Search

The existing multi-target tracking algorithms for underground miners has the problems of slow detection speed and low recognition precision. In order to solve the above problems, a joint algorithm of multi-target detection and tracking algorithm based on the improved YOLOv5s model and the improved Deep SORT algorithm is proposed. In the part of multi-target detection, the YOLOv5s-GAD model is obtained by improving YOLOv5s model. The GhostConv module and the depthwise separable convolution (DWConv) module are introduced to replace the BottleneckCSP module in the YOLOv5s model backbone network and path aggregation network respectively. Therefore, the feature extraction speed is improved. Considering the characteristics of dark underground light and many noisy images, the efficient channel attention neural network (ECA-Net) module is introduced into the minimum feature map to improve the model's overall precision. In the part of multi-target tracking, the omni-scale network (OSNet) is used to replace the shallow residual network in Deep SORT to carry out omni-directional feature learning. Therefore, pedestrian re-identification and target tracking precision are improved. The experimental result shows that on the custom dataset Miner21, the YOLOv5s-GAD model average preciscom (when the intersection of union ratio is 0.5) reaches 97.8%, and the frame rate reaches 140.2 frames/s. The multi-target detection effect is better than the commonly used Faster RCNN, YOLOv3 and YOLOv5s models. On the public miners dataset MOT17, the speed and accuracy of the multi-target detection and tracking joint algorithm are better than those of IOU17, Deep SORT and other common multi-target tracking algorithms. The proposed model has the least number of personnel identity conversions and the best miner re-recognition effect. The joint algorithm of multi-target detection and tracking for underground miners can detect and track underground miners in time, and the multi-target tracking effect is good.

The quasi-in-situ microstructure and texture evolution along the real flow path of pure magnesium during friction stir welding were investigated. Five representative stages were involved from the base metal to the formation of the final stir zone. The material experienced compression, preheating, acceleration, deceleration, and subsequent annealing over the course of the welding process. A highly concentrated (0001) texture, denoted as “orientation convergence”, was initiated at the beginning of the acceleration stage (shearing deformation zone) in front of the tool. Both continuous and discontinuous dynamic recrystallization occurred simultaneously in the stir zone, and continuous dynamic recrystallization was determined to be the primary recrystallization mechanism. The marker material morphology and EBSD data were used to elucidate the in-situ evolution of the shear direction and shear plane along with the real flow path.

In view of problems of premature convergence of application of traditional genetic algorithm in distribution network planning in mining area to affect accuracy of planning, a multi-population simulated annealing and genetic algorithm was proposed. The algorithm takes the minimum year planning cost as objective function , and combines with advantage of simulated annealing algorithm, adds multi-population characteristics at the same time, so as to solve the problem of premature convergence in power distribution network planning and improve search efficiency and convenient to obtain the global optimal solution in the planning. The experimental results show that cost planning, iteration times are significantly reduced by use of multi-population simulated annealing and genetic algorithm to optimize mine distribution network, the running time is reduced by about six percent, the error rate is reduced by about three percent compared with the original algorithm, and the algorithm is more effective and efficient.

In view of problems that traditional risk assessment and reliability analysis methods of power grid focused on relay protection reliability indicators and did not involve change of reliability and actual risk in operation of relay protection device, the paper put forward a method that used three-parameter Weibull distribution to estimate time-varying failure rate, and established Markov model with 10 state based on multiple factors. It obtained time-varying failure rate function model through analyzing and processing data of power grid operation and management by Matlab simulation software; in combination with software and hardware and management factors, it also built process hidden risk assessment system of relay protection device based on Markov model by PSS/E software. The experimental results show that the system can early find and rectify hidden danger in power grid, which improves management level of the power grid, and realizes safe and stable operation of the power grid.

The structure, characteristics and working principle of extrinsic F-P sensor were introduced in details, and application of the extrinsic F-P sensor in gas detection system was expounded. The sensor takes hollow core photonic crystal fiber as isotopic carrier, is not affected by environmental temperature, which has characteristics such as high sensitivity, strong anti-jamming capability and low transmission loss, and can significantly improve accuracy and precision of gas concentration detection.

This paper presents a double frequency conductivity measurement method for measuring mine bursting water, to solve the capacitance effect of the conductivity sensor itself has the help. The core controller of the system is the single chip microcomputer ATMEGA128. This paper introduces the basic principle of the measurement of the existing problems and the dual frequency measurement method, and then introduces and analyzes the hardware. To test and analyze the collected data, the double frequency method is found to have good stability and accuracy in the measurement of the electrical conductivity of mine inrush water. It is proved that the method and the system design of the hardware circuit can accurately measure the electric conductivity of the mine inrush water source.

With the acceleration of urbanization, building energy consumption carbon emissions (BECCE) were increasing, weightily influencing global warming and the social-economic developments in cities. The detailed exploration of the most effective distance of urban three-dimensional (3D) compact form on BECCE guides urban space optimization. In this study, 135 People's Banks of China (PBC) were taken as samples. Based on the banks building features, socioeconomic conditions, macroclimate, and urban 3D compact form at different buffers, the most effective distance on the BECCE was determined. According to Partial least-squares regression (PLSR), the results show that the most effective distance of the urban compact form on the BECCE was 150 m. The study divided the normalized 3D compactness index (NVCI) values from 135 buildings for the distance of 150 m into five categories and used the geographic detector to recognize statistically obvious differences between the subregions. The natural breakpoint method was used to categorize the study area into five classes, specifically: low, medium-low, medium, medium-high, and high compact form. Based on the geographic detector method result, the compact form of medium, medium high, and high compact form was optimized to low, medium low. The optimization can effectively reduce the BECCE due to changing the compact form of the building by 65.2% or 65.7%. The study combined mathematical statistics with spatial analysis approaches to identify the most effective distance for reducing energy consumption. Our results will contribute toward considerable reductions in the BECCE. HIGHLIGHT: Innovatively used mathematical statistics to determine the most effective distance of urban compact forms. Demonstrated the influence of on BECCE to urban compact forms. Determined the most effective distance was 150 m. Verified the most effective distance was 150 m in different building climate zones. Optimized the urban 3D compact form within 150 m to reduce the BECCE. [ABSTRACT FROM AUTHOR]

In smart grid and smart building environments, accurate forecasting of load demand in residential buildings is of critical importance. This helps to enhance the stability of the power system, facilitate the integration of distributed renewable energy sources, and develop efficient demand response strategies. In view of this, this paper proposes a day‐ahead power load regulation assessment model based on the maximum information coefficient (MIC) combined with an improved Marine Predators Algorithm (IMPA) to optimize the extreme learning machine (ELM). The daily lagged load is used to construct the initial feature set, and the MIC is used for feature selection to filter out the top five features with the largest values. The MPA improvement strategy includes self‐mapping to generate chaotic sequence initialization and boundary mutation operations. The experimental results show that the proposed model has the optimal performance in evaluating the load regulation amount compared to other models. Taking Mean Absolute Error (MAE) as an example, compared with ELM and MPA‐ELM, IMPA‐ELM improved by 13.85% and 3.04%, respectively. [ABSTRACT FROM AUTHOR]

ABSTRACT Air conditioning load is a crucial demand response resource for optimizing energy consumption control, and its accurate analysis provides an essential basis for achieving efficient energy management. We aim at solving the problems of scarcity, single type, low accuracy and difficult construction of high‐quality data sets available for air conditioning operation characteristic models at present. This paper proposes a construction method of air conditioning operation characteristic model based on an improved seagull optimization algorithm to optimize deep belief network (ISOA‐DBN). Firstly, the data set for the study of air conditioning operation characteristics is obtained through experiments. Secondly, the Restricted Boltzmann Machine (RBM) and Deep Belief Network (DBN) are used to study the operating characteristics of air conditioning. The results show that the model effect is better when DBN is used to study the operating characteristics of air conditioning, and the coefficient of determination reaches 0.9439. Then, the SOA is improved, and its performance is tested. The results show that ISOA performs better than SOA in the test of 14 standard functions. Finally, the ISOA is used to adjust the DBN parameters finely. The results show that compared with DBN and SOA‐DBN, ISOA‐DBN has a better model effect when used to study the operating characteristics of air conditioners, and the coefficient of determination reaches 0.9534. This can provide strong support for studying air conditioning operating characteristics under different working conditions and has broad application prospects in optimizing energy consumption control. [ABSTRACT FROM AUTHOR]

Additive friction stir deposition (AFSD) is a novel solid-state additive manufacturing method developed on the principle of stirring friction. Benefits from its solid-phase properties, compared with traditional additive manufacturing based on melting–solidification cycles, AFSD solves the problems of porosity, cracks, and residual stress caused by the melting–solidification process, and has a significant improvement in efficiency. In AFSD, the interaction between feedstocks and high-speed rotating print heads suffers severe plastic deformation at high temperatures below the melting point, ending up in fine, equiaxed recrystallized grains. The above characteristics make components by AFSD show similar mechanical behaviors to the forged ones. This article reviews the development of AFSD technology, elaborates on the basic principles, compares the macroscopic formability and material flow behavior of AFSD processes using different types of feedstocks, summarizes the microstructure and mechanical properties obtained from the AFSD of alloys with different compositions, and finally provides an outlook on the development trends, opportunities, and challenges to the researchers and industrial fields concerning AFSD. [ABSTRACT FROM AUTHOR]

A low-roughness ultra-thin copper foil was prepared by pulsed electrodeposition with a duty cycle 60%, pulse frequency of 600 Hz on titanium. The influence of sodium 3,3'-dithiodipropane sulphonate (SPS), hydroxyethyl cellulose (HEC), gelatin and collagen additives on the microstructure, mechanical properties and electrochemical behaviour of electrolytic copper foil was explored. Furthermore, the reaction mechanism of SPS and collagen additives on electrodeposited copper were discussed. The results showed that at 0.08 g L−1 collagen concentration, the lowest thickness, the highest microhardness and the optimal surface roughness were achieved (5.12 μm, 279.63 HV and 1.885 μm, respectively). X-ray diffraction results confirmed that electrolytic copper foils prepared when SPS was introduced into the blank solution had a preferred orientation of (220) texture, which benefitted from the synergistic effect of copper ions and additives. The intermediates formed by the additive and Cu+ occupied the active sites on the cathode surface that increased the nucleation sites for deposition. In addition, the formed complexes can act as a barrier to narrow ion deposition channels and inhibit the growth of Cu ions. [ABSTRACT FROM AUTHOR]

The application of convolutional neural network has been gradually introduced in the field of X-ray bone images. At present, there is less research on methods to automatically detect abnormal parts of the skeleton. The method improves on the Densenet network by adjusting the network structure, adding a convolutional block attention module (CBAM) attention mechanism to the original Densenlayer, fusing spatial attention and channel attention to suppress unnecessary features and strengthen the network's capability to extract image features, and optimizing the Transition block by fusing two pooling strategies, average and maximum, to increase the model's anti-interference capability. The data collected after the experiment show that the improved Densenet has increased accuracy in the detection of skeletal abnormalities at all sites compared with the traditional network. The average accuracy is 1–5% higher than the benchmark method, the highest elbow accuracy reached 77.62%, the lowest hand accuracy also reached 64.28% and the area under the receiver operating characteristic curve is increased by 3–7%, the highest reached 80.83%. [ABSTRACT FROM AUTHOR]