Linked data for enhanced energy management at district and building levels: an indicator-based methodology
- Autores: Yehong Li
- Directores de la Tesis: Sergio Vega Sánchez (dir. tes.), Raúl García Castro (codir. tes.)
- Lectura: En la Universidad Politécnica de Madrid ( España ) en 2017
- Idioma: español
- Tribunal Calificador de la Tesis: Juan Monjó Carrió (presid.), Benito Lauret Aguirregabiria (secret.), Sergio Rodríguez Trejo (voc.), David Martínez Espinosa (voc.), James O`Donnell (voc.)
- Programa de doctorado: Programa de Doctorado en Construcción y Tecnología Arquitectónicas por la Universidad Politécnica de Madrid
- Materias:
- Enlaces
- Tesis en acceso abierto en: Archivo Digital UPM
- Resumen
Energy efficiency at both the district and building levels is imperative for sustainable cities. The current initiatives for smart cities and communities, and the growing use of distributed energy resources, increasingly require the allocation and management of energy from building level to district level. Effective, integrated energy management at the district and building scales is a solution to improve multi-level energy efficiency, but such a solution is an information-driven process which requires the exchange and analysis of energy performance information and data gathered from different stakeholders. Due to the complexity of energy management at the district scale, there may be numerous potential stakeholders and a massive amount of data involved. When seeking to gain insights into such data for potential energy performance improvement and to address stakeholders’ performance goals, the challenges are threefold. Firstly, there is a challenge to identify and analyse stakeholders in the energy management context; secondly, it is essential to define a method to extract the key performance information and core insightful data detailing stakeholders’ concerns; thirdly, an interoperability problem exists in regard to exchanging and sharing cross-domain heterogenous data among various stakeholders.
In this thesis, a key performance indicator (KPI)-based, linked data1 methodology is proposed to systematically support the identification and analysis of stakeholders, the extraction of key performance information and master data that underpin stakeholders’ goals. It also aids the interchange of such information and data among the stakeholders, along with the exploitation and analysis of the cross-domain data for multi-level energy performance improvement. The proposed methodology has been developed using the following steps: ▪ Firstly, a three-task method was developed to identify stakeholders and conduct a prioritisation analysis regarding them; the latter is aimed at identifying the key stakeholders who take precedence in decision-making and achieving their performance goals.
▪ Secondly, a bi-index method was defined to select the KPIs that represent the key performance information and measure the progress towards stakeholders’ goals. KPIs should be selected among a pre-list of energy performance indicators (PIs) through stakeholders’ involvement, considering both their vote regarding PIs and their prioritisation. Thus, the selected KPIs can not only support the performance goals of stakeholders, but also balance their benefits.
▪ Thirdly, the selected KPIs were used to identify the cross-domain master data for energy performance analysis. The identified master data using KPIs ensure that the stakeholders’ concerns are underpinned and avoid the analysis of unnecessary data.
▪ Fourthly, an EM-KPI ontology2 was developed to facilitate the exchange of key performance information and master data among different stakeholders. The ontology describes not only the multi-level KPIs, but also the cross-domain master data; therefore, it supports both performance tracking and improvement analysis.
▪ Finally, the method for linked data generation was defined, using the EM-KPI ontology to realise data interchange and exploitation. Linked data resolve the data interoperability problem and facilitate data exploitation for performance problem identification and informed decision-making in relation to improvement measures.
The proposed methodology has been validated through a representative case study for a small district called Villa Solar. The district was the competition site of Solar Decathlon Europe 2012. It contains 18 solar houses, which are energy prosumers aiming to achieve net zero energy buildings, and several public service buildings, which are net energy consumers. The combined features of the buildings and their connections to a smart microgrid make it one of the best cases for district-scale energy management and suitable for the context of smart cities and communities. Both the buildings and the microgrid in the district were equipped with energy management systems and monitored from September 17th to 28th, 2012, but initially without information exchange among different stakeholders.
Through the use of the proposed methodology, six groups of key stakeholders were identified, 23 KPIs were selected, and a linked dataset was generated. The linked dataset generated validates the assertion that the ontology enables the interchange of the key performance information and master data among various stakeholders. The exploitation of the linked dataset helps to identify the key performance problems and make better decisions for improvement, compared to the methodology currently used in the district. Through stakeholder interaction and information exchange, the district has the potential to save at least 18.24% of the energy cost. The case study demonstrates the feasibility and benefits of the proposed methodology to extract, exchange and analyse the key performance information and master data, in order to improve multi-level energy performance and enhance energy management at both district and building levels.
