In this paper we report a kind of fast-scale instability occurring in the single-ended primary inductance converter （SEPIC） power factor pre-regulator, which is designed to operate in discontinuous conduction mode. ...In this paper we report a kind of fast-scale instability occurring in the single-ended primary inductance converter （SEPIC） power factor pre-regulator, which is designed to operate in discontinuous conduction mode. Main results are given by exact cycle-by-cycle computer simulations as well as theoretical analysis. It is found that the instability phenomenon manifests itself as a fast-scale bifurcation at the switching period, which implies the occurrence of border collision bifurcation, or is related to the transition of the regular operating mode of the SEPIC. According to the theoretical analysis and simulation results, the effects of parameters on system stability, and the locations of the bifurcation points are confirmed. Moreover, the effects of such an instability on power factor and switching stress are also discussed. Finally, the occurrence of the asymmetric bifurcation locations is investigated. The results show that this work provides a convenient means of predicting stability boundaries which can facilitate the selection of the practical parameters.展开更多
针对在无人机高速运动场景下,相关滤波跟踪算法存在边界效应和滤波器退化等问题,提出了一种融合相机运动补偿的自适应时空正则化的无人机跟踪相关滤波算法。首先,在相关滤波器中设计了一种自适应空间正则化方法来缓解边界效应;其次,为...针对在无人机高速运动场景下,相关滤波跟踪算法存在边界效应和滤波器退化等问题,提出了一种融合相机运动补偿的自适应时空正则化的无人机跟踪相关滤波算法。首先,在相关滤波器中设计了一种自适应空间正则化方法来缓解边界效应;其次,为防止滤波器退化,设计了基于高置信度样本的时间正则化方法及优化更新策略,同时利用交替方向乘子法(Alternating Direction Method of Multipliers,ADMM)优化目标函数的求解,保证了算法的运行效率;然后,设计了快速尺度滤波器,实现目标尺度的快速估计;最后,提出了一种鲁棒性的跟踪不确定性规则及基于相机运动补偿的重检测器,使得目标丢失后可及时恢复。实验结果表明,提出的算法在无人机数据集上的精确度和成功率分别可达0.739和0.634,与ECO_HC(Efficient Convolution Operators-Hand Crafted)算法相比分别提升了1.65%和6.9%,与同类主流算法相比,有效地提升了跟踪时的精确度和成功率,经历快速运动、目标遮挡等情况时应用该算法的无人机也表现出良好的跟踪性能。展开更多
文摘In this paper we report a kind of fast-scale instability occurring in the single-ended primary inductance converter （SEPIC） power factor pre-regulator, which is designed to operate in discontinuous conduction mode. Main results are given by exact cycle-by-cycle computer simulations as well as theoretical analysis. It is found that the instability phenomenon manifests itself as a fast-scale bifurcation at the switching period, which implies the occurrence of border collision bifurcation, or is related to the transition of the regular operating mode of the SEPIC. According to the theoretical analysis and simulation results, the effects of parameters on system stability, and the locations of the bifurcation points are confirmed. Moreover, the effects of such an instability on power factor and switching stress are also discussed. Finally, the occurrence of the asymmetric bifurcation locations is investigated. The results show that this work provides a convenient means of predicting stability boundaries which can facilitate the selection of the practical parameters.
文摘针对在无人机高速运动场景下,相关滤波跟踪算法存在边界效应和滤波器退化等问题,提出了一种融合相机运动补偿的自适应时空正则化的无人机跟踪相关滤波算法。首先,在相关滤波器中设计了一种自适应空间正则化方法来缓解边界效应;其次,为防止滤波器退化,设计了基于高置信度样本的时间正则化方法及优化更新策略,同时利用交替方向乘子法(Alternating Direction Method of Multipliers,ADMM)优化目标函数的求解,保证了算法的运行效率;然后,设计了快速尺度滤波器,实现目标尺度的快速估计;最后,提出了一种鲁棒性的跟踪不确定性规则及基于相机运动补偿的重检测器,使得目标丢失后可及时恢复。实验结果表明,提出的算法在无人机数据集上的精确度和成功率分别可达0.739和0.634,与ECO_HC(Efficient Convolution Operators-Hand Crafted)算法相比分别提升了1.65%和6.9%,与同类主流算法相比,有效地提升了跟踪时的精确度和成功率,经历快速运动、目标遮挡等情况时应用该算法的无人机也表现出良好的跟踪性能。
