In the wave of digital transformation, secure flow information and data authenticity verification have become vital challenges. This paper proposes an innovative model based on blockchain technology to build efficient, transparent immutable environment. The uses distributed ledger characteristics guarantee integrity introduces advanced identity management mechanisms trust algorithms enhance security in process. then improved consensus algorithm, combined with contract technology, ensure...

Abstract Research shows that there are some data sets with strong timing in reality. After certain processing, they can predict the changes of transactions future to extent and play a positive role guiding production. However, often have characteristics overall trend variability seasonal fluctuation, also affected by many random factors. Therefore, show non stationarity. In this paper, CEEMD-LSTM-ARIMA combined model is proposed. This method reconstructs variables decomposing time series...

Regular security facilities check plays a key role in preventing potential dangers production safety. Those generally include helmets, protective suits, and fire extinguishers. Amongst them, extinguishers are essential for hazard prevention extinguishing. However, manual checks of extinguishers' presence may suffer from low efficiency error-prone, so we developed real-time deep-learning framework, called FEDNet (Fire Extinguisher Detection Network), the automatic detection on scenes via...

Cut tobacco width plays an essential role in the production of cigarette industry. Current cut mainly has a range 0.9 mm, 0.95 and 1.05 mm. With increase width, proportion medium, as well long filaments final products, will increase. In contrast, when short broken shreds decreases, structural distribution stability products well. However, if leaf is too wide, rolled cigarettes tend to be empty, resulting low flammability slow ignition speed. Therefore, online closed-loop control system...

In the current information age, network security issues are becoming increasingly prominent. The vast amount of log data generated by systems carries crucial that aids in detecting threats and abnormal behavior. This research proposes a clustering method based on DBSCAN algorithm, aimed at identifying anomalies within networks. We introduce characteristics data, discuss limitations traditional methods, provide detailed description our algorithm its application analysis. effectively...

In the production, daily management and operation of enterprises, application computer technology has brought unlimited opportunities competitive advantages to enterprises. With wide in enterprises have stronger competitiveness adaptability, which laid a good foundation for their sustainable development. This paper introduces development software gives trend, disadvantages technology. The mode system is given. error rate this always low, with lowest 0.42% 1.33% traditional system. highest...

This paper studies the detection model of network access data tampering attack based on blockchain technology to solve problem over-dependence central server and easy in traditional environment. The uses decentralization encryption monitor user behavior real time through smart contracts, enhances protection with SHA-256 hash algorithm, combines consensus algorithm ensure consistency security. experimental results show that performs well detecting multiple types an accuracy 99.51% F1 score...

With the development of science and technology, modern agriculture has made great progress, but frequent agricultural safety events seriously affect people's trust in production even economic social development. Therefore, it is very important to trace origin products. Blockchain, as a new technology recent years, been widely used various source tracing fields due its characteristics decentralization, traceability, non-tampering. This paper designs tobacco system based on alliance chain. The...

Abstract Support vector machine (SVM) has good classification effect on small samples and data sets with errors. We propose to improve the Harris Hawk optimization algorithm (HHO) optimize SVM further its performance. The parameters of affect accuracy, but there are certain difficulties in selection parameters, so HHO is used select optimal SVM. easy fall into local optimum convergence accuracy affected some extent. Chaotic perturbation nonlinear time-varying updating strategy introduced...

In construction, coal mining, tobacco manufacturing and other industries, wearing helmets is crucial safety measure for workers, the monitoring of helmet plays a significant role in maintaining production safety. However, manual demands substantial human, material financial resources, will suffer from low efficiency are error prone. Therefore, we proposed lightweight real-time deep learning-based detection framework called YOLO-H, automatic detection. Our YOLO-H model was developed on...

The operation and maintenance data abnormal detection technology is currently hampered by difficult procedures low accuracy. primary purpose of this paper to investigate address the problem accuracy in combining artificial neural network models with deep learning technology. We choose long short term memory (LSTM) Encoder-Decoder LSTM model as object for further optimization because they perform better anomaly identification than other many measures. Also, altering parameters like loss rate,...

To prevent the serious impact on production efficiency caused by service interruption, data loss, and other issues resulting from server faults, this paper proposes a fault prediction model named Wavelet Packet Transform Probabilistic Neural Network(WPT-PNN). WPT-PNN loosely mixes wavelet packet transform probabilistic neural network to achieve quick localization signal denoising effects. The proposed is validated using operational gathered. experimental results suggest that can effectively...

To solve the low frequency, wide variety and changing problems in machine abnormal sound detection, we propose an unsupervised detection method based on residual autoencoder technology. This can automatically learn capture characteristics of sounds without need to mark a large number samples advance. First, Mel-frequency cepstral coefficients(MFCC) is used construct acoustic features normal sound, then reconstruct features, finally anomaly fraction between reconstructed by audio be measured...

The optimal foraging algorithm is a swarm intelligence based on the behavior of animal groups. This paper proposes an optimized with dynamic parameters to improve search performance by changing scaling factor increase convergence accuracy and speed. Testing 10 benchmark functions showed that modified optimization outperforms comparison algorithms in terms used solve 0–1 knapsack problem, experimental results show can achieve better results.

As an important economic pillar of Guangxi, Guangxi Tobacco industry is facing the problem increasing business data and complex management requirements. To solve these problems, a log acquisition system China based on big platform was proposed. The uses Rsyslog to collect large-scale in real time, then Hadoop store process data, hive as warehouse store, query analyze stored HDFS. Finally, MySQL used valuable provide services for front-end application. research results this paper efficient...