Filetype
The proposed algorithm aims at helping diabetic patients which use multi-dose injection (MDI) therapy to determine the dose and time of insulin and glucose intakes to stabilize their blood glucose (BG) concentration. The objective is to spend most of the time in normal BG range (70-180 mg/dl). The control was done by minimizing the risk of hypo- and hyperglycemia over the doses and times of insuli
Semantic segmentation and object detection are nowadays dominated by methods operating on regions obtained as a result of a bottom-up grouping process (segmentation) but use feature extractors developed for recognition on fixed-form (e.g. rectangular) patches, with full images as a special case. This is most likely suboptimal. In this paper we focus on feature extraction and description over free-
This paper describes a new method of finding thin, elongated structures in images and volumes. We use shortest paths to minimize very general functionals of higher-order curve properties, such as curvature and torsion. Our method uses line graphs to find the optimal path on a given discretization, often in the order of seconds on a single computer. The curves are then refined using local optimizat
The first part of the thesis deals with adaptive CPU resource management for multicore platforms. The work was done as a part of the resource manager component of the adaptive resource management framework implemented in the European ACTORS project. The framework dynamically allocates CPU resources for the applications. The key element of the framework is the resource manager that combines feedfor
We consider the problem of trajectory generation for time-optimal path tracking for the class of pseudo-omnidirectional mobile robots. An Euler-Lagrange model of the robot dynamics is derived, and by writing it on special form a convex reformulation of the path-tracking problem can be utilized. This enables the use and regeneration of time-optimal trajectories during runtime. The proposed approach
We consider robust output feedback distributed model predictive control (DMPC). The proposed controller is based on the results in [8] in which nominal stability and feasibility was proven for a DMPC-formulation without terminal constraint set or terminal cost in the optimization. We extend these results to show robust stability under state feedback as well as output feedback when dynamics and mea
Current market trends for industrial automation are the need for customizable production, shorter time to market, and powerful global competitive pressure. Based on these trends two challenges have been identified: 1) flexible production systems and 2) integration and utilization of devices and software. Applications from both process automation, manufacturing, and robotics have been considered.
An approach for merging statically schedulable subr egions in dataflow models is pr esented. The approach combines abstr act int erpr etation, loop analysis, and static scheduling of cyclo-static dataflow networ ks. The approach has been implemented in a Java-based tool that per forms automatic classification of dataflow act or s, generat ion of stat ic schedules using constr aint programming, and
This thesis presents models and methods for feedback-based resource management for cyber-physical systems. Common for the scenarios considered are severe resource constraints, uncertain and time-varying conditions and the goal of enabling flexibility in systems design rather than restricting it. A brief survey on reservation-based scheduling, an important enabling technology for this thesis, is pr
This paper explores a formally grounded ap- proach to solve the problem of dynamic binding in service- oriented software architecture. Dynamic binding is a widely adopted mean to automatically bind exposed software interfaces to actual implementations. The execution of an operation on one or another implementation, though providing the same result, could turn out in different quality of service, e
Based on previous results on linear predictors and Kalman filter, this paper will formulate multi-step-ahead multi-variate predictors for application in multi-step-ahead predictive control.
This thesis explores methods for solving the structure-and-motion problem in computer vision, the recovery of three-dimensional data from a series of two-dimensional image projections. The first paper investigates an alternative state space parametrization for use with the Kalman filter approach to simultaneous localization and mapping, and shows it has superior convergence properties compared wit
We consider the problem of localizing a novel image in a large 3D model. In principle, this is just an instance of camera pose estimation, but the scale introduces some challenging problems. For one, it makes the correspondence problem very difficult and it is likely that there will be a significant rate of outliers to handle. In this paper we use recent theoretical as well as technical advances t
Embedded systems often operate in so variable con- ditions that design can only be carried out for some worst-case scenario. This leads to over-provisioned resources, and undue power consumption. Feedback control is an effective (and not yet fully explored) way to tailor resource usage online, thereby making the system behave and consume as if it was optimized for each specific utilization case. A
This report addresses the related problems of trajectory generation and time-optimal path tracking with online obstacle avoidance. We consider the class of four-wheeled vehicles with independent steering and driving on each wheel, also referred to as pseudo-omnidirectional vehicles. Appropriate approximations of the dynamic model enable a convex reformulation of the path-tracking problem. Using th
There is currently a strongly growing interest in obtaining optimal control solutions for vehicle maneuvers, both in order to understand optimal vehicle behavior and to devise improved safety systems, either by direct deployment of the solutions or by including mimicked driving techniques of professional drivers. However, it is nontrivial to find the right mix of models, formulations, and optimiza
In this paper the impact of different meals and rapid insulin were estimated as Finite Impulse Response Models from a data set of 18 patients. Based on these models short-term individualized predictors were tested for 20 and 60 minute prediction. The predictors were evaluated using Clarke Grid Analysis and had on average more than 94 % and 75 % in the A zone and less than 1 % and 3 % in the erroro
