摘要
针对电气化铁道牵引网的电气特性和无功补偿装置的现实条件,提出基于神经网络和专家系统相结合的晶闸管投切电容器复合控制方案。利用神经网络构造多路自适应噪声对消滤波器,并行地在线检测出牵引网的无功电流、有功电流和高次谐波电流;以无功电流差补值作为控制参量,并以高次谐波电流作为谐振保护信号,由专家系统给出投切无功补偿电容的控制指令。特性分析表明,该系统有效提高了系统功率因数,降低了无功补偿电容的投切频率,安全可靠性高,是牵引电网晶闸管投切电容进行无功补偿的优选控制方案。
In accordance with the features of traction power system and the condition of VAR compensation device, a thyristor switched capacitor hybrid control scheme based on neural network and expert system was proposed. First, a self-adaptive filter was constructed by neural network, which is able to detect reactive current and harmonics current on-line. Then, adopting reactive current margin and harmonics as control and protection variable, an expert system was formed to control thyristor switched capacitor. Analysis of compensation features indicates that the method is valid and feasible. The quality of VAR compensation is high with lower switching frequency. It is a better scheme for VAR compensation on traction power system.
出处
《中国电力》
CSCD
北大核心
2005年第8期7-9,共3页
Electric Power
基金
国家自然科学基金资助项目(50405034)
关键词
复合控制
晶闸管投切电容
无功补偿
hybrid control
thyristor switched capacitor （TSC）
VAR compensation
