General interactive fixed-charge piecewise-linear programming using tabu search
A theoretical framework and a practical algorithm are presented to solve discontinuous piecewise linear optimization problems. The descent algorithm which is elaborated uses active-set and projected gradient approaches. It is a generalization of the ideas used by Conn to deal with non-smoothness in the l{sub 1} exact penalty function. We address a general (non-concave) form of the fixed-charge problem which involves moreover (non-separable) interactive fixed costs, i.e. the fixed cost of undertaking a task depends upon which other tasks are also undertaken. The inescapable non-convexity feature of discontinuous functions gives rise to the existence of several local optima, most of which are not interesting as solutions to practical problems. A tabu-search heuristic is implemented within the above active-set piecewise linear optimization algorithm, in order to look for a global minimum. The resulting algorithm is applied to randomly-generated interactive fixed-charge transportation problems.
- OSTI ID:
- 36056
- Report Number(s):
- CONF-9408161--
- Country of Publication:
- United States
- Language:
- English
Similar Records
A piecewise-linear regression on the ASHRAE time-series data
New Tabu Algorithm for Positioning Mining Drillholes with Blocks Uncertainty