Abstract
A single PDE models the economics and dynamics, in continuous time, of how a building space responds to a stochastic external temperature and to a heating or cooling system. A quadratic function models user discomfort (during intermittent occupation) and we assume a daily cycle of electricity prices (stepwise or continuous). We can rapidly compute the time-varying optimal temperature control rule (the precision of control varies optimally with variations in the cost of control). We can also rapidly compute a purely physical performance parameter (mean or variance) for any variable, over any region of the problem space, plus the time to first exit from that region.
Original language | English |
---|---|
Title of host publication | host publication |
Publication status | Published - 25 Aug 2009 |
Event | 20th International Symposium on Mathematical Programming (ISMP09) - Chicago, Marriott Hotel and Booth School of Business Duration: 24 Aug 2009 → 28 Aug 2009 |
Conference
Conference | 20th International Symposium on Mathematical Programming (ISMP09) |
---|---|
City | Chicago, Marriott Hotel and Booth School of Business |
Period | 24/08/09 → 28/08/09 |
Keywords
- Financial mathematics, partial differntial equations, stochastic dynamic optimisation, emperature control