55results about How to "Discrimination speed is fast" patented technology

The invention discloses a distributing net single phase earth fault type and a distinguishing method of phase identification, relating to a distributing net single phase earth fault distinguishing method in the field of AC distributing net testing and relaying protection technology. According to the invention, through collecting three-phase voltage of the distributing net system and neutral point voltage of the distributing net system, phase angle of each eigenvector is computed through a special relation, the type of the fault and the phase identification can be fast distinguished through the phase angle distinguishing logic. The inventive working system is that a three-phase bus (000), a grounding transformer (100), a linear side of an arc suppression coil (200) are connected in turn with the earth; a controller (400), a thyristor (300) and a secondary side of an arc suppression coil (200) are connected in turn. The invention is simple in criterion, small in collection quantity, fast in distinguishing speed, high in accuracy, reliable in security, which is suitable for 3-66 KV distributing net based on an aerial line or a power cable.

The invention belongs to the field of voice recognition, and particularly relates to a phone-call recording access failure reason recognizing method. The method comprises the following steps: markingaccess failure reasons by signals; if reasons cannot be obtained by signal classification, extracting an audio fingerprint characteristic sequence from to-be-recognized phone-call recording, and searching from an audio fingerprint database by the sequence; if matched fingerprints can be found out, marking the access failure reason for to-be-recognized phone-call according to access failure reasonlabels in fingerprint key values; and if the matched fingerprints cannot be found out, recognizing audio contents into text contents by automatic voice recognition, classifying in an access failure document classifying model by a text classifying method on the basis of the contents, and marking the to-be-recognized phone-call recording by access failure reason classifying results obtained by classification. The method can recognize recording files in an offline manner, streaming phone-call voice can also be recognized, the universality is high, and the phone-call recording access failure reason recognizing method is suitable for different application scenarios of the call center.

The invention relates to a vivo face identification device for a face recognition system and belongs to the face recognition technical field. The vivo face identification device for the face recognition system comprises a light source, a collecting lens, a lantern slide, a projection lens, a video camera and a computer, wherein the computer is respectively connected with the video camera and the light source, a light beam emitted from the light source irradiates on the lantern slide after passing through the collecting lens, and the projection lens projects an image of the lantern slide on an identified object. Compared with the prior art, the vivo face identification device for the face recognition system has the advantages of utilizing parallel lines which are projected out to identify the vivo faces, static graphic photos and dynamic screen images and being high in identification accuracy; not needing identified people to perform facial expression changes to cooperate with the identification, overcoming the defects that the prior art can only identify the static photos but can not identify the screen dynamic images and greatly improving the security and convenience of the face recognition system; being easy in structure, being reliable in work, being rapid in identification speed and being easy to implement.

The invention discloses a load identification method based on an improved probabilistic neural network. The load identification method comprises steps of adopting a binary system to perform coding on a load type in an electricity usage network, establishing a coding library of load types, collecting an electric parameter of each load type, establishing a non-linear mapping relation between the electric parameter and the code of each load type, creating a probabilistic neural network, training the probabilistic neural network to obtain an error function, using the error function as a fitness function of a particle swarm algorithm, adopting the particle swarm algorithm to perform optimization on a smooth factor of the probabilistic neural network to obtain an optimal smooth factor, updating the probabilistic neural network according to the obtained optimal smooth factor to obtain an improved probabilistic neural network and performing recognition on the load type in the electricity usage network on the basis of the probabilistic neural network. The load identification method based on the improved probabilistic neural network can simultaneously identify multiple loads, improves speed of identifying a pernicious load under a multi-load mode, and can better realize recognition of restricted electric appliances and electricity usage control under a condition of multiple electric appliances.

The invention provides a method and a system for distinguishing the language of a document image, comprising the following steps: detecting the round white pixel connection domain in the document block of the document file, and determining whether the document block is a Korean on the basis of the detecting the round white pixel connection domain.

The invention relates to a power grid abnormal state distinguishing method based on ring theory. The method comprises the following steps of: forming an original data matrix by utilizing WAMS (Wide Area Measurement System) data with processing of sliding time window and matrix standardization processing to obtain a high-dimensional random matrix for a ring theory; carrying out spectrum analysis research on the processed random matrix, and analyzing the processed random matrix from the ring theory; data is distributed between an outer ring and an inner ring when a system is in normal operation,and a large number of abnormal data points within the inner ring start to appear when the system starts to have an abnormal situation, therefore constructing a criterion of abnormal states by comparing a length of a minimum characteristic value with an inner ring radius; and judging the severity of the abnormal states according to the proportion of inner ring data points in a ring graph. According to the power grid abnormal state distinguishing method based on ring theory, power grid state is analyzed from a data driving perspective by using the WAMS data, an evaluation index of power grid abnormal state is constructed through the ring theory, so that the power grid abnormal state and the severity thereof can be quickly identified, and power big data is fully utilized.

The invention relates to a human identification device used for a face recognition system, belonging to the technical field of face recognition. The human identification device comprises a left light source, a right light source, a camera and a computer; the left light source and the right light source are respectively arranged at the right front side and the left front side of a to-be-recognized person and used for lighting the face of the to-be-recognized person; and the computer is respectively connected with the left light source, the right light source and the camera. The human identification device provided by the invention identifies the to-be-recognized person and a paper photo and a display screen through utilizing the positive and negative faces formed by a manner that the head of the to-be-recognized person shields the side lighting, the identification accurate rate is high. Compared with the prior art, the identification device does not need the to-be-recognized person to do expression changes to match the recognition, and moreover, the device solves the defects that in the prior art, only the dynamic photos can be identified and the dynamic screen images cannot be identified, thus, the security and convenience of the face recognition system are greatly improved; and meanwhile, by adopting the technical scheme provided by the invention, the system is simple and reliable in structure, fast in identification speed and easy in implementation.

The invention provides a distance measurement method based on a millimeter-wave radar. The method comprises the following steps that: obtaining the radar signal data of the millimeter-wave radar; according to the radar signal data, obtaining a detection matrix, and setting a distance CFAR (Constant False-Alarm Rat) threshold value, wherein the line of the detection matrix represents distance, andthe row of the detection matrix represents speed; according to the CFAR threshold value, grouping the detection matrix; in each group, comparing data in each row to obtain a maximum value; and outputting the maximum value, and determining a minimum distance between the millimeter-wave radar and an object according to the maximum value. By use of the method, the distance data is distinguished in advance, a situation that the radar can not detect the object when the radar and the object are simultaneously static can be eliminated. Meanwhile, the distinguishing speed for the shortest distance between the radar and the object is greatly improved, and the working performance of the millimeter-wave radar is improved. The invention also provides a distance measurement system based on the millimeter-wave radar, a computer readable storage medium and the millimeter-wave radar, and also has the above beneficial effects.

The invention discloses a video stream data acquisition method used for solving a technical problem that delay of a server increases due to response speed decline in a high concurrency scenario. The method comprises the following steps: a virtual set top box client end sends a video stream request to a virtual set top box server, the virtual set top box server sends a page request used for requesting page data to a network server, the network server obtains the page data, the page data is sent to the virtual set top box server after M pieces of path information is added to the page data, all of N pieces of data that N pieces out of the M pieces of information corresponds to is static state data which is stored in a static state data synchronization database via the network server, the page data is processed via the virtual set top box server and therefore video stream data is obtained, and the video stream data is sent to the virtual set top box client end. The invention also discloses a corresponding system, a page data transmission method and the network server.

The invention discloses a method and a device for automatically and quickly extracting newly-added human activity plaques on a marine red line, and belongs to the field of environmental monitoring. The method comprises the following steps: segmenting a background image and a to-be-detected image by utilizing a background vector, and judging segmented plaques one by one by utilizing a brightness statistical value and a spectral characteristic function to obtain changed plaques and suspected changed plaques; carrying out secondary segmentation on the suspected changed plaques and plaques corresponding to the background image according to the characteristics of spectrum, geometry, texture and the like, and judging the segmented plaques one by one by utilizing a brightness statistical value and a spectrum characteristic function to obtain changed plaques; merging all the changed plaques to obtain newly added human activity plaques; and counting the newly added human activity plaques to obtain the number and area of the newly added human activity plaques in the ocean red line. The method can automatically and quickly extract the position, number, distribution area and other informationof the newly added human activity plaques in the marine red line, and provides quick and accurate technical support for marine red line human activity supervision.

The invention relates to a discrimination method and determination device for laser damage on the surface of a film or an optical element and a determination device therof. The prior art has the disadvantages of strong subjectivity, great working intensity and low testing efficiency. The invention provides a discrimination method for laser damage on the surface of a film or an optical element. According to the method, an operating laser beam carries out monopulse laser irradiation on a sample; if the surface of the sample has no damage, the surface does not produce plasma flash; if the surface is damaged, strong plasma flash is generated the moment the damage appears, and if a photoluminescence peak of an element not being N, O, H or C appears according to analysis of a flash spectrum, appearance of damage can be determined. The invention also discloses a determination device realizing the above-mentioned method. The device comprises a laser generator and a test board and is characterized by further comprising a convergent lens and a fibre-optical probe on the imaging surface of the convergent lens, wherein the fibre-optical probe, a fiber spectrometer and a computer are connected sequentially. According to the invention, high precision in the discrimination is obtained; the discrimination speed is fast; a wide variety of films can be discriminated; the construction of the determination device is simple.

The invention discloses a distinguishing method of maturity degrees of tobaccos based on image analysis. The distinguishing method comprises the following detailed steps: (1) classifying tobaccos of different maturity degrees through professional staffs; (2) acquiring different maturity degree images of the tobaccos; (3) acquiring tobacco images and performing normalization on the images; (4) performing normalization on the acquired tobacco images by utilizing an image analysis method; (5) extracting values of three components R, G and B of the images; (6) building maturity degree estimation parameters, screening characteristic parameters of existing known colors and a series of self-built characteristic parameters, and screening the all characteristic parameters by utilizing a person relevant analysis method; (7) acquiring an M value from the images, and judging a maturity degree value through contrasting maturity degree grading table; and (8), realizing image reading and maturity degree output through a phone app and computer software.

The invention discloses a quick recognizing and processing method and system for an ECG signal. The method comprises the following steps: denoising an input original ECG signal to obtain an ECG signal; inputting the ECG signal in a basic characteristic signal recognizing and processing flow or a neural network signal recognizing and processing flow according to the complexity of the ECG signal; and outputting a recognizing and processing result. The provided quick recognizing and processing method and system for the ECG signal improves the signal recognition efficiency and accuracy of analysisresults, and can be applied to various embedded platforms.

The invention discloses an intelligent alarming method for a grid. On the basis of a PMU point arrangement optimization, a grid observer based on a bus-bar node voltage is constructed to enable a failure and a node voltage to correspond to each other, and a decision tree and an algorithm supporting the cooperation of a vector machine are used for giving an alarm for a grid failure. The method utilizes the decision tree to reduce the classifying combination number and constructs a classifying performance value to represent the divisibility of each type; furthermore, the PMU point arrangement optimization is introduced to reduce the dimension number of the observer and greatly improve the judging speed and meet the real-time requirements of the failure alarm.

The invention provides a differential protection locking method for the tail current of a series transformer. The tail current is determined by utilizing the trend of a current sampling value in the phase current, differential protection is locked according to the phase, the tail current can be determined more accurately, protection is not refused during a fault, protection is not carried out formistake when the tail current occurs in the CT, accidents are prevented from expanding, and safe and stable operation of the power grid can be guaranteed more effectively.

The invention discloses a method and a device for identifying laser damages to an optical film. The types of elements contained in damaged materials can be determined by acquiring mass-to-charge ratio distribution and performing mass spectrometric analysis and element judgment through computer software, so that damages to the film can be judged accurately and rapidly on line. The device adopting the method consists of a test sample stage, a mass-to-charge ratio acquisition system and a computer operation system. By adopting the method and the device, the damage state of the film or an optical element under strong laser light can be judged rapidly and accurately on line in real time. By applying the method to laser damage threshold test, integration, automation and intelligence of a test system can be realized.

The invention provides a synchronous phase modifier rotor winding weak turn-to-turn short circuit fault diagnosis method, which comprises the following steps: acquiring air gap flux density, rotor amplitude and stator amplitude of a synchronous phase modifier in real time, and selecting the air gap flux density, the rotor amplitude and the stator amplitude of the synchronous phase modifier before and after a fault as evidence bodies; calculating basic trust distribution functions of the three evidence bodies before and after the fault, fusing the three trust distribution functions by adopting an evidence fusion rule, calculating a fault trust distribution value, comparing the fault trust distribution value with a fault judgment threshold value, if the fault trust distribution value is greater than the fault judgment threshold value, judging that the fault exists, and otherwise, judging that the fault exists; and constructing a field-circuit-network coupling simulation model of high-voltage direct-current power transmission containing the synchronous phase modifier, and performing simulation verification. According to the synchronous phase modifier rotor winding weak turn-to-turn short circuit fault diagnosis method, the problem that the existing weak turn-to-turn short circuit fault diagnosis technology of the synchronous phase modifier is insufficient can be solved, and the detection precision of the turn-to-turn short circuit fault of the rotor winding of the synchronous phase modifier is improved.

The invention discloses a method for distinguishing the optical axis of a camera in real time, relates to a method for distinguishing relative poses of objects in machine vision in real time, and belongs to the field of machine vision. The method comprises steps as follows: Step one, LED flickers are mounted on a game display screen, a high-speed camera for acquiring images of the LED flickers mounted on the screen in real time is arranged in a virtual gun, and backlight removal treatment is performed on the acquired images of the LED flickers mounted on the screen; Step two, the mapping relation between acquired LED flicker positions and actual positions of the LED flickers on the game display screen is established according to parity of point numbers of the LED flickers, acquired by the virtual gun in real time, on the game screen and the LED flickers, and the relative poses of the objects in machine vision are distinguished in real time. According to the method, the position of the optical axis of the camera in machine vision is distinguished through LEDs arranged on the game display screen and calculation of the virtual gun mapping position relation, and the method is high in distinguishing speed and high in result precision and is slightly interfered by the outside.