论文部分内容阅读
神经网络技术在人工智能、自动控制以及模式识别等领域的研究与应用正方兴未艾。而滞环电流控制是一种传统常规的电流控制方式,在功率因数校正和无功补偿等领域有着广泛的应用。该文介绍了三相变流器的BP神经网络滞环电流内环控制,该方案可实现神经网络对快速变量的控制,提高滞环控制的性能,使系统对参数的变化有较强的不灵敏性和鲁棒性。该文分析了三相电源不平衡、某一路电流反馈丢失的工况下,系统的控制特性。为了使系统在轻负载下得到良好的频谱特性,采用实时变误差增益的控制策略,并讨论了容差带下限。同时借助于矢量调制的思想,结合神经网络滞环调节器,优化系统性能,减小系统EMI和开关损耗。
Research and application of neural network technology in the fields of artificial intelligence, automatic control and pattern recognition are in the ascendant. The hysteresis current control is a conventional conventional current control mode, which has a wide range of applications in the fields of power factor correction and reactive power compensation. This paper introduces the BP neural network hysteresis current three-phase converter current loop control, the program can achieve rapid control of neural network to improve the performance of hysteresis control, the system changes in the parameters are not strong Sensitivity and robustness. This paper analyzes the control characteristics of the system under unbalanced three-phase power supply and loss of current feedback on one path. In order to get good spectral characteristics under light load, the control strategy of real-time variable gain gain is adopted, and the lower limit of tolerance band is discussed. At the same time, by virtue of the idea of vector modulation, combined with neural network hysteresis regulator to optimize system performance and reduce system EMI and switching losses.