81results about How to "Short solution time" patented technology

The invention discloses a vision navigating method for a movement carrier based on the laser plane assistance. The method comprises the following steps: building a distance measurement and environment system, determining the position of a rectangular bright spot to be detected in an image, determining the center coordinate of the rectangular bright spot, obtaining a geometric distance measurement model, correcting the distance measurement model, segmenting the visual image, determining a local vectorization environmental space with distance information, positioning the position of the carrier per se, and building a global environmental space. Through the adoption of the method different from the traditional movement carrier autonomous navigation middle distance obtaining method, a laser plane is introduced as an auxiliary means, a single or a plurality of vision sensors are combined, the threadlike bright spot formed by the reflection of the laser plane on an object to be irradiated is fully utilized and taken as an cooperated distance measurement target, and the geometric relationship between the vision sensor and a laser light source is effectively utilized, so that the requirements for the precision, yield ratio and real-time performance of the distance calculating can be met. Therefore, the method is a reliable navigation method, and can be applied to a movement carrier autonomous navigation system.

The invention discloses a multi-quantile estimation method of sea clutter Weibull amplitude distribution parameters. The method includes the steps of firstly, acquiring the cumulative distribution function of a sea clutter Weibull amplitude distribution model according to the probability density function of the sea clutter Weibull amplitude distribution model; secondly, selecting three sample cumulative probability p1, p2 and p3, and acquiring three quantile equations related to the shape parameter beta and the scale parameter eta of the sea clutter Weibull amplitude distribution model according to the definition of quantile; thirdly, using clutter amplitude data to obtain estimation values x^p1, x^p2 and x^p3 of quantiles xp1, xp2 and xp3; fourthly, using the estimation values x^p1, x^p2 and x^p3 and the three quantile equations in the second step to obtain the estimation value beta^ of the shape parameter beta and the estimation value eta^ of the scale parameter eta. By the method applicable to target detection under the sea clutter background, interference of anomalous scattering units to samples can be reduced, and parameter estimation performance is increased.

The invention discloses a silicon carbide ceramic preform, an aluminum-based silicon carbide ceramic material, and a preparation method of the silicon carbide ceramic preform, and relates to the technical fields of materials and preparation thereof, and aims at solving the technical problems that the silicon carbide ceramic preform prepared by an existing preparation process is low in strength, uneven in structure, and low in volume fraction of silicon carbide. According to the main technical scheme, the preparation method comprises the following steps: mixing raw materials for preparing the silicon carbide ceramic preform evenly, so as to prepare a raw material slurry, wherein the raw materials for preparing the silicon carbide ceramic preform comprise silicon carbide, a dispersing agent, an organic monomer and water; introducing an initiator into the raw material slurry, and mixing evenly, so as to obtain ceramic slurry; carrying vacuum degassing treatment, injection molding and demolding drying on the ceramic slurry, so as to obtain a ceramic biscuit; and sintering the ceramic biscuit at 800-1200 DEG C, so as to obtain the silicon carbide ceramic preform. The preparation method is mainly used for preparing the silicon carbide ceramic preform with high strength and uniform structure, so as to prepare the aluminum-based silicon carbide ceramic material with high performance.

The invention relates to a self-navigation method for a planetary power descending branch, and belongs to the technical field of deep space exploration. According to the method, a descending sensor, a radio receiver and a navigation camera are combined; before a planetary landing task is started, a planetary orbiter is used for arranging a beacon with radio measurement and communication functions near a landing area; and a position of the beacon is determined by multi-time observation of the orbiter. In a planetary power descending stage, a radio receiving device arranged on a detector is used for measuring a position of the detection relative to the radio beacon; the position, the speed and the attitude of the detector are estimated through the descending sensor and the navigation camera; the navigation data calculation time is short; the demand for the self-navigation instantaneity is met; and the precision and the instantaneity of navigation are guaranteed.

The invention relates to a planet entering section navigation filtering method based on a layered atmosphere model and belongs to the technical field of deep space detection. The method provided by the invention takes an output of an inertial measurement unit (IMU) as an external measuring value on the basis of a dead reckoning navigation method and simultaneously introduces a multilayered atmosphere model to establish a planet entering section measurement equation and construct a state estimating device; filtering estimation is carried out on positions, speeds and direction information of a landing device; resolving time of navigation data is short and real-time requirements of self navigation are met; compared with a single dead reckoning navigation method, the influences on the state of the landing device by initial deviation can be corrected and the navigation precision is improved.

The invention discloses an improvement method of a three-carrier ambiguity resolution (TCAR) based on the addition of a graded small-sized search space under a short base line. The method disclosed by the invention comprises the following steps of: respectively adding a tiny search space in wide-lane and basic frequency-section ambiguity resolution determining phases according to a standard difference condition in a three-carrier phase linear combination process of a TCAR method; utilizing the length of the base line and structural configuration information to carry out primary screening; determining the unique integer ambiguity resolution through a residual minimum method; and finally, combining a dynamic carrier to solve a real-time posture of the carrier relative to an observation vector of a visible satellite and carrier wave single difference information. The method disclosed by the invention is applicable to positioning and determining the postures of signals of a three-frequency global navigation satellite system.

The invention relates to the technical field of display, in particular to a backlight module, a display device and electronic equipment. The backlight module comprises a rubber frame, a light guide plate, a light source and a photoelectric conversion module arranged departing from the light-outgoing side of the light guide plate, wherein the photoelectric conversion module is used for converting absorbed optical energy into electric energy. The backlight module can be used for converting the optical energy into the electric energy through the additionally arranged photoelectric conversion module, the electric energy can be conveniently supplemented under the condition that the electric energy lacks outdoors, and the service time can be prolonged.

The invention relates to a planetary power descending branch navigation method based on a TDS (total descending sensor) and image measurement and belongs to the technical field of deep space exploration. At a power descending branch, an inertial unit of a landing device navigation system is used for flight position recurrence and is a basic sensor; a Doppler radar on the TDS is used for directly acquiring distances from a landing device to the surface of a target planet in three wave beam directions and relative speeds of the landing device along the three wave beam directions and then calculating the altitude of the landing device and speeds of the landing device in three axial directions of a landing point fixedly-connected coordinate system; a navigation camera serving as a vision auxiliary navigation sensor is used for precisely measuring included angles between sight directions of different characteristic points of a landing area; the positions of the characteristic points relative to the landing point are known; the sight included angles contain horizontal position information of the landing device relative to the landing point, so that the horizontal position precision of the landing device is greatly improved.

The invention relates to a modeling and optimization method for a delayed coking process of residual oil. Based on a lump theory modeling method, raw materials are divided into four lumps of a saturated component, an aromatic component, colloid and asphaltene; products are divided into six lumps of gas, liquefied gas, gasoline, diesel oil, wax oil and coke; stable pressure and temperature of a reaction process are considered; a delayed coking lump kinetics model is established; and raw material configuration and operation temperatures of multiple delayed coking apparatuses are optimized by adopting a differential evolution algorithm based on the model, so that the potential of an existing apparatus is brought into play, the overall benefit is increased, and guidance is provided for improving the running level of the delayed coking apparatus.

The invention relates to a pure electric bus charging power distribution and optimization method based on a column generation framework. Under comprehensive consideration of a series of limiting conditions such as electricity price fluctuation in different time periods and the electric quantity demand of the electric bus, the operation time, the number of the charging continuous charging guns andthe total power of the charging stations, the minimum daily charging cost of the charging stations is taken as the optimal target, and an electric bus charging power distribution strategy and optimization model based on a column generation framework is constructed. The resource limitation of the charging station is clearly considered in the main problem, and the operation demand, the charging demand and the charging power continuity control of the electric bus are considered in the sub-problems. The model is extremely high in calculation efficiency; the corresponding optimal charging strategycan be rapidly formulated for the operation scheme of the large-scale electric bus, objective conditions such as dynamic electricity price and system operation resource limitation in different periodsare considered while the charging requirement of the electric bus is met, and a scientific basis is provided for the charging operation management strategy of the charging station.

The invention relates to a method of solving the balance problem of assembly line welding of a body in white through an improved particle swarm optimization algorithm. The method includes the steps: defining one function Phi (-) to convert the position of a particle swarm into an assembling sequence, providing a constrained condition and a target function for the first class of assembling wire balance problem, and finally constructing the particle swarm optimization algorithm to solve the balance problem of assembly line welding of the body in white. The method of solving the balance problem of assembly line welding of a body in white through an improved particle swarm optimization algorithm has better performance in the solution quality and has higher speed in the solution time.

The invention relates to a dynamic positioning ability analyzing method applicable to the ocean engineering field. The method includes the steps of firstly, building an environment load module, and calculating the total environment load; secondly, building a ship acting force module, and calculating ship acting force borne by a ship; thirdly, building a thrust distribution module, superposing thetotal environment load and the ship acting force, inputting into the thrust distribution module, and using an augmented Lagrange algorithm to perform thruster thrust distribution on external load; fourthly, judging conditions that Ti is larger than Timax and v2-v1 is smaller than delta are satisfied or not, if so, determining that v2 is the limit wind speed of the ship under the current environment load, and if not, using a bisection method to perform circulation calculation so as to obtain the limit wind speed; increasing the wind direction angle by one unit, namely 10 degrees, and repeatingthe previous four steps until the calculation of the limit wind speed of all angles within 360 degrees is completed; drawing the dynamic positioning ability curve graph of the ship limit wind speed according to the calculation of the limit wind speed of the 360 degrees.

The invention discloses a pilot designating method. The pilot designating method comprises the following steps of establishing a local database and performing contract pool creation; selecting pilots to be designated and ships to be piloted and establishing a pilot designation model; solving the pilot designation model by means of a dynamic sub-population based improved genetic algorithm and finding an optimal pilot designation plan; designating the pilots according to the determined optimal pilot designation plan. The pilot designating method conducts a mathematical modeling on pilot designation, has good expansibility, can be used for solving a universal pilot designation model, well solves the pilot designation problem and has the advantages of being high in solving accuracy, short in solving time, reasonable in resource allocation and high in utilization rate. The dynamic sub-population based improved genetic algorithm is adopted on selection of an algorithm for solving an optimal pilot designation model and has the advantage that crossover probability changes along with dynamic sub-populations.

The invention discloses a silicon carbide ceramic prefabricated body, an aluminum-based silicon carbide ceramic material and a preparation method thereof, relates to the technical field of materials and preparation thereof, and solves the problem of low strength and poor structure of the silicon carbide ceramic prefabricated body prepared by the existing preparation technology. Uniform and low volume fraction of silicon carbide technical problems. The main technical scheme of the present invention is: uniformly mixing the raw materials for preparing silicon carbide ceramic preforms to obtain raw material slurry; the raw materials for preparing silicon carbide ceramic preforms include silicon carbide, dispersant, organic monomer and water; Add it to the raw material slurry, and mix it evenly to obtain a ceramic slurry; the ceramic slurry is subjected to vacuum degassing treatment, injection molding, demolding and drying in sequence to obtain a ceramic green body; Sintering treatment to obtain a silicon carbide ceramic preform. The invention is mainly used for preparing a high-strength and uniformly structured silicon carbide ceramic prefabricated body, and further preparing a high-performance aluminum-based silicon carbide ceramic material.

The invention relates to a method for modeling and optimizing separation process of a C8-aromatics simulated moving bed. The method comprises the following steps that the constant pressure and temperature of the adsorption and separation process are considered, a liquid phase is assumed to be a dispersion piston flow, a solid phase is assumed to be a piston flow, a linear pushing force model is a mass transfer rate model, the adsorption balance is described by virtue of Langmuir isotherms, a C8-aromatics adsorption and separation process model is built, and multi-target operation problem of two typical simulated moving beds are optimized by a multi-target teaching optimization algorithm based on the model as follows: (a) the PX yield is maximized and the use amount of a desorption reagent is minimized; and (b) the PX yield and purity are maximized at the same time. The built model is capable of obtaining information of components of each bed layer of an adsorption and separation tower, the yield and the purity of main products in extraction and raffinate; with the combination of analysis of operation variables such as the flow of desorption reagent, the extraction liquid flow and the raffinate flow in the model by the optimization algorithm, the operation level of the simulated moving bed can be guided.

A pouring sealant for wind-driven generator cladding protection comprises a component A and a component B, wherein the mass ratio of the component A to the component B is 100:30; the component A comprises the following raw materials in parts by mass: 45-55 parts of polyol, 35-38 parts of calcium powder, 4-5 parts of a water removing agent, 5-10 parts of titanium dioxide, 0.5-1 part of a defoamingagent, 0.3-0.5 part of an adhesion promoter and 0.2-0.5 part of a wetting and dispersing agent; the component B comprises the following raw materials in parts by mass: 75 parts of carbodiimide modified MDI-50 and 25 parts of polymerized MDI. The pouring sealant is solvent-free, environmentally friendly and low in viscosity, can be cured at room temperature, and has the advantages of low cost, highadhesive strength, long operating time, good weather resistance, excellent high and low temperature resistance, namely no cracking after 120 cycles at -40-110 DEG C according to the frequency of 1 cycle per hour, capability of being used in a relatively wide temperature range, and the like.

The invention discloses a triple acceleration topology optimization method. According to the invention, by improving three aspects of multi-grid density mapping, the conjugate gradient method based onpre-processing and an initial value, and local updating, a design domain is divided into a plurality of layers of grids with different thicknesses through multi-grid density mapping, topological solving is conducted on a coarse grid layer firstly, an obtained optimization result is mapped to a next layer of fine grid layer to serve as an initial value, a topological optimization process on the fine grid layer skips redundant iteration, and the topological optimization process is accelerated; According to the conjugate gradient method based on preprocessing and an initial value, a preprocessing regulator and the initial value are added to iterative solution of a topological optimization equation, and the solution speed is increased; in the local updating, an update unit is selected, and the number of unit updates during iteration is reduced to achieve an acceleration effect. Based on the method, triple acceleration of topological optimization is realized, the topological optimization process is remarkably accelerated, and an optimization result with high precision and low calculation cost is obtained.

The invention discloses a distribution network reconfiguration method based on a complementary constraint full-smooth Newton method, comprising the following steps: 1) simplifying the distribution network structure to generate a branch group; 2) establishing a distribution network optimization model according to the generated branch group, wherein the distribution network optimization model comprises an objective function and a constraint condition; 3) solving the distribution network optimization model by using a full-smooth Newton method, and obtaining a distribution network reconfigurationscheme. Compared with intelligent method and heuristic method, this method has the advantages of high searching efficiency and short solving time. The inequality constraint in the optimization model is transformed into equality constraint by introducing smooth function, and the discrete integer value of the variable is obtained in the iterative computation process. The dimension of branch state variable is greatly reduced by branch grouping, which is suitable for large-scale distribution network.

The invention relates to a high-carbon low-chromium multi-element alloy casting ball material, which contains the following chemical element components by mass: 3.2-3.5% of carbon, 2.1-2.3% of silicon, 3.9-4.2% of manganese, 0.2-0.5% of zinc, 2.6-2.8% of chromium, 0.8-1.1% of titanium, 0.03-0.05% of nickel, 0.4-0.5% of magnesium, 0.13-0.15% of La, 0.06-0.09% of Ho, less than or equal to 0.07% of P, less than or equal to 0.07% of S, and the balance of iron. According to the present invention, the alloy adopts rare earth elements such as La and Ho, such that the strength, the wear resistance and the hardness of the alloy are increased, and the wear resistance of the alloy is 2-3 times the wear resistance of the common casting ball and the common forging ball; the core hardness and the surface hardness are basically the same, and the scrap iron is adopted as the raw material, such that the alloy quality is uniform, and cracks are not easily generated; and the refining agent is used for casting production, wherein the porosity in the cast can be reduced by 1-2 DEG, the oxidation inclusion is at about grade 2, and the yield is significantly improved.

The invention relates to a dynamic flight simulator real time motion control method comprising the following steps: 1, overload discretization: carrying out discretization sampling for a given G(t) in a sampling period [delta]t, recording overloads received by a pilot at tn time (n-1,2,..) as G(tn); 2, overload pretreatment: carrying out pretreatment for discretization overload G (tn) so as to form an overload instruction Gc (tn) curve; 3, non unloading segment large arm motion program: resolving a large arm motion angle velocity [theta]1(tn) and an angular acceleration [theta]1i(tn) of the non unloading segment at tn time; 4, unloading segment large arm motion program: resolving a large arm motion angle velocity [theta]1(tn) and an angular acceleration [theta]1i(tn) of the unloading segment at tn time; 5, centrifugation overload, tangential overload and vertical overload calculation: formulas of the three are listed in the specification; 6, middle frame and hanging basket motion program: solving a middle frame roll angle [theta]2(tn) and a hanging basket pitch angle [theta]3(tn).

The invention discloses a maximum likelihood estimation method for sea clutter amplitude model parameters based on inverse Gaussian texture and mainly solves the problem that the sea clutter amplitude model parameters are estimated inaccurately in the prior art. The method comprises steps as follows: 1), N clutter amplitude data are selected, and a moment estimation value of a scale parameter and a moment estimation value of a shape parameter are calculated; 2), a log-likelihood function is calculated according to a probability density function of a sea clutter amplitude distribution model; 3), a two-parameter iterative formula of the maximum likelihood estimation is obtained with the adoption of the log-likelihood function; 4), the two moment estimation values in the step 1) are taken as iterative initial values of the two parameters in the maximum likelihood estimation respectively; 5), the maximum likelihood estimation value (shown in the specification) of the scale parameter and the maximum likelihood estimation value (shown in the specification) of the shape parameter are obtained according to the two-parameter iterative formula in the step 3). The method can be used for effectively and accurately estimating the sea clutter amplitude distribution model parameters based on the inverse Gaussian texture through sufficient utilization of sample information, and can be used for target detection under the background of sea clutters.

The invention relates to an accurate closed-loop control scheme for fiber-optic gyroscope transit time. The accurate closed-loop control scheme comprises on-line real-time locking of fiber-optic gyroscope transit time parameters, two-stage feedback closed-loop control and four modulation waveform combination schemes given according to different design requirements. The main working process of themethod comprises the steps that high-frequency cosine wave modulation is applied to a phase modulator, spectral analysis is conducted on detector output, online real-time monitoring of the transit time is achieved through a digital phase locking scheme, real-time accumulation and rounding filtering processing are conducted on data, and the data are output to the next stage; a preliminary adjustment result is given through frequency division of control circuits such as an FPGA or a DSP, output of the adjustable crystal oscillator is controlled according to a further difference value, and feedback control with the precision superior to the ppm magnitude is achieved. The scheme is a third key parameter closed-loop control scheme after the fiber-optic gyroscope realizes rate closed-loop and phase modulation coefficient closed-loop, and has very important significance for stabilizing the scale factor and the zero offset parameter and improving the engineering performance of the fiber-opticgyroscope.

The invention discloses a pilot assignment method based on an improved discrete brainstorm optimization algorithm. The pilot assignment method comprises the following steps: determining problem constraint conditions according to problem description, and establishing a pilot assignment model; and solving the pilot assignment model based on an improved discrete brain storm optimization algorithm, and finding out an optimal pilot assignment plan. According to the discrete brainstorm optimization algorithm, a method for learning from individuals to two random individuals in the same class of the brainstorm optimization algorithm is provided, and an implementation method for individual fusion and random interference addition in discretization of the brainstorm optimization algorithm is provided. By adopting the improved discrete brainstorm optimization algorithm, the search efficiency of the optimal path can be improved, the algorithm is prevented from falling into local optimum, and the precision of the algorithm is improved.

The invention provides a bionic mask with built-in fans. The bionic mask with the built-in fans comprises an air suction channel, an air exhaust channel, a mouth and nose sealing ring, the air suction fan, the air exhaust fan, an internal air pressure sensor, an external air pressure sensor and a breathing bionic control unit. An air filtering layer and the air suction fan are arranged in the air suction channel. The air exhaust fan is arranged in the air exhaust channel. The internal air pressure sensor and the external air pressure sensor are arranged in the gas environment formed by a human body and the mouth and nose sealing ring and the atmospheric environment located outside the mask respectively. The breathing bionic control unit comprises a central processing unit and a power management module, and is attached to the exterior of the sealing ring. The bionic mask with the built-in fans can solve the problem that the breathing resistance of an ordinary filtering mask exists.

The invention discloses a Martian atmosphere entrance interruption allowing estimation method based on multiple models, relates to the Martian atmosphere entrance interruption allowing estimation method, and belongs to the technical field of deep space exploration. The method includes the implementation steps of (1) building a Martian atmosphere entrance section dynamics equation; (2) building the Martian atmosphere entrance section wireless measurement model and the wireless measurement interruption model so as to be used for modifying the detector position and speed information in the step (3); (3) resolving the detector state in cooperation with the wireless model and the interruption model in the step (2) and a filtering algorithm, and estimating the detector state. By means of the Martian atmosphere entrance interruption allowing estimation method, the problem of estimation divergence of the detector state under wireless measurement interruption can be effectively solved, and the reliability and the state estimation accuracy of an autonomous navigation system of a detector on the Martian atmosphere entrance section are improved. By means of the Martian atmosphere entrance interruption allowing estimation method, technical supports and references can be provided for adopting the wireless navigation scheme design on the atmosphere entrance section in Martian landing detection in future.

The invention relates to a suspension bridge inhaul cable force finding analysis method. The suspension bridge inhaul cable force finding analysis method is a suspension bridge inhaul cable force finding analysis method based on a target tracking-iterative updating algorithm. According to the method, under the condition that the geometric configuration and boundary conditions of the suspension bridge are known and the cable force in a finished bridge state is unknown, the mid-span deflection of a suspension bridge in a finished bridge state is used as a target value, an ANSYS parameterized programming language is used, and a force finding analysis method of a mid-span deflection-iterative update real constant array is tracked to solve the inhaul cable force meeting the structural design requirement of the suspension bridge. The suspension bridge inhaul cable force finding analysis method has the advantages that the suspension bridge inhaul cable force finding analysis method is simple,rapid, small in iteration frequency, short in solving time, low in requirement for computer configuration, high in solving precision and capable of meeting the requirements for academic research andengineering application to a certain extent, and is an effective method for determining the cable force of the suspension bridge in the bridging state.

The invention discloses a multi-quantile estimation method of sea clutter Weibull amplitude distribution parameters. The method includes the steps of firstly, acquiring the cumulative distribution function of a sea clutter Weibull amplitude distribution model according to the probability density function of the sea clutter Weibull amplitude distribution model; secondly, selecting three sample cumulative probability p1, p2 and p3, and acquiring three quantile equations related to the shape parameter beta and the scale parameter eta of the sea clutter Weibull amplitude distribution model according to the definition of quantile; thirdly, using clutter amplitude data to obtain estimation values x^p1, x^p2 and x^p3 of quantiles xp1, xp2 and xp3; fourthly, using the estimation values x^p1, x^p2 and x^p3 and the three quantile equations in the second step to obtain the estimation value beta^ of the shape parameter beta and the estimation value eta^ of the scale parameter eta. By the method applicable to target detection under the sea clutter background, interference of anomalous scattering units to samples can be reduced, and parameter estimation performance is increased.

The invention relates to a planetary power descending branch navigation method based on a TDS (total descending sensor) and image measurement and belongs to the technical field of deep space exploration. At a power descending branch, an inertial unit of a landing device navigation system is used for flight position recurrence and is a basic sensor; a Doppler radar on the TDS is used for directly acquiring distances from a landing device to the surface of a target planet in three wave beam directions and relative speeds of the landing device along the three wave beam directions and then calculating the altitude of the landing device and speeds of the landing device in three axial directions of a landing point fixedly-connected coordinate system; a navigation camera serving as a vision auxiliary navigation sensor is used for precisely measuring included angles between sight directions of different characteristic points of a landing area; the positions of the characteristic points relative to the landing point are known; the sight included angles contain horizontal position information of the landing device relative to the landing point, so that the horizontal position precision of the landing device is greatly improved.

The invention relates to a feature extraction and classification system and method for gastric magnetic signals and belongs to the field of bioelectromagnetics. The feature extraction and classification system comprises a gastric magnetic fast-wave signal extraction module, a gastric magnetic signal filtering module, a gastric magnetic fast-wave feature extraction module and a gastric magnetic signal classification module. The morphological method and the RLS adaptive algorithm are adopted to effectively filter out baseline drift components in signals; low-frequency components and high-frequency components of the gastric magnetic signals are decomposed by wavelet packets, corresponding bands are adaptively selected according to signal features and analysis requirements to be matched with spectrums of the signals, and noise components proximate to fast-wave components in the gastric magnetic signals are filtered out; a matching pursuit and nonparametric basis function combined algorithm not only overcomes the defect that time-frequency atoms need to be built in advance by independently using a matching pursuit algorithm, and ensures that template signals can be adaptively conditioned not to depend too much on prior knowledge; the K-means algorithm is adopted, so that the solving time is short and the classification effect meets requirements.

The invention discloses a thermoelectric coupling modeling method of a power semiconductor chip cell. The method comprises the following steps: S1, dividing a power semiconductor chip into a multi-cell structure; s2, extracting a voltage-current-temperature three-dimensional model of the chip; s3, extracting a voltage-current-temperature three-dimensional model of the cell; s4, solving a cellular current distribution proportion and corresponding loss under any temperature distribution; s5, superposing the temperature rise generated by each cell to obtain the overall temperature gradient of the chip; s6, repeating the steps S4 and S5 until the temperature gradient deviation is converged; and S7, extracting chip surface temperature peak values under different working conditions, and calculating corresponding power module thermal resistance. According to the method, a chip lumped Fourier series analytical thermal model is combined with a multi-cell distributed electrical model, so that thermoelectric coupling modeling and calculation of power semiconductor chip cells are achieved. Besides, the method has the advantages of small required calculation amount, short solving time and no convergence problem, and is particularly suitable for online prediction of the temperature of the power semiconductor chip.