Operations Research Seminar - 23/3/2016

Wednesday, 23 March 2016, 2:30 p.m. - 4:30 p.m. 

Local: Sala de Seminários, Edifício VII

Faculdade de Ciências e Tecnologia, Quinta da Torre, Caparica

Lecturer: Manuel Matos (INESC TEC & FEUP)

Title: "Smart Grids and Network Management"

Abstract: The foreseen deployment of smart grids will increase overall energy efficiency and favor a smooth integration of large amounts of electricity coming from renewable sources, through the management of DER (distributed energy resources) such as distributed generation, storage and flexible loads. This opens new opportunities and brings new challenges for the different players in the field, namely for the Distribution System Operator. New roles for the DSO, new services, new functionalities and new tools are required to deal with this new reality. The talk will address these issues and some of the work developed at INESC TEC will be described, including new concepts and tools for distribution optimal power flow, state estimation, voltage control, forecasting, EV charging, flexibility estimation and planning. The motto is "new answers to old and new problems".

Short-bio: Manuel A. Matos was born in 1955 in Porto, Portugal. He is with the Faculty of Engineering of the University of Porto (FEUP), Portugal, since 1978 (Full Professor since 2000). He is also coordinator of the Centre for Power and Energy Systems of INESC TEC and President of the Scientific Council of INESC TEC. His research interests include classical and fuzzy modeling of power systems, reliability, optimization and decision-aid methods, with application to renewables’ integration, electric vehicles’ deployment and smart grids.


Lecturer: Carlos Henggeler Antunes (Deptº de Engª Electrotécnica e de Computadores, Universidade de Coimbra + INESC Coimbra)

Title: "Multiobjective evolutionary approaches for the management of residential energy resources"

Abstract: Dynamic tariffs are expected to become a relevant pricing scheme in the context of smart grids, with energy prices varying in short periods of time possibly with significant differences of magnitude. In this framework, active management of residential load can play an important role to help users optimizing the usage of end-use energy resources while minimizing the energy cost. Load management involves deciding which and when control actions should be implemented without jeopardizing the quality of energy services provided. These decisions are strongly influenced by energy costs, end-users’ preferences and requirements, potential dissatisfaction sensed by the end-user when the operation cycle of loads is changed, technical constraints, weather forecasts, the existence of local generation and storage systems. An evolutionary algorithm is presented to optimize the integrated usage of multiple residential energy resources considering a large set of management strategies. These energy resources include local generation, shiftable loads, thermostatically controlled loads and storage systems (either stationary or an electric vehicle). For the different groups of loads, customized solution encoding and operators are used since the detailed knowledge of the physical characteristics of the problem allows tailoring the algorithm to obtain effective results that can be implemented in practice. The multi-objective model considers as objective functions the minimization of the energy cost and the minimization of end-user’s dissatisfaction associated with management strategies.

Short-bio: Carlos Henggeler Antunes obtained his PhD degree Electrical Engineering, specialization in Optimization and Systems Theory, University of Coimbra, 1992. He has been Director of the R&D Institute INESC Coimbra 2003-2014. His research interests are multiple objective optimization, with mathematical programming algorithms and meta-heuristics, multiple criteria decision analysis, and applications in the energy sector, in particular energy efficiency and demand-side management.


Lecturer: João Murta Pina & António Pombo (Dep. Engenharia Electrotécnica, UNINOVA/CTS)

Title: "Otimização Multiobjectivo Para o Aumento da Fiabilidade em Redes de Distribuição Radiais com Incorporação de Geração Distribuída e Sistemas de Armazenamento"

Abstract: O planeamento das redes de distribuição de energia elétrica é considerado de extrema importância para o desenvolvimento de infraestruturas de elevada fiabilidade. A este nível, as empresas elétricas estão a ser confrontadas com objetivos contraditórios de clientes que requerem maior qualidade de serviço e dos que querem preços de energia mais baixos. Para competir neste contexto, é importante que as empresas elétricas estabeleçam um balanço entre os custos de investimento em equipamentos para aumentar a fiabilidade e o nível de fiabilidade alcançado com esse investimento. Assim, foram desenvolvidos modelos matemáticos para a integração das diversas variáveis minimizadas associadas à fiabilidade, deste problema de otimização multiobjectivo. 

Short-bio: António Pombo, Professor Adjunto no Departamento de Engenharia Electrotécnica, da Escola de Tecnologia de Setúbal, Instituto Politécnico de Setúbal (desde 1997). PhD em Engenharia Electrotécnica, pela FCT-UNL (2015), MSc em Gestão Global, pelo ISCTE-IUL (2000) e Licenciatura em Engenharia Electrotécnica, pelo IST-UL (1992). Experiência profissional na área técnico-comercial, na Merlin Gerin Portugal (1992-1993), na LTE - Grupo EDP na área de planeamento de redes de distribuição e na EDP Internacional na fiscalização, especificação de equipamentos e análise de cadernos de encargos (1993 - 1998). 

Dr. João Murta Pina is a senior researcher at Centre of Technology and Systems (CTS, line Energy Efficiency, Intelligent Control and Industrial Systems) and Assistant Professor in the Department of Electrical Engineering at Faculdade de Ciências e Tecnologia from Universidade Nova de Lisboa. He has background in Control Theory, and he holds a five year diploma, an MSc and a PhD in Electrical Engineering, the latter in Electrical Machines. He has been involved in applications of artificial intelligence (multiobjective optimization and automatic learning) to operation of industrial processes, whose concepts he has been extending to other fields, as energy efficiency improvement (by means of providing energy utilization feedback to users), optimization of renewable energy plants (through detailed modeling of elements and systems and correlation with environmental and economic conditions) and application of superconducting technologies to power systems (e.g. fault current limiters or energy storage). He has around 50 publications in peer reviewed journals and conferences.