Most studies about the solar forecasting topic do not analyze and exploit the temporal and spatial components that are inherent to such a task. Furthermore, they mostly focus just on precision and not on other meaningful features, such as flexibility and robustness. With the current energy production trends, where many solar panels are distributed across city rooftops, there is a need to manage all this information simultaneously and to be able to add and remove sensors as needed. Likewise, robust models need to be able to cope with (inevitable) sensor failure and continue producing reliable predictions. Due to all of this, solar forecasting models need to be as decoupled as possible from the number of data sources that feed them and their geographical distribution, enabling also the reusability of the models. This article contributes with a family of Deep Learning models for solar irradiance forecasting complying with the aforementioned features, i.e. flexibility and robustness. In the first stage, several Artificial Neural Networks are trained as a basis for predicting solar irradiance on several locations at the same time. Thereupon, a family of models that work with irradiance maps thanks to Convolutional Long Short-Term Memory layers is presented, obtaining forecast skills between 7.4% and 41% (depending on the location and horizon) compared to the baseline. The latter family comes with flexibility and robustness features, which are required in large-scale Intelligent Environments, such as Smart Cities. Working with irradiance maps means that new sensors can be added (or removed) as needed, without requiring rebuilding the model. Experiments carried out show that sensor failures have a mild impact on the prediction error for several forecast horizons.

The Vehicle Energy Consumption calculation Tool (VECTO) is used for the official calculation and reporting of CO2 emissions of HDVs in Europe. It uses certified input data in the form of energy or torque loss maps of driveline components and engine fuel consumption maps. Such data are proprietary and are not disclosed. Any further analysis of the fleet performance and CO2 emissions evolution using VECTO would require generic inputs or reconstructing realistic component input data. The current study attempts to address this issue by developing a process that would create VECTO input files based as much as possible on publicly available data. The core of the process is a series of models that calculate the vehicle component efficiency maps and produce the necessary VECTO input data. The process was applied to generate vehicle input files for rigid trucks and tractor-trailers of HDV Classes 4, 5, 9 and 10. Subsequently, evaluating the accuracy of the process, the simulation results were compared with reference VECTO results supplied by various vehicle manufacturers. The results showed that the difference between simulated and reference CO2 emissions was on average-0.6% in the Long Haul cycle and 1% in the Regional Delivery. Such a process could be a powerful tool for calculating HDV CO2 emissions for development and analysis purposes, e.g. for new vehicle prototypes or multistage vehicles, and for creating VECTO equivalent models that can be used to assess alternative operating conditions and mission profiles of existing vehicle models. The methodology was applied for creating input of various components in the US tool for HDV certification, GEM, for generic sample-vehicle models available.

There is increasing evidence suggesting that real-world fuel consumption and CO2 improvements in the last decade have been much less than those measured during type-approval tests. Scientific studies have found that the offset between officially reported values and real-world vehicle CO2 emissions in Europe has constantly increased over the last years. The difference between officially reported and actual CO2 emissions of vehicles has three main implications: (i) it undermines the effectiveness of CO2 regulations in reducing greenhouse gas emissions in Europe; (ii) it distorts competition between vehicle manufacturers; (iii) it undermines innovation. As a fundamental step to deal with this issue, the European Commission has already replaced the old and outdated test procedure used so far in the emission type-approval of vehicles with the worldwide harmonized light vehicles test procedure (WLTP). Being a lab-based test procedure, the WLTP, by its nature, can only cover part of the CO2 gap. There is therefore increasing pressure to integrate the current type-approval system with additional measures based on real-world measurements. One of the options under discussion is to use the CO2 emissions measured during the real driving emission test. The objective of the present paper is to assess the validity of this proposal and to propose other possible ways to deal with the CO2/fuel consumption gap. In particular, the paper presents experimental evidence on the variability of the CO2/fuel consumption of a vehicle, questioning the idea that a single central estimate of these quantities may be sufficient.

No abstract

The Vehicle Energy Consumption calculation Tool (VECTO) is used in Europe for calculating standardised energy consumption and CO2 emissions from Heavy-Duty Trucks (HDTs) for certification purposes. The tool requires detailed vehicle technical specifications and a series of component efficiency maps, which are difficult to retrieve for those that are outside of the manufacturing industry. In the context of quantifying HDT CO2 emissions, the Joint Research Centre (JRC) of the European Commission received VECTO simulation data of the 2016 vehicle fleet from the vehicle manufacturers. In previous work, this simulation data has been normalised to compensate for differences and issues in the quality of the input data used to run the simulations. This work, which is a continuation of the previous exercise, focuses on the deeper meaning of the data received to understand the factors contributing to energy and fuel consumption. Fuel efficiency distributions and energy breakdown figures were derived from the data and are presented in this work. Correlation formulas were produced to calculate the energy loss contributions of individual components and resistances (air drag, rolling resistance, axle losses, gearbox losses, etc.) over the Regional Delivery and Long Haul cycles, given a limited number of input parameters such as vehicle characteristics and average component efficiencies. Default values and meaningful ranges of variation of these parameters obtained from the data of the fleet are also reported in this work. The importance of air drag and rolling resistance losses are highlighted since these losses account for about 70% of the energy consumed downstream the engine. Finally, based on the correlation formulas to calculate the individual energy losses, a method is presented that calculates the final energy consumption and CO2 emissions for all the regulated HDTs classes and that does not rely on the use of VECTO.