Yi Gu

Yi Gu

Auto-generated profile, based on publicly available information

Yi Gu

Baidu Online Network Technology (Beijing) Company Ltd., Beijing, China

Research Focus & Metrics

Academic Career Length

11Years (2016-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
28 January 2020 in IEEE Access
Verified
Journal Article
Open access
HJHuaiguang Jiang
Huaiguang Jiang
YGYi Gu
Yi Gu
YXYu Xie
Yu Xie
RYRui Yang
Rui Yang
YZYingchen Zhang
Yingchen Zhang
Global horizontal irradiance (GHI) is a critical index to indicate the output power of the photovaltaic (PV). In traditional approaches, the local GHI can be measured with very expensive instruments, and the large-area GHI collection depends on complex satellite-based models, solargis algorithms, and the high-performance computers (HPC). In this paper, a novel approach is proposed to capture the GHI conveniently and accurately. Considering the nonstationary property of the GHI on cloudy days, the GHI capturing is cast as an image regression problem. In traditional approaches, the image regression problem is treated as two parts, feature extraction (for the images) and regression model (for the regression targets), which are optimized separately and blocked the interconnections. Considering the nonlinear regression capability, a convolutional neural network (CNN) based image regression approach is proposed to provide an End-to-End solution for the cloudy day GHI capturing problem in this paper. The multilayer CNN is based on the AlexNet and VGG. The L2 (least square errors) with regularization is used as the loss function in the regression layer. For data cleaning, the Gaussian mixture model with Bayesian inference is employed to detect and eliminate the anomaly data in a nonparametric manner. The purified data are used as input data for the proposed image regression approach. In the experiments, three-month sky images and GHI data (with 1-min resolution) are provided by the National Renewable Energy Laboratory (NREL) with the HPC system. The numerical results demonstrate the feasibility and effectiveness of the proposed approach.

31 July 2018 in IEEE Transactions on Smart Grid
Verified
Journal Article
Closed access
YGYi Gu
Yi Gu
HJHuaiguang Jiang
Huaiguang Jiang
Jun ZhangJun Zhang
Jun Zhang
YZYingchen Zhang
Yingchen Zhang
Hongyu WuHongyu Wu
Hongyu Wu
EMEduard Muljadi
Eduard Muljadi
This paper proposes a multi-timescale operational approach for a three-phase unbalanced distribution system, which contains hourly scheduling at substation level and minutes power flow operation at feeder level. In the substation scheduling model, the objective is to minimize the system cost with variable renewable generations. The given error distribution model of renewable generation is formulated as a chance constraint, and derived into a deterministic form by Gaussian mixture model (GMM) with genetic algorithm-based expectation-maximization (GAEM). In the feeder scheduling model, the system cost is further reduced with the optimal power flow (OPF) at a timescale of minutes. Considering the nonconvexity of the three-phase unbalanced OPF problem in distribution systems, the semidefinite programming (SDP) is used to relax the problem into a convex problem, and a distributed computation approach is built based on alternating direction method of multiplier (ADMM). The IEEE 123-bus distribution system, DU (University of Denver) campus distribution system, and the IEEE 8500-bus distribution system are used as the test bench for the proposed approach. And the numerical results demonstrate the effectiveness and validity of the proposed method.

21 April 2021 in IEEE Access
Verified
Journal Article
Open access
JLJingwei Liu
Jingwei Liu
YGYi Gu
Yi Gu
SHShumin Han
Shumin Han
ZZZhibin Zhang
Zhibin Zhang
JGJiafeng Guo
Jiafeng Guo
XCXueqi Cheng
Xueqi Cheng
Recent years have witnessed rapid developments on computer vision, however, there are still challenges in detecting tiny objects in a large-scale background. The tiny objects knowledge become sparse and weak due to their tiny size, which makes the tiny objects difficult to be detected with the common approaches. In this paper, a new network named Specific Characteristics based Feature Rescaling and Fusion (SFRF) is designed to detect tiny persons in a broad horizon and massive background. Different from the methods in general, a Nonparametric Adaptive Dense Perceiving Algorithm (NADPA) is designed to automatically select and generate a new resized feature map with the high density distribution of tiny objects. Then, a method called Many-For-One strategy is used for feature fusion of the feature pyramid network (FPN) layers to improve the feature representation and detection. Finally, an ensemble model method named hierarchical Coarse-to-fine mechanism is designed based on the proposed methods to further improve the performance. The experiments demonstrate that the proposed approach achieves an obvious performance improvement on tiny object detection than the existing approaches, and our approach has been awarded as the 1st-place in the first large-scale Tiny Object Detection (TOD) challenge.

21 November 2016
Verified
Conference Paper
Closed access
YGYi Gu
Yi Gu
HJHuaiguang Jiang
Huaiguang Jiang
YZYingchen Zhang
Yingchen Zhang
Jun ZhangJun Zhang
Jun Zhang
TGTianlu Gao
Tianlu Gao
EMEduard Muljadi
Eduard Muljadi
In this paper, a big data visualization platform is designed to discover the hidden useful knowledge for smart grid (SG) operation, control and situation awareness. The spawn of smart sensors at both grid side and customer side can provide large volume of heterogeneous data that collect information in all time spectrums. Extracting useful knowledge from this big-data poll is still challenging. In this paper, the Apache Spark, an open source cluster computing framework, is used to process the big-data to effectively discover the hidden knowledge. A high-speed communication architecture utilizing the Open System Interconnection (OSI) model is designed to transmit the data to a visualization platform. This visualization platform uses Google Earth, a global geographic information system (GIS) to link the geological information with the SG knowledge and visualize the information in user defined fashion. The University of Denver's campus grid is used as a SG test bench and several demonstrations are presented for the proposed platform.

24 March 2021 in IEEE Access
Verified
Journal Article
Open access
JLJingwei Liu
Jingwei Liu
YGYi Gu
Yi Gu
SHShumin Han
Shumin Han
ZZZhibin Zhang
Zhibin Zhang
JGJiafeng Guo
Jiafeng Guo
XCXueqi Cheng
Xueqi Cheng
The deep convolutional networks have a great success in vision classification tasks. For object detection, tasks are divided into two subtasks: localization and classification. The detectors scan the whole image to generate object proposals relying on the predefined anchors or points, then classify and fine trim the proposals. The localization task plays an important role in object detection. The foreground objects and background can be easily confused under complex scenes in the existing approaches. Thus, the localization task results in a bad influence on the performance of classification problems. In order to enhance the object localization perceived, a novel method called Attentional Feature Perceive and Augmentation (Att-FPA) is constructed, which devotes setting up a feature dual perceive for localization and classification. Firstly, an attention mask branch (with an attention mask) is imported to check out if the detecting area is the background or containing one or more objects. By maintaining the feature representation of background, the attention mask is employed to strengthen the feature representation of foreground objects, which significantly enhances the localization perceived ability of objects. Moreover, Att-FPA supplies a novel way on feature re-extraction and augmentation, which further promotes the performance significantly. In the task of Ms COCO object detection, Att-FPA achieves an outstanding promotion(3.0+ mAP) than baseline (Faster-Rcnn) method. It establishes a new efficient method of feature representation, which outperforms the state-of-the-art models in object detection.