1741. Multi-fidelity modelling and optimisation for long-term public sector capacity planning
Invited abstract in session TA-4: Journal of Simulation. Computer Modelling and Simulation, stream OR Journals.
Tuesday, 8:30-10:00Room: Rupert Beckett LT
Authors (first author is the speaker)
| 1. | Graham Burgess
|
| STOR-i Centre for Doctoral Training, Lancaster University | |
| 2. | Luke Rhodes-Leader
|
| Management Science, Lancaster University | |
| 3. | Dave Worthington
|
| The Management School, Lancaster University | |
| 4. | Rob Shone
|
| Management Science, Lancaster University | |
| 5. | Dashi Singham
|
| Operations Research Department, Naval Postgraduate School |
Abstract
In long-term capacity planning, public sector organisations must meet the public need over time with tight financial budgets. Examples of such problems arise in housing, healthcare and prisons. The overall aim of this research is to investigate a multi-fidelity, integer-ordered simulation optimisation approach to long-term public sector capacity planning problems. To this end, we study a problem in housing. Here, decision makers must split their resources between emergency shelter and permanent social housing to alleviate homelessness. We present three models of differing fidelity for the homeless care system in California, USA. These are a fluid model, an analytical queueing model and a discrete-event simulation model. We then propose and solve a deterministic optimisation problem based on the fluid model. Using policy-based shape constraints, we minimise unsheltered homelessness while planning for long-term relief to the system via permanent housing. We then discuss ongoing work on multi-fidelity, integer-ordered simulation optimisation.
Keywords
- Capacity Planning
- Simulation
- OR/MS and the Public Sector
Status: accepted
Back to the list of papers