Peter Langthaler

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

Abstract 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.

New Regressors for the Direct Identification of Tire Deformation in Road Vehicles Via “In-Tire” Accelerometers

The interaction between the tire and the road is crucial for determining the dynamic behavior of a road vehicle, and the road-tire contact forces are key variables in the design of traction, braking, and stability control systems. Traditionally, road-tire contact forces are indirectly estimated from vehicle-dynamics measurements (chassis accelerations, yaw-roll rates, suspension deflections, etc.). The emerging of the ldquosmart-tirerdquo concept (tire with embedded sensors and digital-computing capability) has made possible, in principle, a more direct estimation of contact forces. In this field - still in its infancy - the main open problems are the choice of the sensor(s) and the choice of the regressor(s) to be used for force estimation. The objective of this work is to present a new sensor-regressor choice, and to provide some preliminary experimental results, which confirm the validity of this choice. The idea is to use a wheel encoder and an accelerometer mounted directly in the tire. The measurement of the in-tire acceleration is transmitted through a wireless channel. The key innovative concept is to use the phase shift between the wheel encoder and the pulse-like signals provided by the accelerometer as the main regressor for force estimation.

NOx Virtual Sensor Based on Structure Identification and Global Optimization

On-line measurement of engine NOx emissions is the object of a substantial effort, as it would strongly improve the control of Cl engines. Many efforts have been directed towards hardware solutions, in particular to physical sensors, which have already reached a certain degree of maturity. In this paper, we are concerned with an alternative approach, a virtual sensor, which is essentially a software code able to estimate the correct value of an unmeasured variable, thus including in some sense an input/output model of the process. Most virtual sensors are either derived by fitting data to a generic structure (like an artificial neural network, ANN) or by physical principles. In both cases, the quality of the sensor tends to be poor outside the measured values. In this paper, we present a new approach: the data are screened for hidden analytical structures, combining structure identification and evolutionary algorithms, and these structures are then used to develop the sensor presented. While the computational time for the sensor design can be significant (e.g. 1 or more hours), the resulting formula is very compact and proves able to predict the behaviour of the system at other operating points. The method has been validated with NOx data from a production engine measured with a Horiba Mexa 7000. The approach is able to yield a good prediction behaviour over a whole cycle. The results are consistent with physical knowledge.

MPC for a diesel engine air path using an explicit approach for constraint systems

The air path of an internal combustion engine is a classical example of MIMO system with actuator constraints and high dynamic requirements. While the classical approach consists in using simple, decoupled heuristic controllers and empirical limitations, this paper proposes to state the problem in terms of an optimal control problem with input constraints and to solve it in a model based environment. To this end, a recently developed controller design - explicit MPC - is used to calculate the explicit solution of the state feedback control law offline and to store it in tables for online controller selection. The combination of plant and switched controller leads to a piecewise linear and discrete system - a special form of hybrid system. The approach has been tested on a production diesel engine yielding impressive improvements in terms of soot while compared with the basic application.

Torque generation model for diesel engine

In this paper a combustion model of direct injection diesel engine is proposed to calculate the in-cylinder pressure and a crank-slider mechanism model to calculate instantaneous indicated torque. The crankshaft is modelled as a rigid body. The parameters of both models are identified via non-linear least square optimization algorithm. The data, used for the identification procedure, are purposely obtained through experiments on a diesel turbocharged BMW MD47 1900cc with a dynamic test-bench.

Fast Predictive Oxygen Charge Control of a Diesel Engine

Every control system of a diesel engine tries to supply the combustion chamber with the mixture of fresh gas and recirculated exhaust gas adequate to the requested injection quantity. Usually, this mixture is expressed in particular by a combination of intake manifold pressure, measured locally, and fresh air flow measured at a substantial distance from the combustion chamber. As the amount of exhaust recirculation is subject to many disturbances and uncertainties, standard control approaches are not satysfying, especially during transients, and the emission targets are met essentially by reducing the dynamics of fuel injection by so called smoke maps. In this paper, we discuss the possibility to design an oxygen charge control which estimates the available oxygen amount in the intake manifold - after the mixture of fresh and recirculated exhaust gas - using an oxygen sensor. To reduce the effect of modeling errors or of bandwidth limitations, the comparison with the standard choice of variables is performed using data based multilinear models and an explicit MPC to compute the control signals. Comparisons performed on a production engine on a dynamic test bench confirm that the approach is able to provide a better oxygen control during transients.

Identification of driveline parameters using an augmented nonlinear model

Abstract Modern approaches for engine control assume the knowledge of the dynamic properties connecting the engine via driveline to the road. A formally identical problem arises when a combustion engine is operated on a test bench, with an electrical machine simulating the wheel load. While in some cases design information allows sufficient estimation of the parameters, in many other cases it may prove more adequate to determine them using measurements. This is usually complicated by the fact that measurements of driveline quantities are disturbed by gas exchange in the cylinder. As this paper shows, a suitable representation of the plant allows to concentrate the disturbances arising from the compression into a nonlinear periodic term acting in parallel to a static nonlinear feedback caused by friction. If standard linear parameter identification methods are used, the effect of this concentrated nonlinearity corresponds to a possibly infinite number of additional poles, which, however, depend on the rotational speed of the shaft. Using this property, it turns out possible to use a standard ARMAX identification approach to determine the model of the driveline. This is confirmed by measurements performed on an engine test bench.

Adaptive Inverse Torque Control of a Diesel Engine Using Adaptive Mapping Update

Torque control is a basic element of engine control systems, in particular since it has become a standard interface for different functionalities. Torque control is also a critical requirement emission test cycle simulation on test benches. This torque control is usually reached by extensive, physical based modeling of the vehicle. This paper presents an approach to avoid this effort and to obtain a dramatic reduction of the parametrization work, by first determining an approximated model and then updating it online during operation. This model is than used for a stable inverse control. To handle model uncertainties and perturbation a correction feedback, with robustifying effect, is added to the control structure. This approach is detailed using data and measurements on a BMW M47D production diesel engine on a dynamic test bench.

Robust Model Predictive Control of a Diesel Engine Airpath

Abstract Model predictive control represents one of the most promising methods, also in fast industrial applications like a Diesel engine. But especially Diesel engine control meets problems of uncertainties and disturbances, thus model-plant mismatch is omnipresent, which obviously decreases the performance of model based control. This paper compares different robust predictive control strategies applied to a Diesel engine airpath.

Using formal methods to verify safe deep stall landing of a MAV

The various fields of application for miniature air vehicles often do not provide distinct landing areas or even require additional equipment like nets or parachutes to land the aircraft without damaging it. This work introduces the deep stall landing (DSL) as a maneuver that uses the extraordinary aerodynamic characteristics of a delta wing MAV that come into effect after the angle of attack passes the stall angle. This landing maneuver is modeled based on a longitudinal aerodynamic model that takes lift, drag, thrust, weight, and pitching moment into account. By determining the operational modes that the aircraft has to perform in order to either complete the landing maneuver or abort it in case of a missed approach a hybrid system is identified. This system contains both continuous and discrete state dynamics that model the aircraft in each landing phase. Based on this hybrid system reachability analyses are performed which utilize level set methods to calculate backwards reachable sets. These sets are used to identify transitions within the modeled system that bring the aircraft form one operational mode to another without leaving the safe flight envelope. The final result is a discrete event system that covers all possible transitions within the refined model. Based on this discrete model an autonomous system can be implemented that is able to determine whether the initiation of the landing maneuver is safe in terms of keeping the aircraft within the safe flight envelope during the whole maneuver. Furthermore the results of the reachability analysis determine for which states of the aircraft it would be safe to initiate a recovery maneuver in case of a missed landing approach.