Dongdong Hou
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Dongdong Hou has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- Bo Wang
School of Electrical Engineering and Automation, Wuhan University, Wuhan 430072, China
2 shared publications - Peijia Liu
School of Electrical and Information Engineering, Zhengzhou University, Zhengzhou, China
2 shared publications - Zhengyu Zhu
Zhengzhou University
2 shared publications - Gangqi Chen
School of Marine Science and Technology, Northwestern Polytechnical University, Xian, China
1 shared publication Guanghua Li
Henan Key Laboratory of Underwater Intelligent Equipment, 713th Research Institute, China State Shipbuilding Corporation Ltd., Zhengzhou, China
1 shared publicationJinwei Xiao
Northwestern Polytechnical University,School of Astronautics,Xian,China,710072
1 shared publication
Latest Publications
- 8 July 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 145-150
A Lightweight Deep Neural Network-Based Optical Image Detection Method for Underwater Submerged Rocket Wrecks
Maritime recycling plays a significant role in reusing components from spacecraft. However, the current methods for detecting and recovering sunken spacecraft wrecks suffer from inaccurate detection and low efficiency. To improve the detection effectiveness of rocket wrecks, this paper proposes a method based on underwater optical images which leverage the high-resolution features of images. To address the issue of low accuracy and efficiency in detecting rocket wrecks, a lightweight deep neural network detection model is introduced to enhance accuracy while reducing the computational load and increasing efficiency. Finally, through quantitative and qualitative analyses, our proposed method achieves a detection accuracy of 96.0% with low computational power, which lays a solid foundation for spacecraft recycling.
- 8 July 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 145-150
A Lightweight Deep Neural Network-Based Optical Image Detection Method for Underwater Submerged Rocket Wrecks
Maritime recycling plays a significant role in reusing components from spacecraft. However, the current methods for detecting and recovering sunken spacecraft wrecks suffer from inaccurate detection and low efficiency. To improve the detection effectiveness of rocket wrecks, this paper proposes a method based on underwater optical images which leverage the high-resolution features of images. To address the issue of low accuracy and efficiency in detecting rocket wrecks, a lightweight deep neural network detection model is introduced to enhance accuracy while reducing the computational load and increasing efficiency. Finally, through quantitative and qualitative analyses, our proposed method achieves a detection accuracy of 96.0% with low computational power, which lays a solid foundation for spacecraft recycling.
- 17 June 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 22 (14) , 14366-14377
SINS/DVL Integrated Navigation Method With Current Compensation Using RBF Neural Network
In strapdown inertial navigation system (SINS)/Doppler velocity log (DVL) integrated navigation, the DVL is utilized to restrain the error accumulation of SINS. When operating in mid-water zone, the DVL may lose bottom track due to the limited sensor range. In this case, its measurements will be affected by sea currents. To overcome this problem, this paper proposes a novel current compensation method for SINS/DVL integrated navigation. Based on a new technical route, the proposed method removes the constraints of the existing methods. It is developed as follows. By applying the DVL measurements in the acoustic Doppler current profiler mode, a current measurement module is established to operate in parallel with the SINS/DVL integrated navigation system. On this basis, a radial basis function (RBF) neural network is designed for current compensation. Specifically, when the DVL operates normally, it can simultaneously output the earth-relative and water-relative velocities. The earth-relative velocity is applied for integrated navigation. Meanwhile, it is applied to train the RBF neural network along with the water-relative velocity. Therefore, the RBF neural network approximates the function between the earth-relative and water-relative velocities. When the DVL loses bottom track, the well-trained RBF neural network can be utilized as a velocity predictor. By inputting the water-relative velocity from DVL, it can predict the earth-relative velocity for integrated navigation. In this way, the current compensation is achieved. The effectiveness of the proposed method is verified by realistic semi-physical experiments.
- 17 June 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 22 (14) , 14366-14377
SINS/DVL Integrated Navigation Method With Current Compensation Using RBF Neural Network
In strapdown inertial navigation system (SINS)/Doppler velocity log (DVL) integrated navigation, the DVL is utilized to restrain the error accumulation of SINS. When operating in mid-water zone, the DVL may lose bottom track due to the limited sensor range. In this case, its measurements will be affected by sea currents. To overcome this problem, this paper proposes a novel current compensation method for SINS/DVL integrated navigation. Based on a new technical route, the proposed method removes the constraints of the existing methods. It is developed as follows. By applying the DVL measurements in the acoustic Doppler current profiler mode, a current measurement module is established to operate in parallel with the SINS/DVL integrated navigation system. On this basis, a radial basis function (RBF) neural network is designed for current compensation. Specifically, when the DVL operates normally, it can simultaneously output the earth-relative and water-relative velocities. The earth-relative velocity is applied for integrated navigation. Meanwhile, it is applied to train the RBF neural network along with the water-relative velocity. Therefore, the RBF neural network approximates the function between the earth-relative and water-relative velocities. When the DVL loses bottom track, the well-trained RBF neural network can be utilized as a velocity predictor. By inputting the water-relative velocity from DVL, it can predict the earth-relative velocity for integrated navigation. In this way, the current compensation is achieved. The effectiveness of the proposed method is verified by realistic semi-physical experiments.
- 19 May 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 21 (15) , 17131-17143
An Approach for DVL-Aided SINS In-Motion Alignment Based on Observability Analysis
Currently, in-motion alignment for Strapdown inertial navigation system (SINS) is still a challenge, especially in underwater condition. This paper proposes an approach for DVL-aided SINS in-motion alignment from the perspective of observability. The core of the approach is an improved observability analysis method which overcomes the problems of existing methods. Based on numerical calculations, it can accurately and intuitively analyze the observability of the multi-dimensional complex navigation systems. In response to the problem that numerical calculations may introduce calculation errors, a solution based on prior knowledge of the system is presented. By utilizing the proposed observability analysis method to analyze the DVL-aided SINS in-motion alignment system, this paper has a further research on improving the alignment accuracy through observability incentives. Specifically, the factors which restrict the alignment accuracy are clarified, the reason for how the alignment accuracy is improved by trajectory design is explained, the question of whether the alignment accuracy can be further improved by more complex trajectories is studied, and the guidance for more efficient observability incentives is provided. Consequently, theoretical supports are presented to improve the alignment accuracy efficiently and purposefully. The effectiveness of the proposed approach is verified by simulations and semi-physical experiments.
- 19 May 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 21 (15) , 17131-17143
An Approach for DVL-Aided SINS In-Motion Alignment Based on Observability Analysis
Currently, in-motion alignment for Strapdown inertial navigation system (SINS) is still a challenge, especially in underwater condition. This paper proposes an approach for DVL-aided SINS in-motion alignment from the perspective of observability. The core of the approach is an improved observability analysis method which overcomes the problems of existing methods. Based on numerical calculations, it can accurately and intuitively analyze the observability of the multi-dimensional complex navigation systems. In response to the problem that numerical calculations may introduce calculation errors, a solution based on prior knowledge of the system is presented. By utilizing the proposed observability analysis method to analyze the DVL-aided SINS in-motion alignment system, this paper has a further research on improving the alignment accuracy through observability incentives. Specifically, the factors which restrict the alignment accuracy are clarified, the reason for how the alignment accuracy is improved by trajectory design is explained, the question of whether the alignment accuracy can be further improved by more complex trajectories is studied, and the guidance for more efficient observability incentives is provided. Consequently, theoretical supports are presented to improve the alignment accuracy efficiently and purposefully. The effectiveness of the proposed approach is verified by simulations and semi-physical experiments.