Improving convergence of Iterative Feedback Tuning

Iterative Feedback Tuning constitutes an attractive control loop tuning method for processes in the absence of an accurate process model. It is a purely data driven approach aiming at optimizing the closed loop performance. The standard formulation ensures an unbiased estimate of the loop performance cost function gradient with respect to the control parameters. This gradient is important in a search algorithm. The extension presented in this paper further ensures informative data to improve the convergence properties of the method and hence reduce the total number of required plant experiments especially when tuning for disturbance rejection. Informative data is achieved through application of an external probing signal in the tuning algorithm. The probing signal is designed based on a constrained optimization which utilizes an approximate black box model of the process. This model estimate is further used to guarantee nominal stability and to improve the parameter update using a line search algorithm for determining the iteration step size. The proposed algorithm is compared to the classical formulation in a simulation study of a disturbance rejection problem. This type of problem is notoriously difficult for Iterative Feedback Tuning due to the lack of excitation from the reference.     

Virtual Reference Feedback Tuning for non-minimum phase plants

Model reference control design methods fail when the plant has one or more non-minimum phase zeros that are not included in the reference model, leading possibly to an unstable closed loop. This is a very serious problem for data-based control design methods, where the plant is typically unknown. In this paper, we extend the Virtual Reference Feedback Tuning method to non-minimum phase plants. This extension is based on the idea proposed in Lecchini and Gevers (2002) for Iterative Feedback Tuning. We present a simple two-step procedure that can cope with the situation where the unknown plant may or may not have non-minimum phase zeros.     

Learning nonlinear image manifolds by global alignment of local linear models

Appearance-based methods, based on statistical models of the pixel values in an image (region) rather than geometrical object models, are increasingly popular in computer vision. In many applications, the number of degrees of freedom (DOF) in the image generating process is much lower than the number of pixels in the image. If there is a smooth function that maps the DOF to the pixel values, then the images are confined to a low-dimensional manifold embedded in the image space. We propose a method based on probabilistic mixtures of factor analyzers to 1) model the density of images sampled from such manifolds and 2) recover global parameterizations of the manifold. A globally nonlinear probabilistic two-way mapping between coordinates on the manifold and images is obtained by combining several, locally valid, linear mappings. We propose a parameter estimation scheme that improves upon an existing scheme and experimentally compare the presented approach to self-organizing maps, generative topographic mapping, and mixtures of factor analyzers. In addition, we show that the approach also applies to finding mappings between different embeddings of the same manifold.     

Transcritical CO2 refrigeration cycle with ejector-expansion device

An ejector expansion transcritical CO₂ refrigeration cycle is proposed to improve the COP of the basic transcritical CO₂ cycle by reducing the expansion process losses. A constant pressure mixing model for the ejector was established to perform the thermodynamic analysis of the ejector expansion transcritical CO₂ cycle. The effect of the entrainment ratio and the pressure drop in the receiving section of the ejector on the relative performance of the ejector expansion transcritical CO₂ cycle was investigated for typical air conditioning operation conditions. The effect of different operating conditions on the relative performance of the ejector expansion transcritical CO₂ cycle was also investigated using assumed values for the entrainment ratio and pressure drop in the receiving section of the ejector. It was found that the COP of the ejector expansion transcritical CO₂ cycle can be improved by more than 16% over the basic transcritical CO₂ cycle for typical air conditioning operation conditions.     

Cultivation and engineering of a software metrics program

Abstract. This paper reports from a case study of an organization that implements a software metrics program to measure the effects of its improvement efforts. The program measures key indicators of all completed projects and summarizes progress information in a quarterly management report. The implementation turns out to be long and complex, as the organization is confronted with dilemmas based on contradictory demands and value conflicts. The process is interpreted as a combination of a rational engineering process in which a metrics program is constructed and put into use, and an evolutionary cultivation process in which basic values of the software organization are confronted and transformed. The analysis exemplifies the difficulties and challenges that software organizations face when bringing known principles for software metrics programs into practical use. The article discusses the insights gained from the case in six lessons that may be used by Software Process Improvement managers in implementing a successful metrics program.     

Parallel Globally Consistent Normal Orientation of Raw Unorganized Point Clouds

A mandatory component for many point set algorithms is the availability of consistently oriented vertex‐normals (e.g. for surface reconstruction, feature detection, visualization). Previous orientation methods on meshes or raw point clouds do not consider a global context, are often based on unrealistic assumptions, or have extremely long computation times, making them unusable on real‐world data. We present a novel massively parallelized method to compute globally consistent oriented point normals for raw and unsorted point clouds. Built on the idea of graph‐based energy optimization, we create a complete kNN‐graph over the entire point cloud. A new weighted similarity criterion encodes the graph‐energy. To orient normals in a globally consistent way we perform a highly parallel greedy edge collapse, which merges similar parts of the graph and orients them consistently. We compare our method to current state‐of‐the‐art approaches and achieve speedups of up to two orders of magnitude. The achieved quality of normal orientation is on par or better than existing solutions, especially for real‐world noisy 3D scanned data.     

Why are Tailored Messages More Effective? A Multiple Mediation Analysis of a Breast Cancer Screening Intervention

Past research has found that tailoring increases the persuasive effectiveness of a message. However, the observed effect has been small and the explanatory mechanism remains unknown. To address these shortcomings, a tailoring software program was created that personalized breast cancer screening pamphlets according to risk, health belief model constructs, and visual preference. Women aged 40 and older (N = 119) participated in a 2 (Tailored vs. Stock Message) × 2 (Charts/Graphs vs. Illustrated Visuals) × 3 (Nested Replications of the Visuals) experiment. Participants provided with tailored illustrated pamphlets expressed greater breast cancer screening intentions than those provided with other pamphlets. In a test of 10 different mediators, perceived message relevance was found to fully mediate the Tailoring × Visual interaction.     

The ETH‐MAV Team in the MBZ International Robotics Challenge

This study describes the hardware and software systems of the Micro Aerial Vehicle (MAV) platforms used by the ETH Zurich team in the 2017 Mohamed Bin Zayed International Robotics Challenge (MBZIRC). The aim was to develop robust outdoor platforms with the autonomous capabilities required for the competition, by applying and integrating knowledge from various fields, including computer vision, sensor fusion, optimal control, and probabilistic robotics. This paper presents the major components and structures of the system architectures and reports on experimental findings for the MAV‐based challenges in the competition. Main highlights include securing the second place both in the individual search, pick, and place the task of Challenge 3 and the Grand Challenge, with autonomous landing executed in less than 1 min and a visual servoing success rate of over for object pickups.