Periodic Reporting for period 1 - iFLEX (Intelligent Assistants for Flexibility Management)
Reporting period: 2020-11-01 to 2022-04-30
The iFLEX project aims at empowering the consumers by making it as easy as possible for them to participate in demand response. A core concept of the project is the iFLEX Assistant, a novel software agent that acts between consumer(s), and their energy systems, various stakeholders and external systems helping them to achieve mutual benefits through local energy management and DR. The iFLEX Assistants are designed to provide a common approach to enhance user experience, level of automation and personalization in a wide variety of DR and energy services. Because of different requirements of these services, the project provides a common software framework (i.e. iFLEX Framework) for developing application-specific iFLEX Assistants that are customized for the needs of particular service(s). The focus is especially on households and DR for supporting high penetration of renewables. In addition, there is a need for effective incentives and market structures that encourage consumers to invest in these innovative DR solutions. To this end, the iFLEX Assistants are customizable for different incentive and market mechanisms to allow exploitation of the solution in different countries and climatic regions, as well as, to enable A/B testing of different incentive and user engagement mechanisms with real-users. Although the focus is on electricity, the iFLEX project targets to overcome the current silo-approaches and provide holistic energy management that optimizes across various energy vectors. Co-creation with end-users is inherent in different project phases and coordinated by consumer organisation in the consortium. iFLEX validation is carried out with field pilots in three climatic regions.
The iFLEX project provides benefits for different users, both individual consumer/prosumers and professional users. The consumer/prosumer can be benefited from a smart energy management system, reduction in energy cost, and contribution to a more sustainable environment. Professional users, such as electrical system operators will be benefited from the flexibility provided by the complete iFLEX solution to operate their system in a more cost-effective method. Furthermore, each individual tangible result developed in this project can be used by professional users as part of different solutions to model/manage/interface with flexibility assets, markets, and end-users. The initial prototype of the iFLEX framework for energy & flexibility management has been implemented. The iFLEX Framework is a collection of libraries, tools and configuration scripts that provide means for the development and deployment of iFLEX Assistants into consumer/prosumer premises. The iFLEX Assistant is an innovative software agent solution that facilitates consumer participation in demand response. The iFLEX Assistant can learn consumer behaviour and the dynamics of their premises to provide optimal and personalized flexibility management for the consumer. Furthermore, the iFLEX Assistant provides consumers with natural and seamless ways for communicating their requirements and preference to tailor the flexibility management according to their needs. The iFLEX Framework consists of the following individual exploitable results: Resource Interface Module & Security Data Management, Hybrid Modelling and Flexibility Management, End-user Interface, and Aggregator/Market Interface. The iFLEX Assistant developed on top of the framework will empower consumers/prosumers by facilitating their participation in DR and energy markets. By automating the tedious tasks related to flexibility management while operating according to the individual end-user wishes (and providing them with a customised experience), the iFLEX Assistant tackles several barriers and promotes the increased use of DR and flexibility management in Europe and globally. While the iFLEX Framework is targeted at consumers and stakeholders that provide them with services (e.g. ESCOs, building automation companies, etc.) it also benefits other stakeholders in the power/energy system value chains. In addition to activating consumers, the main benefit provided to aggregators and utilities is the more accurate load and flexibility forecasting methods which improve the effectiveness of DR and holistic energy management services. Furthermore, the deterministic Demand-Side Flexibility Management (DSFM) methods make it feasible to utilize DSFM for continuous balancing in the power grid. The more deterministic DSFM can be utilized for different purposes and market settings, including TSO reserve markets, energy wholesale markets, imbalance settlements and reducing peaks in distribution networks, to name a few.
Short term load forecasting and flexibility modelling
The iFLEX project goes beyond the-state-of-the-art in short term load forecasting by developing hybrid approaches that combine machine learning with physics-based modelling. Combining the best parts of physics-based modelling and machine learning makes it possible to achieve accurate and robust models, which can be also replicated for new consumers with minimum efforts. Two novel hybrid modelling approaches have been investigated in the WP3. The more mature approach, documented in D3.1 utilizes machine learning for modelling the baseline consumption and physics-based models to modify the ML forecast according to the law of energy conversion. Physics based greybox models with few parameters learned from data are also applied for modelling the indoor temperature during the DR event and the total heating demand of the building. Scientific publication of this approach is planned for the next phase of the project. The second innovative approach studied in the project utilizes a physics-based simulator to train a common Artificial Neural Network (ANN) model with a large amount of data from different buildings. The idea is that this forces the ANN to learning the underlying physics and therefore provide more robust results when compared to typical ML approaches. This innovative approach is published as a scientific journal (Kannari et al., 2021).
Artificial Intelligence for automated decision making in flexibility management
The iFLEX project goes beyond the-state-of-the-art in AI-based consumer flexibility management by developing data-efficient and safe approaches and demonstrating them in operational environment. The hybrid Deep Reinforcement Learning approach developed in the project utilizes the consumer digital twins to plan and optimize control actions. This allows the solutions to be much more efficient with respect to the interaction with the real world (i.e. data-efficiency is similar to supervised learning instead of reinforcement learning). Robustness is provided with three layered control architecture that builds upon existing building and smart home control systems, extending them with new expert rules for verifying the action proposals made by the model-predictive control system. The initial prototype of this approach is presented in D3.7. A journal publication about the initial solution at TRL 5 is published in IEEE Access (Kiljander et al., 2021)
The iFLEX project goes beyond the-state-of-the-art in short term load forecasting by developing hybrid approaches that combine machine learning with physics-based modelling. Combining the best parts of physics-based modelling and machine learning makes it possible to achieve accurate and robust models, which can be also replicated for new consumers with minimum efforts. Two novel hybrid modelling approaches have been investigated in the WP3. The more mature approach, documented in D3.1 utilizes machine learning for modelling the baseline consumption and physics-based models to modify the ML forecast according to the law of energy conversion. Physics based greybox models with few parameters learned from data are also applied for modelling the indoor temperature during the DR event and the total heating demand of the building. Scientific publication of this approach is planned for the next phase of the project. The second innovative approach studied in the project utilizes a physics-based simulator to train a common Artificial Neural Network (ANN) model with a large amount of data from different buildings. The idea is that this forces the ANN to learning the underlying physics and therefore provide more robust results when compared to typical ML approaches. This innovative approach is published as a scientific journal (Kannari et al., 2021).
Artificial Intelligence for automated decision making in flexibility management
The iFLEX project goes beyond the-state-of-the-art in AI-based consumer flexibility management by developing data-efficient and safe approaches and demonstrating them in operational environment. The hybrid Deep Reinforcement Learning approach developed in the project utilizes the consumer digital twins to plan and optimize control actions. This allows the solutions to be much more efficient with respect to the interaction with the real world (i.e. data-efficiency is similar to supervised learning instead of reinforcement learning). Robustness is provided with three layered control architecture that builds upon existing building and smart home control systems, extending them with new expert rules for verifying the action proposals made by the model-predictive control system. The initial prototype of this approach is presented in D3.7. A journal publication about the initial solution at TRL 5 is published in IEEE Access (Kiljander et al., 2021)