53results about How to "Realize real-time early warning" patented technology

The invention relates to an early warning method at four-grads of power network operation, and belongs to the automation field of electric power system management. The method includes the following steps: a safe, stable and comprehensive evaluation is carried out at four-grads according to time scales; the first level is to carry out security evaluation to the real time situation of the power network, and to identify the safety level of the power network operation; the second level is to carry out security collation to the power network forecast operation situation in the next 15 minutes (before the power network operation), and to send a security alarm; the third level is to carry out security collation to a power-generating plan and a servicing plan of the next day; and to send an alarm; the fourth level is to carry out security collation to a power-generating plan and a servicing plan of the next week. The early warning method ensures that the power network can run safely and stably through the safety early warning at four time scales, namely real time, pre-operation, previous-day planning and week planning.

The invention belongs to the field of foundation pit engineering, and provides a super deep underground diaphragm wall deformation automatic real-time monitoring device and an operating method thereof. The monitoring device comprises a fiber grating reinforced stress sensor, a fiber grating concrete strain sensor, a denoising analyzing device and an automatic deformation recognizing device. By the aid of the monitoring device and the operating method thereof, horizontal deformation of an underground diaphragm wall can be tracked and monitored in real time, deforming data can be automatically collected and analyzed, real-time capacity, continuity, effectiveness and reliability of monitoring can be guaranteed, processing efficiency of monitoring data of horizontal deformation of the underground diaphragm wall can be improved, and information sharing level of monitoring data of horizontal deformation of the underground diaphragm wall can be raised, the deformation of foundation pits is guaranteed in a controllable range, constructing risks of deep foundation pits can be reduced, and references can be provided for similar foundation pit engineering.

The invention provides an equipment fault intelligent monitoring method based on an event map technology. According to the method, the historical data and the current data can be learned, and the maintenance suggestions and the fault early warning are provided for equipment maintenance. The method comprises the steps of S10, constructing an event graph through the historical data analysis, the event graph comprises the event elements, and the event elements and the event elements form a network-shaped event graph through event relations, and the event element comprises a core node and an eventattribute; S20, comparing the current equipment state with event attributes in the event map, and matching the current equipment state with corresponding core nodes and event elements; S30, updatingthe event map in real time, and learning a new fault, a new state and a corresponding maintenance scheme of the equipment; S40, judging a current equipment fault condition according to the event element, analyzing an equipment fault generation reason, and giving a maintenance scheme; and S50, according to the relationship between the event elements in the event map, predicting faults which may occur to the current equipment, and giving out fault early warning.

The invention discloses an embedded system based on abnormal behavior recognition. The objective of the invention is to provide the embedded system based on abnormal behavior recognition, and the system comprises a monitoring camera, a behavior recognition main unit, a server (data center), and a comprehensive management platform. The system is characterized in that the monitoring camera is used for collecting onsite video data; the behavior recognition main unit is used for judging the posture of a person on site; the server (data center) is connected with the behavior recognition main unit,and carries out the comprehensive analysis according to the data transmitted by the behavior recognition main unit, and then obtains a final conclusion, and transmits the conclusion to the comprehensive management platform and authorized APP. Through the plug-and-play intelligent behavior recognition main unit and the front-end high-definition monitoring camera, the system can achieve the all-weather early warning of a suspected person and event in a key region, and achieves the early warning against the abnormal behaviors of the people and the weapons and dangerous knives held by the people in a controlled region.

The application provides a method, a device and a system for alarm of an abnormal behavior, and a storage medium. The method comprises the steps of: acquiring video streaming data comprising a multi-frame image; loading an anomaly recognition model, processing the multi-frame image through the anomaly recognition model, obtaining a recognition result; when an abnormal behavior is recognized, obtaining alarm information corresponding to the abnormal behavior according to a mapping relationship between abnormal behaviors and alarm information; and transmitting the alarm information to an intelligent gateway, and enabling the intelligent gateway to transmit the alarm information to corresponding target multi-media equipment. Through adoption of the method of the invention, an alarm efficiency of the abnormal behaviors can be improved.

The invention relates to the technical field of artificial intelligence, in particular to a power grid equipment fault diagnosis method and system based on artificial intelligence. The method comprises the following steps: collecting various types of electric parameter data of power grid equipment, wherein the data of each type of electric parameter form a real-time updated data sequence; predicting the prediction data of the electric parameters corresponding to the future time according to the data sequence of each type of electric parameters, and forming a first data sequence together with the data sequence; performing anomaly evaluation on the first data sequence of each type of electric parameters to obtain an anomaly index, and obtaining an anomaly value sequence according to the correlation of each two types of electric parameters and the corresponding anomaly index; and reasoning the real-time abnormal value sequence to obtain a real-time result of equipment fault diagnosis. According to the embodiment of the invention, the accurate real-time fault diagnosis can be carried out on the power grid equipment, and the safety protection capability of the system is improved.

The present invention relates to a forest fire alarm method based on a transmission network. A remote alarm terminal and a field detection terminal are provided. The field detection terminal is arranged on an electric tower rack and is configured to obtain the forest state information of an area to be detected and upload the obtained forest state information to the remote alarm terminal, and the remote alarm terminal performs data analysis to determine whether the fire is generated or not, generate fire alarm information and issue the fire alarm information to inspection personnel hand-held terminals. The forest fire alarm method based on the transmission network realizes the monitoring of forest fire of remote areas and effectively improve the forest fire hazard prevention and management and control.

The invention discloses an air pollution monitoring system based on NB-IoT technology and edge computing. The air pollution monitoring system comprises a sensing layer, a transmission and processing layer and a platform layer. The sensing layer takes an STM32 processor as a main control chip, collects atmospheric pollutant information by using a laser dust sensor and a gas sensor, and sends monitoring data to an edge computing server through an NB-IoT network base station to be cooperatively processed with a cloud server, so that real-time analysis and processing of atmospheric pollutant data are realized; and the processed data are uploaded to an upper computer, so that the position of a pollution source on a map can be positioned on line, and an early warning function can be realized through a short message mode or a mode of checking pollution indexes in real time through a Web page. According to the invention, a mode of cooperative processing of a plurality of edge computing servers and the cloud server is implemented through NB-IoT wireless communication network transmission, the load of a cloud platform is reduced, the data transmission efficiency and the accuracy of atmospheric pollution data monitoring are improved, acquisition, transmission, analysis processing, early warning and positioning of atmospheric pollution monitoring are realized in an all-around manner, and it is convenient for a user to know the air pollution condition in time.

The invention relates to a river tidal bore segment real time early warning method. The present tidewater prediction method has great error. The method of present invention carries early warning forecast by utilizing historical data and real time information, and concretely includes steps of dividing the river by N detecting points, querying time of arriving a early warning point of traditional Chinese calendar date in past 5 to 15 years and time arriving the early warning point of yesterday, monitoring time of the tidewater to arrive at a previous detection point and time to arrive at a further previous detection point, estimating real time arriving time f1(TK-1), lag arriving time f2(Tyesterday, k) and average arriving time f3(Thistory, k) of the tidewater to arrive at a early warning point k, forecasting time Tk of the tidewater to arrive at the early warning point; detecting tidal bore height Hk-1 of the tidewater at a previous detection point, querying proportion relation of mean value of tidal bore height of the tidewater to arrive at the previous detection point at corresponding date in past 5 to 15 years and mean value of tidal bore height to arrive at the early warning point, and obtaining tidal bore height Hk of the tidewater at the early warning point. The method of the present invention not only eliminates indefinite factor influence of last day, but also realizes quantitative forecast early warning.

The invention discloses an early warning device of a vehicle based on binocular panoramic stereoscopic vision, and relates to the field of intelligent driving. The early warning device comprises an external collection device and an internal control device. The external collection device collects panoramic stereoscopic images around the vehicle through three binocular stereoscopic cameras with horizontal visual angles and same included angles, the images are processed by a central processing unit of the internal control device, image splicing, environment recognition and speed detection and judgment are achieved, and an alarm device gives out early warning. According to the early warning device, a vision-based sensing technology is adopted, the data is easy to obtain, and the information amount is large; meanwhile, due to the adoption of a single sensing technology, the complexity of the system is reduced, and therefore the cost is reduced; meanwhile, real-time early warning on potential danger of the vehicle is achieved, and the driving level of the intelligent vehicle is further increased.

The invention relates to a power transmission line forest fire monitoring system based on a satellite receiving system, which comprises a satellite receiving antenna (1), a low noise amplifier (2), a data signal feeder line (3), a data input card (4), a first computer (5), a second computer (6) and a third computer (7) sequentially connected in series. The system has the advantages that various kinds of remote sensing satellite information broadcast by national satellites and meteorological satellites, the type and the quantity of received satellite information can be freely selected according to requirements, and the cost performance is high; the degree of automation is high, the requirement of timeliness can be guaranteed, and real-time early warning can be realized; system hardware is not influenced by satellite change and can be upgraded freely; and high-accuracy time reference can be provided for satellite tracking and data input, and the system stability is very reliable.

The invention discloses a rail vehicle anti-collision system based on wireless ultraviolet light. The rail vehicle anti-collision system comprises a transmitting terminal installed at the tail of a front vehicle and a receiving terminal installed at the head of a rear vehicle. The transmitting terminal comprises a micro control unit (MCU) unit A, and the micro control unit (MCU) unit A is separately connected with a power module A and an ultraviolet LED array. The receiving terminal includes a micro control unit (MCU) unit B, and the micro control unit (MCU) unit B is separately connected witha photomultiplier tube (PMT), a power module B, a display module, an alarm module and a filter. The invention further discloses a rail vehicle anti-collision method based on the wireless ultravioletlight. The rail vehicle anti-collision system and method based on the wireless ultraviolet light solve the problems of poor positioning accuracy and weak signal receiving capability existing in an existing train anti-collision technology.

The invention discloses a urban road intersection risk assessment and early warning method based on driver physiological data. The method comprises the steps of firstly establishing an urban intersection vehicle risk assessment model; then calibrating a parameter wj in the vehicle risk probability model, and determining the vehicle risk probability model corresponding to the currently implementedintersection; constructing a vehicle risk grading matrix, and calculating the risk value of the target vehicle at the currently implemented intersection; determining a target vehicle risk level according to the vehicle risk level division matrix, and determining the risk level of the currently implemented intersection; and finally, determining early warning information according to the risk levelof the currently implemented intersection, and carrying out early warning on vehicles entering the currently implemented intersection. According to the urban road intersection risk assessment and early warning method based on the driver heart physiological data, the data is easy to obtain, the model calculation is relatively convenient, the traffic risk of the urban intersection can be effectivelyassessed, early warning is issued, and the method has relatively high operability.

The invention discloses an electric multiple units train brake pad intelligent diagnosis system which comprises a rail side basic detection unit, a field control center and a remote control center. The rail side basic detection unit comprises a trigger switch, an image acquisition unit, a rail side control unit and a vehicle number identification module. When an electric multiple units train wheelset passes through the trigger switch, the trigger switch is triggered and coming vehicle information is sent to a sleeper side control unit, and after receiving an instruction of the trigger switch,the sleeper side control unit controls an image acquisition unit to perform image acquisition on a brake pad at a bottom of an electric multiple units train and sends a collected image to a field control center. The field control center recognizes the brake pad at the bottom of the electric multiple units train through the feature mode recognition technology and the edge detection technology, andthen the missing condition of the brake pad is judged through the image comparison technology. According to the invention, the working efficiency is improved, serious consequences caused by manual missing detection, missing repair and errors are effectively prevented, and the operation safety of the electric multiple units train is ensured.

The invention relates to an intelligent safety risk early warning method and system for drilling engineering. The method comprises the steps: obtaining adjacent well abnormal data of a to-be-drilled well through a preset knowledge base, and building an abnormal early warning model of the to-be-drilled well according to the adjacent well abnormal data; acquiring real-time data information of the to-be-drilled well in a drilling process; and performing early warning analysis on the real-time data information based on the abnormity early warning model, and sending an early warning analysis resultto a remote terminal. According to the method, the abnormal early warning model is established, early warning analysis is carried out based on the abnormal early warning model; tracking, forecastingand early warning can be carried out in time, the well site risk can be effectively avoided, the accident occurrence rate is reduced, efficient drilling and intelligent decision making are achieved, unattended operation of the well site is achieved, the intelligent degree is higher, and the working efficiency is obviously improved.

The invention discloses a positioning automatic discrimination system of acute myocardial infarction based on CNN neural network, and relates to the technical field of myocardial infarction positioning discrimination. The positioning automatic discrimination system comprises a data acquisition system, a cloud platform data storage system, a positioning discrimination analysis system and a data display system. A wearable ECG monitor is connected with to-be-discriminated personnel to record and generate a 12 lead primitive electrocardiogram; an electrocardiogram acquisition system acquires 12 lead primitive electrocardiogram data including wave amplitude of P wave, wave amplitude of a QRS wave group, wave amplitude of a ST section and wave amplitude of T wave; and the positioning discrimination analysis system utilizes a positioning discrimination model acquired based on CNN neural network training to carry out convolution calculation to acquire discrimination intermediate data, and thediscrimination intermediate data are mapped by a sigmoid function to acquire discrimination result data, so that positioning discrimination of occurrence positions of the acute myocardial infarction of the to-be-discriminated personnel is made. The positioning automatic discrimination system of the acute myocardial infarction based on the CNN neural network makes accurate positioning discrimination of the occurrence positions of the acute myocardial infarction of the to-be-discriminated personnel.

The invention discloses a low-voltage power distribution network cascading failure early warning method based on a risk assessment model, and the method comprises the steps: data collection and processing: collecting load data, weather data and fault data of a low-voltage power distribution network, and carrying out the data preprocessing of the collected data; fault feature selection: adopting an improved G-ReliefF algorithm to carry out optimal selection of fault features of the low-voltage power distribution network; performing cascading failure search: performing subsequent failure search according to the line outage model and the key line model, generating an accident chain and evaluating the accident chain; and fault early warning: calculating a risk assessment coefficient of the element and the line, determining a fault occurrence probability through the risk assessment coefficient, finally carrying out comprehensive risk assessment on the fault, and calculating a risk level. The method is used for predicting and mastering the safe operation state of the low-voltage power distribution network, discovering potential dangers and judging the fault development trend so as to improve the accuracy of fault early warning and further avoid large-area power failure.

The invention provides a device for video analysis. The device is applied to a mobile terminal. The device comprises a video capturing unit, a real-time recognition unit and a video real-time analysis unit, wherein the video capturing unit is used for capturing target video data; the real-time recognition unit is used for loading a real-time analysis algorithm of the captured target video data; the video real-time analysis unit is used for comparing the captured target video data with preset video data according to the real-time analysis algorithm, so that a comparison result is acquired, and when the comparison result meets a preset alarm condition, an alarm is given. The invention further provides a system for video analysis. The system comprises a terminal server and at least one device for video analysis. The terminal server is used for storing, analyzing and counting the captured target video data. The system is applied to a mobile device, the mobility and the flexibility of video capture are greatly improved, a modularized structure is adopted in the system for video analysis, real-time videos and historical videos are separately stored, analyzed and counted, and the video retrieval speed is increased.

The invention discloses an abnormal early-warning method for the earthquake resistance of a building earthquake-resistant support hanger. The method includes the steps that an acceleration sensor is installed in the middle of an earthquake-resistant slant support of an earthquake-resistant support hanger and used for monitoring an earthquake-resistant-slant-support acceleration response caused byan environmental load, original data of the acceleration sensor is obtained, then structural natural periods of vibration and acceleration root mean square values of multiple sets of data are calculated, a relevance model of the structural natural periods of vibration and acceleration mean square roots is established with the linear regression method, the slope of a linear regression model can beobtained with the least square method, and therefore three standard early-warning indexes are established. When the measured early-warning indexes are over then the standard certain range, it is shownthat the earthquake-resistant support hanger needs to be repaired. By means of the abnormal early-warning method, early warning can be accurately provided for the earthquake-resistant support hanger,the whole safety of the earthquake-resistant support hanger under the earthquake effect is kept, the aim of reducing and positively avoiding secondary disaster is thus achieved, and the effect is rarely considered in the past.

The invention discloses a bridge limiting device. A pressure-bearing elastic piece is installed between an upper anchoring device and an steering device of the bridge limiting device. The upper anchoring device, the steering device and a lower anchoring device are provided with tension cables for connection; the deformation range of a support is within the allowable deformation range; the deformation of the support is mainly adapted to pressed deformation of a pressure-bearing elastic piece; when an allowable deformation range is exceeded, that is, the upper anchoring device is in contact withthe steering device, the upper anchoring device adapts to the steering device through deformation of the tension cable, double guarantees are provided for restraining horizontal and vertical relativedisplacement between a beam body and a pier stud, overlarge horizontal sliding of the beam body is limited, and vertical disengaging of the beam body is also prevented; a pre-warning device is connected with the upper anchoring device and the steering device through wires, when the upper anchoring device makes contact with the steering device, a loop is formed, the pre-warning device generates warning information in real time, a bridge structure is effectively prevented from being further damaged, and the pre-warning device is used in the field of bridge engineering.

The invention discloses a bridge pavement icing detection and early warning system and method, and relates to a bridge pavement icing detection and early warning system and method. The invention aimsto solve the problems that dark ice cannot be accurately detected and icing early warning cannot be performed in real time due to the fact that existing detection equipment detects icing by means of icing external characteristics. The method comprises the steps that 1, an atmospheric temperature and humidity acquisition module acquires air temperature data and air humidity data, and a pavement temperature acquisition module acquires pavement temperature data; 2, the data obtained in the step 1 is processed to obtain a dew point temperature; 3, the analysis module performs second preset condition judgment on the current dew point temperature, and if a second preset condition is met, the analysis module issues an early warning command; if the second preset condition is not met, the analysismodule does not issue the early warning command; and 4, the early warning module performs early warning on the pavement after receiving the early warning command issued by the analysis module. The invention is applied to the field of traffic safety.

The invention relates to the field of personnel identification, and particularly discloses a lost personnel identification method, system and device and a storage medium, through adoption of the lost personnel identification method, intelligent cameras are used for deployment and control in various public places where personnel are prone to being lost, and the personnel identification efficiency is improved. Four technologies of a face recognition algorithm, an abnormal mental person recognition algorithm, an old person / child recognition algorithm and a missing judgment algorithm based on social distance detection are combined, missing persons and suspicious persons are automatically recognized, and an alarm is automatically given out. According to the method, the defects that the labor cost is high, the real-time performance is poor, misjudgment and missed judgment are easily caused and early warning cannot be carried out aiming at the condition that the person is lost in a place when the lost person looks up a video in a complex scene at present are overcome; and video monitoring is not limited by manpower and material resources any more, and the recall ratio can be improved.

The invention discloses a ground clearance monitoring system for crushing blades of a cassava stem crushing and field returning machine. The system comprises a profiling detection mechanism for detecting ground fluctuation, a hydraulic actuating mechanism for controlling the blades to lift, and a microprocessor in signal connection with the profiling detection mechanism and the hydraulic actuatingmechanism. The invention further discloses a cassava stem crushing and field returning machine monitoring system loaded with the system, and the monitoring system further comprises a vehicle-mountedterminal and a satellite positioning device. The informatization and the automation level of the cassava stem crushing and field returning machine are effectively improved, the problem that the cassava crushing and field returning effect is affected due to the fact that the ground clearance of the cassava stem crushing and field returning machine cannot be automatically adjusted along with the uneven ground surface due to topographic changes is solved, the operation efficiency of the cassava stem crushing and field returning machine is improved, and technical support is provided for supervision of operation efficiency.

The invention discloses a driver posture detection method based on a video and a skin color area distance. The method comprises the following steps: extracting skin color areas of sampling images in aplurality of sample videos, calculating mass center coordinates of the skin color areas, converting the mass center coordinates into feature distances to represent feature values of each image, and fusing the feature values of a plurality of images corresponding to a section of video into one feature value by adopting a clustering algorithm; constructing a BP neural network, inputting the fused characteristic values and the corresponding driving posture categories as training samples into the BP neural network, and performing training to obtain a driver posture detection model; during detection, a to-be-detected video during driving of a driver is collected, a feature value of the to-be-detected video is calculated according to the method in the above steps, a calculation result serves asinput of a driver posture detection model, and the calculation result is output as the driving posture category of the to-be-detected video. The method can effectively improve the detection rate of the postures of the driver, achieves the recognition and classification of the driving behaviors of the driver, and finally achieves the real-time early warning of the operation driving process.

The invention discloses a non-intrusive electric bicycle monitoring method based on model self-learning. The method comprises the following steps of carrying out preprocessing such as filtering and frequency reduction on the total power of a user; performing density clustering on the difference of the active power to obtain a mean value of a maximum density class and a corresponding reactive power mean value in time so as to reconstruct a signal, and screening out a time period containing a charging gentle slope by adopting bilateral filtering, state conversion removal and piecewise linear representation; finding out all suspected charging load events according to a non-intrusive load event detection algorithm, determining all load events caused by charging of the electric bicycle by adopting a sliding window, and completing model self-learning; a non-intrusive load event detection algorithm detecting a charging behavior in real time and performs early warning; on the basis of the charging model, the real-time sensing of the charge state and the calculation of the charging quantity being realized. Whether an electric bicycle charging behavior exists or not can be judged according to the total power of a user, model self-learning is completed, and charging behavior real-time early warning, charge state real-time sensing and charging electric quantity calculation are achieved.

The invention discloses a vehicle track prediction method and device. The method comprises the following steps: when a vehicle runs, triggering a prediction program used for predicting a driving track of the vehicle; map feature information of the high-precision map is called and loaded to a memory, and the map feature information is static data; responding to the prediction instruction, and capturing dynamic sensing information in the running process of the vehicle; based on the positioning information of the vehicle, map feature information of an area where the vehicle is currently located is inquired from a memory; and inputting the inquired map feature information and the dynamic perception information into the prediction model, and predicting to obtain a predicted trajectory of the vehicle. The technical problem that the vehicle track cannot be predicted in real time in the prior art is solved.

The invention provides a three-dimensional process visual design method based on a webgl technology, and the method comprises the steps: carrying out a three-dimensional visual scene building step and a three-dimensional visual scene interaction step through a 3D cloud platform based on the webgl technology, carrying out the data connection of a three-dimensional visual scene, a database, and an on-site monitoring system, presetting a plurality of scene models, mounting functional scripts of the corresponding scene models, issuing a three-dimensional visual scene to a three-dimensional human-computer interaction interface, and performing secondary disassembling on the data obtained by classifying and disassembling field data by the 3D cloud platform to obtain the three-dimensional visual scene, and loading and displaying the three-dimensional visual scene; and driving a scene model to perform synchronous field state action according to the field data. The method is convenient to operate and use, real-time data three-dimensional visual monitoring in the whole process is achieved, fault data retrieval and tracing are facilitated when storage exception occurs, and the fault debugging rate and the production efficiency can be improved.

The invention provides an image recognition-based early warning method for a safe area of ​​a power distribution room, comprising the following steps: collecting an image sequence of the safe area of ​​a power distribution room through an image acquisition device to obtain an image sequence to be identified. A processed image sequence is obtained by performing image median filter processing on the image sequence to be recognized. Extract the ROI area image of the primary processing image sequence to obtain the secondary processing image sequence. The feature points of the secondary processing image sequence are extracted, and the feature points are matched by the pyramid optical flow method to obtain the optical flow field of the secondary processing image sequence. Perform threshold segmentation on the optical flow field, and judge whether there is a moving object in the safe area of ​​the power distribution room through the threshold segmentation result. If there is no moving object, the judgment will end, and if there is, an early warning will be given. Realize unmanned monitoring and real-time early warning of the safe area of ​​the power distribution room.

The invention relates to a water quality monitoring system based on a cloud platform. The water quality monitoring system comprises a sampling terminal and the cloud platform; the sampling terminal comprises an immersion type probe, a light source, a spectrograph and an upper computer, and the immersion type probe comprises a slit for accommodating a water sample, a reflecting mirror and an optical fiber interface; during working, ultraviolet visible light generated by the light source passes through the water sample in the slit of the probe through an optical fiber, is reflected by the reflecting mirror and then passes through the water sample again, and is transmitted to the spectrograph through the optical fiber to obtain a water sample spectrum; and a water quality index and harmful substance spectrum prediction model is established on the upper computer. Spectral data of multiple sampling points are uploaded to a unified cloud platform through the upper computer, the data of the multiple sampling points are processed on the cloud platform according to the data migration theory, and correction of parameters of a water quality prediction model of the upper computer is achieved. According to the invention, the model is corrected by using the multi-sampling-point data on the cloud platform based on the data migration principle, and the prediction model which is more accurate than that based on single-sampling-point data is provided by using the big data advantage on the cloud platform.