40results about How to "High misjudgment rate" patented technology

The invention discloses a differential privacy-based data exception detection method. According to the method, the clustering is conducted firstly and then the exception detection is conducted. The method comprises the following steps of adding random noise to the weight of the edge in a minimum spanning tree generated by a prim algorithm by using a noise mechanism in the differential privacy so as to hide the relevance between data objects. Meanwhile, the method uses a judgment criterion of fusing dissimilarity and inverse similarity number to detect abnormality. By adopting the method, the problem that parameters need to be preset and the abnormal data selection is inaccurate in the traditional top-n method can be solved. The method provided by the invention has higher robustness and stronger adaptability, and experimental analysis of the simulation data set and the real data set shows that the method provided by the invention can effectively ensure the security of private data in anenvironment with non-uniform data distribution, improve the recall ratio of anomaly detection and reduce the misjudgment rate.

The invention discloses a method and device for determining abnormal circulation of air conditioner refrigerants and an air conditioner. The method includes the steps that in a refrigeration mode of the air conditioner, the indoor environment humidity of the indoor side of the air conditioner is acquired, and the evaporator pipe temperature and the indoor environment temperature of the indoor side are acquired; a first temperature difference between the indoor environment temperature and the evaporator pipe temperature and a second temperature difference between the pre-stored initial evaporator pipe temperature and the evaporator pipe temperature are determined, and whether the first temperature difference and the second temperature difference are smaller than or equal to corresponding preset values under the indoor environment humidity or not is judged; and when the first temperature difference and the second temperature difference are both smaller than or equal to the corresponding preset values, the abnormal circulation of the refrigerants of the air conditioner is determined. By means of the scheme, the defects that in the prior art, the mistaken judgment rate is high, the running reliability is low, and the maintaining difficulty is large can be overcome, and the beneficial effects that the mistaken judgment rate is low, the running reliability is high, and the maintaining difficulty is small can be achieved.

The invention discloses an urban management case classification and identification method. The method comprises the following steps: obtaining a large number of case image samples, classifying and recognizing case images manually, identifying various object objects and positions involved in urban management case content in the images, and inputting the various object objects, positions and imageswhich are identified manually into a deep learning model to be subjected to learning training. According to the invention, the deep learning model is not directly used to determine urban management case classification. The classification and positions of various objects appearing in the urban management case image are obtained by using the deep learning model. Then, category of specific urban management case is judged according to scene classification judgment algorithm, so that recognition misjudgment rate is reduced. The problem that the deep learning algorithm model is not high in judgmentaccuracy in an actual scene is solved, the misjudgment rate can be reduced, the recognition accuracy can be improved, and the urban management case recognition method better conforms to the actual scene of urban management case recognition.

The invention provides an automatic log generating system and method based on data acquisition and keyword excavation. The automatic log generating system comprises a data acquisition part, a data analysis part and a data display part, wherein the data acquisition part comprises a keyboard monitoring module, a clipboard monitoring module, a mouse monitoring module, a process monitoring module and a data storage module, the data analysis part comprises a process data analysis module and a mouse and keyboard analysis module, and the data display part comprises a data instant display module and a timing work log push module. The automatic log generating system has the advantages of having good user experience, intuitively displaying system data and pushing work log and optimization suggestion to a user in a timing way, so that the user is provided with good work management assistance, the work efficiency of the user is improved, and the automatic log generating system is suitable for people studying and engaging in scientific research under a work environment.

The invention discloses a recognition method of an open-circuit fault of a high-voltage capacitor of a double-tuning AC filter. The method includes main steps: acquiring an unbalanced current on an unbalanced bridge and a to-ground voltage of the unbalanced bridge in the high-voltage capacitor of the double-tuning AC filter, and calculating an unbalanced reactive power; indicating that the high-voltage capacitor has the open-circuit fault if the unbalanced reactive power is greater than a threshold; and indicating that the high-voltage capacitor does not have the open-circuit fault if the unbalanced reactive power is less than the threshold. According to the method, the unbalanced reactive power is regarded as a recognition criterion, the open-circuit fault of the high-voltage capacitor isrecognized, the recognition is accurate and reliable, reliable protection can be provided for the high-voltage capacitor of the AC filter, the method is free of the influence of the system operationmode and the voltage level change etc., and the system is applicable to all high-voltage DC power transmission systems adopting the double-tuning AC filter.

The embodiment of the invention provides a cattle illness data analysis method and device. The method concretely comprises the following steps: obtaining state index data representing cattle health states; selecting sample data, performing sample training of selected sample data according to an initialized cost matrix based on illness classification in advance and a multi-classification Bayesian process classifier based on cost sensibility, and obtaining the illness classification model of training samples; and according to the obtained state index data, the cost matrix and the training model, calculating the illness classification probability of the cattle to be detected and outputting the illness classification probability of the cattle to be detected. The method provided by the embodiment of the invention provides a cost-sensitive multi-classification Bayesian and Gaussian process to process the calculation of the illness classification probability of cattle, perform calculation of the state index states of the cattle and output a plurality of illness classification probabilities, and misjudgement cost factors are added to overcome the defects that a current discriminating method only can obtain general low misjudgement rate and only can output single classification results.

The present invention relates to an on-line identification method for detecting foreign matter in article by utilizing grain chracteristics. Said method includes the following steps: (1). shooting image; (2). image transmission; (3). foreign matter identification processing; (4). identification signal transmission; and (5). removing foreign matter. Said invention adopts LED as background light source, and uses the colour close the colour of article as background colour and adopts FFT to make grain analysis of the image so as to raise correct identification rate of articles of tobacco, etc. and foreign matter.

The invention belongs to the technical field of eyelid image processing, and discloses an intelligent eyelid detection system and method.The method comprises the steps that firstly, an eyelid positioning algorithm is used for positioning the position of an eyelid, an eyelid feature point detection method is used for positioning the position of a canthus point, and the position of the eyelid is calculated through the canthus point; the method comprises the following steps: establishing an iris template, then detecting the position of an eye region by using the iris template, positioning the position of an iris center through an iris center fine positioning algorithm, and finally, taking the detected canthus point and the iris center as eye motion vectors; a neural network model is used, the eye motion vector is used as input of a neural network, and a fixation point mapping relation is established. According to the invention, a good effect is achieved, more stable eye movement characteristics under head movement are selected, a single-camera light-source-free system architecture is adopted, and a 2D sight tracking method based on an artificial neural network is used, so that better precision is achieved; the hardware requirement of the system is reduced, the hardware cost is reduced, and the availability of the system is enhanced.

The invention discloses a fault identification method based on artificial intelligence, and aims to solve the problem of how to accurately identify a fault from seismic data and improve calculation efficiency. Through a series of improvements on a traditional AlexNet neural network, including replacing original local response normalization (LRN) with batch normalization (BN), the training speed is increased. Fault recognition is regarded as a dichotomy problem, and Sigmoid is used for replacing Softmax to serve as a classification function. Meanwhile, in the fault identification problem, the proportion of positive and negative samples (fault and non-fault) is seriously unbalanced, so that a balanced cross entropy loss function is introduced on the basis of a dichotomy cross entropy loss function to solve the problem. Finally, a 1 * 1 convolutional layer is used for replacing three full-connection layers of a traditional AlexNet network, so that the number of parameters of a network structure is greatly reduced. Through analysis of prediction results of theoretical data and actual data, it can be obtained that a model trained through the improved AlexNet network has good performance in the field of fault recognition.

The invention discloses a molecular marker of a paddy rice major gene site capable of preventing bacterial leaf streak and a primer pair for amplifying the molecular marker. The inventor subjects Guangxi wild rice resistant resources DY19(male) and 9311(female) to hybridization, backcross, and selfing to obtain BC3F2 offspring; and then resistance identification and molecular genetics linkage analysis are performed to obtain a molecular marker SL41, which is in close linkage with a gene capable of preventing bacterial leaf streak. The location of the major gene site detected by developed molecular marker is definite, the identification is convenient, and the molecular marker has an important meaning for the long term control of bacterial leaf streak. Moreover, the operation of using the molecular marker to detect bacterial leaf streak of a breeding material is simple, the accuracy can reach 95% or more, the production cost is saved, the efficiency of screening paddy rice species capable of preventing bacterial leaf streak is largely improved, the breeding cycle of disease resistant paddy rice species is shortened greatly, and the breeding efficiency is improved.

The invention discloses an optical fiber winding defect detection method and device based on ensemble learning. The method comprises the following steps of S1, acquiring an optical fiber winding image; S2, preprocessing the image acquired in the step S1; S3, carrying out defect detection on the image preprocessed in the S2 through a digital image processing model; S4, carrying out defect detection on the image preprocessed in the S2 through an SVM model; and S5, according to the digital image processing model and the SVM model, performing voting on defect detection results of the two models in the steps S3 and S4 according to corresponding weights by using a linear combination of set weight addition, judging whether voting results for a plurality of continuous pictures are not less than a set threshold score, and if so, judging that the optical fiber winding defect exists. The optical fiber winding defect detection device can reduce the misjudgment rate and the omission ratio, is suitable for optical fibers with various diameters, is more suitable for industrial environment production, and is high in adaptability.

The invention relates to a water turbine rotating wheel abrasion big data analysis system which comprises a first module, a second module and a starting module. The output end of the first module is electrically connected with a unit hydraulic input module. The output end of the unit hydraulic input module is electrically connected with a unit hydraulic and rated hydraulic computing module. The water turbine rotating wheel abrasion big data analysis system has the advantages that by arranging a first module, a second module and a curve generating time module and electrically connecting the first module and the second module in parallel, through the common effect of the first module, the second module and other modules, the force output conditions of a water turbine can be tracked for a long time, the variation tendency of a rotating wheel force output curve is drawn and the efficiency loss is calculated after various interference factors are eliminated, and therefore the abrasion degree of the rotating wheel can be precisely positioned, the efficiency loss variation is reflected in real time, the rotating wheel is analyzed form peripheral data expression and not related to the rotating wheel type and water flow model, and the analysis accuracy is ensured.

The invention relates to an anti-dry-burning gas stove which comprises an inner fire cover, a temperature detection piece and an air guide assembly. The temperature detection piece and the inner firecover are correspondingly arranged, the temperature detection piece is used for detecting the temperature of a target position on a pot placed above the inner fire cover, a through hole is formed in the inner fire cover, and the through hole penetrates through the bottom face and the top face of the inner fire cover; an air guide duct is arranged in the air guide assembly, an air outlet of the airguide duct is located near the target position, and the air guide duct is in butt joint with the through hole, so that airflow in the through hole can only flow to the air outlet end along the air guide duct, so that the air circulation of the target position is improved, heat generated by combustion flames is taken away, and the detection misjudgment probability of the temperature detection piece is reduced. In the whole process, guiding is conducted based on physical structures such as the air guide assembly, the purpose of reducing the misjudgment probability is achieved, the situation that the physical structures lose the guiding function or the guiding function is reduced due to the fact that the temperature in the using environment is high is avoided, and therefore the service lifecan be guaranteed.

The invention discloses a mobile communication locating and tracking system and a locating and tracking method thereof based on voiceprint recognition. The mobile communication locating and tracking system comprises a control module (Q) which is connected with a voiceprint recognition module (P), a video monitoring subsystem (M), a mobile switching center (MSC) and a gateway mobile locating center (GMLC) of a mobile communication network respectively; and the voiceprint recognition module (P) is further connected with media gateway of the mobile switching center (MSC). Whether criminal suspects change phone numbers or not, the system can locate and track the criminal suspects, a video monitoring subsystem is utilized to compare and confirm videos again on the basis of voiceprint comparison, so that the criminal suspects can be arrested rapidly and accurately.

The invention discloses an ultrasonic-based and electromagnetic ultrasonic combined dual-mode nondestructive testing method and device. The method comprises the following steps: coarsely testing a metal member under the action of a non-static magnetic field by utilizing an ultrasonic detection method; and applying pulsed excitation current to the metal member under the action of a static magnetic field. Lorentz force is generated inside the metal member under the action of the static magnetic field, so that an acoustical signal is generated due to the Lorentz force, and ultrasonic signals generated at positions with defects or without defects of the metal member are different. The electric conductivity distribution of the metal member is reconstructed through an ultrasonic probe array attached to the metal member, and the defects of the metal member are detected according to the result generated by combining the electric conductivity distribution and the ultrasonic detection. According to a device by using the detection method, a control system (2) is connected with a pulsed excitation system (1), an ultrasonic probe array (4) and a magnet system (3); the ultrasonic probe array (4) is connected with a signal detection processing system (5); the signal detection processing system (5) is connected with a reconstruction system (6).

The embodiment of the invention discloses a data identification method and device, electronic equipment and a storage medium. The data identification method specifically comprises the following steps: acquiring communication data to be identified; determining an operation domain call ticket feature and a statistical domain call ticket feature of the communication data to be identified; generating to-be-identified communication data features according to the operation domain ticket features and the statistical domain ticket features; performing feature recognition on the to-be-recognized communication data features to generate a feature recognition result of the to-be-recognized communication data features; and determining an identification result of the to-be-identified communication data according to the feature identification result. According to the technical scheme provided by the embodiment of the invention, the data can be identified according to the multi-dimensional data features, so that the misjudgment rate of data identification is reduced, and the accuracy of data identification is improved.

The invention belongs to the technical field of mining monitoring, and discloses a homologous monitoring stress and acoustic emission integrated sensor, a monitoring method and a mounting method, the sensor device comprises an acoustic emission probe, an amplifier, a stress sensor, a tilt angle sensor and a circuit board; the acoustic emission probe comprises a piezoelectric ceramic crystal and a guide head; the front end face of the piezoelectric ceramic crystal and the rear end face of the guide head are in plane contact, and the piezoelectric ceramic crystal and the amplifier at the rear end are integrated. The stress sensor comprises a hydraulic cylinder body, an elastic body and a pressure-bearing sheet; the hydraulic cylinder is connected with the elastomer through the piston and the connecting device; the pressure-bearing sheet is mounted on one side of the elastomer; the circuit board is connected with the amplifier through a power supply line. The technical problems of high randomness and poor accuracy of early warning of a single sensor are discarded, a thought of integrating the sensor is provided, mechanical and acoustic monitoring methods are integrated into the unified sensor, coupling monitoring of multi-source information is realized, and the early warning accuracy is improved.

The invention relates to a method and a device for reducing reverse signaling misjudgment in narrow-band wireless digital communication, and the method comprises the steps: a first reverse signaling is received and obtained in a downlink time slot; a second reverse signaling is received and obtained in a plurality of downlink time slots after the first reverse signaling; comparison is performed between the first reverse signaling and the preset quantity second reverse signalings according to the receiving sequence successively, if the first reverse signaling is same with one of the second reverse signaling, the first reverse signaling is effective is determined, if the first reverse signaling is different with the preset quantity of second reverse signalings, the current first reverse signaling is ineffective is determined, and a first second reverse signaling acquired after the current first reverse signaling serves as the first reverse signaling. According to the invention, if the reverse signaling is effective is determined through comparing reverse signalings received by a plurality of downlink time slots, after the ineffective reverse signaling is judged, the comparison between following reverse signalings is performed, the misjudgment rate of the reverse signalings can be effectively reduced, and the judgment efficiency is improved.

The invention provides a night vision infrared camera detection device based on a near infrared spectrometer, a roller is arranged on the lower surface of a sliding base, two electric winding devicesare respectively arranged on two sides of the upper surface of the sliding base, a telescopic assembly is arranged between the two electric winding devices, and supporting assemblies are uniformly fixed on the telescopic assembly. The supporting assembly comprises a supporting rod and supporting pulleys, the supporting rod is fixed to the telescopic assembly, the supporting pulleys are fixed to the two ends of the supporting rod respectively, the transmission device conducts transmission around the two electric winding devices and the supporting pulleys, and a hanging device used for hanging the detection optical fiber box is arranged on the transmission device. Aiming at the characteristic that infrared rays with the wavelength of 850-940nm supplement light for night visual monitoring, anear-infrared spectrometer is used for quickly and accurately detecting the infrared rays in the wave band, so that the detection accuracy is effectively improved, and the detection misjudgment rate is reduced.