artificial intelligence blockchain building energy management building information modelling data analytics data mining energy efficiency Energy efficient building energy management energy performance indicator flexibility measurement industry 4.0 industry 4.0 maturity assessment Internet of Things knowledge management linked data machine learning manufacturing Ontology ontology engineering ontology population product configuration product lifecycle management production planning and control requirement engineering resource efficient manufacturing smart cities Supply Chain 4.0 sustainability virtual engineering
2023 |
Wicaksono Hendro; Trat, Martin.; Bashyal Atit; Boroukhian Tina; Felder Mine; Ahrens Mischa; Bender Janek; Groß Sebastian; Steiner Daniel; July Christoph; Dorus Christoph; Zoerner Thorsten Artificial Intelligence Enabled Dynamic Demand Response System for Maximizing the Use of Green Electricity in Production Processes, Robotics and Computer-Integrated Manufacturing Journal Article Forthcoming Robotics and Computer-Integrated Manufacturing, Forthcoming. Abstract | BibTeX | Tags: artificial intelligence, Artificial neural network, machine learning, Ontology, production scheduling, reinforcement learning @article{Wicaksono2023, title = {Artificial Intelligence Enabled Dynamic Demand Response System for Maximizing the Use of Green Electricity in Production Processes, Robotics and Computer-Integrated Manufacturing}, author = {Wicaksono, Hendro; Trat, Martin.; Bashyal, Atit; Boroukhian, Tina; Felder, Mine; Ahrens, Mischa; Bender, Janek; Groß, Sebastian; Steiner, Daniel; July, Christoph; Dorus, Christoph; Zoerner, Thorsten}, year = {2023}, date = {2023-12-30}, journal = {Robotics and Computer-Integrated Manufacturing}, abstract = {The transition towards renewable electricity provides opportunities for manufacturing companies to save electricity costs through participating in demand response programs. End-to-end implementation of demand response systems focusing on manufacturing power consumers is still challenging due to multiple stakeholders and subsystems that generate a heterogeneous and large amount of data. This work develops an approach utilizing artificial intelligence for a demand response system that optimizes industrial consumers’ and prosumers’ production-related electricity costs according to time-variable electricity tariffs. It also proposes a semantic middleware architecture that utilizes an ontology as the semantic integration model for handling heterogeneous data models between the system's modules. This paper reports on developing and evaluating multiple machine learning models for power generation forecasting and load prediction, and also mixed-integer linear programming as well as reinforcement learning for production optimization considering dynamic electricity pricing represented as Green Electricity Index (GEI). The experiments show that the hybrid auto-regressive long-short-term-memory model performs best for solar and convolutional neural networks for wind power generation forecasting. Random forest, k-nearest neighbors, ridge, and gradient-boosting regression models perform best in load prediction in the considered use cases. Furthermore, this research found that the reinforcement-learning-based approach can provide generic and scalable solutions for complex and dynamic production environments. Additionally, this paper presents the validation of the developed system in the German industrial environment, involving a utility company and two small to medium-sized manufacturing companies. It shows that the developed system benefits the manufacturing company that implements fine-grained process scheduling most due to its flexible rescheduling capacities.}, keywords = {artificial intelligence, Artificial neural network, machine learning, Ontology, production scheduling, reinforcement learning}, pubstate = {forthcoming}, tppubtype = {article} } The transition towards renewable electricity provides opportunities for manufacturing companies to save electricity costs through participating in demand response programs. End-to-end implementation of demand response systems focusing on manufacturing power consumers is still challenging due to multiple stakeholders and subsystems that generate a heterogeneous and large amount of data. This work develops an approach utilizing artificial intelligence for a demand response system that optimizes industrial consumers’ and prosumers’ production-related electricity costs according to time-variable electricity tariffs. It also proposes a semantic middleware architecture that utilizes an ontology as the semantic integration model for handling heterogeneous data models between the system's modules. This paper reports on developing and evaluating multiple machine learning models for power generation forecasting and load prediction, and also mixed-integer linear programming as well as reinforcement learning for production optimization considering dynamic electricity pricing represented as Green Electricity Index (GEI). The experiments show that the hybrid auto-regressive long-short-term-memory model performs best for solar and convolutional neural networks for wind power generation forecasting. Random forest, k-nearest neighbors, ridge, and gradient-boosting regression models perform best in load prediction in the considered use cases. Furthermore, this research found that the reinforcement-learning-based approach can provide generic and scalable solutions for complex and dynamic production environments. Additionally, this paper presents the validation of the developed system in the German industrial environment, involving a utility company and two small to medium-sized manufacturing companies. It shows that the developed system benefits the manufacturing company that implements fine-grained process scheduling most due to its flexible rescheduling capacities. |
2018 |
Howell Shaun; Wicaksono, Hendro; Yuce Baris; McGlinn Kris; Rezgui Yacine User Centered Neuro-Fuzzy Energy Management Through Semantic-Based Optimization Journal Article IEEE Transactions on Cybernetics, pp. 1-15, 2018, ISSN: 2168-2267. Abstract | Links | BibTeX | Tags: Artificial neural network, building energy management, data mining, Fuzzy logic, Genetic algorithm, middleware, Ontology, optimization, semantic web, WebGL @article{Howell2018, title = {User Centered Neuro-Fuzzy Energy Management Through Semantic-Based Optimization}, author = {Howell, Shaun; Wicaksono, Hendro; Yuce, Baris; McGlinn, Kris; Rezgui, Yacine}, url = {https://ieeexplore.ieee.org/document/8412214/}, doi = {10.1109/TCYB.2018.2839700}, issn = {2168-2267}, year = {2018}, date = {2018-07-19}, journal = {IEEE Transactions on Cybernetics}, pages = {1-15}, abstract = {This paper presents a cloud-based building energy management system, underpinned by semantic middleware, that integrates an enhanced sensor network with advanced analytics, accessible through an intuitive Web-based user interface. The proposed solution is described in terms of its three key layers: 1) user interface; 2) intelligence; and 3) interoperability. The system's intelligence is derived from simulation-based optimized rules, historical sensor data mining, and a fuzzy reasoner. The solution enables interoperability through a semantic knowledge base, which also contributes intelligence through reasoning and inference abilities, and which are enhanced through intelligent rules. Finally, building energy performance monitoring is delivered alongside optimized rule suggestions and a negotiation process in a 3-D Web-based interface using WebGL. The solution has been validated in a real pilot building to illustrate the strength of the approach, where it has shown over 25% energy savings. The relevance of this paper in the field is discussed, and it is argued that the proposed solution is mature enough for testing across further buildings.}, keywords = {Artificial neural network, building energy management, data mining, Fuzzy logic, Genetic algorithm, middleware, Ontology, optimization, semantic web, WebGL}, pubstate = {published}, tppubtype = {article} } This paper presents a cloud-based building energy management system, underpinned by semantic middleware, that integrates an enhanced sensor network with advanced analytics, accessible through an intuitive Web-based user interface. The proposed solution is described in terms of its three key layers: 1) user interface; 2) intelligence; and 3) interoperability. The system's intelligence is derived from simulation-based optimized rules, historical sensor data mining, and a fuzzy reasoner. The solution enables interoperability through a semantic knowledge base, which also contributes intelligence through reasoning and inference abilities, and which are enhanced through intelligent rules. Finally, building energy performance monitoring is delivered alongside optimized rule suggestions and a negotiation process in a 3-D Web-based interface using WebGL. The solution has been validated in a real pilot building to illustrate the strength of the approach, where it has shown over 25% energy savings. The relevance of this paper in the field is discussed, and it is argued that the proposed solution is mature enough for testing across further buildings. |
2017 |
McGlinn, Kris; Yuce, Baris; Wicaksono, Hendro; Howell, Shaun; Rezgui, Yacine Usability evaluation of a web-based tool for supporting holistic building energy management Journal Article Automation in Construction, 84 , pp. 154 - 165, 2017. Abstract | Links | BibTeX | Tags: Artificial neural network, BEMS, Fuzzy logic, Genetic algorithm, IFC, Information visualisation, Ontology @article{MCGLINN2017154, title = {Usability evaluation of a web-based tool for supporting holistic building energy management}, author = {Kris McGlinn and Baris Yuce and Hendro Wicaksono and Shaun Howell and Yacine Rezgui}, url = {https://www.sciencedirect.com/science/article/pii/S0926580516303545}, doi = {https://doi.org/10.1016/j.autcon.2017.08.033}, year = {2017}, date = {2017-03-31}, journal = {Automation in Construction}, volume = {84}, pages = {154 - 165}, abstract = {This paper presents the evaluation of the level of usability of an intelligent monitoring and control interface for energy efficient management of public buildings, called BuildVis, which forms part of a Building Energy Management System (BEMS.) The BEMS ‘intelligence’ is derived from an intelligent algorithm component which brings together ANN-GA rule generation, a fuzzy rule selection engine, and a semantic knowledge base. The knowledge base makes use of linked data and an integrated ontology to uplift heterogeneous data sources relevant to building energy consumption. The developed ontology is based upon the Industry Foundation Classes (IFC), which is a Building Information Modelling (BIM) standard and consists of two different types of rule model to control and manage the buildings adaptively. The populated rules are a mix of an intelligent rule generation approach using Artificial Neural Network (ANN) and Genetic Algorithms (GA), and also data mining rules using Decision Tree techniques on historical data. The resulting rules are triggered by the intelligent controller, which processes available sensor measurements in the building. This generates ‘suggestions’ which are presented to the Facility Manager (FM) on the BuildVis web-based interface. BuildVis uses HTML5 innovations to visualise a 3D interactive model of the building that is accessible over a wide range of desktop and mobile platforms. The suggestions are presented on a zone by zone basis, alerting them to potential energy saving actions. As the usability of the system is seen as a key determinate to success, the paper evaluates the level of usability for both a set of technical users and also the FMs for five European buildings, providing analysis and lessons learned from the approach taken.}, keywords = {Artificial neural network, BEMS, Fuzzy logic, Genetic algorithm, IFC, Information visualisation, Ontology}, pubstate = {published}, tppubtype = {article} } This paper presents the evaluation of the level of usability of an intelligent monitoring and control interface for energy efficient management of public buildings, called BuildVis, which forms part of a Building Energy Management System (BEMS.) The BEMS ‘intelligence’ is derived from an intelligent algorithm component which brings together ANN-GA rule generation, a fuzzy rule selection engine, and a semantic knowledge base. The knowledge base makes use of linked data and an integrated ontology to uplift heterogeneous data sources relevant to building energy consumption. The developed ontology is based upon the Industry Foundation Classes (IFC), which is a Building Information Modelling (BIM) standard and consists of two different types of rule model to control and manage the buildings adaptively. The populated rules are a mix of an intelligent rule generation approach using Artificial Neural Network (ANN) and Genetic Algorithms (GA), and also data mining rules using Decision Tree techniques on historical data. The resulting rules are triggered by the intelligent controller, which processes available sensor measurements in the building. This generates ‘suggestions’ which are presented to the Facility Manager (FM) on the BuildVis web-based interface. BuildVis uses HTML5 innovations to visualise a 3D interactive model of the building that is accessible over a wide range of desktop and mobile platforms. The suggestions are presented on a zone by zone basis, alerting them to potential energy saving actions. As the usability of the system is seen as a key determinate to success, the paper evaluates the level of usability for both a set of technical users and also the FMs for five European buildings, providing analysis and lessons learned from the approach taken. |