In order to realize the real-time detection of an unmanned fishing speedboat near a ship ahead, a perception platform based on a target visual detection system was established. By controlling the depth and width of the model to analyze and compare training, it was found that the 5S model had a fast detection speed but low accuracy, which was judged to be insufficient for detecting small targets. In this regard, this study improved the YOLOv5s algorithm, in which the initial frame of the target is re-clustered by K-means at the data input end, the receptive field area is expanded at the output end, and the loss function is optimized. The results show that the precision of the improved model’s detection for ship images was 98.0%, and the recall rate was 96.2%. Mean average precision (mAP) reached 98.6%, an increase of 4.4% compared to before the improvements, which shows that the improved model can realize the detection and identification of multiple types of ships, laying the foundation for subsequent path planning and automatic obstacle avoidance of unmanned ships.

Considering the high requirements of current kiwifruit picking recognition systems for mobile devices, including the small number of available features for image targets and small-scale aggregation, an enhanced YOLOX-S target detection algorithm for kiwifruit picking robots is proposed in this study. This involved designing a new multi-scale feature integration structure in which, with the aim of providing a small and lightweight model, the feature maps used for detecting large targets in the YOLOX model are eliminated, the feature map of small targets is sampled through the nearest neighbor values, the superficial features are spliced with the final features, the gradient of the SiLU activation function is perturbed, and the loss function at the output is optimized. The experimental results show that, compared with the original YOLOX-S, the enhanced model improved the detection average precision (AP) of kiwifruit images by 6.52%, reduced the number of model parameters by 44.8%, and improved the model detection speed by 63.9%. Hence, with its outstanding effectiveness and relatively light weight, the proposed model is capable of effectively providing data support for the 3D positioning and automated picking of kiwifruit. It may also successfully provide solutions in similar fields related to small target detection.

In order to investigate the mechanism of interaction between a vertical rotary tiller and the soil to obtain the best structural and operational parameters of the tool, improve the operational quality of the rotary tiller, achieve the effect of soil breaking and leveling to meet the agronomic requirements of tillage, a series of simulation tests was conducted through the design of 3D models of rotary tillers with different structural forms and the building of discrete element models of the action between the rotary tiller and the soil. The virtual simulation of different operating parameters and structural parameters of vertical rotary tiller was carried out by EDEM, and the effects of operating parameters, tool bending angle, and cutter installation number on the operating effect were analyzed. The parameter combination with the best operating effect was obtained through orthogonal test analysis and verified by soil trough test. A soil tank test showed that, with the optimal combination of operating and structural parameters, the average soil breaking rate of the vertical rotary tiller was 81.3%, which meets the agronomic requirements of tillage operation, and the actual power consumption was 6% higher than the simulation value, which was sufficiently close to verify the validity of the simulation. This study can lay a foundation for optimization research on the vertical miniature rotary tiller.

The combination of ultra-wide band (UWB) and inertial measurement unit (IMU) positioning is subject to random errors and non-line-of-sight errors, and in this paper, an improved positioning strategy is proposed to address this problem. The Kalman filter (KF) is used to pre-process the original UWB measurements, suppressing the effect of range mutation values of UWB on combined positioning, and the extended Kalman filter (EKF) is used to fuse the UWB measurements with the IMU measurements, with the difference between the two measurements used as the measurement information. The non-line-of-sight (NLOS) measurement information is also used. The optimal estimate is obtained by adjusting the system measurement noise covariance matrix in real time, according to the judgment result, and suppressing the interference of non-line-of-sight factors. The optimal estimate of the current state is fed back to the UWB range value in the next state, and the range value is dynamically adjusted after one-dimensional filtering pre-processing. Compared with conventional tightly coupled positioning, the positioning accuracy of the method in this paper is improved by 46.15% in the field experimental positioning results.