84 results about "Mathematical theory" patented technology

Aiming at better describing the statistical dispersion characteristics of boundary points, the invention relates to a more practical, popular and adaptive method for more accurately diagnosing the fault of a power transformer on the basis of the interval mathematical theory, belonging to the field of the main power equipment fault diagnosis technology. The following technical schemes are adopted in the invention: (1) a syringe with good air tightness is adopted for injecting the oil sample; (2) after the gases in the collected oil sample are subjected to ultrasonic and vacuum removal and separation in an oil-gas separating room, the components and the content of the gases are analyzed with a gas chromatograph; (3) three contrast values on the components of the five characteristic gas are formed through calculation and are represented with different codes; (4) the maximum three contrast values on the content of the five characteristic gas, which is measured in a plurality of experiments, are extracted to form the interval numbers of the three contrast values; and (5) the corresponding type and nature of the fault of the power transformer are determined. The method is mainly used for diagnosing the fault of the power transformer.

The invention relates to a composite insulator ageing state comprehensive evaluation method based on multi-characteristic vectors. According to the method, the mathematical theory of fuzzy variable weights is introduced into state evaluation of composite insulators, parameters which comprehensively and accurately reflect the working states of the insulators are separately selected, an improved analytic hierarchy process (IAHP) is adopted, and opinions of an expert group are processed. On the basis of constant weights, data of the insulators is combined, and variable weights of all factors are determined. Before indexes are adopted to conduct state evaluation, all the indexes are subjected to normalization, quantitative data is transformed according to corresponding formulas, a scoring system is adopted for qualitative data, the score ranks from zero to one, the distribution function which combines half trapezoid and triangle is adopted to determine the membership degree, and the influence of all index quality on the insulator state is synthesized by adopting a multilevel fuzzy comprehensive evaluation method. According to the method, related state parameters are organically and comprehensively combined, the accurate and positive evaluation on the state of the insulators in running is made, and the method has the practical value.

The invention relates to a risk assessment method for refuse landfill polluted groundwater. The method comprises the following steps of: acquiring a landfill risk parameter, an aeration zone anti-pollution property parameter, an aquifer vulnerability parameter, an evaluation object exposure parameter and a groundwater value level parameter; calculating the weight of an evaluation index by an analytic hierarchy process according to the weight and expert grading value of a calculation index, and obtaining a landfill polluted groundwater risk comprehensive assessment result by using the weight of the evaluation index and the expert grading value of the calculation index through a comprehensive index method mathematical theory; performing multi-stage early warning by combining the landfill polluted groundwater risk comprehensive assessment result, and proposing measures, which are needed to be improved, strengthened and monitored, in landfill management; and predicting the risk level of the landfill polluted groundwater after a treatment scheme is implemented according to the level of risk early warning by combining the proposed landfill treatment measure and implementation scheme, and checking the scientificity and reasonableness of the landfill treatment scheme. The invention also provides a risk assessment system for the refuse landfill polluted groundwater.

The invention discloses a site selection method for immigrant settlement areas based on a geographic information system. The method comprises the following steps of: 1) by using a spatial analysis technology, which comprises terrain analysis, buffer area analysis and superposition analysis, of the geographic information system, establishing a spatial site selection model, and thus obtaining a series of candidate points / areas; 2) establishing an evaluation index system of settlement areas, acquiring each index data of the candidate settlement areas, and entering a database; 3) performing dimensionless standardization on an original matrix of each evaluation index of the candidate settlement areas; 4) determining the weight of each evaluation index by using a composite even weighting method based on an information entropy theory; and 5) based on a fuzzy mathematical theory, performing quality sequencing on the candidate settlement areas by using a fuzzy adjacency ratio method. The method has an important practical application value in the site selection of the immigrant settlement areas of water and electricity engineering and has wider application prospect.

The invention discloses a water quality parameter prediction method based on a multilayer RNN and a D-S evidence theory. Aimed at a high-dimension, multi-peak, discontinuous and non-convex complex nonlinear system of the water environment, historical data of water quality parameter content is used to train a deep RNN predication model, the model is constructed, and the future content of the waterquality parameter is predicted preliminarily; and the problem of uncertainty can be solved in the mathematical theory via the D-S evidence theory. Thus, conflict solution and evidence fusion can be realized by utilizing the D-S evidence theory on the basis of the RNN model prediction result, the prediction precision of the single parameter is improved, and early warning of multiple parameters is more accurate.

The invention discloses a method of three-phase unbalance load flow calculation based on the complex affine mathematical theory. The method includes inputting network original parameters including impedance values of each line and transformers, and active power and reactive power of each load node phase; initializing amplitude values and phase angles of each load node phase voltage; setting zero of iteration; calculating each phase injected current complex affine vector of each load node; starting from an end load node, performing summation calculation of branch current, and acquiring starting current complex affine values of each phase of each branch; and utilizing a set original load node voltage as a boundary condition to calculate voltage drop of each phase of each branch and end voltage complex affine values. By the method compared with the prior art, effective optimization of an uncertain load flow calculation interval algorithm is provided, interval load flow result conservative property, defects of an interval method are overcome, load flow of a distribution network containing uncertain information can be figured out effectively, and conservative property in the process of interval load flow calculation can be solved.

The invention discloses a short-term wind speed forecasting method based on local integrated study. The method comprises steps as follows: step one, K wind speed samples most similar to a measured wind speed sample point are found out with a K-nearest neighbor algorithm; step two, a corresponding wind speed predicted value of a local sample of each wind speed sample point is obtained; step three, a plurality of wind speed predicted values obtained in the step two are fused by adopting certain fusion strategies to obtain a predicted value of a current wind speed point. The short-term wind speed forecasting method based on the local integrated study is proposed firstly, and the complexity and the diversity of wind speed sample change are considered; mathematical theory knowledge of the local integrated study algorithm is provided; difference of different wind speed samples is effectively obtained, different wind speed forecasting models are established according to different wind speed sample points, and compared with an overall wind speed forecasting model, a local wind speed forecasting model and a whole integration wind speed forecasting model, the forecasting error rate is reduced by more than 10%.

The invention relates to a unit layout and electromagnetic design method of a fuselage conformal array antenna. The method includes: (1) establishing a body coordinate system; (2) designing a unit arrangement form according to a fuselage structural characteristic, determining a unit spacing, and preliminarily determining a unit number and an array row number according to an electrical performance index; (3) determining position coordinates of the units on a fuselage; (4) selecting a radiation unit, establishing a local coordinate system shown in the description of each unit, and establishing a far-field pattern analysis model of a conformal array by employing a mathematical theory related to coordinate rotation; (5) superposing radiation patterns of the units to obtain a far-field analysis model of the whole array; (6) reducing the number of variables by employing a corrected Bernstein polynomial; and (7) regarding a pitching plane maximum side-lobe level and an azimuth plane maximum side-lobe level as a target function, performing optimization by employing a particle swarm optimization algorithm to obtain a control parameter of the corrected Bernstein polynomial, and sampling the polynomial to obtain an excitation amplitude value of an antenna unit.

The invention relates to a method for designing a dynamic burdening control system, which comprises the following steps of 1) establishing a standard process model for a dynamic burdening system; 2) establishing a system requirement model for the dynamic burdening system; 3) establishing an object application model for a dynamic burdening control system; 4) in combination of the standard process model for the dynamic burdening system in the step 1), the system requirement model for the dynamic burdening system in the step 2) and the object application model for the dynamic burdening control system in the step 3), establishing a feature matching model for the dynamic burdening control system by utilizing the knowledge excavation technique. In this way, the scientificity and the rationality of the design effect of the dynamic burdening control system are ensured. According to the technical scheme of the invention, the method for designing the dynamic burdening control system is realized through the mathematical theory statistical technique and the knowledge excavation technique in combination with a semantic model. In this way, the dynamic burdening control system is designed to be better in scientificity, theoretical instructiveness and practical applicability. Therefore, the method is wide in application prospect.

The invention discloses a reservoir evaluation method for heterogeneous interbedded sand bodies. The comprehensive evaluation index CEI of each small layer is obtained through the method, and then, the reservoir classification and evaluation result of each small layer can be obtained. Compared with the calculation process in the prior art, the method is simple and convenient and can be implementedby ordinary geological researchers. The problem that the existing quantitative reservoir evaluation needs to be based on complex mathematical theories is solved. The comprehensive quantitative evaluation of heterogeneous interbedded sand body reservoirs can truly go out of scientific research institutions and advanced laboratories and go deep into the basic design personnel of oil and gas development. The method provides effective geological guidance for the evaluation and development of heterogeneous interbedded sand bodies, and can reduce the misjudgment in the early stage of development and realize the full integration of theoretical research and practical development.

The invention belongs to the technical field of computer applications, and discloses a community division method of a multi-relation social network. The community division method of the multi-relationsocial network comprises the following steps: firstly, converting original network data into similarity tensor, and then establishing an analysis model; then using a tensor decomposition method to obtain a tensor decomposition result; and finally obtaining a community division result by using a clustering analysis method. According to the community division method, the multi-relation network dataare used for performing the community division, the division accuracy is better than the community division result obtained by just using single-relation data, and compared with the traditional method, the tensor decomposition method has the advantages of being more efficient and accurate; and compared with the existing multi-relation network community division method, data information can be fully used, and a better division result is obtained by using a mature mathematical theory method.

The invention discloses an image feature decomposition-based high spatial resolution remote sensing image prospecting method. The method comprises the following steps of extracting color feature and the texture feature of an image; solving the spectrum second derivative of the image; comprehensively analyzing the three elements i.e. the extracted color feature, the texture feature and the second derivative image; and by using a logic overlay analysis method, solving the intersection by using the algorithm of multiplying the three elements through a fuzzy mathematical theory, wherein the intersection position is the remote sensing prospecting target area. According to the method, appearance shapes such as color, shape, texture and the like shown by the high spatial resolution remote sensing data are fully utilized and the data is fully mined by adopting the techniques such as mathematics, image processing, GIS (Geographic Information System) and the like to comprehensively confirm the prospecting target area; and reliable basis is provided for the periphery prospecting of mining areas and the ore deposit location in the next step.

The invention discloses an evaluation method of an integrated energy system. The method comprises the following steps: establishing an integrated evaluation index system; calculating the subjective weight of the index of each level relative to the index of the previous level; calculating the objective weight of the index of each level relative to the index of the previous level; calculating to obtain a comprehensive weight; and evaluating the integrated energy system. According to the invention, based on four criteria of economy, energy efficiency, environmental protection and reliability, all aspects of performance of the system are considered; a subjective weight is given to each criterion by adopting the FAHP of a triangular fuzzy number, so that fuzziness and randomness of expert scoring are effectively avoided; EWM is adopted to give objective weights to all indexes, and a strong mathematical theory basis is provided; a comprehensive weight is obtained by adopting a linear weighting method, and experience judgment of experts and objective laws of sample data are comprehensively considered; finally, the evaluation result is more scientific and accurate by adopting a method approaching positive and negative ideal solutions, and the comprehensive energy system can be evaluated more objectively and reliably.

The invention provides a method for plane segmented approximation of an arc. The method is suitable for a forming path in forming technologies such as laser plate forming and plate local bending forming. The plane or straight line feature segmented approximation is performed based on sine and cosine theorem and a triangle geometrical relationship to be arc with a given curvature. According to the method for plane segmented approximation of the arc, a plane or a straight line is segmented in advance, a segmentation mode is uniform segmentation or random segmentation, the elevated heights of the end of the plane or the straight line and each segmentation node relative to the initial position of the plane or the straight line after each segmentation section is calculated based on the sine and cosine theorem and the triangle geometrical relationship according to the curvature radius of the to-be-approximated arc, each segmentation section is bent to approximate the arc, and finally the goal of approximating the arc with the given curvature is achieved. A relationship of a laser forming technology and a mathematical theory model can be established, and the method is of very good guiding significance for improving forming precision.

The invention discloses a blind separation method for mobile source signal aliasing. In real life, an acquired aliasing signal has certain variability, an aliasing filter is time-varying, and comparedwith the process of a traditional time-invariant aliasing filter, the separation process of a source signal becomes more difficult. Aiming at the aliasing condition of the time variation, a robust blind separation algorithm is provided. The method includes the steps of locating a source signal through a time difference estimation algorithm, and providing strict mathematical theory derivation; reconstructing the aliasing filter, updating model parameters by utilizing an expectation maximization algorithm, and separating a source signal by utilizing a Wiener filtering method; and finally, verifying the effectiveness of the proposed algorithm through a simulation experiment, and comparing with a robust mobile source signal aliasing blind separation algorithm (Full-rank algorithm) to prove that the algorithm provided by the invention has better robustness for mobile source signal aliasing blind separation.

The invention relates to an online cold-rolled steel strip shape statistic method which mainly includes macroscopic shape data statistics, shape component data statistics and shape data distribution test and fuzzing. By the aid of two present frequently-used shape statistic methods including a statistic method based on a probability theory and a grading percent statistic method, distribution laws that shape data comply with are tested by a K-S (Kolmogorov-Smirnov) test method and a chi-square test method in a goodness-of-fit test theory, critical shape values are fuzzed by a fuzzy theory method, and finally, macroscopic shape statistics is realized. Shape components are recognized by the aid of a pattern recognition theory, shape component data distribution is tested, shape component data are fuzzed, and shape component statistics is finished. The online cold-rolled steel strip shape statistic method has a strict mathematical theory basis, and reliable shape statistics data are provided for production of high-performance cold-rolled steel strips by combining the shape theory with the probability statistics theory.

The invention provides a method in which coordinates where a watermark is embedded in an image is calculated through character type watermark information, the character type information is utilized to generate the watermark through a two-dimensional code generator, then the two-dimensional code watermark is embedded in the image, whether a two-dimensional watermark exists in an original image is detected, and a watermark repeated embedding attack is resisted. Based on mathematical theory and image data characteristics, combined with characteristics of a transform domain watermarking algorithm, the idea of an airspace algorithm is utilized, a watermark information mapping algorithm, a two-dimensional watermark distributed embedding technology and similarity detection algorithm are used as key technologies, and thus the method can achieve the purpose of resisting repeated embedding and have relatively good robustness and relatively high data capacity.

The invention discloses an assessment method of pile sinking of ocean platform taking consideration of risk level classification. The method includes the following steps: classifying risk levels of pile sinking; establishing a system of assessment indexes and classifying the risk levels for all the assessment indexes, confirming the comparison matrix and using the average and range method to construct the judgment matrix to determine the weights of all the assessment indexes, determining the membership grade of each risk level for each assessment index, establishing a fuzzy mathematical model to solve the fuzzy vector of the pile sinking risk level for the ocean platform and based on the principle of the largest membership grade making the assessment about pile sinking risk level of the ocean platform. The invention provides an assessment method based on AHP and fuzzy mathematical theory, applies the method of risk levels classification to give an assessment to the pile sinking of ocean platforms. The assessment results are more close to the actuality. The method is suitable for the assessment of pile sinking risks of ocean platforms.

The invention provides a propulsive force distribution method, device and equipment for a synchronous pushing and splicing system of a shield tunneling machine and a medium. The method comprises the steps that first propulsive force resultant force and first resultant force point coordinates of oil cylinders in an alternate pushing and splicing mode are obtained; constructing a propulsive force constraint equation which enables the second propulsive force resultant force and the second resultant force point coordinate to be equal to the first propulsive force resultant force and the first resultant force point coordinate respectively and enables the propulsive force change of each oil cylinder in a propulsive state to be minimum; and solving the propulsive force constraint equation to obtain the propulsive force of each oil cylinder in a propulsive state in the synchronous pushing and splicing mode. On the basis of mechanics and mathematical theories, a propulsive force constraint equation which enables total propulsive force resultant force and resultant force points to be unchanged and enables propulsive force changes of all oil cylinders in a propulsive state in a synchronous pushing and splicing mode and all oil cylinders in an alternate pushing and splicing mode to be as small as possible is constructed, propulsive force loss can be reduced, adjustment of the propulsive force of all the oil cylinders is more accurate, and the propulsive force of all the oil cylinders can be adjusted more accurately. And the shield attitude and the propulsive force speed are guaranteed.

The invention discloses an anesthesia machine and breathing machine pressure control and adjustment method suitable for different R values and C values, which comprises the following steps of: taking a part of R values and C values as positioning points; performing proportion integration differentiation (PID) control parameter adjustment on the R value positioning points and the C value positioning points; in terms of the congruent relationships of the R values, the C values and corresponding PID parameters, establishing a membership function according to a fuzzy mathematical theory and grading the discourse domains of the membership function; setting a fuzzy rule and a decision method according to a grading condition, and writing a corresponding fuzzy control algorithm according to the set fuzzy rule and decision method; automatically calculating the PID parameters corresponding to the R values and the C values which are not set as the positioning points in the value ranges of R and C according to the fuzzy control algorithm; and performing pressure feedback control on an overall pressure control system by using the calculated PID parameters.

The invention discloses an impact risk dynamic early warning method based on hierarchical analysis and fuzzy mathematics, and the method comprises the following steps: 1, determining an early warningobject and an impact risk impact factor suitable for the early warning object, i.e., an early warning index; 2, determining the impact risk grade of each early warning index to form an impact risk early warning index system; 3, based on an APH (analytic hierarchy process) theory, determining the weight of each early warning index to form a fuzzy weight vector A; 4, based on a fuzzy mathematical theory, establishing a single-factor fuzzy early warning set of each index, and forming a factor early warning fuzzy matrix R; 5, calculating a fuzzy comprehensive evaluation vector B = A * R of the early warning object; and 6, determining the impact danger level of the early warning object according to the maximum membership principle. According to the method, various rock burst early warning indexes are comprehensively considered, the early warning model is adopted to perform impact risk early warning on the early warning object, the early warning is more scientific, the early warning efficiency is higher, and the result is more reliable.

The invention belongs to the technical field of artificial intelligence, and discloses a deep Bayesian network modeling training method, which comprises the following steps: preliminarily determining model nodes of a Bayesian network by adopting features of an analysis object, and determining the model nodes as deep nodes replaced by feature vectors obtained by a deep neural network; dividing the sample data into a training set and a test set to determine a deep neural network model structure and initialize network parameters, and inputting the original data into a deep neural network to obtain output features of the deep neural network, and obtaining a full sample training set after discrete quantization through the training set which is not discretized, obtaining the trained Bayesian network, obtaining the output characteristics of the Bayesian network, obtaining a full sample test set after discrete quantization, and achieving accuracy prediction of the output nodes of the deep neural network. According to the method, modeling reasoning can be carried out through the feature quantity obtained through feature engineering, and original data can be directly processed. The method is a Bayesian network model which has a complete mathematical theoretical basis, is good in interpretability and has an automatic feature extraction capability.

The invention discloses a micro-blog spammer detection method based on a fuzzy multi-class SVM. A one-versus-rest SVM multi-class classifier and a fuzzy mathematical theory are combined for micro-blog spammer detection. According to the conventional micro-blog spammer research, global spammer characteristics are analyzed, the analysis force is not enough, the analysis of the characteristics of certain kinds of spammers is not conducted, and the spammers escape detection of a detection system. According to the method, the characteristics of the micro-blog spammers is conducted, the spammers are divided into three categories, the one-versus-rest SVM multi-class classifier is constructed, fuzzy processing is performed on mixed classified samples generated by the multi-class classifier by employing a fuzzy clustering method, the combined classifier is obtained, and the detection precision is improved.

The invention discloses a method for measuring the elevation of an independent structure, which comprises the following steps: selecting two survey stations, which are in communication with a to-be-measured point and are positioned on the same vertical plane, on the ground far away from the structure to be measured according to the angle measurement and distance measurement integrated function ofa total station and the advantages and characteristics of accurate data and convenience in operation of the total station and by combining a certain mathematical theory; erecting the total station atthe two survey stations respectively; observing two vertical angles and a section of horizontal distance; calculating the height difference between the top end of the structure to be measured and thesurvey stations through mathematical calculation; and obtaining the accurate elevation of the independent structure conveniently and quickly through the relationship between known points on the groundand the survey stations. The method for measuring the elevation of the independent structure can quickly finish the measurement work by only two persons.

The invention discloses a marine ship oil spill accident multi-material emergency scheduling method. The method comprises the following steps: scheduling material type confirmation, scheduling ship type confirmation, oil spill accident demand estimation, oil spill drift diffusion simulation, a triangular fuzzy mathematical theory, theoretical model establishment, algorithm design and actual multi-material scheduling scheme determination. Aiming at the characteristics of continuity and drifting diffusivity of marine ship oil spill accidents, the oil spill emergency material scheduling problem under the characteristics of multiple materials and multiple scheduling ships is researched, a drifting diffusion position is determined through GNOME simulation, a material demand point is determined,and then the material demand in a fuzzy state is determined by utilizing a triangular fuzzy number; and finally, an emergency material scheduling model is built for the marine ship oil spill accident. The application of the model needs to clarify the reserve of the emergency storage, the number of schedulable ships, the distance between the emergency storage and an accident point and other information in advance, and solves the problem of scheduling and distribution of various materials by the emergency material storage.

The invention provides a rectifying magnetoresistance device formed by utilizing discrete devices. The rectifying magnetoresistance device utilizes combination of a rectifying effect of a diode and a device purely having a magnetoresistance effect to realize a rectifying magnetoresistance effect. The invention further provides a preparation method of the rectifying magnetoresistance device. The invention also provides a method for building a mathematical theory model based on the rectifying magnetoresistance device. According to the method, the mathematical theory model built according to electrical properties of each device and operating principles thereof is utilized to simulate electrical transport properties of the devices. The rectifying effect and the magnetoresistance effect are organically combined together through connecting the rectifying device and the magnetoresistance device in parallel, so that rectification voltage changes along with an external magnetic field, that is, a rectifying magnetoresistance effect is realized.

The invention discloses a topological optimization method for structural defect recognition, and the method comprises the steps: taking the pseudo density of a finite element model unit of a to-be-detected structure as a design variable, taking the square of a displacement response error of the to-be-detected structure and a defect-containing structure as a topological optimization target function, and carrying out the finite element analysis of the structure, so as to obtain a real-time structural displacement response; deriving the derivative of the objective function relative to the designvariable, carrying out topological optimization iterative solution of defect recognition, and completing defect recognition when the objective function converges to the minimum value. Defect identification of the structure is achieved by adopting a topological optimization method, a defect identification problem is converted into an inverse problem to be solved, and the method has a rigorous mathematical theory. According to the invention, positioning and shaping identification of structural defects can be carried out, prior information is not needed, the influence of defect quantity is avoided, and structural defect display is visual.

The invention discloses a sandstone stratum vertical shaft freezing construction safety monitoring method, including the steps of: S1. determination of a factor set; S2. determination of an evaluation set; S3. establishment of a weight set; S4. single factor fuzzy comprehensive evaluation; S5. Determination of a fuzzy composition operator; S6. fuzzy comprehensive judgment; and S7. evaluation index analysis. On the basis of analysis of a common safety evaluation method for engineering construction, a fuzzy mathematical theory is adopted, main risk influence factors such as freezing rock strength, freezing wall thickness, freezing temperature, shaft excavation section height, underground water seepage, freezing stratum pressure, freezing pipe deflection and power failure time are considered, and a Luohe River formation sandstone stratum vertical shaft freezing construction safety comprehensive judgment method is proposed, which is of certain theoretical and practical significance to aspects of reducing shaft building cost, improving construction speed and enriching freezing shaft-sinking theories of our country.

A wireless sensor network data fusion method of fusing time delay and survival time evenly is characterized by firstly considering an actual factor that the event occurrence has the unpredictability in the wireless sensor network application, establishing a data fusion problem into a random optimization model taking the time-averaged network energy consumption minimization as the objective, then designing a strategy of real-timely selecting the nodes participating in the data fusion based on a Lyapunov optimization method, and establishing a trade-off relation between the network energy consumption and the maximum fusion time delay of various perceptual information, thereby achieving the purpose of fusing the time delay and the survival time evenly. The wireless sensor network data fusion method of fusing the time delay and the survival time evenly of the present invention is reasonable in model, is supported by the mathematical theory, can satisfy the different fusion time delay requirements of the different types of perceptual information, and can be applied to a wireless sensor network directly.

The invention belongs to the field of fluid mechanical design and particularly discloses a multi-grade centrifugal pump guide vane hydraulic design method of high matching of inner and outer edge flow lines of a blade. The process of the multi-grade centrifugal pump guide vane hydraulic design method comprises the following steps: determining main geometrical parameters of a multi-grade centrifugal pump guide vane; designing the inner and outer edge flow lines of a transition section as a concentric circle; carrying out same-original-point and equal-angle ray equal division on the transition section of the outer edge flow line; measuring each section of arc length on the transition section of a corresponding outer edge mold line and the average radius of each section on the transition section of the outer edge flow line respectively, and calculating a wrapping angle corresponding to each section of the arc length on the transition section of the outer edge flow line; designing the inner edge flow line of the blade by adopting a wrapping angle two-wing conversion method; introducing a deviation coefficient lambda and calculating theoretical lengths of an inlet section and an outlet section of the inner edge flow line; and drawing a blade axial plane graph of the multi-grade centrifugal pump guide vane. The multi-grade centrifugal pump guide vane hydraulic design method not only provides supports for mathematical theory, but also improves the matching effect of the inner and outer edge flow lines of the blade of the multi-grade centrifugal pump guide vane; and the inner hydraulic loss can also be effectively reduced and the operation efficiency is further improved.