Volume 65, Issue 4

The paper deals with methods to derive analytical redundancy relations for diagnosable systems. The starting point is an over-determined set of constraints, which is obtained by a structural analysis of the linear system under consideration. The problem is to find analytical expressions for residuals in terms of the measured signals and the derivatives of these signals. Two new methods are described. The first method provides a representation of the residual in terms of the observability matrix of the diagnosable subsystem. The second method replaces the derivative constraints by a series of equations that include the derivatives of the state variables and the outputs. The result is a matrix-vector representation of the over-determined set of equations, which leads to the analytical redundancy relations. The methods are illustrated by their application to an automotive example.

Detecting early enough the anomalous behavior of technical systems facilitates cost savings thanks to avoiding system downtimes, guiding maintenance, or improving performance. The novel framework proposed in this paper processes event streams originating from system monitoring for anomaly detection purposes. Therefore, statistical models characterizing the normal behavior of the monitored system are learned from the events. Instead of having one coarse normal model for all operational states, the proposed framework contains a mechanism for automatically detecting different conditions of the system allowing for fine-tuned models for every condition. The performance of the framework is demonstrated by means of a real-world application, where the log files of a large-scale printing machine are analyzed for anomalies.

New results on a particular type of state-dependent parameterization for model predictive control (MPC) are presented. Based on such parameterizations efficient MPC algorithms can be formulated, which combine the advantages of explicit and online optimization-based MPC. The new results comprise an offline stability check for the parameterizations to decide if a closed-loop MPC scheme applying the parameterizations is asymptotically stabilizing. Second, a novel way of computing the parameterizations with improved scalability in the state space dimension is included. Furthermore, new results and simplifications of almost explicit MPC schemes based on univariate parameterizations are contributed. The results are presented in a common framework and are illustrated in numerical examples including an almost explicit controller for an eight-dimensional spring-damper system.

Zusammenfassung Es wird eine Methode zur sensorlosen Positionsbestimmung bei elektromagnetisch betätigten Aktoren vorgestellt. Dabei werden basierend auf einer Signalinjektion die positionsabhängigen Parameter bei der injizierten Frequenz bestimmt und daraus über ein geeignetes Modell die Position des Magnetankers ermittelt. Die Eignung des Verfahrens zur sensorlosen Positionsregelung wird an einem bidirektionalen Proportionalventil anhand praktischer Versuche demonstriert.

Zusammenfassung Dieser Beitrag behandelt die praktische Umsetzung einer Regelung der Ammoniumkonzentration in der biologischen Abwasserreinigung der Kläranlage der BASF. Das System wird über zwei Typen von Aktoren beeinflusst, die kontinuierlich bzw. diskret arbeiten, und weist damit das Verhalten hybrider dynamischer Systeme auf. Es wird die praktische Anwendung eines Ansatzes zur hybriden ereignisbasierten Regelung vorgestellt, der um Mechanismen zur besseren Störkompensation und zum Umgang mit Stellgrößenbeschränkungen erweitert wird. Die Ergebnisse der praktischen Erprobung dieses Regelungsansatzes zeigen, dass der Arbeitspunkt des Systems an die Störung adaptiert und somit die Regelgüte gegenüber einem heuristischen Regelungsansatz deutlich verbessert wird.

In this publication, we describe an automatic vision based system for the reliable detection and recognition of green asparagus in variable daylight conditions. The focus is on an algorithm for the estimation and tracking of the 3D-position of green asparagus based on using of Kalman and Particle filters and a combination of both filters.