Search

In order to improve the stability of direct current (DC) bus voltage when it is disturbed, the virtual DC motor (VDCM) control is applied to the DC converters. However, the constant control parameters in traditional VDCM control cannot provide sufficient inertial and damping support for the system while achieving good dynamic performance. Aiming at this issue, the four-parameter adaptive VDCM control is introduced in this paper. Firstly, the small signal model of dual active full bridge (DAB) with VDCM control is built, then the influence of four different VDCM control parameters on the stability of DC system is studied by drawing the bode diagram and the step response curve. On this basis, the four-parameter adaptive VDCM control is proposed. Finally, the superiority of four-parameter adaptive VDCM control in suppressing DC bus voltage fluctuation and improving the system dynamic response speed is verified by simulation and experiment, the experimental results show that, compared with three-parameter adaptive VDCM control, the fluctuation peak value of DC bus voltage can be reduced about 33% under the four-parameter adaptive VDCM control, and the recovery time after receiving a disturbance of DC bus voltage can be reduced about 25% under the four-parameter adaptive VDCM control. The results show that the four-parameter adaptive VDCM control strategy studied in this paper can further suppress the fluctuation peak of DC bus voltage and improve the dynamic response speed of DC system, which is of great significance to improve the stable operation of DC microgrids.

Deep convolutional neural networks (DCNNs) have been driving significant advances in semantic image segmentation due to their powerful feature representation for recognition. However, their performance in preserving object boundaries is still not satisfactory. Visual mechanism theory indicates that image segmentation tasks require not only recognition, like DCNNs, but also local visual attention capability. Considering that superpixel is good at grasping detailed local structure, we propose a probabilistic superpixel-based dense conditional random field model (PSP-CRF) to refine label assignments as a postprocessing optimization method. First, the well-known fully convolutional networks (FCN) and Deeplab-ResNet are employed to produce coarse prediction probabilistic maps at each pixel. Second, we construct a fully connected CRF model based on the PSP generated by the simple linear iterative clustering algorithm. In our approach, an effective refining algorithm with entropy is developed to convert the pixel-level appearance and position features to the normalized PSP, which works well for CRF. Third, our method optimizes the PSP-CRF to obtain the final label assignment results by employing a highly efficient mean field inference algorithm and some quadratic programming relaxation related algorithms. The experiments on the PASCAL VOC segmentation dataset demonstrate the effectiveness of our methods which can improve the segmentation performance of DCNNs to 82% in mIoU while increasing the computational efficiency by 47%.

Due to the development of the computer vision, machine learning, and deep learning technologies, the research community focuses not only on the traditional SLAM problems, such as geometric mapping and localization, but also on semantic SLAM. In this paper, we propose a Semantic SLAM system which builds the semantic maps with object-level entities, and it is integrated into the RGB-D SLAM framework. The system combines object detection module that is realized by the deep-learning method, and localization module with RGB-D SLAM seamlessly. In the proposed system, object detection module is used to perform object detection and recognition, and localization module is utilized to get the exact location of the camera. The two modules are integrated together to obtain the semantic maps of the environment. Furthermore, to improve the computational efficiency of the framework, an improved Octomap based on the Fast Line Rasterization Algorithm is constructed. Meanwhile, for the sake of accuracy and robustness of the semantic map, conditional random field is employed to do the optimization. Finally, we evaluate our Semantic SLAM through three different tasks, i.e., localization, object detection, and mapping. Specifically, the accuracy of localization and the mapping speed is evaluated on TUM data set. Compared with ORB-SLAM2 and original RGB-D SLAM, our system, respectively, got 72.9% and 91.2% improvements in dynamic environments localization evaluated by root-mean-square error. With the improved Octomap, the proposed Semantic SLAM is 66.5% faster than the original RGB-D SLAM. We also demonstrate the efficiency of object detection through quantitative evaluation in an automated inventory management task on a real-world data sets recorded over a realistic office.

Gesture recognition aims to recognize meaningful movements of human bodies, and is of utmost importance in intelligent human-computer/robot interactions. In this paper, we present a multimodal gesture recognition method based on 3-D convolution and convolutional long-short-term-memory (LSTM) networks. The proposed method first learns short-term spatiotemporal features of gestures through the 3-D convolutional neural network, and then learns long-term spatiotemporal features by convolutional LSTM networks based on the extracted short-term spatiotemporal features. In addition, fine-tuning among multimodal data is evaluated, and we find that it can be considered as an optional skill to prevent overfitting when no pre-trained models exist. The proposed method is verified on the ChaLearn LAP large-scale isolated gesture data set (IsoGD) and the Sheffield Kinect gesture (SKIG) data set. The results show that our proposed method can obtain the state-of-the-art recognition accuracy (51.02% on the validation set of IsoGD and 98.89% on SKIG).

Gray code application in distributed video coding (DVC) has been studied in the literature and it was claimed that DVC performance could be improved by Gray code because of stronger bitplane correlation presented. However, two factors are ignored in the literature, which renders the aforementioned claim untenable. These factors are log-likelihood ratio (LLR) computation and robustness to wrongly decoded bits, which may influence the DVC performance when different bit representations are used. This paper comprehensively evaluates the performances of Gray code in different DVC schemes, such as feedback channel-based transform domain Wyner-Ziv (FC_TDWZ) video coding, parallel pixel domain Wyner-Ziv video coding, and encoder rate control-based TDWZ (ERC_TDWZ). Experimental results suggest that the Gray code does not always improve the DVC performance, although stronger bitplane correlation is presented. In particular, the performance of FC_TDWZ with Gray code is exactly the same as that with natural bit code, because the same magnitudes of LLRs are obtained. Moreover, the reconstruction quality of the WZ frames in ERC_TDWZ is improved by Gray code not because of stronger bitplane correlation but because Gray code is more robust to wrongly decoded bits induced by rate underestimation. However, these RD gains are quite limited when the rate is controlled accurately at the encoder.