JGJinliang Gong

Jinliang Gong

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Jinliang Gong has a total of 9 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • A lightweight RT-DETR model for corn seedling and weed detection with adaptive feature fusion

    by Guangze Tian, Jinliang Gong, Yanfei Zhang

    Due to the large size of the existing corn seedling and weed detection models, they are difficult to deploy on resource-constrained platforms such as miniature weeding robots. Therefore, it is necessary to achieve a balance between detection accuracy and computational efficiency for the detection models. This paper proposed a lightweight detection model, RT-DETR-DSA, based on the RT-DETR-r18 model. Firstly, a dual-path convolution-attention block was designed to construct the backbone network, which improved the model’s computational efficiency by parallelizing the convolution and attention mechanisms. Secondly, based on the cross-scale feature interaction layer of the original model, this paper proposed the adaptive feature fusion block to complement and enhance the high and low-level features to improve the model’s detection accuracy. Finally, to further reduce the model’s computational load, a sliding-window self-attention block was introduced into the neck network, which divided the feature map into multiple non-overlapping windows and calculated the attention within each window separately. Experimental results showed that the proposed RT-DETR-DSA model was constructed with only 13.0M parameters, corresponding to a 35.6% reduction compared with the original RT-DETR-r18 model; its computational load was measured as only 34.9G, achieving a 40.6% reduction over the original model; the mAP@0.5 of the improved model was 1.2 percentage points higher than that of the original model; and its FPS (frames per second) reached 85.4. The RT-DETR-DSA model proposed in this paper improved computational efficiency while maintaining high detection accuracy in corn seedling and weed detection, and provided a potential technical solution for the visual perception system of miniature weeding robots.

  • Optimized design and testing of a multi-degree-of-freedom cable-driven underactuated picking manipulator

    by Junhui Sun, Jinliang Gong, Yanfei Zhang

    In order to address the issues of poor flexibility and complex control systems in existing picking manipulators, the present work designs a multi-degree-of-freedom cable-driven underactuated manipulator. Firstly, an end-effector drive module based on the cam mechanism is designed, which can control gripping force without the need for force sensor feedback, and enables the opening, closing, and rotary movements of the end-effector to be controlled by the same motor. Secondly, the influence of the structural parameters of the fingers on the contact force at the phalange is analyzed, and then the end-effector is optimized using a genetic algorithm. Next, the cable pulling force control model is established. Subsequently, a prototype manipulator is fabricated for testing, the results show that the adjustment range of cable pulling force is 11.1–27.4 N, and the output torque of manipulator is 0.64 N·m. Finally, the results of picking test in the orchard demonstrate that the manipulator is capable of accomplishing non-destructive picking operations.


  • Open Access

    Accurate Stiffness Modeling Method for Flexure Hinges with a Complex Contour Curve

    by Jinliang Gong, Yanfei Zhang, Kazi Mostafa, Xiang Li

    Background: Flexure hinges have certain advantages, such as a simple structure, smooth movement, no need for lubrication, frictionless movement and high precision. The flexure hinge’s transfer of force and displacement relies on its deformation. Thus, stiffness is an important index for evaluating hinge flexibility. Objective: Stiffness analysis of the flexure hinges is necessary to be performed. This paper aims to present a unified stiffness model solving method of the flexible hinges with complex contour curves. Methods: The transfer matrix of a flexure hinge was derived based on balance equations and the virtual work principle with consideration of axial, shear, and bending deformations. The element stiffness matrix of a flexure hinge was obtained from the relationship between the transfer and stiffness matrices. In this manner, the unified formula of element stiffness of a general flexure hinge was established. By using this method, rigidity models of parabolic, corner-filled, and the right circular flexure hinge have been deduced. By taking the right circular flexure hinge as an example, the results obtained using this method were compared with those of methods provided in other studies and the finite element results. Results and Conclusion: The comparison results revealed that the proposed method increases the rigidity accuracy because the effect of the uneven distribution coefficient of shear stress was considered. The stiffness error was within 7%, which demonstrates the validity of this method. In contrast to the other methods, the proposed method can be applied by determining the first integral element stiffness of a common flexible hinge. Moreover, the proposed method provides better commonality, flexibility, and ease of programming. In particular, it is much easier for the flexure hinges with a complex contour curve. Transitivity can be used to calculate the rigidity after the flexure hinge has been divided into subunits, thus making it unnecessary to convert to the global coordinate system.