With the continuous progress of intelligent power system technology, in order to meet the needs of substation operation and maintenance, a target detection algorithm is applied to identify the status of equipment switches. YOLOv7, as the latest achievement of YOLO (You Only Look Once) series algorithms, has good speed and accuracy in target detection tasks. However, when the generalized network is applied in a specific scenario, its advantages are not obvious due to its high weight and poor portability. In this paper, an improved GF-YOLOv7 network model is proposed to apply in the recognition of the status of bounce locks in a substation. The MobileViT module is used to improve the feature extraction ability of the backbone network. Referring to the CBAM feature attention mechanism, the channel attention module and the spatial attention module are used to design a more lightweight feature fusion network. The experimental results in the test set show that the proposed network can significantly reduce the network weight and improve the detection accuracy on the basis of a small reduction in the detection speed, and the accuracy reaches 97.8%, which can meet the needs of the detection task of substation bounce locks.

The study of unmanned ground vehicles (UGVs) operating under unstructured roads is of great significance to intelligent transportation, agricultural development and military technology. In order to ensure reliable and stable operation of UGVs on unstructured terrain, it is necessary to identify the current road terrain and perform vehicle stability adjustment. Road terrain identification is a prerequisite for stability control. Most of the existing road terrain identification methods use a single vehicle sensor, which has the problem that complex algorithms need to be applied for data processing, which reduces the real-time performance. Moreover, the single sensor is weak in anti-interference and limited in recognizing the road. To address these problems, a method is proposed to collect vehicle motion data using on-board gyroscope sensors and velocity sensors. Back propagation (BP) neural network is used to identify the category of the road. For the problem that the conventional proportional-integral differential (PID) algorithm cannot be adapted to different road stability control, a multi-loop adaptive proportional-integral differential (PID) control system with the velocity loop as the outer loop and the torque (current) loop as the inner loop is proposed. In order to verify the feasibility and effectiveness of the method, experiments are conducted on a UGV using robot operating system (ROS), and the results verify the feasibility and superiority of the road identification and stability control method proposed in this paper. It provides a good theoretical basis and valuable technical guidance for the UGV operation and control on unstructured roads.

In practical applications involving robotic arms, particularly in tasks such as manipulating door handles, improper strategies often lead to excessive contact forces. Such forces not only jeopardize the integrity of the robotic arm’s joints but also pose a risk of damaging the door handle. This paper delves into a meticulous study aimed at refining the opening techniques employed by manipulators, enhancing their adaptability across various environments. A novel method is introduced, amalgamating force information feedback with the deep deterministic policy gradient algorithm, fostering a more nuanced approach in trajectory planning. This innovative strategy is meticulously evaluated through simulations and physical experiments, proving instrumental in guiding the robotic arm toward the successful completion of the door-opening task. The findings from the experiments underscore the algorithm’s prowess in cultivating a compliant door-opening strategy, harmonizing with the force applied at the manipulator’s end. A comparative analysis with conventional methods reveals a notable reduction in the end force of the manipulator, facilitating a more efficient and secure execution of door-opening operations.