4results about How to "Avoid smooth" patented technology

The invention provides a formation switching obstacle avoidance system and an obstacle avoidance method thereof, and relates to the technical field of robot obstacle avoidance, and the system comprises a formation member classification module, a follower distributed control module, a navigator centralized control module and a real-time obstacle avoidance control module. The formation member classification module is responsible for clearly dividing formation members into two roles of followers and pilots according to specific task requirements of the formation and robot characteristics; and the follower distributed control module is responsible for motion control of a follower, so that the follower can perform self-adaptive adjustment according to the motion of the navigator and the surrounding environment, including adjustment of the position and the speed of the follower. According to the formation switching obstacle avoidance system and the obstacle avoidance method thereof, the characteristics of affine transformation are integrated into formation obstacle avoidance, a new formation obstacle avoidance mode is realized, the formation size can be flexibly adjusted for obstacle avoidance in the obstacle avoidance process, and the integrity of the formation can be well kept in the obstacle avoidance process.

The utility model belongs to the technical field of cryopreservation tubes, and particularly relates to an auxiliary device facilitating opening and closing of a cryopreservation tube, the auxiliary device comprises the cryopreservation tube, a placing frame, a long handle hook and directional clamping tweezers, the cryopreservation tube comprises a tube body and a tube cover, the tube body is detachably connected with the tube cover in a threaded mode, the long handle hook is used for extracting the cryopreservation tube, and the directional clamping tweezers are used for clamping the cryopreservation tube. The directional clamping tweezers are used for opening a cryopreservation tube, and the placing frame is used for placing a tube cover; the cryopreservation tube can be simply and conveniently taken out from the hundred-hole box through the long-handle hook and the extraction ring, the situation that the surface of the tube body is smooth and the tube body is prone to falling off when clamped by tweezers is avoided, and the cryopreservation tube is conveniently opened and closed through the friction design of the directional clamping tweezers and the tube cover clamping lugs; the placing rack is combined with the extraction ring through the hollowed-out holes and the hanging needles, so that the tube cover of the cryopreservation tube is convenient to invert, and the conditions that the tube cover rolls over and is easy to pollute are reduced.

This invention provides a multidimensional collaborative enhanced graph clustering method. It extracts preliminary node embedding representations from each view in multi-view attribute graph data, employs three-level collaborative contrastive learning (3C-CL), calculates contrastive losses at three granular levels (node, group, and global), and obtains a unified embedding representation that integrates multi-view features. Gradient norms are used to dynamically adjust the loss weights of the contrastive losses at each level, and a clustering loss optimization network based on the Student's t-distribution is jointly used to output the final graph clustering assignment result. In sparsely connected views, the receptive field is automatically expanded to capture higher-order neighbor information, while in densely connected views, the aggregation depth is limited to avoid excessive information smoothing. This complementary system of structural information from micro to macro levels significantly improves cross-view semantic consistency, identifies difficult samples with similar attributes to the target node but different potential categories, and achieves more accurate sample selection.

The application relates to the technical field of photoacoustic signal denoising, in particular to a harmful gas monitoring device and method for a smart livestock farm, the method comprising the following steps: collecting a to-be-detected gas in the livestock farm and obtaining a photoacoustic signal of the to-be-detected gas; calculating a feature of the photoacoustic signal affected by a noise superposition signal, so as to determine a penalty factor when the photoacoustic signal is subjected to denoising processing, and realize photoacoustic signal denoising; processing the denoised photoacoustic signal into a photoacoustic signal, so as to detect a harmful gas concentration through a gas sound wave detection module in the harmful gas monitoring device. The application aims to improve the denoising effect of the photoacoustic signal, and finally effectively improve the detection accuracy of the harmful gas monitoring device.