Wuqi Gao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Wuqi Gao has a total of 21 co-authors. Here are the top co-authors by number of shared publications:
Liangliang Li
School of Mechatronic Engineering, Xi’an Technological University, Xi’an 710021 P. R. China
3 shared publicationsJunmin Luo
School of Computer Science and Engineering, Xi'an Technological University, Xi'an, China
2 shared publicationsMan Wang
School of Electronics and Information Engineering, Xi’an Technological University, Xi’an 710021, P. R. China
2 shared publicationsTing Yang
School of Ordnance Science and Technology, Xi’an Technological University, Xi’an 710021, P. R. China
2 shared publications- Xiaoyan Li
School of Electronics and Information Engineering, Xi’an Technological University
2 shared publications - Changyuan Wang
School of Opto-Electronical Engineering, Xi’an Technological University, Xi’an 710021, China
1 shared publication
Latest Publications
- Open Access
- 26 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 33805-33818
Creep Prediction of Fluoroplastics Based on Physics Informed Neural Networks
Fluoroplastics are key sealing materials in rocket engines. They can easily fail due to creep during long-term storage, which may threaten the engines’ safety during operation. Accurate prediction of their time-dependent deformation under varying temperature and pressure conditions remains a critical challenge. Existing methods for predicting polymer creep, such as empirical methods and physically-based constitutive models, often depend on fixed working conditions and require material composition analysis. To address these issues, this paper proposes a physics-informed two-stage framework for predicting the extrapolative creep behavior of fluoroplastics. An Extended Burgers Model (EBM) with adaptive parameters is employed to represent viscoelastic behavior under coupled temperature–pressure variations. A hybrid PSO–NLS strategy is developed to robustly identify EBM parameters from 18 experimental creep curves, overcoming initialization sensitivity and ensuring high-quality training labels. Subsequently, a neural network learns the mapping from service conditions (T,F) to EBM parameters, guided by material-specific physical constraints that enforce long-term consistency and extrapolation capabilities. The resulting model generates full creep strain curves ε(t) that are both data-driven and physically plausible. Experimental results demonstrate that the proposed method surpasses conventional empirical models in extrapolation accuracy, adherence to physical principles, and ease of practical implementation. It combines the strong fitting ability of data-driven models with the interpretability of physics-based models to improve both prediction accuracy and generalization performance.
- 26 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 33805-33818
Open AccessCreep Prediction of Fluoroplastics Based on Physics Informed Neural Networks
Fluoroplastics are key sealing materials in rocket engines. They can easily fail due to creep during long-term storage, which may threaten the engines’ safety during operation. Accurate prediction of their time-dependent deformation under varying temperature and pressure conditions remains a critical challenge. Existing methods for predicting polymer creep, such as empirical methods and physically-based constitutive models, often depend on fixed working conditions and require material composition analysis. To address these issues, this paper proposes a physics-informed two-stage framework for predicting the extrapolative creep behavior of fluoroplastics. An Extended Burgers Model (EBM) with adaptive parameters is employed to represent viscoelastic behavior under coupled temperature–pressure variations. A hybrid PSO–NLS strategy is developed to robustly identify EBM parameters from 18 experimental creep curves, overcoming initialization sensitivity and ensuring high-quality training labels. Subsequently, a neural network learns the mapping from service conditions (T,F) to EBM parameters, guided by material-specific physical constraints that enforce long-term consistency and extrapolation capabilities. The resulting model generates full creep strain curves ε(t) that are both data-driven and physically plausible. Experimental results demonstrate that the proposed method surpasses conventional empirical models in extrapolation accuracy, adherence to physical principles, and ease of practical implementation. It combines the strong fitting ability of data-driven models with the interpretability of physics-based models to improve both prediction accuracy and generalization performance.
- Open Access
- 1 June 2025
- Published by Walter de Gruyter GmbH in International Journal of Advanced Network, Monitoring and Controls
- Vol. 10 (2) , 20-29
Research on Automatic Problem-Solving Technology of Olympic Mathematics in Primary Schools Based on AORBCO Model
This study addresses intelligent problemsolving in elementary math competitions by proposing an AORBCO model-based system. It integrates knowledge graphs, rule-based reasoning, and cognitive optimization to simulate human problem-solving processes. The framework systematically analyzes competition problem types, constructs a structured knowledge base, and implements dual-solving modules: rule-template matching and knowledge graph reasoning, supplemented by question bank similarity retrieval. Experimental results demonstrate 15% higher accuracy and 30% faster solving speed compared to conventional methods, with enhanced interpretability. Key innovations include the first application of AORBCO in educational AI, novel knowledge representation methods, and specialized cognitive optimization algorithms. The research provides technical support for personalized math education and advances intelligent tutoring systems. Future work will focus on improving model generalization and exploring multimodal learning integration.
- 1 June 2025
- Published by Walter de Gruyter GmbH in International Journal of Advanced Network, Monitoring and Controls
- Vol. 10 (2) , 20-29
Open AccessResearch on Automatic Problem-Solving Technology of Olympic Mathematics in Primary Schools Based on AORBCO Model
This study addresses intelligent problemsolving in elementary math competitions by proposing an AORBCO model-based system. It integrates knowledge graphs, rule-based reasoning, and cognitive optimization to simulate human problem-solving processes. The framework systematically analyzes competition problem types, constructs a structured knowledge base, and implements dual-solving modules: rule-template matching and knowledge graph reasoning, supplemented by question bank similarity retrieval. Experimental results demonstrate 15% higher accuracy and 30% faster solving speed compared to conventional methods, with enhanced interpretability. Key innovations include the first application of AORBCO in educational AI, novel knowledge representation methods, and specialized cognitive optimization algorithms. The research provides technical support for personalized math education and advances intelligent tutoring systems. Future work will focus on improving model generalization and exploring multimodal learning integration.
- 29 September 2022
- Published by Springer Nature in International Journal of Control, Automation, and Systems
- Vol. 20 (11) , 3549-3560
Robust Fractional Embedded Cubature Kalman Filter for Fractional Nonlinear Stochastic System
- 29 September 2022
- Published by Springer Nature in International Journal of Control, Automation, and Systems
- Vol. 20 (11) , 3549-3560
Robust Fractional Embedded Cubature Kalman Filter for Fractional Nonlinear Stochastic System
- Open Access
- 7 September 2022
- Published by World Scientific Pub Co Pte Ltd in International Journal of Pattern Recognition and Artificial Intelligence
- Vol. 36 (11)
Driving State Discrimination Algorithm Based on Lightweight Network and Contrast Learning
Driver misbehavior is one of the major traffic safety hazards as car ownership increases year by year. So, it is important to have driver fatigue detection and behavior recognition. Initially, given the fatigue detection problem that the images captured by the visible light camera cannot capture the eyes of a driver wearing sunglasses or eye glasses. As a solution, this paper introduces a DCT-HSV preprocessing algorithm for infrared images, which is believed to enhance the target characteristics of infrared images. The paper also introduces a more efficient lightweight SSD detection model, which achieves a better balance in terms of model size and detection performance. It also shows a better performance in self-built datasets and basic vehicle operation datasets. Secondly, aiming at the problems of high complexity and poor accuracy of the existing driving behavior detection model, this paper designs a driver abnormal behavior discrimination model based on the comparison twin, which has good performance on the Kaggle public dataset. The relevant experimental detection results show that the method constructed in this paper has high detection accuracy, low warning delay, and good practical use value.
- 7 September 2022
- Published by World Scientific Pub Co Pte Ltd in International Journal of Pattern Recognition and Artificial Intelligence
- Vol. 36 (11)
Open AccessDriving State Discrimination Algorithm Based on Lightweight Network and Contrast Learning
Driver misbehavior is one of the major traffic safety hazards as car ownership increases year by year. So, it is important to have driver fatigue detection and behavior recognition. Initially, given the fatigue detection problem that the images captured by the visible light camera cannot capture the eyes of a driver wearing sunglasses or eye glasses. As a solution, this paper introduces a DCT-HSV preprocessing algorithm for infrared images, which is believed to enhance the target characteristics of infrared images. The paper also introduces a more efficient lightweight SSD detection model, which achieves a better balance in terms of model size and detection performance. It also shows a better performance in self-built datasets and basic vehicle operation datasets. Secondly, aiming at the problems of high complexity and poor accuracy of the existing driving behavior detection model, this paper designs a driver abnormal behavior discrimination model based on the comparison twin, which has good performance on the Kaggle public dataset. The relevant experimental detection results show that the method constructed in this paper has high detection accuracy, low warning delay, and good practical use value.
- 5 August 2022
- Published by World Scientific Pub Co Pte Ltd in International Journal of Pattern Recognition and Artificial Intelligence
- Vol. 36 (10)
Face Detection Based on Dynamic Min Size and Dense Connection
Since the face detection technology is widely used, improving the detection speed and accuracy in face detection tasks has become a key challenge. Therefore, this paper takes the MTCNN model as the research object and makes improvements, the purpose of which is to optimize the detection speed and accuracy simultaneously. A dynamic min size algorithm is proposed. According to the size of the input image, it dynamically controls the minimum size of the face to be recognized by the model and reduces the number of iterations of the image pyramid, increasing the number of detection frames per second by 4[ ]fps. Then, the standard convolution structure in the P-Net and R-Net models is replaced by the depth-wise separable convolution, which effectively reduces the number of parameters and computation of the model. Meanwhile, an O-Net model with a densely connected structure is also developed. Our experiments with well-known public datasets have demonstrated that the proposed network structure can improve the detection frame rate. By reusing the features of different levels of the image, the recall rate and the precision rate of the MTCNN model on the validation set are increased by 2.39% and 1.65%, respectively.
- 5 August 2022
- Published by World Scientific Pub Co Pte Ltd in International Journal of Pattern Recognition and Artificial Intelligence
- Vol. 36 (10)
Face Detection Based on Dynamic Min Size and Dense Connection
Since the face detection technology is widely used, improving the detection speed and accuracy in face detection tasks has become a key challenge. Therefore, this paper takes the MTCNN model as the research object and makes improvements, the purpose of which is to optimize the detection speed and accuracy simultaneously. A dynamic min size algorithm is proposed. According to the size of the input image, it dynamically controls the minimum size of the face to be recognized by the model and reduces the number of iterations of the image pyramid, increasing the number of detection frames per second by 4[ ]fps. Then, the standard convolution structure in the P-Net and R-Net models is replaced by the depth-wise separable convolution, which effectively reduces the number of parameters and computation of the model. Meanwhile, an O-Net model with a densely connected structure is also developed. Our experiments with well-known public datasets have demonstrated that the proposed network structure can improve the detection frame rate. By reusing the features of different levels of the image, the recall rate and the precision rate of the MTCNN model on the validation set are increased by 2.39% and 1.65%, respectively.
- RetractedOpen Access
- 18 March 2022
- Published by Wiley in Security and Communication Networks
- Vol. 2022, 1-11
A Method for Detection of Rare Industrial Fake Film Based on Deep Learning
In order to solve the industrial fake film problem produced by repeated imaging in the field of nondestructive testing, this paper proposes an image similarity comparison algorithm of X-ray film based on the Siamese neural network. Firstly, in order to accurately locate and detect weld areas from high-resolution X-ray film, a robust location and detection algorithm based on the Gaussian function is proposed. Next, in order to solve the problem that the deep learning model system extracts features according to squares by default, which results in the disappearance of film features with high horizontal and vertical features, effective weld regions are used and processed in blocks. Then, a small sample data set for weld similarity discrimination is established by using each block region, and an industrial fake film discrimination model based on twin neural networks is constructed. Finally, the corresponding weights are set for different fake feature information areas, and a similarity comparison system for industrial fake film detection is proposed. The experiment shows that the method in this paper can not only accurately locate the weld area but also accurately detect the fake film in the industry, avoiding the influence of the geometric transformation of the image and the insufficient number of data sets in the process of similarity comparison, and has a good detection effect.
- 18 March 2022
- Published by Wiley in Security and Communication Networks
- Vol. 2022, 1-11
RetractedOpen AccessA Method for Detection of Rare Industrial Fake Film Based on Deep Learning
In order to solve the industrial fake film problem produced by repeated imaging in the field of nondestructive testing, this paper proposes an image similarity comparison algorithm of X-ray film based on the Siamese neural network. Firstly, in order to accurately locate and detect weld areas from high-resolution X-ray film, a robust location and detection algorithm based on the Gaussian function is proposed. Next, in order to solve the problem that the deep learning model system extracts features according to squares by default, which results in the disappearance of film features with high horizontal and vertical features, effective weld regions are used and processed in blocks. Then, a small sample data set for weld similarity discrimination is established by using each block region, and an industrial fake film discrimination model based on twin neural networks is constructed. Finally, the corresponding weights are set for different fake feature information areas, and a similarity comparison system for industrial fake film detection is proposed. The experiment shows that the method in this paper can not only accurately locate the weld area but also accurately detect the fake film in the industry, avoiding the influence of the geometric transformation of the image and the insufficient number of data sets in the process of similarity comparison, and has a good detection effect.
- 1 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 47, 69-74
Research on Fitness Action Evaluation System Based on Skeleton
In recent years, with the promotion of the concept of sports to gain health, the cause of national fitness has stepped into the fast lane of development. However, for ordinary fitness people, due to the lack of corresponding supervision and evaluation feedback, resulting in long-term inaccurate or even wrong fitness process. Therefore, this paper designs and implements a skeleton based fitness action evaluation system. Firstly, open pose algorithm is applied to extract the skeleton nodes of each frame in the video, and then the extracted data is prepossessed by key node alignment, skeleton node balancing and other technologies. Then the skeleton nodes are used to evaluate the action duration, speed and scale. Finally, the evaluation model is established, The skeleton node is used to calculate the duration, speed and standard degree of the action, and the standard degree of the action is divided into three grades: excellent and poor, so as to guide the standardization of the trainer's action.
- 1 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 47, 69-74
Research on Fitness Action Evaluation System Based on Skeleton
In recent years, with the promotion of the concept of sports to gain health, the cause of national fitness has stepped into the fast lane of development. However, for ordinary fitness people, due to the lack of corresponding supervision and evaluation feedback, resulting in long-term inaccurate or even wrong fitness process. Therefore, this paper designs and implements a skeleton based fitness action evaluation system. Firstly, open pose algorithm is applied to extract the skeleton nodes of each frame in the video, and then the extracted data is prepossessed by key node alignment, skeleton node balancing and other technologies. Then the skeleton nodes are used to evaluate the action duration, speed and scale. Finally, the evaluation model is established, The skeleton node is used to calculate the duration, speed and standard degree of the action, and the standard degree of the action is divided into three grades: excellent and poor, so as to guide the standardization of the trainer's action.
- Open Access
- 1 February 2021
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 1744 (4) , 042181
A New Method of Measuring the Age of Abalone Based on Data Visualization Analysis
This project uses a new way to count the abalone age, which use abalone’s physical characteristics to predict by multiple linear regression. After the model is trained, when we catch a new abalone, we can use a computer to replace the labor to a certain extent, saving costs to the enterprise. Results are given in a visualization of the data.
- 1 February 2021
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 1744 (4) , 042181
Open AccessA New Method of Measuring the Age of Abalone Based on Data Visualization Analysis
This project uses a new way to count the abalone age, which use abalone’s physical characteristics to predict by multiple linear regression. After the model is trained, when we catch a new abalone, we can use a computer to replace the labor to a certain extent, saving costs to the enterprise. Results are given in a visualization of the data.
- Open Access
- 1 January 2021
- Published by Wiley in Wireless Communications and Mobile Computing
- Vol. 2021 (1)
Improved Byzantine Fault‐Tolerant Algorithm Based on Alliance Chain
Alliance chain is a typical multicenter block chain and is easily implemented, so it is supported by more and more enterprises and governments. This paper analyzes the advantages and disadvantages of the Practical Byzantine Fault Tolerance (PBFT) in the alliance chain application scene. Aiming at the low efficiency of multinode consensus of the PBFT algorithm, the C‐Raft‐PBFT consensus algorithm is proposed. By integrating the Raft algorithm and the PBFT algorithm with the credit mechanism, designing node credit evaluation and grading protocols, and increasing Byzantine node detection based on feedback mechanism and other methods, the system efficiency is improved. The experiment results show that the improved algorithm has better throughput and lower delay, and the system’s fault tolerance is also improved. Among them, the delay is reduced by 1.93 seconds on average; in the case of an increase in system nodes, the number of nodes in the experimental data is between 200 and 225, and the throughput is increased by 6.46% on average.
- 1 January 2021
- Published by Wiley in Wireless Communications and Mobile Computing
- Vol. 2021 (1)
Open AccessImproved Byzantine Fault‐Tolerant Algorithm Based on Alliance Chain
Alliance chain is a typical multicenter block chain and is easily implemented, so it is supported by more and more enterprises and governments. This paper analyzes the advantages and disadvantages of the Practical Byzantine Fault Tolerance (PBFT) in the alliance chain application scene. Aiming at the low efficiency of multinode consensus of the PBFT algorithm, the C‐Raft‐PBFT consensus algorithm is proposed. By integrating the Raft algorithm and the PBFT algorithm with the credit mechanism, designing node credit evaluation and grading protocols, and increasing Byzantine node detection based on feedback mechanism and other methods, the system efficiency is improved. The experiment results show that the improved algorithm has better throughput and lower delay, and the system’s fault tolerance is also improved. Among them, the delay is reduced by 1.93 seconds on average; in the case of an increase in system nodes, the number of nodes in the experimental data is between 200 and 225, and the throughput is increased by 6.46% on average.