40results about How to "Reliable early warning" patented technology

The invention belongs to the field of video detecting, and provides a video-based safety precaution method for vehicle merging from a highway ramp. The method includes arranging two cameras in the same directions as two ways on highway entrances to detect moving vehicles, marking vehicle trajectories, acquiring actual moving distance through calculation according to the vehicle trajectories in a continuous tracking frame, acquiring actual speed of the vehicles, acquiring time difference of vehicle entering the merging areas between the two ways through the speeds of vehicles on the two ways, and judging possibility of vehicle collisions through the time difference. Compared with the prior art, the video-based safety precaution method is not limited by environments, capable of performing real-time and reliable precaution for all vehicles in the video areas, easy to realize, high in accuracy, applicable to real-time traffic precaution for highway ramps, and wide in application prospect.

The invention discloses an active anti-collision system based on an electric automobile and a control method thereof. The active anti-collision system comprises a man-machine interface unit, an information collecting unit, a central processing unit and an executing unit. An electronic control hydraulic brake technology is adopted in the system, a fusion algorithm is adopted to recognize the target according to laser radar and visual sensor signals, safety early-warning distance and safety brake distance are calculated based on road adhesion coefficients, and finally compound decision control execution unit is used for early-warning or active brake. The system is capable of greatly improving the reliability of automobile collision prevention and effectively promoting traveling safety.

The invention provides a vehicle fatigue driving warning method which is characterized by comprising the following steps: step A, an initial image is acquired, features are extracted, models are established, and timing is started; step B, an image of a driver is acquired; step C, when the driver meets the preset facial feature determining conditions and comparison between the current features and the feature models shows that the driver is not the registered one, the facial feature model and the fatigue feature model are reestablished, timing is restarted, and the method goes to step B, and when comparison between the current features and the feature models shows that the driver is the registered one or the driver does not meet the preset facial feature determining conditions, the method goes to step D; and step D, the current fatigue features are extracted, and alarm information is sent when the driving is judged to be fatigue driving according to the current fatigue features and the fatigue feature model, wherein alarm information is sent when the preset timing period is up. The advantage lies in that fatigue driving of drivers is judged comprehensively and fatigue driving warning is carried out more efficiently and reliably so as to ensure the safety of vehicle driving.

The invention discloses an overflow water level warning method for glacial lake outburst, which comprises the following steps: (a) the thickness T of the top of an ice trace dam, the width b of the outburst mouth, the average width w0 of a glacial lake, the length-to-width ratio c of the glacial lake, the average slope Beta of the shore slope of the glacial lake before outburst, and the characteristic particle size d of incipient sediment are surveyed; (b) the critical overflow water depth H of glacial lake outburst is calculated; and (c) the water level of the glacial lake above the overflowmouth rises suddenly for more than 30s, and when HCr<0.6H, there is very small probability that the glacial lake bursts; when 0.6H<=HCr<0.8H, there is small probability that the glacial lake bursts; when 0.8H<=HCr<H, there is medium probability that the glacial lake bursts; and when the HCr>=H, there is large probability that the glacial lake bursts. The method fully considers the factors influencing the critical overflow water depth of glacial lake outburst. The calculated result is more in line with the actual situation. The accuracy of warning is high. The method is of higher applicability.

The invention belongs to the technical field of medical appliances, and relates to an ultrasonic puncturing navigation method, particularly to an ultrasonic puncturing navigation method based on electromagnetic location. The navigation method orderly comprises the following steps: A, tracking and collecting position information of a probe and a puncturing needle through a puncturing location system under the control of a navigation software subsystem, wherein the puncturing location system refers to an electromagnetic puncturing location system; B, transmitting the collected information to an ARM (Advanced RISC Machines) processor in the navigation software subsystem through the electromagnetic puncturing location system, and performing data correction and post treatment; and C, transmitting the information after being corrected and post-processed to an ultrasonic imaging equipment through the navigation software subsystem to realize visual display. The ultrasonic puncturing navigation method based on the electromagnetic location has the advantages of exact location, small volume and light weight; the method is nonradiative and is slightly affected by environmental factor; the puncturing needle path is displayed in real time; and the needle tip position can be clearly observed at random angle, the puncturing accuracy is improved, the hemorrhagic complication is reduced, and the safety is improved.

The invention discloses a vehicle driving behavior detection and early warning system and method based on deep learning. The system includes an information collection module, a transmission module, acontrol processing module, and an early warning module; and the method includes the following steps: firstly collecting signal source information by the information collection module; then transmitting the collected signal source information to the control processing module through the transmission module; then judging a fatigue state and behavior of a to-be-detected object by the control processing module according to the signal source information; and transmitting a obtained result of the fatigue state and behavior to the early warning module and a cloud, and performing corresponding sound and light and vibration early warning by the early warning module according to the result of the fatigue state and behavior. The system and the method combine an image recognition technology of the deep learning, a physiological signal detection algorithm and a driving behavior detection algorithm to perform multi-signal source collection, and the collection accuracy is high; and determination of fatigue and behaviors of the to-be-detected object is performed by utilizing a machine learning-based data fusion algorithm, an obtained judgment result is more accurate, and early warning is more effective and reliable.

A pre-warning method and mechanism for opening the door of car features that before the door is going to be opened, the side lamp is flushing and the pre-warning sound is generated from the horn of car, and when the openness of door reaches 2 / 5-3 / 5 of the most openness, the flash light and pre-warning sound is stopped. Its control switch can be arranged on the handle of door.

The invention relates to the technical field of intelligent control and provides a monocular-vision-based anti-collision early warning system for rail transit, which comprises an image pick-up device, a controller and an alarm; the image pick-up device is mounted at the front end of a vehicle; the signal output end of the image pick-up device is electrically connected with the signal input end of the controller; and the signal output end of the controller is electrically connected with the signal input end of an alarm device. According to the monocular-vision-based anti-collision early warning system for rail transit, analysis and early warning can be performed to rail transit through machine vision, so that the stability is high, and the system can serve as an auxiliary system for the conventional anti-collision early warning system for rail transit; and in the further technical scheme, an ultrasonic range finder is used, so that early warning of the system can be more accurate and more reliable.

The invention discloses a security transaction risk early warning method and system based on instruction stream toxicity indexes; the method comprises the following steps: obtaining transaction information in real time from a security transaction server; continuously storing the obtained real time security transaction information into a plurality of transaction information sets according to the transaction time, wherein each transaction information set contains the total transaction amount equals to a preset transaction amount threshold; obtaining equal transaction amount transaction probability of each transaction information set according to the ratio between the absolute value of buying-selling amount difference of each transaction information set and the total transaction amount contained by the corresponding transaction information set; using the present transaction time as the boundary, solving the average equal transaction amount transaction probability of n continuous transaction information sets most close to the present transaction time before the present transaction time, and using the average equal transaction amount transaction probability as the present transaction time stream toxicity index, wherein n refers to a positive integer; comparing the present transaction time stream toxicity index with a preset stream toxicity index threshold scope, and outputting a risk early warning signal if the present transaction time stream toxicity index is in the preset stream toxicity index threshold scope.

The invention provides a posture information acquisition device. The posture information acquisition device includes an ECG component, and a posture sensing unit which is fixedly arranged on the ECG component; the posture sensing unit senses the direction of the ECG component and outputs the direction information to a client which communicates with the posture information acquisition device; and the ECG component includes two or three electrodes.

The invention provides an egress and ingress early-warning method and device. Egress and ingress time of followers who get in and out of a residential area and target people followed by the followersare analyzed and whether to carry out early-warning or not is determined according to the total number of the followers corresponding to all target people with the same residential address information, so reliable early-warning is realized, and the condition that an illegal event exists due to the fact that a corresponding residual address is likely to be a pyramid selling den or a gathering den can be effectively prevented.

The invention discloses a cerebral stroke warning genetic circuit model which comprises a joint detection circuit module and a feedback circuit module. Output of the joint detection circuit module is input of the feedback circuit module, output of the feedback circuit module acts on the output of the joint detection circuit module via internal feedback branch circuits, and in other words, the feedback circuit module is a next-level circuit of the joint detection circuit module. The cerebral stroke warning genetic circuit model has the advantages that the cerebral stroke warning genetic circuit model is high in structural universality, cerebral stroke can be quickly monitored by the simple and effective joint detection circuit module, interference due to uncertain factors such as environments can be reduced by the introduced feedback circuit module, and accordingly the cerebral stroke can be accurately and reliably warned by systems.

The invention relates to a fine automatic prediction method and a system for atmospheric horizontal visibility with a high spatial resolution. The fine automatic prediction method comprises steps of comprehensively analyzing and processing atmospheric numerical model data and observation data at an automatic meteorological station, establishing a large data set with the principle of minimum spatial horizontal distance; inputting the data set into a deep neural network framework, processing with a plurality of hidden layers to obtain a nonlinear relationship between a predictor and a tag , continuously adjusting a weight by feedback learning of a deep neural network neuron to train a visibility model that can distinguish overall characteristics of the predictor; and then cooperating with aninterpolation method to obtain a fine prediction of visibility with a high spatial resolution. The fine automatic prediction method effectively solves the problem of poor accuracy of predicting the visibility, lower spatial resolution and lower automation level in the prior art, thus provides a reliable warning for local low-visibility disasters occurred at sea, inland, and ports, and is expectedto become a powerful tool for visibility prediction in the meteorological station.

The invention provides a real-time accumulated rainfall rainstorm mountain torrent early warning method based on mountainous area small watershed rainfall runoff response. The method comprises the following steps: (1) selecting a small watershed river reach of a mountainous area needing rainstorm mountain torrent early warning as a target river reach, and searching flood characteristic data of n-stage rainstorm mountain torrent of the target river reach and rainfall characteristic data of stage rainstorm corresponding to the n-stage rainstorm mountain torrent; Determining a flood abrupt rise time point of each field rainstorm mountain flood according to the flood characteristic data; (2) analyzing the n times of rainstorm mountain torrent; (3) determining a critical cumulative rainfall index for sending out rainstorm mountain torrent early warning; (4) 1, when rainstorm occurs at the upstream of the target river reach, collecting time period rainfall data of field rainstorm, and calculating accumulated rainfall from the beginning of rainfall to different time period rainfall data collection time points in real time; And 2, when the accumulated rainfall capacity in the step (1) reaches a critical accumulated rainfall capacity index for giving out rainstorm mountain torrent early warning, immediately giving out the rainstorm mountain torrent early warning.

The invention discloses a power system operation abnormal state estimation method which comprises the steps of creating a historical case library for expert diagnosis, wherein the historical case library is divided into a voltage level, a substation level and an interval level, and each level is provided with a plurality of level nodes; respectively filling the data content of each historical caseinto a hierarchical node corresponding to each level; in the historical case library, carrying out big data analysis on the operation state value of each hierarchical node before the disturbance of each type occurs, and calculating a disturbance early warning threshold value of each operation state value; acquiring panoramic information of a user power grid in real time; and when a certain running state value of the panoramic information exceeds the disturbance early warning threshold corresponding to the running state value, sending out disturbance early warning. According to the estimationmethod, the estimation of disturbance can be provided for the user, the user is reminded to take measures in time, and disturbance is prevented.

The invention provides a hydraulic turbine governor servomotor action rate deviation degree early warning method and system. The hydraulic turbine governor servomotor action rate deviation degree early warning method comprises the steps of collecting the early data of an governor in a healthy state, obtaining the travel of a servomotor through conversion according to the guide vane opening degree,and obtaining action rate data of the servomotor ; establishing a servomotor action rate standard model based on a deep artificial neural network through servomotor action rates under different guidevane opening degrees, different pressures of a pressure oil tank, different positions of a pressure distributing valve and different upstream and downstream water levels; substituting the acquired working condition data into the servomotor action rate standard model to obtain a standard value of the action rate under the working condition; obtaining the servomotor rate deviation degree through the ratio of the standard value to the actually measured servomotor rate; and conducting prediction on the obtained servomotor action rate deviation degree sequence to obtain a future change trend, andgenerating early warning if the future deviation degree exceeds a set threshold value. The precision of the action rate deviation degree prediction result is high, and reliable early warning can be provided.

The invention relates to an online water quality monitoring device and method. The device comprises an integrated water quality detector which is provided with a water inlet pipe and a water outlet pipe; the water inlet pipe is divided into three branches, namely a municipal water source inlet pipe, a water tank water source inlet pipe and a standard water source inlet pipe; electromagnetic valvesare arranged on the municipal water source inlet pipe, the water tank water source inlet pipe and the standard water source inlet pipe; and the electromagnetic valves on the municipal water source inlet pipe, the water tank water source inlet pipe and the standard water source inlet pipe are electrically connected to a controller through wires. The device and the method are flexible, timely and effective; the turbidity of a water source of a water tank is monitored in real time, and comprehensive water quality parameter detection analysis is started once when turbidity early warning exists; and even if there is no turbidity early warning, comprehensive water quality parameter detection can be carried out periodically. The controller provided by the invention regularly performs online calibration on a water quality detection instrument, so that the accuracy and reliability of data acquisition, comparative analysis and early warning alarm are ensured.

The invention discloses a multifunctional intelligent live broadcast system for scenic spots, which comprises a management server, and further comprises: a video acquisition module for acquiring video data by using a video acquisition device of each scenic spot and uploading the video data to the management server; a passenger flow acquisition module which is used for acquiring people number distribution data of each scenic spot and people flow distribution data between two adjacent scenic spots by using a passenger flow statistical system, and uploading the data to the management server; a weather information acquisition module which is used for acquiring weather data of the scenic area and uploading the weather data to the management server; a background sound effect acquisition module which is used for acquiring a required background sound effect and uploading the background sound effect to the management server; and a user terminal which is used for acquiring and displaying the video data, the people number distribution data, the people stream distribution data and the weather data from the management server, and acquiring and playing the required background sound effect from the management server. According to the invention, the functionality of the scenic spot live broadcast system is improved, more information is provided for tourists and scenic spot managers, and the tourists and the managers can be linked.

A method for detecting an activity of an object disposed within a medium at a depth ranging from about 0 to about 100 cm using a radar system, the method including establishing a baseline radar power level of the object in the medium; and detecting one or more radar data anomalies in radar data received of the medium with respect to the baseline radar power level, wherein a presence of the one or more anomalies indicates a presence of the activity of the object.

The invention provides a method and a system for early warning of the deviation of the action rate of a hydraulic turbine governor relay, comprising: collecting data in the early stage of the health state of the governor, and converting the relay stroke according to the guide vane opening, and then obtaining the relay action Rate data; through different guide vane openings, pressure tank pressure, pressure distribution valve position, and the action rate of the relay under the upstream and downstream water levels, a standard model of the action rate of the relay is established based on the deep artificial neural network; the collected working condition data is substituted into In the standard model of the relay action rate, the standard value of the action rate under this working condition is obtained; the ratio of the standard value to the measured relay speed is used to obtain the relay speed deviation; Change trend, if the future deviation exceeds the set threshold, an early warning will be generated. In the present invention, the prediction result of the deviation degree of the action rate has high precision and can provide a reliable early warning.

The invention discloses a comprehensive online monitoring device and method for a switch cabinet. The switch cabinet comprehensive online monitoring device comprises an intelligent early warning mechanism, wherein a base is fixedly installed at the bottom part of the intelligent early warning mechanism, a top plate is fixedly installed at the top part of the base through an installation column, and the front end face of the intelligent early warning mechanism is provided with a traveling wave signal collector, a fault point range finder and a partial discharge signal collector. According to the comprehensive online monitoring device and the method, when in use, a fastener between mounting piece on the outer surface walls of a plate IV and a plate I is loosened, and the operation of a maininterface of the intelligent early warning mechanism can be carried out; and a plate II, a plate III and the plate IV can be wrenched in sequence from the left side to separate cards on the inner surface walls from the installation column, at the moment, a cover closing mechanism composed of the plate I, the plate II, the plate III and the plate IV can be separated from the intelligent early warning mechanism and integrated with the intelligent early warning mechanism, an operator can operate and overhaul the device from a proper position, and therefore the using effect of the comprehensive online monitoring device for the switch cabinet is improved.

The invention provides a safety signal valve automatically discharging dust for a dust remover dust hopper in a severe dust accumulation state. The safety signal valve is installed on a dust discharging pipeline. The dust discharging pipeline is provided with a dust inlet pipe stretching in a hopper body and a dust discharging pipe stretching out of the hopper body. The hopper body is provided with a design bearing material level and a high-material-level warning material level, wherein the dust inlet pipe is located between the design bearing material level and the high-material-level warning material level. The safety signal valve comprises a detection opening flange fixedly arranged at a pipe opening of the dust discharging pipe, an elastic pressure applying assembly fixedly connected with the detection opening flange, and a door rod, wherein one end of the door rod is connected with a valve deck used for sealing the pipe opening, and the other end of the door rod is connected with the elastic pressure applying assembly. The safety signal valve can be matched with the dust discharging pipe to conveniently judge whether the dust remover dust hopper is in a normal state or an abnormal state; and in addition, automatic dust discharging can be conducted after the design dust storing bearing capability of the dust remover is exceeded, and meanwhile, sound-light alarming of the dust remover in the accidental dust accumulation state is achieved.

The invention provides an adaptive active anti-collision braking system with independent training and learning. It is mainly composed of a millimeter-wave radar sensor, a monocular optical sensor, a brake motor controller, a precise stepping brake motor, and a Beidou high-precision system. The combined navigation communication terminal and display terminal are composed of six parts; the present invention can solve the problem that the existing AEBS control parameters are single, which cannot meet the load change of commercial trucks, the change of adhesion coefficient of commercial vehicles in rainy and snowy weather, and the long-term wear and aging of the vehicle brake system. Single unchanging disadvantages. Provide drivers with more comfortable, safe and reliable early warning and emergency braking experience.

A high-pressure gas well wellhead sealing simulation device. It includes a power system, a feedback control unit and a gas leak detection unit. The power system includes a motor, a crankshaft connecting rod mechanism, a simulated pipe string and a blowout preventer rubber cartridge; the gas leak detection unit includes a gas storage tank, a supercharger, A gas detection sensor and an alarm; the feedback control unit includes a gas detection sensor, a controller for controlling the loading of a hydraulic pump, and a hydraulic pump for providing a loading for the blowout preventer. The device can not only simulate the wellhead sealing of high-pressure gas wells, but also provide early warning for gas leakage, and realize adaptive adjustment after gas leakage through detection and feedback. experimental research in the case.

The invention discloses a cerebral stroke warning genetic circuit model which comprises a joint detection circuit module and a feedback circuit module. Output of the joint detection circuit module is input of the feedback circuit module, output of the feedback circuit module acts on the output of the joint detection circuit module via internal feedback branch circuits, and in other words, the feedback circuit module is a next-level circuit of the joint detection circuit module. The cerebral stroke warning genetic circuit model has the advantages that the cerebral stroke warning genetic circuit model is high in structural universality, cerebral stroke can be quickly monitored by the simple and effective joint detection circuit module, interference due to uncertain factors such as environments can be reduced by the introduced feedback circuit module, and accordingly the cerebral stroke can be accurately and reliably warned by systems.

The invention relates to a power system online prewarning method based on microdisturbance signal low-frequency oscillation mode identification, which comprises the following steps of: carrying out measuring point low-frequency oscillation mode analysis and calculation on a single data window of a single measuring point by adopting a slide average autoregression model; calculating and obtaining acalculating result of a plurality of data windows of a plurality of measuring points by a cluster analysis tool and storing a total-network range low-frequency oscillation mode oscillation frequency and damping ratio result; and then judging whether prewarning information needs to be transmitted or not according to the prewarning threshold value condition and realizing matching with a large disturbance monitoring alarm way. The method of the invention can fully excavate oscillation mode identification contained in a plurality of actual measurement microdisturbance signals when a power system is run and provide online prewarning for system running. Meanwhile, the method of the invention realizes continuous monitoring on the low-frequency oscillation mode of the power system; and combined with a statistic analysis method, the method can be used for disclosing the low-frequency oscillation mechanism and influence factor of a large-scale practical power system.