【摘要】 电力系统发生故障时会在短时间内产生大量的警报信号,这给运行人员正确地分析系统状态带来了困难。为了解决这个问题,已提出了基于知识专家系统对电力系统进行警报处理和故障分析。但是,当电力系统规模很大时,由于设备数量很多,故障情况比较复杂,因而使知识库的建立和维护都很困难。针对这一点,本文是运用神经网络来对电力系统故障进行分析。试验结果证明,神经网络在用于故障分析时具有快速并行处理、模糊判断、自学习等优点,是一种可行的方法。更多还原

【Abstract】 A neural-network approach was using in building and testing in experimental inteligent alarm processor which could anelyze the multiple alarm associated with a power system and identify the particular probtem causing these alarms.The test results, althongh prelimary, suggest the new approach could be the long—sought answer for developing a generic intelligent alarm processor which could be inplemented by utilities with minimal customization effort.更多还原