| 基于人工神经网络的空间电压矢量脉宽调制设计及其实现 |
| |
| 引用本文： | 潘庭龙 陈尔奎 纪志成 谢林柏. 基于人工神经网络的空间电压矢量脉宽调制设计及其实现[J]. 中小型电机, 2005, 32(7): 29-34 |
| |
| 作者姓名： | 潘庭龙 陈尔奎 纪志成 谢林柏 |
| |
| 作者单位： | 江南大学控制科学与工程研究中心电气自动化研究所,无锡214122 |
| |
| 摘 要： | 在分析空间电压矢量脉宽调制（SVPWM）原理的基础上，提出了一种基于人工神经网络（ANN）的SVPWM新算法。为了解决SVPWM控制技术中复杂的计算问题，设计的网络系统包括多个具有不同训练策略的前向子网络，从而把任务分散化，并充分地利用ANN的快速并行处理能力、学习能力，缩短了计算时间，降低了谐波成分。在dSPACE平台上对该算法进行了实现，仿真及实验结果表明了该方法的可行性、有效性。
|
| 关 键 词： | 空间电压矢量脉宽调制 人工神经网络 硬件平台 |
| 收稿时间： | 2005-02-22 |
Design and Implementation Using dSPACE of SVPWM Based on ANN |
| |
| PAN Ting-long, CHEN Er-kui, JI Zhi-cheng, XIE Lin-bo. Design and Implementation Using dSPACE of SVPWM Based on ANN[J]. S&M Electric Machines, 2005, 32(7): 29-34 |
| |
| Authors: | PAN Ting-long CHEN Er-kui JI Zhi-cheng XIE Lin-bo |
| |
| Affiliation: | Institute of Electrical Automatic, Control Science and Engineering Research Center, Southern Yangtze Univ. , Wuxi 214122, China |
| |
| Abstract: | The basic principle of space voltage vector PWM is analyzed, and a novel algorithm for SVPWM based on ANN is proposed. In order to cope with the complex calculations required in SVPWM, the designed ANN system employs several feedforward neural networks with different training strategies to separate the task. The calculation time is shortened and the harmonic is reduced because of the fast parallel computation and learning capability of ANN. It is implemented in the dSPACE platform, and the operation results show its feasibility and efficiency. |
| |
| Keywords: | SVPWM artificial neural network hardware plat |
| 本文献已被 维普 等数据库收录！ |
