2379results about How to "Reduce distractions" patented technology

The invention provides a bridge girder erection machine intelligent construction control method and system based on machine vision, and relates to the technical field of intelligent control. Image information of a construction site is collected through a plurality of high-definition cameras arranged on a bridge girder erection machine; inputting the image information into a pre-trained deep learning model, identifying and positioning a target in the construction scene, and constructing a digital twinborn model of the construction scene in combination with a graph structure of a construction scene topological structure subjected to spatial perception graph convolution operation; the intelligent decision agent generates an initial construction strategy according to the current construction scene state, an improved Monte Carlo tree is adopted to search and optimize the initial construction strategy, and an optimal construction strategy is generated through multi-objective solution; and after the construction is completed, based on the digital twin model and the whole process data, analyzing the construction quality and efficiency by using a graph neural network, and generating a multi-dimensional evaluation report and optimization suggestions.

The invention belongs to the technical field of biodegradable materials, and particularly relates to an environment-friendly degradable modified composite material, a preparation method and application. The material comprises PBAT, PBS, PLA, an inorganic mineral filler, natural plant fibers, PBAT-PTMO, nano-silica, an auxiliary agent and a plasticizer. In the preparation process, PLA is prevented from being hydrolyzed in vacuum drying, calcium carbonate and nano silicon dioxide are preheated and activated, natural plant fibers are dispersed through steam explosion and ethanol treatment, and raw material pretreatment comprises high-speed mixing and agglomeration breaking and medium-speed secondary premixing so as to realize uniform distribution. The co-rotating twin-screw extruder accurately controls the temperature, the speed, the rotating speed and the length-diameter ratio, and ensures melting plasticizing, uniform mixing and stable extrusion molding. According to the invention, by adopting accurate proportioning of multiple raw materials and a modification treatment process, the properties of mechanics, degradation, processing and the like of the material are synergistically improved, and the material is widely applied to multiple fields such as packaging bags, disposable tableware and agricultural plastic mulching films.

The invention discloses a collaborative operation system and method for a photovoltaic cleaning robot and a carrying robot, relates to the technical field of automatic control, and solves the defects that an existing collaborative operation system is unstable in communication and insufficient in emergency handling capacity. According to the invention, the environmental information of the photovoltaic power station is collected in real time through the environmental perception module, and the reliability of data transmission is ensured through an automatic retransmission request mechanism and error detection and correction codes based on 5G and LoRa communication protocols; communication parameters are adaptively adjusted through the communication self-adjusting module, so that the communication quality is ensured; the path planning module dynamically allocates cleaning tasks and plans an optimal path, the collaborative obstacle avoidance module dynamically adjusts the motion trails of the robots and avoids obstacles, the operation progress tracking module records the operation states of the robots in real time, and comprehensive operation progress information is provided through data processing and visual display. According to the invention, the communication stability, the autonomous decision-making capability and the operation efficiency of the system are obviously improved.

The invention relates to the technical field of stone structure ultrasonic detection, in particular to a stone structure building detection method based on a concrete ultrasonic fault detection scanner, and the method comprises the steps: obtaining the initial signal intensity when an ultrasonic signal is transmitted by using the ultrasonic fault detection scanner; obtaining an ultrasonic reflection signal after the ultrasonic signal is reflected at each reflection point under each scanning layer, the intensity of the reflection signal after reflection, and the signal propagation duration; the method comprises the following steps of: obtaining a signal propagation time difference, a signal intensity difference, a signal difference index, a signal disorder degree, a defect evaluation coefficient, a seam recognition degree and a seam correction coefficient of each reflection point under each scanning layer, correcting the intensity of a reflection signal, performing three-dimensional reconstruction on a constructed tomography graph, and detecting internal defects of a stone structure building. According to the method, the interference of the reflection points on the splicing seams during stone splicing is reduced, the accuracy of detecting the internal defects of the stone structure building is improved, and the stability detection of the stone structure building is improved.

The invention relates to the technical field of automobile data updating, in particular to a data updating method of automobile diagnosis equipment and the automobile diagnosis equipment. The method comprises the following steps: acquiring a historical running log of a vehicle; vehicle owner driving behavior multi-point sampling is carried out on the vehicle historical operation log, and different scene behavior characteristic evolution is carried out, so that different scene driving behavior models of a vehicle owner are obtained; extracting vehicle time sequence state parameters based on the vehicle historical operation log; performing multi-scene operation simulation on the vehicle time sequence state parameters according to different scene driving behavior models of the vehicle owner, and generating vehicle state operation simulation data of each scene; and carrying out future time period vehicle state rolling prediction and fault time window probability calculation on the vehicle state operation simulation data of each scene to obtain a vehicle abnormal fault part positioning point and a fault time window probability curve. According to the invention, the data updating efficiency and group accuracy of the automobile diagnosis equipment are realized.

The invention relates to a big data processing technology, in particular to an intelligent data labeling method and system. The intelligent data labeling method comprises the following steps: preprocessing public opinion data; semantic deduplication of the historical sample library is carried out, and the sample library is updated; an intelligent auxiliary labeling method based on a historical work order library and a large model; model updating and sample increment updating; compared with the prior art, the method has the advantages that a multi-dimensional data cleaning method is adopted, the quality of the to-be-labeled data is improved, and interference of repeated semantic information is reduced through semantic duplicate removal; potential values of similar work orders in historical work orders are mined, a large model cue word project used for auxiliary labeling is constructed, an available auxiliary labeling cue word template is provided, and the cue word construction and auxiliary labeling process is opened; a PDCA thought is adopted, and a data-annotation-model-work order closed-loop management and PDCA mechanism spiral data quality method is constructed.

The invention discloses a private domain user management method and system for multiple touch channels, a medium and a program product, and the method comprises the steps: obtaining the historical behavior data of a user in each touch channel; identifying the behavior relevance of the user among the arriving channels; constructing an inter-channel behavior mapping relation of migration paths of the user among the touch channels based on behavior relevance; according to the behavior mapping relation between the channels, the overlapping ratio of user behaviors between the reaching channels, the difference of conversion effects and the overlapping ratio of time sequences are calculated; based on the overlapping rate, the difference and the coincidence degree, determining the reaching channel complementation degree; grouping and clustering the reaching channels according to the reaching channel complementation degree to obtain a plurality of channel collaboration units; and the optimal time difference of the arriving users in each channel cooperation unit is calculated, and the marketing resources are allocated among the arriving channels based on the optimal time difference, so that the accuracy of private domain operation management is improved.

The invention provides an electric energy meter correction electric quantity calculation method and device, electronic equipment and a storage medium. The method comprises the steps of obtaining metering historical data of a target electric energy meter; analyzing the metering historical data by adopting a phase angle feature portrait algorithm, and identifying a wiring error type corresponding to the target electric energy meter; determining a corresponding correction transformation matrix based on the wiring error type; and calculating the corrected electric quantity of the target electric energy meter based on the corrected transformation matrix. The invention provides a novel calculation scheme for the corrected electric quantity of the electric energy meter, the corrected electric quantity of the electric energy meter is calculated more accurately by adopting an advanced algorithm and a mathematical model, the interference of human factors is reduced, various complex wiring error conditions can be adapted, and the calculation efficiency is improved. Therefore, the accuracy and reliability of the correction electric quantity calculation method of the electric energy meter of the power system are improved.

The invention discloses a method for correcting a laser gas concentration sensor, which is used for measuring gas concentration by using laser with different wavelengths and correcting by combining environmental parameters. The method comprises the following steps: extracting laser absorption characteristics by using a first feedforward neural network, extracting environmental matrix characteristics by using a convolutional neural network, fusing the characteristics through an attention mechanism, and predicting the gas concentration through a second feedforward neural network. The loss function not only considers the difference between the predicted value and the true value, but also comprises the physical constraint between the light intensity value and the theoretical light intensity value. According to the method, the measurement accuracy and the model robustness under different environmental conditions are improved, and the influence of measurement errors and environmental fluctuation is reduced through the combination of multi-wavelength measurement and physical knowledge.

The invention relates to the field of forestry pest prevention and control, and discloses a juglans sigillata forest intelligent pest prevention and control system and method.The juglans sigillata forest intelligent pest prevention and control system comprises a data acquisition module used for collecting basic information of a juglans sigillata forest, and the basic information comprises meteorological environment parameters, soil physical and chemical properties and in-forest visual image data; the identification and analysis module is used for deeply analyzing data characteristics of the basic information, discriminating types and severity of diseases and pests, predicting potential occurrence risks of the diseases and pests and forming an evaluation result; and the early warning response module is used for presetting a risk level threshold value and rule logic and generating an early warning signal of a corresponding level according to an evaluation result. A data acquisition module integrating multiple types of monitoring units and communication units is constructed, so that the juglans sigillata forest can be sensed in all directions, the limitation of manual inspection is eliminated, and the problems of inaccurate monitoring and difficulty in data backtracking are solved; and the identification analysis module comprises sample management, model training and data analysis and prediction.

The invention relates to the technical field of equipment control, in particular to a self-adaptive switching method and system of a current transformer. The method comprises the following steps: collecting primary side parameters and secondary side parameters of a current transformer to obtain primary side collection parameters of the current transformer and secondary side collection parameters of the current transformer; performing current transformer state segmentation initialization according to the current transformer primary side acquisition parameters and the current transformer secondary side acquisition parameters so as to generate a current transformer operation state classification result; switching result judgment is conducted on the current transformers according to the current transformer operation state classification result, and when the current transformer operation state classification result is a current transformer critical state section, parallel buffering is conducted on the current transformers, and critical current transformer switching data are generated. According to the invention, through accurate state classification, a dynamic switching mechanism, time sequence optimization and visual display, the switching sensitivity and stability of the current transformer are improved.

The invention discloses a reinforcement learning automatic driving method and system based on man-machine cooperation, and belongs to the technical field of automatic driving. The method includes: acquiring a current state of the autonomous vehicle; inputting the current state of the automatic driving vehicle into the trained automatic driving network for processing so as to obtain a decision-making action of the automatic driving vehicle at the current moment; a reward-free strategy learning method is introduced, and an automatic driving network is trained through an explicit intervention mechanism and an implicit intervention mechanism by taking human takeover cost minimization and action distribution entropy and stability maximization as targets. An explicit and implicit intervention value function is designed, and an intelligent agent is trained to drive in an invisible traffic scene with minimum human takeover cost and comfort maximization as targets; the strategy optimization problem of an automatic driving automobile in a complex dynamic traffic environment is solved.

The invention discloses a single tank transportation risk early warning method and system based on an intelligent sensor, and relates to the technical field of risk early warning, and the method comprises the steps: collecting data in a single tank for preprocessing, employing a two-dimensional shallow water wave model to calculate liquid fluctuation, and generating iteration sequence data; applying a Hanning window function, calculating spectrum energy distribution according to the spectrum amplitude, extracting primary frequency and secondary frequency data, and calculating the amplitude ratio of the primary frequency to the secondary frequency; according to the method, through coupling of the continuity equation and the momentum equation, the height change and the velocity field distribution of the liquid can be comprehensively described, the dynamic behavior of the liquid in the single tank can be accurately reflected, the dynamic change of the liquid height gradient and the velocity field gradient is considered, and the complex liquid fluctuation phenomenon can be captured; a window function is applied to perform weighting processing on the data to weaken the weight of the data at two ends of the time sequence, so that the middle part of the data contributes more remarkably to the spectrum.

The invention discloses a frequency source synthesis system and method based on a frequency agility radar, and belongs to the technical field of microwave radio frequency, and the method specifically comprises the steps: generating a target clock signal, carrying out the frequency processing of the target clock signal to generate a frequency multiplication signal, and dividing the frequency multiplication signal into two clock signals through a power divider; performing frequency multiplication processing on one path of clock signal, taking the processed signal as a clock reference signal of a direct digital synthesizer, generating an analog signal according to the clock reference signal, and calibrating the analog signal as an output signal; performing frequency multiplication processing on the other path of clock signal, and taking the processed clock signal as a local oscillation signal of the frequency mixer; performing frequency conversion processing to obtain a frequency mixing signal, performing quadruplicated frequency processing on the frequency mixing signal, and calibrating the processed frequency mixing signal as a final output signal; according to the invention, the frequency source has the characteristics of low phase noise, rapid agility, fine resolution and miniaturization at the same time.

The invention discloses an intelligent compilation system and method of a building construction scheme, and the system comprises a data preprocessing module which is used for obtaining historical construction scheme text data of each building engineering project category; the knowledge base construction module is used for constructing a construction scheme knowledge base of each construction engineering project category by establishing a vector index of a construction scheme vector database of each construction engineering project category; the retrieval information extraction module is used for performing keyword extraction on the construction scheme retrieval statement of the construction engineering project input by the user, and mapping and recalling the construction scheme text data of a plurality of corresponding construction engineering projects; and the construction scheme compiling module is used for obtaining a construction scheme compiling basis of the construction engineering project. Associated storage and value utilization of previous engineering construction schemes are achieved, the scheme compiling efficiency and reliability are improved through intelligent generation of the scheme compiling basis, and effective guarantee is provided for the scheme compiling quality.

The invention discloses a chamfer defect detection method, device and equipment and a storage medium, and relates to the technical field of defect detection.The method comprises the steps that an obtained to-be-detected device image is preprocessed, and chamfer contour points and chamfer contour point pixel coordinates are determined in the preprocessed image; correcting the pixel coordinates of the chamfer contour points, and fitting a chamfer edge straight line by using the corrected pixel coordinates of the chamfer contour points; according to the chamfer edge straight line, screening the corrected chamfer contour point pixel coordinates, and determining a target chamfer end point; and calculating the actual chamfer length based on the target chamfer endpoint, and comparing the actual chamfer length with the preset chamfer length to judge whether the chamfer has defects or not. According to the method, the detection requirements of chamfers of different shapes and sizes can be met by preprocessing and correcting the images, fatigue and subjective judgment errors possibly existing in traditional manual detection are avoided through automatic image processing and data analysis processes, and the detection efficiency is improved.

The invention discloses a cell image segmentation method, and relates to the technical field of cell analysis, and the method comprises the following steps: S1, image acquisition; s2, processing optical artifacts generated by imaging of the phase contrast microscope; s3, processing interference of impurities and debris in the cell image; in the method, a phase contrast microscope and an optical coherence tomography microscope are combined to obtain a cell image, three-dimensional structure information of an OCT image is utilized to assist in correcting artifacts generated by imaging of the phase contrast microscope, and the artifacts are processed based on phase recovery and filtering and texture feature matching and repairing. According to the method, the interference of impurities and debris is removed by applying a diffusion and repair method based on a partial differential equation and a machine learning classification and reconstruction method, so that the cell image quality is effectively improved, the interference of artifacts and impurities on subsequent analysis is reduced, and a reliable image data basis is provided for cell image segmentation and cell morphology analysis.

The invention discloses an electromagnetic transient simulation and mapping method and system oriented to multi-physics field coupling, and belongs to the technical field of substation digital operation and maintenance and electromagnetic simulation. Deploying multiple types of sensors in a transformer substation, preprocessing the acquired multi-physical field data, and fusing with the three-dimensional twinborn model; a coupling algorithm is adopted to carry out multi-physical field coupling analysis, multi-physical quantity display and thermodynamic diagram superposition are realized, and scene evolution change simulation under various typical working conditions is supported; a micro-service architecture is adopted to construct a digital twin system, double closed-loop verification of simulation and actual measurement is performed through actual measurement data and simulation data of a real-model test field, model optimization is completed, and real-time mapping of electromagnetic transient dynamic simulation and a transformer substation physical entity is realized. The method improves the precision of model and entity mapping, meets the requirements of real-time tracking and response of the electromagnetic transient change of the transformer substation, adapts to the digital transformation of different transformer substations, and provides a basis for the operation and maintenance of the transformer substations.

The invention discloses a train communication network intrusion detection method based on a multi-scale residual network, and the method comprises the following steps: building a train communication network topological structure model, and building a traffic model for the mixed transmission of normal traffic and intrusion traffic; collecting a flow data packet in a link as an original data sample in the intrusion detection data set, and performing preprocessing to obtain a data sample containing statistical characteristics; a train communication network intrusion detection MSRNet-GLAM model is constructed, and the performance of the train communication network intrusion detection MSRNet-GLAM model is checked and evaluated, so that the MSRNet-GLAM model can learn key feature information of a data sample in a multi-scale manner; and training and optimizing the MSRNet-GLAM model by using the improved focus loss function, and performing intrusion detection on the traffic data of the train communication network to be detected to obtain a traffic detection result of the train communication network. The method can effectively solve the problems of low security and difficulty in detection of intrusion traffic under the trend of mass data communication of the train communication network, and ensures the security of the train network.

The invention discloses a method for predicting the aging degree of an anti-corona layer of a stator bar, and relates to the field of prediction of the aging degree of the anti-corona layer of the stator bar, and the method comprises the following steps: S1, data sampling; image data and real-time temperature information are obtained; s2, data processing; preprocessing the acquired image data to obtain a data set, and performing evaluation weight division and state division on the data set; s3, constructing a model; constructing a model by using a convolutional neural network; s4, training the model; s5, testing the model; s6, practical application; and predicting the aging degree of the anti-corona layer of the stator bar by using the anti-corona layer life prediction model. According to the method, the machine learning algorithm based on learning is adopted, the infrared thermal image, the ultraviolet sub-image and the temperature data in the aging process of the anti-corona layer of the stator bar are collected, the aging state of the stator bar is classified by applying multi-modal feature fusion, and the method has a good guiding effect on evaluation of the aging degree of the anti-corona layer of the stator bar.

The present invention proposes a communication-sensitive multi-agent collaboration method, which is applied to the field of vehicle network communication technology. The method of the present invention includes: the agent uses an information filtering module to encode the locally observed message into a hidden state vector, that is, a compressed filtered message, and before sending the message, the shared medium access control module is used to score the message, and the edge computing node sorts the message value according to the score, and sends a confirmation message to the agent with the topK message value. After the agent receives the confirmation, it sends the message; the edge computing node receives all messages, uses the message distillation module to extract valid information, and summarizes and distributes it to each agent; in the process of summarizing the agent messages, the graph convolution aggregation method based on the graph information bottleneck is used to retain the agent topology structure information, and the summarized message is distilled. The present invention reduces the transmission of redundant and invalid information between agents, saves communication resources, and improves the decision-making ability and collaboration efficiency of multiple agents.

The invention relates to the technical field of deep foundation pit monitoring, and discloses a deep foundation pit micro-deformation InSAR remote sensing method and system and a storage medium. The method comprises the following steps: synchronously acquiring data through a ground radar and an InSAR satellite, and obtaining a point target set and an original scattering characteristic sequence; obtaining a net deformation phase sequence without influence through registration residual correction, phase deduction and weighted fusion processing, and carrying out phase consistency quality control and unwrapping processing to obtain an accumulated deformation quantity sequence; constructing a deformation mechanism correlation characteristic matrix based on correlation analysis of the cumulative deformation quantity sequence and radar reflection intensity amplitude attenuation characteristics; and fusing the deformation mechanism correlation feature matrix and the real-time differential interference image group, extracting deformation rate and phase gradient change, and further generating a deep foundation pit local micro-deformation risk early warning index set. According to the invention, the tiny deformation of the surrounding area of the foundation pit can be effectively detected, potential risks are warned in advance, and the safety and monitoring efficiency in the construction process are improved.

The invention relates to the technical field of hollow part production, and discloses a metal-plastic composite car lamp frame shell integrated injection molding device which comprises an injection molding machine, a movable mold and a fixed mold are installed in the injection molding machine, an external manipulator and an external cooling module are installed outside the injection molding machine, and two fixing holes are symmetrically formed in the outer surface of the movable mold; an injection molding assembly is arranged in the injection molding machine, the injection molding assembly comprises not less than two metal pipes and not less than four extension pipes, and in the cooling process, the metal pipes are cooled from inside to outside based on the physical characteristics of the metal pipes; by means of the cooling mode, the middle of the molding material is cooled and shrunk firstly, the situation that the outer layer is solidified firstly due to the fact that cooling is conducted from the outside to the inside of the cavity traditionally is changed, when the internal material is cooled and shrunk, limitation of the solidified outer layer material is avoided, and therefore the problem that cavities are generated due to inconsistent shrinkage is effectively prevented.

The invention relates to the technical field of hydrological survey, in particular to an intelligent survey device for hydrological water resources, which comprises a survey main body, a power assembly arranged in an inner cavity of the survey main body, a sampling assembly arranged in the inner cavity of the survey main body, a buffer groove formed in the outer surface of the survey main body, and an anti-collision assembly arranged on the outer surface of the survey main body. Water resources of different depths can be collected through the sampling assembly, the real-time performance of samples is improved, meanwhile, impurities in the water source can be prevented from entering the samples to interfere sample evaluation during sampling, guiding can be conducted through the anti-collision assembly when the exploration device collides with the water, and the detection accuracy is improved. According to the technical scheme, damage generated when the surveying device collides is reduced, meanwhile, when vertical collision occurs and guiding is not conducted, the surveying device can give an alarm in time, the direction which the surveying device is blocked can be observed, the surveying device can be conveniently controlled to change the direction, and the situation that surveying work is delayed due to the fact that the surveying device is damaged is avoided.

The invention discloses an outdoor optical cable that is easy to connect quickly, and belongs to the technical field of outdoor optical cables. The invention discloses an outdoor optical cable that is easy to connect quickly. The outdoor optical cable that is easy to connect quickly drives an extrusion rod into a circular groove of a first optical cable through a second optical cable, and the extrusion rod contacts a circular plate and slides inwards. At the same time, the extrusion rod drives a limiting rod to slide along the L-shaped limiting groove and slides to the end of the L-shaped limiting groove. Subsequently, the second connecting head is rotated to drive the extrusion rod and the limiting rod to rotate slightly, and then the second connecting head is released, so that the circular plate can be pushed in the opposite direction under the action of a first spring to contact the extrusion plate and slide to the outside of the circular groove, so that the limiting rod slides to another position of the L-shaped limiting groove, and the limiting rod drives the extrusion rod to limit, thereby fixing the first connecting head and the second connecting head, thereby achieving the effect of quickly connecting the optical cable, which is helpful to improve the working efficiency of connecting between optical cables and reduce the labor intensity of workers.

The invention discloses a ground resistance comprehensive test system and method, and belongs to the field of ground test, and the ground resistance comprehensive test method comprises the steps: determining a test monitoring area of ground resistance, drawing a power grid distribution three-dimensional model on the test monitoring area, and marking each ground point in the test monitoring area; detecting a grounding resistance value on the grounding point, and evaluating whether the grounding resistance value of the grounding end on the grounding point is abnormal or not; and for the grounding point with the abnormal grounding resistance value, calculating the equivalent grounding resistance value of the grounding point according to the soil parameters on the grounding point, and evaluating whether the grounding system on the grounding point is damaged or not. The test system comprises a grounding resistance tester, a wireless data module, a soil parameter acquisition module and a far-end monitoring center. According to the invention, the grounding resistance of the grounding end is tested by using the multi-frequency alternating-current electric signal, so that the test requirements of different grounding systems under various working conditions can be met, and the universality and the practicability are ensured.

The invention belongs to the technical field of federated learning, and more specifically relates to a privacy protection personalized federated learning method based on model hierarchical optimization. The method comprises the following steps: each client obtains a global model of a server side, updates a local model by using a training data set of a self-side, then performs hierarchical optimization on the local model, and determines partial model parameters to be uploaded of each client; partial model parameters to be uploaded by each client are cut, differential privacy noise is introduced, and then the model parameters are sent to a server side; the server side allocates a weight to each client according to the size of the data volume of each client, then aggregates the model parameters uploaded by each client to obtain a new global model, and sends the new global model to each client; and repeating the steps S1 to S3 until a set training round or convergence is achieved. According to the method, the problems of model performance influence and communication efficiency caused by heterogeneity in federal learning are solved while user privacy is guaranteed.

The invention discloses a preparation method and application of a sensor array based on luminous carbon dots, and relates to the technical field of sensors, the preparation method comprises the following steps: preparing a carbon source which comprises a precursor containing a targeting functional group or a biomolecule derived carbon source, preparing the luminous carbon dots through a chemical synthesis or physical method, modifying the surfaces of the carbon dots, and preparing the sensor array based on the luminous carbon dots. A core-shell structure is added on the surface of a carbon dot by using an in-situ growth method or a layer-by-layer self-assembly method, the carbon dot with the added core-shell structure is fixed on a substrate of a sensor array, and a performance test is carried out. By synthesizing the carbon dots with a targeted recognition function, the selectivity and the sensitivity of the sensor to a target substance can be remarkably improved, and in the synthesis process, the carbon dots can generate specific interaction with the target substance by adjusting factors such as the structure, the shape, the size and the surface chemical property of the carbon dots; therefore, accurate identification of the target substance is realized, and detection of the marker is realized.

The invention discloses a mine safety risk early warning method and system, and relates to the technical field of mine safety early warning, and the method comprises the following steps: in the operation process of a fresh air system, obtaining various parameter information generated in the operation process of the fresh air system, and the parameter information comprises air flow characteristic information and gas diffusion characteristic information; according to the method, the airflow turbulence index and the gas diffusion delay index generated by the fresh air system are intelligently analyzed, the system automatically generates an operation state coefficient based on a machine learning model, high and low risk states are accurately divided, corresponding countermeasures are taken for burst type, unstable type and continuous type high risk states, and early warning or records are given out. According to the grading early warning mechanism, the risk identification precision is effectively improved, man-made misjudgment is reduced, it is ensured that serious risks are handled in time, and the safety and reliability of the system are improved.

The invention relates to the technical field of optical fiber manufacturing, and discloses a heavy water vapor replacement device, a spectral measurement system and a method for reducing water peak absorption loss of a hollow-core optical fiber. The method for reducing the water peak absorption loss of the hollow-core optical fiber is realized by using a heavy water vapor replacement device, the effectiveness is verified through a spectral measurement system, the heavy water vapor replacement device comprises a vapor generation assembly, a first gas cavity assembly and a gas extraction assembly, the gas extraction assembly firstly vacuumizes the interior of a fiber core, and the vapor generation assembly generates heavy water vapor; heavy water vapor is filled in a fiber core of the hollow-core optical fiber through the first gas cavity assembly, the heavy water vapor is prevented from being mixed with gas reserved in the fiber core, it is guaranteed that the fiber core of the hollow-core optical fiber is completely filled with the heavy water vapor, an absorption peak is weakened, especially interference to S-band communication is greatly reduced, hydroxyl formed by heavy water molecules has stable chemical properties, and the service life of the hollow-core optical fiber is prolonged. Secondary adsorption of common water vapor is effectively delayed, and the transmission performance of the optical fiber is improved.