Optimal operation of flood control systems

Abstract

The management and control of multiple-reservoir flood control systems is studied. Our objective is to devise operating policies which minimize flood damages as determined by the flood peaks. Methodologies are presented that employ dynamic programming and stochastic dynamic programming for the optimal operation of multiple-reservoir flood control systems with deterministic and stochastic inflows, respectively. The methodologies are applied to a number of real-world problems involving river basins in Illinois and elsewhere. We then study the effects of parametrically varying a number of the input parameters. A Markov renewal flood synthesis model and a methodology for determining the optimal capacity of a new flood control reservoir are also presented.