In order to resolve the problem that the sample of image for internal detection of DN100 buried gas pipeline microleakage is single and difficult to identify, a recognition method of microleakage image of the pipeline internal detection robot is proposed. First, nongenerative data augmentation is used to expand the microleakage images of gas pipelines. Secondly, a generative data augmentation network, Deep Convolutional Wasserstein Generative Adversarial Networks (DCWGANs), is designed to generate microleakage images with different features for detection in the pipeline of gas pipelines to achieve sample diversity of microleakage images of gas pipelines. Then, a bi-directional feature pyramid network (BiFPN) is introduced into You Only Look Once (YOLOv5) to retain more deep feature information by adding cross-scale connecting lines in the feature fusion structure; finally, a small target detection layer is constructed in YOLOv5 so that more shallow feature information can be retained to achieve small-scale leak point recognition. The experimental results show that the precision of this method for microleak identification is 95.04%, the recall rate is 94.86%, the mAP value is 96.31%, and the minimum size of identifiable leaks is 1 mm.

For the problem of difficult tracking of small leaks in buried gas pipelines, a video detection of small leaks in buried gas pipelines method is proposed for the detection robot inside buried gas pipelines. Firstly, collecting images and videos of leaks inside buried gas pipelines to establish a dataset. Secondly, building a video detection model for small leaks in buried gas pipelines, introduce a bidirectional feature pyramid network into YOLOv5s (You Only Look Once), and build a feature fusion network to enhance the model’s ability of fusing small leaks. Thirdly, building a small target detection layer and a small target detection head in YOLOv5s classification prediction network to enhance the model’s ability of fusing small leaks. Fourthly, the video detection model for small leaks in buried gas pipelines is trained using the dataset. Lastly, the video detection effect of small leaks of the model is verified through leak detection experiments in various situations. The experimental results show that the precision rate of this method is 94.1%, the recall is 94.8%, and the average precision is 94.5%, which has a good detection effect and strong generalization ability.

To address difficulties in the early detection of small gas pipeline leaks, a method is proposed for the image recognition of micropores inside small gas pipelines. First, we design a feature fusion network (Neck) for a small gas pipeline internal micropores identification network model, introducing BiFPN (bi-directional feature pyramid network) into the Neck, and enhancing the micropores feature fusion capability of the network model using cross-scale feature information fusion. Second, we design a classification prediction network (Head) for a small gas pipeline internal micropores identification network model, constructing a small target detection layer in the classification prediction network, and improving the microporosity detection rate of the network model by improving the detection capability for small leakages. Then, micropore datasets are fed into the network model separately to train the model. Finally, comparison and ablation experiments are used to verify the micropore recognition ability of the network model. The experimental results show that the precision of the image recognition method for micropores inside small gas pipelines is 94.7%, the detection rate is 96.6%, and the average precision is 95.5%, which can better realize the early detection of leakages in small gas pipelines.

In response to the localization challenges posed by multi-modal motion mode transitions during the locomotion of tracked-wheel robots in sheltered spaces, this paper presents a combination localization method based on Inertial Navigation System/Odometry (INS/OD). Leveraging the unique locomotion structure of tracked-wheel robots, a multi-modal odometry motion model and non-holonomic constrained (NHC) velocity model are established. A closed-loop extended Kalman filter is employed to construct the combination localization model, facilitating real-time computation and correction of the robot’s heading angle using the odometry motion model, effectively mitigating system errors and preventing heading divergence. Practical localization tests were conducted on tracked-wheel robots to assess the performance of the proposed method. The results demonstrate that the combination localization method achieves position errors within 0.3578 meters, representing only 0.31% of the total travel distance, with an average error of 0.1611 meters, thus satisfying the practical localization requirements of tracked-wheel robots.