Search

The automation of coal mine production processes has largely relied on artificial intelligence (AI) technology to analyze industrial data. However, AI models developed for single application scenarios prove inadequate for the complex environments in coal mining. Relying solely on distributed computing to process the input features of AI models has led to decreased application efficiency. To address these challenges, an automatic reasoning technology for AI models in coal mine industrial data was developed. The system architecture consisted of three layers: the data layer, the computation-driving layer, and the model reasoning layer. The data layer gathered and stored various types of monitoring data, supplying raw data to the computation-driving layer. The computation-driving layer converted this vast amount of raw data into input features for AI models tailored to coal mining applications. An automatic switching mechanism between two computational engines—based on the input feature values—intelligently selected either Spark-based distributed computing or Python-based local computing, depending on the data volume, thereby resolving the issues of slow processing speeds and high latency in large-scale data applications. In the model reasoning layer, the input features were fed into the AI models for reasoning. A collaborative reasoning mechanism, with multiple triggering methods—scheduled, manual, and feedback-triggered—was introduced to enhance the effectiveness of AI models in complex coal mining scenarios. The results demonstrate that this technology enables rapid calculation of input features for multiple AI models across different application scenarios, as well as fast, automatic, and collaborative reasoning.

A precise and fast digital twin mapping method for hydraulic support attitude in fully mechanized working face is proposed to address the problems of low precision, large time delay, and difficulty in balancing precision and time delay in implementing attitude mapping using digital twins. The study designs a digital twin system architecture for hydraulic support attitude, which includes physical perception layer, data layer, business logic layer, and presentation layer. At the business logic layer, by retaining the external shape of the high-precision model of the hydraulic support and merging key parts into lightweight structural components, a digital twin model of the hydraulic support attitude is established. It reduces rendering delay. By establishing a conversion relationship between the angle measured by the inclination sensor and the rotation angle of the digital twin model, consistent of virtual-real mapping of the digital twin model of the hydraulic support attitude and real entities of the hydraulic support is achieved, ensuring the precision of the hydraulic support attitude mapping. At the data layer, transmission intervals are divided from each data transmission link and parameter constraints are applied. Parameter programming is applied to optimize the data update interval of adjustable links, reducing resource waste and latency in the data transmission process of digital twin systems. A precise and fast digital twin mapping platform for hydraulic support attitude is built, and the time delay and precision of virtual real mapping for hydraulic support attitude are tested. The results show that this method has a lower time delay while ensuring the precision of hydraulic support attitude mapping.

The control of information is a defining feature of the information age, and the optical modulator likewise has a crucial role in optical networks. The transmission, processing, and storage of data have demanded low energy consumption and high speed for photonic systems, promoting the development of electro-optic modulators to all-optical modulators. Although these all-optical modulation methods eliminate the photoelectric conversion, the disadvantage of volatile materials requiring continuous power supply when processing and retaining data in new materials-based devices increase energy consumption. We propose a Ge2Sb2Te5 (GST) integrated all-optical, nonvolatile optical modulator for in-fiber operation. The pulse-induced GST phase transition changes the reflectivity of the fiber end face, and this difference affects the result of the interference, achieving a modulation of output light intensity in interference spectra. The experimental results reveal that the device has obtained 13 dB interference intensity contrast in the telecommunications bands, and its response to a pump pulse is around 100 ns. Furthermore, we demonstrated the operation of the device as a scalar multiplication unit and a logic operation unit. The signal can be transmitted, processed, and stored in the fiber without photoelectric conversion. With the benefits of the switching power consumption of less than 100 nJ and the nonvolatile nature of GST, the device will be more energy-efficient in synchronous processing and storing. This in-fiber operating modulator lays the foundation for developing all-optical devices and networks.

The deep integration of the coal industry and artificial intelligence (AI) is an important path for modern mines to achieve intelligent personnel reduction, cost reduction, and efficiency improvement. AI empowerment in the entire process and business application scenarios of the coal industry is a specific technical measure to achieve coal mine intelligence. In the context of the current development of intelligent coal mines, a basic paradigm for the evolution of primary intelligent coal mines to new generation intelligent coal mines has been proposed. The composition, functions, and technical connotations of primary intelligent coal mines and new generation intelligent coal mines have been compared and analyzed. It is pointed out the importance of AI empowerment technology for new generation intelligent coal mine and its two key applications and implementation: the coal mine industry mechanism AI model and the coal mine Industry internet platform. The paper summarizes the current research status of industrial mechanism AI models for complex operations such as coal mine geology, mining, excavation, and safety monitoring. The paper clarifies the rapid development trend of industrial mechanism AI analysis in intelligent coal mine construction. A new generation of intelligent coal mine multi-level cloud edge collaborative industrial Internet platform architecture is designed. Using industrial information software and hardware facilities such as group data center, mine data center, production system centralized control center, and combining the features of massive data cloud computing and small amount of data edge computing, a multi-level cloud edge collaborative mechanism of group cloud, mine cloud and link edge, scene edge is proposed. It is pointed out that further research directions in the future should continue to strengthen the development and software research of AI models for coal mining industry mechanisms. Gradually a knowledge software system empowered by AI throughout the entire process of coal mining will be formed. It is suggested to fully utilize the digital resources and information facilities of the coal mining industry Internet platform to gradually realize the AI technology support of the coal mining industry Internet platform.

Background Methods Results Conclusions The erector spinae plane block (ESPB) and quadratus lumborum block (QLB) are two novel interfascial plane block techniques with possible analgesic effects for both incisional and visceral pain. However, the results of the intercomparison of the two techniques for analgesia after abdominal surgery remain controversial.A systematic literature search was performed on five databases for randomized controlled trials comparing the analgesic efficacy of ESPB and QLB in abdominal surgery. The primary outcome was the 24 h postoperative intravenous IV morphine‐equivalent consumption. A meta‐analysis was performed using a random‐effects model, with subgroup analyses based on the types of surgery and approaches of QLB.The 24 h postoperative IV morphine‐equivalent consumption was lower in patients receiving ESPB than in those receiving QLB (MD –2.307 mg; 95% CI ‒4.577 to −0.038; p = 0.046; and I2 = 96.5%), though the reduction did not reach clinically meaningful difference. Static and dynamic pain at different postoperative time points, the time to first rescue analgesia, and the incidence of PONV showed no significant difference between the two groups. However, the QLB group demonstrated a significantly prolonged time in performing the block compared to the ESPB group (MD –2.985 min; 95% CI –4.608 to –1.363; p < 0.001; and I2 = 97.5%).Based on the available evidence from a systematic search of databases, the single‐shot ESPB exhibits similar analgesic effects as QLB in abdominal surgery. The incidence of PONV was similar between the two techniques, with no block‐related complications identified though QLB might be a more technically challenging approach than ESPB. [ABSTRACT FROM AUTHOR]

The characterization and analysis of rock types based on acoustic emission (AE) signals have long been focal points in earth science research. However, traditional analysis methods struggle to handle the influx of big data. While signal processing methods combined with deep learning have found widespread use in various process analyses and state identification, effective feature extraction using progressive fusion technology still faces challenges in the field of intelligent rock type identification. To address this issue, our study proposes a novel framework for rock type identification based on AE and introduces a new signal identification model called 3CTNet. This model integrates convolutional neural networks (CNNs) and Transformer encoder, intelligently identifying AE of different rock fractures by establishing dependencies between adjacent positions within the data and gradually extracting advanced features. Furthermore, we experimentally compare five oversampling methods, ultimately selecting the adaptive synthetic sampling method (ADASYN) to balance the dataset and enhance the model's robustness and generalization ability. Comparison of the internal structure of our model with a series of time series processing models demonstrates the effectiveness of the proposed model structure. Experimental results showcase the high identification accuracy of the intelligent rock type identification model based on 3CTNet, with an overall identification accuracy reaching 98.780%. Our proposed method lays a solid foundation for the efficient and accurate identification of formation rock types in geological exploration and oil and gas development endeavors. [ABSTRACT FROM AUTHOR]

Rock fracture acoustic emission (AE) signals are commonly used non-destructive testing data in geological exploration, resource exploitation, and engineering fields. However, these signals are often accompanied by noise interference caused by environmental factors. In this study, we propose an enhanced model for denoising rock fracture AE signals, called simplified fully convolutional denoising autoencoder (SFCDAE). This model is based on the denoising autoencoder principle in the field of deep learning neural networks. The SFCDAE model consists of only seven layers, with minimal preprocessing of data input. By comparing denoising performance evaluation indicators, higher peak signal-to-noise ratio (PSNR) and lower root mean square error (RMSE) were achieved. On average, PSNR increased by 5.575% and RMSE decreased by 22.225%. Using simulated environmental noise to validate the model, it was found that the model has good robustness and can remove artefacts from sudden noise. The practical application value of the LSTM classification model was validated using data containing real experimental noise, resulting in a higher classification accuracy of 80.083%. These results indicate that the proposed model has better denoising performance compared to existing intelligent models and has certain practical value. [ABSTRACT FROM AUTHOR]