电压空间矢量(磁链追踪)PWM控制研究与仿真

为了提高电机的功率因数,降低开关损耗,基于气隙磁通控制原理,以电磁量组合来逼近圆形磁链轨迹,而电压矢量的选择对应不同开关模式,因此构成电压矢量控制ＰＷＭ逆变器。利用Ｃ语言仿真,该法输出电压较一般ＳＰＷＭ逆变器提高１５％,每次状态切换只涉及一个元件,开关损耗降低,且模型简单,适用于各种ＰＷＭ调速装置。     

基于多边形边界约简的在线时序表示算法

时间序列是一类重要的复杂数据类型。首先分析了Douglas-Peucker和多边形约简时序表示算法。然后借鉴滑动窗口思想，提出了基于多边形边界约简的启发式在线时间序列表示算法。该算法具有多边形约简算法很好的近似质量，同时可以在固定数据缓冲空间内在线运算。此外，根据最大允许误差，算法用启发式搜索方法自动获取最佳分段数。在随机时间序列上仿真试验证明算法有很高的逼近质量和较低的计算复杂性。     

感应电机转子磁链自适应观测器的研究与仿真

介绍了自适应控制原理应用于系统状态观测器的方法,应用MRAS设计了感应电机的转子磁链观测器,并在矢量控制系统上对控制算法进行仿真,给出了仿真波形。     

Research on time series mining based on shape concept time warping

Time series is an important kind of complex data, while a growing attention has been paid to mining time series knowledge recently. Typically Euclidean distance measure is used for comparing time series. However, it may be a brittle distance measure because of less robustness. Dynamic time warp is a pattern matching algorithm based on nonlinear dynamic programming technique, however it is computationally expensive and suffered from the local shape variance. A modification algorithm named by shape DTW is presented, which uses linguistic variable concept to describe the slope feather of time series. The concept tree is developed by cloud models theory which integrates randomness and probability of uncertainty, so that it makes conversion between qualitative and quantitive knowledge. Experiments about cluster analysis on the basis of this algorithm, compared with Euclidean measure, are implemented on synthetic control chart time series. The results show that this method has strong robustness to loss of feature data due to piecewise segment preprocessing. Moreover, after the construction of shape concept tree, we can discovery knowledge of time series on different time granularity.