Yanwen Li

Monocular one-dimensional galvanometer scanning systems have been widely used in the field of non-contact 3D scanning. However, this system has extremely high requirements for the temporal synchronization between motor control and camera acquisition, as well as the assembly accuracy of the mechanical structure. This paper proposes an online calibration method driven by image feature point data: by introducing an auxiliary structured light system, the intersection points of the Auxiliary Structured Light (ASL) and Measurement Structured Light (MSL) are used as the feature points input to the system, and a residual neural network (ResNet) is employed to realize the mapping process from feature point data to the MSL light plane normal vector. The acquisition of driving data required for the MSL calibration model and the measurement process are completely synchronized on the time axis, which avoids the problem of requiring synchronization between image and encoder data acquisition, and bypasses the rotation axis calibration process in the traditional physical model. In addition, based on the invariance of projective lines, this paper also proposes a dataset construction method and a two-stage training method suitable for this system. This dataset construction method can simplify the acquisition process of data required for neural network training, and the two-stage training mode can suppress the overfitting phenomenon during the training process and improve the generalization ability of the network. Experiments show that the diameter fitting error of the 3D reconstruction of the target sphere by this system is less than 40 μm.

This paper suggests an adaptive car-following strategy for autonomous connected vehicles (ACVs) that integrates a robust controller with an extended disturbance estimator (EDE) and reinforcement learning (RL) to improve performance in dynamic traffic environments. Traditional car-following methods struggle to handle external disturbances and uncertainties in vehicle dynamics. The suggested method addresses this by dynamically adjusting the EDE gain using RL, enabling the system to optimize its control strategy in real time continuously. Simulations were conducted in two scenarios, a single following vehicle and two following vehicles, each tracking a leading vehicle. Results showed significant improvements in velocity tracking, with the RL-based control method reducing velocity error by over 50% compared to conventional approaches. The technique also led to smoother acceleration control, enhancing stability and driving comfort. Quantitative metrics, such as total reward, velocity error, and acceleration magnitude, indicate that the suggested EDE-RL-based strategy provides a robust and adaptable solution for autonomous vehicle control. These findings indicate that RL, combined with robust control, can improve the performance and safety of ACV systems, making it suitable for broader applications in autonomous vehicle platooning and complex traffic scenarios, including vehicle-to-vehicle (V2V) communication.

Facing the development trend of electrification of construction machinery, in view of the drawbacks of the existing electro-hydraulic-load-sensitive system in terms of dynamic characteristics and usage of energy, based on the drive source of a servo motor-driven quantitative pump, an electro-hydraulic-load-sensitive system on the basis of torque open-loop control was proposed. Firstly, the working principle of the system was introduced and the system’s operating characteristics and energy consumption characteristics were theoretically analyzed. Secondly, in order to balance the system’s energy usage and maneuverability, a control strategy with a variable pressure margin was designed. Meanwhile, in order to solve the problem that the hydraulic pump’s mechanical efficiency causes system pressure control deviation, a torque compensation method based on offline data and speed prediction was proposed. Finally, simulation and testing were used to confirm the viability of the control strategy. The test results show that: the system could realize stable pressure margin control, and the response rise time was within 0.7 s under a variety of flow circumstances; the system could follow the control instruction to change linearly and the flow rate changed smoothly in the adjustable pressure gap control; after using the compensation method, the deviation of the pressure gap control was within 2%.

The electro-hydraulic servo pump control system (EHSPCS) is a volume control system that uses a permanent magnet synchronous motor (PMSM) with a fixed displacement pump to directly drive and control the hydraulic cylinder. The energy transmission law of the system is very complicated due to the transformation of electrical, mechanical and hydraulic energy as well as other energy fields, and qualitative analysis of the energy transfer efficiency is difficult. Energy transfer analysis of the EHSPCS under different working conditions and loads is proposed in this paper. First, the energy flow transfer mechanism was analyzed, and the mathematical and energy transfer models of the key components of the system were established to explore the energy characteristic state transition rule. Second, a power bond diagram model was built, its state equation and state matrix were deduced, and a system simulation model was built. Finally, combined with the EHSPCS experimental platform, simulation experiments were carried out on the dynamic position following and steady-state position holding conditions of the system, and the variation rules of the power of each energy characteristic state and the system energy transfer efficiency under different loads were obtained. The research results provide a foundation for the study of power matching and energy-saving mechanism of the EHSPCS.

Eight-degree-of-freedom redundant manipulators are considered more suitable for performing special tasks in unstructured environments due to their high degree of freedom and flexibility. The 8-DOF tunnel shotcrete manipulator is a kind of equipment that is applied to the concrete spraying support in the process of tunnel excavation. Since the mechanism itself is a redundant manipulator with link offset, its inverse kinematics is extremely difficult to solve. To address the problems that the conventional redundant manipulator inverse kinematics solution is difficult to solve analytically and the numerical algorithm requires several iterations with accumulated errors, a synthesis inverse kinematics algorithm based on the weighted minimum norm method and the joint angle parameterization method is proposed to overcome these defects. Firstly, a set of optimized joint angle approximation solutions is obtained based on the weighted least norm method of joint angle limit avoidance optimization function, and secondly, two joint variables are parameterized as known parameters in the optimized approximation solutions based on the joint angle parameterization method, and further, the analytical expressions of the remaining joint angles are derived by algebraic methods. Finally, the superiority of the inverse kinematic synthesis algorithm over the joint angle parameterization method in terms of joint motion is verified by comparing numerical examples of linear motion trajectories. Numerical simulation results show that the algorithm has the advantages of avoiding joint angle limits, high computational accuracy, no cumulative error, and the algorithm can meet the general real-time motion control requirements as demonstrated by the experimental analysis of algorithm time consumption.

In this paper, a novel two-degree-of-freedom (2DOF) spherical parallel mechanism(SPM) with symmetrical structure is proposed and analyzed. First, the models of forward kinematics and inverse kinematics are established based on D-H parameters, and the Jacobian matrix of the mechanism is obtained and verified. Second, the workspace of the mechanism is obtained according to inverse kinematics and link interference conditions. Next, rotational characteristics analysis shows that the end effector can achieve continuous rotation about an axis located in the mid-plane and passing through the rotation center of the mechanism. Moreover, the rotational characteristics of the mechanism are proved and motion planning is carried out. A numerical example is given to verify the kinematics analysis and motion planning. Finally, some variant mechanisms can be synthesized. This work lays the foundation for the motion control and practical application of this 2DOF SPM.