The interest in fisheye cameras has recently risen in the autonomous vehicles field, as they are able to reduce the complexity of perception systems while improving the management of dangerous driving situations. However, the strong distortion inherent to these cameras makes the usage of conventional computer vision algorithms difficult and has prevented the development of these devices. This paper presents a methodology that provides real-time semantic segmentation on fisheye cameras leveraging only synthetic images. Furthermore, we propose some Convolutional Neural Networks(CNN) architectures based on Efficient Residual Factorized Network(ERFNet) that demonstrate notable skills handling distortion and a new training strategy that improves the segmentation on the image borders. Our proposals are compared to similar state-of-the-art works showing an outstanding performance and tested in an unknown real world scenario using a fisheye camera integrated in an open-source autonomous electric car, showing a high domain adaptation capability.

No abstract

Since the advent of the microgrid (MG) concept, almost two decades ago, the energy sector has evolved from a centralized operational approach to a distributed generation paradigm challenged by the increasing number of distributed energy resources (DERs) mainly based on renewable energy. This has encouraged new business models and management strategies looking for a balance between energy generation and consumption, and promoting an efficient utilization of energy resources within MGs and minimizing costs for the market participants. In this context, this paper introduces an efficient management strategy, which is aimed at obtaining a fair division of costs billed by the utilities, without relying on a centralized utility or MG aggregator, through the design of a local event-based energy market within the MG. This event-driven MG energy market operates with blockchain (BC) technology based on smart contracts for electricity transactions to both guarantee veracity and immutability of the data and automate the transactions. The event-based energy market approach focuses on two of the design limitations of BC, namely the amount of information to be stored and the computational burden, which are significantly reduced while maintaining a high level of performance. Furthermore, the prosumer data is obtained by using IEC 61850 standard-based commands within the BC framework. By doing so, the system is compatible with any device irrespective of the manufacturer implementing the IEC 61850 standard. The advantages of this management approach are considerable for: MG participants, in terms of financial benefits; the MG itself, as it can operate more independently from the main grid; and the grid since the MG becomes less unpredictable due to the internal energy exchanges. The proposed strategy is validated on an experimental setup employing low-cost devices.

This paper describes a practical approach to the transformation of Base Transceiver Stations (BTSs) into scalable and controllable DC Microgrids in which an energy management system (EMS) is developed to maximize the economic benefit. The EMS strategy focuses on efficiently managing a Battery Energy Storage System (BESS) along with photovoltaic (PV) energy generation, and non-critical load-shedding. The EMS collects data such as real-time energy consumption and generation, and environmental parameters such as temperature, wind speed and irradiance, using a smart sensing strategy whereby measurements can be recorded and computing can be performed both locally and in the cloud. Within the Spanish electricity market and applying a two-tariff pricing, annual savings per installed battery power of 16.8 euros/kW are achieved. The system has the advantage that it can be applied to both new and existing installations, providing a two-way connection to the electricity grid, PV generation, smart measurement systems and the necessary management software. All these functions are integrated in a flexible and low cost HW/SW architecture. Finally, the whole system is validated through real tests carried out on a pilot plant and under different weather conditions.

No abstract