Abstract
This article presents a novel mixed-integer second-order cone programming model to increase the photovoltaic (PV) hosting capacity and optimize the operation of distribution networks. The problem considers voltage and reactive (Volt/VAr) control through the optimal operation of capacitors banks, substations' on-load tap changers, voltage regulators, and network reconfiguration with radial and closed-loop operation topologies. The proposed formulation considers voltage-dependent models for loads and capacitor banks. The objective function maximizes the PV hosting capacity of the network. Numerical experiments are carried out using the 33-node and the 85-node networks. Results demonstrate the effectiveness of the proposed formulation to increase the penetration of PV sources, especially when the closed-loop operation is allowed, together with network reconfiguration and Volt/VAr control.

Abstract
In the healthcare domain, datasets are often private and lack large amounts of samples, making it difficult to cope with the inherent patient data heterogeneity. As an attempt to mitigate data scarcity, generative models are being used due to their ability to produce new data, using a dataset as a reference. However, synthesis studies often rely on a 2D representation of data, a seriously limited form of information when it comes to lung computed tomography scans where, for example, pathologies like nodules can manifest anywhere in the organ. Here, we develop a 3D Progressive Growing Generative Adversarial Network capable of generating thoracic CT volumes at a resolution of 1283, and analyze the model outputs through a quantitative metric (3D Muli-Scale Structural Similarity) and a Visual Turing Test. Clinical relevance - This paper is a novel application of the 3D PGGAN model to synthesize CT lung scans. This preliminary study focuses on synthesizing the entire volume of the lung rather than just the lung nodules. The synthesized data represent an attempt to mitigate data scarcity which is one of the major limitations to create learning models with good generalization in healthcare.

Abstract
A method is presented to determine the appropriate number of photovoltaic panels that should be installed in an end-user photovoltaic installation to guarantee the supply of energy to the load during the hours of solar radiation, according to factors such as the installation area and global solar radiation. Solar radiation is predicted by approximating the daily distribution of global irradiance through a Gaussian function, which is subsequently corrected using a heuristic approach. Meteorological parameters are used as input data such as the daily solar insolation and the maximum global irradiance for each day; this last parameter is obtained through an expert system based on fuzzy logic that was programmed and trained with the data of ambient temperature and relative humidity that were obtained in the processing stage. Output from this expert system is the predicted values of maximum radiation obtained for each day for a selectable time interval. With the predicted solar radiation, the generation of electrical energy from the photovoltaic panels is calculated. The load is randomly modeled from a pattern of the energy demand of the building to be powered by the photovoltaic system. The number of photovoltaic panels needed is found with the information acquired in the previous stages and the information of the energy demand of the load and the installation area. The results are the number of solar panels that would be needed at all hours of the day from which the radiation prediction was made.

Abstract
Team automata describe networks of automata with input and output actions, extended with synchronisation policies guiding how many interacting components can synchronise on a shared input/output action. Given such a team automaton, we can reason over communication properties such as receptiveness (sent messages must be received) and responsiveness (pending receives must be satisfied). Previous work focused on how to identify these communication properties. However, automatically verifying these properties is non-trivial, as it may involve traversing networks of interacting automata with large state spaces. This paper investigates (1) how to characterise communication properties for team automata (and subsumed models) using test-free propositional dynamic logic, and (2) how to use this characterisation to verify communication properties by model checking. A prototype tool supports the theory, using a transformation to interact with the mCRL2 tool for model checking.

Abstract
Building rehabilitation is a reality, and all phases of rehabilitation work need to be efficient and sustainable. Current procedures for assessing construction conditions are time-consuming, laborious and expensive and pose threats to the health and safety of engineers, especially when inspecting locations not easy to access. In an initial step, a survey of the condition of the building is carried out, which subsequently implies the elaboration of a report of existing pathologies, intervention solutions and associated costs. This survey involves an inspection of the site (through photographs and videos). This work aims to detect and locate cracks defects in images of painted facade walls of buildings. A VGG16 pre-trained model was evaluated first on a public database with cracked and not cracked concrete surfaces and then on a private database of images of painted building facades with and without cracks. The predicted activation maps were analysed with Grad-CAM methods to validate the models’ prediction. The proposed model achieved 99% accuracy on the concrete public dataset and 78% on the building's facade private dataset. The limitations and the future works are identified. © 2022, University of Cantabria - Building Technology R&D Group. All rights reserved.

Abstract