Low-loss stacking closed-loop control method for same-series multi-specification lamination core

Abstract

The application discloses a low-loss stacking closed-loop control method for same-series multi-specification lamination core, and belongs to the technical field of transformer manufacturing. In view of the problem of poor stacking no-load loss consistency of a same-step joint design family across multiple capacity grades, a commercially available typical specification stacking first piece core is selected and the no-load loss is actually measured, and the optimal joint parameter combination is inversely deduced through a loss mapping model; then, the optimal parameters are deduced to the remaining capacity grades according to the geometric similarity scaling law of the same series; the closed loop of "measurement-inversion-scaling-issuance-retest" is formed through the closed-loop issuance and retest correction of the stepping distance controller of the stacking production line transverse shearing line, the lamination machine grade program and the clamping torque executor. The application only needs to measure a single typical specification first piece, and can be popularized to the whole series, so that the no-load losses of the grades are averagely reduced compared with the standard baseline, and the dispersion is narrowed to 5%.

Description

Technical Field

[0001] This invention relates to a method for manufacturing transformer cores, and more particularly to a method for inverting stepped joint parameters and implementing closed-loop control for cross-specification stacking of multi-specification laminated cores of the same series. Background Technology

[0002] Stepped joint lamination can reduce no-load losses and noise in distribution transformers, and is a common core structure used in the same series of oil-immersed distribution transformers. The same series of cores covers multiple standard capacities (capacity spans can reach tens of times), all belonging to the stepped joint design family; when the capacity is increased, the number of joint stages... Step distance Clamping torque The optimal combination varies significantly with capacitance, while the batch magnetic properties of silicon steel sheets... Fluctuations further amplify the dispersion of no-load losses.

[0003] Existing stacking processes mostly use fixed methods. The static process cannot guarantee consistent low loss across specifications. Some existing technologies (such as CN113552852A "Transformer No-Load Performance Control Method") only rely on the normal distribution statistics of the no-load loss coefficient based on the design value for quality inspection, which belongs to statistical quality QC and does not involve the adaptive inversion of joint parameters with capacity. Other methodological literature (such as CN116884759A "Iron Core Stacking Process Scheme Generation System and Method") selects existing process schemes for single-specification iron cores based on historical process database screening and theoretical calculation, without involving a complete method of first-piece actual measurement inversion based on typical specifications, and then automatic derivation and closed-loop distribution to the whole series through geometric similarity scaling law. Although there are a high density of patents on stepped joint structures / step number directions (such as CN101640107A / B, CN105280356A, CN116130217A, etc.), they all focus on the static structure / process of "how to do the joint geometry" and do not propose a closed-loop control method of "driving with inversion based on benchmark specifications and scaling down across specifications". Summary of the Invention

[0004] This invention aims to solve the technical problem that when the same series of iron cores in the same stepped joint design family spans multiple capacity levels (capacity span can be tens of times), the optimal combination of joint parameters is amplified and dispersed due to capacity drift and batch magnetic characteristic fluctuations of silicon steel sheets, which makes it impossible for traditional static stacking processes to guarantee low loss consistency across specifications. In particular, it optimizes representative working conditions based on commercially available typical specifications and simultaneously promotes them to other grades in the same series.

[0005] The technical solution adopted by this invention to solve its technical problem is: a low-loss stacking closed-loop control method for multi-specification laminated iron cores of the same series, wherein the same series refers to a family of iron core designs using the same grade of oriented silicon steel sheets, the same design magnetic flux density, and the same stepped joint form; characterized by including the following steps: S1, reference selection step: selecting a commercially available typical specification as the reference specification from the standard capacity sequence of the same series; S2, first piece test step: stacking the first iron core of the reference specification according to the predetermined stepped joint process, and placing it on an unloaded test bench to measure its no-load loss. With clamping torque S3, Parameter Inversion Steps: Mapping the model using no-load loss Inversion to find the optimal combination of joint parameters that minimizes no-load loss ;in For seam grade, The step distance, For clamping torque, As a benchmark for material specific loss, Due to batch-to-batch fluctuations in the magnetic properties of silicon steel sheets, , , These are the loss correction factors for the seam level, step distance, and clamping torque, respectively; S4, scaling derivation steps: based on the geometric similarity scaling law of the same series. The optimal combination of joint parameters of the reference specification. By applying this to other capacity specifications in the same series, the optimal joint parameter set for each specification is obtained. S5, Closed-loop delivery step: Write the optimal seam parameter set of each specification into the stacking station process instruction, and execute it in a closed loop on the actual production line through the cross-cutting line step distance controller, stacking machine stage program and clamping torque actuator; S6, Retesting and archiving step: Retest the no-load loss of the first piece of each specification, compare it with the model prediction value, and trigger S3 local re-optimization when the deviation exceeds the threshold, forming a "measurement-inversion-scaling-delivery-retesting" closed loop, and archive the entire process data.

[0006] The loss correction factor described in S3 of the above method takes the following calculable form, and parameter inversion is performed accordingly: Discrete table lookup: Time to take , Time to take , Time to take ; ,in Optimal step distance ; ,in Optimal clamping torque ; Fixed joint order during inversion In step distance by Step length, clamping torque by Perform a grid search for the step size, and take Minimum combination .

[0007] The above method, specifically the geometric similarity scaling law described in S4. Derived from geometric similarity constraints of the same design family: target specification capacity Corresponding cross-sectional area of ​​iron core column Stack width ,in , , These are the capacity, cross-sectional area, and lamination width of the selected baseline specification; based on this, the optimal parameters for each specification are derived: step distance. according to Scaling proportionally and rounding to the nearest integer Clamping torque according to Proportional scaling and rounding .

[0008] The above method, seam level in S4 Determine by referring to the table based on capacity: and The specification is grade 5. Specifications of 7 and above are grade 7.

[0009] The above method, in S4, for non- Voltage rating specifications incorporate an insulation margin correction factor. The clamping torque obtained by scaling Adjust upwards: Specifications , Specifications The insulation margin correction factor The following calibration method is used to obtain the calibration results: (for the same cross-sectional area) Next, press respectively Reference clamping torque Clamping torque with special voltage target Stacked comparison samples and retested no-load loss With the rigidity of the iron core, take Not inferior to The minimum torque required for the core rigidity to meet the assembly requirements of this insulation class under operating conditions. ,but .

[0010] The clamping torque measured on the first piece according to the reference specifications described in S2, as described above, is obtained through the above method. Pick This corresponds to the M12 clamp.

[0011] In the above method, the clamping torque actuator mentioned in S5 is an electric wrench or a hydraulic cylinder pressure controller. During stacking, the stack thickness, verticality, and torque values ​​are collected in real time and fed back to the controller.

[0012] In the above method, the retest deviation threshold mentioned in S6 is: When the threshold is exceeded, a second inversion is triggered to form a second closed loop.

[0013] The above method refers to a series that covers multiple specifications of the standard capacity sequence, and the inversion result of the benchmark specification is extended to the other capacity specifications by a scaling law.

[0014] The above method is applied to the manufacturing of laminated cores of oil-immersed distribution transformers, wherein the silicon steel sheets are cold-rolled oriented silicon steel sheets, and the stepped joints are discrete symmetrical odd-numbered joints of level 3, level 5, or level 7.

[0015] The beneficial effects of this invention are:

[0016] (1) Significantly improved consistency of no-load loss across specifications: After optimization by this method, the no-load loss of each specification in the same series of multiple capacity ranges decreased on average compared with the corresponding standard baseline. The dispersion is due to the conventional process Narrow to Within.

[0017] (2) Enhance the universality of the representativeness of the benchmark specifications: The typical specifications commonly used in the market are used as the benchmark for inversion, and the scaling law is applied to other grades in the same series, so that the optimization effect is not limited to individual specifications; and only the first piece test of a single typical specification is required to promote the entire series, avoiding the cost and sample size limitations of large sample tests for each specification.

[0018] (3) Digital closed loop of manufacturing process, implemented in real production line execution link: From the first piece measurement to parameter inversion and then to the closed loop of cross shearing / stacking / torque actuator, the whole process is implemented on the physical production line through sensors and actuators, ensuring the consistent reproduction of stacking process parameters across specifications. Attached Figure Description

[0019] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0020] Figure 1 This is an overall flowchart of the control method of the present invention (the dashed box represents the closed-loop feedback path).

[0021] Figure 2 This is a diagram illustrating the loss mapping model and inversion optimization of the present invention.

[0022] Figure 3This is a block diagram of the geometric similarity scaling law mapping of the same series of inventions;

[0023] Figure 4 This is a block diagram of the closed-loop execution link of the stacked assembly production line of the present invention. Detailed Implementation

[0024] The following description uses the laminated iron core of the S11-M series oil-immersed distribution transformer as an example to further illustrate the present invention. It should be noted that the S11-M series is only one preferred embodiment of the present invention (same design family, 13 standard capacity). The method of this invention is also applicable to other multi-specification laminated iron cores of the same series with the same grade of oriented silicon steel, the same design magnetic flux density, and the same stepped joint form. The scope of protection is defined by the claims.

[0025] Example 1 (Standard specification, S11-M-630 / 10 - a commonly used typical specification in the market)

[0026] Select 23QG090 high-permeability magnetic orientation cold-rolled silicon steel sheets (thickness) Unit iron loss As the core material, the laminates are first cut along the cross-cutting line using a seven-step joint, with the initial step distance set to [value missing]. The length of each lamination is cut progressively shorter according to the cross-sectional diagram of the iron core column. After cutting, the laminations are sent to the stacking station and processed according to... The core pillars and yoke laminations are stacked alternately layer by layer. After each layer is completed, the stacking thickness tolerance is checked using a stacking thickness gauge (controlled within a specified range). (Inner), and use a right-angle ruler to check the perpendicularity of the end face (deviation). ).

[0027] After stacking, assemble the upper / lower clamps and apply clamping torque using a calibrated torque wrench (or an electric wrench with torque feedback). (Corresponding to M12 clamping screw), the clamping sequence follows the principle of "middle first, then both ends, symmetrical and cross" to evenly distribute stress. Then, the iron core is erected, the coil (simulator or actual coil) is installed, and it is placed on the no-load test bench to apply the rated voltage. (Phase voltage), read the no-load loss after stabilizing for 3 minutes. and no-load current .

[0028] Actual measurement Input loss mapping model ,fixed ,by and Inverse calculations are performed to solve for the decision variables. The material specific loss in the model... Take the nominal value of 23QG090 Batch fluctuations Take the deviation of the sampling inspection value of the incoming materials in the current batch ( The optimal step distance is obtained through iterative solution. (Increasing the value by 1mm from the initial value makes the magnetic flux transition at the joint smoother), optimal clamping torque (The consistency with the initial value indicates that the torque is within a reasonable range.) Substitute the values ​​into the model to verify the expected values. It decreased compared to the standard baseline of 810W. .

[0029] Then, the scaling law of geometric similarity was applied. This optimal parameter set (7, 11, 80) is then extended to the remaining capacity levels (see Examples 2 and 3 for details). For batch stacking of the baseline specifications, the stacking station controller automatically issues the parameters. To the cross-cutting line stepper controller, To torque actuators, to achieve consistent stacking of products of this specification.

[0030] Example 2 (Lower bound of small capacity, S11-M-50 / 10 - Verifying the effectiveness of scaling at extremely small capacities)

[0031] S11-M-50 / 10 is the smallest capacity in this series. The cross-sectional area of ​​the iron core column is approximately , width Due to the scaling law Estimation: Step distance Round to the nearest integer (Boundary correction: The step distance should not be less than 6mm, otherwise the mechanical strength at the joint will be insufficient, so adjust upwards); Clamping torque Round to the nearest integer (Boundary correction: Although the cross-section of the small-capacity iron core is small, the specification of the clamping screw is reduced to M10, and the minimum reliable torque is approximately...) ); the seam grade is reduced to (The total thickness of the small-capacity iron core laminations is limited, and the five-level stepped joints are sufficient to disperse the joint magnetic resistance.)

[0032] Stack the first piece of S11-M-50 / 10 according to parameters (5,8,65) and retest. It decreased from the standard baseline of 170W. The results show that the scaling law can still effectively predict the optimal parameters even after boundary correction at the small capacity end, verifying the cross-specification applicability of the method.

[0033] Example 3 (High Capacity + Special Voltage, S11-M-3150 / 35 - Verifying the Effectiveness of Scaling in High Capacity and High Voltage Conditions)

[0034] S11-M-3150 / 35 is the largest capacity in this series ( It is a 35kV high voltage level, and the cross-sectional area of ​​the iron core column is... Due to the scaling law Estimation: Step distance Round to the nearest integer (Boundary correction: If the step distance exceeds 14mm, the joint will be too long, which will increase the eddy current path, so it is capped at 14mm); Clamping torque Considering the requirement for greater clamping force in 35kV insulation structures, a voltage correction factor is introduced. , obtain the target clamping torque This specification adopts an M20 twin-screw parallel clamping structure, with each bolt distributing approximately [amount missing]. Rounding down (Total of double bolts) , and the target Consistent). Silicon steel sheets were changed to 27QG100 ( (Suitable for large capacity) Material loss ratio .

[0035] Stack the first piece of S11-M-3150 / 35 according to parameters (7,14,125) and retest. A decrease from the standard baseline of 2700W The results show that the scaling law remains effective even when combined with a voltage correction factor at the large capacity end, verifying the method's adaptability to non-standard voltage specifications.

[0036] A table summarizing the optimal joint parameters and expected no-load losses for each specification

[0037] 1 S11-M-50 / 10 23QG090 170 (5,8,65) 158 7.1% Small capacity, Take level 5 2 S11-M-200 / 10 23QG090 330 (5,9,70) 306 7.3% 3 S11-M-315 / 10 23QG090 480 (7,10,75) 445 7.3% 4 S11-M-400 / 10 23QG090 570 (7,10,78) 528 7.4% 5 S11-M-500 / 10 23QG090 680 (7,10,80) 629 7.5% 6 S11-M-630 / 10 23QG090 810 (7,11,80) 748 7.7% ★Standard Specifications 7 S11-M-800 / 10 23QG090 980 (7,11,85) 905 7.7% 8 S11-M-1000 / 10 23QG090 1150 (7,12,88) 1060 7.8% 9 S11-M-1000 / 6 23QG090 1150 (7,12,91) 1055 8.2% 6kV, 10 S11-M-1250 / 10 23QG090 1380 (7,12,92) 1270 8.0% 11 S11-M-1600 / 10 23QG090 1650 (7,13,95) 1515 8.2% 12 S11-M-2000 / 10 23QG090 1980 (7,13,98) 1815 8.3% Maximum capacity (10kV) 13 S11-M-3150 / 35 27QG100 2700 (7,14,125) 2470 8.5% 35kV

[0038] Note: step distance Unit: mm; clamping torque unit ;"expected "This is the expected value after model inversion, which is approximately lower than the standard baseline." All parameters were obtained through inversion based on the first piece of the benchmark specification and then scaled using the geometric similarity law. The results were derived rather than obtained from individual tests of each specification, so the effectiveness verification of the entire series does not depend on the sample size of a single specification.

Claims

1. A low-loss stacked closed-loop control method for multi-specification laminated iron cores of the same series, wherein the same series refers to a family of iron core designs using the same grade of oriented silicon steel sheets, the same design magnetic flux density, and the same stepped joint form; characterized in that, Includes the following steps: S1, Benchmark Selection Steps: Select a commonly used typical specification from the standard capacity series of the same series as the benchmark specification; S2, First-piece test procedure: Assemble the first core of the aforementioned standard specifications according to the established stepped joint process, and place it on the no-load test bench to measure its no-load loss. With clamping torque ; S3, Parameter Inversion Steps: Mapping the No-Load Loss Model Inversion to find the optimal combination of joint parameters that minimizes no-load loss ;in For seam grade, The step distance, For clamping torque, As a benchmark for material specific loss, Due to batch-to-batch fluctuations in the magnetic properties of silicon steel sheets, , , These are the loss correction factors for the joint level, step distance, and clamping torque, respectively. S4, Scaling Derivation Steps: Based on the scaling law of geometric similarity in the same series... The optimal combination of joint parameters of the reference specification. By applying this to other capacity specifications in the same series, the optimal joint parameter set for each specification is obtained. ; S5, Closed-loop delivery step: Write the optimal seam parameter group of each specification into the stacking station process instruction, and execute it in a closed loop on the actual production line through the cross-cutting line step distance controller, stacking machine stage program and clamping torque actuator of the stacking production line. S6, Retesting and Archiving Steps: Retest the no-load loss of the first piece of each specification, compare it with the model prediction value, and trigger S3 local re-optimization when the deviation exceeds the threshold to form a closed loop of "measurement - inversion - scaling - distribution - retesting", and archive the data of the whole process.

2. The method according to claim 1, characterized in that, The loss correction factor described in S3 takes the following calculable form, and parameter inversion is performed based on it: Discrete table lookup: Time to take , Time to take , Time to take ; ,in Optimal step distance ; ,in Optimal clamping torque ; Fixed joint order during inversion In step distance by Step length, clamping torque by Perform a grid search for the step size, and take Minimum combination .

3. The method according to claim 1, characterized in that, Geometric similarity scaling law described in S4 Derived from geometric similarity constraints of the same design family: target specification capacity Corresponding cross-sectional area of ​​iron core column Stack width ,in , , These are the capacity, cross-sectional area, and lamination width of the selected baseline specification; based on this, the optimal parameters for each specification are derived: step distance. according to Scaling proportionally and rounding to the nearest integer Clamping torque according to Proportional scaling and rounding .

4. The method according to claim 3, characterized in that, S4 seam grade Determine by referring to the table based on capacity: and The specification is grade 5. Specifications of 7 and above are grade 7.

5. The method according to claim 1, characterized in that, S4 for non Voltage rating specifications incorporate an insulation margin correction factor. The clamping torque obtained by scaling Adjust upwards: Specifications , Specifications The insulation margin correction factor The following calibration method is used to obtain the calibration results: (for the same cross-sectional area) Next, press respectively Reference clamping torque Clamping torque with special voltage target Stacked comparison samples and retested no-load loss With the rigidity of the iron core, take Not inferior to The minimum torque required for the core rigidity to meet the assembly requirements of this insulation class under operating conditions. ,but .

6. The method according to claim 1, characterized in that, The clamping torque measured on the first piece according to the reference specifications in S2 Pick This corresponds to the M12 clamp.

7. The method according to claim 1, characterized in that, The clamping torque actuator mentioned in S5 is an electric wrench or a hydraulic cylinder pressure controller. During stacking, the stack thickness, perpendicularity and torque values ​​are collected in real time and fed back to the controller.

8. The method according to claim 1, characterized in that, The retest deviation threshold mentioned in S6 is When the threshold is exceeded, a second inversion is triggered to form a second closed loop.

9. The method according to claim 1, characterized in that, The same series covers multiple specifications of the standard capacity sequence, and the inversion result of the benchmark specification is extended to the other capacity specifications by scaling law.

10. The method according to claim 1, characterized in that, The method is applied to the manufacture of laminated cores of oil-immersed distribution transformers, where the silicon steel sheets are cold-rolled oriented silicon steel sheets, and the stepped joints are discrete symmetrical odd-numbered joints of level 3, level 5, or level 7.

Images