David Blanco-Rodriguez

A bias correction method for fast fuel-to-air ratio estimation in diesel engines

λ probes in turbocharged diesel engines are usually located downstream of the turbine, exhibiting a good dynamic response but a significant delay because of the exhaust line transport and the hardware itself. With the introduction of after-treatment systems, new sensors that can measure the exhaust concentrations are required for optimal control and diagnosis. Zirconia-based potentiometric sensors permit the measurement of nitrogen oxides and oxygen with the same hardware. However, their dynamic response is slower and more filtered than that of traditional λ probes and, in addition, the sensor location downstream of the after-treatment systems increases this problem. The paper uses a Kalman filter for online dynamic estimation of the relative fuel-to-air ratio λ−1 in a turbocharged diesel engine. The combination of a fast drifted fuel-to-air ratio model with a slow but accurate zirconia sensor permits the model bias to be corrected. This bias is modelled with a look-up table depending on the engine operating point and is integrated online on the basis of the Kalman filter output. The calculation burden is alleviated by using the converged gain of the steady-state Kalman filter, precalculated offline. Finally, robustness conditions for stopping the bias updating are included in order to account for the sensor and model uncertainties. The proposed algorithm and sensor layout are successfully proved in a turbocharged diesel engine. Experimental and simulation results are included to support validation of the algorithm.

A learning algorithm concept for updating look-up tables for automotive applications

Abstract Look-up tables are commonly used in the automotive field for handling operating point variations. However, constant maps cannot cope with systems variations and ageing. Methods, such as Kalman filter or Extended Kalman filter for non-linear cases, can be used for table adaptation providing an optimal solution to the problem. But these methods are computationally intensive, making difficult to implement them on commercial engine control units. The current paper proposes a learning method for online updating of look-up tables or maps. This algorithm uses precalculated membership functions based on a standard Kalman filter observer for weighting the adaptation. The main contribution of the method is the derivation of a steady-state Kalman filter observer that lowers the calculation burden and simplifies the implementation against the standard Kalman filter implementation that requires higher computational cost. As far as table is updated online while engine runs, this allows correcting drift errors and the unit-to-unit dispersion. The method is illustrated for mapping engine variables such as λ − 1 and N O x in a Diesel engine by using an adaptive look-up table, and its characteristics make it suitable for implementing in commercial engine electronic control units for online purposes.

Cycle by Cycle Trapped Mass Estimation for Diagnosis and Control

The development of one cycle resolution control strategies and the research at HCCI engines demands an accurate estimation of the trapped mass. In contrast to current methods for determining the mass flow, which are only able to determine averaged values of the flow entering the cylinders, the present paper proposes a methodology based on the in-cylinder pressure resonance. The determination of such frequency allows inferring the cylinder mass with one cycle resolution. In addition, the method permits determining error metrics based on the mass conservation principle. Validation results for a reactivity controlled compression ignition (RCCI) engine equipped with electrohydraulic variable valve timing (VVT) are presented to illustrate the performance of the method.

Modelling driving behaviour and its impact on the energy management problem in hybrid electric vehicles

Perfect knowledge of future driving conditions can be rarely assumed on real applications when optimally splitting power demands among different energy sources in a hybrid electric vehicle. Since performance of a control strategy in terms of fuel economy and pollutant emissions is strongly affected by vehicle power requirements, accurate predictions of future driving conditions are needed. This paper proposes different methods to model driving patterns with a stochastic approach. All the addressed methods are based on the statistical analysis of previous driving patterns to predict future driving conditions, some of them employing standard vehicle sensors, while others require non-conventional sensors (for instance, global positioning system or inertial reference system). The different modelling techniques to estimate future driving conditions are evaluated with real driving data and optimal control methods, trading off model complexity with performance.

ECU-oriented models for NOx prediction. Part 1: a mean value engine model for NOx prediction:

The implantation of nitrogen oxide sensors in diesel engines was proposed in order to track the emissions at the engine exhaust, with applications to the control and diagnosis of the after-treatment devices. However, the use of models is still necessary since the output from these sensors is delayed and filtered. The present paper deals with the problem of nitrogen oxide estimation in turbocharged diesel engines combining the information provided by both models and sensors. In Part 1 of this paper, a control-oriented nitrogen oxide model is designed. The model is based on the mapping of the nitrogen oxide output and a set of corrections which account for the variations in the intake and ambient conditions, and it is designed for implementation in commercial electronic control units. The model is sensitive to variations in the engine’s air path, which is solved through the engine volumetric efficiency and the first-principle equations but disregards the effect of variation in the injection settings. In order to consider the effect of the thermal transients on the in-cylinder temperature, the model introduces a dynamic factor. The model behaves well in both steady-state operation and transient operation, achieving a mean average error of 7% in the steady state and lower than 10% in an exigent sportive driving mountain profile cycle. The relatively low calibration effort and the model accuracy show the feasibility of the model for exhaust gas recirculation control as well as onboard diagnosis of the nitrogen oxide emissions.

ECU-oriented models for NOx prediction. Part 2: adaptive estimation by using an NOx sensor

The implantation of nitrogen oxide sensors in diesel engines is necessary in order to track emissions at the engine exhaust line for diagnosis and control of the after-treatment devices. However, the use of models is still necessary since the sensor outputs are delayed and filtered. The present paper deals with the problem of the nitrogen oxide estimation in two parts; Part 1 deals with a control-oriented model for the nitrogen oxide estimation, while Part 2 presents data fusion of the model and the sensor to improve the estimation, which is presented in the following. The use of models for the nitrogen oxide estimation is an alternative but the drift and the ageing are still issues. In order to overcome this problem, the fusion of different signals can be carried out in a smart way by means of a Kalman filter. There exist different ways of presenting this fusion, from directly tracking the bias to updating the model parameters. For this, different algorithms are proposed in this paper with the aim of correcting the model output. Furthermore, the estimation of the actual nitrogen oxide concentration, by preventing sensor delay and filtering, is also integrated in the algorithm, which is a suitable strategy for combining nitrogen oxide sensors and models on an onboard basis.

On a Stochastic Approach of the ECMS Method for Energy Management in Hybrid Electric Vehicles

Abstract This paper examines the problem of optimal energy management in vehicles with hybrid powertrains. There is a large literature addressing this problem, with successful application of different optimal control techniques, however, many of them optimise the control strategy for certain driving cycles. This paper takes into account that the driving cycle is not known a priori. The proposed method is based on saving the current driving conditions in a database with different clusters in order to use this information to estimate future driving conditions. Once the future driving conditions are estimated, the ECMS technique is used to minimise the fuel consumption, taking into account that the expected value of the battery energy consumption should be zero at the infinity. Simulations show that the proposed method allows charge sustainability providing near-optimal results.

System Setup and Sensors Characterisation

This chapter is divided into two parts. The first part, which comprises Sects. 3.2 and 3.3, is devoted to the description of the experimental configuration and tests used for this work. Section 3.2 describes the experimental set-up used in the present work, including the relevant characteristics of the engine, sensors and test cell equipment, while the Sect. 3.3 presents the steady-state and dynamic tests performed to tune and validate the methods. In the second part of the chapter, the use of onlinemethods for characterising \(\mathrm {NO_{x}}\) and \(\lambda \)-1 output from exhaust gas concentration sensors is emphasised. \(\mathrm {NO_{x}}\) output is characterised by a novel method based on SOI steps (Galindo et al. 2011), while \(\lambda \)-1 output is characterised by performing injection steps (Guardiola et al. 2013).

Control Models for Engine-Out \mathrm{{NO}}_\mathrm{{x}} and \lambda ^{-1}

This Chapter presents the design and validation results of ECU-oriented models for \(\lambda \)-1 and \(\text {NO}_{\text {x}}\) prediction. The first is just based on the calculation of the fuel-to-air ratio by the injected fuel mass flow and the air mass flow signals from the ECU while the second is based on a nominal set-point relative fitting of the \(\text {NO}_{\text {x}}\) with a series of corrections for accounting with variations on \(\lambda \)-1, temperatures and other signals. The \(\text {NO}_{\text {x}}\) model combines look-up tables with physical-based equations and is designed for being implemented on commercial ECUs.

Conclusions and Future Works

The present dissertation covers the topic of the online dynamic estimation of \(\lambda ^{-1}\) and NO\(_\text {x}\) in diesel engines. In this chapter, the main contributions and conclusions are presented and organised according to the thematic.