Applications: Sustainability, Health Care
Applications: Sustainability, Health Care
Contributed Session
Time Slot: Thursday Morning
Room: AUD_B
Chair: Gabriele Zangara
CHP Systems Optimization in Presence of Time Binding Constraints
Time: 11:30
Caterina Tamburini (Optit), Bettinelli Andrea, Pozzi Matteo
As decarbonization becomes a global priority, there is a need to increase the efficiency of energy production. Combined Heat and Power (CHP) is an energy efficient technology that generates electricity and captures the heat that would be wasted otherwise in order to provide thermal energy, often used to feed district heating networks. Unit Commitment (UC) is a key problem in this context. The goal in UC is to determine a schedule for the machines that maximize the operative margin, satisfying a forecasted heat demand coming from a district heating network as well as functional and regulatory constraints deriving from system composition and placement. This gives rise to Unit Commitment (UC) problems.In this work, we formulate and solve Mixed Integer Linear Problems for the short-term optimization of real world CHP systems. We present some key features that make the problem hard to be solved in practice with particular attention to time binding constraints, e.g., limitations on the number of turn-ons over a period, minimum time on/off permanence, ramp-up and ramp-down constraints, heat storage, regulatory limitations defined over the legislative year. As for the latter, we study how Primary Energy Saving (PES) limitations introduced by the European Parliament Directives can be modeled in a MILP formulation and which are the critical aspects to consider. We then present a time-based decomposition metaheuristic for the optimization of mid-term (i.e., one year budgeting) UC problems, with computational results obtained over real world problem optimizations.A relevant factor for the accuracy of the energy production plan is the ability to forecast the heat demand of the network. Thus, we developed a forecasting module that, given the historical heat demand and other relevant data, automatically builds accurate prediction models. Specifically, each predictive model is trained and tuned using the most appropriate machine learning technique among a pool of state-of-the-art algorithms. The methodology proposed is implemented within a Decision Support System currently in use by plant managers of several CHP systems connected to district heating networks dislocated in Italy. We give a overlook of the DSS interfaces and relative use cases.Further, we will introduce current development directions, such as the support for the integration with the power markets sessions, and enhancement to deal with plants where machines are operated in series and operating temperatures become important decision variables along with the amount of energy produced.
Multi-objective optimisation for constructing cyclic appointment schedules for elective and urgent patients
Time: 11:50
Tine Meersman (Ghent University), Maenhout Broos
In this paper, we study the construction of a cyclic appointment schedule in an outpatient department. In particular, we determine the capacity distribution between elective and urgent patients and the scheduling of the time slots reserved for these patients such that the operational waiting times of elective and urgent patients are minimised. The proposed solution methodology devises a Pareto set of cyclic appointment schedules based on these waiting times with different capacity allocations for urgent patients. An approximation of the Pareto set of non-dominated schedules is obtained using a multi-objective archived simulated annealing heuristic. To accurately validate the cyclic appointment schedules, we incorporate operational decision-making via scheduling individual patients. To this end, we simulate operational variability, i.e., patient arrivals, no-show behaviour, punctuality and scan durations, based on real-life input data. The patients are assigned one-by-one using an online scheduling rule. Computational experiments are conducted with a real-life case study. We compare different appointment scheduling rules and discuss the impact of the capacity distribution between elective and urgent patients and the timing of urgent slots in the cyclic appointment schedules. The results show that the distribution of capacity between patient types, the timing of urgent slots and appointment rules all have significant impacts on patient waiting times. Appointment waiting times improve when urgent slots are spread equally over and throughout the days considered and when the Bailey–Welch rule is used to schedule patients. Trade-offs between elective and urgent waiting times resulting from different capacity distributions or slot timing are exemplified via a Pareto front. The proposed method outperforms relevant single-pass methodologies, and we demonstrate that its performance is strengthened thanks to the integrated optimisation of strategic, tactical and operational decisions.
Optimization for surgery department management: an application to a hospital in Naples
Time: 12:10
Andrea Mancuso (Department of Electrical Engineering and Information Technology, University of Naples “Federico II”), M. Boccia, A. Masone, F. Messina, A. Sforza, C. Sterle
Surgery department with its operating rooms represents the financial backbone of modern hospitals accounting the main part of a hospital cost and revenue. Therefore, maximizing its efficiency is of vital importance since it can have important implications on cost saving and patient satisfaction. In this context, Operations Research methodologies can play a relevant role supporting hospital executives in operating room management and surgery scheduling issues. In particular, great relevance has been given in literature to the Surgery Scheduling Problem. In its general form, it consists in determining a day, an operating room and a starting time of a set of surgeries. In this work, we address the Surgery Scheduling Problem faced by a local hospital of Naples. The aim of the hospital is to determine a surgery schedule capable of handling unforeseeable events (e.g., the arrival of an emergency) while maximizing the number of performed surgeries, according to some medical guidelines. This problem has been modelled by an original integer linear programming formulation that has been tested and validated on several instances derived from real data provided by the hospital. Finally, the proposed formulation can be used to simulate different surgery operating scenarios. The results of this simulation can be used to provide useful managerial insights for an efficient schedule of the hospital surgeries.
A flexible job shop scheduling model for Sustainable Manufacturing
Time: 12:30
Gabriele Zangara (Department of Mechanical, Energy and Management Engineering, University of Calabria), Rosita Guido, Giuseppina Ambrogio, Domenico Conforti
The research work was carried out with particular attention to the use of energy in production processes and the costs deriving from its consumption. The discussion was verticalized within a case study in the manufacturing sector, developing and implementing an optimization model to support the planning and scheduling of processing activities of heat exchangers with the aim of minimizing costs due to energy consumption. The formulated model was in fact tested on real data collected considering the products made at multinational corporate operating in the manufacturing industry. Following the formulation and testing of the customized optimization model, the results obtained showed that the cost due to the energy required for production is about 10% lower than the current production.