In recent years, a lot of progress has been reported in the field of energy disaggregation, also referred to as Non-Intrusive Load Monitoring (NILM). Despite the fact that there are many studies focusing on the residential sector, there is considerably less research interest for the industrial sector. In this paper, we present a deep neural network based on Transformers, targeted towards capturing complex patterns in long sequences of data. The proposed transformer architecture employs 1D spatial convolutions in self-attention, and modifications inside the attention computations manage to reduce computational complexity without any loss in predictive accuracy. In order to evaluate the performance of the proposed deep learning architecture, a set of experiments has been conducted using a publicly available dataset. The experimental results indicate that the proposed model achieves better disaggregation accuracy compared to other state-of-the-art NILM models.

The reduction of road congestion requires intuitive urban congestion-control platforms that can facilitate transport stakeholders in decision making. Interactive ITS visual analytics tools can be of significant assistance, through their real-time interactive visualizations, supported by advanced data analysis algorithms. In this paper, an interactive visual analytics platform is introduced that allows the exploration of historical data and the prediction of future traffic through a unified interactive interface. The platform is backed by several data analysis techniques, such as road behavioral visualization and clustering, anomaly detection, and traffic prediction, allowing the exploration of behavioral similarities between roads, the visual detection of unusual events, the testing of hypotheses, and the prediction of traffic flow after hypothetical incidents imposed by the human operator. The accuracy of the prediction algorithms is verified through benchmark comparisons, while the applicability of the proposed toolkit in facilitating decision making is demonstrated in a variety of use case scenarios, using real traffic and incident data sets.

Tourism demand forecasting comprises an important task within the overall tourism demand management process since it enables informed decision making that may increase revenue for hotels. In recent years, the extensive availability of big data in tourism allowed for the development of novel approaches based on the use of deep learning techniques. However, most of the proposed approaches focus on short-term tourism demand forecasting, which is just one part of the tourism demand forecasting problem. Another important part is that most of the proposed models do not integrate exogenous data that could potentially achieve better results in terms of forecasting accuracy. Driven from the aforementioned problems, this paper introduces a deep learning-based approach for long-term tourism demand forecasting. In particular, the proposed forecasting models are based on the long short-term memory network (LSTM), which is capable of incorporating data from exogenous variables. Two different models were implemented, one using only historical hotel booking data and another one, which combines the previous data in conjunction with weather data. The aim of the proposed models is to facilitate the management of a hotel unit, by leveraging their ability to both integrate exogenous data and generate long-term predictions. The proposed models were evaluated on real data from three hotels in Greece. The evaluation results demonstrate the superior forecasting performance of the proposed models after comparison with well-known state-of-the-art approaches for all three hotels. By performing additional benchmarks of forecasting models with and without weather-related parameters, we conclude that the exogenous variables have a noticeable influence on the forecasting accuracy of deep learning models.

Accurately forecasting power generation in photovoltaic (PV) installations is a challenging task, due to the volatile and highly intermittent nature of solar-based renewable energy sources. In recent years, several PV power generation forecasting models have been proposed in the relevant literature. However, there is no consensus regarding which models perform better in which cases. Moreover, literature lacks of works presenting detailed experimental evaluations of different types of models on the same data and forecasting conditions. This paper attempts to fill in this gap by presenting a comprehensive benchmarking framework for several analytical, data-based and hybrid models for multi-step short-term PV power generation forecasting. All models were evaluated on the same real PV power generation data, gathered from the realisation of a small scale pilot site in Thessaloniki, Greece. The models predicted PV power generation on multiple horizons, namely for 15 min, 30 min, 60 min, 120 min and 180 min ahead of time. Based on the analysis of the experimental results we identify the cases, in which specific models (or types of models) perform better compared to others, and explain the rationale behind those model performances.

Incidents produce heavy congestion in large urban traffic networks and therefore real time information about them (e.g. location, timestamp, type) can be very useful for the drivers. An efficient way of gathering this type of information is through a crowd sourcing reporting system that multimodal travellers may utilise for providing information about various incidents they witness to other interconnected users in the same network. After the incoming traffic reports are evaluated, they can be shared to other travellers who are approaching the location of the reported incidents. Travelers can use the reported information for improving their mobility status. Collecting information using crowd sourcing techniques has implications and risks that need to be addressed. One of the most important challenges in this regard is the estimation of the reliability of the incoming information, usually related to individual user reputation. To this end, the exploitation of a reliability assessment system is of profound importance for assuring that only accurate information is shared between interconnected users. This paper introduces an innovative crowd sourcing information assessment mechanism for urban travellers. The purpose of the proposed probabilistic framework is to estimate if a user-generated report is true or false, given a set of static and dynamic parameters. The latter describe contextual conditions occurring at the time when an incident is reported. The proposed model takes into account the current location and speed of the reporting user due to their impact on the reliability of an incoming report. The proposed probabilistic model was evaluated in a simulation environment. Preliminary results show that, based on a set of rational assumptions, the estimated reliability decreases with the distance from the reported event and the speed of the reporting user. Based on the estimates that our model produces, a reliable true/false recommendation system can be devised for evaluating the user generated reports.

Residential load forecasting is one of the most important tasks of the overall supply management process in electrical grids, since it enables smart grid services such as demand response (DR). Hence, several approaches for accurate residential load forecasting have been proposed in the relevant literature. However, most of the existing methods focus on the forecasting performance and neglect other aspects of the problem like training time and model size (i.e. memory usage). In this paper, we introduce a new model for both short-term and day-ahead residential load forecasting. The model synthesizes an heterogeneous feature set, which is constituted by both automatically-selected lagged values from the load time series and manually-extracted temporal features. Then, the tree-based algorithm light gradient boosting machine (LGBM) is fed with the constructed feature set and used as a regression model. Finally, a data-lightweight strategy is used for retraining the proposed model, which leads to both high forecasting accuracy and low training times. The proposed model has been extensively evaluated on a large real-world residential load dataset. The experimental results indicate that the proposed model achieves both higher forecasting performance and lower training times and model sizes compared to state-of-the-art solutions.