摘要
采用何种控制策略和如何实现控制方法是决定正弦波逆变电源输出波形质量和动态性能的主要因素。在详细分析逆变电源常用控制方法优缺点的基础上,提出了一种基于对角递归神经网络在线辨识自学习整定PID控制和重复调节相结合的新型复合控制策略,并在一台以DSP为核心控制器件的逆变电源装置上进行了实验论证。实验结果表明,该方法能同时实现正弦波逆变器的高精度稳态输出波形和快速动态响应性能,适用于感应电源、UPS不间断电源等需要高性能输出电压波形的场合。
Generally, the key determining sine wave inverter’s output waveform quality and dynamic characteristics is what control strategy is selected,and how to implement the control law . On the basis of analysis of the conventional strategies for the inverter,a novel multiple control strategy is put forward,which is based on diagonal recurrent neural network adaptive PID and repetitive control technique. the experiments on a inverter based on DSP were implemented. The results show that the system can achieve near perfect output voltage waveform, and fast transient responses. This control scheme is best suited for inductive power supply and UPS applications where high performance voltage waveforms are needed.
出处
《电工技术学报》
EI
CSCD
北大核心
2008年第4期87-92,118,共7页
Transactions of China Electrotechnical Society
基金
安徽省自然科学基金资助项目(2006KJ032B)
关键词
逆变器
对角递归神经网络
重复控制
复合控制
Inverter,diagonal recurrent neural network,repetitive control,multiple control
