Synchronous motor digital twin model self-correction method based on response surface method
Through the two-wheel correction method of the response surface method, the internal coefficients of the digital twin model of the synchronous motor are optimized, and the problem of insufficient model accuracy in the existing technology is solved, and the high accuracy operation of the motor model is achieved.
Patent Information
- Application Number
- CN202411211440.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-12-13
AI Technical Summary
There is a lack of effective methods in the prior art to correct the internal coefficients of the digital twin model of the synchronous motor, resulting in insufficient accuracy in the model during operation.
Using the two-wheel correction method of the response surface method, the actual values of the stator current and excitation current are collected within the shaft power and frequency range, and the motor iron consumption, stray loss, mechanical loss and leakage correction coefficient are gradually optimized to form the optimal the correction coefficient vector ensures the accuracy of the model.
The precise correction of key electrical quantity of the synchronous motor digital twin model is achieved, ensuring that the model maintains high accuracy during operation, and avoiding the internal coefficient correction problem that cannot be solved by the single-wheel response surface method.
Smart Images
Figure CN119135010A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of motor technology, and in particular to a self-correction method for a digital twin model of a synchronous motor based on a response surface method. Background Art
[0002] Response surface method is an optimization method that integrates experimental design and mathematical modeling. By testing representative local points, the functional relationship between factors and results in the global scope is regressed and fitted to obtain the optimal level value of each factor. Its basic idea is to approximately construct a polynomial with a clear expression form to express the implicit function. In essence, response surface method is a set of statistical methods used to find the best response value after considering the uncertainty of the input variable value. At present, the correction of the digital twin model of synchronous motor is usually carried out by the result correction method, that is, the method of correcting the model results according to the measured values obtained by the experiment, which can be achieved by machine learning methods such as neural networks. Compared with the result correction, the internal coefficient correction of the digital twin model of synchronous motor has three advantages: first, the initial model of the digital twin of the main motor has a high accuracy and can be applied after layout; second, it requires fewer levels for iterative calibration and training, consumes less resources, and has low sensitivity to the algorithm; third, the mechanism model "bottoms out" under extreme data, and can still maintain rationality and accuracy.
[0003] At present, there are few studies on the application of response surface methodology to the internal coefficients of the digital twin model of synchronous motors. As a motor equipment manufacturer, we have the advantage of being familiar with the internal operating mechanism of synchronous motors. Based on our own advantages, how can we apply the response surface methodology to correct the internal coefficients of the digital twin model of synchronous motors to ensure the accuracy of the digital twin model of synchronous motors?
[0004] Therefore, a self-correction method for a digital twin model of a synchronous motor based on the response surface method is provided. Summary of the invention
[0005] The purpose of the present invention is to overcome the existing defects and provide a self-correction method for a digital twin model of a synchronous motor based on the response surface method, thereby ensuring that the digital twin model of the synchronous motor maintains a high accuracy during operation.
[0006] The technical solution to achieve the above purpose is:
[0007] A self-correction method for a digital twin model of a synchronous motor based on a response surface method, comprising:
[0008] Step S1, within the normal operating range of the shaft power P and the rotational frequency n of the motor, a certain number of shaft powers P and rotational frequencies n are selected at a certain interval as coordinate points to form M operating conditions (P, n);
[0009] Step S2: for each set of working conditions, the shaft power P and the speed value n are directly obtained by collecting the stator current I under each working condition through the previous test. 1 and the excitation current I f1 , as the actual value;
[0010] Step S3, summarize the working conditions and output results under each working condition, [working condition, actual value] [P, n, I 1 ,I f1 ] an array of M rows and 4 columns;
[0011] Step S4, through a round of correction, obtain the four correction coefficients that are optimal (x i1 、x i2 、x i3 、x i4 )Correction coefficient vector;
[0012] Step S5, for each group (P, n) of working conditions, obtain the corresponding optimal (x 1 、x 2 、x 3 、x 4 ) correction coefficient vector, forming [x 1 , x 2 , x 3 , x 4 ] an array of M rows and 4 columns;
[0013] Step S6, through the second round of correction to form [x 1 , x 2 , x 3 , x 4 ] uses the response surface method to select the optimal results of the four correction coefficients whose calculated values are close to the actual values under any set of (P, n) working conditions required, thereby completing the self-correction of the internal coefficients of the motor digital twin model.
[0014] Preferably, in step S1, when Box-Behnken (response surface method) design is used, M is less than 29.
[0015] Preferably, in step S4, the four correction coefficients include: motor iron loss correction coefficient x 1 , Motor stray loss correction factor x 2 , Motor mechanical loss correction factor x 3 and motor leakage correction factor x 4 ; Among them, the initial value of the correction coefficient is known and obtained through theoretical formulas and engineering experience.
[0016] Preferably, in step S4, the first round of correction includes:
[0017] Step S41, using the four correction coefficients in the motor digital twin model to calculate the motor iron loss correction coefficient x 1 , Motor stray loss correction factor x 2 , Motor mechanical loss correction factor x 3 and motor leakage correction factor x 4 As input;
[0018] Step S42, taking the actual values of the stator current and the excitation current I under a set of working conditions 1 ,I f1 As an optimization goal;
[0019] Step S43, design different motor iron loss correction coefficients x 1 , Motor stray loss correction factor x 2 , Motor mechanical loss correction factor x 3 and motor leakage correction factor x 4 The output results of the motor digital twin model under is used as the test group;
[0020] Step S44, using the response surface method to calculate the optimal (x i1 、x i2 、x i3 、x i4 )Correction coefficient vector.
[0021] Preferably, in step S43, the test group contains N groups of data, and when the Box-Behnken design is used, N is less than 29.
[0022] Preferably, in step S6, the second round of correction includes:
[0023] Step S61: 1 , x 2 , x 3 , x 4 ] array as the experimental group input for the second round of correction;
[0024] Step S62, using the corresponding M groups of (P, n) working conditions as output, applying the response surface method to select any required group of (P, n) working conditions as the optimization target value;
[0025] Step S63, calculate and obtain the optimal results of the four correction coefficients that make the calculated values under the group (P, n) working conditions close to the actual values, thereby completing the self-correction of the internal coefficients of the motor digital twin model.
[0026] The beneficial effects of the present invention are:
[0027] The present invention adopts two rounds of response surface method, wherein the first round of response surface method takes the correction coefficient as input, the actual value under a specific set of working conditions is the response value, the optimal correction coefficient mapping relationship of the specific working condition is obtained, and multiple groups of working conditions are traversed. In the second round of response surface method, the specific working condition group is converted from the precondition in the first round to the response value. The two rounds of response surface method are closely linked, and the internal coefficient correction problem of the motor digital twin model that cannot be solved by the single round response surface method is solved;
[0028] After obtaining each set of working conditions and the corresponding optimal correction coefficient array, if the correction coefficient closest to the actual value corresponding to any set of (P, n) working conditions is to be known, there is no need to repeat the previous operation, and only the required (P, n) working condition is used as the target response value and the response surface method is used for calculation;
[0029] The present invention realizes the precise correction of key electrical twin quantities of the digital twin model of the synchronous motor, ensuring that the digital twin model of the synchronous motor maintains high accuracy during operation. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] Figure 1 It is a flow chart of a self-correction method of a digital twin model of a synchronous motor based on a response surface method according to the present invention;
[0031] Figure 2 It is a specific flow chart of the first round of revision in the present invention;
[0032] Figure 3 It is a specific flow chart of the second round of correction in the present invention;
[0033] Figure 4 is a schematic diagram of error comparison before and after stator current correction in the present invention;
[0034] Figure 5 It is a schematic diagram of error comparison before and after the excitation current is corrected in the present invention. DETAILED DESCRIPTION
[0035] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. In the description of the present invention, it should be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside" and the like indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings, which are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as a limitation on the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance.
[0036] The present invention will be further described below in conjunction with the accompanying drawings.
[0037] like Figure 1 As shown, a self-correction method of a synchronous motor digital twin model based on a response surface method comprises:
[0038] Step S1, within the normal operating range of the motor's shaft power P and rotational frequency n, a certain number of shaft powers P and rotational frequencies n are selected at certain intervals as coordinate points to form M operating conditions (P, n).
[0039] In an embodiment, when the Box-Behnken design is used, M is less than 29.
[0040] Step S2: for each set of working conditions, the shaft power P and the speed value n are directly obtained by collecting the stator current I under each working condition through the previous test. 1 and the excitation current I f1 , as the actual value.
[0041] Step S3, summarize the working conditions and output results under each working condition, [working condition, actual value] [P, n, I 1 ,I f1 ] is an array of M rows and 4 columns.
[0042] Step S4, through a round of correction, obtain the four correction coefficients that are optimal (x i1 、x i2 、x i3 、x i4 )Correction coefficient vector.
[0043] In the embodiment, the four correction factors include: motor iron loss correction factor x 1 , Motor stray loss correction factor x 2 , Motor mechanical loss correction factor x 3 and motor leakage correction factor x 4 ; Among them, the initial value of the correction coefficient is known and obtained through theoretical formulas and engineering experience.
[0044] like Figure 2 As shown, the first round of amendments included:
[0045] Step S41, using the four correction coefficients in the motor digital twin model to calculate the motor iron loss correction coefficient x 1 , Motor stray loss correction factor x 2 , Motor mechanical loss correction factor x 3 and motor leakage correction factor x 4 as input.
[0046] Step S42, taking the actual values of the stator current and the excitation current I under a set of working conditions 1 ,If1 as optimization goal.
[0047] Step S43, design different motor iron loss correction coefficients x 1 , Motor stray loss correction factor x 2 , Motor mechanical loss correction factor x 3 and motor leakage correction factor x 4 The output results of the motor digital twin model under are used as the test group.
[0048] In the embodiment, the test group contains N groups of data, and when the Box-Behnken design is used, N is less than 29.
[0049] Step S44, using the response surface method to calculate the optimal (x i1 、x i2 、x i3 、x i4 )Correction coefficient vector.
[0050] Step S5, for each group (P, n) of working conditions, obtain the corresponding optimal (x 1 、x 2 、x 3 、x 4 ) correction coefficient vector, forming [x 1 , x 2 , x 3 , x 4 ] is an array of M rows and 4 columns.
[0051] Step S6, through the second round of correction to form [x 1 , x 2 , x 3 , x 4 ] uses the response surface method to select the optimal results of the four correction coefficients whose calculated values are close to the actual values under any set of (P, n) working conditions required, thereby completing the self-correction of the internal coefficients of the motor digital twin model.
[0052] like Figure 3 As shown, the second round of amendments includes:
[0053] Step S61: 1 , x 2 , x 3 , x 4 ] array as the experimental group input for the second round of revision.
[0054] Step S62, using the corresponding M groups of (P, n) operating conditions as output, applying the response surface method, and selecting any required group of (P, n) operating conditions as the optimization target value.
[0055] Step S63, calculate and obtain the optimal results of the four correction coefficients that make the calculated values under the group (P, n) working conditions close to the actual values, thereby completing the self-correction of the internal coefficients of the motor digital twin model.
[0056] In another embodiment, in step S4 or step S6, the operating conditions and input-output of the response surface method may be in a form other than an array, and the correction process remains unchanged.
[0057] In another embodiment, the numbers M and N of the first and second round test groups are appropriate values for the response surface method Box-Behnken design, and different values of M and N do not change the overall correction process.
[0058] pass Figure 4 , Figure 5 It can be seen that after the internal coefficients of the motor digital twin model are self-corrected, the errors of the stator current and the excitation current are significantly reduced, which effectively ensures that the synchronous motor digital twin model maintains a high accuracy during operation.
[0059] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments may still be modified, or some or all of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A self-correction method for a digital twin model of a synchronous motor based on response surface methodology, characterized in that: include: Step S1, within the normal operating range of the shaft power P and the rotational frequency n of the motor, a certain number of shaft powers P and rotational frequencies n are selected at a certain interval as coordinate points to form M operating conditions (P, n); Step S2: for each set of working conditions, the shaft power P and the speed value n are directly obtained by collecting the stator current I1 and the excitation current I2 under each working condition through the previous test. f1 , as the actual value; Step S3, summarize the working conditions and output results under each working condition, [working condition, actual value] [P, n, I1, I f1 ] an array of M rows and 4 columns; Step S4, through a round of correction, obtain the four correction coefficients that are optimal (x i1 、x i2 、x i3 、x i4 )Correction coefficient vector; Step S5, for each group of (P, n) working conditions, obtain the corresponding optimal (x1, x2, x3, x4) correction coefficient vector through the step S4 to form an array of M rows and 4 columns of [x1, x2, x3, x4]; Step S6, through the second round of correction to form an M-row and 4-column array of [x1, x2, x3, x4], the response surface method is used to select the optimal results of the four correction coefficients whose calculated values are close to the actual values under any set of (P, n) working conditions required, thereby completing the self-correction of the internal coefficients of the motor digital twin model.
2. The self-correction method of a synchronous motor digital twin model based on response surface methodology according to claim 1 is characterized in that: In step S1, when the Box-Behnken design is used, M is less than 29.
3. The self-correction method of a synchronous motor digital twin model based on response surface methodology according to claim 1 is characterized in that: In step S4, the four correction coefficients include: motor iron loss correction coefficient x1, motor stray loss correction coefficient x2, motor mechanical loss correction coefficient x3 and motor magnetic leakage correction coefficient x4; wherein the initial value of the correction coefficient is known and obtained through theoretical formulas and engineering experience.
4. The self-correction method of a synchronous motor digital twin model based on response surface methodology according to claim 3 is characterized in that: In step S4, the first round of correction includes: Step S41, taking the four correction coefficients in the motor digital twin model, namely, the motor iron loss correction coefficient x1, the motor stray loss correction coefficient x2, the motor mechanical loss correction coefficient x3 and the motor magnetic leakage correction coefficient x4, as input; Step S42, taking the actual values of stator current and excitation current I1, I2 under a set of working conditions f1 As an optimization goal; Step S43, designing the output results of the motor digital twin model under different motor iron loss correction coefficients x1, motor stray loss correction coefficients x2, motor mechanical loss correction coefficients x3 and motor magnetic leakage correction coefficients x4 as a test group; Step S44, using the response surface method to calculate the optimal (x i1 、x i2 、x i3 、x i4 )Correction coefficient vector.
5. The self-correction method of a synchronous motor digital twin model based on response surface methodology according to claim 4 is characterized in that: In step S43, the test group contains N groups of data, and when the Box-Behnken design is used, N is less than 29.
6. The self-correction method of a synchronous motor digital twin model based on response surface methodology according to claim 1 is characterized in that: In step S6, the second round of correction includes: Step S61, inputting the array [x1, x2, x3, x4] as the test group for the second round of correction; Step S62, using the corresponding M groups of (P, n) working conditions as output, applying the response surface method to select any required group of (P, n) working conditions as the optimization target value; Step S63, calculate and obtain the optimal results of the four correction coefficients that make the calculated values under the group (P, n) working conditions close to the actual values, thereby completing the self-correction of the internal coefficients of the motor digital twin model.