Publications
-
Leveraging Auto-Distillation and Generative Self-Supervised Learning in Residual Graph Transformers for Recommender System
This paper introduces a cutting-edge method for enhancing recommender systems through the integration of generative self-supervised learning (SSL) with a Residual Graph Transformer. Our approach emphasizes the importance of superior data enhancement through the use of pertinent pretext tasks, automated through rationale-aware SSL to distill clear ways of how users and items interact. The Residual […]
-
Strategic Predictive Maintenance for Internet System Security and Risk Management: A Roadmap
Recent technological advancements have profoundly transformed companies and businesses, enabling them to achieve high levels of performance. Today, system and network infrastructures are crucial and indispensable, requiring continuous corrections and maintenance to ensure superior security, reliability, and availability. However, various risks related to software, hardware, and malicious threats can cause failures, attacks, data loss, service […]
-
EnergysavingsandcoverageoptimizationinedgeWiFisensordeployment for buildings: A multi-objective evolutionary approach
Edge sensor nodes are used to ensure informed decisions in several fields, including smart buildings, supply chain management, sustainability, mobile robotics in industry and logistics, and applications, including the Internet of Things (IoT). However, designing an optimal and cost-effective deployment of edge sensor nodes in a complex environment with different types of walls and interferences […]
-
Developmental approach of safety in ergonomics/human factors: insights of constructed safety in six work environments
This article brings together works on the notion of constructed safety in ergonomics, carried out over the last fifteen years. Firstly, we situate this approach to safety in relation to existing models previously developed (e.g. regulated or managed safety) in regard with the development of activity-centred ergonomics. We then present six research actions in activity-centred […]
-
Architecture proposal of a VR operating procedure system for construction training based on BIM and QHSE rules
This research fosters collaboration in developing VR training simulations for construction procedures using semantic web technology. It introduces SWRP, an intuitive, web-based graphic editor for QHSE rule editing, addressing the difficulties that construction experts face with SWRL. The main contribution is the VROPTS architecture, which integrates BIM, VR, and QHSE rules, as detailed in the […]
-
DuoBrain-HVAC: A Hybrid Autonomous Agent for Controlling HVAC Systems
The HVAC system consumes a significant portion of building energy, contributing to CO 2 emissions. This paper presents DuoBrain-HVAC, a hybrid deep reinforcement learning agent, designed to optimize energy use by controlling a discrete (ON/OFF) air handling unit and a continuous (temperature setpoint) actuator. Simulated in an amphitheater using EnergyPlus, the hybrid agent combines DDQN […]
-
New heuristic order based on tree-decomposition for solving CSPs
The Constraint Satisfaction Problem(CSP)represents a pivotal area of study within artificial intelligence, offering a broad spectrum of applications from scheduling to resource allocation. Despite significant advancements in the development of algorithms and heuristics, the efficient resolution of large and intricate CSP instances continues to pose a considerable challenge to the research community. Traditional heuristics often […]
-
Energy optimization of building design using Genetic Algorithm and RMP ranking method
This study presents a simulation-based multi-objective optimization approach for identifying the optimal building envelope design to minimize energy consumption (for cooling and for heating), and carbon emissions. In this regard, the implemented methodology includes three basic phases: building energy simulation (EnergyPlus), optimization (non-dominant sorting genetic algorithm NSGA II), and multi-criteria decision analysis. The methodology is […]
-
WiFi-Visual Data Fusion For Indoor Robot Localization
In this paper, we propose a WiFi-Visual robot localization method for limiting the unbounded error of imageonly localization due to visual environment similarity. The localization problem is modeled as a classification problem based on the WiFi-Visual data collected at labelled positions. The heterogeneous WiFi-Visual data is harmonized by representing the WiFi features in image form […]
-
Conception and calibration of low-cost diagnostic tool for the evaluation of the thermal, acoustic and visual comforts in buildings
Abstract The comfort of indoor occupants is the primary factor infl uencing the energy consumption in a building. Even if it is a highly subjective notion, a better understanding of its links with the environmental parameters represents a promising way to optimize building energy consumption. The environmental physical parameters can be easily measured but usually […]
-
Analysis of one-dimensional structures using Lie groups
This article presents a formulation for one-dimensional structures. It uses the structure of Lie groups and the associated differential calculus to describe deformations and the dynamic equation of the structure. Three levels of equation setting are explored: level one is the most abstract where a single equation is obtained using the Lie algebra of the […]
-
Cooperation and synchronization of robotic tasks using a digital twin
Industry 4.0 marks a significant advancement in the manufacturing process by integrating advanced digital technologies. Robotics is one of the nine pillars defining the contours of Industry 4.0. These robots must be able to perform tasks safely, especially when working simultaneously in shared areas. However, robots only have a partial view of the production environment […]
Chargement en cours…
Erreur : tout le contenu a été chargé.