No_x estimation in diesel engines is an up-to-date problem but still some issues need to be solved. Raw sensor signals are not fast enough for real-time use while control-oriented models suffer from drift and aging. A control-oriented gray box model based on engine maps and calibrated off-line is used as benchmark model for No_x estimation. Calibration effort is important and engine data-dependent. This motivates the use of adaptive look-up tables. In addition to, look-up tables are often used in automotive control systems and there is a need for systematic methods that can estimate or update them on-line. For that purpose, Kalman filter (KF) based methods are explored as having the interesting property of tracking estimation error in a covariance matrix. Nevertheless, when coping with large systems, the computational burden is high, in terms of time and memory, compromising its implementation in commercial electronic control units. However look-up table estimation has a structure, that is here exploited to develop a memory and computationally efficient approximation to the KF, named Simplified Kalman filter (SKF). Convergence and robustness is evaluated in simulation and compared to both a full KF and a minimal steady-state version, that neglects the variance information. SKF is used for the online calibration of an adaptive model for No_x estimation in dynamic engine cycles. Prediction results are compared with the ones of the benchmark model and of the other methods. Furthermore, actual online estimation of No_x is solved by means of the proposed adaptive structure. Results on dynamic tests with a diesel engine and the computational study demonstrate the feasibility and capabilities of the method for an implementation in engine control units.     

A model-based gain scheduling approach for controlling the common-rail system for GDI engines

The progressive reduction in vehicle emission requirements have forced the automotive industry to invest in research for developing alternative and more efficient control strategies. All control features and resources are permanently active in an electronic control unit (ECU), ensuring the best performance with respect to emissions, fuel economy, driveability and diagnostics, independently from engine working point. In this article, a considerable step forward has been achieved by the common-rail technology which has made possible to vary the injection pressure over the entire engine speed range. As a consequence, the injection of a fixed amount of fuel is more precise and multiple injections in a combustion cycle can be made. In this article, a novel gain scheduling pressure controller for gasoline direct injection (GDI) engine is designed to stabilise the mean fuel pressure into the rail and to track demanded pressure trajectories. By exploiting a simple control-oriented model describing the mean pressure dynamics in the rail, the control structure turns to be simple enough to be effectively implemented in commercial ECUs. Experimental results in a wide range of operating points confirm the effectiveness of the proposed control method to tame efficiently the mean value pressure dynamics of the plant showing a good accuracy and robustness with respect to unavoidable parameters uncertainties, unmodelled dynamics, and hidden coupling terms.     

Coal-fired utility boiler modelling for advanced economical low-NOx combustion controller design

This paper focuses on developing a control-oriented coal-fired utility boiler model for advanced economical Low-NO_x combustion (ELNC) controller design. Two boiler combustion models are proposed in this paper: one is a mathematical model describing the key dynamics of the real-time boiler thermal efficiency and the furnace one-dimensional NO_x concentration distribution under conventional fuel and overfire air operations; the other recast from the first model is a control-oriented grey-box model with a data-driven furnace combustion submodel. Simulation studies on static and dynamic properties of the first mathematical model indicate that the model can function as a real-time simulator for both advanced boiler combustion control laws testing and generating training and validation data for the control-oriented grey-box model. At the end of this paper, the control-oriented grey-box modelling procedure as well as an optional discrete time linear state-space model are summarised to facilitate model-based advanced combustion controllers design.     

Predictive control of a real-world Diesel engine using an extended online active set strategy

In order to meet tight emission limits Diesel engines are nowadays equipped with additional hardware components like an exhaust gas recirculation valve and a variable geometry turbocharger. Conventional engine control units use two SISO control loops to regulate the exhaust gas recirculation valve and the variable geometry turbocharger, although their effects are highly coupled. Moreover, these actuators are subject to physical constraints which seems to make an advanced control approach like model predictive control (MPC) the method of choice. In order to deal with MPC sampling times in the order of milliseconds, we employed an extension of the recently developed online active set strategy for controlling a real-world Diesel engine in a closed-loop manner. The results show that predictive engine control based on online optimisation can be accomplished in real-time – even on cheap controller hardware – and leads to increased controller performance.     

Hybrid modelling and control of the common rail injection system

We present an industrial case study in automotive control of significant complexity: the new common-rail fuel-injection system for Diesel engines under development at Magneti Marelli Powertrain. In this system, an inlet metering valve, inserted before the high pressure (HP) pump, regulates the fuel flow that supplies the common rail according to the engine operating point (e.g., engine speed and desired torque). The standard approach in automotive control based on a mean-value model for the plant does not provide a satisfactory solution as the discrete-continuous interactions in the fuel injection system, due to the slow time-varying frequency of the HP pump cycles and the fast sampling frequency of sensing and actuation, play a fundamental role. We present a design approach based on a hybrid model of the Magneti Marelli Powertrain common-rail fuel-injection system for four-cylinder multi-jet engines and a hybrid approach to the design of a rail pressure controller. The hybrid controller performs significantly better when compared with the classical mean-value based approach.     

Identification of a control-oriented energy model for a system of fan coil units

In recent years, application of advanced control, fault detection and diagnosis algorithms for building heating and cooling systems has been intensively investigated with the aim to improve their energy efficiency and bring the buildings sector into the smart city arena. Hindering the trend, hysteresis and proportional–integral–derivative controllers are still a common practice for temperature control in buildings with Fan Coil Units (FCUs). Introduction of more sophisticated controllers for additional savings requires a cost-effective approach for identification of an energy model which accurately resembles thermal and hydraulic performance of a system of FCUs. In the present work, the control-oriented energy model of a system of FCUs is developed and accompanied with replicable, robust and simple methodologies for its identification derived by consolidating the advantages of physical modelling, identification methods and manufacturer’s catalogue data. The validity of the developed approach is tested on the 248-office living-lab. The introduced simple and accurate dynamic characterization of energy transmitted from a FCU to zone air fills the gap between thermal and energy management for buildings. This enables implementation of predictive building controls and unleashes significant energy and cost-saving potentials of a smart building in a smart city.     

Model-based temperature control of a diesel oxidation catalyst

The problem studied in this article is the control of a DOC (diesel oxidation catalyst) as used in aftertreatment systems of diesel vehicles. This system is inherently a distributed parameter system due to its elongated geometry where a gas stream is in contact with a spatially distributed catalyst. A first contribution is a model for the DOC system. It is obtained by successive simplifications justified either experimentally (from observations, estimates of orders of magnitude) or by an analysis of governing equations (through asymptotic developments and changes of variables). This model can reproduce the complex temperature response of DOC output to changes in input variables. In particular, the effects of gas velocity variations, inlet temperature and inlet hydrocarbons are well represented. A second contribution is a combination of algorithms (feedback, feedforward, and synchronization) designed to control the thermal phenomena in the DOC. Both contributions have been tested and validated experimentally. In conclusion, the outcomes are evaluated: using the approach presented in this article, it is possible to control, in conditions representative of vehicle driving conditions, the outlet temperature of the DOC within ±15 °C.     

基于RBF网络的柴油机运行工况分析

本文介绍了神经网络中运用广泛的RBF神经网络的模型及其学习算法．将其引入到柴油机运行工况分析领域，在MATLAB环境下，建立了柴油机喷油特性的RBF神经网络模型，并用以模拟柴油机的喷油特性，模拟结果得到了实验验证。     

Modeling cycle-to-cycle dynamics and mode transition in HCCI engines with variable valve actuation

Due to dilution limits, homogeneous charge compression ignition (HCCI) engines will need to switch to a conventional spark ignition (SI) or diesel mode at very low and high load conditions. This paper outlines a simple and accurate control-oriented model of a single-cylinder multi-mode HCCI engine using exhaust reinduction. An exhaust manifold model ties the exhausted gas from one cycle to that reinducted on the next cycle. The results show that the cyclic dynamics which occur during an SI-to-HCCI mode transition are essentially just an extension of the inherent cycle-to-cycle dynamics of HCCI. Multi-cycle, multi-mode simulations demonstrate the steady state and dynamic behavior seen on an experimental system. Predicted combustion timing, in-cylinder pressure, work output and exhaust temperature values agree very well with experiment. This model provides a basis for developing and validating controllers capable of controlling HCCI during mode transitions.     

Optimal energy management for a diesel hybrid electric vehicle considering transient PM and quasi-static NOx emissions

In this paper, optimal energy management strategies are derived to balance fuel consumption, raw particulate matter (PM) emissions, and raw nitrogen oxide (NOx) emissions for a Diesel hybrid electric vehicle. Two methods for the derivation of these strategies are compared. One method is based on dynamic programming and steady-state engine maps only. The second method is based on dynamic programming, steady-state engine maps, and a validated transient PM emission model. As a result, only the second method allows for the generation of smooth engine set point trajectories to reduce transient PM emissions without compromising fuel consumption and NOx emissions.     

A decision-making module for aiding ship system automation design: A knowledge-based approach

The use of elements of artificial intelligence, including knowledge-based systems, becomes more and more widespread in aiding design problem solutions. The authors have been working on problems of control systems for many years. A design process involves many decision problems connected with, for example, a choice of a subsystem structure, subunits or particular elements selection. Because of such regards, it was decided to extend knowledge-based system with a module for support of such decision making.In this paper, an elaborated module for decision-making support is considered. The basic theoretical assumptions concerning the accepted method of multiattribute decision making based on pairwise comparison in categories of hierarchical decision process (AHP) is presented. Accepted knowledge representation in AHP method and pairwise comparison method and methods of expert knowledge acquisition are discussed. The module functioning is illustrated by an example of choice of temperature sensors in a system of fuel transport to Diesel engine of a main propulsion unit of a ship.     

Model-based analysis for the thermal management of open-cathode proton exchange membrane fuel cell systems concerning efficiency and stability

In this work we present a dynamic, control-oriented, concentrated parameter model of an open-cathode proton exchange membrane fuel cell system for the study of stability and efficiency improvement with respect to thermal management. The system model consists of two dynamic states which are the fuel cell temperature and the liquid water saturation in the cathode catalyst layer. The control action of the system is the inlet air velocity of the cathode air flow manifold, set by the cooling fan, and the system output is the stack voltage. From the model we derive the equilibrium points and eigenvalues within a set of operating conditions and subsequently discuss stability and the possibility of efficiency improvement. The model confirms the existence of a temperature-dependent maximum power in the moderate temperature region. The stability analysis shows that the maximum power line decomposes the phase plane in two parts, namely stable and unstable equilibrium points. The model is capable of predicting the temperature of a stable steady-state voltage maximum and the simulation results serve for the design of optimal thermal management strategies.     

基于KULI的发动机热管理瞬态模型的参数设置与仿真

以某重型柴油机为原型,利用KULI软件建立了发动机冷却系统模型,阐述了模型中发动机参数的设置原理,进行了瞬态工况下冷却液温度以及润滑油温度变化的仿真。结果说明：其仿真结果与实验数值吻合较好,模型精度可靠,模型中发动机的参数设置原理为发动机热管理瞬态模型的设置提供了一种方法,使用该方法可用来设置不同类型的发动机,为发动机热管理系统的仿真计算提供较为准确的放热边界条件,并进一步用来仿真完整的发动机热管理系统。