64results about How to "Reduced modeling effort" patented technology

The invention discloses an interior noise analysis and prediction method of a high speed train. The method comprises the following steps: establishing a train reconditioning train body model, a statistical energy analysis model of a body in white structure and an interior and exterior vocal cavity statistical energy analysis model, and carrying out simplification and subsystem division; obtaining the statistical energy analysis parameters of the train body structure and an interior vocal cavity model, and loading the statistical energy analysis parameters onto a train body structure model plate subsystem and a vocal cavity model subsystem; and obtaining exterior vocal excitation source energy borne on the train body, applying the exterior vocal excitation source energy onto the exterior vocal cavity statistical energy analysis model, causing the exterior vocal excitation source energy to reach an interior vocal cavity after the exterior vocal excitation source energy is attenuated by the sound insulation property of a structural plate in a body in white structure model so as to obtain structural noise energy radiated into the train by the reconditioning train body under the function of the two-line suspension force of a compartment, and then, carrying out interior noise analysis and prediction. The problem that the interior noise of the train is difficult in prediction and the problems of the upper limit boundedness of a frequency domain, complex calculation flow, incomplete motivation consideration in the traditional method are overcome, calculation efficiency and prediction accuracy are improved, and development and test cost is lowered.

The invention discloses a multi-model disturbance estimation predictive-control method for superheated steam temperature of a thermal power generating unit. According to the method, a superheated steam temperature system serves a controlled object, cooling spray water flow serves as system control input, superheated steam temperature serves as system output, a plurality of disturbance models are established to respectively expand the system state on the basis of a controlled object state space model, and a disturbance model set is formed. Sub-models in the disturbance model set are subjected to weighted averaging, thus the influence of immeasurable disturbance upon the system output is fitted. According to the concept of predictive control, optimization solution is performed once for each step so that optimal system controlled input is obtained. Compared with common predictive control algorithms, the method has the advantages that immeasurable disturbance can be inhibited more effectively, and steam temperature is kept around a set value stably; when no immeasurable disturbance occurs, the method is as effective as the common predictive control algorithms, and tracking regulation performance is fine.

The invention discloses a decoupling control method based on a composite disturbance-rejection diesel engine VGT-EGR system. The method mainly comprises the steps of static feedforward control, dynamic feedforward control and active disturbance-rejection control (ADRC). According to the VGT-EGR system, a two-input two-output system is converted into two single-input single-output channels; the coupling influence between the two channels, changes of dynamic characteristics of each channel along with changes of the working condition of an engine, and disturbance of the external environment are totally regarded as the total disturbance. The ADRC is used for performing real-time observation and compensation on the total disturbance, and precise modeling is not needed; however, the bandwidth of the ADRC is limited by the sampling frequency and noise, and transient tracking is slow. Therefore, the static feedforward control is added into the method, and the simple and efficient dynamic feedforward control based on a model is specially designed. A simulation result and an experimental verification result show that control parameters of the method are easy to calibrate, and rapid tracking of the boost pressure and the air input in the whole working condition can be achieved only through one set of parameters.

The invention discloses a 10-kV three-core cable finite element thermal analysis method. The method comprises the steps of 1, determining the geometric structure, size and material parameters of a 10-kV three-core cable; 2, utilizing the Solid Works for building a three-dimensional model according to the geometric structure of the 10-kV three-core cable; 3, importing the three-dimensional model built by the Solid Works into ANSYS and converting the three-dimensional model into a geometric model in the ANSYS; 4, calculating the heat production rate of a heating portion according to the current loaded by the cable and parameters of the heating portion of the cable; 5, loading material parameters of layers in the ANSYS, dividing units and applying loads to obtain the temperature field distribution result of the 10-kV three-core cable. According to the 10-kV three-core cable finite element thermal analysis method, the three-dimensional model building function of the Solid Works and the finite element analysis function of the ANSYS are combined, three-dimensional temperature field simulation of the 10-kV three-core cable is achieved, the operational process is visual and convenient, the model is precise and reliable, the operation efficiency is high, and the method can be popularized in finite element thermal analysis of complex structures like cable intermediate connectors.

The invention discloses a pure signal analyzing method in a multivariable system, which relates to chemometrics and a process analyzing method. The method comprises the following steps of: describing the multivariable system by adopting a data space, wherein the data space M is divided into a sub space S spanned by the vector s of a concerned variable and adjacent sub spaces H spanned by the vectors of the other variables, namely M is equal to SH, the data space M is the multivariable system, the sub space S of the concerned variable is a pure signal of the concerned variable in the multivariable system, and the adjacent sub spaces H of the other variables are the other signals in the multivariable system; and establishing a pure signal separation model according to the core relation between the vector s of the concerned variable and the adjacent sub spaces H of the other variables in a known modeling sample, so that the pure signal of the concerned variable can be separated out from the other unknown data spaces M through the pure signal separation model. By the method, a concerned signal can be acquired and purified from mixed signals, so the method is very suitable for chemical analysis.

The invention discloses a method for dynamic prediction of air-conditioning loads based on a multi-factor chaos support vector machine. For a cool storage air-conditioning system, a combined cooling heating and power supply air-conditioning system, a solar air-conditioning system and other complicated building energy systems, accurate short-term prediction is conducted on dynamic air-conditioning loads based on a support vector machine technology by conducting phase-space reconstruction on historical load sequences and considering the internal factors of the air-conditioning loads. According to the technical scheme, the method comprises the steps that firstly, a training sample set taking outdoor temperature, relative humidity, solar radiation intensity and other meteorological factors and chilled water temperature as principal elements is established; phase-space reconstruction parameters are optimized by adopting an artificial intelligence method, and phase-space reconstruction is conducted on load sequences after chaos characteristic identification; model related parameters of a load model of the multi-factor chaos support vector machine are found by adopting a grid search method; load model of the multi-factor chaos support vector machine is established, and prediction is performed. By adopting the method, the prediction accuracy of dynamic air-conditioning loads is improved.

The invention discloses an aircraft engine part modeling method based on a control system development platform. A dynamic link library A and a dynamic link library B are constructed, data link among the Pascal programming language, the C / C++ programming language and the M programming language is achieved, meanwhile, the general characteristic of height customization is guaranteed, a simple interaction way is provided for transition from aircraft engine overall performance analysis under GasTurb to control system development under MATLAB, overall performance analysis and control system development are linked, and interaction efficiency between departments is improved; according to the method, a part level dynamic model under the engine overall performance analysis platform GasTurb is directly called through the dynamic link library technology under the control system development platform MATLAB, a repeated modeling process under the control system development platform can be avoided, and the aircraft engine part level modeling workload under the control system development platform and the risk that potential errors occur are reduced.

The invention discloses a dynamic analysis method for a spacecraft propulsion system based on a Modelica model. The method comprises the following steps: 1) determining a model target; 2) constructing a concept model; 3) decomposing a system; 4) building a basic model; 5) designing a connector; 6) establishing an assembly model; 7) building a medium model; 8) building a system model; 9) making simulated analysis. The dynamic analysis method combines a modeling simulation theory of the spacecraft propulsion system with a system Modellica technology and employs a statement method directed at objects to build a dynamic model of the propulsion system. Due to non-causal characteristics, modeling engineers do not need to decouple the propulsion system without assigning input and output variants of the model and the equation solving sequence. During solving, a simulation system can determine the equation solving sequence according to the environment of equation system data flows. Therefore, modeling difficulty and complexity of a dynamic model of the propulsion system are effectively reduced. The reusability and expansibility of the model are improved.

Disclosed is a method for designing a die molded surface for forming a groove-shaped part. An inner molded surface M of the part is extracted, and section lines Pj of the extracted inner molded surface of the part are manufactured; the various section lines Pj are corrected sequentially through correction basis points, and new section lines are obtained; and surface fitting is carried out on the obtained section lines, and a new mold outer molded surface is obtained. The maximum deformation of the groove-shaped part formed through the mold outer molded surface determined through the method is reduced to 0.2mm from 1.5mm, and the design requirements of the part are met. Compared with the traditional technology, when the mold molded surface is designed, rebound deformation obtained after part deformation is considered in the method, and the problem that according to a traditional method, an inner molded surface of a composite material groove-shaped part is directly extracted to serve as the mold molded surface for manufacturing the composite material part, the part size cannot meet the design requirement is solved; and the mold repairing frequency of a follow-up mold is reduced, and therefore the production manufacturing cost is reduced, and the production efficiency is improved.

The invention discloses a power transmission line three-dimensional modeling method and system based on an oblique image. The method comprises the steps of obtaining the oblique image and the POS information of an unmanned aerial vehicle during a surveying and mapping scanning process; constructing a three-dimensional map model according to the oblique image; detecting and identifying a tower in the oblique image, and matching the identified tower with a tower model in a tower standing book system according to the POS information; and implanting the matched and returned tower model into a corresponding position in the three-dimensional map model to generate a three-dimensional model of a power transmission line. According to the power transmission line three-dimensional modeling method andsystem based on the oblique image provided by the invention, the establishment of the power transmission line three-dimensional model is realized through a terrain model and a tower model classification processing and synthesis method, the reuse rate of the tower model is improved, and the flight complexity and the surveying and mapping and modeling workload are reduced; and meanwhile, the problems of large deformation, shadow and shielding of the multi-view image inclined image are solved, and the matching quality is improved.

The invention discloses a virtual terminator model modeling and automatic ligature method based on machine learning. The virtual terminator model modeling and automatic ligature method based on machine learning comprises the following steps: acquiring at least SCD configuration files which have been finished at first; then analyzing the various acquired SCD configuration files to obtain IED information and virtual terminator ligature information in various SCD files; and establishing a knowledge set which comprises a typical IED model base, a virtual terminator typical ligature matrix and a virtual terminator typical ligature confidence matrix on the basis of an analysis result of the SCD files, then configuring the new SCD files, and carrying out automatic ligature on a virtual terminator on the basis of a current knowledge set. By the virtual terminator model modeling and automatic ligature method based on machine learning, the achievement of the configuration operation which has been finished can be utilized effectively, the modeling workload during actual operation is reduced, meanwhile, and compared with a typical IED judgment method, the virtual terminator model modeling and automatic ligature method based on machine learning can avoid interference caused by the reason that various IED factories have difference on standard understanding.

The invention discloses a processing method and system for cable laying modeling and a storage medium, and the method comprises the following steps: building a project file which comprises an electrical equipment model, a cable bridge model and a civil engineering model; importing a cable inventory file and an xml configuration file into the project file; obtaining a cable laying rule; performingcalculating to obtain a plurality of cable laying paths according to the cable laying rule, the cable inventory file and the xml configuration file; and modeling according to the plurality of cable laying paths to obtain a cable three-dimensional model. According to the invention, the cable inventory file and the xml configuration file are imported into the project file; the cable three-dimensional model is obtained by modeling the plurality of cable laying paths calculated according to the cable laying rule, and manual calculation and manual drawing of the cable paths are not needed in the whole process, so the accuracy of the cable laying paths is improved, and the modeling workload of modeling personnel is reduced. The method can be widely applied to the technical field of computer aided design.

The invention relates to a method for modeling an underground roadway three-dimensional scene by CAD (Computer Aided Design) drawings, which is used for reducing the modeling workload and improving the modeling efficiency and belongs to the fields of three-dimensional graphic modeling and software development. According to the technical scheme, the method comprises the following steps: firstly, converting a CAD drawing into a vector data format, and carrying out reading, storage, analysis and formatting processing on point data, line data and text data in the vector data format to obtain three-dimensional graph basic data; and then rendering the three-dimensional graph basic data by utilizing a graph engine to obtain a three-dimensional tunnel model, and finally constructing a graph editorto edit a three-dimensional scene. Manual drawing of an underground roadway by means of modeling software is avoided, and the purposes of reducing the modeling workload and improving the modeling speed are achieved.

The invention discloses a sensor measuring method based on double support vector machines, belonging to the field of sensor information processing. The method comprises the following steps: establishing a data sample by sensor calibration data under the working condition environment; constructing a sensor inverse model based on the support vector machines by applying the data sample; establishing a second data sample by applying error data of sensor output and sensor inverse model output to construct an error model based on the support vector machines; selecting and optimizing the parameters of the sensor inverse model and the error model through a quantum-behaved particle swarm optimization algorithm, as well as minimum standards of root-mean-square error of the model and maximum absolute error; and taking the difference of the sensor inverse model output and the error model output as a measured true valve. The sensor measuring method has the advantages that the influence of non-linearity property of the sensor on the measuring result can be effectively lowered; the measured high-precision measurement is realized; the modeling working amount is low; and the sensor measuring method based on the double support vector machines can be widely applied to the field of high-precision measurement of various sensors.

The invention discloses a method for designing a switchyard structure. The method is used for complex transformer substation frameworks with multiple steel beams and complex modeling. The method comprises the following steps that A, steel beams are grouped according to the lengths of steel beams of the switchyard structure; B, reference steel beams are selected to be analyzed, and the overall rigidity of the reference steel beams is calculated; C, an equivalent steel beam model which is identical with the reference steel beams in overall rigidity and simple in structure is built; D, a structure model composed of equivalent steel beams and structure columns is built, finite element analysis is carried out, and section design of the structure columns is completed; E, a space lattice type steel beam model is built, internal force of ends of the equivalent steel beams and loads borne by the equivalent steel beams are exerted to the space lattice type steel beam model, finite element analysis is carried out, and section design of the reference steel beams is completed; F, design results are applied to other steel beams of the same group, and design of the switchyard structure is completed. The method can simplify the complex structure, reduce modeling workload, shorten the design cycle and improve work efficiency.

The invention relates to a method for modeling three-dimensional scenes of underworkings by using a CAD drawing, capable of reducing the modeling workload and improving the modeling efficiency, and belongs to the fields of three-dimensional graph modeling and software development. According to the technical scheme, the method comprises the following steps: firstly, converting a CAD drawing into avector data format, and carrying out reading, storage, analysis and formatting processing on point data, line data and text data in the vector data format to obtain three-dimensional graph basic data;and then rendering the three-dimensional graph basic data by utilizing a graph engine to obtain a three-dimensional tunnel model, and finally constructing a graph editor to edit a three-dimensional scene. Manual drawing of an underground roadway by means of modeling software is avoided, and the purposes of reducing the modeling workload and improving the modeling speed are achieved.

The invention relates to the technical field of 3D mesh model construction. The invention particularly relates to a method for constructing a 3D mesh model of a large-scale underground engineering under complex geological conditions, comprising the following steps: firstly, constructing or importing a geometric model of a excavated body with a relatively regular shape according to a point, a lineand a surface in other data formats in a mesh generation software ANSYS; and secondly, constructing a 3D mesh model of the excavated body according to the geometric model. 2, establishing a partitionbody with a certain range and simple geometrical shape around the excavation body as a key analysis area; Step 3: outside the key analysis area, delimiting a simple and regular separator as the secondary key concern area, and the mesh can be roughly divided appropriately. The invention has the advantages that: in the excavation process of layered, step-by-step and subsection, if the previously unknown discontinuous surface occurs or the occurrence of the known discontinuous surface changes, the method can change the original 3D mesh model in time without re-establishing / updating the geometricmodel and then dividing the mesh to establish the 3D mesh model.

The invention discloses a convolution self-encoding fault monitoring method based on batch imaging, and belongs to the technical field of batch process fault monitoring. The method comprises two stepsof off-line modeling and on-line monitoring. The offline modeling step comprises the following steps: firstly, normalizing three-dimensional data of the intermittent process; then, taking the two-dimensional array of each batch as an image to be directly input into a convolutional auto-encoder (CAE) to carry out deep unsupervised feature learning; and finally, constructing statistics and corresponding control limits for the features learned by the CAE by utilizing a support vector machine. The online monitoring step includes: normalizing the collected data, and carrying out batch filling; inputting the normalized and filled batch graph into the trained CAE to learn features; and calculating an online statistic, and comparing the online statistic with an offline control limit. Compared with the prior art, the technical scheme provided by the invention avoids information loss caused by data expansion, does not need to divide stages to reduce modeling workload, deeply extracts change characteristics of process variables, and reduces false alarm and missing report rate of intermittent process monitoring.

The invention provides a modeling method and system for a data analysis model, and relates to the field of data analysis modeling. The modeling method for the data analysis model comprises the following steps: obtaining a service demand and a data set set set by a user on a user interface; determining a corresponding time sequence prediction scene according to data characteristics of the data set, and determining a simulation data analysis target according to service requirements; determining a modeling strategy through the data characteristics and the corresponding time sequence prediction scene, selecting at least one analysis angle for the target, and analyzing the target through different angles; and constructing a data analysis and calculation model according to the modeling strategy and the target analysis result. Data visualization analysis modeling can be achieved, models generated by modeling are managed in a unified mode, the modeling workload is reduced, and the modeling efficiency and effectiveness are improved. In addition, the invention further provides the modeling system for the data analysis model. The modeling system comprises an acquisition module, a determination module, an analysis module and a construction module.

The invention provides a visual modeling and conversion method and a system of a Monte Carlo program, comprising the following steps: generating each conversion geometry corresponding to the geometrymodel to be converted according to a model file of the geometry model to be converted stored locally; and converting the geometry model to be converted. Each of the conversion geometries has geometrical information, the geometrical information includes absolute spatial information and geometrical shape information; Acquiring attribute information of each conversion geometry from a locally stored parameter file of a geometry model to be converted; The attribute information includes: material information and parent-child relationship information; According to each conversion geometry, attributeinformation and geometric information of each conversion geometry, the geometric model to be converted is converted according to the geometric description markup language standard. The geometrical model and the conversion method created by the invention have the beneficial effects of improving modeling efficiency, lightening user modeling workload, and improving model conversion accuracy and spatial relationship description.

The invention provides an oil pump performance curve estimation method based on an ADMGA (Adaptive Dissortative Mating Genetic Algorithm) and in particular relates to a method for building an oil pump working characteristic curved surface model. The method comprises the steps: 1) measuring the test data of an oil pump on an oil pump performance test bench; 2) according to the experimental data, estimating the value range of parameters a, b, c and d in the formula (1); 3) updating the parameter domain of genetic iterations; and 4) estimating the oil pump performance curve by adopting the ADMGA. The method used for building the oil pump working characteristic curved surface model has the advantages of unitary expression, less modeling work load and simplicity in operation. According to the invention, not only can the global search capability improved, but also the search time is shortened, and the search performance of the genetic algorithm is improved.

The invention provides a steel bar modeling method and device based on a BIM model and computer equipment. The method comprises the steps of acquiring a BIM model, wherein the BIM model comprises oneor more structural components; determining steel bar arrangement information on a working surface of the structural member, the working surface comprising a geometric surface and / or a profile of the structural member; drawing a steel bar structure on the working surface according to the steel bar arrangement information; and drawing a steel bar model of the structural member according to the steelbar structure on the working surface. By means of the method and device, when steel bar modeling is conducted on the basis of the BIM, steel bar primitives do not need to be drawn and arranged one byone, and the drawing workload is reduced.

The invention relates to the technical field of gradient material mechanical analysis, in particular to a calculation method of a gradient material macro equivalent elastic modulus, and the problem that the acquisition efficiency of an existing gradient material macro equivalent elastic modulus is low is solved. The calculation method comprises the steps that relation expressions of the average stress sigma bar of a homogeneous medium layer and the internal stress of materials of all phases are obtained; a relation expression of a disturbance strain difference epsilon' of an inclusion phase and equivalent eigen strain epsilon* is obtained; a disturbance stain difference sigma', disturbance stress sigma bar of a primary phase and disturbance stain epsilon bar of the primary phase are obtained; relation expressions of the homogeneous medium layer average stress sigma bar and the average strain epsilon bar are determined; equivalent moduli L of all homogeneous medium layers are obtained; equivalent bulk moduli K and equivalent shear moduli G of all the homogeneous medium layers are obtained. According to the calculation method of the gradient material macro equivalent elastic modulus, macro equivalent elastic models and the Poisson ratio can be obtained, the material property distribution situation in the gradient region can be obtained to conduct detail stress analysis conveniently, and the modeling workload for conducting detail analysis on the gradient region can be greatly reduced.

The invention discloses a multi-level whole-process monitoring method for power equipment. The method comprises the following steps: rendering and displaying a three-dimensional scene in a designatedarea of a two-dimensional picture of the power equipment for displaying an online operation state; switching to a three-dimensional scene of corresponding equipment by activating equipment pictures inthe two-dimensional picture; wherein the three-dimensional scene is sequentially divided into a plurality of levels of three-dimensional scene models according to the connection relationship of the power equipment; the switching among the three-dimensional scene models of multiple levels is realized by activating the movable parts with the hierarchical link information in the three-dimensional scene models, and the display of the equipment information in the three-dimensional scene models of multiple levels is realized through instantiated variable transmission. According to the multi-level whole-process monitoring method for the power equipment, the whole-process monitoring from the whole structure of the power equipment to the details of the single equipment is realized, and the equipment perception capability, the defect discovery capability, the state management and control capability and the emergency disposal capability of operation and maintenance personnel on the power equipment of a converter station are improved.

The invention is based on an emerging Modelica language in recent years in the field of computer modeling and simulation, and particularly provides a missile-borne electrical system modeling method. The method mainly comprises the following steps of decomposing a missile-borne electrical system into a plurality of unit-level models; carrying out modeling on the unit-level model by adopting a modeling mode of differential equation description or finite element method description or function fitting method description; identifying the unit-level model; assembling the unit-level models into an on-missile electrical system model; and completing simulation of the on-missile electrical system model. According to the method, modeling difficulty and the modeling workload of the missile-borne electrical system can be reduced, so cost is saved, and research and development efficiency of the missile-borne electrical system is improved.

The present invention relates to an electric power real-time digital simulation system fault point control system. The system comprises N fault point control modules which are cascaded in order, and any one of 1-N fault point control modules is configured to allow the sum of the fault points to subtract the number of the fault points of the current fault point control module through a cascading unit and then output the date to another fault point control module after the fault point control module; the first time fault time, the first time fault position and the first time fault type of the fault point of the current fault point control module are arranged through a setting unit; an operation unit is configured to perform multiplication operation of the first time fault time, the first time fault position and the first time fault type of the fault point of the current fault point to obtain the first time fault control variation of the fault point; and a control unit is configured to obtain the fault control parameter of the fault point according to the first time fault control variation and control the fault point according to the fault control parameter. According to the technical scheme, the uniform control of infinite fault points of an electric power real-time digital simulation system is realized.

The invention relates to the field of collaborative robots and particularly relates to a seven-degree-of-freedom collaborative robot rigidity modeling and identification method. The method comprises the following steps of step 1, kinematics modeling being performed on a robot, and robot joint parameters being defined; 2, rigidity modeling being conducted on the robot; 3, selecting an inverse condition number as an observability index of the optimal method, solving the individual influence of each joint on the inverse condition number, and obtaining a good recognition area in the joint space according to the influence of each joint on the inverse condition number; and 4, calculating joint rigidity. According to the method, the seven-degree-of-freedom robot rigidity identification pose selection is studied, the inverse condition number is used as an observation index to determine a good identification area with relatively high flexibility of the robot, identification precision of the rigidity model is improved, and optimal configuration selection can be effectively carried out.

The invention provides a simulation method for the dynamic performance of a loop accelerated loading device, and the method comprises the steps: respectively building three-dimensional solid models ofa plurality of sub-assemblies in the loop accelerated loading device in three-dimensional solid modeling software, and simplifying the three-dimensional solid models; respectively importing the simplified three-dimensional solid model of the sub-assembly of the loop accelerated loading device into multi-body dynamics simulation software; adding an assembly body internal constraint to each sub-assembly body in multi-body dynamics simulation software; establishing an overall file of the loop accelerated loading device, importing files corresponding to the sub-assemblies into the overall file, and adding constraints among the sub-assemblies; adding a three-dimensional solid model of the pavement and the tire; applying traction force and hydraulic cylinder pressure to the overall file according to different working conditions, simulating the operation condition of the loop acceleration loading device, and exporting various characteristic parameters representing dynamic performance after simulation is finished. The test cost can be reduced, the design efficiency is improved, and the product development time is shortened.

The invention discloses a virtual human quick modeling method oriented to the ship virtual reality field. The method comprises the following steps: 1), determining a needed module according to the modeling demands of a customized virtual human; 2) judging whether each module in the model database meets a modeling requirement or not; 3) determining modules required for virtual human modeling and lacking in the model library, and judging whether a module with reference value exists or not; 4) performing parameter modification modeling on the module with the reference value; 5) performing model reconstruction on the missing modules without reference values; 6) performing quick combined modeling on the modules meeting the customization requirements to generate a virtual human model; accordingto the virtual human modeling method, a virtual human modeling mode combining modular modeling, parametric modeling and personalized feature modeling is adopted, technicians are liberated from bottomlayer model wiring modeling work, the modeling period is shortened, and the modeling workload and cost are reduced.

The invention relates to a reactive power control method and system for a photovoltaic power generation system, comprising the steps of obtaining a reference value signal of the photovoltaic power generation system in the reactive power control mode based on a set reactive power control mode of a photovoltaic power generation system model; controlling the output of the converter according to a reference value signal; The reactive power control mode comprises a voltage control mode, a reactive power control mode and a reactive power-voltage coordination control mode. Through the scheme, the photovoltaic power generation system can adapt to different reactive power regulation requirements of the power grid.