Model order reduction system and model order reduction method

Abstract

The invention is suitable for the technical field of model simulation optimization, and provides a model order reduction system and a model order reduction method.The model order reduction system comprises a model import module, a method selection module and an order reduction processing module which are in communication connection in sequence; the model importing module is used for importing a model to be subjected to order reduction; the method selection module is used for displaying identification information of the multiple model order reduction methods and obtaining a first method selection instruction, the first method selection instruction is used for indicating a first target model order reduction method, and the first target model order reduction method is any one model order reduction method in the multiple model order reduction methods; and the order reduction processing module is used for performing order reduction processing on the model to be subjected to order reduction based on the first target model order reduction method to obtain a first order reduction model. According to the technical scheme, the multiple model order reduction methods are provided, the target model order reduction method selected by the user from the multiple model order reduction methods is used for carrying out order reduction processing on the model to be subjected to order reduction, and the accuracy of model order reduction is improved.

Description

Technical Field

[0001] The present application relates to the technical field of model simulation optimization, and in particular to a model order reduction system and a model order reduction method. Background Art

[0002] With the development of science and technology, more and more accurate mathematical models can be constructed to describe real physical processes. These models usually contain a large number of variables and parameters, and therefore require huge computing resources. However, in practical applications, such as real-time control and rapid decision-making, rapid response is required, which requires the model to be simple and computationally small in order to quickly provide useful information.

[0003] At present, a fixed model reduction method is usually used to reduce the model order, simplifying the high-dimensional complex model into a low-dimensional model to reduce the complexity of the model while maintaining the main features and dynamic behavior of the model. However, the accuracy of model reduction according to the current technical solution is low.

[0004] Therefore, how to improve the accuracy of model reduction has become an urgent problem to be solved. Summary of the invention

[0005] The embodiments of the present application provide a model order reduction system and a model order reduction method, which aim to solve the problem of low accuracy of model order reduction.

[0006] In a first aspect, an embodiment of the present application provides a model reduction system, characterized in that the system comprises a model import module, a method selection module and a reduction processing module which are communicatively connected in sequence; the model import module is used to import the model to be reduced; the method selection module is used to display identification information of multiple model reduction methods, and obtain a first method selection instruction, wherein the first method selection instruction is used to indicate a first target model reduction method, wherein the first target model reduction method is any one of the multiple model reduction methods; the reduction processing module is used to reduce the model to be reduced based on the first target model reduction method to obtain a first reduced-order model.

[0007] In a possible implementation, the system also includes a model input module and a model conversion module, and the model conversion module is communicatively connected to the model input module and the method selection module; the model input module is used to input the transfer function or state equation of the model to be reduced; the model conversion module is used to generate the model to be reduced based on the transfer function or the state equation.

[0008] In a possible implementation, the model input module includes: a transfer function input unit, used to input the coefficients of the numerator and the denominator in the transfer function; and a state equation input unit, used to input the coefficient matrix of the state equation.

[0009] In one possible implementation, the system also includes a parameter setting module, which is communicatively connected to the method selection module and the order reduction processing module; the parameter setting module is used to display parameter identification information corresponding to the first target model order reduction method, and obtain a first parameter setting instruction, wherein the first parameter setting instruction is used to indicate first parameter information of the first target model order reduction method; accordingly, the order reduction processing module is specifically used to reduce the order of the model to be reduced based on the first target model order reduction method and the first parameter information to obtain the first reduced-order model.

[0010] In a possible implementation, the system also includes a graphics selection module and a graphics display module, wherein the graphics selection module is communicatively connected to the order reduction processing module and the graphics display module; the graphics selection module is used to display identification information of multiple characteristic curves and obtain a graphics selection instruction, wherein the graphics selection instruction is used to indicate a target characteristic curve, wherein the target characteristic curve is any one or more of the multiple characteristic curves; and the graphics display module is used to display the target characteristic curves corresponding to the model to be reduced and the first reduced-order model respectively.

[0011] In a possible implementation, the graphic display module includes: a region selection unit, used to obtain a region selection instruction, wherein the region selection instruction is used to indicate a target region, and the target region is a region in any target characteristic curve; and a graphic update unit, used to update the target characteristic curve based on the target region.

[0012] In a possible implementation, the system further includes an error display module, which is communicatively connected to the order reduction processing module; the error display module is used to calculate and display error information between the model to be reduced-order and the first reduced-order model.

[0013] In a possible implementation, the parameter setting module is also used to: obtain a second parameter setting instruction based on the error information, and the second parameter setting instruction is used to indicate second parameter information of the first target model order reduction method; accordingly, the order reduction processing module is also used to reduce the order of the model to be reduced based on the first target model order reduction method and the second parameter information to obtain a second reduced order model.

[0014] In a possible implementation, the method selection module is also used to: obtain a second method selection instruction based on the error information, the second method selection instruction is used to indicate a second target model order reduction method, the second target model order reduction method is any one of the multiple model order reduction methods except the first target model order reduction method; accordingly, the order reduction processing module is also used to reduce the order of the model to be reduced based on the second target model order reduction method to obtain a third reduced order model.

[0015] In a second aspect, an embodiment of the present application provides a model reduction method, which is applied to a model reduction system as described in the first aspect or any one of the implementation methods thereof, and the method includes: importing a model to be reduced; obtaining a method selection instruction based on the model to be reduced, the method selection instruction being used to indicate a target model reduction method, the target model reduction method being any one of a plurality of model reduction methods; and reducing the model to be reduced based on the target model reduction method to obtain a reduced model.

[0016] Compared with the prior art, the embodiments of the present application have the following beneficial effects: the model to be reduced is imported through the model import module, and multiple model reduction methods are provided through the method selection module, so that the user can select a model reduction method from multiple model reduction methods as the target model reduction method according to the characteristics of the model to be reduced or actual needs, and then the model to be reduced is reduced based on the target model reduction method through the reduction processing module to obtain a reduced model. Compared with using the same fixed model reduction method to reduce different models to be reduced, the accuracy of model reduction is improved.

[0017] It can be understood that the model reduction method provided in the embodiment of the present application has the same beneficial effects as the above-mentioned model reduction system, which will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 A schematic diagram of the architecture of a model order reduction system provided in one embodiment of the present application;

[0020] Figure 2 A schematic diagram of the processing flow of a balanced truncation model reduction method;

[0021] Figure 3Schematic diagram of the process flow for selecting a reduction method for a pattern;

[0022] Figure 4 A schematic diagram of the processing flow of a zero-pole simplification and order reduction method;

[0023] Figure 5 A schematic diagram of the architecture of another model order reduction system provided in one embodiment of the present application;

[0024] Figure 6 A schematic diagram of the architecture of another model order reduction system provided in an embodiment of the present application;

[0025] Figure 7 A schematic diagram of parameter settings for a balanced truncation model reduction method provided in one embodiment of the present application;

[0026] Figure 8 A schematic diagram of parameter settings for a mode selection reduction method provided in an embodiment of the present application;

[0027] Fig. 9 A schematic diagram of parameter settings for a zero-pole simplification and order reduction method provided in one embodiment of the present application;

[0028] Fig.10 A schematic diagram of the architecture of another model order reduction system provided in one embodiment of the present application;

[0029] Fig.11 A schematic diagram of a target characteristic curve obtained by reducing the model order using a balanced truncation model reduction method provided in an embodiment of the present application;

[0030] Fig.12 A schematic diagram of a target characteristic curve obtained by reducing the model order using a mode selection reduction method provided in an embodiment of the present application;

[0031] Fig.13 A schematic diagram of the architecture of another model reduction system provided in one embodiment of the present application;

[0032] Fig.14 A schematic diagram of an error type selection interface provided by an embodiment of the present application;

[0033] Fig.15 A schematic diagram of an absolute error between a to-be-reduced-order model and a first reduced-order model provided in one embodiment of the present application;

[0034] Fig.16 A schematic diagram of a relative error between a to-be-reduced-order model and a first reduced-order model provided in one embodiment of the present application;

[0035] Fig.17A schematic diagram of a flow chart of a model reduction method provided in one embodiment of the present application;

[0036] Fig.18 A flowchart of another model reduction method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0037] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0038] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0039] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0040] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0041] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0042] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0043] With the development of science and technology, more and more accurate mathematical models can be constructed to describe real physical processes. These models usually contain a large number of variables and parameters, and therefore require huge computing resources. However, in practical applications, such as real-time control and rapid decision-making, rapid response is required, which requires the model to be simple and computationally small in order to quickly provide useful information.

[0044] In recent years, model reduction has been successfully applied to many fields such as control, fluid mechanics, structural mechanics, thermal analysis, multiphysics, medicine, etc. Given the high computational efficiency of reduced-order models, model reduction is mainly used to speed up product design and optimization, or as a digital twin in real-time control tools. It is an important technology in mathematical modeling and computational science, especially in the simulation and control design of complex systems.

[0045] At present, a fixed model reduction method is usually used to reduce the model order, simplifying the high-dimensional complex model into a low-dimensional model to reduce the complexity of the model while maintaining the main characteristics and dynamic behavior of the model. These low-dimensional models are more efficient in calculation and convenient for analysis, design and control. However, the accuracy of model reduction according to the current technical solution is low.

[0046] In order to solve the above technical problems, the present application proposes a model reduction system, which includes a model import module, a method selection module and a reduction processing module that are communicatively connected in sequence; the model import module is used to import the model to be reduced; the method selection module is used to display the identification information of multiple model reduction methods, and obtain a first method selection instruction, the first method selection instruction is used to indicate a first target model reduction method, and the first target model reduction method is any one of the multiple model reduction methods; the reduction processing module is used to reduce the model to be reduced based on the first target model reduction method to obtain a first reduced model, thereby improving the accuracy of model reduction.

[0047] For ease of understanding, the technical solution of the present application will be described in detail below with reference to the accompanying drawings.

[0048] Figure 1 Schematic diagram of the architecture of a model reduction system provided in one embodiment of the present application. Figure 1 As shown, the model order reduction system 10 includes a model import module 101, a method selection module 102 and an order reduction processing module 103 which are communicatively connected in sequence.

[0049] Specifically, the model importing module 101 is used to import the model to be reduced.

[0050] In a specific implementation, the model file is directly imported through the model import module 101, the model file includes the model to be reduced in order, and the imported model file is parsed to obtain the model to be reduced in order.

[0051] Specifically, the method selection module 102 is used to display identification information of multiple model order reduction methods, obtain a first method selection instruction, and the first method selection instruction is used to indicate a first target model order reduction method, and the first target model order reduction method is any one of the multiple model order reduction methods.

[0052] In one possible implementation, the method selection module 102 displays identification information of multiple model order reduction methods, and a selection component is set corresponding to the identification information of each model order reduction method. The user selects the first target model order reduction method by clicking the selection component. The method selection module 102 obtains the first method selection instruction through the user's clicking operation, and the model order reduction method corresponding to the selection component clicked by the user is the first target model order reduction method.

[0053] In another possible implementation, the method selection module 102 displays identification information of multiple model order reduction methods and provides an input component, in which the user inputs identification information corresponding to the first target model order reduction method. The method selection module 102 obtains the first method selection instruction through the user's input operation, and the model order reduction method corresponding to the identification information input by the user is the first target model order reduction method.

[0054] As an example, the identification information of the multiple model order reduction methods displayed by the method selection module 102 includes the names, abbreviations, numbers or English abbreviations of the multiple model order reduction methods.

[0055] As an example, model order reduction methods include balanced truncation model order reduction method, mode selection order reduction method, zero-pole simplification order reduction method, etc. Among them, balanced truncation model order reduction is a technology commonly used for linear time-invariant system dynamic system dimensionality reduction. Its core idea is to retain the "primary" dynamic characteristics of the system while removing the "secondary" dynamic characteristics to obtain a low-order model, which can approximate the main behavior of the original system in a sense; the mode selection order reduction method focuses on a specific frequency region and discards the region outside the frequency range; the zero-pole simplification order reduction method accurately reduces the order of the model by canceling zero-pole pairs or eliminating states that have no effect on the response of the entire model.

[0056] Exemplarily, the method selection module 102 displays the abbreviations of multiple model reduction methods, namely balanced truncation, mode selection and zero-pole simplification. The user clicks to select the selection component corresponding to balanced truncation in the method selection module 102. The method selection module 102 obtains the first method selection instruction through the click operation. The balanced truncation model reduction method is the first target model reduction method.

[0057] Specifically, the order reduction processing module 103 is used to perform order reduction processing on the model to be reduced based on the first target model order reduction method to obtain a first reduced order model.

[0058] As an example, when the first target model reduction method is a balanced truncation model reduction method, the specific reduction process is as follows: Figure 2 As shown, it includes starting to input the system {A, B, C, D}; solving the fully controllable Gram matrix P and the fully observable Gram matrix Q through the Lyapunov equation; selecting the Bartels-Stewart method to solve the Lyapunov equation; decomposing P = Lc*Lc', Q = Lo*Lo', where Lc and Lo are lower triangular matrices; performing singular value decomposition (Singular Value Decomposition, SVD) on the matrix Lc'*Lo Lc'*Lo = U*∑*V, where U and V are orthogonal matrices, and ∑ is a diagonal matrix; setting T = Lc*U*K (K is the inverse square root of ∑), then T is the balanced transformation matrix of the system; performing balanced transformation on the system through the T matrix to obtain {A', B', C', D'}, at this time P'=Q'=diag{σ1,σ2,…σ n}; select the reduction order through the P and Q matrices, and divide {A', B', C', D'} into blocks to obtain the first reduced-order model.

[0059] Exemplarily, the algorithm principle of the balanced truncation model reduction method is as follows:

[0060] By solving the Lyapunov equations (1) and (2) for continuous systems, or (3) and (4) for discrete systems, the square root factors S and R of the controllability and observability Grammian matrices in (5) are calculated.

[0061] AP+PA T +BB T =0 (1)

[0062] A T Q+QA+C T C=0 (2)

[0063] APA T +BB T =P (3)

[0064] A T QA+C T C=Q (4)

[0065] P=S T SQ=R T R (5)

[0066] Calculate SR using formula (6) T SVD decomposition of

[0067] SR T =[U1 U2 U3]diag(Σ1 Σ2 0)[V1 V2 V3] T (6)

[0068] Calculate the orthogonal triangular decomposition (QR decomposition) for i=1,2

[0069] S T U i =M i X i R T V i =N i Y i

[0070] Among them, M i , N i is a matrix with orthogonal columns, X i , Y i is a non-singular matrix;

[0071] Calculate SVD for i=1,2 S : Calculate Z in formula (7) + and Z,

[0072]

[0073] in The minimum system (Z + AZ,Z + B, CZ, D),

[0074]

[0075] If the matrix of the system (A, B, C, D) is divided into the following form:

[0076]

[0077] With Z -1 PQZ=diag(X1,X2,0) has the same structure, so the result of the system (A,B,C,D) is:

[0078]

[0079] Among them, the parameter definitions are shown in Table 1:

[0080] Table 1

[0081]

[0082]

[0083] As another example, when the first target model order reduction method is a mode selection order reduction method, the specific order reduction process is as follows: Figure 3 As shown, it includes importing the model; converting the model into zero-pole form; performing partial fraction expansion through the residue algorithm; planning the frequency interval; canceling the poles outside the frequency interval; converting the reduced-order model, and exporting the reduced-order model.

[0084] As another example, when the first target model reduction method is a zero-pole simplification reduction method, the specific reduction process is as follows: Figure 4 As shown, it includes importing the model that needs to be reduced in order; converting the model into the form of zeros and poles; setting the tolerance for identifying and canceling zero-pole pairs; searching for pole pairs to find zeros that meet the tolerance range, where the tolerance range is the product of the real part of the pole and the tolerance value; canceling the zero-pole pairs; converting the uncanceled zeros and poles into the reduced-order model; and exporting the model.

[0085] The technical solution provided by this embodiment imports the model to be reduced in order through the model import module, and provides multiple model reduction methods through the method selection module, so that the user can select a model reduction method from multiple model reduction methods as the target model reduction method according to the characteristics of the model to be reduced in order or actual needs, and then reduce the model to be reduced in order based on the target model reduction method through the reduction processing module to obtain a reduced model. Compared with using the same fixed model reduction method to reduce the order of different models to be reduced in order, the accuracy of model reduction is improved.

[0086] Figure 5 A schematic diagram of the architecture of another model reduction system provided in an embodiment of the present application. Figure 5 As shown, based on the above embodiments, this embodiment further illustrates and optimizes the technical solution. Specifically, the model reduction system 50 in this embodiment also includes a model input module 501 and a model conversion module 502. The model conversion module 502 is communicatively connected with the model input module 501 and the method selection module 102.

[0087] Specifically, the model input module 501 is used to input the transfer function or state equation of the model to be reduced; the model conversion module 502 is used to generate the model to be reduced based on the transfer function or state equation.

[0088] In a specific implementation, when the input of the model input module 501 is the transfer function of the model to be reduced, the model conversion module 502 first converts the transfer function into a state equation, and then generates the model to be reduced based on the state equation; when the input of the model input module 501 is the state equation of the model to be reduced, the model conversion module 502 directly generates the model to be reduced from the state equation.

[0089] As an example, the model input module 501 includes a transfer function input unit and a state equation input unit. The transfer function input unit is used to input the coefficients of the numerator and the denominator in the transfer function; the state equation input unit is used to input the coefficient matrix of the state equation.

[0090] The technical solution provided in this embodiment inputs the transfer function or state equation of the model to be reduced order through the model input module, and then the model conversion module generates the model to be reduced order according to the transfer function or state equation, thereby improving the flexibility of inputting the model to be reduced order.

[0091] Figure 6 A schematic diagram of the architecture of another model reduction system provided in an embodiment of the present application. Figure 6As shown, based on the above embodiments, this embodiment further illustrates and optimizes the technical solution. Specifically, the model reduction system 60 in this embodiment also includes a parameter setting module 601, and the parameter setting module 601 is communicated with the method selection module 102 and the reduction processing module 103.

[0092] Specifically, the parameter setting module 601 is used to display parameter identification information corresponding to the first target model reduction method, and obtain a first parameter setting instruction, where the first parameter setting instruction is used to indicate first parameter information of the first target model reduction method; accordingly, the reduction processing module 103 is specifically used to perform reduction processing on the model to be reduced based on the first target model reduction method and the first parameter information to obtain a first reduced order model.

[0093] In a specific implementation, after the first target model reduction method is determined by the method selection module 102, the parameter setting module 601 displays the parameter identification information corresponding to the first target model reduction method, for example, the parameter identification information is the parameter name; an input component is provided in the parameter setting module 601, and the user inputs the parameter value corresponding to the parameter identification information in the input component, and the parameter setting module 601 obtains the first parameter setting instruction through the user's input operation, and the parameter value input by the user in the input component is the first parameter information; the reduction processing module 103 performs reduction processing on the model to be reduced based on the first target model reduction method and the first parameter information to obtain a first reduced-order model.

[0094] As an example, Figure 7 As shown, when the first target model reduction method is the balanced truncation model reduction method, the parameter setting module 601 displays the parameter identification information corresponding to the balanced truncation model reduction method as the order, and the user enters the order value, for example 20, in the input component provided by the parameter setting module 601; the parameter setting module 601 obtains the first parameter setting instruction based on the input operation, and determines that the first parameter information is the order 20, and the reduction processing module 103 uses the balanced truncation model reduction method to reduce the order of the model to be reduced to the order 20 to obtain the first reduced order model.

[0095] As another example, Figure 8As shown, when the first target model order reduction method is the mode selection order reduction method, the parameter setting module 601 displays the parameter identification information corresponding to the mode selection order reduction method as the cutoff frequency lower limit and the cutoff frequency upper limit, and the user inputs the cutoff frequency lower limit value and the cutoff frequency upper limit value in the input component provided by the parameter setting module 601, for example, the cutoff frequency lower limit value is 2, and the cutoff frequency upper limit value is 50; the parameter setting module 601 obtains the first parameter setting instruction based on the input operation, and determines that the first parameter information is the cutoff frequency lower limit value 2 and the cutoff frequency upper limit value 50, and the order reduction processing module 103 uses the mode selection order reduction method to reduce the order of the model to be reduced with the cutoff frequency lower limit value 2 and the cutoff frequency upper limit value 50, and obtains the first reduced order model.

[0096] As another example, Fig. 9 As shown, when the first target model reduction method is the zero-pole simplification reduction method, the parameter setting module 601 displays the parameter identification information corresponding to the zero-pole simplification reduction method as tolerance, and the user enters the tolerance value, such as 0.1, in the input component provided by the parameter setting module 601; the parameter setting module 601 obtains the first parameter setting instruction based on the input operation, and determines that the first parameter information is the tolerance 0.1, and the reduction processing module 103 uses the zero-pole simplification reduction method to reduce the order of the model to be reduced with a tolerance of 0.1 to obtain a first reduced order model.

[0097] The technical solution provided in this embodiment can set the parameters involved in the model order reduction method through a parameter setting module according to the actual needs of the user, thereby improving the accuracy of the model order reduction.

[0098] Fig.10 A schematic diagram of the architecture of another model reduction system provided in an embodiment of the present application. Fig.10 As shown, based on the above embodiments, this embodiment further illustrates and optimizes the technical solution. Specifically, the model reduction system 100 in this embodiment also includes a graphic selection module 1001 and a graphic display module 1002. The graphic selection module 1001 is communicatively connected with the reduction processing module 103 and the graphic display module 1002.

[0099] Specifically, the graphic selection module 1001 is used to display the identification information of multiple characteristic curves and obtain the graphic selection instruction, the graphic selection instruction is used to indicate the target characteristic curve, and the target characteristic curve is any one or more of the multiple characteristic curves; the graphic display module 1002 is used to display the target characteristic curves corresponding to the model to be reduced and the first reduced-order model respectively.

[0100] In a specific implementation, the graphic selection module 1001 displays identification information of multiple characteristic curves, and a selection component is set corresponding to the identification information of each characteristic curve. The user selects the corresponding characteristic curve by clicking the selection component. The graphic selection module 1001 obtains the graphic selection instruction through the user's clicking operation. The one or more characteristic curves corresponding to the one or more selection components clicked by the user are the target characteristic curves. The graphic display module 1002 displays the target characteristic curves corresponding to the model to be reduced and the first reduced-order model, respectively.

[0101] As an example, the identification information of the multiple characteristic curves displayed by the graphic selection module 1001 includes the names, abbreviations, numbers or English abbreviations of multiple model reduction methods.

[0102] As an example, the multiple characteristic curves displayed by the graph selection module 1001 include a step graph, a Bode graph, a pulse graph, a Nyquist graph, a Nichols graph, a singular value graph, a zero-pole graph, and the like.

[0103] In a possible implementation, the graphic display module 1002 includes a region selection unit and a graphic update unit. The region selection unit is used to obtain a region selection instruction, the region selection instruction is used to indicate a target region, and the target region is a region in any target characteristic curve; the graphic update unit is used to update the target characteristic curve based on the target region.

[0104] For example, Fig.11 and Fig.12 They are respectively a schematic diagram of the target characteristic curve obtained by reducing the model order using the balanced truncation model reduction method and a schematic diagram of the target characteristic curve obtained by reducing the model order using the mode selection reduction method, as shown in Fig.11 and Fig.12 As shown, taking a 48-order state equation as an example, the balanced truncation model reduction method and the mode selection reduction method are selected respectively, and the parameters corresponding to the balanced truncation model reduction method are set to 25 through the parameter setting module 601, and the parameters corresponding to the mode selection reduction method are set to 10 and 40 respectively. The Bode diagram is selected through the graphic selection module 1001. After clicking "Run", the graphic display module 1002 will automatically calculate and draw the Bode diagram of the model to be reduced and the first reduced model. Fig.11 In the diagram, the first picture is the amplitude diagram, the second picture is the phase diagram, and the third picture is the Hankel singular value diagram. In the equilibrium stage, a histogram of Hankel singular values ​​is also displayed. Click the histogram of Hankel singular values ​​through the area selection unit. The histograms correspond to different orders. After clicking the corresponding order, the graphic update unit automatically updates the above two Bode diagrams. Fig.12In the figure, the first figure is the amplitude diagram, the second figure is the phase diagram, and the third figure is the frequency range diagram. After dragging the blue area in the frequency range diagram through the area selection unit, the two Bode diagrams above are automatically updated through the graphic update unit.

[0105] The technical solution provided in this embodiment selects the target characteristic curve through a graphic selection module, and displays the target characteristic curves corresponding to the model to be reduced and the first reduced-order model respectively through a graphic display module, so that the user can intuitively see the changes in the model reduction order, thereby improving the user experience.

[0106] Fig.13 A schematic diagram of the architecture of another model reduction system provided in an embodiment of the present application. Fig.13 As shown, based on the above embodiments, this embodiment further illustrates and optimizes the technical solution. Specifically, the model reduction system 130 in this embodiment also includes an error display module 1301, and the error display module 1301 is communicatively connected to the reduction processing module 103.

[0107] Specifically, the error display module 1301 is used to calculate and display error information between the to-be-reduced-order model and the first reduced-order model.

[0108] In the specific implementation, Fig.14 As shown, the error display module 1301 displays two error types, namely, absolute error and relative error, and a circular selection component is provided before each of the absolute error and the relative error. The user selects the corresponding error type by clicking the selection component. When the user selects the absolute error, the error display module 1301 displays a schematic diagram of the absolute error between the to-be-reduced-order model and the first reduced-order model, as shown in FIG. Fig.15 When the user selects relative error, the error display module 1301 displays a schematic diagram of the relative error between the to-be-reduced model and the first reduced-order model, such as Fig.16 shown.

[0109] Further, the user determines whether it is necessary to adjust the parameters of the first target model order reduction method or adjust the model order reduction method through the error information displayed by the error display module 1301. For example, when the absolute error or relative error between the to-be-reduced model and the first reduced-order model is greater than a preset threshold, the operation of adjusting the parameters of the first target model order reduction method or adjusting the model order reduction method is performed.

[0110] In a possible implementation, the parameter setting module 601 is also used to obtain a second parameter setting instruction based on the error information, and the second parameter setting instruction is used to indicate second parameter information of the first target model reduction method; accordingly, the reduction processing module 103 is also used to reduce the model to be reduced based on the first target model reduction method and the second parameter information to obtain a second reduced order model.

[0111] In a specific implementation, when the user determines that the parameters of the first target model reduction method need to be adjusted based on the error information displayed by the error display module 1301, the second parameter value of the first target model reduction method is input in the parameter setting module 601, and the parameter setting module 601 obtains the second parameter setting instruction in response to the user's input operation. The reduction processing module 103 performs reduction processing on the model to be reduced based on the first target model reduction method and the second parameter information indicated in the second parameter setting instruction to obtain a second reduced order model.

[0112] In another possible implementation, the method selection module 102 is further used to obtain a second method selection instruction based on the error information, where the second method selection instruction is used to indicate a second target model reduction method, where the second target model reduction method is any one of the multiple model reduction methods except the first target model reduction method; accordingly, the reduction processing module 103 is further used to perform reduction processing on the model to be reduced based on the second target model reduction method to obtain a third reduced order model.

[0113] In a specific implementation, when the user determines that the model reduction method needs to be adjusted based on the error information displayed by the error display module 1301, the user clicks on the selection component corresponding to the second target model reduction method in the method selection module 102. The method selection module 102 obtains the second method selection instruction in response to the user's click operation, and the reduction processing module 103 reduces the model to be reduced based on the second target model reduction method to obtain a third reduced-order model.

[0114] In addition, the parameters of the second target model reduction method can be set through the parameter setting module 601, and then the reduction processing module 103 will reduce the order of the model to be reduced based on the second target model reduction method and the corresponding parameter values ​​set. The specific operation steps can refer to the above-mentioned parameter setting and reduction processing methods, which will not be repeated here.

[0115] The technical solution provided in this embodiment displays the error information between the model to be reduced and the reduced-order model through an error display module, which makes it easy for users to understand the reduction effect. The parameters of the model reduction method can be adjusted or the model reduction method can be replaced through the error information, thereby improving the accuracy of the model reduction.

[0116] Fig.17 A flow chart of a model reduction method provided in an embodiment of the present application is provided. Based on the above embodiment, this embodiment further illustrates and optimizes the technical solution. Specifically, the method in this embodiment is applied to the above Figure 1 , Figure 5 , Figure 6 , Fig.10 as well as Fig.13The model is reduced to a system such as Figure 7 As shown, the method provided in this embodiment includes the following steps:

[0117] S1701, import the model to be reduced.

[0118] In a possible implementation, a model file is imported, the model file includes the model to be reduced in order, and the imported model file is parsed to obtain the model to be reduced in order.

[0119] In another possible implementation manner, a transfer function or a state equation of the model to be reduced order is obtained, and then the model to be reduced order is generated based on the transfer function or the state equation.

[0120] As an example, when the transfer function of the model to be reduced is obtained, the transfer function is first converted into a state equation, and then the model to be reduced is generated based on the state equation.

[0121] As another example, when the state equation of the model to be reduced is obtained, the state equation is directly converted into the model to be reduced.

[0122] S1702, obtaining a method selection instruction based on the model to be reduced, where the method selection instruction is used to indicate a target model reduction method, where the target model reduction method is any one of a plurality of model reduction methods.

[0123] In a specific implementation, the method selection instruction carries identification information of the target model order reduction method, and the target model order reduction method is determined based on the identification information carried in the method selection instruction.

[0124] As an example, the model order reduction method includes a balanced truncation model order reduction method, a mode selection order reduction method, a zero-pole simplification order reduction method, and the like.

[0125] As an example, the identification information of the target model order reduction method includes the name, abbreviation, number or English abbreviation of the target model order reduction method.

[0126] S1703, reducing the order of the model to be reduced based on the target model reduction method to obtain a reduced order model.

[0127] In the specific implementation, after determining the target model reduction method, the model to be reduced is reduced based on the target model reduction method, and the high-dimensional complex model to be reduced is simplified into a low-dimensional model to reduce the complexity of the model to be reduced while maintaining the main characteristics and dynamic behavior of the model to be reduced.

[0128] It should be noted that for the specific steps of model reduction, please refer to the above content and will not be repeated here.

[0129] The technical solution provided in this embodiment selects a model reduction method from multiple model reduction methods to reduce the model to be reduced according to the characteristics of the model to be reduced and actual needs, thereby improving the accuracy of model reduction.

[0130] Fig.18 A flow chart of another model reduction method provided in an embodiment of the present application. Based on the above embodiment, this embodiment further illustrates and optimizes the technical solution. Specifically, the method in this embodiment is applied to the above Figure 1 , Figure 5 , Figure 6 , Fig.10 as well as Fig.13 The model is reduced to a system such as Fig.18 As shown, the method provided in this embodiment includes the following steps:

[0131] Reading a state equation or a transfer function of a model to be reduced in order, and generating a model to be reduced in order based on the state equation or the transfer function;

[0132] Selecting a target model order reduction method from multiple model order reduction methods, and setting parameters corresponding to the target model order reduction method, performing an order reduction operation on the model to be reduced according to the target model order reduction method and the set parameters to obtain a reduced order model;

[0133] Plot the model to be reduced and the reduced-order model to obtain a characteristic curve and an error curve;

[0134] It is determined whether the order reduction effect reaches the expected result according to the error curve. If so, the order reduction operation is terminated; if not, it is further determined whether it is necessary to reset the parameters. If so, the parameters of the first target model order reduction method are reset before the order reduction operation is performed on the model to be reduced; if not, the target model order reduction method is reselected from multiple model order reduction methods, and subsequent operations are performed again until the order reduction effect reaches the expected result.

[0135] The technical solution provided in this embodiment can select any one of the model reduction methods from multiple model reduction methods to perform the reduction operation according to actual needs, and can customize the reduction parameters, thereby improving the accuracy of the model reduction; in addition, after completing the model reduction operation, the model to be reduced and the reduced model are plotted to obtain characteristic curves and error curves so that users can analyze the reduction results, thereby improving the user experience; further, it is determined whether the reduction effect meets expectations based on the plotted error curve. If it does not meet expectations, the reduction operation is performed again on the model to be reduced by adjusting the reduction parameters or changing the model reduction method, thereby further improving the accuracy of the model reduction.

[0136] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0137] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0138] It should be noted that the above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A model reduction system, characterized in that: The system comprises a model import module, a method selection module and an order reduction processing module which are sequentially connected in communication; The model import module is used to import the model to be reduced; The method selection module is used to display identification information of multiple model order reduction methods, and obtain a first method selection instruction, where the first method selection instruction is used to indicate a first target model order reduction method, and the first target model order reduction method is any one of the multiple model order reduction methods; The order reduction processing module is used to perform order reduction processing on the model to be reduced based on the first target model order reduction method to obtain a first reduced order model.

2. The system according to claim 1, characterized in that The system further comprises a model input module and a model conversion module, wherein the model conversion module is communicatively connected with the model input module and the method selection module; The model input module is used to input the transfer function or state equation of the model to be reduced; The model conversion module is used to generate the model to be reduced based on the transfer function or the state equation.

3. The system according to claim 2, characterized in that The model input module includes: A transfer function input unit, used for inputting coefficients of the numerator and the denominator in the transfer function; The state equation input unit is used to input the coefficient matrix of the state equation.

4. The system according to claim 1, characterized in that The system further comprises a parameter setting module, wherein the parameter setting module is in communication connection with the method selection module and the order reduction processing module; The parameter setting module is used to display parameter identification information corresponding to the first target model order reduction method, and obtain a first parameter setting instruction, where the first parameter setting instruction is used to indicate first parameter information of the first target model order reduction method; Correspondingly, the order reduction processing module is specifically used to perform order reduction processing on the model to be reduced based on the first target model order reduction method and the first parameter information to obtain the first reduced order model.

5. The system according to claim 1, characterized in that The system further comprises a graphics selection module and a graphics display module, wherein the graphics selection module is in communication connection with the order reduction processing module and the graphics display module; The graphic selection module is used to display identification information of multiple characteristic curves and obtain a graphic selection instruction, wherein the graphic selection instruction is used to indicate a target characteristic curve, and the target characteristic curve is any one or more of the multiple characteristic curves; The graphic display module is used to display the target characteristic curves corresponding to the to-be-reduced-order model and the first reduced-order model respectively.

6. The system according to claim 5, characterized in that The graphic display module comprises: A region selection unit, used for acquiring a region selection instruction, wherein the region selection instruction is used for indicating a target region, and the target region is a region in any of the target characteristic curves; A graphics updating unit is used to update the target characteristic curve based on the target area.

7. The system according to claim 4, characterized in that The system further comprises an error display module, wherein the error display module is communicatively connected with the order reduction processing module; The error display module is used to calculate and display the error information between the to-be-reduced-order model and the first reduced-order model.

8. The system according to claim 7, characterized in that The parameter setting module is also used for: Based on the error information, obtaining a second parameter setting instruction, where the second parameter setting instruction is used to indicate second parameter information of the first target model order reduction method; Correspondingly, the order reduction processing module is further used to perform order reduction processing on the model to be reduced based on the first target model order reduction method and the second parameter information to obtain a second reduced order model.

9. The system according to claim 7, characterized in that The method selection module is also used for: Based on the error information, a second method selection instruction is obtained, where the second method selection instruction is used to indicate a second target model order reduction method, where the second target model order reduction method is any one of the multiple model order reduction methods except the first target model order reduction method; Correspondingly, the order reduction processing module is further used to perform order reduction processing on the model to be reduced based on the second target model order reduction method to obtain a third reduced order model.

10. A model reduction method, characterized in that: Applied to the model reduction system according to any one of claims 1 to 9, the method comprises: Import the model to be reduced; Based on the model to be reduced in order, a method selection instruction is obtained, where the method selection instruction is used to indicate a target model reduction method, where the target model reduction method is any one of a plurality of model reduction methods; The model to be reduced is subjected to order reduction processing based on the target model order reduction method to obtain a reduced order model.

Images