Local Flexibility Markets (LFMs) play a crucial role in modern electrical grids as they can address grid congestion by efficiently exploiting demand-side flexibility through market-clearing algorithms and incentive signals. However, demo projects worldwide offer limited access to detailed information that enables a good understanding of market design, estimating computational requirements, and exploring different algorithms at multiple levels. Accordingly, This paper introduces ORFLEX, an open-source platform for rapid testing and deploying local flexibility markets. ORFLEX exhibits salient features, including scalability, agile deployment, and secure interactions by leveraging cloud and edge infrastructure and Functions as a Service (FaaS). The platform aims to provide a go-to guide for LFM implementation. The proposal was validated using a straightforward market with a hierarchical configuration involving a market operator, a spot market aggregator, and many flexumers. Experimental results revealed the systems’s ability to assist in large-scale deployments while increasing the number of participants on the demand side. The methodical simulation insights will help market designers and regulators to properly develop LFMs and analyze the impacts of different clearing and demand-side flexibility estimations.

The escalating energy demand and peak consumption periods during winter pose strenuous challenges. The rapid growth of electricity demand intensifies the risk of energy shortages. Strategic measures are imperative to mitigate the impact of winter peak periods on the grid and the increasing electrical load from conventional heating systems. To address such a challenge, this paper presents a methodology for determining the optimal size and operation of electric thermal storage (ETS) in multi-zone residential apartment buildings by forming an optimization problem. The economic implications of employing ETS with time-of-use (TOU) price signals and its impact on load shifting are explored. A synthetic energy consumption dataset of a residential building in Quebec, Canada, is utilized to evaluate the proposed method. Simulation results demonstrate that the judicious use of ETS, especially with dynamic electricity rates, enhances consumer economic savings by 66.54% and satisfies the energy demand.

Distributed generation and energy storage technologies have helped SmartGrid projects gain great momentum over the last decade. However, despite a large number of pilot and demonstration projects, low-level information is often unavailable. Therefore, tools for defining and building different operation scenarios are required. These tools can facilitate adopting novel approaches to multi-domain energy management. This paper proposes a distributed, flexible co-simulation framework to integrate simulators from separate domains and platforms. Particularly, the proposed scheme enables the development of hybrid thermal-electric systems for smart buildings. In this study, an object-oriented approach to modeling electrical thermal storage (ETS) units is also suggested. The evaluation process is carried out using real-world data. A case study is practiced by designing a residential agent that performs model predictive control (MPC) of residential heating load in the presence of ETS. The results show that proper integration of ETS into Home Energy Management Systems (HEMSs) can achieve economic savings of up to 45 %. The findings of this study demonstrate ETS's high potential for reducing customer bills while satisfying users' comfort. Furthermore, they recommend practical strategies for short-term planning of smart grids by increasing their flexibility based on ETS-integrated Demand Response (DR) programs. © 2022 IEEE.

Abstract As an essential feature of autonomous road vehicles, obstacle detection must be executed on a real‐time onboard platform with high accuracy. Cameras are still the most commonly used sensors in autonomous driving. Most detections using cameras are based on convolutional neural networks. In this regard, a recent teacher–student approach, called transfer learning, has been used to improve the neural network training process. This approach has only been used with a neural network acting as a teacher to the best of our knowledge. This paper proposes a novel way of improving training data based on attention transfer by getting the attention map from a human. The proposed method allows the dataset size reduction by 50%, which leads to up to a 60% decline in the training time. The experimental results indicate that the proposed method can enhance the F1‐score of the network by up to 10% in winter conditions.

The choice of explicative variables could influence the efficiency of power consumption prediction. Different techniques based on a sensitivity analysis permit to choose the best set of variables among all candidates. Therefore, studies introduce a different set of factors like the temperature to construct a useful prediction system. The goal of this paper is to define the best variable candidate that can efficiently describe power consumption. Accordingly, a model is developed to integrate both meteorological and complementary variables for load forecasting. We use, therefore, powerful tools based on, namely ANOVA, ANCOVA, and Backward Elimination to identify the significant factors that will be inserted in the Principal Component Analysis (PCA) for household and aggregated levels. Mainly, the objective of PCA is to minimize the dimension of the data-set and maximize the information entropy over distinct un-correlated principal extractions. Consequently, the application of polynomial regression to principal components for load forecasting conducts to better results providing a useful forecast by capturing the best explanatory variables.

In order to support smart grid strategies and home energy management systems, a crucial task is to be able to retrieve a well-approximated model that allows the implementation of advanced adaptive and predictive techniques. However, implementing control and energy efficiency approaches to the residential sector could be complex and filled with technical difficulties for deployment. Unlike commercial and industrial sectors, the residential environments lack information and would rely on machine learning methods to effectively control and predict the required variables. Thus, this work explores the use of the Genetic Algorithms (GA) to the task of estimating thermal parameters of an equivalent physical representation model of temperature behavior.

This paper explores the performance of an Advanced Driving Assistance System (ADAS) during navigation in urban traffic and a winter condition. The selected ADAS technology, Mobileye, has been integrated into a hydrogen electric vehicle. A set of three cameras (visible spectrum) has also been installed to give a surrounding view of the test vehicle. The tests were carried out during the dusk as well as in the night in winter condition. Using Matlab, the messages provided by Mobileye system have been analyzed. More than 2800 samples (short sequences of 5s Mobileye messages) have been processed and compared with the corresponding video samples recorded by the three cameras. In average, the selected ADAS device was able to provide 99% of true positive vehicle detection and classification, even in poor ambient lighting condition in winter. However, 72% of samples involving a pedestrian was correctly classified.

The integration of Fuel Cell (FC) technology in power systems as grid-tied or stand-alone source is a solution to accelerate the transition to cleaner energy production. Proton Exchange Membrane Fuel Cells (PEM-FC) can be used in residential applications where more effort should be employed to facilitate the integration of renewable energy sources. These PEM-FCs are generally used in association with a power-conditioning system to meet the specifications of residential loads. However, supplying alternative current loads generates low and high frequency ripples in the FC current and therefore, impacts negatively the lifetime of the FC. Particularly, thermostatic residential loads, e.g. baseboard heaters mainly used in Nordic countries, impose high and fast variations on the power profile which directly influences the FC performance. This paper focuses on the development of a control strategy for FC power conditioning system to supply residential loads. The proposed strategy minimizes the ripples of the fuel cell current with reduced use of energy storage compared to typical solutions. The proposed control strategy has been implemented in Field Programmable Gate Arrays (FPGA) and validated by simulations and experiments using thermostatic loads.

Detailed information about load behavior and home's occupancy is important to implement Home Energy Management Systems (HEMS) capable of reducing energy consumption while maintaining user comfort. This is why Non-intrusive Appliance Load Monitoring (NIALM) and Non-intrusive Occupancy Monitoring (NIOM) have an important role to play in the new context of smart grid. This paper shows the implementation of two algorithms for transient event detection and classification, which is the first key step of a NIOM process. The first method employs Wavelet transform for the feature extraction and Artificial Neural Networks for the classification problem. The second method is based on the theory of Matched Filters to achieve the transient event detection and classification. Experiments permitted to validate the proposed methods using a dataset of occupied residential building.

Desertification due to global warming has a negative impact on the agricultural production capacity of hundreds of thousands of people in the world. To remedy this situation, one of the best ideas is the use of renewable energy sources to pump water and irrigate the fields. The aim of this research is the management of renewable energy hybrid power system (RE-HPS) for agricultural applications. The RE-HPS consists of a Photovoltaic (PV) system and a Wind Turbine (WT). A lead-acid battery bank is used to increase the reliability of the power system. Three Power Management Strategies (PMSs) have been evaluated on their ability to meet the requirements of the pump and loads. The dynamic behavior of the hybrid system was tested under various wind speed, solar radiation, and load demand conditions. The wind speed and solar radiation data are based on real data from Dakar in Senegal. The simulation model was developed using MATLAB/Simulinkl Stateflow. The adopted approach consists in the development of an energy management algorithm capable not only of ensuring the regulation of the water level in the reservoir but also of satisfying the demand for the load and of protecting the batteries from overcharging and deep discharging.

Electric water heaters have been regarded as a load to be exploited in residential energy management applications due to their potential energy storage capacity. Nevertheless, the implementation of control strategies for water heater systems requires high-performance models which must be capable of reproducing a water heater's internal operation dynamics, especially their inner water temperature variations. Therefore, appropriate water heater model selection and design is a challenge. This paper presents physical parameter estimation methods for two typical water heater models based on experimental data. The first method is based on the evaluation of the on-off times of the heating elements when there is no water consumption; the second method uses an extended Kalman filter to estimate the model's states and physical parameters. Additionally, by leveraging these estimated parameters, a comparative study of the temperature estimation and electric power consumption in both water heater models has been produced and is included. Experimental results show that a precise estimation of the physical parameters in the model allows the water thermal process to accurately identify and predict future power and energy consumption values.

This paper presents a new methodology for house-hold electricity demand forecast using a hybrid method combining nonparametric model and time series analysis. The relationship between outdoor temperature and electricity demand is studied in order to develop an approach to modeling and analyzing the electricity demand in response to temperature changes. First, the kernel density estimation as a nonparametric method is used to examine the mentioned relationship and provide probability distribution of future possible demand values. Second, two autoregressive (AR) models are applied to forecast total power demand using different information sources. These sources comprise results of forecasting temperature/electricity demand relationship as well as nonparametric residual data. The performance of the methodology represented in this work is evaluated using a comparison to the results of an ARMAX model. The forecasting accuracy of the models is compared on the basis of mean absolute error (MAE) and mean absolute percentage error (MAPE) metrics. The forecasting results demonstrate that the hybrid model performs remarkably well and thus is more favorable than ARMAX model. This study employs real data for numerical analysis of proposed methods to increase the results efficacy.

Lead acid batteries are an important part of most energy storage applications. Managing these applications is complicated by the inherent difficulty of following the evolution of important battery parameters; more specifically, the State of Charge (SOC), State of Health (SOH) and Ampere-hour capacity (AHC). This article shows that the Two-Pulse Load Test is a simple and reliable method to determine SOC and SOH of a Lead acid storage battery. The Two-Pulse Load Test is based upon the hypotheses that certain voltage readings are linear in regard to the SOC and SOH, The results presented in this article validate these hypotheses for a 180 ampere hour lead acid battery. Although the battery characterisation requires some time consuming preliminary testing, SOC and SOH can thereafter be determined in less than 5 minutes.

Power electronics converters like Voltage source inverters (VSI) are normally employed as the power interface in many applications including the uninterruptible power systems (UPS) and the renewable energy systems. Both, stand alone operation and grid connected operation of VSI need an accurate voltage and frequency control (V-F control) to keep a stable operation and a good output power control. A correct V-F control is mandatory for the parallel operation of inverters like in UPS and micro-grid systems. This paper presents a control structure for voltage and frequency control of VSI based on the Adaptive Linear Neuron with Frequency Locked Loop (ADALINE-FLL). The proposed structure offers a good transient and steady state response. The proposed structure has been implemented in a Field Programmable Gate Array (FPGA) device. Simulation and experimental results demonstrate the validity of the proposition.

A global strategy which aims at providing a method to manage a PEMFC thermal behavior of an electric vehicle is proposed. The Energy Management System (EMS) plays the role of the power flow supervisor and provides the time-to-start the fuel cell based on the battery energy depletion during the trip. Taking into account the freezing operating conditions and given this time-to-start estimation, the fuel cell temperature management system calculates the most appropriate time to start heating the stack in order to reduce heat loss through the natural convection. When the stack is started, the exothermic reaction heat is used as a self-heating power source to further increase the stack temperature. The simulation results have shown that the proposed approach is efficient and can be implemented in real-time in Fuel Cell Electric Vehicles.

This work is based on the road-trip knowledge in order to reduce the energy consumption of a vehicle. The proposed algorithm computes the globally optimal speed profile and the associated energy consumption profile from the road-trip information. This profile is the lowest boundary of the energy required to move the vehicle. The energetic plan obtained is independent of the powertrain architecture, it can be used with any vehicle (conventional internal- combustion engine vehicle, hybrid electric vehicle and fuel cell vehicle). This optimization algorithm runs offline and (in future works) the results will be integrated into a real-time energy management system. Algorithm results are compared to experimental data obtained with a low-speed vehicle. For the same trip duration, the presented preliminary results show that the energy consumption is lower with the optimal driving cycle.