37results about How to "Fitting results are accurate" patented technology

The invention discloses a method based on monocular vision for detecting and roughly positioning the edge of a road, and relates to the field of machine vision and intelligent control. Aiming at a continuous road with different edge characteristics, two road edge detection methods are supplied and suitable for semistructured and nonstructured roads and can be applied to vision navigation and intelligent control over a robot. The invention provides a method for detecting the edge of the road based on colors and a method for detecting the edge of the road based on threshold value partitioning. On the basis of the obtained edge of the road, an image is subjected to inverted perspective projection transformation, so that a front view is transformed into a top view; and according to a linear corresponding relation between a pixel and an actual distance in the image, a perpendicular distance from the current position of the robot to the edge of the road and a course angle of the robot can be calculated. The method is easy to implement, high in anti-interference performance, high in instantaneity and suitable for the semistructured and nonstructured roads.

The two-dimensional image area positioning method based on the grating projection can determine the source and characteristic of image noise for noise inhibition. The method includes the following steps: taking the datum picture with CCD camera before projecting one coarse set or fine set of grating to the object before taking one image and obtaining coarse positioning image through threshold parting, image filtering, etc; filtering the image with fine grating image to obtain the fine grating grey image; non-interlaced traversing the pixels in the coarse positioning image and processing in compensation algorithm to divide the measured object space into narrow zonal areas and obtain the corresponding coarse positioning result image; and superposing the above boundary information onto all the fine grating grey images to be analyzed, non-interlaced traversing, performing quadratic curve fitting and final superposing the extreme points of all the fine grating grey images as the ultimate object space dividing result.

The invention belongs to the technical field of laser spectroscopy, and mainly relates to a calibration-free wavelength modulation spectroscopy gas detection method based on an S[2f] method. Corresponding scientific analysis shows that A, i[1], i[2], concentration and phase difference psi[1] and psi[2] have little influence on S[2f] / S[2f]<c> signals, and only WMRF has obvious influence on the S[2f] / S[2f]<c> signals. Therefore, under conditions of uncertain concentration and phase shift, WMRF is obtained by inverting the S[2f] / S[2f]<c> signals, and because fitting parameters are greatly reduced, the method can be guaranteed to be simple, rapid and accurate. On the basis, the concentration of to-be-measured gas is fitted, and a measurement value with higher accuracy degree is obtained. The new method is simple, rapid and accurate, and is suitable for wide promotion in the field of gas detection.

The invention relates to a rapid evaluation method for a state of health of an ex-service battery module of an electric vehicle. According the rapid evaluation method, the state of health (SOH) of thebattery is compared with voltage Lorenz dispersion sigma<-> in a discharging process, the correlation between the SOH and a sigma<-> value during the discharging process are analyzed, and the workload is not increased since an operating voltage during the charging and discharging processes of the battery can be acquired in real time by means of a battery management system without additional acquisition; the acquired operating voltage is not limited to a certain state of charge (SOC) value as long as it is in a corresponding SOC interval, and the method is more convenient; and the calculationof the voltage Lorenz dispersion sigma<-> is based on a result of voltage equalization in the SOC interval, and the calculation result is more precise. The rapid evaluation method can realize rapid evaluation of the SOH of an ex-service battery of the electric vehicle by just calculating the voltage Lorenz dispersion sigma<-> during the discharging process of the ex-service battery module to be detected, so as to carry out consistent rapid sorting, thereby achieving the purposes of simple, convenient and low-cost recycling of the ex-service battery.

The invention relates to an antenna reflection surface fitting method. The method comprises the following steps of: 1) by taking a paraboloid as a rigid body and an original point in a standard coordinate system as a reference point, describing the change of a position and a direction of the paraboloid by using rotary motion and translational motion in the standard coordinate system and a measurement coordinate system, and inputting the initial values of fitting parameters by selecting rotation angles (a, b), translation (Dx, Dy and Dz) and a focal length f of the paraboloid as the fitting parameters of the rigid body; 2) importing measured data of a measured point on the paraboloid; and 3) writing a parabolic equation in the format of a vector, converting a coordinate of the measured point in the measurement coordinate system into that in the standard coordinate system, working out the equation of the measured point which is converted to that in the standard coordinate system, substituting the measured point which is converted to that in the standard coordinate system into a standard equation of the paraboloid for comparison, working out an error equation, linearizing the error equation, and working out a parameter value according to a non-linear least square principle to obtain the optimum parabolic equation.

The invention belongs to the technical field of laser spectroscopy, in particular relates to a WMRF new model-based calibration-free wavelength modulation spectrum gas detection method, and solves the technical problem of low precision of a calibration-free wavelength modulation spectrum technology currently used for gas concentration detection. An etalon of a 0.02 cm<-1> free spectral range is used to analyze wavelength change WMRF caused by signal modulation, a first harmonic component with the amplitude linearly changed and a second harmonic component with an amplitude as a constant are observed, and a new model for describing WMRF is described. By using the new model, a system can acquire a more accurate fitting result; compared with a conventional system, hardware module addition and manual operation are not needed, only program improvement is needed, and development and application are facilitated. The method is simple in technical scheme and convenient in operation, consumption of lots of manpower and material resources is not needed, and the cost and convenience degree in use are easily accepted by most application departments.

The invention discloses a chest non-invasive blood pressure detection method based on pulse wave conduction time and belongs to the field of non-invasive blood pressure monitoring methods. The problemthat a time parameter, namely the blood pre-ejection period, is ignored in pulse wave conduction time PTT in traditional detection approaches is considered in the method, correspondingly the measurement position is fixed to an outlet of the thoracic aorta, the cardiac ejection period PEP is obtained through electrocardio signals ECG and aorta acceleration tracing signals APG, the influence of PEPin PTT is removed, the more accurate pulse wave conduction time PTT' is obtained, the detected conduction time is more proximate to the actual conduction time, and it is ensured that accurate aorta blood pressure can be obtained. Meanwhile, according to the method, by taking the defect that the accuracy of traditional non-invasive blood pressure detection is not ideal into consideration, a bloodpressure model is improved, and the heart rate HR, the pulse wave morphological parameter C and the skin conductance rho are added, so that a model fitting result is more accurate.

The invention provides a wind power scene generation method and system based on probability distribution, and the method comprises the steps of obtaining the historical data of a prediction value andan actual measurement value of wind power, and extracting the difference between the prediction value and the actual measurement value of the wind power, i.e., a prediction error; selecting a wind speed, a wind direction and a fitting error which directly influence wind power, and respectively fitting distribution of prediction errors under the influence of the wind speed, the wind direction and the fitting error; combining the prediction error distribution under the three influence factors according to corresponding weight parameters, taking the minimum residual sum of squares of a probability density function and a distribution histogram as an objective function, solving an optimal solution of the weight, and obtaining a prediction error probability density function; discretizing the probability density function to obtain a large number of random scenes, and selecting typical scenes; according to the invention, mixed distribution is adopted, weights are set for error distribution under different influence factors, the influence of the influence factors on the error distribution is comprehensively considered, and an obtained fitting result is more accurate than single error distribution.

The invention discloses a three-dimensional crosslinking network-based calculation method of dynamic mechanical parameters of an asphalt mixture. A three-dimensional network linking model is established according to stress-strain characteristics of different components, load test data are combined, corresponding dynamic mechanical functions of the asphalt mixture are obtained through fitting parameters of energy of all single-chain structures in the network linking model under different stress-strain load states, and then the dynamic mechanical parameters thereof are calculated and determined.According to the method, influences of the different components of the asphalt mixture are fully considered, three-dimensional crosslinking networks are respectively and independently established forthe different components in the asphalt mixture, and influences of the different components on the overall stress-strain characteristics of the mixture are evaluated thereby. According to the method,macroscopic performance characterization of the material under dynamic load action can be linked from the molecular chain energy of the microstructures through establishment of the crosslinking networks, expansion from a microlevel to a macrolevel can be realized, and fitting results are more precise. At the same time, dependence on the test data is also accordingly reduced, test costs are lower,and the method is suitable for promotion.

The invention relates to a wavelength calibration method for an atomic fluorescence spectrometer based on a digital micro mirror array, the method comprises the steps of: a basic function relation between positions of the digital micro mirror array and standard wavelengths is calculated; fitting errors are calculated, and the fitting errors are stored at a corresponding position in a ''wavelengtherror'' Column; for a sample to be tested, a user selects a pre-detection fluorescence wavelength to be measured, the pre-detection fluorescence wavelength is corrected by the wavelength errors storedin the ''wavelength error'' Column to obtain a calibrated wavelength, the calibrated wavelength is substituted into the basic function relation in step 2 for inverse computation to obtain a digital micro mirror inverted array so as to obtain a measurement wavelength of the sample to be tested to obtain an accurate fluorescence intensity value. The accurate fluorescence intensity value can be obtained by substitution of the calibrated wavelength into the basic function relation in the step 2 for inverse computation to obtain the position of the digital micro mirror inverted array for measuringthe measurement wavelength of the sample to be tested. The method can accurately measure the fluorescence intensity information of the wavelength position of the sample to be tested.

The invention relates to a device and a method for correcting an error of a triaxial magnetometer. The device at least comprises a satellite-borne calculation module (40) which is in communication connection with a star sensor (10), an orbit measurement assembly (20) and a triaxial magnetometer (30). The correction mode of the satellite-borne calculation module (40) is that nominal geomagnetic field parameters of a current position are calculated based on attitude measurement parameters sent by the star sensor (10) and orbit parameters sent by the orbit measurement assembly (20); and a first geomagnetic field parameter sent by the triaxial magnetometer (30) is fit into a corrected second geomagnetic field parameter based on a least square method by taking the nominal geomagnetic field parameter as a reference value. By generating the real-time magnetic compensation parameters, the errors caused by real-time changes of a satellite electromagnetic environment are effectively eliminated.

The invention relates to a dynamic rotating speed fine tuning method for ship speed optimization, belongs to the technical field of ships, and solves the problems that a model cannot be updated in real time, the calculation amount is large and the efficiency is low in the prior art. Comprising the steps of preprocessing ship navigation data according to a fixed time period; dividing the processed navigation data into a plurality of groups of data corresponding to different working conditions by using a clustering algorithm; based on a polynomial function, introducing an index forgetting factor, and establishing a function relationship between the marine fuel consumption and the host rotating speed in each group of data to obtain a plurality of fitting models; the method comprises the following steps: acquiring current navigation data, clustering to obtain an actual working condition, acquiring a minimum value point closest to a current main engine rotating speed in a current main engine rotating speed set range in the current navigation data according to a fitting model corresponding to the actual working condition, and if the sea oil consumption at the minimum value point is less than the current sea oil consumption obtained according to the current main engine rotating speed, determining that the sea oil consumption is not less than the current sea oil consumption. If so, adjusting the current host rotating speed to a minimum value point, otherwise, keeping the current host rotating speed unchanged. And dynamic rotating speed fine tuning is realized.

The invention relates to two die bodies withe the same inventive concept and discloses a hardening calibration method for an X-ray CT beam and based on the die bodies. One X-ray CT beam hardening calibration die body comprises an upper part and a lower part, the upper part is a measurement structure, the main body of the upper part is a cylinder, the outer edge of the bottom face of the cylinder is an Archimedean spiral, step-shaped circular holes are formed in the outer edge of the bottom face of the cylinder, the diameter of the adjacent layers of the circular holes is equal to or slightly larger than the difference of the maximum radius and the minimum radius of the spiral, and an outer spiral parameter and an inner layer number are corrected according to the actual measurement demands;the lower part is an auxiliary structure which is divided into a holding area, a centering area and a clamping area from top to bottom, and the surface of the holding area is rough, convenience is provided for operation; the centering area is a smooth cylinder used for conducting centering operation on a detector; in the operation process, the radius of the corresponding inner layer is selected according to the required penetration distance, firstly, the centering area in the auxiliary structure is placed in the center of a view field for centering, then the die bodies are downwards moved, the steps of the inner layer, corresponding to the selected radius, in the measurement structure are placed in the center of the view field for 360-degree scanning.

The invention discloses a method based on monocular vision for detecting and roughly positioning the edge of a road, and relates to the field of machine vision and intelligent control. Aiming at a continuous road with different edge characteristics, two road edge detection methods are supplied and suitable for semistructured and nonstructured roads and can be applied to vision navigation and intelligent control over a robot. The invention provides a method for detecting the edge of the road based on colors and a method for detecting the edge of the road based on threshold value partitioning. On the basis of the obtained edge of the road, an image is subjected to inverted perspective projection transformation, so that a front view is transformed into a top view; and according to a linear corresponding relation between a pixel and an actual distance in the image, a perpendicular distance from the current position of the robot to the edge of the road and a course angle of the robot can be calculated. The method is easy to implement, high in anti-interference performance, high in instantaneity and suitable for the semistructured and nonstructured roads.

The invention relates to an SDT improvement method applied to the field of numerical control machine tool monitoring. The SDT improvement method comprises the following steps: segmenting the internalstate original data of the numerical control system and performing standard SDT compression; performing function fitting of the compressed data by using the multi-model optimization method; saving thepoints having high influence on the compression precision according to the fitting error; and dynamically adjusting the tolerance of the next interval to be compressed according to the fluctuation state of the adjacent intervals. According to the method, the tolerance can be dynamically adjusted according to the fluctuation state of the data; function fitting is performed by using the multi-modeloptimization method; and the points having high influence on the compression precision are saved according to the fitting error. Therefore, the partially redundant machining state process informationof the machine tool can be deleted, and less machine tool state data can be stored without losing the effective information as far as possible so that the compression effect is better.

The invention discloses a mobile vibration testing device and a cable force testing method. The mobile vibration testing device comprises a body, a control unit and a storage unit, wherein a travelingunit and a testing unit are arranged on the body; the traveling unit is used for driving the body to travel; the testing unit comprises a vibration sensor for collecting vibration signals; the control unit is connected with the traveling unit, the testing unit and the storage unit; the control unit is used for controlling the traveling unit, the testing unit and the storage unit; the storage unitis used for storing vibration signals. The cable force testing method comprises the following steps: starting the testing device; driving the body to travel along a cable member by the traveling unit, continuously testing vibration of the cable member by the testing unit in the traveling process and sending the vibration to the control unit; sending the testing time and the vibration signal to the storage unit by the control unit for storage; identifying constraint conditions at two ends of the cable member according to the signals to obtain cable force of the cable member.

A calculation method of dynamic mechanical parameters of asphalt mixture based on three-dimensional cross-linking network. According to the stress-strain characteristics of different components, a three-dimensional network model is established. Combined with the load test data, the energy of each single chain structure in the network model is The parameters under different stress-strain loading states are fitted to obtain the corresponding dynamic mechanical functions of the asphalt mixture, and then the dynamic mechanical parameters are calculated and determined. The invention fully considers the influence of different components of the asphalt mixture, independently establishes three-dimensional cross-linking networks for different components in the asphalt mixture, and evaluates the influence of different components on the overall stress-strain characteristics of the mixture. Through the establishment of the cross-linking network, the present invention can realize the expansion from the microcosm to the macrocosm from the molecular chain energy connection of the microstructure to the macroscopic performance characterization of the material under the action of dynamic load, and the fitting result is more accurate. At the same time, the dependence on test data is correspondingly reduced, the test cost is low, and it is suitable for promotion.

The invention relates to an antenna reflection surface fitting method. The method comprises the following steps of: 1) by taking a paraboloid as a rigid body and an original point in a standard coordinate system as a reference point, describing the change of a position and a direction of the paraboloid by using rotary motion and translational motion in the standard coordinate system and a measurement coordinate system, and inputting the initial values of fitting parameters by selecting rotation angles (a, b), translation (Dx, Dy and Dz) and a focal length f of the paraboloid as the fitting parameters of the rigid body; 2) importing measured data of a measured point on the paraboloid; and 3) writing a parabolic equation in the format of a vector, converting a coordinate of the measured point in the measurement coordinate system into that in the standard coordinate system, working out the equation of the measured point which is converted to that in the standard coordinate system, substituting the measured point which is converted to that in the standard coordinate system into a standard equation of the paraboloid for comparison, working out an error equation, linearizing the error equation, and working out a parameter value according to a non-linear least square principle to obtain the optimum parabolic equation.

The invention relates to a rapid evaluation method for a state of health of an ex-service battery module of an electric vehicle. According the rapid evaluation method, the state of health (SOH) of thebattery is compared with voltage Lorenz dispersion sigma<-> in a discharging process, the correlation between the SOH and a sigma<-> value during the discharging process are analyzed, and the workload is not increased since an operating voltage during the charging and discharging processes of the battery can be acquired in real time by means of a battery management system without additional acquisition; the acquired operating voltage is not limited to a certain state of charge (SOC) value as long as it is in a corresponding SOC interval, and the method is more convenient; and the calculationof the voltage Lorenz dispersion sigma<-> is based on a result of voltage equalization in the SOC interval, and the calculation result is more precise. The rapid evaluation method can realize rapid evaluation of the SOH of an ex-service battery of the electric vehicle by just calculating the voltage Lorenz dispersion sigma<-> during the discharging process of the ex-service battery module to be detected, so as to carry out consistent rapid sorting, thereby achieving the purposes of simple, convenient and low-cost recycling of the ex-service battery.

The invention discloses a chest non-invasive blood pressure detection method based on pulse wave conduction time and belongs to the field of non-invasive blood pressure monitoring methods. The problemthat a time parameter, namely the blood pre-ejection period, is ignored in pulse wave conduction time PTT in traditional detection approaches is considered in the method, correspondingly the measurement position is fixed to an outlet of the thoracic aorta, the cardiac ejection period PEP is obtained through electrocardio signals ECG and aorta acceleration tracing signals APG, the influence of PEPin PTT is removed, the more accurate pulse wave conduction time PTT' is obtained, the detected conduction time is more proximate to the actual conduction time, and it is ensured that accurate aorta blood pressure can be obtained. Meanwhile, according to the method, by taking the defect that the accuracy of traditional non-invasive blood pressure detection is not ideal into consideration, a bloodpressure model is improved, and the heart rate HR, the pulse wave morphological parameter C and the skin conductance rho are added, so that a model fitting result is more accurate.

The invention relates to a method for analytic calculation of statistic characteristics of disturbance motions. Specific parameters of self-correlation and power spectral density functions of carrier motions through the steps of turbulence information extraction, self-correlation sequence calculation, self-correlation function fitting, power spectrum density calculation and the like are quantitatively given out, the power spectrum density of external turbulence frequency points can be directly calculated, the shortcoming that in the prior art, severe power spectral density peak and jumping cause inaccurate description of disturbance characteristics is overcome, quantified description of the carrier motion disturbance characteristics is achieved, the accuracy is high, and data is smooth. By adopting the method, the disturbance characteristics of the carrier motions can be accurately, simply and quantitatively worked out by adopting the numerical computation method.

The invention relates to an improved SDT method applied in the field of monitoring of numerical control machine tools, comprising: segmenting the original data of the internal state of the numerical control system and performing standard SDT compression; adopting a multi-model optimization method to perform function fitting on the compressed data; According to the fitting error, the points that have a great impact on the compression accuracy are saved; according to the fluctuation state of the adjacent interval, the tolerance of the next interval to be compressed is dynamically adjusted. The method can dynamically adjust the tolerance according to the fluctuation state of the data; use the multi-model optimization method to fit the function; and save the points that have a great influence on the compression accuracy according to the fitting error. In this way, part of the redundant machine tool processing state process information is deleted, and less machine tool state data is stored while ensuring that effective information is not lost as much as possible, which has a better compression effect.

The invention relates to two phantoms with the same inventive concept, and provides an X-ray CT beam hardening calibration method based on the phantoms. One of the X-ray CT beam hardening calibration phantoms includes upper and lower parts, the upper part is a measuring structure, the main body is a cylinder, the outer edge of the bottom surface of the cylinder is an Archimedes spiral, and there is a stepped circular hole in the center. The diameter of the round hole in the adjacent layer is equal to or slightly larger than the difference between the maximum and minimum radius of the spiral. The parameters of the outer spiral and the number of inner layers are modified according to the actual measurement requirements; the lower part is an auxiliary structure, which is divided into a holding area from top to bottom , Centering area and clamping area, the surface of the holding area is rough and easy to operate; the centering area is a smooth cylinder, used for detector centering operation; during the operation, select the corresponding inner layer radius according to the required penetration distance , first place the centering area in the auxiliary structure in the center of the field of view for centering, and then move down the phantom to place the inner step corresponding to the selected radius in the measurement structure in the center of the field of view for 360-degree scanning.

The invention discloses a circle detection method based on edge detection and fitted curve clustering. The method includes the following steps that first, the edge of a grey-scale map is detected, edge curves with the pixel number smaller than a pixel threshold are removed, and then a curve set T is obtained; second, the gradient direction vector of each edge point in the curve set T in the grey-scale map is calculated and recorded; third, candidate curve marking and parameter fitting based on edge gradient direction constraint are performed, wherein in the third step, firstly, estimated circle centers and estimated radii are calculated, secondly, the direction vectors of the edge points of each curve to the corresponding estimated circle center are calculated, thirdly, candidate points are marked based on the gradient direction constraint, fourthly, a candidate curve set is determined, and fifthly, circle parameters of the candidate curves are calibrated; fourth, the candidate curves are clustered and fused, and then the fused circle center and the fused radius are obtained; fifth, the detection result is output according to the completeness degree of the circle. The method has the advantages of being good in adaptability, high in calculation speed, good in noise immunity, little in resource consumption and capable of being widely applied in the field of circular curve detection.

The invention discloses a mesoscale test method for determining the salt diffusion coefficient of bottom mud in a dredging reclamation area, which belongs to the field of environmental technology. Several groups of bottom mud with the same thickness and test platforms for salt release from bottom mud at different water levels are built; The sensor is placed in the bottom mud, and the water body conductivity sensor is placed in the middle of the bottom mud-water interface and the water body. Using the calibrated correlation curve between the conductivity and salt content of the soil and water body in the research area, the monitored conductivity data is converted For the salinity, use GMS software to build a numerical model corresponding to the test platform, and obtain the molecular diffusion coefficient of salt release from mesoscale sediment and the molecular diffusion coefficient of salt in water by fitting the salinity-time series data . The present invention comprehensively builds a mesoscale test platform for the salt diffusion coefficient of mesoscale sediment and corresponding numerical models, monitors the release and migration of salt in sediment and water in real time, and obtains the salt in sediment and water through numerical model fitting Diffusion coefficient, high precision.

The invention discloses a method for measuring the temperature of a divertor target plate of a tokamak device based on single-wave filtering, and relates to the technical field of infrared temperature measurement and magnetic confinement nuclear fusion. The radiant energy measured by the thermal imager is at the wavelength λ 0 The relational expression under the 0 The relationship between the temperature of the lower divertor target plate and the emissivity, and finally use the gray value of the image measured by the thermal imager during the discharge process of the tokamak device to calculate the real temperature of the divertor target plate at this time; the present invention uses a single The wave filter reduces the influence of the wavelength range on the temperature measurement accuracy, and also takes into account the transmittance of the filter, the attenuation of radiation in the special optical path, and the different temperature conditions of the tokamak device. The relationship between combined temperature and emissivity is more accurate than the linear model and the constant model.