33results about How to "Reduce the impact of calculations" patented technology

Embodiments of the present invention provide a method and a device for text clustering, and relates to the technical field of text processing. After determining k text data which was obtained text feature vector concentrated as the center point of k cluster, and traversing the text feature vectors set, dividing each text datum into the cluster with the highest similarity, and then executing for each cluster: calculating a mutual information between each word in a text data of a cluster and the clustered data in the cluster, and ranking the calculated amount of mutual information, based on a preset evaluation function of the mutual information according to the method. Extracting T words with high mutual information quantity as the feature vector of the text data, and determining an averagevalue of the feature vectors of all text data in the calculated cluster as a new center point of the cluster, until the preset condition is satisfied. The method and device for text clustering reducing the influence of an isolated data point on the calculation of the new center point, and the method is more accurately, and in the condition of cluster characteristics are ensured to be complete, reducing the dimension, so that the calculation efficiency is improved.

The invention discloses an onboard 360-degree around-viewing display method and system. The method comprises the following steps of: S1, enabling a projection module to subject images acquired by respective cameras to viewpoint transformation by looking up a lookup table and to project the images onto a vehicle top view; S2, subjecting the four camera images to optical alignment by using an exposure gain algorithm; S3, splicing the splicing seams of the images acquired by the adjacent cameras by using a splicing algorithm to obtain the vehicle top spliced view. The method reuses the splicing seam of a last frame when the splicing seam at a certain position is not large in adjacent timing sequential difference so as to reduce an influence of noise on the calculation of the splicing seam greatly for improving stability. In addition, the direct use of the splicing seam of the last frame also shortens the calculation time of splicing seam and improves algorithm efficiency.

The invention discloses a calculation method for determining a heterogeneous rock crack propagation path based on an FFT method, and compared with an original phase field method based on a finite element framework, the phase field fracture calculation method based on the FFT homogenization method can overcome the defects of the finite element framework, reduce the influence of the size effect on calculation, and improve the number calculation efficiency and reduce the calculatoin cost while ensuring the calculation precision. from the microcosmic perspective, the relation between microcosmic-scale crack development and macroscopic-scale material damage is established through the homogenization theory, the material fracture mechanism can be accurately reflected, and a reliable basis is provided for predicting heterogeneous material fracture. In addition, the calculation method for determining the heterogeneous rock crack propagation path based on the FFT method is easy to program and convenient to implement through a programming language, fracture prediction is conducted through a computer, and the manual calculation amount is greatly reduced.

The invention discloses a quasi-static ultrasonic elastography displacement calculation method so as to solve the technical problems that linkage is mistaken and two-dimensional cross-correlation research calculated quantity is excessively large due to iteration in the prior art.The embodiment of the invention further provides a corresponding device.According to the feasible embodiment, the method comprises the steps of calculating a central point for any displacement, and utilizing prior time shifting for building an iteration cross-correlation function based on a neighborhood window for iteration cross-correlation calculation; obtaining a cross-correlation reliability exponential value of iteration cross-correlation calculation, and judging whether the cross-correlation reliability exponential value is smaller than a first credible threshold value or not; if the cross-correlation reliability exponential value is larger than or equal to the first credible threshold value, determining transverse time shifting according to the prior time shifting, and determining longitudinal time shifting according to the prior time shifting and an iteration cross-correlation result; if the cross-correlation reliability exponential value is smaller than the first credible threshold value, using two-dimensional comprehensive cross-correlation research inside the neighborhood to obtain the longitudinal time shifting and the transverse time shifting of the central point of displacement calculation; traversing all ultrasonic radio-frequency signals, and generating displacement field distribution of tested biological tissue.

The invention discloses a navigation satellite integrity key point comprehensive identification method based on a grey relational degree. The invention takes the navigation satellite system as a research object and brings forward three properties of the satellite integrity key point which include node reliability, integrity and topology importance degree, and comprehensively and effectively evaluates the importance of various nodes of the navigation satellite system. The navigation satellite integrity key point comprehensive identification method comprises steps of choosing an appropriate identification index, on the basis, calculating identification index values from the angles of the topology structure, reliability and integrity, performing irrelevant screening on the identification indexes, calculating the weight values of various identification indexes, obtaining a grey relevant value of each node through quantitative calculation and identifying integrity key point. The invention overcomes the influence on the system by node pairs, overcomes the insufficient evaluation which takes the node degree as a single index, effectively identifies the integrity key point and effectively guides the design of the navigation satellite system.

The invention discloses a method and device for predicting a ground leakage fault of a train power supply system. The method steps include: S1, acquiring an output signal during the movement of the train power supply system in real time; S2, determining whether there is a ground leakage fault according to the output signal obtained in real time, obtaining the target parameter value of the power supply system in a current fault state and the fault data of the target parameter value if so, otherwise obtaining the degradation data of the target parameter value; and S3, obtaining the fault data and the degradation data of the target parameter value in a specified time period, obtaining the degradation trajectory of the target parameter value, and evaluating the degradation trajectory of the target parameter value to predict the fault state of the power supply system. The device includes a signal acquisition module, a fault real-time monitoring module and a fault status prediction module. The method is easy to achieve, can monitor the ground leakage fault of the power supply system and predict the fault state, and achieves high prediction precision.

The invention discloses an amplitude and phase correction device and method suitable for a three-dimensional camera sonar array signal receiver. The device mainly comprises a signal source, an analogreceiving module, a signal acquisition and storage module and a signal processing module. The device is characterized in that when amplitude and phase correction is carried out on the three-dimensional camera sonar array signal receiver, the signal source simultaneously inputs in-phase signals with the same amplitude and the frequency being the sonar working frequency in all channels of the analogreceiving module; the analog receiving module carries out low-noise amplification, time gain control and band-pass filtering processing on the signals and outputs the signals to the signal acquisition and storage module; the signal acquisition and storage module carries out A/D sampling, converts the signals into digital signals and stores the digital signals in sampling data; and the signal processing module reads the sampling data, carries out complex demodulation and digital low-pass filtering, obtains amplitude differences between different receiving channels and reference channels by calculating effective values of all channel data, obtains phase differences between all the channels and the reference channels through a complex cross-correlation algorithm, and compensates the amplitude and phase differences.

The invention relates to a flow detection device for a bathroom. The flow detection device is arranged in a pipeline of a bathroom accessory and connected with a power generator in the pipeline of thebathroom accessory, and comprises a controller, a power supply circuit and a flow detection circuit; the power supply circuit is connected with the power generator and the controller separately for supplying electric power generated by the power generator to the controller, the flow detection circuit includes a frequency signal collection circuit for collecting rotation frequency signals of the power generator and a voltage signal collection circuit for collecting output voltage signals of the power generator, the frequency signal collection circuit and the voltage signal collection circuit are both connected between the signal collection end of the power supply circuit and the signal input end of the controller, and the flow detection circuit obtains the rotation frequency signals and voltage signals of the power generator and transmits the signals into the controller so that the controller can obtain water flow data through calculation. By means of the flow detection device for thebathroom, the obtained water flow data is more accurate, and the utilization rate of the power generated by the power generator is high.

The invention discloses a fog drop sample image processing method in a camera, and relates to the field of spray quality detection. The method comprises the following steps: collecting a sample imageuniformly irradiated by a light source; intercepting a sampling image with uniform brightness from the sample image; carrying out bilateral filtering noise reduction and graying processing on the sampled image; correcting the distortion of the grayed image; carrying out adhesion fog drop judgment on the corrected image; fully segmenting and marking the judged adhered fog drops by adopting a morphological opening operation segmentation method; separating and counting the fog drops by using a watershed segmentation method, binarizing the corrected grayscale images, classifying and summing the binary images to obtain the total number of pixel points of the fog drops and the total number of pixel points of the sampled images, and substituting the total number into a formula to calculate a spray quality parameter. The problems that in the prior art, an instrument is heavy, a fogdrop image segmentation algorithm is inaccurate, and noise pixel points reduce result accuracy are solved. The method can be widely applied to real-time analysis of spraying quality in severe field environments.

The invention discloses a partial excitation loss fault identification method for pulse loss. According to the partial excitation loss fault identification method, a partial excitation loss fault of pulse loss is identified by analyzing and comparing differences among six pulse waves in one cycle of an excitation voltage, and during comparison, 10% of maximum data is removed from data of each pulse wave, so that the influence of excitation voltage peaks on calculation during rectification and commutation can be reduced; and 50% of minimum data is removed, therefore, sampling data of a normal pulse wave without lost pulse is basically a relatively large positive value, an average value cannot be reduced or even close to 0 due to the excitation voltage negative value part, the average valueof the pulse waves of the lost pulse is generally close to 0 or smaller than 0, a large difference exists between the average value and the pulse waves without the lost pulse, and the difference is used for identifying the excitation loss fault of the partial excitation loss fault.

The invention relates to a CT (Computed Tomography) spatial resolution measurement method for improving a disc card method and belongs to the technical field of processing of images. The method comprises the following steps: S1, installing and starting a CT detection device; S2, carrying out fan-beam ray scanning; S3, rebuilding a two-dimensional image of a disc card by using a fan-beam FBP algorithm according to the obtained projection data; S4, obtaining a denoised grayscale image by using an RTV algorithm; S5, calculating the rebuilt two-dimensional image and the denoised gray image by using the improved disc card method to obtain an average gray value of the edge of the disc card; and S6, calculating ERF and PSF curves to obtain an MTF curve and a CT spatial resolution of the disc card. When the CT spatial resolution measurement method is used for calculating the edge gray value of the disc card, the influence of noise and insufficient card disc production accuracy on calculation of the edge average gray value is reduced, so that a more accurate and reasonable result can be obtained when calculating spatial resolution of a CT system.

The invention discloses a rotor grounding protection direct current quantity anti-interference algorithm which is used for sampling direct current quantity pre-low-pass filtering such as direct-current voltage and leakage current of a rotor winding and comprises the following steps: firstly, acquiring N sample direct current voltage and sample leakage current before switching of an electronic switch, wherein N accounts for 20-40 percent of the total sample S after low-pass filtering; grouping the sample direct-current voltage and the sample leakage current, calculating the average value of each group of samples, calculating the change rate of the average value of the sample leakage current, sorting the change rates, selecting the average value of the sample leakage current with the minimumcorresponding change rate, and finding out the average value of the sample direct-current voltage at the same moment; and calculating the grounding resistance and the grounding position of the rotoraccording to the average value of the sample leakage current with the minimum change rate and the average value of the direct current voltage at the same moment, thereby effectively reducing the influence of fluctuation and interference on the calculation of the direct current quantity, adding dC voltage drastic fluctuation locking discrimination and effectively reducing the influence of drastic fluctuation and interference on protection logic discrimination.

The invention discloses a coupon limit dynamic adjustment method based on user behavior characteristics, and the method comprises the following steps: 1, collecting and recording operation variables of a user for a commodity in a time period t, and carrying out the combination of the operation variables to form a user behavior weight in the time period t; 2, calculating a user adjustment weight according to the coupon shopping information of the user; 3, performing clustering analysis on the personal information of each user to obtain a feature weight of each user; 4, calculating an incentive score of each user according to a formula in combination with the behavior weight, the adjustment weight and the feature weight; and 5, dividing the incentive scores and the preferential quotas of the users into M grades, enabling the divided incentive scores and preferential quotas to be in one-to-one correspondence, and finally obtaining preferential quota grades of the users. According to the invention, the incentive level of the user is calculated according to the behavior characteristics of the user, and different preferential quotas are obtained according to the level, so that different users can obtain different preferential quotas of the same kind of coupons, and personalized user incentive is formed.

The invention discloses a rotor grounding protection DC amount mean value calculating data selection method. The rotor grounding protection DC amount mean value calculating data selection method selects a data window by using the principle that the standardized data variance is the smallest for each DC amount involved in the calculation, selects a data window with less fluctuation and less interference, and removes (1-T%)/2 of the data of the maximum and the minimum for the selected data windows, and selects the intermediate T% data to calculate the average value of the DC amount, so as to further reduce the influence of fluctuations and interference on the calculation of the average value of the DC amount.