767results about How to "The judgment result is accurate" patented technology

The invention provides a dust removal method and device applied to an air conditioner outdoor heat exchanger. An outdoor unit of an air conditioner comprises the outdoor heat exchanger and an air duct system composed of a motor and vanes. The dust removal method includes the following steps that whether dust removal processing needs to be carried out on the outdoor heat exchanger is judged; when it is judged that the outdoor heat exchanger needs dust removal processing, at least the motor is controlled to rotate reversely, and when the outdoor unit works normally, the motor rotates forwards. By means of the technical scheme, the effect that the air conditioner automatically carries out dust removal on the outdoor heat exchanger is realized, a higher heat exchange efficiency of a condenser is guaranteed, and the complex operation of detaching the outdoor unit when the condenser is cleaned manually is avoided.

The invention discloses a traffic image multi-type vehicle detection method based on deep study. The traffic image multi-type vehicle detection method based on deep study firstly combines a nerve network characteristic with a region generation algorithm, realizes two processes-region generation and region determination-at the same time through using a nerve network conventional layer, uses a background model to determine a motion area of a discrete series image targeting a specific scene and provides an extra reference basis to region generation, and combines with a vehicle detection result to perform upgrading and correction on the background model according to different conditions. Besides, the traffic image multi-type vehicle detection method also brings forward a network model compression scheme to reduce model parameters and calculation time, brings forward a new detection result optimization means based on grouped error calculation to replace a conventional non-maximum-value inhibition scheme, and improves overall detection accuracy.

The invention discloses a wavelet transformation and deep learning-based rolling bearing weak fault diagnosis method. The method comprises the following steps: a rolling bearing vibration signal is obtained, and the acquired vibration signal is subjected to continuous wavelet transformation to obtain a time-frequency diagram; autocorrelation operation is performed on wavelet coefficients corresponding to each frequency on the time-frequency diagram to filter out noise interference and extract periodic fault components; the Hilbert transformation is used to perform envelope demodulation to obtain a fault characteristic frequency; a processed time-frequency diagram is input as a feature diagram, and categories of early faults are determined by training a deep learning classification model. Theoretical and experimental results show that fault categories can be determined accurately at an early stage when weak fault of a rolling bearing occur based on an improved wavelet time-frequency diagram used as the classification model for input training, right determination results can be given when the method is applied to different bearings, high training speed can be realized, and high actual application value can be achieved.

The invention discloses a method for rapidly and accurately judging false wiring of a metering device. The method comprises the following steps that firstly, it is ensured that the measuring conditions that the voltage of a three-phase and three-wire circuit is symmetrical, and the three-phase load power factor angles are equal are met; secondly, six phase angle differences on the three-phase, three-wire and two-element metering device are measured through a phase position voltammeter; the relation of the current phase angle of the three-phase, three-wire and two-element metering device is analyzed, the current phase difference is determined, the voltage phase angle relation is analyzed, the voltage phase sequence is determined, the relation of the voltage and the current phase angle is analyzed, the voltage phase difference is determined through the phase crossing method, the voltage and current phase angle relation is analyzed finally, and the current direction is determined. The method has the advantages that the algorithm is simple, the data size is small, using is convenient, and the method is rapid and accurate.

The present invention discloses a method and an apparatus for automatically identifying a region of interest in a three-dimensional CT image. The method includes: training a 3D convolution neural network through a three-dimensional CT image marked with a region of interest to obtain a 3D convolutional neural network with a high identification rate for the region of interest; then, through the trained 3D convolution neural network, identifying the obtained three-dimensional CT image and obtaining the region of interest; and calculating the volume, the expansion direction and the expansion diameter of the interest of interest. With the method of the embodiments in the invention, the region of interest representing the lesion can be obtained automatically and efficiently, and the identifying accuracy is also higher. In addition, the volume, the expansion direction and the expansion diameter of the lesion can be automatically calculated, therefore, helping the doctor to make a more accurate judgment of the condition of a patient.

The invention discloses a speech quality detection method and device, which relates to the technical field of speech detection, and mainly aims at improving the comprehensiveness and accuracy of speech quality detection. The main technical scheme of the invention is characterized by converting the speech to be quality-checked into the text to be quality-checked; extracting quality inspection itemsone by one from a quality inspection item collection, wherein the quality inspection items are used for detecting whether the text to be inspected conforms to requirements; using a pre-trained machine learning model to detect whether the text to be inspected contains content matching with the quality inspection item; according to the matching result and the detection rule corresponding to the quality inspection entry, determining whether the speech to be inspected is qualified or not. The invention is used for the normative detection of the voice data content.

The invention relates to a multi-stage phishing website detection method and detection system based on supervised learning. A first detection layer performs phishing website judgment based on a blacklist database or a white list database, the to-be-detected website is directly output if the website is matched with the blacklist database or the white list database; or a second detection layer extracts features of the to-be-detected website URL and constructs a classifier model to perform detection according to the URL features of the known phishing website; if the website is detected as a suspicious website, a third detection layer downloads the page of the to-be-detected website to acquire the page content features, and the classifier model is constructed according to the known phishing website content features so as to perform the detection; an output end outputs the result that to-be-detected website is the phishing website or the normal website, and adds the data to the blacklist database and the white list database. The primary black-white list judges the known website and reduces the detection cost; the secondary URL detects and distinguish the clear phishing website or the normal website; and the third-stage page content detection identifies the suspicious website of the secondary detection, the judgment result is precise; an identification result is accurate, and the detection time is short.

The invention relates to the field of air conditioner control, in particular to an electric heating structure and an air conditioner compressor starting system and method. The electric heating structure comprises an electric heating control system and an electric heating device. The electric heating control system comprises a driving circuit and an electronic switch circuit. The driving circuit is used for adjusting the duty ratio of the electronic switch circuit. The electronic switch circuit is used for adjusting the output power of the electric heating device according to the duty ratio. The electric heating device is used for preheating an air conditioner compressor. According to the technical scheme, the output power of the electric heating device can be adjusted according to the exhaust temperature of the compressor; compared with the method that preheating is conducted by adopting a constant-power electric heating device in the prior art, the heating speed of the compressor is high, and the preheating process safety can be ensured; and accordingly, the starting waiting time of an air conditioner compressor system under a low-temperature environment is reduced, energy consumption is reduced, and meanwhile, the user experience effect is improved.

The invention belongs to the field of data mining and provides a method and system for advertisement recommendation based a microblog. The method comprises the steps that microblog data are read; the microblog data are initialized and a microblog text lexical item set is obtained; stop words of the microblog text lexical item set are deleted and a microblog text original feature lexical item set is obtained; mapping is conducted on the microblog text original feature lexical item set and a feature lexical item dictionary, whether lexical items in the microblog text original feature lexical item set exist in the feature lexical item dictionary or not is judged, and the tf-idf values of the appearing lexical items are calculated and serve as the feature values of the lexical items; whether the lexical items of the feature lexical item dictionary exist in the microblog text original feature lexical item set or not is judged and the feature values of the lexical items which do not appear are marked to be zero; feature vectors of the feature values obtained through calculation are automatically classified to classifications divided in advance; according to an automatic classification result, advertisements are recommended to a user. The advertisements recommended by the method and system are accurate and the effect is good.

The invention discloses a direct-current single-pole ground fault determination method of a flexible direct-current power transmission system. Voltages to ground of positive and negative direct-current buses of a direct-current side are obtained; if an absolute value of one voltage value is less than a low-voltage limiting value and an absolute value of the other voltage value is larger than a high-voltage limiting value and a condition that duration time is larger than a set screening time is met, a single-pole ground fault is determined to occur in a direct-current bus corresponding to the voltage value with the absolute value less than the low-voltage limiting value. According to the invention, on the basis of the changing feature of the voltage during the direct-current single-pole ground fault development process, the voltages to ground of the positive and negative direct-current buses are selected as direct-current single-pole ground fault identification parameters. Besides, overtime of the voltage parameter is also used as another main criterion, thereby improving the fault identification accuracy and avoiding wrong determination caused by other faults. With the method, the direct-current single-pole ground fault of the flexible direct-current power transmission system of the modularized multi-level converter can be identified rapidly and accurately.

The invention discloses a security judging method of a rail irregularity state in the technical field of railway security operation control. The security judging method comprises the following steps of: establishing a vehicle rail model in multi-body dynamic simulation software; determining a rail irregularity parameter; inputting the rail irregularity parameter into the vehicle rail model and generating an executable code in the multi-body dynamic simulation software; inputting actually-measured rail irregularity data into the executable code, and operating the executable code to obtain wheel-rail acting force; calculating a security evaluation index of the rail irregularity state according to the wheel-rail force; processing the calculated security evaluation index of the rail irregularity state; and comparing the processed security evaluation index of the rail irregularity state with a pre-set security standard so as to obtain a security judging result of the rail irregularity state. Compared with the conventional security judging method of the rail irregularity state, the security judging method provided by the invention is more accurate in judging result.

The invention discloses an audio similarity detecting method, an audio similarity detecting device and electronic equipment. The detecting method comprises the following steps that: an audio stream to be detected is obtained; a feature parameter value of the audio stream to be detected is determined; a previously known feature parameter value of a reference audio stream corresponding to the audio stream to be detected is compared with the feature parameter value of the audio stream to be detected; and according to a comparison result, the similarity between the audio stream to be detected and the reference audio stream is calculated. According to the technical scheme disclosed by the invention, the feature parameter values between the audio streams are compared, so that the judgment of the similarity is more accurate and quantitative, and the use experience of users is favorably improved.

The invention provides a detection method and system for target domain name hijacking. The method comprises the steps that a server side adopts a mode of sending a domain name analysis request to an authority DNS server of the target domain name and receiving at least one analysis result of the target domain name returned by the authority DNS server to establish a standard analysis result list; the server side allocates the domain name to multiple specific client sides to execute a target domain name analysis probing task, and each client side sends the received target domain name to an LDNS server configured for the client side to request domain name analysis and reports the received analysis result to the server side; the server compares each received analysis result which is adopted as a probing task with a standard analysis result recorded by the standard analysis result list, and therefore whether a hijacking behavior exists in the LDNS server configured for the server side or not is determined.

The invention relates to the technical field of parking management, in particular to a vehicle management method and system and a server. The method comprises the following steps: receiving a vehiclemoving request sent by a first terminal through scanning an identification code on a target vehicle, wherein the vehicle moving request comprises a license plate number of the target vehicle; Determining a second terminal corresponding to the license plate number according to a corresponding relation between the license plate number and pre-stored vehicle owner contact information and license plate number information; And sending a vehicle moving prompt message to the second terminal to prompt a user of the target vehicle to move the vehicle. when receiving a vehicle moving request which is sent by a first terminal through scanning an identification code on a target vehicle and comprises a license plate number of the target vehicle; a second terminal corresponding to the license plate number is determined according to the license plate number and the corresponding relation between the pre-stored owner contact information and the license plate number information, and a prompt message for prompting the user of the target vehicle is sent to move the vehicle to the second terminal, so that the vehicle moving problem can be effectively solved under the condition that the privacy information of the owner is not leaked.

The invention provides a method for detecting steel tube void of concrete filled steel tube (CFST) arch bridge by utilizing a transient impact method; the method provided by the invention is characterized by comprising the following steps: 1, arranging measuring points, axially arranging a plurality of measuring points at the outer wall of a steel tube which is made of a CFST and is to be detected, and placing an acceleration transducer on each measuring point; 2, arranging a vibration excitation point at the outer wall near to the head end or tail end of the steel tube, horizontally knockingthe vibration excitation point by a vibration excitation device, and detecting the acceleration response signals of the measuring points by the acceleration transducer; 3, outputting the response signals of the measuring points by the acceleration transducer, inputting the signals into a data collector by a signal amplifier, obtaining the natural frequency and acceleration power spectral density of each measuring point by the data collector, and storing in a computer. In the invention, the steel tube void of the CFST arch bridge can be located fast and exactly according to the natural frequency and power spectral density of acceleration response and via calculating and comparing the acceleration spectral energy values of the measuring points based on an expression.

The invention discloses a method for detecting a human motion state based on accelerated speed changes. The method includes the following steps that A, a mobile device where an accelerated speed sensor is embedded is placed at any position of a human body, an accelerated speed sampling value of the accelerated speed sensor is compared with a preset accelerated speed threshold value, and according to the compared result, whether a human moves or not is judged; B, according to the step A, if the human moves, a self-correlation function is used for calculating the self-correlation degree of the accelerated speed sampling value; C, the self-correlation degree is compared with a preset self-correlation threshold value, and according to the compared result, whether the human walks or not is judged. According to changes of the accelerated speed value, whether the human walks or not can be judged. Therefore, the method not only is simple and accurate in judgment result, but also has very great practical significance and polarization value.

The invention relates to a fault arc online detecting method based on wavelet first-layer high-frequency component correlation. The method comprises the following steps of 1) extracting the first-layer wavelet decomposition high-frequency component of series current during normal operation of a load to obtain the reference sequence of sampling current; 2) collecting the sampling current of a low-voltage AC (alternate current) series circuit, where the load is disposed, at a set frequency, performing wavelet decomposition on the sampling current to extract a first-layer high-frequency detail component; 3) segmenting the first-layer high-frequency detail component extracted in the step 2) to obtain a sequence to be detected; 4) according to the correlation coefficient of the sequence to be detected and the reference sequence, determining whether fault arcs exist, if so, performing the step 5), and if not, returning to the step 2); 5) determining whether the number of the fault arcs meets the UL1699 standards, if so, outputting tripping signals. Compared with the prior art, the fault arc online detecting method based on wavelet first-layer high-frequency component correlation overcomes the dependency of existing low-voltage AC series fault arc determining methods on specificity of the load and shortens the fault determining time.

The invention relates to a multifunctional fire fighting emergency device with a redundancy judgment function, which is a treble redundancy judgment device. The device provided by the invention comprises three smog detecting units, signal processing units corresponding to the three smog detecting units, and an action unit, as well as a comparing unit and a logic and control unit, wherein the comparing unit calculates the signal superposing degree of the second signal processing unit over the first signal processing unit, that of the third signal processing unit over the first signal processing unit, and that of the third signal processing unit over the second signal processing unit, and then sends the results to the logic and control unit where threshold determination is carried out; if the superposing degree is smaller than the threshold, that the smog concentration of the positions detected by two smog detecting units exceeds standard is detected. The emergency device adopts treble redundancy judgment, and as a result, the result is more accurate, and alarming measures can be promptly taken while misjudgment is avoided.