72 results about "Numerical fitting" patented technology

A tomography reconstruction method of fluorescence molecules based on a linearity relation belongs to the technical field of the image reconstruction in biomedicine. The reconstruction method is characterized in that the method sequentially comprises the following steps: 1. the outside surface outline is obtained by picture processing and numerical fitting based on the map data of a mouse used in experiment; 2. the outside surface outline is led into a finite element software package for meshing; 3. every excitation light source, and finite element grid nodes which are corresponding to corresponding fluorescence monitoring points are found; 4. aiming at one given excitation light source sk, a linear relation matrix Wsk between fluorescence intensity which is detected from surface and unknown fluorescent probe distribution is established; 5. aiming at all excitation light sources, a general linear relation matrix is generated; 6. after the linear relation is obtained, a algebra reconstruction method is used for iterating the fluorescent probe distribution which approaches the reality. The method is characterized in that non-uniform optical parameters can be handled and the precise linear relation between the fluorescence intensity which is detected from surface and the unknown fluorescent probe distribution is rapidly established.

A device and method of detecting surface plasmon resonance for sensing molecules or conformational changes in molecules with high resolution and fast response time is disclosed. Light from a light source (14) is focused through a prism onto a metal thin film (15) on which sample molecules to be detected are adsorbed. The total internal reflection of the laser / incident light is collected with a differential position or intensity sensitive photo-detecting device instead of a single cell or an array of photo-detectors (12) that are widely used in previous works. The ratio of the differential signal to the sum signal of the differential position or intensity sensitive photo-detecting device (12) provides an accurate measurement of the shift in the surface plasmon resonance angle caused by the adsorption of molecules onto the metal films (15) or by conformational changes in the adsorbed molecules. The present invention requires no numerical fitting to determine the resonant angle and the setup is compact and immune to background light, The methods and sensors of this invention can be used in numerous biological, biochemical, and chemical applications such as measuring subtle conformational changes in molecules and electron transfer reactions can be studied.

The invention discloses a wind power plant optimal control method based on a wake flow effect. Based on a Jensen wake flow model, wake flow air speed influence factors are analyzed, the wind turbine thrust coefficient is finally determined, and power output of a wind turbine can be optimized; and in particular, according to the wind turbine wake flow characteristic, a wind turbine wake flow speed stacking model is built, a numerical fitting method is then used for conducting numerical fitting on the wind turbine thrust coefficient and the wind-power utilization coefficient, fitting results are joined with the Jensen wake flow model, a wind turbine output power model is refined, and an optimal model of the wind turbine output power model is obtained. By means of the wind power plant optimal control method, a visualization method can be used for detecting change conditions of the wind power plant total output power at different wind speeds after a control optimization algorithm is added, and the effectiveness of the control algorithm is indicated compared with the total output power in the natural state.

The invention discloses a high-precision satellite-borne orbit prediction method based on numerical fitting. The method comprises the following steps: firstly, obtaining atmospheric environment parameters including the latest solar radiation flow and a geomagnetic index, satellite star parameters and the latest precise orbital elements of the satellite, combining the star and space environment parameters, carrying out orbit prediction by utilizing a ground high-precision numerical method so as to obtain a set of satellite instantaneous orbital elements in a prediction arc section, then, converting into an average element to perform numerical fitting so as to obtain a change item coefficient of each orbital element, arranging multiple sets of primary orbits to carry out segmented extrapolation according to an satellite-borne predication requirement, carrying out orbit prediction according to the arranged initial orbital element and the change item coefficient to obtain the average element of the satellite at the prediction moment, if the prediction precision requirement is high, arranging a plurality of sets of initial values for staged prediction, converting the average element ofthe satellite at the prediction moment into the instantaneous element, so that the instantaneous position of the satellite is obtained. According to the method in the invention, the problem that the precision of satellite-borne orbit prediction is poor in the prior art is solved.

The invention provides a method for calculating the oil content in an oil tank of an unmanned plane, belonging to the field of unmanned plane flight data detection. The invention is characterized by comprising an engine rotating speed sensor and the corresponding curve of the engine and oil consumption measured by an experiment method in advance. According to the engine rotating speed measured by the engine rotating speed sensor and oil consumption acquired through numerical fitting and corresponding to the current rotating speed, the consumed oil content after the engine of the unmanned plane is started can be obtained through integral. According to the primary oil content before the engine is started, the residual oil content in the oil tank can be obtained. The calculating method can obtain the residual oil content in the oil tank without other sensors according to the corresponding relationship of the engine oil consumption and the rotating speed for realizing measurement, i.e. the engine rotating speed and working time.

A polynomial complete homomorphic encryption method based on the coefficient mapping transformation. A plaintext is expressed as a polynomial consisting of a set of random values, two sets of random coefficient factors and a random constant of a specified mapping function, and in the polynomial: the expression and a set of random coefficient factors of the specified mapping function are taken as a key; another set of random coefficient factors, a set of random arguments and random constants of the mapping function are taken as the ciphertexts for homomorphic operations, so that the part of function key performs three different mappings and then undergoes numerical fitting to obtain the family of operational support functions consisting of three sub-functions respectively, which are used to perform the homomorphic operation of the ciphertext based on the family of operational support functions and return to the locality for decryption by the key.

The invention discloses a non-contact non-destructive detection method of the depth of a steel surface hardened layer, which is characterized by comprising the following steps: (1) a laser beam which is modulated is radiated on a sample to be measured, an infrared detector is used for receiving the thermal radiation of the sample, and a measured value which is changed along with the modulation frequency of the incident laser beam of a thermal radiation signal is obtained; (2) a profile curve of the thermal conductivity which is changed along with the depth is obtained by adopting the forward numerical fitting algorithm according to the obtained measured value of step (1); (3) and the depth of the hardened layer of the sample to be measured is determined according to the limit value of the change of the thermal conductivity. The non-contact non-destructive detection method of the invention adopts the thermal radiation technology and obtains the depth of the surface hardened layer of the steel sample by the reconstruction of the thermal physical property depth profile; the experiments verify that the obtained reconstructed curve of the thermal conductivity has great consistency with the microhardness curve.

The invention provides a modeling method for a converter valve saturable reactor iron core loss circuit model, which comprises the following steps: exerting impulse excitation at two ends of a saturable reactor; measuring the terminal voltage of the saturable reactor and the current passing the saturable reactor; obtaining the relationship between the resistance of the iron core and the magnetic flux density of the iron core through strict mathematical derivation and reasonable approximation hypothesis; and establishing the functional relationship between the resistance of the iron core and the magnetic flux density of the iron core through numerical fitting. According to the invention, the functional relationship between the resistance of the iron core and the magnetic flux density of the iron core is established through numerical fitting; as the nonlinear resistance is adopted to present the loss change process of the iron core, the actual physical process can be reflected accurately; and meanwhile, the linear function is adopted to present the resistance of the iron core, so that the fitting result is highly accurate as well, the physical conception is clear, and the modeling method is suitable for engineering application.

The invention provides a wireless sensor network node selection method based on blind source separation. The method comprises the following steps: determining the relation between node average observation noise and blind source separation performance through a numerical fitting method based on the two factors of observation noise and residual energy of a sensor node; modeling a node selecting problem into a combination optimization problem based on comprising of the node residual energy and the blind source separation performance, and solving the optimization problem on a sink node by adopting a heuristic algorithm based on convex optimization to determine a selected node; transmitting an acquired mixed signal from the selected node to the sink node, and performing blind source separation at the sink node to recover each independent source signal. By adopting a node optimization selection algorithm, the node residual energy can be effectively increased, the life cycle of a wireless sensor network is prolonged, and meanwhile excellent signal separation performance is ensured.

The invention discloses a method for constructing high-temperature plastic flow densification constitutive models of powder sintering materials. The method comprises a step of carrying out numerical fitting analysis by taking the flow stress and instant density data of a powder sintering material as basis, so as to establish a nonlinear constitutive relationship model among a high-temperature flow stress sigma, an instant density rho, a deformation temperature T and a strain rate xi of the powder sintering material. The model is suitable for the plastic processing technologies such as hot rolling, hot extrusion, hot forging and the like of powder sintering bodies with ingredients of pure iron, carbon steel and alloy steel and suitable for the mould design; and the model not only can accurately represent the high-temperature plastic flow densification behaviors of the powder sintering materials, but also can accurately predict the flow stresses and instant densities of the powder sintering materials under different hot processing conditions, has favorable stability and can be applied to the plastic processing technologies such as hot rolling, hot extrusion, hot forging and the like of powder sintering materials and the computer-assisted design of moulds.

The invention discloses a measurement method of the depth of a metal surface hardening layer with a curve surface, which is characterized by comprising the following steps of: adopting a photothermal radiation measurement method for measurement, respectively obtaining the amplitude of the surface temperature field of the sample before and after hardening and the measurement value of frequency variation relative to modulated light; then carrying out self normalization processing to obtain normalized amplitude and phase value corresponding to the frequency value; utilizing plan theory and adopting a forward numerical fitting algorithm for fitting so as to obtain a profile curve with thermal conductivity changing along the depth; and the hardening depth is analyzed by the profile curve of the thermal conductivity. The measurement method obtains the depth of the surface hardening layer of the metal sample, thus conveniently realizing non-destructive measurement to the thickness of the surface hardening layer of the metal sample with the curve surface.

The invention provides a piecewise extrapolation policy-based method for estimating a dynamic response of a mooring system. For a mooring positioning cable rope containing a composite component, the influence of factors such as gravity, tension, ocean current force, elastic elongation and the like of a mooring line is considered, and based on a piecewise extrapolation thought, the dynamic response of the whole mooring system is solved. Each component section is divided from a mooring point; then each component section is subdivided into different anchor chain units according to a discrete numerical method; in combination with a hydrostatic equation, a tension and direction span of the mooring line of each unit is calculated; and finally superposition is performed to obtain a total vertical span, and the total vertical span is compared with water depth, so that static characteristics of the whole mooring line can be obtained on the premise of meeting a precision condition. System excitation is given; a static characteristic curve of the mooring line is subjected to numerical fitting by utilizing spline interpolation; and force synthesis is performed by an included angle relationship between the distributed mooring line and a ship bow, so that the dynamic response of the mooring system is realized.

The invention discloses a dynamics modeling method for a multiple-input-multiple-output gear transmission system. The dynamics modeling method comprises the following steps of S1, dividing the multiple-input-multiple-output gear transmission system into a plurality of sub transmission systems; S2, performing rigid body modeling on components in each sub-transmission system, and assembling the components to form a sub-system model; S3, simplifying the coupling into a multi-degree-of-freedom spring with rigidity and damping, and connecting the subsystem models through the coupling to form an overall model of the multi-input multi-output gear transmission system; s4, performing multi-body dynamics simulation analysis on the overall model to obtain discrete data of wave power and fluctuation torque at the multi-degree-of-freedom spring; s5, performing numerical fitting on the discrete data to obtain a time-varying fluctuation function; and S6, establishing a centralized mass dynamic modelof each subsystem, and adding the wave function into a degree-of-freedom equation of a node corresponding to the centralized mass dynamic model to obtain a decoupled subsystem dynamic model. The method is simple in modeling and high in solving speed.

The invention discloses a downhole flexible composite pipe service life prediction method, which can effectively evaluate service lives of downhole flexible composite pipes to provide a theoretical basis for the string inspection cycle of the downhole flexible composite pipe water injection process. The main steps are as follows: different test blocks are chosen, and test wells to be tracked regularly are determined; according to different service times, samples are collected at different positions of wellbores, and after sampling, strings are tripped into the original wells; indoor tensile resistance, bursting and hydrostatic pressure are tested on the extracted sample pipes; numerical fitting is conducted on the same test item of the same well at different service times to draw a fitted curve; according to water injection string performance requirements, service lives corresponding to corresponding values, i.e. the service lives of the downhole flexible composite pipe water injection strings, are determined on the fitted curves.

The embodiment of the invention discloses a method of correcting a height coefficient formula of a flame center. Through a numerical simulation, a flue gas temperature of a furnace outlet and height coefficients of the flame center more accurate than those obtained from a traditional thermodynamic calculation can be obtained, an empirical formula of the height coefficients of the flame center in the traditional thermodynamic calculation is taken as a template, numerical fitting is conducted on the height coefficients of a first flame center obtained by the numerical simulation, and a calculation formula of the height coefficients of the flame center more adapted to an existing oxygen-enriched combustion boiler but similar to the traditional empirical formula is obtained. The method is of great significance to the thermodynamic calculation under the condition of oxygen-enriched combustion boiler. The method solves the technical problem that the traditional empirical formula of the height coefficients of the boiler flame center cannot obtain accurate height coefficients of the existing oxygen-enriched combustion boiler center.

The invention discloses an experimental method for quickly measuring a wheel-track adhesion-creeping curve. After test starting, servo-driven motors drive two rotating shafts to make same-speed rotation at a set simulation speed, wherein the creep rate is kept to be 0; a normal load is applied to a wheel sample shaft downwardly and thus the wheel and track samples make rolling contact, wherein interms of the interface contact stress, a field wheel track interface contact stress is horizontal. During the follow-up experiment process, a track sample shaft drive motor keeps a stable rotating speed and a wheel sample shaft drive motor carries out adjustment according to a set creep rate s changing curve; and the creep rate between the wheel and track samples is controlled to change accordingto a set curve, and an adhesion coefficient between the wheel and track samples during the creep rate changing process is measured and recorded. According to the method disclosed by the invention, a complete adhesion-creeping curve can be obtained in one experiment process; compared with the traditional method, the provided method enables the experiment time to be reduced effectively; the measuredcurve is uniform and continuous and contains more curve changing details; and the curve feature analysis can be carried out by using a numerical fitting method effectively.

The invention discloses an online Duvel whole temperature interval cold loss measuring method. By means of the method, a function relation between Duvel cold end temperature change, radiation heat leakage and transfer heat leakage is utilized to conduct numerical fitting on a temperature change curve of the cold end, and a cold loss expression can be exported in a reversed mode. The method overcomes the shortcoming that radiation heat leakage and transfer heat leakage cannot be obtained respectively in the common liquid nitrogen weighting method or the standard refrigerator calibration method, only a total cold loss value can be obtained, and measurement cannot be conducted in the state where a refrigerator is coupled. The method achieves the effect that radiation heat leakage and heat transfer heat leakage can be measured simultaneously under the condition that the original coupling state is not damaged, heat leakage of the Duvel cold end under all temperature points can be obtained, and the method provides an effective measuring means for heat design and heat failure analysis of Duvel.

Provided is an incoming flow parameter determining method applicable to an appearance of a body of revolution. The method comprises the steps of 1, building a surface pressure approximation model applicable for the appearance of the body of revolution, wherein firstly a surface pressure computational formula is determined, and surface pressure is expressed as a sum of a product of incoming flow dynamic pressure qc and a pressure coefficient Cpi and incoming flow static pressure p infinity; then a polynomial form is adopted to express the pressure coefficient Cpi, and polynomial factors are a flight attack angle alpha, a sideslip angle beta and an incoming flow pressure ratio R; finally the factors in the polynomial are obtained through numerical fitting, regression or a least square method to surface measuring point pressure of the body of revolution in multiple sets of states; 2, obtaining the surface measuring point pressure of the body of revolution in a flight experiment, conducting back calculation according to the surface measuring point pressure combining the approximation model, and accordingly obtaining an incoming flow parameter. According to the incoming flow parameter determining method applicable for the appearance of the body of revolution, the model can be used for an embedded air data system, and the prediction accuracy of the system can be effectively improved.

The invention discloses a method for high-accuracy measurement of a linear refractive index of a material, and belongs to the fields of photonics materials and optical information processing. The method can be used for performing high-accuracy measurement on the linear refractive index of an ultra-thin material, and comprises the following steps of: replacing a phase object in the conventional 4f system with a sample to be measured, and replacing a nonlinear sample to be measured in the conventional 4f system with a known optical nonlinear standard sample; performing numerical fitting on acquired reference faculae and nonlinear faculae, wherein only the linear refractive index of the sample to be measured is variable in the numerical fitting; and changing the value of the linear refractive index of the sample to be measured to ensure that simulated nonlinear faculae are inoculated with experimental nonlinear faculae, so that the linear refractive index of the material can be acquired very conveniently with a high accuracy. According to the method, an optical path is simple, data processing is performed conveniently, and the test speed is high, so the method can be widely applied in the research fields of measurement of the linear refractive index or thickness of the material, the photonics materials, the optical information processing and photonics devices.

The invention discloses a time-frequency second-order cross ambiguity function calculation method based on numerical fitting. The time-frequency second-order cross ambiguity function calculation method comprises the following steps: setting time delay and frequency search parameters of initial coarse estimation; Obtaining a time delay difference initial estimation value and a frequency differenceinitial estimation value according to the time delay and frequency search parameters in combination with a second-order cross ambiguity function; carryinbg out Data sampling in the sampling range, andthe sampling frequency and a corresponding second-order mutual fuzzy function value are recorded to form a data set; Performing curve fitting by utilizing a radial basis function according to the data set to obtain a fitting function; In the sampling range, obtaining a frequency value corresponding to the maximum value obtained by the fitting function; Wherein the time delay difference initial estimation value and the frequency value are a time delay difference final estimation value and a frequency difference final estimation value of two paths of signals; According to the method, the symmetry of the second-order cross-ambiguity function in the frequency domain is utilized, the complexity of time-frequency joint estimation of the second-order cross-ambiguity function is reduced through double estimation of coarse estimation and fine estimation and a numerical fitting method, and fine estimation of the frequency difference is completed.

The invention discloses a PIV flow field recovery method based on deep transfer learning, and the method combines a small part of experiment data with a large part of numerical fitting data, and completes the recovery process from local experiment flow field data to global experiment flow field data. The method has the advantages that when the PIV technology is used for measuring the flow field, good global flow field distribution can be obtained only through reconstruction according to the experimental result of the local flow field, the application range of the PIV technology is expanded, and the use technical threshold of the PIV is reduced; compared with a traditional data processing method, the method has the advantages that deep migration learning is adopted, the accuracy of PIV flowfield reconstruction is further improved by means of a large amount of numerical fitting data, and although the training time is long, the calculation speed of the trained migration network is far higher than that of the traditional data processing method.

The invention discloses a quality index soft measurement method and device of a caprolactam product. The method comprises the following steps: establishing a computation mechanism mathematic model about physical-chemical indexes of the caprolactam; computing according to the computation mechanism mathematic model to respectively obtain analysis numerical value about the product quality index of the caprolactam entering a distillation tower; performing numerical fitting and online correction on the computation mechanism mathematic model to acquire the corrected soft measurement computation result about the product quality index of the caprolactam entering the distillation tower. After establishing the computation mechanism mathematic model, the physical index and the chemical index of the caprolactam can be computed in real time, and the computation result can be corrected to obtain the soft measurement computation result about the product quality index of the caprolactam, the aim of acquiring the product quality information of the caprolactam entering the distillation tower through real-time computation is realized.

The invention provides a new method for determining a virtual machine migration opportunity. The method comprises the following steps: when any resource utilization rate of a virtual machine reaches acertain value, starting a prediction module; performing numerical fitting according to the resource utilization rates corresponding to an interval fixed time length within the closer time before themoment; deducing the function to obtain a function corresponding to a change slope; solving a change value of the function in a future period of time to obtain a change trend of the utilization rate;if the change trend is always increasing, determining as the virtual machine migration opportunity; if the change trend is in a decreasing state, continuously monitoring the utilization rate; if the change trend is in a fluctuation state, calculating an integral value of a slope function in the future period of time to obtain the change value of the utilization rate; and comparing the change valuewith a set threshold value to determine the migration opportunity. After the prediction stage, if the utilization rate is still increasing and reaches a limit value, the duration time of the utilization rate reaching the limit value is counted as a judgment condition of virtual machine migration. According to the method, by predicting the change trend of each utilization rate of the virtual machine, a tight resource is located; corresponding judgments are made according to different change trends to implement accurate positioning of the migration opportunity.

The invention discloses a method for utilizing a mathematic dominant model to solve solar cell physical parameters. Through experiment data, two mathematic model parameters of open-circuit voltage and short-circuit voltage of a photovoltaic panel are directly read, then through a numerical fitting method, two another parameters are solved, therefore the mathematic model is obtained, through the mathematic model parameters, a maximum power point and a current voltage slope are solved, through the parameters, an equation set related to physical parameters is obtained, and five physical parameters are solved. The method for utilizing the mathematic dominant model to solve the solar cell physical parameters has the advantages that the problem is solved that in the physical parameter solving process, the maximum power point and the current voltage slope which are obtained through the experiment data are not precise, and meanwhile the problem of the large calculated amount brought by simultaneous fitting of the five parameters in a traditional physical parameter solving process is solved.

The invention discloses a roadbed underlying hidden hole critical stability calculation method for a loess area. Firstly, a calculation model is established according to the physical properties of thehole diameter of the loess hidden holes, and the yield stress is calculated; a power equation is established, circular frequency of the vibration type is calculated and a modal analysis result is combined, and then the maximum equivalent stress under a dynamic load condition is calculated; numerical calculation is carried out by changing buried depth, and the relation between the maximum equivalent stress and the stable burial depth is obtained, the yield stress in the model is substituted, and the stable burial depth of the hidden hole in the model soil is obtained; and the center distance is changed in sequence, the stable burial depth is calculated, and the result is subjected to numerical fitting to obtain a roadbed underlying hidden hole stability discriminating curve of the loess area. The method combines the advantages of flexibility of numerical analysis and integrity and continuity of analytical solution, and hole diameter and center distance double-variable critical burial depth regression prediction is carried out on the stability of the loess hidden hole under the dynamic load, so that the method is high in speed, high in accuracy, easy to popularize and capable of well meeting the requirements of specific engineering.

The invention provides a calculation method of a thermal power plant heating load, and belongs to the fields of thermal power plant heating design and computer technology. According to the method provided by the invention, different engineering instances are classified into different cases such as engineering instances with same figures at different regions and engineering instances of different figures but at same regions; each case is analyzed separately to obtain a different calculation model; calculation complexity and accuracy of the models are considered during acquisition of the calculation models; a linear relationship between a product of a temperature difference between indoor and outdoor and an area, a volume or length and the heating load is acquired by using a method of numerical fitting, so as to acquire a plurality of calculation models; and thus when calculating a particular thermal power plant, a corresponding calculation model can be selected for calculation. According to the scheme provided by the invention, problems in the prior art that the calculation is complex, inaccurate, not dedicated to the thermal power plant and the like are solved; the calculation method provided by the invention has advantages of being simple and quick, being accurate in result, and the like.

The invention belongs to the technical field of temperature measuring and discloses a temperature measuring method based on an HF first overtone emission spectrum, for the purpose of solving the temperature measuring problem of the ultrasonic flow field of an HF chemical laser. The method comprises: using a near-infrared spectrometer to measure HF first overtone emission spectrum signals, and then using the vibration-rotation intensity distribution formula of the HF first overtone emission spectrum to carry out numerical fitting on obtained data so as to calculate the temperature of a measuring system. The method can be applied to temperature distribution testing of different positions (a jet pipe outlet, an optical cavity and an optical cavity downstream ) and different stages (non-laser output and laser output) of an HF chemical laser flow field and can be popularized to other systems generating high vibration state HF (V>=2). The method has the advantages of simplicity, high precision, good stability and non interference with a testing body.

The invention relates to a method for detecting a single atmospheric high-risk chemical leakage source based on terahertz. The method comprises collecting and disposing atmospheric high-risk chemicalconcentration information and corresponding spatial coordinates, carrying out numerical fitting on the concentrations to space coordinates in a computation system to obtain a spatial distribution function of concentrations, transmitting the spatial distribution function into an imaging system and showing a spatial distribution map of concentrations in the imaging system through different colors and different curve surfaces. The method realizes 3D display of toxic gas components, toxic gas concentrations and toxic gas flowing directions in the complex atmospheric high-risk chemical environmentin which rapid judgments are required, realizes continuous processing on the concentration data obtained by discrete measurement, produces a distribution map closer to the real situation of the site and fast and effectively establishes the people evacuation measures and poison source suppression method at the gas leakage site.

The invention discloses a rock long-term strength determination method based on numerical fitting, which comprises the steps of carrying out a conventional compression test to obtain rock compressivestrength and a total stress-strain curve, carrying out a graded loading creep test to determine a creep curve equation of a rock test piece and a final strain value when rock deformation tends to be stable, and taking a stress value corresponding to an intersection point of a final strain value fitting curve and the total stress-strain curve of the rock test piece as long-term strength. Accordingto the method, the intersection point of the final strain fitting curve and the total stress-strain curve is obtained through numerical fitting, the long-term strength of the rock is determined, the problems that inflection points are fuzzy, subjective judgment is difficult and the like in the long-term strength determining process are solved, and the method is simple, effective, accurate and reliable in result, rapid in process and programmable.

The invention discloses an evacuation traffic flow distribution model based on a traffic basic graph, being suitable for construction of large-scale activities, natural disasters or major accident emergency plans and emergency response when emergency events occur, and particularly relates to an evacuation traffic flow distribution model based on a traffic basic graph. The evacuation traffic flow distribution model includes the steps: analyzing a map into a Cartesian coordinate system, and determining an evacuation range circle layer according to an evacuation center point through an analysis method; performing traffic simulation on a specific area road network by utilizing a traffic simulation method and a numerical fitting method, and performing fitting to obtain a traffic basic map of akey road section on a corresponding evacuation circle layer; after the basic graph is obtained, establishing an unconstrained optimization model, and obtaining the maximum flow of the key road sectionthrough derivation; and finally, according to the obtained maximum flow, distributing the probability of selecting a destination by the evacuation traffic flow.