Magnetic-thermal constraint optimization design and operation method for high overload reactor of network-type energy storage converter

Abstract

This invention discloses a magnetic-thermal constraint optimization design and operation method for a high-overload reactor in a grid-type energy storage converter. Belonging to the field of magnetic components in power electronic converters, the method includes: constructing a magnetic-thermal coupling constraint model with an improved iGSE iron loss model and dynamic heat dissipation constraint equations; performing magnetic constraint structure evolution based on this model, employing distributed multi-air gaps and amorphous / magnetic powder core materials to suppress edge magnetic flux; and thermal constraint structure evolution, employing flat copper wire vertical winding and microchannel active heat dissipation; and outputting optimal geometric parameters through multiphysics iteration. During operation, the switching frequency is dynamically adjusted based on real-time current and temperature estimates to achieve loss redistribution.

Description

Technical Field

[0001] This invention belongs to the field of magnetic components for power electronic converters, and particularly relates to a magnetic-thermal constraint optimization design and operation method for a high overload reactor in a grid-type energy storage converter. Background Technology

[0002] Currently, grid-forming (GFM) energy storage converters have become core equipment in high-proportion renewable energy power systems due to their ability to actively support grid voltage and frequency. Unlike traditional grid-following converters, which only need to withstand 1.1 to 1.2 times the short-term overload, GFM converters must provide up to 3 times or even higher rated short-circuit current during grid short-circuit faults and maintain this current for several seconds to activate downstream protection devices. Existing filter reactor designs still rely on empirical magnetic-thermal parameters under rated operating conditions, employing a single centralized air gap, traditional silicon steel sheet materials, and natural air-cooling structures, making them unsuitable for the aforementioned extreme overload scenarios.

[0003] However, existing technologies still face the following technical challenges when dealing with 3x short-circuit overloads in grid-type converters: First, the large current surge causes the magnetic core to rapidly enter the deep saturation region, resulting in a sharp drop in inductance close to zero, which can lead to overcurrent-induced tube failure. Second, the superposition of high-frequency ripple and large-amplitude power frequency current significantly exacerbates the skin effect and proximity effect, causing the temperature of hot spots inside the windings and magnetic core to exceed the insulation material's tolerance limit (e.g., 180°C) within seconds, leading to insulation breakdown failure. Furthermore, there is a physical contradiction between "low-frequency saturation resistance" and "high-frequency low loss" in single magnetic materials, and traditional fixed-frequency control strategies cannot dynamically adjust the heat source under extreme high-temperature environments, further limiting the fault ride-through capability of grid-type converters. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a method for the magnetic-thermal constraint optimization design and operation of a high-overload reactor in a grid-type energy storage converter, comprising: Based on the rated parameters of the grid-type energy storage converter and the preset short-circuit overload conditions, a magnetic-thermal coupling constraint model is established, which includes an improved iGSE core loss model and dynamic heat dissipation constraint equations. Based on the magnetic-thermal coupling constraint model, the evolution of the magnetic constraint structure and the evolution of the thermal constraint structure are performed. The evolution of the magnetic constraint structure includes adopting a multi-air gap dispersion structure to suppress the edge magnetic flux effect, and the evolution of the thermal constraint structure includes adopting a conductor-optimized structure and an active heat dissipation structure. Based on the iterative results of the evolution of the magnetic confinement structure and the thermal confinement structure, the geometric parameters of the high overload reactor are output; Based on the real-time current value and real-time temperature estimate of the high overload reactor during operation, the switching frequency of the converter is dynamically adjusted to achieve loss redistribution under the short-circuit overload condition.

[0005] Optionally, establishing the magnetic-thermal coupling constraint model further includes: Based on the duty cycle sequence of the grid-type converter under extreme operating conditions, the time-domain magnetic flux density waveform is obtained by Fourier fitting mapping; Based on the time-domain magnetic flux density waveform and the characteristic constants of the core material, the core loss per unit volume is calculated using the improved iGSE model. The dynamic heat dissipation constraint equation is established based on the core loss, the copper loss corrected by the skin effect gain coefficient, and the high overload power penalty coefficient of the mesh type.

[0006] Optionally, performing the magnetically confined structure evolution further includes: The total number of distributed air gap segments is determined based on the vacuum permeability, number of coil turns, effective cross-sectional area of ​​the magnetic core, maximum short-circuit current, minimum allowable inductance, and magnetic core saturation magnetic induction intensity. The total air gap length is divided into multiple micro-air gaps, and the length of each micro-air gap is set separately to suppress the edge magnetic flux diffusion factor.

[0007] Optionally, performing the magnetically confined structure evolution further includes: Based on the calculation results of the magnetic core loss using the magnetic-thermal coupling constraint model, amorphous alloy or iron-silicon-aluminum magnetic powder cores with high saturation magnetic induction intensity are selected as the magnetic core material.

[0008] Optionally, dynamically adjusting the switching frequency of the converter further includes: The phase current of the high overload reactor is collected at a preset period, and the temperature of the reactor core area is estimated in real time based on the built-in thermal resistance-capacitance network model. When the phase current is greater than a preset overload threshold and the core area temperature is greater than a preset temperature threshold, a frequency reduction strategy is executed.

[0009] Optionally, implementing the frequency reduction strategy further includes: The frequency reduction is calculated based on the difference between the core area temperature and the temperature threshold using a preset proportionality coefficient. A new switching frequency is determined based on the rated switching frequency and the frequency reduction, and the new switching frequency is limited to a preset minimum frequency.

[0010] Optionally, dynamically adjusting the switching frequency of the converter also includes: When the phase current recovers to below the overload threshold or the core area temperature drops below the preset recovery threshold, the switching frequency is smoothly restored to the rated switching frequency through the hysteresis control loop.

[0011] Optionally, performing the magnetically confined structure evolution further includes: A heterogeneous hybrid magnetic circuit consisting of an inner column and an outer yoke is constructed, wherein the inner column is selected from iron-silicon-aluminum magnetic powder cores with soft saturation characteristics, and the outer yoke is selected from manganese-zinc ferrite for suppressing high-frequency eddy current losses. The volume ratio of the inner column to the outer yoke is determined based on a multi-objective optimization algorithm.

[0012] On the other hand, the present invention also provides an electronic device including a memory, a processor, and a computing program stored in the memory and executable on the processor, wherein the processor implements the method when executing the computing program.

[0013] On the other hand, the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method.

[0014] Compared with the prior art, the present invention has the following advantages and technical effects: This invention completely breaks through the design flaw of traditional filter reactors that "simply rely on sacrificing heat dissipation volume to obtain safety margin" when dealing with extremely high overload conditions, and successfully defines the "high overload survival rule" for the core magnetic components of grid-type converters.

[0015] At the underlying logic level, this invention creatively constructs a three-in-one collaborative optimization framework of "magnetic circuit-structure-heat dissipation": In terms of hardware, the spatial architecture of "distributed micro-air gap + hybrid high-saturation composite magnetic material + embedded micro-channel vertical winding flat wire" is reconstructed, which eliminates deep magneto-saturation and high-frequency eddy current thermal blind zone from the physical source. In the software dimension, by introducing a dynamic temperature-frequency coupling operation strategy based on the modified iGSE loss model, traditional passive magnetic components are given the intelligent characteristic of "active defense" for the first time, enabling them to achieve peak shaving and valley filling under extreme short-term heat flow impact.

[0016] In summary, this invention represents not only an improvement on a single component but also a significant upgrade to the underlying key supporting technology of grid-connected power storage systems (GFM-PCS). It achieves a perfect decoupling and unification of "300% extreme short-circuit overload tolerance" and "high system power density (15% volume reduction)," avoiding the fatal problem of converters being forced to shut down due to inductance loss or insulation breakdown during grid faults. This provides a more robust and efficient hardware-level safety defense for new power systems, possessing extremely disruptive engineering potential and broad industrialization prospects. Attached Figure Description

[0017] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic diagram of the method flow according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the active frequency conversion control logic of a grid-type converter based on a digital twin thermal model according to an embodiment of the present invention. Figure 3 The following is a diagram illustrating the implementation steps and dynamic control logic of an embodiment of the present invention. Detailed Implementation

[0018] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0020] Example 1 like Figure 1 As shown, this embodiment provides a magnetic-thermal constraint optimization design and operation method for a high-overload reactor in a grid-type energy storage converter, including: Step S1 (Boundary Injection): Input converter parameters, including: rated power Rated current Maximum short-circuit current (e.g., 3 times) ), switching frequency Extract the negative sequence current component and the high-frequency grid-connected resonant frequency point under asymmetrical fault conditions.

[0021] Step S2 (Loss Separation): Using rapid loss characterization techniques, the total current is decomposed into a "low-frequency high DC bias component" (mainly causing magnetic saturation) and a "high-frequency ripple component" (mainly causing core heating). Calculations are then performed separately. (Iron loss) and (Copper loss).

[0022] Step S3 (Evolution of Magnetic Confinement Structure): Anti-saturation design and material selection.

[0023] (1) Anti-saturation design: The traditional single-air-gap structure is abandoned in favor of a multi-air-gap distributed structure (Distributed AirGaps). The total air-gap length is... Divided into micro air gap This is to suppress the fringing flux effect and prevent overheating of the windings near the air gap.

[0024] (2) Material selection: High saturation magnetic induction intensity ( Amorphous alloy or iron-silicon-aluminum magnetic powder core.

[0025] Step S4 (Evolution of Thermally Constrained Structure): Conductor Optimization and Active Heat Dissipation.

[0026] (1) Conductor optimization: Flat wire edge-wise winding is adopted to maximize slot fill factor and reduce the impact of conductor optimization. .

[0027] (2) Active heat dissipation: Microchannel heat dissipation plates or insulating thermally conductive ceramic sheets are implanted between the winding layers.

[0028] Step S5 (Multiphysics Iteration): Substitute the above formulas and iterate. If Then increase Or adjust the cross-sectional area of ​​the magnetic core; if the inductance decreases... If this is increased, the number of air gaps will be increased. The final output will then be the optimal geometric parameters.

[0029] The core difference between this invention and traditional empirical design lies in the establishment of a high-precision mathematical model for loss and heat dissipation.

[0030] (1) An improved primary core loss (iGSE) model considering multiple operating conditions of the network; The output of a grid-type converter is not a standard sine wave, but a distorted waveform containing high-frequency PWM harmonics. The traditional Steinmetz equation (OSE) fails. This invention uses an improved iGSE model to calculate the iron loss per unit volume. : ; The variables and their physical meanings in the formula are explained below: (W / m³) represents the transient loss power density of the magnetic core per unit volume; (s) is the switching period, which corresponds to the reciprocal of the converter's carrier frequency; (T) represents the time-domain magnetic flux density waveform. This waveform is obtained by Fourier fitting mapping of the duty cycle sequence under extreme operating conditions using a mesh control algorithm (SVPWM or SPWM). (T) is the peak-to-peak flux density, which directly determines the area of ​​the hysteresis loop; (Dimensionless) represent the characteristic constants of the magnetic core material. For amorphous or nanocrystalline materials, they are typically... , ; (Dimensionless) is the correction coefficient for non-sinusoidal waveforms, and its calculation formula is: ; In the formula, is the Steinmetz coefficient under a sine wave.

[0031] (2) Dynamic heat dissipation constraint equations of magnetothermal coupling; To address the "short-term high heat flux" impact of short-circuit overload, the following thermal balance constraint inequality is established: ; In the formula, (W / (m²·°C)) is the comprehensive convective heat transfer coefficient. Under forced air cooling conditions, this invention corrects this value through fluid dynamics simulation. (m²) represents the effective contact area of ​​the reactor's heat dissipation structure (including the equivalent area of ​​the external heat dissipation fins). (°C) represents the maximum allowable surface temperature of the reactor (usually taken as the insulation class minus a 20°C margin) and the ambient temperature, respectively. (m³) represents the effective volume of the magnetic core; (Dimensionless) is the AC resistance gain coefficient caused by the skin effect and proximity effect. For high frequency and high current, this value can reach 1.5-3.0. () represents the DC resistance of the winding; (Dimensionless) is the high overload power penalty coefficient for grid-type structures. In this invention, this coefficient is set as a dynamic value that increases non-linearly with the overload multiple (for example, it is 1.0 under rated conditions, and 2.5 instead of 9 under 3 times overload, because the thermal inertia buffering effect of heat capacity is taken into account).

[0032] On the other hand, this embodiment also provides an electronic device, including a memory, a processor, and a computing program stored in the memory and executable on the processor, wherein the processor implements the method when executing the computing program.

[0033] On the other hand, this embodiment also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method.

[0034] Example 2 This embodiment provides a magnetic-thermal constraint optimization design and operation method for a high-overload reactor in a grid-type energy storage converter, including: This embodiment details the physical design process of a filter reactor for a 2MW grid-forming PCS (Grid-Forming PCS) used in a wind-solar-storage integrated base, utilizing the "magnetic-thermal constraint optimization design method" described in this invention.

[0035] (a) Scene boundaries and design inputs (corresponding to step S1); First and foremost, the boundary conditions of the design object need to be defined. This converter must possess extremely strong grid support capabilities, requiring it to provide reactive power support of three times its rated current for at least three seconds during a three-phase short-circuit fault in the grid. Specific input parameters are as follows: (1) Rated power ( ): 2.0MW; (2) Rated current ( ): 2000A (RMS); (3) Maximum short-circuit current ( Duration: 6000A (RMS) / 3.0s; (4) Switching frequency ( ): 3.0kHz; (5) Target inductance ( ): 300 H(±5%); (6) Allowable inductance sag ( ):exist Under operating conditions, (Right now H).

[0036] (ii) Evolution design of magnetically confined structure (corresponding to step S3); (1) Selection of core material and geometry; Traditional designs typically use 0.35mm cold-rolled grain-oriented silicon steel sheets, which, while exhibiting high saturation magnetic induction, suffer from significant losses and rapid attenuation of high-frequency permeability under high-frequency ripple. This embodiment, based on the "loss separation" results, selects an FeSiB amorphous alloy C-core. The saturation magnetic induction of this material... However, its magnetic permeability stability at high frequencies is better than that of silicon steel, and its iron loss ( It is only 1 / 5 the size of silicon steel.

[0037] (2) Distributed air gap anti-saturation design; To address the problems of severe fringing flux effect leading to localized overheating and premature magnetic circuit saturation due to high local magnetic reluctance in traditional single-air-gap designs under 6000A impact, this embodiment employs a multi-air-gap distributed structure. Based on the magnetic circuit optimization model proposed in this invention: ; In the formula, The permeability of free space, ; This refers to the number of coil turns. This represents the effective cross-sectional area of ​​the magnetic core. The maximum allowable length of a single air gap (usually limited to 2 mm to suppress edge effects).

[0038] The design results are explained below: The total air gap length... Divided into Segment, length of each air gap Total effective air gap length This structure disperses the intense edge magnetic flux, which was originally concentrated in a single air gap, into eight regions, thus increasing the air gap diffusion factor. The linearity of the reactor at 6000A is significantly improved by reducing the value from 1.35 in the traditional design to 1.08.

[0039] (III) Thermally constrained structural evolution design (corresponding to step S4); (1) Optimization of winding conductors; To address the skin effect and proximity effect at a switching frequency of 3000Hz, traditional round wires are abandoned in favor of a flat wire edge-wise winding process. A thickness of 4mm is selected. The specifications for 12mm (width) oxygen-free copper flat wire are optimized by using a vertical winding structure that allows the single-layer winding to directly contact the heat dissipation channel. Furthermore, compared to round wire, the flat wire has a lower AC resistivity. At 3kHz, the ripple loss decreased from 1.8 to 1.15, significantly reducing high-frequency ripple loss.

[0040] (2) Composite active cooling system; For the massive instantaneous Joule heat generated by a 3x overload for 3 seconds, natural air cooling alone cannot meet the insulation requirements. This embodiment constructs a dual "inner-outer" heat conduction channel: Thermal conduction channel 1: Internal filling. A high thermal conductivity potting compound is filled between the winding and the core, and between the winding turns. Eliminate the blind spot of internal thermal resistance.

[0041] Heat conduction channel 2: interlayer microchannel. A 3mm thick aluminum microchannel heat pipe is embedded between the two sets of coils, and the heat pipe is directly connected to the external liquid cooling radiator.

[0042] The thermal equilibrium equation is: ; In the formula, For conducting thermal resistance (through potting compound and microchannel plate). To reduce convective thermal resistance. This design utilizes extremely high thermal conductivity and microchannel design to minimize the total thermal resistance. It decreased by 60%.

[0043] (iv) Comparative analysis of implementation effects; Based on multiphysics simulation (Ansys Maxwell + Fluent) and physical prototype testing, the design scheme of Embodiment 2 of the present invention is compared with the traditional design (0.35mm silicon steel sheet + round wire + air cooling) under the same working conditions.

[0044] Table 1 As shown in Table 1, the magnetic flux density of the conventional design is approximately 4500A. That is, it reaches 1.8T and enters the deep saturation region, and the inductance drops sharply to 65. H, this will cause the grid-type converter to lose its filtering capability during fault ride-through, which in turn will cause the overcurrent protection to malfunction or even cause the unit to fail.

[0045] In contrast, Embodiment 2 of the present invention weakens the edge effect and delays the saturation inflection point on the magnetic circuit through a "distributed air gap"; and constructs a rapid heat conduction highway on the thermal circuit through "flat wire vertical winding + microchannel heat dissipation". Ultimately, it successfully withstood an extreme impact of 6000A / 3s while maintaining an inductance of 245 while reducing the volume by 15%. Above Class H, the maximum temperature is controlled at 115°C (far below the 180°C limit for Class H insulation), completely solving the industry problem of "overload equals saturation, overload equals overheating".

[0046] Example 3 This embodiment provides a magnetic-thermal constraint optimization design and operation method for a high-overload reactor in a grid-type energy storage converter, including: This embodiment is an advancement and supplement to Embodiment 2 (Physical Structure Optimization) of the present invention. Addressing the "control-thermal" coupling constraint problem of grid-type energy storage converters under extreme environmental conditions, a dynamic frequency conversion operation method without incurring any additional hardware costs is proposed, corresponding to the "dynamic adjustment of multi-physics operating domain" part in the core steps of the invention.

[0047] (I) Description of extreme operating scenarios and pain points; In environments with extremely high temperatures, such as deserts and Gobi, At a new energy base, the converter suffered a prolonged three-phase short-circuit fault caused by a deep grid dip. In this situation, the grid-based control algorithm requires the converter to continuously output power. (i.e., 3 times the rated current) This process takes several seconds to allow the grid voltage to recover. At this critical boundary, with the ambient temperature already reaching 50°C, the heat dissipation margin of the physical heat dissipation structure (passive + active heat transfer) in Example 1 is severely compressed. If the converter's rated switching frequency is maintained... With the core loss caused by the surge in alternating magnetic flux remaining unchanged, the reactor temperature will rapidly exceed the tolerance limit of Class H insulation in about 2.0 seconds. This forces the converter to shut down, failing to meet the stringent grid fault ride-through (FRT) standard's technical guideline requirement that "low-voltage / zero-voltage ride-through must maintain full-load support for at least 3 to 5 seconds."

[0048] (II) Core control mechanism and theoretical model; The innovation of this embodiment lies in its introduction of a "magnetic-thermal-electrical multidimensional loss real location" mechanism, transcending the traditional electrical domain. During high overload periods, the total loss of the reactor... Iron loss ( ) and copper loss ( Together they constitute the whole.

[0049] (1) Strong high-frequency correlation of core loss (iron loss); According to the improved Steinmetz equation (iGSE), the core loss density under high-frequency non-sinusoidal excitation can be equivalent to: ; In the formula, For core loss (W); These are inherent constants of the material; The converter switching frequency (Hz); The peak-to-peak value of the alternating magnetic flux density (T); The frequency index (for amorphous alloys, usually...) ); Magnetic flux density index (usually) ); For the effective volume of the magnetic core ( ).because Iron loss affects switching frequency The decline is extremely sensitive.

[0050] (2) Weak correlation between ripple current and copper loss; Reduce switching frequency This can cause high-frequency ripple in the inductor current. Increased, and high-frequency ripple in the inductor current The expression is: ; AC copper loss The expression is: ; In the formula, It is a high-frequency AC resistor; This is the effective value for high-frequency AC.

[0051] Although the increase in ripple current increases the AC copper loss of the conductor There was a slight increase, but the high-frequency AC resistance was greatly reduced due to the use of "flat wire vertical winding" in Example 2. Compared to the drastic drop in iron losses, the slight increase in copper losses is entirely manageable. Therefore, reducing... It can achieve the optimal minimization of total heat source power.

[0052] (III) Implementation steps and dynamic control logic; Figure 2 This diagram illustrates the active frequency conversion control logic of a grid-type converter based on a digital twin thermal model.

[0053] Figure 2 The first diagram in the image shows the current state (steps P201 & P202, condition A), illustrating... The current suddenly jumps to 2.5 pu. The blue dashed line marks the 2.0 pu threshold, at which point the system identifies a "severe short-circuit ride-through condition".

[0054] Figure 2 In the diagram, the second image represents the temperature state (steps P201 & P202, condition B). In the initial stage of the fault, the temperature rises rapidly. The red dashed line (100℃) is the trigger point for step P202. Notice that although the current was already overloaded, the frequency did not change immediately until the temperature line crossed the 100℃ red line, at which point the logic was activated (demonstrating the "AND" logic). The green dashed line (90℃) is the hysteresis recovery point for step P204.

[0055] Figure 2In the diagram, the third figure represents the frequency response (steps P203 & P204). In the falling segment (P203), the blue line begins to decrease after the temperature exceeds 100°C. Since the formula is... The higher the temperature, the lower the frequency, until it is clamped at 1.5kHz. This visually demonstrates the "loss reduction redistribution execution". The plateau segment is when the temperature is high during the fault duration, and the frequency is maintained at a low level to protect the equipment. The rising segment (P204) is when... After the fault is cleared (current restored), although the temperature may still be high, according to the logic (fault clearing OR temperature <90), the system begins to execute "low-speed hysteresis control", and the blue line smoothly ramps back up to 3000Hz, avoiding sudden jumps.

[0056] Therefore, the controller (DSP / FPGA) incorporates a digital twin thermal model of the reactor, which executes the following steps periodically: Step P1 (Real-time Thermal Status Sensing): The control system collects phase current at a period of 1ms. The temperature of the reactor core region is estimated in real time based on the thermal resistance-capacitance network (Cauer / Foster model). Precision control within .

[0057] Step P2 (Dynamic Threshold Determination): The system performs an AND operation. The condition is met only if: Condition A: Overload characteristic. (Determined to have entered a severe short-circuit ride-through state); Condition B: Thermal constraint top-reaching characteristics (The physical heat dissipation threshold is approaching).

[0058] Step P3 (Loss Reduction and Redistribution Execution): Once the above conditions are triggered, the underlying PWM modulator executes a smooth linear frequency reduction strategy to avoid severe current distortion caused by sudden jumps. The control law is as follows: ; In the formula, the proportionality coefficient Set as As temperature accumulates, the switching frequency decreases from... Smoothly pull down to the lowest limit frequency .

[0059] Step P4 (Self-recovery logic): When the power grid fault is cleared or Falling back to In the following cases, the low-speed hysteresis control loop smoothly restores the frequency to the rated frequency. In order to regain optimal grid-connected harmonic power quality.

[0060] Figure 3The main comparison focuses on the dynamic process of using a traditional fixed-frequency strategy (fixed at 3kHz) and the variable-frequency strategy of this invention (dynamically down-frequency to 1.5kHz) under "3 times extremely severe short-circuit overload" and "50℃ extremely high temperature environment". The core idea of ​​this invention is to take the physical mechanism of "loss redistribution" as the starting point. Addressing the predicament of grid-connected converters under full-load support during extreme fault ride-through (FRT) conditions, where reactors are highly susceptible to rapid "overheating and explosion" due to high-frequency alternating magnetic fields, this invention innovatively proposes an electro-thermal coordinated control strategy of "temperature feedback and active frequency conversion." Specifically, when the system detects severe overload and the core temperature exceeds the safety threshold, it immediately breaks away from conventional rigid fixed-frequency control and actively and smoothly reduces the switching frequency. This cleverly sacrifices a slight increase in low-frequency copper losses to achieve a sharp drop in exponentially increasing high-frequency iron losses, thereby significantly reducing the total heat generation and forcibly slowing the temperature rise rate. Ultimately, this successfully "physically extends the life" of the converter, ensuring it can safely withstand the long and harsh grid fault ride-through period before approaching the insulation limit.

[0061] Figure 3 The first image in the diagram represents the sensing layer (condition triggering), corresponding to step P201 and the condition I_out>2.0I_rated in the patent. It demonstrates... In the event of a sudden short-circuit fault, the underlying control requires the converter output to surge instantaneously from 1.0 PU and maintain a peak output of 3.0 PU. The process of crossing a harsh power grid.

[0062] Figure 3 The second diagram in the image represents the execution layer (controller actions), corresponding to the core formula in the patent: The red line represents rigid, inflexible control (traditional); the blue line represents the dynamic intervention of this invention. When the temperature exceeds 100°C, the algorithm immediately takes effect... The slope smoothly applied the brakes to the carrier frequency, pulling it down to 1.5kHz.

[0063] Figure 3 The third figure in the diagram represents the physical layer (loss transfer mechanism), visually demonstrating the physical phenomenon of "loss redistribution" in the example table. The magenta (iron loss) decreases exponentially, releasing a huge amount of heat dissipation space; although the orange (copper loss) increases slightly due to the increased ripple, overall, the total heat generation rate (thick blue line) is reduced by about 37%.

[0064] Figure 3 The fourth image represents the target layer (demonstrating the rescue effect), which is the physical limit pain point that the patent ultimately aims to solve. The traditional route (red dotted line) reaches its limit at 2.6 seconds (just after the support passes). He was killed in the crash. On the red line (marked with a dozen red five-pointed stars: machine crash and shutdown).

[0065] The patented circuitry of this invention (thick blue line) shows a significantly slower rate of temperature increase after the frequency reduction "cooling down" intervention. After enduring the full 5 seconds, the temperature only reaches a maximum of [missing value]. Around 10:00, the 5-second crossing task specified in the grid connection guidelines was safely completed. At this point, the fault was cleared and the temperature began to drop.

[0066] (iv) Implementation effect verification and data comparison; At ambient temperature Current In the boundary test (3 times the nominal value), the thermodynamic trajectory was fundamentally changed after the active frequency conversion control of the present invention was introduced.

[0067] Table 2 shows a comparison of parameters between fixed-frequency operation and active frequency conversion operation under specific extreme conditions.

[0068] Table 2 (v) Conclusion; This embodiment demonstrates that, relying on the "active operating state machine intervention" proposed in the patent, when physical heat dissipation approaches its limit, the excitation frequency of the heat source is changed by digital control, successfully finding a transfer trough (SweetSpot) between copper and iron losses. This enables the grid-type energy storage converter to extend the extreme operating condition tolerance time from 2 seconds to 5 seconds without increasing any component cost or size under extreme conditions of high ambient temperature and deep overload, perfectly meeting the stringent requirements of the new generation of grid-type large-scale base projects for extreme transient stability.

[0069] Example 4 This embodiment provides a magnetic-thermal constraint optimization design and operation method for a high-overload reactor in a grid-type energy storage converter, including: This embodiment details the application of the method of the present invention in the harsh combined working condition of "requiring both high overload resistance to saturation and facing high-frequency broadband harmonics", corresponding to the "magnetic-thermal structure collaborative optimization based on multi-objective algorithm" in the core steps of the present invention.

[0070] (a) Scenarios and technical challenges; With an extremely low short-circuit ratio ( In weak grid or even off-grid scenarios, when a grid-connected converter operates as a virtual voltage source, it is highly susceptible to coupling with grid impedance. The high-frequency resonance. These resonant currents are superimposed on the fundamental switching frequency (e.g., the high-frequency resonance). Above the equivalent ripple, the current flows through the filter reactor. If conventional high-saturation magnetic flux density silicon steel is used, its significant high-frequency eddy current effect will cause extreme local overheating (high-frequency iron loss increases exponentially); if low-loss ferrite is used, its low saturation magnetic flux density (equivalent ripple) will prevent the flow of iron loss from increasing. When subjected to a short-circuit current three times the rated current, it will instantly reach deep saturation and lose its supporting capacity. Single magnetic materials face the physical paradox of "high-frequency overheating" and "low-frequency saturation".

[0071] (II) The design mechanism of heterogeneous composite magnetic circuit topology; To address the aforementioned pain points, this embodiment innovatively proposes a heterogeneous hybrid magnetic core topology architecture consisting of a "soft saturated magnetic powder core inner pillar + extremely low loss ferrite outer yoke".

[0072] (1) Inner Core Design: Select iron-silicon-aluminum alloy magnetic powder core (Sendust Core).

[0073] The inner column is directly surrounded by the winding and bears the maximum magnetomotive force. The iron-silicon-aluminum material particles have a natural insulating layer, forming a microscopic "distributed air gap," exhibiting superior "soft saturation" characteristics. Its effective permeability... With magnetic field strength The dynamic evolution model can be expressed as: ; In the formula, The initial permeability; The half-decay offset field strength of the material ( ); The characteristic softening coefficient of the material (usually 100%) This feature ensures that the inductance of the reactor decreases smoothly and slowly under a 3x overload (6000A) current surge, avoiding the "cliff-like" sudden change that occurs with traditional silicon steel and ensuring the robustness of the control system.

[0074] (2) Outer Yoke Design: Manganese-zinc high-frequency soft magnetic ferrite (MnZnFerrite) is selected.

[0075] The outer yoke serves as a return path for the closed magnetic flux. During alternating high-frequency ripple flux, eddy current losses are extremely severe. Manganese-zinc ferrite possesses extremely high resistivity ( It can completely block the flow path of high-frequency eddy currents and is specifically responsible for efficiently "absorbing and conducting" high-frequency oscillating magnetic flux.

[0076] (III) Multi-physics, multi-objective optimization based on NSGA-II algorithm; To determine the optimal volume ratio of Sendust and Ferrite in the entire set of magnets to achieve the global optimum of magnetic-thermal performance, this invention employs a non-dominated sorting genetic algorithm (NSGA-II) with an elite retention strategy for optimization.

[0077] (1) Definition of decision variables; Define heterogeneous volume ratio coefficient (in ), combined with the core window width Core diameter As a three-dimensional decision variable space .

[0078] (2) Establish a multi-objective optimization function; The system aims to simultaneously minimize the "overall volume of magnetic components" and "severe high-frequency heat generation (in order to...)" (Represented by equivalent harmonic loss). Construct a bi-objective fitness function: ; In the formula, For the primary objective, the total physical volume of the magnetic core ( ); For the second objective, the total core loss (W) of the magnetic circuit at high frequencies; are the Steinmetz coefficients for Sendust and Ferrite, respectively.

[0079] (3) Physical boundary constraints; Constraint 1 (Magnetic Constraint): At any given time, inductance retention rate ; Constraint 2 (Thermal Constraint): The highest hot spot in the three-dimensional thermal field calculation under steady-state and transient combined heat generation. .

[0080] (4) Finding the optimal solution and then the inverse solution; After 500 iterations, the algorithm converged to the Pareto Front. The "inflection point solution" that balances volume boundary conditions and heat dissipation efficiency was selected. The final optimal physical volume ratio was determined to be: iron-silicon-aluminum magnetic powder core: ferrite = 7:3 (i.e., ...). ).

[0081] (iv) Implementation effect verification and data comparison The heterogeneous hybrid magnetic circuit reactor of the present invention was compared with the traditional single magnetic circuit scheme of the same specification in an experiment. The results are shown in the table below.

[0082] Table 3 shows a comparison of the comprehensive performance of different magnetic circuit design schemes under this extreme combined working condition.

[0083] Table 3 (v) Conclusion; This fourth embodiment fully demonstrates the "multi-objective optimization method" of the present invention. By constructing a revolutionary topology structure of "70% soft-saturated inner column to bear short-circuit energy storage + 30% high-frequency low-loss outer yoke to control ripple heating," the contradiction between low-frequency large excitation and high-frequency eddy current heating is successfully decoupled. Data proves that, while ensuring strong short-circuit overload maintenance capability, this design reduces the parasitic heating amplitude caused by wide-frequency high-frequency eddy currents by more than 60%, completely eliminating the safety hazards of container fires and explosions in energy storage power stations induced by weak grid resonance.

[0084] In summary, this invention completely breaks through the design flaw of traditional filter reactors that "simply rely on sacrificing heat dissipation volume to obtain safety margin" when dealing with extremely high overload conditions, and successfully defines the "high overload survival rule" for the core magnetic components of grid-type converters.

[0085] At the underlying logic level, this invention creatively constructs a three-in-one collaborative optimization framework of "magnetic circuit-structure-heat dissipation": In terms of hardware, the spatial architecture of "distributed micro-air gap + hybrid high-saturation composite magnetic material + embedded micro-channel vertical winding flat wire" is reconstructed, which eliminates deep magneto-saturation and high-frequency eddy current thermal blind zone from the physical source. In the software dimension, by introducing a dynamic temperature-frequency coupling operation strategy based on the modified iGSE loss model, traditional passive magnetic components are given the intelligent characteristic of "active defense" for the first time, enabling them to achieve peak shaving and valley filling under extreme short-term heat flow impact.

[0086] In summary, this invention represents not only an improvement on a single component but also a significant upgrade to the underlying key supporting technology of grid-connected power storage systems (GFM-PCS). It achieves a perfect decoupling and unification of "300% extreme short-circuit overload tolerance" and "high system power density (15% volume reduction)," avoiding the fatal problem of converters being forced to shut down due to inductance loss or insulation breakdown during grid faults. This provides a more robust and efficient hardware-level safety defense for new power systems, possessing extremely disruptive engineering potential and broad industrialization prospects.

[0087] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for magnetic-thermal constraint optimization design and operation of a high overload reactor of a grid-forming energy storage converter, characterized in that, include: Based on the rated parameters of the grid-type energy storage converter and the preset short-circuit overload conditions, a magnetic-thermal coupling constraint model is established, which includes an improved iGSE core loss model and dynamic heat dissipation constraint equations. Based on the magnetic-thermal coupling constraint model, the evolution of the magnetic constraint structure and the evolution of the thermal constraint structure are performed. The evolution of the magnetic constraint structure includes adopting a multi-air gap dispersion structure to suppress the edge magnetic flux effect, and the evolution of the thermal constraint structure includes adopting a conductor-optimized structure and an active heat dissipation structure. Based on the iterative results of the evolution of the magnetic confinement structure and the thermal confinement structure, the geometric parameters of the high overload reactor are output; Based on the real-time current value and real-time temperature estimate of the high overload reactor during operation, the switching frequency of the converter is dynamically adjusted to achieve loss redistribution under the short-circuit overload condition.

2. The method according to claim 1, characterized in that, Establishing the magnetic-thermal coupling constraint model further includes: Based on the duty cycle sequence of the grid-type converter under extreme operating conditions, the time-domain magnetic flux density waveform is obtained by Fourier fitting mapping; Based on the time-domain magnetic flux density waveform and the characteristic constants of the core material, the core loss per unit volume is calculated using the improved iGSE model. The dynamic heat dissipation constraint equation is established based on the core loss, the copper loss corrected by the skin effect gain coefficient, and the high overload power penalty coefficient of the mesh type.

3. The method according to claim 1, characterized in that, Performing the magnetically confined structure evolution further includes: The total number of distributed air gap segments is determined based on the vacuum permeability, number of coil turns, effective cross-sectional area of ​​the magnetic core, maximum short-circuit current, minimum allowable inductance, and magnetic core saturation magnetic induction intensity. The total air gap length is divided into multiple micro-air gaps, and the length of each micro-air gap is set separately to suppress the edge magnetic flux diffusion factor.

4. The method according to claim 1, characterized in that, Performing the magnetically confined structure evolution also includes: Based on the calculation results of the magnetic core loss using the magnetic-thermal coupling constraint model, amorphous alloy or iron-silicon-aluminum magnetic powder cores with high saturation magnetic induction intensity are selected as the magnetic core material.

5. The method according to claim 1, characterized in that, Dynamically adjusting the switching frequency of the converter further includes: The phase current of the high overload reactor is collected at a preset period, and the temperature of the reactor core area is estimated in real time based on the built-in thermal resistance-capacitance network model. When the phase current is greater than a preset overload threshold and the core area temperature is greater than a preset temperature threshold, a frequency reduction strategy is executed.

6. The method according to claim 5, characterized in that, Implementing the frequency reduction strategy further includes: The frequency reduction is calculated based on the difference between the core area temperature and the temperature threshold using a preset proportionality coefficient. A new switching frequency is determined based on the rated switching frequency and the frequency reduction, and the new switching frequency is limited to a preset minimum frequency.

7. The method according to claim 6, characterized in that, Dynamically adjusting the switching frequency of the converter also includes: When the phase current recovers to below the overload threshold or the core area temperature drops below the preset recovery threshold, the switching frequency is smoothly restored to the rated switching frequency through the hysteresis control loop.

8. The method according to claim 1, characterized in that, Performing the magnetically confined structure evolution also includes: A heterogeneous hybrid magnetic circuit consisting of an inner column and an outer yoke is constructed, wherein the inner column is selected from iron-silicon-aluminum magnetic powder cores with soft saturation characteristics, and the outer yoke is selected from manganese-zinc ferrite for suppressing high-frequency eddy current losses. The volume ratio of the inner column to the outer yoke is determined based on a multi-objective optimization algorithm.

9. An electronic device comprising a memory, a processor, and a computing program stored in the memory and executable on the processor, characterized in that, When the processor executes the computing program, it implements the method of any one of claims 1-8.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1-8.

Images