30results about How to "Shorten the warning time" patented technology

The invention provides a YOLOV3-based smoke and fire automatic detection and early warning method. The method comprises the steps of S1, constructing a sample set of a training detection model; S2, establishing a deep learning target detection network architecture based on YOLOV3; S3, configuring training parameters, and training a detection model; S4, acquiring to-be-detected image information; acquiring image frames of an on-site video picture from monitoring equipment on a to-be-detected site, and processing frame-by-frame images by utilizing an image preprocessing method; S5, detecting smoke and flame targets, and sending the video image frames processed in the step S4 into a pre-trained detection model in the step S3 for target detection, and outputting a detection result; S6, carrying out post-processing on a detection result; and S7, continuously analyzing the detection results of the multiple frames of images, confirming that the target is valid and outputting an alarm. According to the YOLOV3-based smoke and fire automatic detection and early warning method, second-level detection and alarm can be realized, the fire early warning time is greatly shortened, timely notification and timely rescue are realized, and fire spreading is effectively prevented.

A rapid early-warning system based on a phase-sensitive optical time-domain reflectometer includes a disturbance monitoring and positioning module, and a disturbance signal identifying module. The disturbance monitoring and positioning module: a disturbance monitoring part performs difference on the beat signal intensity of a phase-sensitive optical time-domain reflectometer, the rapid early-warning system based on the phase-sensitive optical time-domain reflectometer can monitor the difference of the beat signal intensity all the time, and once the difference of the beat signal intensity changes, it means that a disturbance signal generated by disturbance occurs; and a positioning module can determine position information of a disturbance point through the time difference of an optical time-domain reflectometer echo. The disturbance signal identifying module includes a pre-treatment module, a feature extraction module, a template training module, a pattern classification module, and a post-treatment module; template training means that templates of a plurality of types of signals can be constructed according to features of a series of different types of disturbance signals; the features of the disturbance signals and preset templates of the plurality of types of signals are matched, and the types of the signals can be determined according the matching degrees; and a warning measure or a no-warning measure can be preformed according to a result of pattern classification.

The invention relates to a method and device for warning power-fail of base station by interrupting. After a power module tests the power-fail of base station, a signal of warning power-fail is outputted to a programmable logic device; the programmable logic device generates an interrupt source signal according to the signal of warning power-fail, and writes a value that stands for the signal of warning power-fail in a memory module; the interrupt source signal and another interrupt source signal carry out AND operation, and the result of the AND operation is transmitted to a processor to trigger the processor to interrupt; after the processor is triggered to interrupt, the memory module is inquired; after the value that stands for the signal of warning power-fail is inquired, the information of warning power-fail is transmitted to a controller of the base station. The invention warns power-fail of base station by interrupting; while improving the reliability, the application of the method of interrupting the processor with other interrupt source signal can save the interrupt resource of the processor, and can conveniently implement the function of warning power-fail of base station in the application that the interrupt resource of the processor is in stress.

A road surface division mark recognition apparatus includes: a vehicle-mounted camera that takes an image of a road surface ahead of a vehicle; an image processing portion which has a plurality of image processing modes that correspond respectively to a plurality of kinds of road surface division marks, and which recognizes a road surface division mark in a selected image processing mode; a temperature measurement portion that measures the temperature of the vehicle-mounted camera; and a restriction portion that restricts the action of the vehicle-mounted camera if the temperature measured by the temperature measurement portion is higher than or equal to a threshold value. The threshold value differs between the image processing modes.

The invention discloses a satellite navigational alarming device of vehicle accident, which comprises the following parts: vehicle locating navigational device, alarming signal transmitting device, receiving device at relative salvage department to receive the alarming signal and locating navigational signal, triggering switch cascaded on the alarming signal transmitting device, wherein the alarming signal transmitting device sends wireless signal to the satellite to transmit the wireless signal salvage department, which makes the relative department to receive the salvage alarming signal. The invention can save the salvage time and life and property loss due to delay, which makes the injured person survived in time.

The invention discloses a gas detection alarm which comprises an upper cover, a lower shell, a sensor seat, a sensor, an alarm connecting piece, an alarm and a processing circuit; threaded holes in corresponding positions are arranged around the upper cover and the lower shell; the upper cover and the lower shell are fastened into an explosion-proof outer shell through bolts in the threaded holes; the side surface of the lower shell is respectively provided with the sensor seat and the alarm connecting piece which are communicated with the explosion-proof outer shell; the sensor is arranged in the sensor seat; the sensor seat is connected to a sensor breathing apparatus; an explosion-proof piece and a breathing port are arranged in the sensor breathing apparatus; the alarm connecting piece is connected to the alarm; the explosion-proof piece is arranged in the alarm; and the processing circuit is electrically connected between the sensor and the alarm, and is used for receiving the electric signals of the sensor and triggering the alarm.

The invention provides a disease early warning method and system, relates to the technical field of the Internet of Things, and regards an Internet of Things server as an execution main body. The method includes steps: receiving disease description information transmitted by an acquisition terminal; processing the disease description information, and generating disease information; assessing the disease information according to a preset rule, and generating disease early warning information; and performing format conversion on the disease early warning information and transmitting the converted information to a user terminal to enable the user terminal to display the disease early warning information. According to the method and system, the Internet of Things server is adopted so that thedisease early warning time is shortened, and the disease prevention and control range is extended.

The invention discloses a fluorescence spectrum technology-based dissolved organic matter (DOM) water quality monitoring and early warning method, and belongs to the technical field of water environment monitoring. The fluorescence spectrum technology-based dissolved organic matter water quality monitoring and early warning method is a surface water organic pollution rapid early warning method based on three-dimensional fluorescence spectrum, and is used for solving problems that rapid early warning method of dissolved organic matter (DOM) water quality is low in efficiency, treatment is time-consuming, and treatment period is long. The fluorescence spectrum technology-based dissolved organic matter water quality monitoring and early warning method is high in operation efficiency, and short in consumed period. The fluorescence spectrum technology-based dissolved organic matter (DOM) water quality monitoring and early warning method is unique; efficiency is increased by 30 to 40% compared with that of the prior art, and early warning time is reduced by 50 to 60%.

The invention discloses a road target rapid detection and early warning method and system, and belongs to the field of computer vision. The road target rapid detection and early warning method comprises the steps: employing an image collection device to collect road scene images, and extracting N levels of regions of interest from the road scene images according to a target distance from near to far; and sequentially carrying out target detection on the N levels of interested regions according to the sequence of the target distances from near to far, carrying out early warning after a certainlevel of interested region with a relatively short target distance detects that a target exists, and not detecting the subsequent level of interested region, wherein N is greater than or equal to 2, and the target distance is the distance between the target and the image collection device. Compared with an existing detection and early warning algorithm for targets such as pedestrians and vehicles,the road target rapid detection and early warning method has the advantages that the targets appearing on the driving route can be detected and early warned more quickly and accurately, and the practical value is high.

The invention provides a GPU (graphics processing unit)-based radar signal intra-pulse characteristic real-time analysis realizing scheme which mainly comprises a data acquisition module, a GPU signal analysis module and a CPU (central processing unit) signal analysis module, wherein the data acquisition module is used for receiving digital bit stream from a digitalizer, and dividing the data into a plurality of short-time sequences to be subsequently processed; the GPU signal analysis module is used for analyzing and processing each short-time sequence datum, and outputting the datum obtained in a processing way, namely the frequency diversity characteristic, the frequency, the frequency root-mean-square error, the phase jump characteristic and the like; and the CPU signal analysis module is used for analyzing the GPU output data, counting diversity proportion in each short-time sequence belonging to the same impulse, carrying out the least square fit on an instantaneous frequency value in each short-time sequence to obtain a frequency modulation gradient and fit the root-mean-square error, counting the instantaneous frequency mean root-mean-square error and the effective phase jump, and judging the intra-pulse modulation types according to each characteristic parameter. The GPU-based radar signal intra-pulse characteristic real-time analysis realizing scheme is better in processing instantaneity, and further has the advantages of being simple in structure, high in expansibility, easy to maintain, and capable of being upgraded.

The invention discloses a device and method for predicting coal rock burst through hydrogen bond rupture, and the device comprises a structural body extending into a coal rock drill hole, and a prediction analysis module, wherein the structural body is internally provided with a voltage signal collection module which is used for collecting transient voltage values of the inner wall of the coal rock drill hole at different depths; a temperature acquisition module which is used for acquiring temperature changes in the coal rock drill holes at different depths; a signal transmitting module whichis used for transmitting the acquired information to the prediction analysis module; and a prediction analysis module which is connected with the signal transmitting module and is used for comparing and analyzing the received signals and data acquired by the coal rock fracture experiment. According to the invention, transient voltage values generated by the inner wall during coal rock drilling fracture at different depths are collected; the collected data is transmitted to the computer in real time according to the temperature change of the inner wall of the coal rock drill hole at different depths, the computer compares and analyzes the collected data with laboratory simulated fracture experiment data, and therefore the coal rock burst danger level and strength are predicted.

The invention discloses a method and a device for constructing a river diatom water bloom early warning model. The method comprises the steps of: carrying out analysis according to river hydrologicalelement data, water quality element data and algae density to obtain key impact factors of a river diatom water bloom early warning model; performing fusion according to a key influence factor data set to obtain a sample set of a support vector machine; performing optimization calculation on parameters of the support vector machine by using a genetic algorithm to obtain an optimal solution of theparameters of the support vector machine; and constructing the river diatom water bloom early warning model according to the sample set of the support vector machine and the optimal solution, and carrying out water bloom early warning. By implementing the method, a support vector machine algorithm is applied, so that the operation is more efficient, the early warning timeliness is improved, meanwhile, the adopted genetic algorithm can consider the dynamics of population evolution, the optimal solution can be quickly and accurately found, the model efficiency and the calculation precision are improved, and the method has good practicability.

The invention provides a side slope dangerous rock early warning method, device and equipment and a readable storage medium, and relates to the technical field of side slope dangerous rock early warning. A rock model set and a side slope virtual analysis model which conform to the actual engineering situation are established in advance, and then the two models are combined to generate a virtual side slope model; an overload method and a strength reduction method are adopted to enable the natural slope model to be unstable, the stability coefficient of the rock block outside the slope is obtained through analysis and calculation, then a neural network model is utilized, the rock model of the block outside the slope and instability simulation data serve as input values, the stability coefficient serves as an output value, and a neural network is trained; and identifying other rocks on the slope by using the trained neural network to obtain a corresponding stability coefficient so as to carry out early warning.

The invention relates to a virtualization escape defense method based on a kernel debugger, which mainly comprises the following steps: command reading: adding a virtual machine process into a Windbgkernel debugger, and reading an RPCI command received by the process; analyzing the data packet: analyzing a suspicious instruction in the RPCI command, and judging whether the length of the data packet has overflow possibility or not; memory allocation protection: when overflow possibility exists, carrying out automatic judgment on an allocation mode of the target block, and realizing predictability of an address; and overflow defense: automatically judging an address relationship between an overflow point and an overflowed object, and carrying out early warning on an overflow attack behavior. By using the kernel debugger Windbg under the Windows operating system, suspicious overflow behaviors under the virtualization environment are automatically judged, and the heap layout is accuratelypredicted, virtual machine escape attacks can be effectively defended, more efficient and complete heap overflow prevention is achieved, and the safety of the virtualization environment is guaranteed.

The invention, which relates to the technical field of symptom monitoring, provides a symptom monitoring system and method. The method comprises: an information management module establishes basic user information according to a management operation; a user supervision module generates symptom information through a selection operation, manages symptom information according to basic user information, and generates symptom monitoring information; and a symptom early-warning module sets an early-warning rule, evaluates the symptom monitoring information according to the early-warning rule, and generates early-warning information. According to the invention, a monitoring starting point is moved forward to a symptom occurrence time, so that the early-warning time of the epidemic situation is shortened and the accuracy of the epidemic prevention and control is improved; and the operation difficulty is simplified by using the selection operation and the application range of the applicable population is extended.

The invention relates to a refinery fire identification method based on a convolutional neural network, and belongs to the technical field of fire video identification. The method comprises the following steps of: 1, detecting a running target through a background difference method, and judging whether a moving object exists in a current picture or not; 2, transmitting the moving target into a flame detection model, and outputting a flame probability and a flame region; 3, transmitting the suspicious area into a false alarm model, and outputting the probability of real flames; and 4, judging whether flame or flame false alarm exists according to the probability of the real flame. Through verification, the method has higher recognition accuracy and lower false alarm rate, and the model can be updated according to the actually running picture, so that the method is more suitable for the actual scene of the scene; through real-time monitoring of a control area, a rapid response can be made at the initial stage of a fire, and real-time analysis and post-processing are carried out, so that early forecasting and control of the fire can be realized, and the influence and loss caused by the fire can be reduced.

An industrial control system-oriented intrusion detection rule matching optimization method comprises the steps of arranging information such as IP, ports, vulnerabilities and security configuration of industrial control system assets, selecting a plurality of feature attributes from rule heads and rule options of matching rules, setting a clustering center to be 3, dividing rules in a rule base into three levels by adopting a mean value clustering algorithm, calculating threat values after rule matching succeeds according to the rule header feature data, the vulnerability score, the configuration vulnerability score and the threat score of the rule action, and sorting in the hierarchical cluster according to the threat values; and when intrusion detection is carried out on a data packet, after feature information of the data packet is extracted, matching the feature information with a rule according to grading and sorting. According to the method, the rules in the rule base can be graded and sorted according to real threats faced by the industrial control system, the rules corresponding to harassment attacks are matched and deduced, the problem of low detection efficiency caused by a large number of rules is solved, and the detection efficiency is improved.

The invention discloses the technical field of respiratory disease monitoring and discloses a wearable respiratory disease monitoring system and method based on Internet of Things. The system comprises a respiratory signal collection module, a blood pressure collection module, a cardioelectric collection module, a wireless communication module, a signal processing module, a main control module, adisease analysis module, a prewarning module, a cloud service module and a display module. Respiratory rate and respiratory intensity are calculated by adopting a 'time-frequency domain' approximationmethod through the signal processing module, so that the system has high resistance to movement interference; online real-time calculating of the respiratory rate and the respiratory intensity can berealized; the respiratory rate and the respiratory intensity can be acquired at the same time; respiratory disease prewarning information is converted in format and then sent to a user terminal through the prewarning module to enable the user terminal to display the respiratory disease prewarning information; a could server is adopted to shorten disease prewarning time and expanding disease prevention and control range.

The invention discloses a parachute-opening state monitoring method and system based on feature recognition of a resistance parachute. Firstly, it is judged whether the guide parachute has been sent out. Points are used as feature points, and the gradient of the total number of feature points is used as the threshold for judging whether the drag parachute is successfully opened. The invention judges the parachute opening situation of the drag parachute by measuring the change gradient of the total number of feature points on the drag parachute, and sets a corresponding threshold. When the change gradient of the total number of feature points is lower than a certain threshold, the voice radio station is automatically triggered. Provide voice prompts to realize the pilot's quick judgment and emergency treatment of the parachute opening status of the aircraft. Based on picking up the characteristic points of the parachute opening of the aircraft, the invention can quickly judge whether the parachute opening is successful, shortens the alarm time, improves the reaction time of the pilot and the flight test safety margin, and has good practicability.

The invention discloses a method and system for rapid detection and early warning of road targets, which belong to the field of computer vision. Area of ​​interest: according to the order of the target distance from near to far, target detection is performed on the N-level area of ​​interest in sequence. After a target is detected in a certain level of interest area with a relatively short distance to the target, an early warning will be issued, and subsequent levels will not be detected. The region of interest is detected; wherein, N≥2, the target distance is the distance between the target and the image acquisition device. Compared with the existing detection and early warning algorithms for pedestrians, vehicles and other targets, the present invention can more quickly and accurately detect and warn the targets appearing on the driving route, and has good practical value.

The invention discloses a fire source detection system and detection method, and the system comprises a fire source detection device which is disposed at a fire source detection point and is used fordetecting and outputting a fire source signal; and a signal analysis processor which is in communication connection with the fire source detection device and is used for judging whether a fire occursor not according to the received fire source signal and forming a fire judgment result. The fire source detection device composed of the infrared flame sensor or a plurality of different types of smoke detectors is adopted to detect the fire source, an alarm can be given at the first time when the fire source occurs, the early warning time when a fire occurs is greatly shortened, fire spreading can be effectively prevented, losses are reduced, and safety is ensured. After the system is in communication connection with external equipment, the fire source detection condition and the fire behavior judgment condition can be sent to the external equipment so that external personnel can know the fire scene condition at the first time.

The invention provides a slope dangerous rock early warning method, device, equipment and readable storage medium, and relates to the technical field of slope dangerous rock early warning. Then the two models are combined to generate a virtual slope model. The overload method and strength reduction method are used to make the natural slope model unstable. After analysis and calculation, the stability coefficient of the rock block outside the slope is obtained, and then the neural network model is used. The rock model of the block outside the slope and the instability simulation data are used as input values, and the stability coefficient is used as the output value to train the neural network, and the trained neural network is used to identify other rocks on the slope to obtain the corresponding stability coefficient, so as to carry out early warning.

The invention discloses a parachute opening state monitoring method based on drag parachute feature recognition and a system thereof, and the method comprises the steps: judging whether a guide parachute is emitted or not, judging whether the drag parachute is successfully opened or not if the guide parachute is emitted normally, and taking the intersection point of a drag parachute reinforcing band as a feature point; and using the change gradient of the total number of the feature points is used as a threshold value for judging whether the drag parachute is successfully opened or not. The parachute opening state of the drag parachute is judged by measuring the change gradient condition of the total number of the feature points on the drag parachute, the corresponding threshold value is set, and when the change gradient of the total number of the feature points is lower than the certain threshold value, the voice radio station is automatically triggered to carry out voice prompt, so that rapid judgment and emergency treatment of a pilot on the parachute opening state of the aircraft are realized. According to the method, whether parachute opening succeeds or not can be rapidly judged on the basis of pickup of the parachute opening feature points of the airplane, the alarm time is shortened, the response time of a pilot is shortened, the test flight safety margin is improved, and the method has good practicability.

The invention relates to a design method of the points of interest of a vehicle navigation system. The method comprises the following steps: the forms of two dynamic linked lists are located, namely a broadcast linked list and a point location linked list; the point location linked list contains the broadcast linked list; the point location linked list contains the longitude and latitude data of all the points of interest, the broadcast linked list contains the longitude and latitude data of the points of interest to be warned early; the longitude and latitude data of the points of interest are converted to the integer mode; the localization information of the vehicle is updated instantly, the points of interest in the broadcast linked list are started to compare; when the data of the points of interest completely are monitored to satisfy the set range, the vehicle is shown to enter the early warning range; and after the vehicle enters the early warning range, the information of the broadcast linked list is directly imported in the point location linked list, and the early warning information is prompted through an external loudspeaker. By using the method of the invention, the location speed of the points of interest is fast and the early warning time of the vehicle navigation system is shortened extremely.

The invention discloses a voice guide system based on image recognition. The system comprises a wireless communication system and a monitoring alarm system. The wireless communication system comprisesa wireless receiving module and a wireless transmitting module; the wireless receiving module comprises an earphone and a wireless receiver, the wireless transmitting module comprises a microphone and a wireless transmitter, the monitoring alarm system comprises a monitoring module and an alarm module, the monitoring module comprises a wireless camera, and the alarm module comprises an audible and visual alarm and a microprocessor. The invention discloses a voice guide system based on image recognition. The audible and visual alarm and the microprocessor are arranged, so the facial expressions of the tourists can be monitored, whether the tourists are in danger or not is judged by analyzing the facial expressions; when the tourists are in danger, early warning can be sent out at the firsttime to remind a tour guide principal to rush to an accident site for rescue, the early warning time is greatly shortened, and therefore real-time monitoring on the safety of the tourists is achieved.

The invention discloses an online monitoring device for a high steep slope in a highway, and relates to the technical field of highway high steep slopes. The online monitoring device comprises a supporting column, a protection frame and a fixed column; the supporting column is fixed in an immovable rock-soil body on two sides of a slope monitoring point; a monitoring assembly is fixedly installedat the top of the supporting column and comprises a box body; an inner cavity is formed in the box body; two fixed bases are fixedly installed on the inner wall of the inner cavity; and an insulatingcylinder is fixedly installed between the two fixed bases. The online monitoring device can achieve online monitoring of the high steep slope in the highway, save manpower and achieve real-time monitoring; when the high steep slope in the highway slides down, fallen soil can be blocked by the protection frame, the situation that the fallen soil impacts the road surface and the fixed column can beavoided, an early warning device and a flashing lamp jointly remind passing vehicle drivers, dual reminding is achieved, the situation that a driver can notice the abnormity of the high steep slope isguaranteed, the early warning time is saved, and the early warning effect is improved.

The invention relates to a jet propulsion superspeed torpedo. A torpedo body (1) is composed of a head (2), a front cabin (3), a middle cabin (8), a rear cabin (11) and a tail jet pipe (14). A heavy rocket engine and a fuel tank on the most advanced 'snowstorm' torpedo at present are omitted, main compressed air is used for replacing fuel gas generated by the rocket engine, and the torpedo body (1) is propelled in a jet mode; the supercavitation technology is continuously adopted, and the auxiliary compressed air is used for generating artificial cavitation; the middle cabin (8) is also used as a container for storing auxiliary compressed air, and the rear cabin (11) is also used as a container for storing main compressed air; the mass of the torpedo body (1) is greatly reduced, the noise is greatly reduced, the cost-effectiveness ratio is greatly improved, the speed is further improved, the penetration capability is higher, and the power is larger.

A rapid early-warning system based on a phase-sensitive optical time-domain reflectometer includes a disturbance monitoring and positioning module, and a disturbance signal identifying module. The disturbance monitoring and positioning module: a disturbance monitoring part performs difference on the beat signal intensity of a phase-sensitive optical time-domain reflectometer, the rapid early-warning system based on the phase-sensitive optical time-domain reflectometer can monitor the difference of the beat signal intensity all the time, and once the difference of the beat signal intensity changes, it means that a disturbance signal generated by disturbance occurs; and a positioning module can determine position information of a disturbance point through the time difference of an optical time-domain reflectometer echo. The disturbance signal identifying module includes a pre-treatment module, a feature extraction module, a template training module, a pattern classification module, and a post-treatment module; template training means that templates of a plurality of types of signals can be constructed according to features of a series of different types of disturbance signals; the features of the disturbance signals and preset templates of the plurality of types of signals are matched, and the types of the signals can be determined according the matching degrees; and a warning measure or a no-warning measure can be preformed according to a result of pattern classification.