Bounding box regression is a crucial step in object detection, directly affecting the localization performance of the detected objects. Especially in small object detection, an excellent bounding box regression loss can significantly alleviate the problem of missing small objects. However, there are two major problems with the broad Intersection over Union (IoU) losses, also known as Broad IoU losses (BIoU losses) in bounding box regression: (i) BIoU losses cannot provide more effective fitting information for predicted boxes as they approach the target box, resulting in slow convergence and inaccurate regression results; (ii) most localization loss functions do not fully utilize the spatial information of the target, namely the target’s foreground area, during the fitting process. Therefore, this paper proposes the Corner-point and Foreground-area IoU loss (CFIoU loss) function by delving into the potential for bounding box regression losses to overcome these issues. First, we use the normalized corner point distance between the two boxes instead of the normalized center-point distance used in the BIoU losses, which effectively suppresses the problem of BIoU losses degrading to IoU loss when the two boxes are close. Second, we add adaptive target information to the loss function to provide richer target information to optimize the bounding box regression process, especially for small object detection. Finally, we conducted simulation experiments on bounding box regression to validate our hypothesis. At the same time, we conducted quantitative comparisons of the current mainstream BIoU losses and our proposed CFIoU loss on the small object public datasets VisDrone2019 and SODA-D using the latest anchor-based YOLOv5 and anchor-free YOLOv8 object detection algorithms. The experimental results demonstrate that YOLOv5s (+3.12% Recall, +2.73% mAP@0.5, and +1.91% mAP@0.5:0.95) and YOLOv8s (+1.72% Recall and +0.60% mAP@0.5), both incorporating the CFIoU loss, achieved the highest performance improvement on the VisDrone2019 test set. Similarly, YOLOv5s (+6% Recall, +13.08% mAP@0.5, and +14.29% mAP@0.5:0.95) and YOLOv8s (+3.36% Recall, +3.66% mAP@0.5, and +4.05% mAP@0.5:0.95), both incorporating the CFIoU loss, also achieved the highest performance improvement on the SODA-D test set. These results indicate the effectiveness and superiority of the CFIoU loss in small object detection. Additionally, we conducted comparative experiments by fusing the CFIoU loss and the BIoU loss with the SSD algorithm, which is not proficient in small object detection. The experimental results demonstrate that the SSD algorithm incorporating the CFIoU loss achieved the highest improvement in the AP (+5.59%) and AP75 (+5.37%) metrics, indicating that the CFIoU loss can also improve the performance of algorithms that are not proficient in small object detection.

Model predictive power control (MPPC) has been receiving plenty of concern owing to its fast dynamic response. However, the varying switching frequency makes the harmonics of output voltage vary over a wide scope, which adds difficulty in designing alternating current (AC) filter. Here, on the basis of the active and reactive power derivatives of each vector and the effect of switching states on neutral point (NP) voltage; we present a MPPC with dual vectors (MPPCDV) to control the three-level (3L) inverter. In this algorithm, a new switching sequence is provided to balance NP voltage of 3L inverter, and two cost functions are derived to reduce the number of cycle computation. According to the first cost functions, the sector containing optimal switching sequence (OSS) is obtained. The second cost function is assigned to achieve the multi-objective optimization. Simulations and experiments are built to test the effectiveness of MPPCDV. When the 3L inverter adopts the MPPCDV, the majority of the harmonic frequencies in output voltage are concentrated near the sample frequency, the steady-state performances are improved remarkably and the dynamic performance is as fast as the conventional MPPC.

For a three-phase isolated AC-DC matrix converter, the slight volt-second imbalance caused by variable input voltage is an inherent defect and has been neglected in most PWM schemes. The analysis in this paper shows that the imbalance lies in a single direction during the entire input cycle, and it might generate a large DC bias to severely decrease the efficiency of the steady-state operation. Therefore, the imbalance phenomenon is first investigated, and the generation mechanism of the steady-state DC bias is derived in detail. To eliminate such a DC bias, a novel transformer primary-side voltage pulse is originally defined. Based on the special crisscrossing operation mode with the above pulse, a simple but effective four-segment space vector PWM (4S-SVPWM) scheme is proposed. Compared with the conventional more-segment SVPWM schemes, the novel 4S-SVPWM scheme can achieve volt-second self-balancing without auxiliary methods, as well as a 1/2 reduction in the common-mode voltage (CMV) peak value and a 1/3 reduction in the switching actions. The overall efficiency and power density of the converter are greatly improved. Finally, an experimental prototype is built to validate the correctness and effectiveness of the novel scheme.

In this study, a novel field-oriented control (FOC) algorithm was proposed in a 60° coordinate system for controlling the speed of permanent magnet synchronous motors. The FOC algorithm consists of several parts in which the reference currents and feedback currents are transformed into the representation form in a 60° coordinate system. Current regulators are typically used in a 60° coordinate system to directly obtain the reference voltage vector. The proposed FOC structure was established by incorporating the space vector pulse width modulation algorithm in a 60° coordinate system. The proposed FOC structure simplified the FOC algorithm and reduced its computational burden. The feasibility of the proposed method was verified through simulations and experiments.