To address issues in traditional forest fire detection models, such as large parameter sizes, slow detection speed, and unsuitability for resource-constrained devices, this paper proposes a forest fire detection method, FEDS-YOLOv11n, based on an improved YOLOv11n model. First, the C3k2 module was redesigned using the FasterBlock module, replacing C3k2 with C3k2-Faster in both the Backbone network and Neck section to achieve a lightweight model design. Second, an EMA attention mechanism was introduced into the C3k2-Faster module in the Backbone, replacing C3k2-Faster with C3k2-Faster-EMA to compensate for the accuracy loss in small-object detection caused by the lightweight design. Third, the original upsampling module in the Neck was replaced with the lightweight dynamic upsampling operator DySample. Finally, the detection head was improved using the SEAM attention module, replacing the original Detect head with SEAMHead, which enables better handling of occluded objects. The experimental results show that compared to YOLOv11n, the proposed FEDS-YOLOv11n achieves improvements of 0.9% in precision (P), 1.9% in recall (R), 2.1% in mean precision at IoU 0.5 (mAP@0.5), and 2.3% in mean precision at IoU 0.5–0.95 (mAP@0.5–0.95). Additionally, the number of parameters is reduced by 21.32%, GFLOPs are reduced by 26.98%, and FPS increases from 48.2 to 71.8. The FEDS-YOLOv11n model ensures high accuracy while maintaining lower computational complexity and faster inference speed, making it suitable for real-time forest fire detection applications.

Aiming at the shock absorption and stability under complex terrain conditions, a six-wheeled food delivery robot with a suspension damping structure was designed. The food delivery robot is mainly composed of a take-out box, a chassis mobile structure and a suspension damping structure. The dynamic analysis of the suspension damping structure of the robot is carried out through the Lagrangian equation, and the offset of the wheels in different driving environments is obtained. Then, the zero-moment point method is used to analyze the stability of the food delivery robot in two non-horizontal movement environments of the left and right wheels and the front and rear wheels. Based on this, taking speed bumps and ramp terrain with different slopes as examples, the stability of the food delivery robot is analyzed by ADAMS simulation. The simulation and experiment results verify the rationality of the structure design of the food delivery robot with a suspension damping structure and the correctness of the theoretical analysis.

Path planning is one of the research hotspots for outdoor mobile robots. This paper addresses the issues of slow convergence and low accuracy in the Double Deep Q Network (DDQN) method in environments with many obstacles in the context of deep reinforcement learning. A new algorithm, Improve Double Deep Q Network (IDDQN), is proposed, which utilizes second-order temporal difference methods and a binary tree data structure to improve the DDQN method. The improved method evaluates the actions of the current robot using second-order temporal difference methods and employs a binary tree structure to store the results obtained from these methods, replacing the traditional experience pool structure. The environment is constructed using a grid method, programmed in the Python language, with two two-dimensional grid maps created for simple and complex environments. DDQN and four related deep reinforcement learning methods, such as Multi-step updates and Experience Classification Double Deep Q Network (ECMS-DDQN), are compared through simulation experiments with the IDDQN method. Simulation results indicate that the IDDQN method improves various path planning metrics compared to the DDQN method and other relevant reinforcement learning methods. In the simple environment, IDDQN method exhibits a 26.89% improvement in step convergence time, a 22.58% improvement in reward convergence time, and a 10.30% improvement in average reward value after convergence compared to the original DDQN algorithm. It also outperforms other simulated methods in the simple environment, although the difference is not significant. In the complex environment, the IDDQN method avoids falling into local optima compared to other methods, demonstrating the accuracy of its strategy in complex environments. Other methods show artificially high average reward values after converging in local optima, lacking reference value. In the complex environment, IDDQN method exhibits a 33.22% improvement in step convergence time and a 25.47% improvement in reward convergence time compared to the original DDQN algorithm, clearly surpassing other participating simulated methods. The data above indicate that the IDDQN method improves both convergence speed and accuracy compared to the DDQN method and the relevant improvement methods simulated in this paper. Particularly in environments with many obstacles, the performance improvement is evident, allowing for effective path planning in such environments.

The urban street is a congested environment that contains a large number of occluded and size-differentiated objects. Aiming at the problems of the loss of the target to be detected and low detection accuracy resulting from this situation, a newly improved algorithm, based on YOLOv4, DCYOLO is proposed. Firstly, a Difference sensitive network (DSN) is introduced to extract the edge features of objects from the original image. Then, assign the edge features back to increase the edge intensity of the object in the original image and ultimately improve the detection performance. Secondly, the feature fusion module (CFFB) based on context information is introduced to realize the cross-scale fusion of shallow fine-grained features and deep-level features, to strengthen the cross-scale semantic information fusion of feature maps and eventually improve the performance of object detection. At last, in the network prediction part, the SIOU loss function replaces the original CIOU loss function to improve the convergence speed and accuracy of object detection. The experiments based on MS COCO 2017 and self-made datasets show that, compared with the YOLOv4, the detection accuracy of DCYOLO models is greatly improved with an increase of 9.1 percentage points in AP and 10.4 percentage points in AP_s. Compared with YOLOv5x and Faster R-CNN, DCYOLO shows higher accuracy and better detection performance. The experiment result proves that the DCYOLO algorithm can adapt to the dense object detection requirements in the congested environment of urban streets.

Due to the inherent limitations of detection networks and the imbalance in training data, small-target detection has always been a challenging issue in the field of target detection. To address the issues of false positives and missed detections in small-target detection scenarios, a new algorithm based on STDA-YOLOv8 is proposed for small-target detection. A novel network architecture for small-target detection is designed, incorporating a Contextual Augmentation Module (CAM) and a Feature Refinement Module (FRM) to enhance the detection performance for small targets. The CAM introduces multi-scale dilated convolutions, where convolutional kernels with different dilation rates capture contextual information from various receptive fields, enabling more accurate extraction of small-target features. The FRM performs adaptive feature fusion in both channel and spatial dimensions, significantly improving the detection precision for small targets. Addressing the issue of a significant disparity in the number of annotations between small and larger objects in existing classic public datasets, a new data augmentation method called Copy–Reduce–Paste is introduced. Ablation and comparative experiments conducted on the proposed STDA-YOLOv8 model demonstrate that on the VisDrone dataset, its accuracy improved by 5.3% compared to YOLOv8, reaching 93.5%; on the PASCAL VOC dataset, its accuracy increased by 5.7% compared to YOLOv8, achieving 94.2%, outperforming current mainstream target detection models and small-target detection algorithms like QueryDet, effectively enhancing small-target detection capabilities.