Recently, sound-based diagnosis systems have been given much attention in many fields due to the advantages of their simple structure, non-touching measurement style, and low-power dissipation. In order to improve the efficiency of coal production and the safety of the coal mining process, accurate information is always essential. It is indicated that the sound signal produced during the cutting process of the coal mining shearer contains much cutting pattern identification information. In this paper, the original acoustic signal is first collected through an industrial microphone. To analyze the signal deeply, an adaptive Hilbert–Huang transform (HHT) was applied to decompose the sound to several intrinsic mode functions (IMFs) to subsequently acquire 1024 Hilbert marginal spectrum points. The 1024 time-frequency nodes were reorganized as a 32 × 32 feature map. Moreover, the LeNet-5 convolutional neural network (CNN), with three convolution layers and two sub-sampling layers, was used as the cutting pattern recognizer. A simulation example, with 10,000 training samples and 2000 testing samples, was conducted to prove the effectiveness of the proposed method. Finally, 1971 testing sound series were recognized accurately through the trained CNN and the proposed method achieved an identification rate of 98.55%.
The phenomenon of blocking pipe would appear for a trailing suction hopper dredger (TSHD) rake arm pipe in the construction process. Thus, based on the theory of particle flow mechanics, a transient three-dimensional two-phase hydrodynamic and sediment mixture model was established in this paper. Besides, different particle sizes of sediment concentration and velocity in the piping in the rake arm of the law were analyzed to solve the problem of blocking pipe and the prevention of dredging to provide theoretical support. Then, a model predictive controller was designed to regulate the mudflow in the pipeline by controlling the mud pump speed and compared with the proportional-integral (PI) controller. According to the results, the particle size of sediment affects concentration distribution in the pipeline under the same construction conditions for the TSHD. Besides, the larger the particle size of the sediment, the more significant the difference in the sediment concentration distribution in the pipeline. Similarly, the flow velocity is another influencing factor behind the change in concentration distribution in the pipeline, the increase in the flow velocity will improve the uniformity of concentration distribution in the pipeline, and the critical velocity will maintain the concentration distribution within a reasonable range. In terms of transportation control, the designed predictive model controller is capable of reducing overshoot and shorten the time required for adjustment. To sum up, the research result not only provides a valuable reference for the theoretical analysis and a controller design of the pipeline transportation system in the TSHD but also verifies the feasibility of the computational fluid dynamics model used in the study of pipeline transportation mechanisms.
Currently, detectors have made significant progress in inference speed and accuracy. However, these detectors require Non-Maximum Suppression (NMS) during the post-processing stage to eliminate redundant boxes, which limits the optimization of model inference speed. We first analyzed the reason for the dependence on NMS in the post-processing stage. The result showed that a score loss in a one-to-many label assignment leads to the presence of high-quality redundant boxes, making them difficult to remove. To realize end-to-end object detection and simplify the detection pipeline, we propose herein a mixed label assignment (MLA) training method, which uses one-to-many label assignment to provide rich supervision signals, alleviating the performance degradation, and we eliminate the need for NMS in the post-processing stage by using one-to-one label assignment. Additionally, a window feature propagation block (WFPB) is introduced, utilizing the inductive bias of images to enable feature sharing in local regions. Through these methods, we conducted experiments on the VOC and DUO datasets; our end-to-end detector MA-YOLOX achieved 66.0 mAP and 52.6 mAP, respectively, outperforming the YOLOX by 1.7 and 1.6. Additionally, our model performed faster than other real-time detectors without NMS.
In order to solve the Instability of uninterrupted ship power connect to shore power and increase shore power usage, this paper adopts a strategy called RBF neural network adaptive sliding mode control. It can improve the large error of the traditional sliding mode control. The strategy adopted can suppress the occurrence of large disturbance of the system and improve the stability of the grid-connected inverter output during the process of grid connection. The simulation model is built in MATLAB/SIMULINK. Under the condition of low voltage 440V, the simulation results are compared with the results of traditional sliding mode control without RBF neural network. It verifies the effectiveness of sliding mode control and enhances its reliability.