The EEMD(Ensemble Empirical Mode Decomposition) algorithm faces challenges in aligning the generated IMF(Intrinsic Mode Function) components during the decomposition process. To address this issue, a novel denoising method that combines EEMD with the PRT(Phase Randomization Technique) is proposed, enhancing the denoising performance of the improved EEMD algorithm. By incorporating PRT, the method effectively handles nonlinear and nonstationary signals, significantly improving the stability and reliability of the IMFs, and enhances the performance of the EEMD algorithm in noisy environments. The experimental results strongly demonstrate the innovation’s value, as the EEMD-PRT algorithm shows superior performance compared to traditional methods by improving the signal-to-noise ratio and correlation coefficient of noisy signals, reducing the mean square error and mean absolute error. Furthermore, its effectiveness has been thoroughly validated in pipeline leak detection for pipes with varying diameters.

In order to encrypt and store communication network faster and more effectively, a communication network data encryption and the data of storage method based on improved RSA(Rivest-Shamir-Adleman) algorithm is proposed. Firstly, wavelet transform method is combined with empirical mode decomposition method of complementary set to denoise communication network data and improve the accuracy of communication network data. Then, the fuzzy C-means clustering algorithm is used to cluster communication network data, and similar data is uniformly encrypted to improve the efficiency of data encryption storage. Finally, the conventional distributed access management mode is replaced in communication networks with hash access data algorithms to enhance the security of data storage and prevent data loss. By improving the encryption process of RSA encryption algorithm, encrypted storage of communication network data is achieved. The experimental results show that the proposed method has good denoising effect, high security, and high encryption storage efficiency, making it suitable for encrypted storage of communication network data. 

Due to the lack of practical teaching resource of optical communication network, an all-optical access network practical teaching system based on virtual-real combination is designed. The communication service opening is regarded as main task, a real practical teaching system is built based on all-optical access network commercial equipment. The application layer server and communication user terminal are simulated by software and hardware combination. Configuration information from application layer server to communication user terminal through all-optical access network is planned and the success of various communication services opening is verified. The network troubleshooting is regarded as main task, a virtual simulation experiment teaching system of all-optical access network is developed by utilizing virtual reality technology. The virtual campus application scenarios including central office, fiber distribution terminal and university-enterprise joint laboratory are constructed. Concentrated presentation of optical distribution network, multi-service gateway, optical line terminal and application layer server related network troubles is realized. With the design and application of this practical teaching system, the teachers’ construction ability of optical network practical teaching resources is improved, and the students’ practice and innovation ability to solve complex problems in the application of optical network is cultivated. 

With the emergence of advanced communication technologies such as high-throughput satellites, the integration of power satellite and smart grid has become an inevitable trend, and its related security certification has become the focus of research. However, the existing authentication schemes are mainly studied under the isomorphic architecture focusing on ensuring the authenticity verification of entities. Therefore, a heterogeneous authentication scheme for smart grid based on power satellite is introduced. The proposed scheme realizes the authenticity verification of each communication entity in a heterogeneous environment, and provides a validity verification method for terminals. Security analysis confirms the correctness and security of the scheme, and performance analysis shows that the proposed scheme effectively reduce the time cost of authentication phase than the existing schemes. 

Due to the limitation of environmental and economic factors, the collected seismic records usually have a lot of noise interference, which may cause some obstacles to the subsequent seismic data processing. Effectively attenuating noise is a key issue in seismology. In recent years, CNNs ( Convolutional Neural Networks) have achieved some success in the field of seismic data denoising. However, weak signal recovery in the presence of strong background noise is insufficient for existing convolutional neural networks. To address the above issues, a denoising network called MAUnet(Multi-Scale U-Net and An Attention Fusion Mechanism) is proposed. based on a multi-scale U-Net and an attention fusion mechanism. MAUnet innovatively introduces a dual-mechanism architecture, where a multi-scale module enables the network to learn features at different scales. And an attention-based feature fusion mechanism allows the network to combine shallow high-frequency details with deep semantic information, enhancing its learning capability and achieving feature complementarity. Experimental results demonstrate that our method has better noise attenuation and recovery capability for weak signals than competitive methods. 

S-Band Polymer Waveguide Amplifier Based on Thulium Ytterbium Co-Doped Nanocrystals LIU Tingting, WANG Jiahe, ZHAO Dan, WANG Fei (College of Electronic Science and Engineering, Jilin University, Changchun 130012, China) Abstract: In order to solve the problem of insufficient communication bandwidth caused by the constantly growing network capacity, expanding the optical communication band from the main C-band (1 530 ~1 565 nm) to the S-band (1 460 ~1 530 nm) and L-band (1 565 ~1 630 nm) has become an effective way to address this issue. Tm3+can emit light at a wavelength of ~1 490 nm under pump excitation, and doping it into a waveguide can achieve S-band optical amplifier. The NaYF4 : Tm3+, Yb3+nanocrystals are introduced into polymer materials as amplifier gain media, a planar waveguide amplifier structure is designed, and the gain performance of the device is simulated in the S-band. Four waveguide amplifier samples with different Yb3+doping concentrations (x=5,10,15,20) are prepared using NaYF4 :1% Tm3+ and x% Yb3+ nanocrystals, and the device gain performance is tested. The experimental results show that when the Yb3+doping concentration is 10%, the gain of the waveguide amplifier is maximum, and the 1 cm long device achieves a relative gain of 8.2 dB at a pump power of 400 mW. The modeling and simulation of the Tm3+and Yb3+co doped system proposed in this paper provide theoretical guidance for the development of high-performance S-band waveguide amplifiers. Using NaYF4 , the polymer optical waveguide amplifier prepared by Tm3+ and Yb3+ nanocrystals has achieved S-band optical amplification and is expected to be widely used in optical communication.

In order to make up for the lack of selectivity of a single gas sensor in the face of a variety of gases and to identify a variety of gases more accurately, an improved convolutional neural network based on structural reparameterization technology and depth-separable convolution technology is proposed. It integrates the multi- branch convolution structure during model training into the single branch simple convolution layer during inference. In addition to simplifying the complexity of the inference model, the feature extraction ability of the model for gas response data is greatly enhanced. When this method is applied to a common data set of gas sensor array containing 10 common VOCs, the recognition accuracy reaches 96. 46%, and the accuracy reaches 97. 44% after adjusting the complexity of the model and adding the convolutional layer.

 In order to enhance the overall efficiency of the laser pump source system, a design for a load- adaptive semiconductor laser array constant current driver is presented. This driver can adaptively adjust the supply voltage of the load based on the number of series loads in the laser array and the changes in the drive current, thereby optimizing work efficiency. The maximum output voltage of the driver is 22 V, and the maximum output current is 1 200 mA. When driving eight series loads of the laser array, the efficiency can exceed 91. 6%. This load-adaptive technology provides a new approach for designing efficient semiconductor laser drivers. 

In order to solve the problem that most of the existing microelectronics major courses are based on theory and lack simulation experiments, a set of FPGA(Field Programmable Gate Array) microelectronics and integrated circuit design experiment system are designed based on RISC-V(Reduced Instruction Set Computer) CPU(Central Processing Unit). The ModelSim software compiler is used to simulate and verify, and FPGA is used as development platform to realize CPU system functions. Taking RISC-V reduced instruction set as the instruction set of the CPU and modularization as the design idea, the five-level pipeline CPU is designed from the local microprocessor to the whole. The five-level pipeline includes value, decoding, execution, memory access and write back. The system integrates software and hardware development to stimulate students' interest in learning. The experimental platform built gradually realizes the configuration and instruction set of CPU to the architecture, programming, simulation, writing and debugging of the whole CPU, enabling students to have a deep understanding of the design of integrated circuit system with FPGA, which is conducive to the study of professional theoretical courses. The design simulation content comes from the application of OBE(Outcomes- Based Education) teaching theory to integrated circuit EDA(Electronic Design Automation) course. This design method and content can also be applied to the combination of industry, university and research to improve innovation and entrepreneurship ability of students. 

To address the problem of determining the optimal strategy for air combat maneuver decision making with incomplete information, an evaluation method based on interval projection pursuit is proposed. Firstly, the interval number is introduced to represent the incomplete air combat data, the Euclidean distance in the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) is replaced by the grey relational degree, the positive and negative ideal solutions are given, and the RIGRA-TOPSIS(Referential Interval Grey Relation Analysis- Technique for Order Preference by Similarity to Ideal Solution) method is proposed to quantitatively obtain the objective weight of air combat indicators. Then, an improved interval projection pursuit method is introduced, which takes the obtained objective weights as the initial projection vector of the interval projection pursuit method, uses Gini coefficient instead of standard deviation to calculate the projection density, uses the particle swarm optimization algorithm to calculate the projection value, and determines the optimal strategy corresponding to the maximum projection value in the air combat game. The simulation results show that the proposed method has a good effect in the case of small sample air combat strategy set.

The anti-disturbance ability and dynamic performance are poor in the inverter stage of PET(Power Electronic Transformer)based on traditional VSG(Virtual Synchronous Generator)control. In order to address the problem, a VSG control combined with linear active disturbance rejection control is proposed. A second-order active disturbance rejection controller is added to the active part of the traditional VSG control. State observation and feedback are introduced to the controller to estimate the system error in real time, so that the frequency and power of the inverter can be stably tracked under different operating conditions, while the oscillation phenomenon is weakened. With the improved control strategy, the maximum overshoot is reduced by 54. 5 percent and the frequency fluctuation time is reduced by 0. 22 s, which make the PET inverter stage have stronger dynamic response and steady-state performance. The simulation demonstrates that the control strategy is feasible and effective. It also shows that the improved VSG control strategy accelerates the recovery speed of the PET inverter stage after being disturbed, and has better dynamic performance.

Magnus rotor is a new type of auxiliary propulsion device for ships. Addressing the disparity between the lift model of Magnus rotors and traditional lift formulae, a combined approach of theory and numerical simulation is employed for investigation. Initially, the geometric and flow domain models of the Magnus rotor are established. Subsequently, computational fluid dynamics software is utilized for grid independence verification and numerical validation. A numerical simulation method is then employed to analyze the aerodynamic characteristics of the Magnus rotor model, considering the impact of various wind speeds and rotor rotation rates on the generated thrust. Finally, based on the obtained data, adjustments are made to the traditional lift model, and the reliability of the modified model is verified by comparing simulation data under different conditions with literature data. The results indicate that the propulsion force of the Magnus rotor increases with the increase of rotation ratio, and the optimized lift model exhibits a high degree of fit with simulated values, demonstrating higher accuracy. 

For a class of uncertain networked stochastic T-S(Takagi-Sugeno) fuzzy systems with time-varying delay, quantization error, the problem of hybrid-triggered robust H∞ control system with two-terminal quantization is studied. Firstly, in order to alleviate the burden of networked communication, a hybrid-triggered scheme is adopted to reduce data transmission. The construction of quantizers at the sensor side and the actuator side is stadied to quantify the sampling and control signal respectively, and the quantization error is considered to improve the control accuracy of the system. By taking the effect of network-induced delay, uncertainty and quantization error are taken into consideration, the T-S fuzzy systems based on the hybrid-triggered mechanism is remodeled. Secondly, by selecting the delay-dependent and fuzzy basis-dependent Lyapunov function, and introducing the free weight matrix, the sufficient conditions for the double-terminal quantized fuzzy system are satisfied, and the influence of energy bounded noise signals on the output is suppressed under the H∞ performance index γ. Finally, the simulation proves that the proposed scheme can reduce data transmission effectively, improve the system control accuracy, and reduce the conservatism of design. 

Aiming at the problem of limited communication conditions in the control of severe slug flow in the horizontal pipe-downward inclined pipe-riser system of submarine mixed transportation, a MPC (Model Predictive Control) strategy based on DET (Dynamic Event Triggering) is proposed. Firstly, the causes and processes of slug flow are described, and the simplified model for control is analyzed. Secondly, a dynamic event triggering mechanism based on the deviation between predicted value and actual value is designed to monitor the running state of the system in real time and adjust the event triggering conditions to improve the steady-state performance of the system and reduce the triggering frequency. Finally, the comparative experiment analysis is carried out. The results show that the proposed method can effectively reduce the triggering frequency of the system on the basis of ensuring the control performance. Compared to the small opening valve, the oil production increases by about 2. 8%.

 In order to solve the problem of constructing the overall impedance model of electric vehicle cascade- type charging module, a cascade-type charging module impedance modeling method based on VIENNA rectifier and full-bridge LLC resonant converter is proposed. Firstly, a typical topology of the charging module is determined, and the small signal model of the front VIENNA rectifier based on the state space method and the rear full-bridge LLC resonant converter based on the equivalent circuit method are constructed respectively. Secondly, the closed-loop output impedance of VIENNA rectifier and the closed-loop output impedance and input impedance of full-bridge LLC resonant converter are obtained by combining the control strategy. The small Signal circuit model of the charging module can be obtained by integrating the front and rear small signal models and control strategies, and then the overall impedance model of the charging module can be derived. According to Nyquist stability criterion, the influence of system parameters on the stability of charging module is analyzed. The charging module simulation system is built based module. The proposed modeling method realizes the overall impedance modeling of single-stage to two-stage charging modules, and provides a theoretical basis for analyzing the parallel stability of charging modules in the future.

In order to solve the problem of poor obstacle avoidance and low efficiency of robots in obstacle environments, a wheeled industrial robot obstacle avoidance control algorithm based on neighborhood rough sets is proposed. Firstly, the environment perception method based on 2D Laser rangefinder is used to analyze the local environment of the robot and the obstacle environment, so as to provide accurate environmental information for subsequent obstacle avoidance control; Secondly, based on the initial obstacle avoidance decision rules of robots, the knowledge reduction effect of neighborhood rough sets is utilized to reduce them to the minimum obstacle avoidance decision rules, and a feasible path set is obtained; Finally, based on the feasible path set, the ND +(Nondimensional) algorithm is used to determine the direction of the robot’s obstacle avoidance decision, thereby achieving obstacle avoidance control for wheeled industrial robots. The experimental results show that this method can achieve high accuracy and control efficiency of machine obstacle avoidance control while ensuring the stability of obstacle avoidance control. 

The recognition rate of gesture recognition is low because of the poor segmentation effect. Therefore, a dynamic multi-point gesture recognition method based on improved support vector machine is proposed. The depth threshold method is used to segment the dynamic multi-point gesture image, extract the largest circular fine hand area in the palm, obtain 7-dimensional HOG(Histogram of Oriented Gradients) feature vector of the hand, complete the gesture action image preprocessing, introduce support vector machine, and improve the algorithm by error term, and adopt the optimized linear classification feature vector of the improved support vector machine. The dynamic multi-point gesture recognition is realized by using the gesture feature vector after input classification by support vector machine. The experimental results show that the recognition rate reaches more than 92. 5% under the condition of illumination, while the recognition rate is still higher than 90. 0% under the condition of no illumination. The proposed method has little fluctuation under the influence of illumination, and the image segmentation information is complete and the recognition accuracy is high.

A modified detection algorithm of electric bicycle helmet based on YOLOv5s(You Only Look Once version 5 small) is proposed to address the issues of small target missed detection and low accuracy in electric bicycle helmet wearing detection. CBAM ( Convolutional Block Attention Module) is introduced into the backbone network enhancing attention to clustered targets and effectively solving the problem of poor detection performance caused by occlusion. The PANet structure in the neck network is changed to a feature fusion structure that combines the idea of cross-scale feature fusion network (BiFPN: Bidirectional Feature Pyramid Network) enhances the multi-scale fusion ability of the model in different directions and effectively fuses multi- scale features of the target. Using SIoU(Structured Intersection over Union)localization loss function instead of CIoU(Complete Intersection over Union)loss function improves the accuracy of bounding box regression. The experimental results show that the accuracy P and recall R of the improved YOLOv5s model are 94. 7% and 91. 2%, respectively, and the average accuracy value mAP is 95. 6%, which is 6%,7%, and 6. 5% higher than that of the original YOLOv5s model, respectively. The method has significantly improved the accuracy of electric bicycle helmet wearing detection.

 In order to improve the accuracy and efficiency of face recognition results of smart access control system, and thus enhance the intelligent service of smart door security, a smart access control face recognition algorithm based on twin neural network is proposed. The wavelet coefficients of the face image signal are obtained by wavelet transform, the appropriate threshold is selected to process the wavelet coefficients, and the inverse transform of the wavelet coefficients is carried out again to obtain the de-noised face image. After the face image is de-noised, the output value of the face image is mapped and processed in the twin neural network to form a feature vector with a dimension of 128. The contrast loss function is introduced to determine the similarity of the face image by comparing the Euclidean distance between the output feature vectors of the sample network, and finally realize intelligent access control face recognition. The experimental results show that the intelligent access control face recognition results and recognition efficiency of the proposed algorithm are significantly better than other algorithms. 

Currently, the lack of standardized support for ontology quality assessment models in the field of data governance is a significant issue. Building a standardized ontology quality assessment model is of utmost importance in addressing this challenge. By studying the dimensions under ISO/ IEC 25012 data quality standards, the GQM (Goal-Question-Metric) methodology is used as a guide to define metrics under the dimensions and realize the mapping from metrics to dimensions. Finally, based on the DQV(Data Quality Vocabulary)data quality model proposed by W3C(World Wide Web Consortium), a scalable and robust ontology quality model is constructed. The proposed quality assessment model provides a complete, unified, and standardized terminology system to describe the various elements of ontology quality, and provides a standardized quality knowledge representation model for ontology quality assessment. Finally, taking the completeness dimension as an example,the corresponding quality assessment model is constructed, and the feasibility of the model is verified by using the downhole operation data set. It effectively solves the problem of the lack of standardization of ontology quality assessment model in data governance field, and provides a unified and standardized term system to describe each element of ontology quality in data governance field. 

A crowd counting model based on background suppression and noise monitoring is proposed to solve the problems of large-scale change of crowd, complex background, and label noise. In the coding stage, the first 13 layers of VGG16_bn are used as the backbone, and the initially extracted features are sent to the two-branch feature extraction module and the background information aggregation module respectively, to mitigate the large- scale changes of the population and improve the discriminability of the background. Finally, the information processed by the two modules is fused, and the predictive density map is generated by decoder regression, which is supervised with the ground truth density map to achieve noise suppression. Compared with other algorithms, the counting accuracy of this model has been improved. MAE(Mean Absolute Error) and MSE(Mean Squared Error) on ShanghaiTech PartA are 58. 1 and 95. 9 respectively. Ablation experiments conducted on ShanghaiTech PartA also verified the effectiveness of the modules. Experimental results show that the algorithm can effectively improve the accuracy of crowd counting. 

Integrated passenger transport hubs usually involve a large number of vehicles and routes, and the traffic flow, passenger demand, and traffic conditions of passenger transport hubs are dynamically changing, which can easily lead to schedule conflicts and make multi route vehicle scheduling difficult. Therefore, a comprehensive passenger transportation hub vehicle multi line scheduling method based on improved ant colony algorithm is proposed. Considering the reduction of operating costs, waiting time, and overall travel time, with the goal of minimizing the operating costs and passenger travel time of the integrated passenger transport hub system, a scheduling optimization model is constructed. Ant colony algorithm is used to the model, introducing search hotspots, optimizing pheromone update strategies and heuristic factors to improve the ant colony algorithm, and the multi line scheduling of comprehensive passenger transportation hub is completed. The experimental results show that the proposed method can more comprehensively carry out multi line scheduling of vehicles, with a waiting rate of less than 5% and an average scheduling time of only 5. 8 s, effectively improving convergence rate, accuracy, and efficiency.

In order to improve the data utilization rate and avoid the influence of risk factors in information on intelligence analysis, a risk feature screening algorithm for intelligent intelligence analysis data based on SVM (Support Vector Machine) is proposed. The continuous wavelet transform method is used to eliminate the influence of noise signals in intelligence data on the analysis results, and the projection matrix is established by combining principal component analysis method to extract the main features of various types of noise-free intelligence data. The main feature extraction results of various kinds of intelligence data are input into support vector machine, and the classification plane in support vector machine is established by using optimization theory, and the classification rules of feature data in the classification plane are defined to screen the risk features of intelligence data. The experimental results show that the proposed method can accurately classify intelligence data, and the risk data detection efficiency is high, which can realize effective screening of risk data.

To protect the valuable field geological resources that have become increasingly vulnerable to natural or human damage in recent years, and to facilitate teaching reform in online virtual geological practice in the post-pandemic era, an automatic field geological section reconstruction technology based on 3D laser scanning is proposed. Using the FARO 3D laser scanner, 3D high-density point cloud data is collected from field geological outcrops of various scales. By processing, optimizing, and visually modeling the high-quality data collected from the field, this technology can protect precious field geological resources and enrich students’ virtual geological practice resources. Taking some field geological outcrops from the Longhuitou Scenic Area in Xingcheng City, Liaoning Province, as examples, virtual geological modeling and visualization applications are carried out, enhancing the quality of geological practice teaching. This approach serves the preservation of field geological legacies and significantly contributes to the advancement of geological education by providing a more interactive and enriched virtual learning environment.

The electricity demand of industry users is usually affected by seasonal and cyclical factors, and sometimes the data obtained is incomplete, missing, or incorrect, which can have a negative impact on the accuracy of predictions. In order to achieve accurate prediction of industry user electricity demand, a data-driven and heterogeneous computing method for predicting industry user electricity demand is proposed. The Lagrange interpolation algorithm is used to fill in the missing part of user electricity data, the standardized preprocessing of electricity data is used to make electricity demand prediction accurate enough, denoising autoencoders and sparse constraint functions are used to extract electricity data features. The long-term memory neural network’s forgetting gate layer, input gate layer, update gate layer, and output gate layer are used to obtain the future trend of electricity demand, the task of industry user electricity demand prediction is completed. The experimental results show that the proposed method is suitable for long-term and short-term industry user electricity prediction, and the prediction results have high accuracy and short time consumption. 

At present, the need for high-quality and semantically rich data has become increasingly urgent as data-driven artificial intelligence is being applied in a wide and in-depth manner in a variety of fields. Data governance works in various fields focus on the governance of data models. However, both internationally and domestically, research on data semantic governance is still insufficient. In particular, there is a lack of systematic exploration of semantics from the underlying basic theory. Therefore, the essence of semantic organization and representation from the basic theory is revealed, and a conceptual system model of conceptual world is put forward. A nature characteristic conceptual semantic registration metamodel and a relational characteristic conceptual semantic registration metamodel is constracted based on the ISO/ IEC(International Organization for Standardization/ International Electrotechnical Commission) 11179 series of standards, achieving the registration and management of rich semantic knowledge. Finally, a metadata registration and governance system is designed and developed in the context of data governance in the field of oil and gas exploration and evaluation. The two types of semantic models based on the MDR(Metadata Registry) standard have been verified, reflecting their effectiveness in practical applications. 

Texture feature extraction is the key to texture classification, and there are various factors such as rotation, illumination, and scale variations in texture images. To enhance the robustness of the texture feature extraction algorithm for rotation, illumination, and scale variations, the FWLMP ( Fractal Weighted Local Morphological Pattern) is proposed. First, a scale-invariant descriptor is constructed by using the relative invariance of fractal dimensionality to scale variation. Then, it is sampled and analyzed using the expansion, erosion, and opening-closing operations in mathematical morphology, and its weights are calculated by using the fractal dimension image. This algorithm is scale-invariant and robust to rotation and illumination changes. To achieve the classification of Qing Dynasty costume images, the Qing Dynasty Buzi image dataset is constructed. The FWLMP and similar algorithms are tested on four public texture datasets and a private dataset constructed by ourselves. The experimental results show that the FWLMP algorithm performs well in texture image classification and in Buzi image classification for Qing Dynasty civil and military officials. 

 In order to solve the problems of low brightness, serious noise, and color distortion of color polarization images in low-illumination scenes, an unsupervised learning algorithm for color enhancement of low- illumination color polarization images is proposed, which is named LPEGAN(Low-Light Polarization Enhance Generative Adversarial Network). Firstly, a double-branch feature extraction module is designed and used different branches to extract features from Stokes parameters S0 and S1 ,S2 , respectively. Secondly, the residual void convolution module is constructed. And the different expansion rates can expand the receptive field to improve the model extraction ability and reduce the image color distortion. The edge texture loss function is constructed to ensure the structural similarity between the enhanced image and the input image. Experimental verification is carried out on the public datasets LLCP(Low-Light Chromatic Intensity-Polarization Imaging), IPLNet(Intensity-Polarization Imaging in Low Light Network), and self-built datasets. The experimental results show that the proposed algorithm has better visual effects, and all evaluation indicators are significantly improved. Polarized image brightness is enhanced, noise is significantly suppressed, and image colors are more realistic and natural.

 The low efficiency of video surveillance in identifying emergencies results in that the recognition system is unable to detect and respond to emergencies in a timely manner, increasing the risk of potential hazards. Therefore, a recognition algorithms of monitoring image abnormal behavior based on computer vision is proposed. Based on the initial background of the monitoring image, a differential operation is used to obtain the differential image between the background image and the monitoring image, and the background subtraction method is used to perform binary processing on the combined sorted new monitoring image to complete target area recognition. Then, a rectangle is used to traverse the target area, collect effective motion blocks from the target area, extract the feature vectors of the motion blocks, and complete the extraction of abnormal behavior features in the monitoring image. And the identification of abnormal behavior in monitoring images through Kuhntak conditions is completed. The experimental results show that the proposed method has an abnormal behavior recognition time of less than 1. 0 s, and the recognition accuracy remains above 94%. It can accurately identify abnormal behavior in monitoring images, effectively improving recognition efficiency and recognition rate.

 In order to solve the problems of low efficiency of manual laboratory management, slow transmission speed of monitoring data and insufficient level of informatized management, an intelligent laboratory management and control system based on all optical access network is designed and implemented. It consists of laboratory front-end monitoring module, all optical access network data transmission module and laboratory management and control center module. The ESP 8266 microcontroller is used as main control chip in the laboratory front-end monitoring module. The acquisition unit of laboratory operation status data is designed. It consists of fingerprint recognition module, temperature and humidity sensor, smoke sensor and network camera. Real-time data acquisition of personnel management and control, video surveillance and environmental monitoring is thus realized. Optical network units, optical fiber distribution network and optical line terminal are utilized to construct all optical access network data transmission module, realizing remote high-speed transmission of each laboratory monitoring data. The web page of laboratory management and control platform is developed, real-time operation status data of each laboratory is displayed in terminal management and control server. The joint debugging result shows that real-time personnel management and control, video surveillance and environmental monitoring are realized, and the laboratory management and control system provides reliable stability in a long time. 

Geomagnetic navigation utilizes geomagnetic field information for positioning and navigation. However, due to the slow changes in the geomagnetic field and small variations in the total geomagnetic field gradient, the available information is limited and lacks identification accuracy during geomagnetic matching, thereby restricting improvements in positioning accuracy. To enhance the identification accuracy of the geomagnetic field and improve positioning precision, a novel approach employing a combined artificial magnetic source is proposed. A cylindrical permanent magnet composed of rare-earth neodymium iron boron material is chosen as the magnetic source. The spatial distribution of the magnetic field is analyzed through modeling and simulation using COMSOL software to determine an optimal design scheme for the magnetic beacon. Evaluation metrics such as standard deviation, roughness, and information entropy are employed to assess enhancement in characteristics of the geomagnetic map resulting from this magnetic beacon scheme, ultimately leading to a more reasonable design that improves positioning accuracy. Experimental results demonstrate that within a specific test area, centimeter-level positioning accuracy can be achieved with this combined magnetic beacon scheme.