No abstract

To avoid weight, size and cost constraints, on-board integrated charging devices use the traction system components. In this paper, the topologies of on-board integrated electric vehicles (EVs) chargers are reviewed. The topologies of on-board integrated chargers can be classified into three groups based on the integrated components. The first type is that only converter is integrated as one part of the on-board charger. While the grid side also needs filter inductors. The second group is based on the switched reluctance motor, which can integrate both the converter and the motor windings into the on-board charger. This is followed by the integrated charger based on permanent magnet motor (PM) and induction motor (IM). The last group also integrates both the converter and the motor windings. Then this paper gives a new single-phase on-board integrated charger based on AC motor. Parts of inverter are used as single-phase rectifier, parts of inverter coupled with one capacitor and inductor reconstruct an active power filter (APF) circuit, and the motor windings are used as grid-side filter inductors. As the existence of APF circuit, the proposed integrated charger can eliminate the second harmonic ripple and the quasi-Z-source network just need small capacitor and inductor to suppress ripples. Using the proposed topology, the motor can keep still in charging mode. A 2.2-kW prototype is designed to give the experimental results.Index Terms-Active power filter (APF), integrated charger, quasi-Z-source.

Temporal action proposal generation is an important and challenging task in video understanding, which aims at detecting all temporal segments containing action instances of interest. The existing proposal generation approaches are generally based on pre-defined anchor windows or heuristic bottom-up boundary matching strategies. This paper presents a simple and end-to-end learnable framework (RTD-Net) for direct action proposal generation, by re-purposing a Transformer-alike architecture. To tackle the essential visual difference between time and space, we make three important improvements over the original transformer detection framework (DETR). First, to deal with slowness prior in videos, we replace the original Transformer encoder with a boundary attentive module to better capture temporal information. Second, due to the ambiguous temporal boundary and relatively sparse annotations, we present a relaxed matching loss to relieve the strict criteria of single assignment to each groundtruth. Finally, we devise a three-branch head to further improve the proposal confidence estimation by explicitly predicting its completeness. Extensive experiments on THUMOS14 and ActivityNet-1.3 benchmarks demonstrate the effectiveness of RTD-Net, on both tasks of temporal action proposal generation and temporal action detection. Moreover, due to its simplicity in design, our RTD-Net is more efficient than previous proposal generation methods without non-maximum suppression post-processing. The code will be available at https://github.com/MCG-NJU/RTD-Action.

No abstract

Single-phase quasi-Z-source rectifiers have second harmonic currents and corresponding ripple voltages on the dc bus. To filer the voltage ripple, bulky capacitor bank is needed. This paper proposed an active power filter (APF) for single-phase quasi-Z-source rectifier. It eliminated the second harmonic power with small capacitor and inductor. This topology is suitable to integrated electric vehicle (EV) on-board charger, which can use EV's inverter hardware as its own rectifier hardware. Thus, this proposed topology can save much space and weight. Simulation and experimental results verified the proposed system.