摘要
提出了基于波形匹配的方法预测一次电流过零点,来提高高频电源输出功率。然而直接使用归一化互相关算法计算量大,速度慢。为了提高匹配速度,本文采用爬山法启发式搜索策略,实现了非遍历性搜索,并能快速找到最优匹配点。实验结果表明,爬山搜索算法性能优越,用于高频电源过程控制中是可行的,且具有快速性和较高的准确性,相比直接互相关算法,速度提高了一个数量级左右。
This paper proposes a method based on waveform matching to forecast the zero-crossing point of primary current and increase the output power of high-frequency power supply. However, it has complex computation and slow matching speed using normalized cross-correlation algorithm in waveform matching directly. To improve the speed of image matching,a heuristic hill-climbing searching strategy are presented in this paper,which realized non-ergodic searching and can be used to find the best matching point very quickly. The experimental results show that the algorithms presented enjoy an excellent performance and is feasible for high-frequency power supply process control with high precision and quick speed. Compared with direct cross-correlation algorithm,speed increases about an order of magnitude.
出处
《科技通报》
北大核心
2011年第5期720-725,共6页
Bulletin of Science and Technology
基金
国家自然科学基金(61034005)资助
关键词
控制理论与控制工程
爬山法
波形匹配
高频电源
control theory and control engineering
Hill-climbing waveform matching
high-frequency power supply
