| 摘要 |
A method and system for optimizing performance of a Kaplan turbine power generating unit are provided wherein an optimal "N-dimensional" virtual cam is populated with gate and blade positional settings producing maximum power output for a set of N operating parameters. The preferred parameters include head, flow, power generation level, physical situation of the unit, operating state of neighboring units, trash rack loss and a parameter indicative of cavitation (such as relative submersion level). When the system detects a unique set of operating conditions for which optimization is needed, an iterative optimization search is performed for the best gate and blade positional settings, which are then saved in memory (populating the cam matrix) for subsequent use when the same operating conditions are encountered. The optimal matrix divides parameters having ranges into range segments and treats other parameters as having discrete states. The optimization search is preferably performed while maintaining flow constant, and a flow characteristic search is performed to determine candidate gate and blade settings resulting in constant flow. <IMAGE> |
