发明名称 METHOD AND SYSTEM FOR NEURAL NETWORK MODELING OF TURBINE COMPONENT
摘要 PROBLEM TO BE SOLVED: To solve the problem that, since there are possibilities that a large temperature gradient is generated depending on the loading rate, high stress and uneven thermal expansion are generated, and a turbine clearance is changed during the operation, there is a need to dynamically monitor and predict the turbine clearance condition and the expansion to operate a turbine at the most effective range. SOLUTION: The method for control the turbine clearance includes a step 704 of applying at least one operation parameter as an input to at least one neural network model; a step 706 of modeling the thermal expansion of at least one turbine component through the neural network model; and a step 708 of performing a control task based on at least part of the modeled thermal expansion of one or more turbine components. COPYRIGHT: (C)2009,JPO&INPIT
申请公布号 JP2009162231(A) 申请公布日期 2009.07.23
申请号 JP20090001237 申请日期 2009.01.07
申请人 GENERAL ELECTRIC CO <GE> 发明人 MINTO KARL DEAN;ZHANG JIANBO;KARACA ERHAN
分类号 F01D11/08;F01D25/00;F01K13/02;F02C7/00;F02C9/00;F02C9/28 主分类号 F01D11/08
代理机构 代理人
主权项
地址