We present a prototype of a decentralized power trading system based on the use of distributed ledger technology. This sort of efficient, decentralized marketplace is needed to empower prosumers and make them first-class members of a smart, decentralized power grid in order to drive further renewable energy adoption. Unlike the bulk of previous work in this field, we focus on private permissioned distributed ledgers rather than conventional blockchains. The proposed solution is entirely independent of cryptocurrency, with an explicit design capability of being adapted piecemeal without any fundamental changes to the present regulatory environment. To be economical, efficient, and scalable, our prototype is based on a lean, Corda-based private permissioned distributed ledger. It allows for instant, automatic bidding on and trading of ‘power promises’ and the robust implementation of short-term, small-scale liquid electrical power futures. We demonstrate that the prototype performs well and presents several clear advantages over existing solutions based on conventional blockchains. Therefore, the proposed approach represents a promising, robust solution to the smart grid decentralized power trading problem.

Power quality disturbances (PQD) have a negative impact on power quality-sensitive equipment, often resulting in great financial losses. To prevent these losses, besides detecting a PQD on time, it is important to classify it, so that appropriate recovery procedures are employed. The majority of research employs machine learning model PQD classifiers on manually extracted features from simulated or real-world signals. This paper presents an end-to-end approach that circumvents the manual feature extraction and uses signals generated from mathematical voltage sag type formulas. We developed a configurable voltage sag generator that was used to form training and validation datasets. Based on the synthetic three-phase voltage signals, we trained several end-to-end LSTM classifiers that classify voltage sags according to ABC classification. The best-performing model achieved an accuracy of over 90% in the real-world dataset.

Automatic blood vessel segmentation from retinal images plays an important role in the diagnosis of many systemic and eye diseases, including retinopathy of prematurity. Current state-of-the-art research in blood vessel segmentation from retinal images is based on convolutional neural networks. The solutions proposed so far are trained and tested on images from few available retinal blood vessel segmentation datasets, which might limit their performance when given an image with retinopathy of prematurity signs. In this paper, the performance of three high-performing convolutional neural networks for retinal blood vessel segmentation is evaluated in the context of blood vessel segmentation on retinal images exhibiting retinopathy of prematurity. The main motive behind the study is to test if existing public datasets suffice to develop a high performing predictor that could assist an ophthalmologist in retinopathy of prematurity diagnosis. To do so, the authors created a dataset consisting solely of retinopathy of prematurity images. Retinal blood vessels are manually labeled in the images by two observers, including an ophthalmologist experienced in retinopathy of prematurity treatment. Experimental results show that all three solutions have difficulties in detecting the retinal blood vessels of infants due to a lower contrast compared to images from public datasets as demonstrated by significant drop in classification sensitivity. All three solutions segment alongside retinal also choroidal blood vessels which are not used to diagnose retinopathy of prematurity, but instead represent noise and are confused with retinal blood vessels. Through observations and numerical analysis of experimental data, it is concluded that existing solutions for retinal blood vessel segmentation need improvement oriented toward using more detailed datasets or deeper models in order to assist the ophthalmologist in retinopathy of prematurity diagnosis.

Distributed ledger technology (DLT) and blockchain-based energy trading solutions are often labeled as resource-hungry, excessive power consumers which consequently create a significant carbon footprint. Using this approach to energy trading, without simultaneously minimizing energy consumption, defeats the purpose of using such trading solutions to encourage renewable energy production and may even cancel out any ecological benefits. This paper demonstrates that it is possible to create a DLT-based energy trading system which provides security, transparency, autonomy, scalability, and decentralization, provided by using a DLT, but with significantly lower penalties than other previously known solutions. This is best illustrated through the fact that the carbon footprint per transaction of the solution presented in this paper is 0.19 grams of CO ₂ compared to 20 grams of CO ₂ created by single transaction of similar complexity on the Ethereum blockchain. This makes per-transaction carbon footprint of our system approximate 100 times smaller than that of the most commonly used Ethereum blockchain. We present the architecture and features of the proposed platform, as well as a thorough analysis of its performance, including power consumption and estimated carbon footprint. All experiments are done on a dedicated Beowulf cluster comprised of general-purpose computers. The cluster mimics a microgrid environment and presents a testing ground for real-world performance and power consumption analysis of a system used for trading energy predominantly produced from prosumers and their renewable sources.