Publikationen

Due to their unique mobility, Low Earth Orbit (LEO) satellites can provide network services in areas with little to no existing infrastructure and due to their low orbit they enable communication at low latencies. We thus motivate the use of LEO satellites for supporting multi-hop Vehicle-to-Everything (V2X) communication as an alternative to dense base station deployment and present CLEO, an approach that can opportunistically use LEO satellites for information dissemination among vehicles. Using extensive computer simulations, we show that not only are relaying latencies low enough for such an approach to be feasible, but we can actually improve performance of information dissemination in terms of all of Packet Delivery Ratio (PDR), Channel Busy Ratio (CBR), and end-to-end latency.

In this paper we address the modeling and tracking control problem of an overactuated lift and tilt vertical nanopositioning system. We derive a model based on the rigid body dynamics which adequately describes the first resonance mode of each motion axis. This model is validated with measured data in the frequency domain to illustrate that it adequately reflects the real behavior. We further device an abstract model for the derivation of advanced control strategies. By virtue of the single axis model, three SISO controllers are implemented. The control strategy is accomplished comprising a nominal feedforward and LQG-type controller plus an L1 adaptive augmentation with output feedback. The baseline (or nominal) controller features sufficiently high bandwidth for the mere stabilization, decoupling, and disturbance rejection, while the L1 adaptive component plays a central role for recovering the nominal closed-loop dynamics in the presence of parametric uncertainties w.r.t. the input gain which are quite difficult to handle in the nominal design. The effectiveness and robustness of the proposed control strategy is verified via real-time experiments featuring subnanometer and nanoradian tracking errors which seem to be fully-dominated by the measurement noise.

As processes become more complex and the need to measure each and every variable becomes more critical, the ability of physical sensors to always provide the sufficient accuracy and sampling time can be difficult. For many complex systems, such as nonideal mixtures, multiphase fluids, and solid-based systems, it may not be possible to even use a physical sensor to measure the key variables. For example, in a multiphase fluid, the concentration or density may only be able to be accurately estimated using a laboratory procedure that can only produce a limited number of samples. Similarly, the quality variables of steel may only be determinable once the final steel product has been produced, which limits the ability to effectively control the process with small time delays. In such cases, recourse has to be made to soft sensors, or mathematical models of the system that can be used to forecast the difficult-to-measure variables and allow for real-time process monitoring, control, and optimisation. Although the development of the soft-sensor model is well-established, the various applications and use cases have not been often considered and the key challenges examined. It can be seen that soft sensors have been applied to a wide range of processes from simple, chemical engineering systems to complex mining processes. In all cases, major improvements in the process operations have been observed. However, key challenges remain in updating the soft-sensor models over time, combining laboratory measurements, especially when they are infrequent or of uncertain quality, and the development of soft sensors for new conditions or processes.

In modern process industries, elaborate monitoring and isolation of various disturbances and faults are needed for reliable and efficient system operation. The classic process-monitoring and fault-diagnosis methods can grasp the correlation between variables, and thus, only take care of abnormal situations caused by the corruption of the correlation relationship. However, dynamics anomalies are even more noteworthy as they reflect more internal details of the system dynamic behaviour under specific situations, and more importantly, can cause severe failures and spread to a broader range of areas while evolving over time. In this paper, a monitoring-and-isolation strategy is proposed to concurrently detect and isolate faults of static deviations and dynamic anomalies. The natural sparsity of the faulty variables is used to overcome the limitations of unknown fault directions and insufficient erroneous measurements, thereby translating the isolation problem into a quadratic programming problem with a sparsity constraint and solved by the least absolute shrinkage and selection operator (LASSO). The case study shows the advantages of the proposed method in monitoring and isolating static deviations and dynamic anomalies.

This paper introduces a learning-based Nonlinear Model Predictive Control (NMPC) method that combines NMPC with a Reinforcement Learning (RL) algorithm to achieve automatic parameter tuning of the NMPC optimizer, resulting in better control performance. In this study, two learning-based models were designed based on the tabular Q-learning algorithm but with different definitions of state and action spaces. To test the effectiveness of the proposed model, we conducted two kinds of experiments in which the models were applied to optimize the lane-keeping performance of an autonomous driving system. The case study results from simulations showed that the agent could match a proper parameter matrix for the NMPC within one minute. In real-world experiments, we extended the proposed control scheme to practical driving tasks using a 1:8 scale Audi model car in a specific experimental field. The model exhibited acceptable robustness in the face of relatively large deviations from the sensors and other real-time interference. These results demonstrate that the proposed learning-based NMPC method is a promising direction for solving real-time control problems.