34results about How to "Reflect the distribution" patented technology

The invention relates to an unbalanced data sampling method in an improved C4.5 decision tree algorithm. The method comprises the steps as follows: firstly, initial weights of various samples are determined according to the number of various samples; the weights of the samples are modified through the training result of the improved C4.5 decision tree algorithm in each round; the information gain ratio and misclassified sample weights are taken into account by a division standard of the improved C4.5 algorithm; the final weights of the samples are obtained after T iterations; the samples in minority class boundary regions and majority class center regions are found out according to the sample weights; over-sampling is carried out on the samples in the minority class boundary regions by an SMOTE algorithm; and under-sampling is carried out on majority class samples by a weight sampling method, so that the samples in the center regions are relatively easily selected to improve the balance degree of different classes of data, and the recognition rates of the minority class and the overall data set are improved. According to the unbalanced data sampling method in the improved C4.5 decision tree algorithm, weight modification is carried out through the improved C4.5 decision tree algorithm; and over-sampling and under-sampling are specifically carried out according to the sample weights, so that the phenomena of classifier over-fitting, loss of useful information of the majority class and the like are effectively avoided.

The invention discloses a method and an apparatus for classifying an image. The method comprises the following steps of: extracting a group of characteristics from an image to be used as a characteristic vector, wherein the extraction comprises the following steps of: determining a plurality of first regions arranged along a direction of a first shaft and a plurality of second regions arranged in a direction of a second shaft crossed with the first shaft for each characteristic of the characteristic vector; calculating a first difference between pixels and/or mean values of the plurality of first regions and a second difference between pixels and/or mean values of the plurality of second regions; calculating a gradient size and a gradient direction according to the first difference and the second difference to form each characteristic; and classifying the image according to the extracted characteristic vector.

A dual-waveband infrared instrument for diagnosing mastopathy is composed of a hand-held laser source consisting of single-source laser probe and multi-source laser probe with multiple laser heads and their positions regulator, an infrared detecting system consisting of optical splitting system, near infrared camera head and far infrared camera head, and a data processing and computer control system.

The invention provides an automatic generating method for a three-dimensional virtual city. The automatic generating method comprises the steps that trunk roads are generated according to an initial city road network in a land use model and land use intensity; block roads are generated according to the trunk roads and the land use intensity; three-dimensional buildings are generated according to the block roads; the three-dimensional virtual city is generated according to the trunk roads, the block roads and the three-dimensional buildings. In the automatic generating method for the three-dimensional virtual city, on the basis of the initial city road network provided by the land use model, the trunk roads are generated according to the land use intensity, the block roads and the three-dimensional buildings are further generated by utilizing the trunk roads and the land use intensity, and then the three-dimensional virtual city corresponding to the land use model is generated. Besides, the distribution conditions of the roads and the building in the city can be reflected comprehensively, accurately and visually.

The invention discloses an image processing method and system for detecting geometric impurities in tire rubber. The method includes: step 1, obtaining a binarized image of the tire rubber; step 2, according to different characteristics of rubber pixel points and carbon black pixel points value, obtain the connected domain of carbon black pixels in the binarized image; step 3, calculate the area A and width d of each connected domain; step 4, calculate each connected domain according to the area A and width d of each connected domain The slender length S of the connected domain; Step 5, compare the calculated slender length of each connected domain with a preset threshold, and the connected domain whose slender length exceeds the preset threshold is a geometry with carbon black pixels in the binarized image shape impurities. The beneficial effects of the present invention are: processing the binarized image by using the slender length of the connected domain, the geometric impurities such as straight lines, circles and curves in the binarized image can be identified and image-segmented, and the obtained binarized image is more accurate. It can reflect the distribution of carbon black in rubber.

The invention belongs to the field of aero-engine air system design, and particularly relates to a probability analysis method for functional parameters of an air system, which comprises the following steps: 1, determining an input variable; step 2, obtaining input variable measurement data; step 3, analyzing correlation between input variables; 4, analyzing input variable distribution; 5, analyzing input variable distribution; step 6, calculating functional parameters of the air system; step 7, analyzing the functional parameter probability of the air system; and 8, analyzing the sensitivity of the functional parameters of the air system. According to the probability analysis method for the air system function parameters, air system function parameter probability distribution analysis considering uncertainty is achieved for the first time, and compared with an air system function parameter limit value obtained in the prior art, the distribution condition of the air system function parameters under the actual condition can be better reflected, and an important basis can be provided for carrying out various types of analysis on the functional parameters of the air system under actual conditions.

The invention discloses a method for risk assessment and safety detection and evaluation of atmospheric pressure storage tanks, and the method comprises the following steps of S1, screening of storagetanks; S2, risk ranking: ranking all storage tanks according to the RBI risk assessment results, determining key storage tanks that require key detection and evaluation, and preliminarily determiningcorresponding detection methods for key storage tanks based on the assessment results and field condition limitations; S3, detection of the key storage tanks; and S4, solving of storage tank problems: using finite element analysis to determine the anti-settling stability and fatigue life prediction of the key storage tanks. The method firstly carries out RBI risk assessment of the storage tanks,sorts all storage tanks to determine the key storage tanks that require key detection and evaluation, inspects and evaluates the key storage tanks by using the method of acoustic emission detection ortank opening inspection according to the characteristics of the key storage tanks, and determines the anti-settling stability of the key storage tanks by finite element analysis.

The invention discloses a universal integrated collection system and method for resource and environment load data. The acquisition system comprises a data acquisition module, a data communication module and a data processing module, and by integrating sensors such as dust, noise, power, volatile organic compounds, sulfur dioxide, nitric oxide, carbon monoxide, ozone and an electronic weightometer, integrated acquisition of various data is achieved, and data processing is conducted on the acquired data; the acquisition method comprises the following steps: setting gas index concentration sampling points, and arranging sensors at points to be measured; determining an index quantity needing to be collected in real time; after the data indexes are stable, data acquisition is carried out; missing value compensation and data abnormal value elimination are carried out on the data, and weighted calculation is carried out on the measurement data after data missing value compensation and abnormal value elimination processing. According to the invention, the problems of low collection efficiency and inaccurate collection result of related indexes of a process workshop at present can be solved.

The invention provides a double-layer drainage asphalt pavement quality detection method which comprises the following steps: step 1, sampling and incinerating an asphalt mixture before pavement laying to determine the asphalt consumption; the remaining mineral aggregate after incineration is screened, and the grading composition of the mixture is detected; 2, taking the asphalt mixture on site, forming a Marshall test piece, and testing performance indexes of the Marshall test piece; 3, the bituminous mixture is subjected to actual pavement laying, and after laying is completed, sampling is conducted to detect the pavement performance; and 4, carrying out pavement core drilling sampling on the detection road section, and carrying out industrial CT scanning on the core sample to obtain theinternal pore characteristics and spatial distribution of the core sample. The detection method disclosed by the invention can be used for timely and effectively monitoring the material consumption deviation in construction and providing timely feedback for mixture mixing; meanwhile, on-site construction links are managed and controlled, and the construction quality is guaranteed; the test resultis accurate and reliable, and practical guidance is provided for further optimization of the grading design of the permeable asphalt concrete.

The invention discloses a portable ECT (Electronic Computed Tomography) system with an image correction function. The portable ECT system comprises a detachable ECT sensor module used for being mounted on a pipeline to be detected, a data acquisition unit is connected behind the ECT sensor system, the data acquisition unit is used for primarily processing acquired data by a data processing and correction module, and the image is further corrected according to the sensitivity provided by a cloud platform sensitivity online computing server connected with the 5G communication module or a local sensitivity database memory, and the corrected image is displayed by a local image display system. According to the invention, the nonlinear problem and the portable application problem of the electrical tomography technology can be effectively improved, and effective phase flow pattern measurement can be carried out on pipelines within a certain range.

The invention relates to a method for detecting the graphic precision of a laser-doped SE battery. The method comprises the following steps of: S1, respectively measuring the PT value of laser rays, the spacing value of the laser rays, an edge distance offset value, a diagonal difference value, a single-side offset value of each laser rays and the center line offset value of each laser ray of a to-be-detected battery piece; and S2, respectively judging whether the six parameters measured in the previous step meet respective corresponding precision requirements or not, the precision requirements being determined according to the design size of the pattern of the SE battery. According to the method provided by the technical schemes of the invention, distribution of all the laser rays can be accurately reflected by detecting the single-side deviation values of the laser rays and the center line deviation values of the laser rays, silicon wafers with abnormal SE battery patterns are found in time before a printing process, the situation that batch overprinting is poor during printing alignment due to the fact that silicon wafers with insufficient precision enter the printing process is prevented, and the production of screen printing graph and SE graph overprinting defective products is effectively avoided, and the economic benefits are improved.

The invention discloses a storage unit threshold voltage reading method and device, electronic equipment and a storage medium, and the method comprises the following steps: programming a whole chip, and writing checkerboard data; adjusting the internal reference voltage of the chip, and recording the number of the storage units with data change information in the adjustment process of the internal reference voltage of the chip and the addresses of the corresponding storage units; drawing a threshold voltage distribution diagram according to the numerical value of the reference voltage and the number of the storage units with the data change information under the corresponding reference voltage. According to the storage unit threshold voltage reading method provided by the embodiment of the invention, the distribution condition of the threshold voltage in the storage unit in the chip can be intuitively and accurately reflected, and the data retention capability of the chip can be visually represented.

The invention provides a method for designing working resistance of a solid filling hydraulic support, and belongs to the technical field of filling coal mining mechanical equipment. According to themethod, a mechanical model of interaction of a top plate, a bracket and a filling body of a multi-span statically indeterminate structure is adopted; joint numerical model, an overlying strata stresscurve is used as a top plate load; the load distribution of the top plate is reflected more truly. The mechanical relationship between the support and the supporting acting force of the filling body is more accurate, the supporting strength of the front top beam and the rear top beam of the hydraulic support of the double-top-beam structure can be obtained according to different elastic foundationcoefficients of the filling body and different top plate deformation limiting conditions, and compared with an existing support design method, the method is more scientific and reasonable.

The invention discloses a porosity inversion method of a time-difference sound speed logging curve, and belongs to the technical field of coal exploration. The method comprises that data of the time-difference sound speed logging curve is obtained by utilizing a geophysical prospecting means; practical quantified data of the time-difference sound speed of rocks is exported according to the time-difference sound speed logging curve; according to the practical time-difference sound speed quantified data, the porosity is inversed via a formula; according to porosity inversion data, AutoCAD is used to reconstruct a porosity curve; and the porosity curve is analyzed to deduce the growth degree of rock cracks. The logging reversion method commonly used in crack growth in the field of petroleum exploration is applied to the coal field, the logging is utilized fully, the rock porosity in certain depth range of a mining area is reflected more effectively, the crack growth condition is represented more visually to certain extant, and a new idea and method are provided for research on crack growth in the coal field.