462 results about "Distribution law" patented technology

The invention relates to a method for detecting lane lines of complex road conditions, which includes: firstly converting colored road images into gray level images and performing median filtering, then performing area of interest division on the filtered images, performing binaryzation processing on road images in the area of interest by using a symmetrical local threshold dividing method so as to extract lane line feature information, and finally performing parallel recognition respectively on a left lane line and a right lane line by adopting a decorrelation random sample consensus algorithm (RANSAC) method based on extracted lane line feature point distribution law. Compared with a traditional RANSAC method, the method for detecting lane lines of complex road conditions greatly improves instantaneity, remarkably strengths robustness and can fast and effectively detect lane lines in terms of complex road conditions caused by illumination changing, shadow shielding, lane line damaging, splotch covering and the like.

A network driving environment integrated perception model based on a convolutional and hollow convolutional structure simultaneously realizes object detection and semantic segmentation. A video imageof a road environment is shot through a forward-looking camera system mounted on a vehicle. The residual network model is used to get the bottom feature map of the image. The converged network is designed, which includes two sub-modules: object detection and semantic segmentation. The two modules share the bottom feature map. Among them, the object detection module is responsible for predicting the confidence level of the object frame and the category, and the semantic segmentation module is responsible for predicting the pixel level of each category. The appropriate loss function is selectedfor each of the two modules, and the perceptual model tends to converge in both modules after alternate training. Finally, the joint loss function is used to train the two modules simultaneously to get the final perceptual model. The model can simultaneously complete object detection and semantic segmentation with small computation amount, and the perceptual model uses a large amount of data of object detection to assist the semantic segmentation module to learn the image distribution law.

The invention proposes a steelmaking-continuous casting scheduling method utilizing a priority policy hybrid genetic algorithm. The method comprises the following steps: establishing a production scheduling plan target function; establishing a constraint condition set; performing iterative calculation on the target function by utilizing the priority policy hybrid genetic algorithm; and calculating decision variables. The calculation of the decision variables specifically comprises the steps of performing model initialization; calculating a feasible solution: designing a segmented combined real number code composed of casting time information of a continuous casting machine and information of a heat machining device, randomly generating operation time according to a distribution law, and obtaining a time conflict-free scheduling plan by backward inference calculation and conflict elimination methods; and performing population genetic optimization: quantitatively describing a matching relationship among machining devices in reality by using processing weight assignment of a task executable device, and introducing the matching relationship to genetic operation in the form of a device selection priority policy to perform population evolution. The method can solve the problem of uncertainty of device selection and operation time in production to obtain an optimized executable production scheduling plan.

An FB distribution rule 20 determines an actual vehicle actuator operation control input and a vehicle model operation control input such that a difference between a reference state amount determined by a vehicle model 16 and an actual state amount of an actual vehicle 1 (a state amount error) approximates to zero, and the control inputs are used to operate an actuator device 3 of the actual vehicle 1 and the vehicle model 16. In the FB distribution law 20, when an actual vehicle feedback required amount based on the state amount error exists in a dead zone, then an actual vehicle actuator operation control input is determined by using the required amount as a predetermined value. A vehicle model manipulated variable control input is determined such that a state amount error is brought close to zero, independently of whether an actual vehicle feedback required amount exists in a dead zone. This enhances linearity of a control system and also enhances the robustness against disturbance factors or changes therein while carrying out operation control of an actuator that suits a behavior of an actual vehicle as much as possible.

The invention provides a virtuality-reality integrated three-dimensional display method, which is used for displaying virtual network elements and real entity objects in a geographic information system (GIS) to form a composite space. The display method includes the following steps: (1) the network elements are GISized; (2) the composite space is three-dimensionally visualized. The invention also provides another display method and a display system. The invention is used for virtually and omnidirectionally displaying virtual network elements in a real space and accurately obtaining the spatial positions and attribute information of the virtual network elements for the research of the distribution law and movement law of the virtual network elements.

Provided is a vehicle-mounted GPS space-time trajectory big data optimization method and system. The vehicle-mounted GPS space-time trajectory big data optimization method comprises the steps that map matching processing is performed according to original GPS trajectory data and a road-grade traffic network, and a road section corresponding to each trajectory point is determined; the initial optimization rate is determined according to a GPS trajectory error distribution law; preliminary optimization is performed based on density and includes the preliminary optimization performed according to a density evaluation model and the initial optimization rate, so that the trajectory points gathered on the road surfaces are selected, and the points away from the road surfaces are regarded as drifting points to be abandoned; secondary optimization is performed based on angle similarity and includes similarity value calculation conducted on preliminary optimization results by utilizing an angle similarity evaluation model, clustering processing is performed, then the results having highest similarity values in clustering categories are selected as final optimization results, and selected high-quality data are obtained. By means of the vehicle-mounted GPS space-time trajectory big data optimization method and system, overall positioning accuracy based on low-accuracy vehicle-mounted GPS trajectory big data is improved. The optimization method is simple and easy to achieve.

The invention discloses an AFC data-based public transport passenger flow OD real-time estimation method. The method utilizes card swiping data of a public transport AFC system, performs sorting to obtain bus station passenger flow, on this basis excavates a mapping relation between public transport passenger flow OD and the bus station passenger flow, and builds a public transport passenger flow OD real-time estimation model based on Kalman filtering, thereby realizing real-time estimation of the public transport passenger flow OD. The AFC data-based public transport passenger flow OD real-time estimation method provided by the invention performs deep excavation on passenger flow data of the public transport AFC system, and obtains an optimal estimation value of a public transport passenger flow distribution probability through the Kalman filtering model, thereby realizing real-time estimation of the public transport passenger flow OD, which is of great importance to accurately describing the public transport passenger flow demand distribution law and realizing real-time optimization of bus route operation dispatching.

The invention relates to a lower limit method for the reservoir forming physical property of an effective reservoir layer under the restraint of reservoir forming power and pore structure, comprising the following step of: on the basis of building a function relationship between the tension force of an oil-water interface and the temperature of a stratum system, building a function relationship between the semidiameter and the permeability of a maximum through hole throat of a reservoir layer, and building a function relationship between K / phi and K of the reservoir layer under the restraint of hole throat structure type, computing the reservoir forming physical property lower limit of the effective reservoir layer under the restraint of the reservoir forming power and the pore structure under different stratum temperature conditions. The research result of the method is combined with the research result of the physical property recovery and the reservoir forming power recovery of the reservoir layer in the geological historical period, so that the effectiveness of the reservoir layer in the geological historical period can be evaluated, and the oil gas distribution law in the reservoir forming period can be analyzed, therefore, the method has an important meaning for guiding the oil-gas exploration and deployment.

The invention discloses an orchard pest and disease damage general investigation system and method based on UAV (unmanned aerial vehicle) remote sensing. The method comprises steps as follows: hyperspectral image data and RGB images of fruit tree canopies are collected by a UAV, and the collected hyperspectral image data is corrected; spectral reflectance, crown breadth, blade density and textural features of corresponding band images of the fruit tree canopies are calculated according to the corrected hyperspectral image data; size, shape and color features of fruits on fruit trees, color distribution law of the fruit trees and a three-dimensional view of orchard earth surface are calculated according to the RGB images; the spectral reflectance, the blade density and the textural features of the spectral images of the fruit tree canopies as well as the size and the color features of the fruits are input into a discrimination model, and identification of pest and disease damage of the fruit trees in the orchard is realized. Based on the hyperspectral images and the RGB images with more abundant information, the demand of high-frequency regular monitoring for insect situation information of the orchard can be met, and the general investigation efficiency of the pest and disease damage of the orchard is improved.

The invention relates to a landslide space-time prediction method. A landslide catastrophe process of a research position is subjected to space-time prediction by combining a physical conception-based hydrological model and a Bishop slope stability analysis method, and the landslide space-time prediction method is a landslide prewarning and forecasting method which combines soil mechanics, hydraulics and computer numerical simulation. The landslide prediction method has the advantages that a landslide deterministic model and the physical conception-based hydrological model are integrated effectively, the space-time distribution laws of the basin slope stability under the action of rainfall and the qualitative and quantitative relationship of the space-time distribution laws are analyzed, and the basin landslide is analyzed and predicted rapidly and effectively. By adopting the Bishop slope stability analysis method and simultaneously considering the spatial distribution difference in water content of soil mass, the precision and the applicability of the model are improved. The landslide prediction method provided by the invention is suitable for the landslide prediction of a simple side slope, a complex side slope and a basin scale.

A control input for operating an actual vehicle actuator and a control input for operating a vehicle model are determined by an FB distribution law based on a difference between a reference state amount determined by a vehicle model and an actual state amount of an actual vehicle such that the state amount error is approximated to zero, and then an actuator device of the actual vehicle and the model vehicle are operated based on the control inputs. The FB distribution law determines a control input for operating the model such that a state amount error is approximated to zero while restraining a predetermined restriction object amount from deviating from a permissible range. A vehicle control device capable of enhancing robustness against disturbance factors or their changes while performing operation control of actuators that is as suited to behaviors of an actual vehicle as possible is provided.

The invention discloses a detection method of two dimensional nanostructured material film surface detect distribution and relates to the field of material surface defect and its distribution detection. The method comprises the following steps of: firstly drying a sample beforehand, removing water molecules on the surface of the sample, allowing a gas mixture of steam and other gases to flow on the surface of the sample by a steam injection device, carrying out condensation operation, acquiring morphological, distribution and quantitative features of droplets on the surface of the film material under single and multiple condensation conditions, and determining distribution law of the sample surface defect. As the absorption and condensation process of the water molecules on the surface ofthe two dimensional film material is usually completed in milliseconds, the defect detection is very rapid; in addition, deionized water vapor which is mild is used on the surface of the material, and the sample can be restored to the original state just by a subsequent drying step. Therefore, the method is lossless for the sample to be detected.

The invention relates to a single-trap level favorable region optimal selection method under multi-geological factor quantitative constraints. The method includes the following steps that: based on a high-resolution stratigraphic framework, a sedimentary unit-level isochronous sedimentary stratum can be formed through division in a longitudinal direction; five kinds of two-sandstone section oil and water splitting models can be established according to well logging data, mud logging data, well testing data, well logging interpretation oil and water data and seismic data; single-sand body-level oil and water splitting can be realized under the sedimentary unit-level isochronous stratigraphic framework; a single-sand body-level sub-layer roof surface structural graph, a sedimentary microfacies planar graph and an oil and water distribution law can be compiled; oil-containing traps can be determined according to river boundaries, faultage, micro-amplitude structures, and sandstone updip wedge out lines; trap area superimposition and trap thickness accumulation are performed on the oil-containing traps in different single-sand body-level sub-layers; favorable region evaluation indexes of trap type, reservoir porosity, reservoir depth, organic matter type, maturity, hydrocarbon-generating intensity and source-reservoir-cap combination are combined together so as to build an oil reservoir forming element probability and oil reservoir forming matching probability double-factor favorable evaluation model; and favorable regions can be selected optimally.

A power distribution method for electric vehicle driven by two power sources includes executing a first power distribution law by a vehicle control unit, wherein the first power distribution law sets a distribution ratio according to a vehicle speed of the electric vehicle, a torque demand, and a current gear of a transmission. The powers outputted by first and second motors are adjusted according to the distribution ratio such that the first and second motors can operate at the most efficient operation points to increase the operational efficiency of the electric vehicle driven by two power sources.

Provided is a webpage text classification method based on feature selection. Firstly, data sets formed by a large number of webpages are divided into a training set and a testing set; secondly, different weights are endowed to labels according to webpage content expression capacity of information in a webpage label domain, and the weights (the product of a normalized word frequency and an inverse document frequency) of feature words in each webpage in the training set are calculated; on the basis of the obtained weights and through the combination of an intra-class distribution law and inter-class deviation, feather vectors of all webpages in the training set are calculated, and accordingly feature vectors of all classes in the training set are calculated; finally, the word frequency of feature words in each webpage in the testing set is calculated, the similarity between a webpage to be classified and each class in the training set is calculated, the class with the maximum similarity serves as the class where the webpage to be classified belongs, and a classification result is obtained.

The invention discloses a method for predicting the plane distribution law of the effective natural fissures of an oil reservoir and belongs to the field of natural fissure research. The method comprises the following steps: according to the conventional fissure recognition and description technologies, preliminarily determining the characteristic parameters of the natural fissures of the ultra-permeability dense oil reservoir; acquiring the preliminary distribution law of the natural fissures according to the finite element method; according to the characteristic parameters and the preliminary distribution law, building a reservoir-bed fissure geologic model; according to the sequence stratigraphy and the phase-controlled modeling technology, building a reservoir bed matrix geologic model; according to development conditions of the natural fissures in different matrixes, combining the reservoir bed fissure geologic model and the reservoir bed matrix geologic model and building a preliminary reservoir bed geologic model taking the natural fissures into consideration; on the basis of the preliminary reservoir bed geologic model taking the natural fissures into consideration, acquiring the plane distribution law of the effective natural fissures by adjusting the characteristic parameters to be fitted with production data. The method provides a reliable basis for optimizing a development technology policy.

The invention discloses a regional landslide risk evaluation method based on slope body units and machine learning. The method comprises the steps that first, to-be-evaluated slope units in a research region are divided according to a digital elevation model, and a landslide risk evaluation indicator system suitable for the research region is established; second, the distribution law of evaluation indicators in the slope units where a landslide happens is analyzed, and a landslide risk grading table of all the indicators is constructed based on the distribution law; and last, a landslide disaster risk evaluation model based on an LM-BP neural network is established, the degree of risk of each evaluation unit is calculated, and risk grading is performed. Through the method, calculation is precise, the principle is reliable, the operation process is easy and convenient to learn, and the method can meet the general requirement of regional landslide risk evaluation.

Disclosed is a beam steerable communication apparatus comprising a focusing element with a focal plane, a plurality of antenna elements disposed on the focal plane of the focusing element, a beamforming network which has a plurality of antenna ports and a plurality of beamforming ports, at least one radio frequency transceiver capable to control the phase of carrier frequency signals on its independent outputs, and a baseband modem, wherein each antenna port of the beamforming network is coupled to at least one antenna element, each beamforming port of the beamforming network is coupled to an independent transceiver output, and the phase distribution law formed on the beamforming ports is determined by a baseband modem control commands to the transceivers, said phase distribution law is to combine or distribute signal power on all beamforming ports from or to at least one antenna port of the beamforming network, wherein each radio frequency transceiver includes at least a phase shifter, a Tx / Rx switch, LNA, PA, up and down-conversion mixers, LO generation circuit, receive power combining and transmit power distribution schemes. The technical result of the invention is in the capability of providing efficient beam steering with high gain antennas with all the prospect features of phased antenna arrays, such as the ability to combine / distribute signal power between several RF amplifiers, thus, facilitating linearity requirements and increasing an output power of a transmitter. The present invention can be used, without limitations, in radio relay point-to-point communication systems, e.g. for forming backhaul networks of cellular mobile communication, in car radars and other radars, in local and personal communication systems, in satellite and intersatellite communication systems, etc.

The invention discloses a coalbed methane gas drainage testing method during the combined mining of multiple coalbeds. The testing method comprises a first step of previous preparation, a second step of gas absorption, a third step of gas drainage, a fourth step of finishing a first test and a fifth step of changing one testing parameter, such as the simulation mining depth or the advance speed of a model working face, remaining other testing parameters unchanged and repeating the second step, the third step and the fourth step. The method can simulate the coalbed methane exploitation in the spot working condition in a laboratory, reappear the change of a coal bed parameter in the actual working condition, study the changes of a stress field, a fracture field, a seepage field and a temperature field and analyze the spatial-temporal evolution and the distribution law of the gas flow of the coalbed and the fracture coupling of the coal and rock.

The invention discloses methods for building a power consumption factor model and estimating following-up driving range of an electric car. The model building method comprises the following steps of: introducing aggregate variable, calculating a comprehensive variable value of per second based on the instantaneous speed and acceleration of the electric car obtained in advance; forming a basic database by combining corresponding power consumption rate of per second; dividing a driving segment according to a certain time interval; counting the mean speed of each segment; merging the comprehensive variable value of the driving segment at the same mean speed; counting a distribution law, and calculating the average power consumption rate of per kilometer, namely a power consumption factor; and determining the final electric car power consumption factor model based on the mean speed by applying a mathematical statistic method according to the counted mean speed and the power consumption factor. The invention simultaneously discloses a method for estimating following-up driving range of the electric car.

The invention discloses a rapid simulation method of leakage of a high-sulfur natural gas gathering and transportation device for a marine gas field. Through analysis of leakage and diffusion distribution laws of high-sulfur natural gas in a densely populated area and optimization of a Gaussian diffusion model parameter selection method, a leakage and diffusion simulation procedure for sulfur-containing natural gas in a gathering and transportation pipeline and natural gas contained in a desulfurization device in a purification plant is acquired, and the gas leakage and diffusion result is rapidly simulated. A Gaussian model is selected, the influences of the ground roughness, the atmospheric stability and the leakage amount on gas leakage and diffusion are sufficiently considered, related parameters are corrected in combination with surroundings of a well field, the leakage and diffusion result under the four conditions of leakage of a sulfur-containing natural gas gathering and transportation pipeline, leakage of hydrogen-sulfide-containing pregnant solutions of the desulfurization device in the purification plant, leakage of hydrogen-sulfide-containing pregnant solutions of pipelines and gas leakage of pipelines of a station is rapidly and quantitatively simulated, and the longest distance that hydrogen sulfide with the concentration being 100 ppm and 1,000 ppm can reach after leakage and diffusion is obtained.

The invention provides a numerical data compression method for a column storage database, which includes: extracting one column of data from a data table, dividing each column of data into multiple subsidiary columns, counting characteristic values of data distribution laws in each subsidiary column, and compressing by selecting the corresponding lightweight compression algorithm according to the characteristic values of the data distribution laws, wherein different data in each subsidiary column occupies identical space. The numerical data compression method is simple in operation and applicable to compression of data of regular distribution and irregular distribution.

The invention discloses a power-saving control method, comprising the following steps of: determining parameters of a power-saving mode, wherein the parameters include at least one of the length of an initial dormancy window, the length of a final dormancy window, the length change law of the dormancy window and the distribution law of monitoring windows; and carrying out power saving control according to the parameters. Through the technical scheme of the invention, the mean value of delay of the service can be kept in the QoS (Quality of Service) range required by the service.

The invention discloses a piston heat load test device with a laser as a heat source, and a piston heat load test method. The device comprises a bracket, a control device, a heating device, a temperature sensing device, a monitoring device and a cooling device, wherein the heating device consists of a laser device and a beam shaper which changes a laser beam emitted by the laser device into a laser beam with a plurality of concentric ring-shaped spots and projects the laser beam onto a piston surface vertically. The method comprises the following steps of: firstly, optimizing a laser energy distribution ratio of areas on the piston surface according to the temperature field distribution of a piston under an actual working condition; secondly, designing an optical conversion plate according to the energy distribution ratio; thirdly, converting an incident beam into a concentric ring beam which meets the requirement on the laser energy distribution ratio of the areas; and fourthly, simulating the temperature field distribution law of the piston under the actual working condition by loading the beam on the piston. The device and the method can simulate the temperature field distribution on the piston surface under the actual working condition and have the advantages of short experiment period, high reliability and high controllability.

The invention discloses a drawing architectural element identification method and system based on machine learning. The drawing architectural element identification method is based on the method of combining a rule and machine learning. Rule matching and mathematical modeling are conducted on vector diagram metainformation obtained from a drawing, the feature distribution law of an architectural element in the drawing is strictly and accurately analyzed through the training process, the data sets of a primitive database, an architectural element rule base and an architectural element feature base are formed, the architectural element identification method in the probability sense is obtained, and the recognition rate of an architectural element and a property parameter of the architectural element are further improved. The drawing architectural element identification system based on machine learning tries to break through the limitation of an existing drawing identification method, provides an efficient drawing analysis and identification tool for a metro design and construction department, and has the three advantages of being high in precision ratio, identification efficiency and intelligent degree.

The invention discloses a stability dynamic evaluation method for a strain softening slope. The method comprises the following steps of (1) building a slope model and determining linear strain softening Mohr-Coulomb model parameters; (2) calculating an elastic initial stress field of the slope model and calculating an elastoplastic stress field of each unit of the slope model; (3) calculating a plastic shear strain value of each unit of the slope model; (4) calculating a shear strain increment of the slope model, and obtaining corresponding shear strength parameters of the slope model and a space-time distribution law of the shear strength parameters; (5) performing re-assignment on mechanical strength parameters of a rock-soil body of the slope, and analyzing a stability dynamic change process of the slope model; and (6) exporting a displacement distribution cloud chart, extracting coordinates, and performing fitting through a least square method in statistical analysis software, wherein a final fitting curve reflects a potential slip surface of the to-be-analyzed slope. According to the method, a dynamic change situation of a stability coefficient and the potential slip surface of the slope in a slope failure process can be obtained.

The invention provides an insulator string fault detection method based on infrared image temperature distribution law and a BP neural network. The method comprises the steps that the rectangular target area of an insulator string in an infrared image is extracted through manual and color space conversion; the length feature of the target area is calculated through image processing techniques such as binarization, Hough transform and rotation to extract insulator steel caps and disk surface areas respectively; an image temperature matrix and a segmentation result are used to calculate the average value of each steel cap and disk surface area as the feature quantity, and a K-Means clustering algorithm is introduced in the calculation process to remove background pixels; in the order of insulators from the low voltage end to the high voltage end, temperature eigenvectors of the steel caps and disk surfaces are formed respectively; and based on the temperature distribution law of the insulator string, an insulator string fault diagnosis model based on the BP neural network is established. The detection method provided by the invention has high recognition accuracy.

The invention discloses a court topology identification method. The method comprises: voltage fluctuation of a low-voltage side of a court transformer is made; a voltage of the low-voltage side of thecourt transformer and voltages of to-be-identified user ammeters are collected respectively to obtain a corresponding reference number sequence and a corresponding comparison number sequence; and a gray correlation analysis is carried out to determine a court topology and the like. The court topological identification method has the following beneficial effects: problems that lots of manpower andmaterials are consumed by using a court communication instrument and court crossing caused by using a carrier technology are solved; the court topology identification becomes accurate and convenient;and compared with the data analysis methods like the regression analysis and primary cost analysis of mathematical statistics, the gray correlation analysis method needs a few of samples and does notrequire that the samples conform to a typical distribution law and the computing load is reduced.

The invention relates to a synchronization active type islanding detection system and a method based on M sequence modulation and variance estimation. The detection system comprises a synchronization inverter, a synchronization switch, a voltage Hall sensor arranged on a synchronization communal join point, a digital signal processor connected with the synchronization inverter, the synchronization switch and the voltage Hall sensor. The detection method is to utilize the pseudo-random signal M sequence of the communication technique, modulate the modulation wave of the inverter according to the distribution law of the M sequence, so as to make the output have the characteristics of the M sequence, and thus the output signals also have the characteristics of random signals. During synchronizing, the voltage of the synchronization communal join point is controlled by a power network and the randomness is small; therefore the obtained variance value of the random sequence is very small; when islanding, the voltage of the synchronization communal join point is determined by the output of the inverter, the output of the inverter is modulated through the digital signal processor and have the characteristics of the random signals, so the variance value is significantly increased. During detecting, a variance threshold value is set, and islanding can be determined when the detected variance is more than the threshold value. The synchronization active type islanding detection system and the method has the advantages of simple arithmetic, low cost, small effects to harmonic waves of the output power and small islanding detection blind zone.