67results about How to "Reliable reference data" patented technology

The invention relates to the field of route selection, in particular to a route selection method under the reliability restraint condition. The method includes the obtaining step for obtaining topological information, the analyzing step for conducting reliability analysis on node equipment and optical cable links according to the topological information, the selecting step for selecting respective shortest routes for different types of services to form the shortest route list with the bandwidth resource utilization rate as the main measurement according to the result of reliability analysis, and the determining step for conducting inspection and comparison on the shortest route list for different service types according to the distribution condition of the services of the types on the basis of the service balance degree and the route reliability and selecting the shortest route with the highest reliability according to service requirements. According to the route selection method, the network bandwidth resource utilization rate is taken into consideration, and meanwhile the reliability restraint condition of nodes and the links is introduced in; the shortest route is selected, and meanwhile the requirement for the route reliability of the services is met.

The invention relates to a steam system in field of chemical industry, metallurgy and the like and in particular relates to an operation optimizing system and an operation optimizing method of a steam system of the fields. The operation optimizing system of the steam system comprises a real-time database, a relation database, an optimization model constructing model, an optimization model calculating model and a man-machine interface module. The invention further provides the operation optimizing method of the steam system, which can construct an optimized model according to actual conditions of the steam system and can utilize an optimization algorithm to calculate to obtain an optimization operation scheme, so that field scheduling personnel and managers can schedule according to the optimization scheme, the economical efficiency of the operation of the steam system is improved, the energy consumption of an enterprise is reduced and the production cost of the enterprise is reduced.

The invention relates to a multifunctional membrane separation device for treating pulping and paper-making waste water and a separation method thereof. The device is provided with four groups of membrane assemblies which respectively correspond to four membrane separation processes, wherein each group of membrane assemblies are connected with an inflow pipeline, an outflow pipeline and a circulating backflow pipeline, the inflow pipeline is connected with a blow down pipeline and a chemical washing pipeline, and the outflow pipeline is connected with a backwashing pipeline; different membrane assemblies can be singly used, serially used or used in parallel; each inflow pipeline is connected with a storage pot type heat exchanger to achieve the purpose of simultaneously carrying out temperature regulation and feed buffering; and a pressure sensor is arranged in front of a reverse osmosis membrane assembly for automatically controlling and separating required constant pressure. The device can realize the following working modes: the single working of the membrane assemblies; the serial working of the membrane assemblies; the parallel working of the membrane assemblies; the chemicalwashing of the membrane assemblies; the backwashing of the membrane assemblies; and the blow down of the membrane assemblies. The device can meet the requirement on the membrane separation of variouskinds of pulping and paper-making waste water with different components and properties and has the advantages of high experiment efficiency and small data error.

The invention provides a money collection method, a money collection device and an ERP (Enterprise Resource Planning) system used for a physical store. The method comprises the following steps of: under a commodity shopping guide mode, receiving the commodity query request of a customer, calling commodity information corresponding to the commodity query request, and displaying the commodity information to the customer; under a commodity money collection mode, adding commodities selected by the customer into a shopping cart, selecting member information corresponding to the customer, generating a bill according to a settlement way selected by the customer and the member information when the customer carries out commodity settlement, collecting money according to the bill, and generating order form information and sales record information after collection is confirmed; when a goods returning application or a refunding application of the customer is received, calling the order form corresponding to the goods returning application or the refunding application, and finishing an operation corresponding to the goods returning application or the refunding application according to the order form; and according to a set period, carrying out order form conversion rate statistics, and generating a conversion rate analysis statement according to the order form conversion rate statistics. By use of the method, a money collection system in a traditional physical store is improved.

The invention discloses a weather image intelligent identification method and system based on multi-depth convolutional neural network fusion, and the method comprises the following steps: A, collecting an image, and carrying out the preprocessing of the image; B, establishing four deep convolutional neural network models with different structures, respectively improving full connection layers ofthe deep convolutional neural network models, adding a feature layer, and performing network training based on the high-quality training data obtained in the step A; C, extracting features of newly added feature layers of the four obtained deep learning models, and performing training by adopting an Xgboost ensemble learning model to obtain a fusion model; and D, amplifying a weather image to be identified, identifying the amplified weather image by the obtained fusion model, and performing voting to obtain an identification type with the highest final vote number. According to the method, theproblem of low efficiency of manually extracting the image features in an existing traditional weather recognition method is solved, and the recognition accuracy of the deep learning model is effectively improved.

The invention relates to an identification method of battery pack inconsistency. A frequency division model is adopted. An average battery model and a differential battery model are included. An extended Kalman filter EKF algorithm is combined to realize identification of battery pack inconsistency. The method comprises the following steps of firstly, using a current sensor, a voltage sensor and a temperature sensor to collect work data of a battery pack; further, using the average battery model, combining an EKF algorithm and using a high frequency mode to estimate an average charge state of the battery pack; and finally, near an average SOC of the battery pack, using the differential battery model, combining the EKF algorithm and using a low frequency mode to estimate a difference between a monomer battery SOC and the average SOC of the battery pack. In the invention, the battery pack inconsistency can be accurately identified through a low calculated amount, then battery pack equalization or fault diagnosis is realized, and dynamic performance, safety and endurance of the battery pack are increased.

The invention discloses an internet-based electronic commerce transaction system. The system comprises an e-commerce platform, a cloud platform and a user management terminal. The system is characterized in that the cloud platform is connected with the e-commerce platform and the user management terminal through a wireless communication network. The user management terminal comprises a backgroundinformation interception module, a resource management module and a transaction database. The e-commerce platform comprises an e-commerce resource overall planning module, a transaction commodity acquisition module and an e-commerce resource configuration module. The cloud platform comprises a transaction screening processing module, a resource comprehensive screening module, a platform managementmodule and a cloud server. According to the invention, the sales credit evaluation coefficient of a store and the purchase credit evaluation coefficient of a purchase user are counted respectively. According to the counted sales credit evaluation coefficient and purchase credit evaluation coefficient, the corresponding store or buyer user is pushed to the user or the store. Therefore, the matching degree of the purchase user and the store is improved. The bidirectional sales and shopping credit of the store and the buyer are increased. The satisfaction degree of the store and the buyer is greatly improved.

The invention discloses a road and bridge safety intelligent monitoring and early warning management system based on big data. The system comprises a region division module, a detection point distribution module, a pressure detection module, a wind speed detection module, a wind direction detection module, a wind direction angle analysis module, an early warning reminding module, a handrail heightdetection module, a handrail height analysis module, a remote service center, an analysis server and a storage database. According to the invention, the pressure of crowds after extruding the guardrail is detected; the wind speed and the wind direction around the guardrail are detected; a wind direction angle corresponding to the wind direction is analyzed; the total pressure borne by the guardrail is calculated; whether a guardrail is at risk is analyzed, early warning and reminding are carried out on the guardrails with dangers, meanwhile, the comprehensive safety coefficient of the guardrails is comprehensively analyzed by detecting the heights of the guardrails and combining the number of the guardrails, the guardrails with unqualified safety are blocked, and meanwhile inspection andmaintenance are carried out, so that the occurrence rate of guardrail collapse events is reduced, and the sense of safety is brought to people.

The present invention relates to an electronic jacquard machine jacquard needle position detection mechanism. It includes a connecting component which is connected with said jacquard needle and can be upwards and downwards moved together with said jacquard. The described connecting component includes at least an electroconducting portion and an insulating portion. Said detection mechanism also includes a detection switch, said detection switch includes an insulating plate, first electrode fixedly set on the described insulating plate and second electrode positioned under said first electrode. The connecting component is slide contacted with the described first electrode and second electrode. Besides, said invention also provides the working principle of said detection mechanism and its concrete operation method and steps.

The invention discloses a packaged optical fiber sensor based on a single capillary glass tube and a testing method thereof. The optical fiber sensor comprises a single-mode optical fiber, a first thermal regeneration fiber grating, a second thermal regeneration fiber grating and a capillary glass tube. The first thermal regeneration fiber grating and the second thermal regeneration fiber gratingare both inscribed on a fiber core of the single-mode fiber; a preset distance is formed between the first thermal regeneration fiber grating and the second thermal regeneration fiber grating; the capillary glass tube sleeves the outside of the second thermal regeneration fiber grating; and one end of the capillary glass tube is fixed on the single-mode fiber between the first thermal regenerationfiber grating and the second thermal regeneration fiber grating through high-temperature glue. Therefore, differentiated measurement of the temperature and strain can be realized.

The invention discloses a water scaling detection method, device and system. The detection method comprises the following steps: controlling the operation of a target device, which can simulate the operation of a water heater, detecting the status data of the target device during the operation process, and finally obtaining the working condition data of water scaling. The problem that in the prior art the water heat design is usually deviated from the actual situation can be solved by the provided method, the water scaling situation in a water heater can be faithfully represented by the working condition data, and a reliable reference is provided for the water heater design.

The invention discloses a determination method for sympathetic explosion distance of an explosive. The method comprises the following steps: an explosive is prepared into main triggering explosive cartridges and explosive cartridges to be triggered with a diameter of 35 mm, with a specified density, and with a weight of 150 g or 200 g; the main triggering explosive cartridge and the explosive cartridge to be triggered are fixed on a rope or a rod; the two explosive cartridges are arranged at a certain distance; a No.8 detonator is inserted in the main triggering explosive cartridge and the insertion depth is two-thirds of the detonator length; the energy focusing cave of the detonator is arranged toward the explosive cartridge to be triggered during testing; after initiation of explosion, the sympathetic explosion condition of the explosive cartridge to be triggered is checked, after three successive sympathetic explosions, the longest distance is taken as the sympathetic explosion distance of the explosive. When the method is combined with the explosive application condition tests, the obtained results can directly reflect sympathetic explosion safety distance of an explosive under application conditions. The method can provide reliable reference data for storage, transportation and usage of explosives.

The invention provides a wind speed prediction method based on STCP-BP. The wind speed prediction method based on the STCP-BP includes: firstly, utilizing wind speed and wind direction information of an upwind direction neighborhood point of the prevailing wind direction of a target prediction point to build a spatial and temporal correlation prediction model, and obtaining a wind speed prediction value of the target prediction point; secondly, utilizing historical wind speed data of the target prediction point to build a BP neural network direct multi-step prediction model, and obtaining a future multi-step wind speed prediction value of the target prediction point; and finally, performing mean combination prediction on the two prediction values, and obtaining a wind speed direct multi-step prediction value of the target prediction point. According to the wind speed prediction method based on the STCP-BP, wind speed information of an upwind direction observation point is added on the basis of a traditional neural network prediction method, spatial and temporal correlation properties of wind speed evolution is considered, a statistical method is adopted for achieving multi-step prediction, and wind speed prediction accuracy of wind power plants is improved greatly. The wind speed prediction method based on the STCP-BP can provide reliable reference data for a dispatch department to make and adjust plans, thereby reducing influences of wind power intermittency on power grids.

The invention provides an instrument and method for monitoring a cavity beneath a tunnel passing through a ground fissure zone. The monitoring instrument comprises a cloud deck close to a ground fissure at the bottom of the tunnel, and a laser range finder on the cloud deck, wherein connecting pieces are pre-buried in the central position of the bottom of a subsection of the tunnel passing through the ground fissure, the cloud deck is installed on the connecting pieces, and then, the laser range finder is arranged on the cloud deck; and data control software in a data processing system is utilized to control the laser range finder to automatically acquire cavity monitoring data at fixed time. The monitoring instrument utilizes the range finding principle of the laser range finder, is simple in structure, easy to operate and low in cost, and can monitor the cavity area in real time and obtain the variation conditions of the cavity area timely.

The invention relates to a forced oil circulation transformer oil-submerged pump fault monitoring device and monitoring method. According to the technical scheme, an oil outlet of an oil-submerged pump is connected with an oil inlet of a main transformer through an oil conveying pipeline, the oil conveying pipeline is provided with a sampling port for collecting an oil sample, an oil-submerged pump outlet valve is arranged on a pipeline between the sampling port and the main transformer, and an oil-submerged pump inlet valve is arranged on a pipeline connected with an inlet of the oil-submerged pump; and the online monitoring device further comprises a transformer oil chromatography tester used for detecting the total hydrocarbon content of dissolved gas in an oil sample taken out from thesampling port. The device can realize online measurement of the oil sample at the sampling port of the oil-submerged pump in an operation state for analysis of dissolved gas in oil, intuitively and effectively judges whether the oil-submerged pump is in a fault state or not through data, is high in monitoring accuracy, provides reliable reference data for equipment operation, and greatly shortensthe fault judgment time length of the oil-submerged pump of a forced oil circulation transformer.

The invention discloses an intelligent anesthesia monitoring system, and belongs to the technical field of medical equipment. The intelligent anesthesia monitoring system comprises an anesthesia monitor, a video monitor, a server, a repeater and a mobile terminal, wherein the server is connected to the anesthesia monitor and the video monitor through the Internet; the repeater is connected to theserver through the Internet; the mobile terminal comprises a query communication module, a query control module, a query display module and a query selection module; the query communication module isconnected to the repeater through a local area network; the query communication module, the query display module and the query selection module are in communication connection with the query control module; and a user of the mobile terminal can query and perform real-time observation on vital sign data or video data of patients with different operating table numbers through the mobile terminal. The intelligent anesthesia monitoring system solves a problem that anesthetists cannot obtain monitoring data of patients in different operating rooms in real time.

The invention discloses a system and method for testing the surface quality of Wolter-I-type mandrel by adopting a contact-type contourgraph. The test system consists of two portions, i.e., a countourgraph and an auxiliary adjusting apparatus, wherein the auxiliary adjusting apparatus comprises an axis-X rectilinear displacement platform, a horizontal rotary turntable, a vertical surface rotary turntable and a chuck for clamping a workpiece, the axis-X rectilinear displacement platform is fixedly installed on a table-board of a contourgraph detection platform, the horizontal rotary turntable is fixedly installed on the axis-X rectilinear displacement platform, the vertical surface rotary turntable is fixedly installed on the horizontal rotary turntable, and the chuck is fixedly installed on the vertical surface rotary turntable and used for installing and clamping the mandrel. By adopting the system and method, performance indexes of the Wolter-I-type mandrel such as dimension errors,surface roughness, corrugation degree and shape errors can be detected, no other additional ultra-precision measurement instrument is needed, the purchasing cost of the ultra-precision measurement instrument can be reduced, the measurement times can be reduced, and the measurement efficiency can be increased.

The invention provides an electric car charging-discharging behavior prediction method. The method includes the steps of a, setting electric car charging-discharging basic constraint; b, setting a target function; c, calculating MSN influence omega 2; d, using omega 2 to correct cross inheritance particle swarm algorithm parameters; e, using the electric car charging-discharging basic constraint as premise, and using the corrected cross inheritance particle swarm algorithm so solve the target function so as to obtain an electric car charging-discharging plan and power distribution network load. The method has the advantages that the cross inheritance particle swarm algorithm is used to predict the charging-discharging plan of an electric car user, influence of a mobile social network on the charging-discharging plan is fully considered, prediction result accuracy is increased greatly, and reliable reference data can be provided to the power supply department during power grid load adjusting.

A new EMU roof insulation detection device is composed of a processor, a control / display panel, an inverter, a voltage regulator, a current / voltage sampling module, a step-up transformer, a permanent magnet mechanism, and a vacuum interrupter. The device can be divided into a high-voltage signal generation part and a control part, the former is installed on the roof, and the latter is installed in the electrical cabinet of the car. The invention introduces the composition principle of the insulation detection system, the structure of main units and the characteristics of the new insulation detection system. The device has the function of automatic data storage. These data provide a scientific and accurate basis for judging the insulation performance of the device, and also provide real and reliable reference data for the insulation design of the high-voltage equipment on the roof. reference. The device has the function of repairing the insulation performance degradation of high-voltage equipment caused by pollution and smog, thereby reducing the flashover accidents caused by pollution and smog.

The invention discloses a tacrolimus metabolic gene detection kit. The tacrolimus metabolic gene detection kit comprises a primer pair shown as SEQ ID No.1-2, a probe pair shown as SEQ ID No.3-4, a primer pair shown as SEQ ID No.5-6, a probe pair shown as SEQ ID No.7-8, a primer pair shown as SEQ ID No.9-10, and a probe pair shown as SEQ ID No.11-12, wherein the primer pair shown as SEQ ID No.1-2and the probe pair shown as SEQ ID No.3-4 are used for detecting CYP3A5 6986A>G gene locus mutation, the primer pair shown as SEQ ID No.5-6 and the probe pair shown as SEQ ID No.7-8 are used for detecting ABCB1 1236C>T gene locus mutation, and the primer pair shown as SEQ ID No.9-10 and the probe pair shown as SEQ ID No.11-12 are used for detecting POR 1508C>T gene locus mutation. The kit can accurately judge the tacrolimus metabolic type of a to-be-detected person, meanwhile the tacrolimus metabolic degree can further be judged by combining data, and the more reliable reference basis is provided for clinical safe and effective medication.

The present invention discloses a method and a system for extracting data parameters of left auricle based on CT images. The method comprises the following steps: acquiring a coronary artery CTA imagesequence; performing three-dimensional reconstruction on the coronary artery CTA image sequence to obtain a three-dimensional coronary artery image; utilizing the three-dimensional coronary artery image and a pre-established left atrium segmentation model to obtain a left atrium three-dimensional image; extracting data parameters of left auricle from the left atrium three-dimensional image, wherein the data parameters of left auricle at least comprise a left auricle opening position, a left auricle opening area, a left auricle opening longest diameter, a left auricle opening shortest diameterand a left auricle depth; and generating an image analysis report according to the extracted data parameters of left auricle. The method and system for extracting data parameters of left auricle based on CT images can obtain diversified three-dimensional data of left auricle efficiently, rapidly and accurately and provide effective reference for follow-up clinical evaluation and scientific research analysis.

The invention discloses a BIM technology-based building steel structure performance analysis method and device and a computer storage medium, and the method comprises the steps: obtaining the coordinates of each connection node of a BIM steel structure building information model in each grade of earthquake simulation experiment, and analyzing the displacement distance of each connection node in each grade of earthquake simulation experiment; meanwhile, detecting stress-strain parameter data and bearing capacity parameter data of each connection node in each grade of earthquake simulation experiment, monitoring a welding cracking area and a welding deformation degree coefficient of each connection node in each grade of earthquake simulation experiment, and calculating an anti-seismic performance estimation influence coefficient of each connection node in each grade of earthquake simulation experiment; and analyzing the anti-seismic performance state of each connection node in each grade of earthquake simulation experiment, and displaying the position of each connection node with anti-seismic hidden danger in each grade of earthquake simulation, thereby determining the influence of different grades of earthquakes on the anti-seismic performance of the steel structure building, and further carrying out anti-seismic reinforcement on the steel structure of the building in advance.

The invention discloses a bus intelligent control system. The system includes a data server, a monitoring terminal and a plurality of bus terminals; the plurality of bus terminals are connected with the monitoring terminal through the data server; each bus terminal includes a controller as well as an environment detection module, a vehicle speed detection module, a vehicle distance detection module, a people statistics module, an electrical equipment detection module and a wireless communication module which are all in control connection with the controller; and the data server includes a datastorage module, a safety diagnosis module, a risk assessment module and an emergency command module. According to the system of the invention, the speeds, distances and electrical equipment of busesare detected through the bus terminals, so that reliable reference data can be provided for the data server in safety diagnosis and risk assessment; and the emergency command module sends corresponding emergency plans to the bus terminals according to the bus risk assessment of the buses, and therefore, an early warning function is realized, and the safety of passengers and the buses can be improved, and casualties can be reduced.

The invention discloses a wheat plant real-time counting method based on deep learning image segmentation, and the method comprises the steps: collecting a plurality of wheat images through an unmanned plane, the images being three-dimensional tensors and comprise the image height, the image width and the waveband number, and a header file corresponding the images comprising a longitude coordinate, a latitude coordinate and a photographing height; the unmanned aerial vehicle comprising a deep learning module, carrying out deep learning model processing on the collected wheat image in real time, predicting the number of wheat plants in a phase amplitude range, and interpolating the spatial distribution of the number of wheat plants to form a plant number distribution diagram of a target area. The integrated technology of unmanned aerial vehicle remote sensing and computer vision target recognition is used, spatial distribution of the number of wheat plants is monitored in real time based on a deep learning algorithm, the wheat plants can be accurately recognized and positioned, and the number of the wheat plants is calculated.

The invention provides an e-commerce video live broadcast platform based on cloud computing data analysis. The e-commerce video live broadcast platform comprises a specialty product classification andstatistics module, a database, a specialty product resource configuration module, a transaction screening processing module, a transaction specialty product acquisition module, an analysis cloud platform, a user pushing module and a specialty product video live broadcast module. By counting sales a credit evaluation coefficient of a shop owner, a buyer with the high sales credit evaluation coefficient is preferentially pushed to the shop owner, conversion between mandarin and regional dialect in e-commerce live broadcast can be realized in the specialty video live broadcast module, the problems that in the prior art, the sales credit evaluation coefficient of a store cannot be effectively evaluated, conversion between mandarin and regional dialects cannot be achieved, and the matching degree between a buyer and the store is poor are solved. conversion between mandarin and regional dialects cannot be achieved, and the matching degree between a buyer and the store is poor are solved, the satisfaction degree and trust degree of shopkeepers to buyers are greatly improved, and the richness of e-commerce live broadcast is increased.

The invention relates to a method and a system for analyzing an eye picture and attribute information. The method comprises the following steps: acquiring the eye picture of a user suffering from keratitis and the attribute information of the user; judging classification information of the eye picture according to the inflammatory infiltration and / or ulcer shape of the eye picture; and according to the attribute information and the classification information, obtaining the probability that the keratitis suffered by the user belongs to different categories. A professional ophthalmologist is notneeded. The probability of the keratitis type of the user can be obtained only by uploading the eye picture and the attribute information, reliable reference data are provided for the user or doctorsto make diagnosis and treatment decisions, the method can be popularized to primary hospitals, more primary crowds are popularized, the defect that existing medical resources are not evenly distributed is overcome, and burdens of professional ophthalmologists are relieved.

The invention discloses a testing method of the peeling strength between a sealing gum and a stuck sample of a stuck base material, which manufactures the stuck base material sample into two T-shaped sample blocks, and a cavity with specified size is reserved in the middle for injecting the sealing gum, thus integrating concrete and the sealing gum into a whole stuck sample which is placed in theenvironment with the temperature of 20 DEG C minus or plus 2 DEG C and the relative humidity of 65 percent minus or plus 5 percent for being conserved for 28 days. When in test, an opening with the length not larger than one tenth of the total seaming interface length is cut between the bonding surfaces of the sealing gum and a cement mortar by a sharp blade and is used as a peeling opening, the sample is placed in a clamping tool, the clamping tool is connected with a tensile machine which is operated at the speed of 5 millimeters per minute, the stuck base material is stretched in an opposite direction, and until the sealing gum 2 is totally peeled off the concrete 3, the largest tensile value is the peeling strength between the sealing gum and the stuck base material sample. The testing method of the peeling strength between the sealing gum and the stuck sample of the stuck base material has simple operation, convenient use and favorable popularization, has no need of new additional equipments, and has strong guidance and remarkable industrial practice significance to specific industrial practices.

The invention discloses a bridge health and safety online monitoring cloud platform based on big data comparative analysis. The bridge health and safety online monitoring cloud platform comprises a region division module, a detection point layout module, a bridge opening radius detection module, a bridge opening radius analysis module, a vault arc length acquisition module, a vault arc length analysis module, an analysis server, a cloud monitoring platform and a storage database. The method comprises the following steps: dividing a bridge opening area of a to-be-monitored arch bridge into a plurality of sub-areas, carrying out detection point arrangement on each sub-area in the bridge opening area, detecting the bridge opening radius at the detection point position of each sub-area, comparing to obtain the bridge opening radius difference value of each sub-area, and simultaneously obtaining the arch crown arc length of each sub-area in the bridge opening area, comprehensively calculating the bridge opening health safety influence coefficient of the to-be-monitored arch bridge, analyzing whether the bridge opening health safety of the to-be-monitored arch bridge is qualified or not, and if not, sending an early warning instruction to inform related personnel to carry out blocking and processing, so that the travel potential safety hazards of people are reduced, and the life and property safety of people is guaranteed.

The invention relates to an insulation on-line monitoring device and method for an impact capacitor in a generator outlet circuit breaker, a capacitor shell and the ground are ingeniously separated through an insulation cushion block and connected through an annealed copper wire, current passing through the capacitor is conveniently collected, a PT in a GCB is directly used, reliable reference data are provided for equipment operation, and the monitoring accuracy is improved. The risk of unplanned shutdown of equipment is reduced, and online monitoring of the inter-electrode insulation of the impact capacitor which is closed in the GCB and cannot be observed directly can be realized through capacitance testing. The method provides big data support and reference for operation, inspection and maintenance of a power plant, can quickly and accurately judge whether tripping is caused by impact on insulation of a capacitor when a power generation enterprise has unplanned shutdown, is helpful for the power generation enterprise to quickly check tripping reasons, completes unit grid-connected operation in the shortest time, and improves the power generation efficiency. And electric quantity assessment of a power grid dispatching department to power generation enterprises is reduced.

The invention discloses an auxiliary lesion judgment device which comprises a receiving module, used for receiving an endoscope picture; a comparison module, in signal connection with the receiving module, and used for comparing the endoscope picture with each focus detection and identification model for determining the focus category in the focus identification library, and determining whether asuspected focus exists on the endoscope picture or not; a labeling module, in signal connection with the comparison module and used for correspondingly labeling each suspected focus at the position ofeach suspected focus on the endoscope picture after the comparison module determines that the suspected focus exists on the endoscope picture; and a display module, in signal connection with the labeling module and used for displaying the endoscope picture labeled by the labeling module. In the device, the focus detection and recognition model is used for recognizing the focus in the endoscope picture, the suspected focus is marked to reflect the position of the suspected focus, and then the endoscope picture marked with the suspected focus is displayed, so that the automation degree of the endoscope picture recognition link can be improved, and the working intensity of doctors is reduced.