Periodic Reporting for period 1 - RESONANCE (Replicable and Efficient Solutions for Optimal Management of Cross-sector Energy)
Période du rapport: 2023-01-01 au 2024-06-30
The RESONANCE project develops an innovative software framework that provides means for rapid development and plug-and-play deployment of standard-compliant Customer Energy Manager (CEM), Resource Manager (RM), and their aggregation solutions. The CEM, specified in the EN 50491-12 standard family, is the next-generation demand-side flexibility management (DSFM) solution in Europe. CEM is a software agent that automates DSFM by interacting with smart appliances (represented by RMs), aggregators, and the markets to maximize customer benefits. According to the new EN 50491-12-2 standard, CEMs are envisioned to 1) provide a more deterministic demand response, and 2) be able to optimize consumer benefits with respect to multiple incentives and optimization targets. To achieve this, there is a need for accurate models of flexible assets (smart appliances) and model predictive control techniques to automate the decision-making within the customer premises. The RESONANCE Framework will facilitate the adoption of CEMs as the next generation DSFM system by significantly reducing the development efforts and costs. This is achieved with 1) a standard-compliant and modular system architecture, and 2) an innovative modeling pipeline that combines automated machine learning (AutoML) with physics-based modeling to provide accurate and robust models of flexible assets with minimum effort. The project brings together 19 partners (including a cluster with 40 organizations) with inter-disciplinary expertise and forms a basis for a cross-sector energy ecosystem that significantly contributes to the mobilization of DSFM at a large scale. Large scale piloting in six member states with a variety of consumer sectors, flexible assets (e.g. electric vehicles, HVAC systems, and white goods), stakeholders, and market settings (including sector integration with district heating) is utilized for demonstrating and validating the scalability and replication potential of the solutions.
Main achievement 1 – Initial RESONANCE Framework
The main achievement of the first reporting period is the initial version of the RESONANCE Framework. The initial RESONANCE Framework consists of three catalogues of software services (i.e. Resource Manager, Customer Energy Manager and Aggregation and Market Integration catalogue), and two marketplaces (i.e. Data and Service marketplaces) supported with configuration and deployment tools.
Main achievement 2 – Initial Resource Manager Catalogue
The initial version of the Resource Manager Catalogue provides software libraries and tools for the development of standard-compliant Resource Managers (RM). The initial RM Catalogue is described in detail in D3.1 Initial Resource Manager Catalogue. RM is a software component that represents a flexible asset such an HVAC system or EV charging. In alignment with the EN 50491-12-2 standard, RMs must support advanced control mechanisms and deliver baseline load forecasts and flexibility data to the Customer Energy Manager. The RM Catalogue is created to simplify the development and deployment of RM solutions across various flexible assets.
Main achievement 3 – Initial Customer Energy Manager Catalogue: The initial version of the CEM Catalogue, described in detail in D4.1 Initial Customer Energy Manager Catalogue, provides software modules for rapid development and deployment of standard-compliant Customer Energy Managers. The CEM is an intelligent software agent designed to optimize the operation of flexible resources (via RMs) within buildings, enhancing benefits for consumers.an be easily customized to match the preferences and requirements of both consumers and prosumers.
Main achievement 4 – Initial Aggregation and Market Integration Catalogue
The main achievements of the first reporting period regarding aggregation and market integration catalogue was to study functional and non-functional requirements coming from the relevant use cases as well as the current implementations in different pilots.
Main achievement 5 – First phase pilots in six countries
Demand-side flexibility management systems (including RMs, CEMs, and aggregation platforms) were deployed into six countries with a total of 74 sites including large apartment buildings, industrial sites, commercial buildings, public buildings and large group of single-family houses. In total the number of people living or working in the pilot buildings was over 4100 in phase 1. The pilots demonstrated and validated the RESONANCE Framework according to the initial plans described in DoA. Please refer to section 1.1.3 for further a short summary of the main achievements and activities in each pilot and to the D6.1 First phase validation and guidelines for large-scale replication for a detailed description of each pilot.
The main achievement of the first reporting period is the initial version of the RESONANCE Framework. The initial RESONANCE Framework consists of three catalogues of software services (i.e. Resource Manager, Customer Energy Manager and Aggregation and Market Integration catalogue), and two marketplaces (i.e. Data and Service marketplaces) supported with configuration and deployment tools.
Main achievement 2 – Initial Resource Manager Catalogue
The initial version of the Resource Manager Catalogue provides software libraries and tools for the development of standard-compliant Resource Managers (RM). The initial RM Catalogue is described in detail in D3.1 Initial Resource Manager Catalogue. RM is a software component that represents a flexible asset such an HVAC system or EV charging. In alignment with the EN 50491-12-2 standard, RMs must support advanced control mechanisms and deliver baseline load forecasts and flexibility data to the Customer Energy Manager. The RM Catalogue is created to simplify the development and deployment of RM solutions across various flexible assets.
Main achievement 3 – Initial Customer Energy Manager Catalogue: The initial version of the CEM Catalogue, described in detail in D4.1 Initial Customer Energy Manager Catalogue, provides software modules for rapid development and deployment of standard-compliant Customer Energy Managers. The CEM is an intelligent software agent designed to optimize the operation of flexible resources (via RMs) within buildings, enhancing benefits for consumers.an be easily customized to match the preferences and requirements of both consumers and prosumers.
Main achievement 4 – Initial Aggregation and Market Integration Catalogue
The main achievements of the first reporting period regarding aggregation and market integration catalogue was to study functional and non-functional requirements coming from the relevant use cases as well as the current implementations in different pilots.
Main achievement 5 – First phase pilots in six countries
Demand-side flexibility management systems (including RMs, CEMs, and aggregation platforms) were deployed into six countries with a total of 74 sites including large apartment buildings, industrial sites, commercial buildings, public buildings and large group of single-family houses. In total the number of people living or working in the pilot buildings was over 4100 in phase 1. The pilots demonstrated and validated the RESONANCE Framework according to the initial plans described in DoA. Please refer to section 1.1.3 for further a short summary of the main achievements and activities in each pilot and to the D6.1 First phase validation and guidelines for large-scale replication for a detailed description of each pilot.
Modelling of flexible assets and baseline loads with an innovative hybrid approach
In WP3 we have developed innovative hybrid approaches for modelling flexible assets such as Heating, Ventilation and Air-conditioning (HVAC) systems. The key idea in hybrid modelling is to combine the best parts of machine Learning and physics-based modelling. While ML models often outperform physics-based models and require less effort, they demand extensive data, making it difficult to model building responses to control signals accurately.
Artificial Intelligence for automated demand response
In WP3 and WP4 initial methods for AI-based model predictive control have been developed. The idea is to utilize the innovative hybrid models to provide optimal control at three levels of the hierarchy: RM, CEM and Aggregation platform. The combination of hybrid modelling with MPC makes the approach magnitudes more data-efficient than current reinforcement learning solutions. This is important for demand-side flexibility management. Initial approaches have been demonstrated in the first phase pilots and improvements will be provided for phase 2. Scientific publications of the approaches demonstrated in the pilots are planned for phase 2.
Interoperability for demand side flexibility management
Based on the AIME methodology developed in the OMEGA-X project, conceptual synthesis and ontology mapping with selected reference ontologies have been performed to allow the creation of a customized RESONANCE ontology, tailored to fit the unique requirements of the use cases of the project’s pilots. This process ensures that the ontology effectively captures and represents the specific concepts, relationships, and semantics relevant to each pilot. The customized RESONANCE ontology is currently being developed together with an API to support semantic data exchange in the project.
Trust, security and privacy for demand-side flexibility management
The trust, security and privacy (TSP) is addressed in the project from multiple angles. Initially, along with the project system architecture the TSP perspective has been studied providing basic TSP analysis of the system architecture and TSP requirements. Developed use cases are studied from trust and security perspective and Data Privacy Impact Assessment (DPIA) of the use cases is in progress. These two efforts will provide additional view on the TSP requirements and will improve TSP posture of the developed system.
In WP3 we have developed innovative hybrid approaches for modelling flexible assets such as Heating, Ventilation and Air-conditioning (HVAC) systems. The key idea in hybrid modelling is to combine the best parts of machine Learning and physics-based modelling. While ML models often outperform physics-based models and require less effort, they demand extensive data, making it difficult to model building responses to control signals accurately.
Artificial Intelligence for automated demand response
In WP3 and WP4 initial methods for AI-based model predictive control have been developed. The idea is to utilize the innovative hybrid models to provide optimal control at three levels of the hierarchy: RM, CEM and Aggregation platform. The combination of hybrid modelling with MPC makes the approach magnitudes more data-efficient than current reinforcement learning solutions. This is important for demand-side flexibility management. Initial approaches have been demonstrated in the first phase pilots and improvements will be provided for phase 2. Scientific publications of the approaches demonstrated in the pilots are planned for phase 2.
Interoperability for demand side flexibility management
Based on the AIME methodology developed in the OMEGA-X project, conceptual synthesis and ontology mapping with selected reference ontologies have been performed to allow the creation of a customized RESONANCE ontology, tailored to fit the unique requirements of the use cases of the project’s pilots. This process ensures that the ontology effectively captures and represents the specific concepts, relationships, and semantics relevant to each pilot. The customized RESONANCE ontology is currently being developed together with an API to support semantic data exchange in the project.
Trust, security and privacy for demand-side flexibility management
The trust, security and privacy (TSP) is addressed in the project from multiple angles. Initially, along with the project system architecture the TSP perspective has been studied providing basic TSP analysis of the system architecture and TSP requirements. Developed use cases are studied from trust and security perspective and Data Privacy Impact Assessment (DPIA) of the use cases is in progress. These two efforts will provide additional view on the TSP requirements and will improve TSP posture of the developed system.