163results about How to "Avoid false positives" patented technology

The invention discloses a loss diagnosis and health management system and method for a transformer, and belongs to the technical field of intelligent transformers, and the system comprises a thermal fingerprint module which is used for recording a temperature drop curve of each part of the transformer after planned power failure, and obtaining a thermal inertia fingerprint of the transformer based on the fitting of multiple temperature drop curves; the theoretical temperature module is used for calculating the theoretical cold body temperature of each part of the transformer based on the real-time environment temperature and the thermal inertia fingerprint when the transformer runs; the net temperature rise module is used for acquiring the actual temperature value of each part of the transformer and subtracting the corresponding theoretical cold body temperature from the actual temperature value; weak thermal characteristic signals caused by insulation aging and loss abnormity of the transformer can be accurately extracted from strong background noise such as load fluctuation and environment temperature through technologies such as thermal inertia fingerprint construction and net temperature rise separation, early diagnosis of progressive health deterioration is realized, support is provided for intervening potential faults in advance, and the fault diagnosis accuracy is improved. And the passive situation of post-treatment of operation and maintenance is changed.

The invention discloses a wear online identification and compensation control system and method for a high-pressure distribution plunger and barrel assembly, and the system comprises a data collection module which is used for collecting working condition data and state data; a health state dynamic baseline model obtained through offline training is arranged in the data processing module, and the wear level of the plunger and barrel assembly is recognized by calculating the residual error between actual state data and health state data predicted by the model under the current working condition; and the compensation control module is used for generating a self-adaptive compensation control instruction according to the identified wear grade and the current working condition and sending the self-adaptive compensation control instruction to an execution mechanism. According to the method, the dynamic base line synchronous with the real-time working condition is constructed, early weak wear characteristics can be separated from strong working condition interference, the recognition sensitivity and accuracy are improved, the influence of wear on the system performance is actively counteracted through closed-loop adaptive compensation control, the service life of equipment is expected to be prolonged, and the system performance is improved. And technical support is provided for predictive wear monitoring management.

This invention provides a method for detecting sensor anomalies in nuclear power plants. The method includes: constructing a principal component analysis model, calculating the Q-statistic, and determining a first threshold; training a Bayesian long short-term memory neural network model and calibrating its prediction interval; calculating a second and third threshold based on the predicted value and uncertainty output by the calibrated model; inputting test data into the two models to obtain the Q-statistic, predicted value, and uncertainty, and comparing them with the corresponding thresholds to obtain three judgment results; and making a final judgment on the three results using a logical judgment table. This invention integrates spatiotemporal characteristics and uncertainty assessment, achieving high-precision and high-reliability sensor anomaly detection with only normal operating data, exhibiting high reliability and robustness.

The application discloses a Chinese pear psylla MIRA-CRISPR / Cas12a detection kit and method, and belongs to the technical field of rapid detection of pest resistance. In view of the target resistance mutation (GluCl gene V267I site) of Chinese pear psylla to abamectin, the application first combines the CRISPR / Cas12a system with the MIRA isothermal amplification technology, and establishes a visual detection system with high specificity and high sensitivity. The kit comprises MIRA amplification primers, specific crRNA, Cas12a protease and a fluorescent reporter probe. During detection, only a single head of pear psylla sample is needed, after rapid DNA extraction, MIRA amplification and CRISPR cutting reaction are carried out under the condition of constant temperature of 37 DEG C, the total time consumption is not more than 1 hour, and the result is directly visually judged by a portable blue light lamp (green fluorescence is displayed for a positive sample). The application is simple in operation, does not need large instruments, and the result is intuitive, so that real-time monitoring of the abamectin resistance of Chinese pear psylla in the field can be realized, and technical support is provided for scientific rotation of drugs and precise prevention and control.

The application discloses a high-pressure gas particle counter with a laser diode as a light source, which comprises an optical detection unit, a gas path system and a standard particle reference system, wherein the optical detection unit is used for generating laser and dividing the laser into two independent light paths, i.e., a main light path and an auxiliary light path, and is used for irradiating a main detection area and an auxiliary detection area respectively and receiving a particle scattering light signal and converting the particle scattering light signal into an electric signal; the gas path system is used for selectively conveying a high-pressure gas to be measured and a standard particle gas to the main detection area and the auxiliary detection area; the application belongs to the technical field of gas particulate pollution monitoring, and aims to solve the problem that the output light power of a laser diode in the prior art will drift with the increase of working time and the change of temperature, thereby causing the measurement reference to be inaccurate, and achieves the technical effect that a real-time online dynamic calibration mechanism is constructed by introducing an auxiliary reference light path and the standard particle reference system, and the problem of the measurement reference drift caused by factors such as light source aging and temperature drift is effectively overcome.

The utility model provides an AGV forklift goods detection stabilizing device, and belongs to the technical field of automatic guided vehicles. The device comprises a fixing plate, a main goods detecting part, a central goods detecting part, an auxiliary goods detecting part and a control unit. The main goods detecting part comprises two goods detecting assemblies, each goods detecting assembly is provided with a fixing frame, a goods detecting baffle, an elastic limiting piece and a first proximity switch, and when the included angle between the goods detecting baffle and the fixing frame is smaller than or equal to 5 degrees, a first detecting signal is output; the central cargo detection part comprises an inductive sensor, detects the cargo insertion state and outputs a second detection signal; the auxiliary goods detecting component comprises a second proximity switch, and when goods enter the sensing range, a third detection signal is output. The control unit receives the three paths of signals at the same time, and outputs a signal of detecting goods in place only after the three paths of signals are all continuously conducted for at least preset time. According to the utility model, the cargo insertion stability is ensured through multiple detection mechanisms, and the safety and reliability of the operation of the AGV forklift are improved.

The utility model discloses a kind of anti-falling alarm sickbed, comprising: monitoring fence is at least set in the both sides of bed body;Monitoring fence includes being set on the baffle several first grid bars and being set on the substrate several second grid bars, several first grid bars and several second grid bars one-to-one, elastically sliding connection;Second grid bar is internally provided with distance sensor;Processor is connected with several distance sensor circuit;Alarm and processor wireless signal connection.The utility model sets monitoring fence in the periphery of bed body, and trigger is formed by the first grid bar, second grid bar and the distance sensor located in second grid bar of elastically sliding connection, when patient exists the behavior of climbing fence, patient's own weight triggers and sends to alarm, can form reminder to accompanying personnel, the utility model only forms alarm when monitoring fence is pressed, avoid the false alarm situation that exists when normal use occurs, can improve alarm accuracy, reduce the situation that medical resource waste occurs.

The application relates to the technical field of automobile transmission shaft production, in particular to a safety management system for automobile transmission shaft production based on big data, which comprises a data acquisition module, which is used for collecting production parameter data, environment parameter data, operator behavior data and quality inspection data of multiple production devices in the automobile transmission shaft production process in real time, and constructing an original safety data set; a big data processing platform, which is in communication connection with the data acquisition module, is used for cleaning, fusing and standardizing the original safety data set, and generating a standardized safety data set; a safety risk research and judgment module, which is connected with the big data processing platform, is used for storing a dynamic safety threshold model, and identifying abnormal states and safety risks in the transmission shaft production process in real time based on the standardized safety data set, and outputting quantitative research and judgment results; the safety risk research and judgment module realizes the whole domain monitoring, hierarchical early warning, millisecond-level emergency shutdown and closed-loop control of the transmission shaft production safety, and significantly improves the response speed and management efficiency.

The invention provides a radar point cloud-based foreign object intrusion detection method and system in a working area, and the method comprises the steps: setting a detection area range according to the working space of a mechanical connecting rod, and dynamically generating a direction bounding box (OBB) of the connecting rod through obtaining the posture state information and size data of the working connecting rod in real time; spatial screening is carried out on original radar point cloud data, only point cloud data which are located in a detection area and are not located in an OBB are reserved, and background point clouds outside the detection area and point clouds generated by a connecting rod are filtered out; and performing multi-dimensional anomaly detection on the filtered point cloud data, and comprehensively judging whether a continuously existing foreign object intrusion event exists or not. Through a linkage mechanism of point cloud filtering, background modeling and clustering verification, high-precision, low-false-alarm and real-time industrial intrusion detection is realized, and the problems of poor adaptability, high false alarm rate and the like of a traditional method in a complex environment are solved.

This utility model relates to the field of mining engineering equipment technology, and in particular to an intelligent control device for mine filling. It includes a base, a main control box fixedly connected to the top of the base, and two conveying barrels mirror-distributed around the main control box. Flange pipes communicating with the inner walls of the two conveying barrels are fixedly connected to the tops of the two conveying barrels. Flow meters are fixedly connected to the inner walls of the flange pipes. A mixing tank is installed on the outer wall of the base, and a stirring rod is rotatably connected to the inner wall of the mixing tank. A liquid level sensor is fixedly connected to the inner wall of the mixing tank. The flow meter collects the material conveying volume of the main material bin and the auxiliary material bin in real time. The controller automatically compares the actual conveying ratio with the set value, avoiding the mixing ratio error caused by manual observation of material level and manual speed adjustment, ensuring that the ratio of main material to auxiliary material remains stable within a reasonable range. A concentration sensor monitors the mixing state of the filling slurry in the mixing tank, ensuring the basic quality of the filling slurry and improving the strength stability of the filling material.

This disclosure provides a distance determination method, apparatus, and storage medium. The method includes: determining whether data acquired by an infrared distance sensor of an electronic device is greater than a distance threshold and less than a first proximity threshold; if so, acquiring light-sensing data; determining whether the light-sensing data is greater than a preset light-sensing value; if so, determining a second proximity threshold; and determining the positional state between an obstruction and the screen of the electronic device based on the second proximity threshold and the data acquired by the infrared distance sensor. This disclosure, based on data acquired by the infrared distance sensor, the first proximity threshold, the distance threshold, the light-sensing data, and an oil stain algorithm, performs multi-level precise differentiation of the positional state between the obstruction and the screen of the electronic device, effectively avoiding false alarms caused by misinterpreting obstructions as oil stains for optimization, and improving the response accuracy of the electronic device's screen.

This application relates to the field of fire early warning technology, and in particular to an intelligent fire early warning method and device for student dormitories. The method includes: collecting smoke concentration data, gas concentration data, flame signal data, and human movement status data; determining whether human movement exists based on the human movement status data; if no human movement exists, performing a fire risk assessment based on comparisons between smoke concentration data and preset smoke thresholds, comparisons between gas concentration data and preset gas thresholds, and whether the flame signal data is non-zero; if human movement exists, performing a fire risk assessment based on comparisons between flame signal data and preset flame thresholds, and whether the flame signal data is continuously rising. Therefore, this method is applicable to situations involving spontaneous combustion or electrical fires in student dormitories, avoiding false alarms caused by student activities, proactively identifying when students ignite other substances with candles or lighters, and avoiding false alarms due to dust or smog caused by heavy fog or dormitory cleaning.

The invention provides a vehicle-mounted highway tunnel brightness real-time monitoring and early warning system, and belongs to the technical field of vehicle driving, and the system comprises a data obtaining module which obtains tunnel brightness test data through a real vehicle driving experiment; the image preprocessing module obtains a tunnel driving image and preprocesses the tunnel driving image to obtain a seven-dimensional spatial-temporal feature vector; the brightness identification monitoring module establishes a tunnel brightness identification model according to the test data and the seven-dimensional spatial-temporal feature vector, and extracts a tunnel brightness value; the threshold generation and early warning module classifies the brightness values by using an improved clustering algorithm to obtain a tunnel dangerous brightness threshold and performs early warning; and the vehicle-mounted integrated module connects the tunnel brightness identification model and the tunnel dangerous brightness threshold to the vehicle to realize real-time monitoring and early warning. The problems that tunnel safety protection is high in deployment threshold and high in cost due to the fact that a professional brightness detection instrument is needed to obtain accurate brightness detection capacity when a vehicle enters and exits a tunnel, and an existing early warning mechanism lacks dynamic environment continuous quantitative evaluation are solved.

The application provides a mountainous large bridge comprehensive analysis method based on an intelligent inspection unmanned aerial vehicle, and belongs to the technical field of bridge health management. The method comprises the following steps: performing spatial overlap degree matching on each bridge disease entity identified in the current inspection and the same-position disease entity in a historical inspection archive, and determining the continuous evolution sequence of each bridge disease; determining the visual rendering coefficient of the corresponding bridge disease in the preset bridge three-dimensional model from the time sequence variation of the disease characteristic parameters in the continuous evolution sequence, and constructing a disease density distribution hot area; when the time sequence variation of the bridge disease extracted in the disease density distribution hot area exceeds the allowable evolution interval of the corresponding disease type, performing difference analysis on the neighborhood structure surface information of the bridge disease in the bridge three-dimensional model and the original design parameters of the mountainous large bridge; and updating the allowable evolution interval according to the difference analysis result. The technical scheme provided by the application can dynamically distinguish the normal damage development and abnormal accelerated degradation in the mountainous large bridge.

The utility model discloses a kind of toxic and harmful gas detection alarm control system, it is related to the technical field of instrument automatic control.The system includes field alarm, toxic gas detector, gas alarm controller, air switch, AC contactor, thermal overload relay and exhaust fan.Toxic gas detector real-time monitoring toxic gas concentration in environment, and signal transmission to gas alarm controller;Two early warning thresholds are provided in controller, when concentration reaches first threshold, trigger sound light alarm, reach second threshold, drive relay control exhaust fan start, quickly exclude leakage gas.The utility model integrates multipoint centralized monitoring, hierarchical response and overload protection function, supports 4-20mA anti-interference signal transmission, applicable to metallurgy, chemical industry and other flammable and explosive places.Has short response time, false alarm rate is low and the like characteristics, can effectively prevent personal injury and fire explosion accident, improve industrial safety protection capability.

This invention belongs to the field of analytical detection technology and discloses an analytical method for detecting N-nitrosobaricitinib impurities in baricitinib tablets. The method includes solution preparation and detection steps: blank, 0.9 ng / ml reference standard, and test sample solutions are prepared using a 40%–60% acetonitrile aqueous solution as a diluent; an Agilent Poroshell 120 EC-C18 column is used with 0.1% formic acid-methanol gradient elution; mass spectrometry is performed using an AJS ESI ion source in MRM scanning mode; and the content is calculated using the external standard method. This method is simple to operate, rapid in analysis, highly specific, highly sensitive (detection limit 4.30 ppb), and exhibits excellent precision, accuracy, and linearity. It can accurately quantify target impurities, providing reliable technical support for the quality control of baricitinib tablets.

The invention discloses an underground water multi-pollutant rapid prediction and diagnosis system, which comprises a multi-parameter sensing and data acquisition module used for acquiring underground water quality parameters in real time; the edge pretreatment unit is used for carrying out edge pretreatment on the underground water quality parameters; the off-line synthetic data engine is used for generating synthetic training data of transient extreme working conditions; the feature engineering processing unit is used for constructing interaction features based on a geochemical mechanism; the AutoML reasoning unit is used for deducing and outputting a pollutant concentration predicted value through a model; and the physical-data dual-drive fault diagnosis module is used for diagnosing the authenticity of pollution according to three-dimensional calculation results of the physical residual calculated by the physical residual calculation unit, the mean square error of the sensor data calculated by the data reconstruction error unit and the difference between the observed value and the predicted value calculated by the statistical test unit. And judging whether to give an alarm according to the diagnosis result. According to the scheme, the specific pollutants can be accurately identified, and the authenticity of pollution can be accurately identified.

The application discloses a kind of rail vehicle pipe joint loosening detection method and system, belong to rail transit vehicle pipe joint detection field;The method comprises: local posture decoupling, detects pipe joint structure features and establishes local coordinate system, estimates local transformation;Marked line skeleton extraction extracts anti-loose mark line candidate area and skeleton, obtains the detected mark line track and template mark line track;Homologous track comparison, according to the boundary division fixed side, active side and cross-seam neighborhood section, template mark line track is mapped into reference mark line track, and the joint deviation feature of the detected mark line track is calculated;Loosening determination output, according to joint deviation feature and registration confidence, calculate loosening risk score, output pipe joint state result.Template image homologous reference, local coordinate system eliminates visual angle deviation, curve topological structure represents anti-loose mark line, distinguishes line drawing non-standard and real loosening, avoids false alarm and miss report, improves detection accuracy and robustness.

The application relates to a railway foreign matter alarm method, device and equipment and a storage medium, and relates to the technical field of image processing. The method comprises the following steps: acquiring a to-be-detected image comprising a railway; determining a to-be-detected region comprising the railway in the to-be-detected image according to the to-be-detected image and a semantic segmentation model; in the case where the number of to-be-detected regions is one, judging whether a target foreign matter exists in the to-be-detected region according to the to-be-detected image and a foreign matter classification model; the foreign matter classification model is used for detecting foreign matters of a preset type in the to-be-detected image, and the type of the target foreign matter is a type other than a train in the preset type; and in the case where the target foreign matter exists in the to-be-detected region, an alarm is given. The application is used for improving the accuracy of railway foreign matter detection.

The invention discloses a flow type single molecule detection probe and method based on rolling circle amplification and application in the technical field of biomolecule detection. The flow type single molecule detection probe comprises a plurality of groups of magnetic bead detection probes and antibody binding probes which are used in pairs, each magnetic bead detection probe comprises a superparamagnetic microsphere, and the surface of the superparamagnetic microsphere is covalently connected with a plurality of fixed antibody probes; each fixed antibody probe comprises a first antibody and a first fixed DNA chain; each antibody binding probe comprises a second antibody and a second fixed DNA chain; aiming at the same target antigen, the first antibody and the second antibody in the same group can be specifically combined with different epitopes of the target antigen to form a sandwich immune complex. According to the invention, the hairpin structure DNA probe is combined with the adjacent ligation and rolling circle amplification technology to form dual specific screening of immune recognition and DNA ligation, so that the signal-to-noise ratio and sensitivity of detection are remarkably improved, and the design complexity, development cost and cross reaction risk of a multi-target detection system are greatly reduced.

The application provides an engineering quality detection data tamper-proofing and abnormal early warning method and system, relates to the technical field of engineering quality detection, and comprises the following steps: obtaining original detection data and tamper-proofing and storing evidence; generating a detection report based on the original data and tamper-proofing and storing evidence; matching a deviation allowable threshold from a dynamic parameter library according to the engineering type, grade and equipment accuracy; comparing the detection report with the original data, comparing the self-inspection data with the sampling inspection data, and obtaining comparison results; constructing a capability portrait based on historical data of a detection institution and performing abnormal scoring by using an unsupervised algorithm; determining an abnormal grade and grading early warning by comprehensively considering the comparison results and the comprehensive abnormal score; and storing evidence by chaining the hash values of the detection report and the original data and generating a verification two-dimensional code. The application realizes multi-source cross verification and can effectively identify data abnormal situations such as inconsistency between report values and original records.

The application relates to the technical field of data processing, and discloses a rain and sewage mixed connection tracing and spatial positioning method for a drainage pipe network of a micro monitoring station network, which comprises the following steps: according to the flow connection relationship of the drainage pipe network and the layout nodes of the micro monitoring station, the upstream confluence pipe network range of each micro monitoring station is defined, and a unique mapping relationship database between each pipe section and the corresponding multi-level monitoring topological sequence thereof is established; water quality and water level data of the monitoring station are acquired, and an analysis time period is intercepted; based on the space-time propagation correlation attribute of the data between the stations, a target monitoring topological sequence currently producing a response is identified in the mapping relationship database; a homomorphic mapping candidate pipe section set corresponding to the sequence is locked, and a pollutant source discharge inversion model of each candidate pipe section is constructed according to the fluid transport and diffusion mechanism of the pipe network; the model is used to calculate the fitting deviation degree of each candidate pipe section source hypothesis to the monitoring data, and the pipe section with the minimum deviation degree is selected as the positioning result.

The application relates to a tire pressure monitoring alarm method and system, which comprises the following steps: acquiring monitoring data of a first tire pressure monitoring system, and judging whether the monitoring data meets a preset strategy condition; if the monitoring data meets the preset strategy condition, determining a target alarm strategy as a first alarm strategy; if the monitoring data does not meet the preset strategy condition, acquiring auxiliary judgment information, determining the target alarm strategy from a plurality of alarm strategies based on the auxiliary judgment information, wherein the plurality of alarm strategies comprise a first alarm strategy associated with the first tire pressure monitoring system and a second alarm strategy associated with a second tire pressure monitoring system; and generating tire pressure alarm information based on the target alarm strategy. According to the application, the determination of the target alarm strategy no longer depends on the signal state of the first tire pressure monitoring system alone, but can dynamically adapt between the first alarm strategy associated with the first tire pressure monitoring system and the second alarm strategy associated with the second tire pressure monitoring system in combination with the auxiliary judgment information.

The application provides a building fire-fighting early warning system and method, and belongs to the technical field of building fire-fighting safety. The application synchronously collects temperature, smoke concentration and carbon monoxide concentration data in each monitoring area of a building, calculates the deviation degree of sensor sliding window variance and baseline variance, evaluates the reliability index of each sensor in real time and generates adaptive fusion weight, simultaneously detects and isolates the fault sensor through time sequence deviation residual analysis, evaluates the consistency between sensors and generates a correction weight, relies on the dynamic update of the environment baseline based on the exponential weighted moving average, generates a graded early warning threshold for each area and triggers the corresponding response measures. The application effectively improves the robustness of the system under the condition of sensor failure, enhances the fusion accuracy in the multi-sensor conflict scene, realizes the early prediction of the fire spread situation, reduces the false alarm rate while maintaining a high fire detection sensitivity and a short detection response time.

This invention relates to a container cluster protection method, apparatus, and storage medium. When a Node in a Kubernetes cluster crashes, all container instance names on the crashed Node are recorded in a container instance-crash node mapping table. When a Node running a container instance corresponding to a given container instance name crashes, the Node is recorded under the corresponding container instance name in the container instance-crash node mapping table. The currently crashed Node and the previously crashed Nodes are arranged chronologically to form a crashed Node list. If the length of the crashed Node list corresponding to a container instance name in the container instance-crash node mapping table exceeds a set length threshold, the container instance corresponding to that container instance name is determined to be abnormal. The abnormal container instance is then circuit-broken, preventing it from being rebuilt on the Node and avoiding further crashes of Nodes in the Kubernetes cluster caused by the reconstruction of the abnormal container instance.

The invention relates to the technical field of railway facility monitoring, in particular to a GIS-based railway facility abnormal data intelligent alarm method and system. The method comprises the steps of obtaining multi-source detection data, determining a first abnormal area and a second abnormal area, calculating a real-time pull-out value deviation, adjusting a correlation degree threshold value, correcting a track state threshold value and generating a high-risk abnormal area alarm prompt. According to the method, multi-source data structured management and spatial correlation analysis are realized through a unified data interface and an intelligent preprocessing mechanism, abnormal distribution is presented in combination with a GIS display module, the data analysis intuition and decision support capability are improved, a hierarchical alarm strategy is designed, and the timeliness and accuracy of railway facility abnormal response are remarkably improved.

The invention provides a molecular diagnosis method for avian leukosis virus and Marek's disease virus mixed infection. The molecular diagnosis method comprises the following steps: (S01) collecting diseased material tissues; (S02) carrying out nucleic acid extraction on the sick material tissue of the sick chicken; (S03) detecting the nucleic acid by using specific primers of avian leukosis and Marek's disease; (S04) preparing a PCR (Polymerase Chain Reaction) detection system, and respectively adding ALV-env gene amplification primers and MDV-meq gene amplification primers for gene amplification; (S05) carrying out 1% agarose gel electrophoresis detection on the amplification product in the step (4), and carrying out sequence determination after determining a target band; (S06) carrying out BLAST comparison on the sequences obtained by sequencing, carrying out gene sequence analysis and drawing an evolutionary tree; and (S07) determining the virus type according to the homology and the evolutionary branches, so as to realize rapid diagnosis of ALV and MDV mixed infection, strain characteristic analysis and exclusion of other similar epidemic diseases, and provide technical support for accurate prevention and control of AL and MD in the poultry industry, provenance purification and epidemiological investigation.

The present application relates to the technical field of battery management, in particular to a lithium battery pack intelligent management system based on multi-modal sensing. It comprises: a multi-modal sensing and fusion unit that collects battery pack operating environment and internal state data in real time, and fuses the operating environment and internal state data through weighted average to generate fused multi-modal data; an intelligent management master unit that predicts battery health state calculation based on cumulative gas production life mapping model, and predicts thermal runaway risk through a long short-term memory network model to generate control instructions, state information and alarm signals; a power execution unit that performs overcharge and overdischarge protection operations on the battery pack based on the control instructions. The present application collects multi-dimensional data such as voltage, current, temperature and gas concentration through a multi-modal sensing module, and extracts key features such as voltage consistency, temperature distribution and current fluctuation by combining statistical analysis and time-frequency transformation methods, thereby achieving comprehensive perception of the battery pack operating state.

The invention relates to the technical field of filling automation control, in particular to a cosmetic filling control method based on a large model knowledge graph, which comprises the following steps: acquiring an actual filling rate to determine a dynamic filling efficiency index, and determining whether the filling efficiency is stable based on the dynamic filling efficiency index. Determining the abnormal level of the filling efficiency in combination with the nitrogen filling response delay; determining that the filling efficiency is reduced due to the fact that moisture absorption raw material components are influenced by environmental factors to change the viscosity of a material body based on the combination of the viscosity-temperature sensitivity coefficient and the actual viscosity of a product and the environmental humidity, and adjusting a temperature set value of a compression area; the actual response time for adjusting the nitrogen flow is obtained to determine a response efficiency index, whether the comprehensive efficiency of nitrogen charging meets the expectation or not is determined based on the response efficiency index, and the calibration delay for determining the abnormal level of the filling efficiency is adjusted. The filling process is regulated and controlled on the basis of the product characteristics of cosmetics needing to be filled and the characteristics of filling equipment.