Long-Term Joint Operation of Cascade Reservoirs Using Enhanced Progressive Optimality Algorithm and Dynamic Programming Hybrid Approach

Dynamic Programming (DP) is one of the most classical methods adopted for reservoir operation. It reduces computational efforts complex high-dimensional problems by piecewise dimensionality reduction and provides global optimums problems, but it suffers “curse dimensionality”. Progressive Optimality Algorithm (POA) has been used repeatedly in operation studies during last decades because alleviates dimensionality” DP good convergence extensive applicability. Nonetheless, POA encounters two difficulties multireservoir applications. One transfer interrupt problem that makes search procedure hard to achieve free allocation water between non-adjacent stages, another leads a low rate. In order overcome these deficiencies, this study made some enhancements proposed hybrid approach combining Enhanced (EPOA-DP) long-term cascade systems. EPOA-DP, EPOA employed improve quality solutions reduce effort two-stage solution. The was tested using real-world four-reservoir system ten-reservoir benchmark test example. results demonstrated outperforms both time