Title: Two-Stage Stochastic Energy Scheduling for Multi-Energy Rural Microgrids Incorporating Irrigation Systems and Biomass Fermentation
Presenter: Li Zhengmao (Aalto University)
Abstract:
Rural regions often face energy challenges due to unstable renewable generation, limited grid access, and inefficient use of local biomass resources. This talk presents a novel framework for multi-energy rural microgrids (MERMs) that integrate electricity, heat, biomass, and irrigation systems. A detailed anaerobic biomass fermentation (ABF) model is developed to capture realistic biochemical conversion from polymer substrates to methane, while an irrigation model is formulated to represent the interaction between environmental and agricultural factors. To handle diverse uncertainties in renewable generation, load demands, and weather conditions, a two-stage stochastic optimization approach is proposed and solved efficiently via a Progressive Hedging (PH)-based decomposition algorithm. Simulation results demonstrate that the proposed model achieves up to 68.4% cost reduction and significantly improves system resilience and flexibility. The talk concludes with discussions on future directions, including large-scale rural microgrids and simplified linearized ABF modeling for real-time applications.
When/where: 27.11.2025 (14:00 - 15:00) / Vuorimiehentie 2 (Circular Raw Materials Hub), C104 Platinium
Open to the public. No registration required.
Title: Data assimilation for dynamic systems
Presenter: José Augusto F. Magalhães (Aalto University)
Abstract:
A data assimilation (DA) problem for a dynamic system refers to the construction of a set of equations given a differential description of the evolution of the system and a collection of observations related to quantities of interest. The quantity of interest is often the state variable of such a system denoted as signal, with inferences about the signal being obtained from indirect and noisy observations. DA theorists work within a rather mature field, typically analysing the governing differential equations as mathematical constructs inspired by real problems. DA practitioners have adopted a particular formulation of these equations to accommodate real-world observations, developing algorithms that turn their implementation into a feasible task across a wide range of computational settings. The first part of this talk provides an overview of the major developments in these two branches over the past 60-70 years of the field. The second part focuses on Monte Carlo applications in DA for dynamic systems, particularly a class of problems involving stochastic (and possibly high-dimensional) nonlinear models. For these problems, Monte Carlo methods offer significant advantages over conventional DA algorithms. We illustrate these ideas through a series of examples, presenting the problem and its solution in the context of bioreactors, reaction-diffusion systems, and atmospheric models.
When/where: 23.10.2025 (14:00 - 15:00) / Vuorimiehentie 2 (Circular Raw Materials Hub), C104 Platinium
Open to the public. No registration required.
Title:
Predictive control and feedback equilibrium-seeking for sustainable water resource recovery
Presenter:
Otacilio B. L. Neto (Aalto University)
Abstract:
In this presentation, we investigate automation solutions to transition wastewater treatment plants (WWTPs) into sustainable water resource recovery facilities (WRRFs), that is, plants that produce goods and energy using wastewater as a raw material. We investigate optimal control and game-theoretical methods (i) for repurposing the already existing infrastructure to these emerging objectives, and (ii) for promoting a wastewater-centered market through an interconnected network of WRRFs. Our findings contribute a step towards zero-waste water resource recovery infrastructures, with our study on networked systems also benefiting applications from different domains, such as smart grids and supply chain management.
When/where: 25.09.2025 (14:00 - 15:00) / Kemistintie 1, Room D311 (KE5)
Open to the public. No registration required.