A design method for flat wire windings of electric motors with high slot fill factor and low AC loss.

Abstract

This invention discloses a method for designing flat wire windings for motors with high slot fill factor and low AC loss, comprising: acquiring stator parameters and high-frequency operating parameters of the target motor, constructing an electromagnetic simulation model of the stator core containing multiple stator slots; analyzing the leakage magnetic field density gradient curve distributed radially in the stator slots, generating a configuration sequence of the ladder cross-sectional thickness of the multi-layer flat wires in the stator slots; generating a U-shaped winding unit with a first span and a second span, wherein the two ends of the winding unit are respectively positioned at different wiring layers determined based on the first span and the second span; calculating the deflection matrix of the flat wire pins at the motor ends using an end topology reconstruction algorithm model with equal cross-sections; outputting the winding unit spatial configuration and winding wiring process data of the deflection matrix; decoupling the physical conflict between high-frequency eddy current loss and DC copper loss, and improving the high-frequency electromagnetic operating efficiency of the motor while maintaining the high slot fill factor constraint boundary.

Description

Technical Field

[0001] This invention relates to a method for designing flat wire windings for motors, and more particularly to a method for designing flat wire windings for motors with high slot fill factor and low AC loss. Background Technology

[0002] As modern motors continue to evolve towards higher efficiency and higher power density, flat wire winding technology has been widely used in electric vehicle drive motors, industrial servo motors, and other fields due to its ability to significantly improve slot fill factor, optimize heat dissipation, and enhance mechanical strength. However, in the pursuit of higher performance, especially for high-frequency operating conditions, traditional flat wire winding design faces a prominent contradiction. On the one hand, to fully utilize the limited slot space to increase the total cross-sectional area and current carrying capacity of the conductor, it is common to use flat wires with larger cross-sections and uniform dimensions, which can indeed achieve a very high initial slot fill factor. On the other hand, as the motor operating frequency increases, the skin effect and proximity effect in AC losses intensify dramatically. Under these conditions, the internal current distribution of large-section conductors becomes severely uneven, the equivalent AC resistance increases significantly, leading to a surge in high-frequency copper losses in the winding and a significant decrease in efficiency.

[0003] Simply reducing the overall conductor diameter or thickness to solve high-frequency loss problems, while suppressing AC effects to some extent, directly reduces the total conductor cross-sectional area, lowering slot space utilization (i.e., reducing the global slot fill factor), and consequently affecting the motor's continuous output capability and material utilization. Therefore, effectively suppressing AC losses at high frequencies while maintaining the inherent advantage of high slot fill factor in flat wire motors has become a key technical challenge in current flat wire motor design. Existing single-gauge conductor designs often fail to meet both requirements simultaneously, necessitating the exploration of new winding configurations and conductor arrangement methods. Summary of the Invention

[0004] This invention overcomes the shortcomings of the prior art and provides a design method for flat wire windings of motors with high slot fill factor and low AC loss.

[0005] To achieve the above objectives, the technical solution adopted by this invention is: a design method for a flat wire winding of a motor with high slot fill factor and low AC loss, comprising the following steps: S1: Obtain the stator parameters and high-frequency operating condition parameters of the target motor, and construct an electromagnetic simulation model of the stator core containing multiple stator slots; S2: Analyze the leakage magnetic field density gradient curve distributed radially in the stator slot. Based on the leakage magnetic field density gradient curve, generate the configuration sequence of the thickness of the stepped cross-section of the multi-layer flat wire in the stator slot through the stepped cross-section allocation algorithm model. The configuration sequence satisfies the following condition: the thickness of the flat wire cross-section near the groove opening layer is less than the thickness of the flat wire cross-section near the bottom of the groove. S3: Generate a U-shaped winding unit with a first span and a second span. The two ends of the winding unit are respectively positioned in different wiring layers determined based on the first span and the second span, and the two ends respectively have a target cross-sectional thickness corresponding to the wiring layer in the configuration sequence. S4: Calculate the deflection matrix of each layer of flat wire pins at the motor end using the equal cross-section end topology reconstruction algorithm model, so that any two pins that need to be connected by welding at the motor end can meet the equal cross-section matching condition of equal wiring layer thickness. S5: Outputs winding unit spatial configuration and deflection matrix winding wiring process data.

[0006] In a preferred embodiment of the present invention, the step of executing the tiered cross-section allocation algorithm model and generating the configuration sequence in S2 specifically includes: Construct a total loss assessment objective function model that includes DC copper loss and high-frequency AC loss; The stator slots are divided into strong leakage magnetic field region and weak leakage magnetic field region. Under high frequency operating conditions, the skin effect and proximity effect loss coefficients induced by the leakage magnetic field density gradient curve in each layer are calculated. Under the constraint of maintaining the total stator slot fill factor not lower than a preset threshold, the objective function model is evaluated by minimizing the total loss, the optimal cross-sectional thickness of each wiring layer is dynamically solved, and a configuration sequence is generated. The configuration sequence shows a thickness gradient distribution characteristic from the wiring layer closest to the stator slot opening to the wiring layer closest to the slot bottom, with the thickness increasing layer by layer or remaining partially unchanged.

[0007] In a preferred embodiment of the present invention, step S3, generating a winding unit having a first span and a second span, specifically includes: The first span and the second span are configured with different span values ​​to form a non-full pitch winding arrangement with alternating long and short pitches, thereby using the short pitch torque effect to suppress low-frequency circulating currents between parallel branches. Based on the configuration sequence, continuously formed variable cross-section conductors are generated as winding units; In the top U-shaped bend area of ​​the winding unit, a geometric model of the cross-section gradient transition segment is generated, which allows the cross-section to smoothly transition from the target cross-section thickness at one end to the target cross-section thickness at the other end.

[0008] In a preferred embodiment of the present invention, step S4, which involves executing an end topology reconstruction algorithm model based on equal cross-sections, specifically includes: Construct an initial connection adjacency matrix that represents the mapping relationship between the interconnections of all winding pins at the motor end; In the traversal optimization of the initial connection adjacency matrix, an equal cross-section matching constraint model is incorporated: it is determined whether the cross-sectional thickness of the wiring layer where the two pins to be connected in the initial mapping are located is equal. If the determination result is that there is a difference in cross-sectional thickness, the variable step size torsion adjustment logic algorithm is triggered to break the fixed pitch of the first span or the second span, generate a non-linear pin deflection angle fine adjustment amount to reconstruct the connection topology path, until a pin that meets the equal cross-sectional matching condition is found to connect the guide section or series section. The set of all pin deflection angles that have been reconstructed and verified is output as the deflection matrix.

[0009] In a preferred embodiment of the present invention, the design method further includes: generating an asymmetric high-frequency oil-cooled microchannel, specifically: When each wiring layer uses a configuration sequence, calculate the redundancy of the lateral gaps relative to the maximum available slot width of the stator slot; The algorithm model of spatial rearrangement in the stator slot is used to shift the redundancy generated near the slot opening layer to the same side of the stator slot. In the thin-section flat wire region of the slot, the redundancy after offset is used to generate a three-dimensional model of the high-frequency oil-cooled microchannel with an asymmetrical distribution, and the three-dimensional model data is attached to the winding wiring process data.

[0010] In a preferred embodiment of the present invention, in S5, the output winding wiring process data is used to directly input to the control system of the automated robotic arm and laser welding equipment to guide the non-destructive insertion positioning and equal cross-section alignment welding process of the irregular cross-section flat wire.

[0011] A high slot fill factor and low AC loss flat wire winding design system for electric motors includes: Memory, used to store computer programs; A processor is a step in implementing a design methodology when executing a computer program.

[0012] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of a design method.

[0013] A motor stator assembly is manufactured using winding wiring process data output from a design method; The motor stator assembly includes a stator core and multiple layers of flat wire windings passing through the stator slots. The cross-sectional thickness of the flat wires near the slot opening is less than that near the bottom of the slot, and the top of all flat wire winding units is in a non-equal thickness state across layers, while the two pins at their end welded connections are in a cross-layer equal thickness butt joint state.

[0014] In a preferred embodiment of the present invention, an asymmetrical gap is provided between the thin-section flat wire sidewall of the stator slot opening and the inner wall of the stator slot, with one side tightly attached to the insulating paper and the other side left open. The asymmetrical gap extends through the axial direction of the stator core, forming a built-in high-frequency oil cooling microchannel.

[0015] This invention addresses the shortcomings of the prior art and has the following beneficial effects: (1) This invention provides a design method for a flat wire winding of a motor with high slot fill factor and low AC loss. By combining the leakage magnetic field density gradient curve with the ladder cross section allocation algorithm model, a configuration sequence and irregular winding unit with thin flat wire cross section at the stator slot opening and thick flat wire cross section at the slot bottom are generated. The microscopic non-uniformity of the alternating magnetic field in the stator slot is transformed into a data sequence that guides the variation of the physical shape of the winding. Thinner wires are allocated in the area of ​​dense leakage flux to cut off the eddy current loop, and thicker wires are retained in the area of ​​sparse leakage flux to reduce DC resistance. Compared with the defects of the prior art, which is that the use of uniform large cross section flat wire leads to a surge in high frequency AC loss or the reduction of the wire diameter leads to a decrease in global slot fill factor, this solution decouples the physical conflict between high frequency eddy current loss and DC copper loss, and improves the high frequency electromagnetic operating efficiency of the motor under the condition of maintaining the high slot fill factor constraint boundary.

[0016] (2) This invention provides a design method for flat wire windings of motors with high slot fill factor and low AC loss. Through the equal cross-section end topology reconstruction algorithm model, the pin deflection angle fine adjustment is calculated and the deflection matrix is ​​output so that the pins to be welded at the end meet the equal cross-section matching condition. The pins are connected in three-dimensional space and transformed into node topology optimization logic. The nonlinear deflection step size is used to break the fixed pitch and drive the non-equal thickness irregular winding pins across layers. Equal thickness docking is achieved at the end welding point. Compared with the existing technology that introduces variable cross-section wires in the stator slot to generate thick and thin pin docking, which is prone to laser focus shift and cold welding defects, this solution eliminates the structural interference caused by variable cross-section without adding non-standard laser welding equipment, and realizes high-yield automated non-destructive welding manufacturing of complex flat wire windings.

[0017] (3) This invention provides a design method for a flat wire winding of a motor with high slot fill factor and low AC loss. By using the stator slot space rearrangement algorithm model, the lateral gap redundancy generated by the configuration sequence of the wiring layer is shifted to the same side of the stator slot and an asymmetric high-frequency oil cooling microchannel is generated. The conductors of the slot area with dense heat generation are directly in the boundary of the cooling medium fluid. Compared with the prior art, after using thin flat wire, the assembly gap is disordered or the stator space resources are idle due to the decrease in local slot fill factor. This invention directly transforms the electromagnetic physical loss reduction space into an active thermal management cooling channel, which expands the continuous output boundary of the motor at high frequency without changing the stator peripheral water jacket structure. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Figure 1 This is a flowchart of a preferred embodiment of the present invention; Figure 2 This is a schematic diagram of the stepped cross-section configuration sequence of a preferred embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein. Therefore, the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0021] Application Overview: In the field of high-speed, high-fundamental-frequency drive motor technology, a double-span flat wire winding structure with a uniform cross-section is adopted. This structure can improve the stator slot fill factor, reduce DC copper loss, and suppress low-frequency circulating current between parallel branches. However, under high-frequency operating conditions, the leakage magnetic field distribution in the stator slot exhibits significant radial non-uniformity, with the leakage magnetic flux density near the slot opening being much higher than that at the slot bottom. The uniform large-section flat wire will generate a strong skin effect and proximity effect in the strong leakage magnetic field area at the slot opening, causing a surge in AC loss and local heat accumulation.

[0022] If the cross-sectional area of ​​the flat wire is reduced to suppress high-frequency AC loss, the space utilization rate of the weak leakage magnetic field area at the bottom of the slot will decrease, resulting in a decrease in the overall slot fill factor and an increase in DC resistance. If variable cross-section flat wires are introduced directly into different layers in the stator slot, the winding pins will have cross-layer thickness differences when connected at the ends, resulting in cross-sectional mismatch of the pins to be welded, which cannot meet the process requirements of automated laser welding alignment. Therefore, there is a technical contradiction in the current technology between maintaining a high overall slot fill factor, suppressing local high-frequency AC loss, and adapting to the automated welding process with equal cross-sections at the ends.

[0023] This invention addresses the issues of high local eddy current losses and mismatched welding at the ends of variable cross-section conductors in high-frequency motor stator windings by employing a stepped cross-section allocation algorithm model and an end topology reconstruction algorithm model based on equal cross-sections. The invention analyzes the stator slot leakage magnetic field density gradient curve to generate a configuration sequence for multi-layer flat wires with progressively increasing thickness from the slot opening to the slot bottom, and generates irregularly shaped U-shaped winding units. Simultaneously, it constructs an end-connection adjacency matrix and incorporates an equal cross-section matching condition model. By triggering a variable step-size torsion adjustment logic algorithm, it reconstructs the end-connection topology path and calculates the deflection matrix to ensure that the cross-sectional thickness of the pins to be connected at all welding nodes is equal. Furthermore, it laterally offsets the lateral gap redundancy in the thin cross-section region of the slot opening, generating asymmetrically distributed high-frequency oil-cooled microchannels.

[0024] Compared with the prior art, this invention decouples the conflict between high slot fill factor and low AC loss by spatial matching of leakage magnetic field gradient and cross-sectional thickness; and overcomes the process obstacle of standardized automated alignment welding of irregular cross-section conductors by using end topology reconstruction algorithm; and reconstructs the thermal management structure in the stator slot by rearranging the redundant space generated by the decrease in winding thickness, thereby improving the electromagnetic efficiency and overall power density of the motor stator assembly.

[0025] Example 1: A method for designing flat wire windings for motors with high slot fill factor and low AC loss, comprising the following steps: S1: Obtain the stator parameters and high-frequency operating condition parameters of the target motor, and construct an electromagnetic simulation model of the stator core containing multiple stator slots; S2: Analyze the leakage magnetic field density gradient curve distributed radially in the stator slot. Based on the leakage magnetic field density gradient curve, generate the configuration sequence of the thickness of the stepped cross-section of the multi-layer flat wire in the stator slot through the stepped cross-section allocation algorithm model. The configuration sequence satisfies the following condition: the thickness of the flat wire cross-section near the groove opening layer is less than the thickness of the flat wire cross-section near the bottom of the groove. S3: Generate a U-shaped winding unit with a first span and a second span. The two ends of the winding unit are respectively positioned in different wiring layers determined based on the first span and the second span, and the two ends respectively have a target cross-sectional thickness corresponding to the wiring layer in the configuration sequence. S4: Calculate the deflection matrix of each layer of flat wire pins at the motor end using the equal cross-section end topology reconstruction algorithm model, so that any two pins that need to be connected by welding at the motor end can meet the equal cross-section matching condition of equal wiring layer thickness. S5: Outputs winding unit spatial configuration and deflection matrix winding wiring process data.

[0026] It should be noted that, in order to suppress AC losses at the stator slot opening under high-frequency operating conditions, the flat wire with a uniform cross-section in the prior art needs to be replaced with a variable cross-section winding based on the leakage magnetic field gradient distribution when the two ends of the U-shaped winding unit are distributed on wiring layers with different cross-sectional thicknesses. If the wiring logic with fixed span and fixed deflection angle is used to perform the end twisting and welding process, it will cause a difference in cross-sectional thickness between the two pins to be welded, resulting in structural interference due to mismatch in cross-sectional dimensions. This difference in cross-sectional thickness will cause focal shift, incomplete welding, and weld explosion defects in the laser welding process, thereby reducing the yield rate of automated manufacturing of motor stators. Therefore, the key technical problem to be solved in the implementation of this invention is to achieve equal thickness matching and alignment of the end pins of all variable cross-section windings at the welding node through the reconstruction of the end spatial topology path without adding non-standard welding equipment.

[0027] It should also be noted that the aforementioned deflection matrix refers to the set of numerical values ​​representing the target angle or displacement parameters of the flat wire pins at the stator end in three-dimensional space during the end-twisting forming process of flat wire motor manufacturing. This data set is represented and calculated in the form of a mathematical matrix in the underlying algorithm model. In actual automated CNC machining, it directly corresponds to and is transformed into the physical twisting angle and deflection stroke parameters of the CNC robotic arm or twisting fixture. In S1 of this embodiment, the conventional parameters of the target motor, such as the pole slot fit, core size, and highest operating fundamental frequency, are obtained. A three-dimensional electromagnetic simulation model of the stator core is constructed using conventional electromagnetic simulation methods of finite element analysis, which serves as the physical load basis for subsequent magnetic field analysis.

[0028] In S2 of this embodiment, improvements are made from the perspective of spatial decoupling of the physical mechanism of loss. One-dimensional transverse leakage flux data of stator slots is extracted, and a distribution mapping relationship between leakage flux density and radial slot depth is established. Based on this, the stepped section allocation algorithm model mainly performs multi-objective optimization from the dimension of balancing low-frequency DC copper loss and high-frequency AC eddy current loss. It dynamically outputs a set of discrete thickness value sequences, forcibly constrains the flat wire thickness of the strong leakage flux region to be the minimum, and gradually thickens it towards the weak leakage flux region, thereby weakening the local proximity effect within the boundary framework of maintaining the overall slot fill factor.

[0029] In S3 of this embodiment, since the present invention introduces the physical structure of the U-shaped winding unit with irregular cross-section, the continuous variable cross-section conductor is prepared by a stepped drawing or stamping process. In the three-dimensional spatial forming stage, the top U-shaped bending area is set as a smooth cross-section gradient transition section, so that after the U-shaped winding unit is inserted into the stator slot, the straight slot section located in the wiring layer of the first span and the straight slot section located in the wiring layer of the second span are assigned independent target cross-section thicknesses in the configuration sequence.

[0030] In S4 of this embodiment, the improvement is made from the software aspect of the wiring connection relationship. A topology mapping space is constructed with all the winding pins to be connected at the stator end as nodes. The physical thickness of the corresponding wiring layer is absolutely equal as a mandatory optimization boundary condition for end reconstruction. When the algorithm detects that there is a cross-sectional difference between two pins connected by the original rule, it breaks the original uniform symmetrical twisted connection form by dynamically fine-tuning the span step size and spatial deflection angle of the local pins, and re-plans the convergence path of the end pins until all the node pairs to be welded meet the equal cross-sectional matching condition. Finally, the nonlinear deflection data of each pin is extracted to form a deflection matrix.

[0031] In S5 of this application, the three-dimensional spatial configuration data and deflection matrix containing the physical parameters of the irregular U-shaped winding unit are converted into standard machine-readable process data and directly sent to the control system of the automated CNC forming equipment, the wire insertion robotic arm and the laser welding equipment to guide the complete winding wiring process manufacturing.

[0032] Example 2: S1: Obtain the stator parameters and high-frequency operating condition parameters of the target motor, and construct an electromagnetic simulation model of the stator core containing multiple stator slots; It should be noted that in S1, the operation of acquiring the stator parameters and high-frequency operating condition parameters of the target motor and constructing an electromagnetic simulation model of the stator core containing multiple stator slots constitutes the physical environment modeling and data input stage of the high slot fill factor and low AC loss flat wire winding design method of this invention. By reading the stator structure parameters of the motor stator core, such as the outer diameter, inner diameter, slot type structure, number of slots, and pole-slot matching, and simultaneously reading the high-frequency operating condition parameters such as the highest operating fundamental frequency and phase current amplitude, the processing device uses the above input parameters to establish a two-dimensional or three-dimensional geometric boundary of the mapped physical stator core and its multiple internal stator slots in the virtual space, and performs finite element meshing on the internal region of the stator slots to generate an electromagnetic simulation model of the stator core. This provides the necessary boundary constraints and initial excitation conditions for subsequent analysis of the leakage magnetic field density gradient curve distributed radially along the stator slots. By establishing an accurate electromagnetic simulation physical space, subsequent steps can accurately extract the non-uniform distribution characteristics of leakage magnetic flux in different wiring layers inside the stator slots, thus laying a calculable underlying data support for generating the configuration sequence of the ladder cross-sectional thickness.

[0033] In the specific execution process of constructing the stator core electromagnetic simulation model, this invention employs a time-harmonic electromagnetic field finite element mathematical model to analyze the physical field characteristics within the slot. The core principle of this model is to solve Maxwell's equations within the closed stator slot boundary under high-frequency alternating excitation conditions, converting macroscopic stator parameters and operating condition data into a microscopic spatial magnetic flux density matrix. In specific implementation details, the spatial magnetic vector potential within the stator slot region is governed by partial differential equations. The constraint is that, in this calculation formula, A is the spatial magnetic vector position matrix; The permeability parameter characterizing the material of the stator core and slot region is determined by looking up the nonlinear BH curve of the selected silicon steel sheet material in a table. The source current density vector is calculated from the phase current amplitude and the initial winding cross-sectional area in the obtained high-frequency operating condition parameters. The angular frequency corresponding to the high-frequency operating condition parameters; Characterizes the conductivity of the conductive material inside the stator slot.

[0034] The system sets the outer boundary of the stator core as a Dirichlet boundary condition and iteratively applies the aforementioned governing equations to the gridded stator slot region, calculating and outputting the magnetic vector potential distribution of each grid node. Then, at any position within the radial depth of the stator slot, the curl equation is used to... Let's derive the leakage flux density vector. The introduction of this time-harmonic electromagnetic field finite element mathematical model enables accurate numerical quantification of the high-frequency alternating magnetic field distribution within the confined space of the stator slot. This improvement solves the problem that traditional static or low-frequency empirical models cannot accurately assess the severe skin effect and proximity accumulation effect of leakage flux in the slot after the fundamental frequency exceeds a certain threshold. The output flux density vector directly provides a highly reliable physical input variable for subsequent accurate matching of the thickness of the stepped flat wire cross section.

[0035] S2: Analyze the leakage magnetic field density gradient curve distributed radially in the stator slot. Based on the leakage magnetic field density gradient curve, generate the configuration sequence of the thickness of the stepped cross-section of the multi-layer flat wire in the stator slot through the stepped cross-section allocation algorithm model. The configuration sequence satisfies the following condition: the thickness of the flat wire cross-section near the groove opening layer is less than the thickness of the flat wire cross-section near the bottom of the groove. In a preferred embodiment of the present invention, the step of executing the tiered cross-section allocation algorithm model and generating the configuration sequence in S2 specifically includes: Construct a total loss assessment objective function model that includes DC copper loss and high-frequency AC loss; The stator slots are divided into strong leakage magnetic field region and weak leakage magnetic field region. Under high frequency operating conditions, the skin effect and proximity effect loss coefficients induced by the leakage magnetic field density gradient curve in each layer are calculated. Under the constraint of maintaining the total stator slot fill factor not lower than a preset threshold, the objective function model is evaluated by minimizing the total loss, the optimal cross-sectional thickness of each wiring layer is dynamically solved, and a configuration sequence is generated. The configuration sequence shows a thickness gradient distribution characteristic from the wiring layer closest to the stator slot opening to the wiring layer closest to the slot bottom, with the thickness increasing layer by layer or remaining partially unchanged.

[0036] It should be noted that, in this embodiment, this step, by analyzing the leakage magnetic field density gradient curve distributed radially along the stator slot and generating a sequence of stepped cross-sectional thickness configurations, constitutes the core physical field decoupling stage of the present invention for resolving the technical contradiction between high-frequency AC loss and high slot fill factor. This step extracts the spatial magnetic vector potential distribution inside the stator slot from different radial depths from the slot opening to the slot bottom by reading the stator core electromagnetic simulation model data output from the previous step, and fits it to form a leakage magnetic field density gradient curve characterizing the non-uniform distribution of leakage flux.

[0037] Utilizing the aforementioned leakage magnetic field density gradient curve, which reflects the physical characteristics of high leakage magnetic flux density in the slot opening region and low leakage magnetic flux density in the slot bottom region, the step-by-step cross-section allocation algorithm model is triggered. This algorithm model transforms the single and fixed full-slot winding size parameter into multi-layered decoupled independent spatial variables. Smaller cross-sectional thickness parameters are allocated to the stator slot opening region with dense leakage magnetic flux, while larger cross-sectional thickness parameters are allocated to the stator slot bottom region with sparse leakage magnetic flux. This transforms the non-uniformity of the alternating magnetic field within the stator slot into a specific data sequence guiding the winding. This not only cuts off the physical conditions for the exponential increase of high-frequency skin effect and proximity effect in the slot opening region at the source, but also provides a customized geometric dimension input matrix for the subsequent generation of irregularly shaped U-shaped winding units, enabling the subsequent end topology reconstruction to have a clear objective function for equal cross-section matching.

[0038] In the specific process of executing the above-mentioned ladder cross-section allocation algorithm model and generating the configuration sequence, this embodiment adopts microscopic loss analysis under high-frequency alternating magnetic field and geometric boundary multi-objective optimization algorithm model. The core optimization principle of this model is to quantitatively evaluate the difference in AC eddy current loss caused by different radial spatial positions of stator slots, and to seek the global minimum value of the sum of DC copper loss and high-frequency AC loss of all wiring layers in the slot within the constraint boundary while ensuring that the total conductive cross-sectional area meets the constraint boundary.

[0039] The system establishes an objective function model for evaluating the total loss within the stator core slot region. This objective function is expressed as follows: In this computational model, the independent variable T represents a configuration sequence containing 2n layers of flat wire cross-sectional thickness. Where k is the radial layer index number of the wiring layer increasing from the slot opening to the slot bottom; I is the rated phase current amplitude under high-frequency operating conditions. denoted as the resistivity of the flat wire conductor material; L is the effective axial length of the stator core; w is the fixed cross-sectional width of the flat wire, limited by the stator slot width. The optimal cross-sectional thickness of the k-th layer flat wire is the solution required by the algorithm. The operating angular frequency of the alternating current; The average local leakage magnetic field density of the k-th layer is extracted based on the leakage magnetic field density gradient curve. The forward accumulation term in the formula represents the DC copper loss distribution of each layer, and the backward accumulation term represents the high-frequency AC loss coefficient distribution induced by the skin effect and proximity effect induced by the leakage magnetic field.

[0040] The algorithm model sets strict physical space and performance boundary logic judgment conditions when iteratively solving the total loss evaluation objective function. The system executes the space constraint equation to maintain the total stator slot fill factor not lower than a preset threshold. Candidate solutions are filtered, where d is the total thickness of interlayer insulation, H is the effective radial depth of the stator slot, and S is the preset slot full rate threshold. During the objective function differentiation and gradient descent optimization process, due to the extremely large [value] in the strong leakage magnetic field region... Numerical value and thickness Parameters are adjusted to suppress the exponential growth of AC losses, while simultaneously increasing the weak leakage flux region near the stator slot bottom under the constraint of maintaining the slot fill factor space constraint equation. Parameters were adjusted to compensate for the reduction in cross-sectional area and decrease DC copper losses. The system was further integrated with... The monotonicity of the logic constraints forces the solution output to have a thickness gradient distribution sequence with the characteristic of increasing thickness layer by layer or partially remaining unchanged. This quantitative loss mapping, combined with the computer mechanism of nonlinear parameter optimization, overcomes the traditional design limitation that the cross-sectional size of the conductor in the stator slot is determined by a uniform empirical value. It solves the technical problems of reducing the size of the uniform flat wire to suppress the sharp reduction of the global slot fill factor caused by high-frequency AC loss and the serious deterioration of the DC resistance of the motor, and realizes the physical parameter configuration that maximizes the overall operating efficiency and power density of the motor.

[0041] S3: Generate a U-shaped winding unit with a first span and a second span. The two ends of the winding unit are respectively positioned in different wiring layers determined based on the first span and the second span, and the two ends respectively have a target cross-sectional thickness corresponding to the wiring layer in the configuration sequence. In a preferred embodiment of the present invention, step S3, generating a winding unit having a first span and a second span, specifically includes: The first and second spans are configured with unequal values ​​to form a non-full-pitch winding arrangement with alternating long and short pitches. This utilizes the short-pitch torque effect to suppress low-frequency circulating currents between parallel branches. For example, the first span is configured as span 11, and the second span is configured as span 8; Based on the configuration sequence, continuously formed variable cross-section conductors are generated as winding units; In the top U-shaped bend area of ​​the winding unit, a geometric model of the cross-section gradient transition segment is generated, which allows the cross-section to smoothly transition from the target cross-section thickness at one end to the target cross-section thickness at the other end.

[0042] It should be noted that in this embodiment, this step reads the ladder section thickness configuration sequence output by the previous step, as well as the preset stator core span arrangement rules. Based on the set first span and second span, the system defines in the three-dimensional virtual space the specific radial wiring layers that need to be inserted into the straight slots on both sides of each U-shaped winding unit. Due to the existence of the span, the straight slots on both sides of the U-shaped winding unit must fall into spatial layers with different magnetic field densities. The system calls the target section thickness values ​​corresponding to the two different wiring layers in the configuration sequence and assigns them to the straight slots on both sides of the winding unit, thereby rendering and generating an irregular U-shaped conductor model with unequal thickness at both ends. In this way, the independent thickness parameters of each layer decoupled due to the non-uniform distribution of leakage magnetic field in the stator slot are reintegrated into a continuous winding entity, so that a single U-shaped flat wire has heterogeneous properties that match the magnetic field gradient.

[0043] In performing the above steps of generating a winding unit with a first span and a second span, this scheme specifies in detail the specific values ​​of the double span and the continuous forming transition characteristics of the variable cross-section conductor. The system is configured to have unequal span values ​​for the first span and the second span. Taking the first span as span 11 and the second span as span 8 as an example, this combination of alternating long and short pitches constitutes a non-full-pitch winding arrangement. It can utilize the short-pitch effect to weaken the high-order harmonics in the air gap magnetic field and form a reverse induced electromotive force phasor in the winding end connection circuit to offset the potential difference between parallel branches in the same phase, thereby suppressing the low-frequency circulating current generated between parallel branches.

[0044] Based on the different thickness values ​​corresponding to the above-mentioned span and layer positions and configuration sequences, the system generates a continuously formed variable cross-section conductor as a winding unit. This variable cross-section conductor is a complete continuous conductor integrally formed by a stepped drawing or rolling process. Considering the physical characteristic that there is a difference in thickness at both ends of the same winding entity, the system generates a geometric model of a gradually changing cross-section transition section at the top U-shaped bend area outside the stator slot of the winding unit. Under the constraint of maintaining a constant radial width of the flat wire, the geometric model drives the cross-sectional thickness of the flat wire to continuously and smoothly evolve from the smaller target cross-sectional thickness corresponding to the straight slot section at one end to the larger target cross-sectional thickness corresponding to the straight slot section at the other end, along the three-dimensional spatial curve of the U-shaped bend trajectory.

[0045] Regarding the generation process of the geometric model of the aforementioned cross-section gradient transition section, this optimized embodiment adopts a spatial geometric modeling interpolation algorithm model with a continuous equal width gradient cross-section. The core optimization principle of this algorithm model is to construct a spatial nonlinear transition curve with first derivative continuity under the premise of ensuring volume continuity and the integrity of the conductive channel, so as to eliminate the geometric discontinuities caused by the abrupt change in thickness.

[0046] In the specific implementation details, the system extracts the curve arc length coordinate variable 's' based on the spatial center axis of the U-shaped bend area, and sets the total arc length of the entire cross-section gradual transition section as... System construction of cross-sectional thickness spatial evolution function equation In this calculation formula, This is the initial target cross-sectional thickness mapped to the wiring layer containing the first span based on the configuration sequence; The target cross-sectional thickness is mapped to the wiring layer containing the second span; the range of the arc length coordinate variable s is constrained to [0, ].

[0047] The system uses this cubic polynomial interpolation model logic to force the boundary conditions to satisfy s=0 and The derivative of the thickness change rate is completely zero. This improvement achieves a smooth transition effect in the cross section, eliminates the mechanical stress concentration caused by the step-like thickness jump, and solves the manufacturing defect problem of local copper wire breakage or insulation film damage that is very easy to occur in the process of automatic robotic gripping and three-dimensional bending of irregular cross-section flat wire. At the same time, it avoids the high-frequency current skin impedance distortion and local electromagnetic hot spot accumulation caused by the drastic change of conductive cross section in the top bending area.

[0048] S4: Calculate the deflection matrix of each layer of flat wire pins at the motor end using the equal cross-section end topology reconstruction algorithm model, so that any two pins that need to be connected by welding at the motor end can meet the equal cross-section matching condition of equal wiring layer thickness. In a preferred embodiment of the present invention, step S4, which involves executing an end topology reconstruction algorithm model based on equal cross-sections, specifically includes: Construct an initial connection adjacency matrix that represents the mapping relationship between the interconnections of all winding pins at the motor end; In the traversal optimization of the initial connection adjacency matrix, an equal cross-section matching constraint model is incorporated: it is determined whether the cross-sectional thickness of the wiring layer where the two pins to be connected in the initial mapping are located is equal. If the determination result is that there is a difference in cross-sectional thickness, the variable step size torsion adjustment logic algorithm is triggered to break the fixed pitch of the first span or the second span, generate a non-linear pin deflection angle fine adjustment amount to reconstruct the connection topology path, until a pin that meets the equal cross-sectional matching condition is found to connect the guide section or series section. The set of all pin deflection angles that have been reconstructed and verified is output as the deflection matrix.

[0049] It should be noted that in this embodiment, this step reads the spatial arrangement information of the irregular U-shaped winding units generated and located in each wiring layer in the previous step. Without changing the optimal ladder cross-section allocation state in the stator slot, the three-dimensional pin connection problem is transformed into a spatial connection matching problem at the mathematical level. The system uses this algorithm model to identify all pin pairs to be welded at the stator end that intersect according to the conventional span rule, removes the connection nodes that produce cross-sectional size differences, and replans the deflection path to output the final deflection matrix in which all pins meet the equal cross-section alignment. This completely eliminates the cross-layer docking conflict of thick and thin pins caused by the leakage magnetic field adaptation and variable cross-section conductor introduced in the previous step. While retaining the electromagnetic optimization loss reduction and resistance reduction effect, it uses flexible reconstruction to compensate for the process obstacles caused by hardware structure variation, providing a feasible process path for subsequent output of standardized winding wiring process data and realization of automated non-destructive alignment welding.

[0050] In the specific process of executing the end topology reconstruction algorithm model based on equal cross section and generating the deflection matrix, this optimization embodiment adopts the variable step size end path optimization algorithm model based on restricted graph theory. This model maps all flat wire pins at the motor end to discrete nodes of the topology graph and maps the spatial connection paths between pins to directed edges. Under the global constraint of ensuring that the electrical phase logic connectivity of the three-phase windings of the motor is not destroyed, the physical connection edges are reorganized through the heuristic nonlinear step size iteration law until the cross section thickness attributes of the nodes at both ends of all connection edges are absolutely consistent.

[0051] The system constructs an initial connection adjacency matrix A, which maps the interconnections of all winding pins at the motor ends. Its internal matrix elements... Based on the initially set first or second span, the i-th pin and the j-th pin have an initial electrical connection and physical intersection mapping; when traversing and optimizing this initial connection adjacency matrix A, the system incorporates the equal cross-section matching constraint condition model equation. ,in and These are, respectively, the system extracts from the ladder cross-sectional thickness configuration sequence and maps to the cross-sectional thickness parameters of the wiring layer where the i-th pin and j-th pin are located.

[0052] The system executes logical judgment conditions, and when the judgment result is... When there is a difference in cross-sectional thickness, the variable step size torsion adjustment logic algorithm is immediately triggered. This algorithm establishes a calculation model for pin deflection angle reconstruction. In this computational model, The target deflection angle for the reconstructed pin; The initial spatial deflection angle is based on a fixed pitch; Q is the total number of stator slots in the stator core. This is a dynamically generated integer variable step size adjustment value for the algorithm.

[0053] The system adjusts continuously The value of breaks the original fixed pitch, generates a non-linear pin deflection angle fine adjustment, drives the connection topology path, performs redirection mapping search between adjacent slots and different wiring layers until a new target pin k is found, so that the constraint equation satisfies C(i,k)=0 and the electrical phase is legal, thereby establishing a new connection relationship of the guide section or series section.

[0054] The system ultimately extracts the set of all pin target deflection angles after reconstruction and verification as the deflection matrix output. The introduction of this algorithm model and the improvement of its dynamic space optimization mechanism fundamentally eliminate the welding hot melt focus offset and end face non-fusion phenomenon caused by direct docking of thick flat wires and thin flat wires. It solves the technical problem that high full-slot ratio variable cross-section motors cannot be mass-produced using standard laser welding production lines. Under the condition of not needing to develop complex non-standard welding equipment, it achieves extremely high forming yield and structural reliability of complex asymmetric winding end connections.

[0055] S5: Outputs winding unit spatial configuration and deflection matrix winding wiring process data.

[0056] In a preferred embodiment of the present invention, in S5, the output winding wiring process data is used to directly input to the control system of the automated robotic arm and laser welding equipment to guide the non-destructive insertion positioning and equal cross-section alignment welding process of the irregular cross-section flat wire.

[0057] It should be noted that in this embodiment, this step reads the three-dimensional irregular winding entity model data generated in the previous step, as well as the end pin deflection matrix verified by topology optimization. The system compiles the geometric topology results containing parameters and spatial path information into structured data that can be recognized by the industrial control system. This completes the leap from front-end virtual electromagnetic field parameter optimization and flexible spatial topology reconstruction to the bottom-level control instructions of the back-end physical manufacturing production line. By outputting the integrated winding wiring process data, the complex ladder cross-section distribution shape and the asymmetric twisting path with non-fixed step size are removed from the scope of pure digital simulation. This provides the downstream automated processing equipment with three-dimensional spatial positioning basis and action execution logic, and establishes the feasibility closed loop of the high-frequency loss reduction design scheme of the present invention in a real industrial mass production environment.

[0058] In the specific process of executing the output winding wiring process data and directly inputting it into the control system of the automated robotic arm and laser welding equipment, this optimized embodiment adopts a machine instruction coordinate transformation and laser energy adaptive mapping algorithm model. This algorithm model converts the relative geometric spatial relationship and angle parameters generated in the previous steps into the absolute physical execution coordinates of the multi-axis processing equipment, and dynamically gives the corresponding laser heat source output parameters for the end pin thickness characteristics that have been matched with equal cross-sections, so as to ensure the non-interference of physical wiring assembly and the consistency of end welding fusion.

[0059] The system first establishes a spatial positioning transformation calculation model. In this computational model, Execute the coordinate matrix for the three-dimensional target at the end of the gripper of the automated robotic arm; This is the homogeneous translation transformation matrix from the local reference coordinate system of the stator core to the global reference coordinate system of the equipment; The rotation transformation matrix is ​​generated to extract the target deflection angle parameter of the i-th pin in the deflection matrix of the system. The initial positioning coordinates of the irregular cross-section flat wire determined according to the aforementioned configuration sequence and winding unit span rules; A preset spatial avoidance margin matrix to avoid complex interlayer physical collisions at the end.

[0060] The system further extracts the pin thickness parameters after the equal cross-section verification is completed, and constructs an adaptive matching equation for the thermal input of laser welding. In this equation, The target command value for the output line energy density of the laser welding equipment applied to the kth uniform cross-section aligned welding node; This represents the physical thickness of the cross-section at which the two pins at this node are absolutely equal. To pre-calibrate and store the specific wavelength laser absorptivity and dynamic thermal conductivity constant of the copper conductor material.

[0061] The control system of this equipment, by analyzing and loading the aforementioned spatial coordinates and energy matching commands, drives the robotic arm along the optimal spatial trajectory to perform clamping and non-destructive insertion positioning of irregularly shaped cross-section flat wires. The improvements to the aforementioned algorithm model and process data mapping mechanism have solved the problem of insulation varnish film extrusion and damage caused by mechanical interference in the complex twisted space of irregularly shaped winding conductors with varying step lengths. At the same time, by utilizing the calculated single laser line energy density, it synchronously melts through the aligned pins of equal thickness, completely eliminating the defects of thin wire overheating and collapse or thick wire cold welding failure that inevitably occur when facing pins with unequal thickness in traditional fixed welding processes. This achieves high yield and fully automated manufacturing of complex asymmetric motor flat wire winding configurations.

[0062] In a preferred embodiment of the present invention, the design method further includes: generating an asymmetric high-frequency oil-cooled microchannel, specifically: When each wiring layer uses a configuration sequence, calculate the redundancy of the lateral gaps relative to the maximum available slot width of the stator slot; The algorithm model of spatial rearrangement in the stator slot is used to shift the redundancy generated near the slot opening layer to the same side of the stator slot. In the thin-section flat wire region of the slot, the redundancy after offset is used to generate a three-dimensional model of the high-frequency oil-cooled microchannel with an asymmetrical distribution, and the three-dimensional model data is attached to the winding wiring process data.

[0063] It should be noted that in this embodiment, this step extracts the configuration sequence output by the aforementioned ladder cross-section allocation algorithm model, compares the maximum available space boundary of the stator core with the actual allocated slot thin-section flat wire physical dimensions, and quantifies the idle space data inside each wiring layer. Through the stator slot space rearrangement algorithm model, the assembly alignment rule of the conventional centrally symmetrical arrangement of motor windings is broken, driving the thin-section windings in the slot area to move towards a specific side of the stator slot as a whole. The tiny assembly gaps that were originally discretely distributed on both sides of the windings are concentrated and aggregated on the other side of the stator slot, thereby constructing a continuous three-dimensional asymmetric oil-cooling microchannel inside the stator slot. The structural model of this microchannel is then added to the final output winding wiring process data. In this way, the idle space generated by reducing the radial thickness of the conductor in the previous step to reduce high-frequency AC loss is directly transformed into an effective cooling channel of the motor active thermal management system. This achieves deep reuse of the electromagnetic loss reduction structure and the fluid heat dissipation structure in the same confined space, making up for the space utilization defects caused by the decrease in the apparent value of local slot fill factor due to the variable cross-section design.

[0064] In the specific process of performing the above spatial rearrangement and generating asymmetric high-frequency oil-cooled microchannels, this embodiment adopts a fluid channel geometry optimization algorithm model based on discrete boundary aggregation. This model uses coordinate translation transformation to perform unilateral spatial superposition of the area of ​​the scattered redundant non-conductive regions in the stator slot to obtain the cooling medium flow boundary with the maximum connected cross-sectional area and the optimal hydraulic diameter.

[0065] The system first establishes a calculation model for the redundancy of the transverse gap. In this computational model, This represents the lateral gap redundancy generated by the k-th wiring layer. This represents the maximum available physical slot width of the stator slot at the corresponding position in the kth layer. The inherent mechanical width of the flat conductor; The thickness of the insulating layer surrounding the conductor.

[0066] The system further defines the winding side displacement constraint equations based on the geometric center of the cross section. ,in The translation coordinate parameter is the geometric center of the k-th winding section relative to the longitudinal symmetry center plane of the stator slot; To control the sign function of the bias direction, the system executes the logical judgment of the sign function to drive all wiring layer windings belonging to the slot area to converge and approach the left or right inner wall.

[0067] After completing the above translation calculations, the system extracts the continuous fluid cavity formed between the opposite side of the bias winding and the slot wall insulation paper, and constructs a three-dimensional spatial integral equation. To reconstruct the volume of the oil cooling channel, in the formula... The final three-dimensional total volume of the asymmetric high-frequency oil-cooled microchannel is given; L is the effective axial length of the stator core; m is the total number of wiring layers that define the thin-section flat wire region of the stator slot. The target cross-sectional thickness of the k-th layer extracted from the configuration sequence.

[0068] The introduction of the above algorithm model and the improvement of its single-sided aggregation offset logic, without occupying additional effective electromagnetic volume of the motor and without changing the original main cooling water jacket structure, directly exposes the area of ​​thin conductors in the slot where the skin effect heats up most densely under high-frequency conditions to the physical scouring path of the cooling medium. This solves the problem of heat accumulation in the stator slots and rapid aging and degradation of local insulation caused by local leakage magnetic field distortion at extremely high speeds in high-frequency motors. By using space substitution, the thermal balance limit and continuous peak output power boundary of the motor stator assembly are significantly improved.

[0069] Example 3: The present invention also provides a motor flat wire winding design system with high slot fill factor and low AC loss, comprising: Memory, used to store computer programs; A processor is a step in implementing a design methodology when executing a computer program.

[0070] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of a design method.

[0071] A motor stator assembly is manufactured using winding wiring process data output from a design method; The motor stator assembly includes a stator core and multiple layers of flat wire windings passing through the stator slots. The cross-sectional thickness of the flat wires near the slot opening is less than that near the bottom of the slot, and the top of all flat wire winding units is in a non-equal thickness state across layers, while the two pins at their end welded connections are in a cross-layer equal thickness butt joint state.

[0072] In a preferred embodiment of the present invention, an asymmetrical gap is provided between the thin-section flat wire sidewall of the stator slot opening and the inner wall of the stator slot, with one side tightly attached to the insulating paper and the other side left open. The asymmetrical gap extends through the axial direction of the stator core, forming a built-in high-frequency oil cooling microchannel.

[0073] It should be noted that the high slot fill factor and low AC loss motor flat wire winding design system and computer-readable storage medium constitute the underlying physical computing platform and program execution carrier built by the present invention based on the above algorithm model. The design system includes a processor and a memory that are coupled together. The memory adopts physical devices such as non-volatile solid-state storage media or flash memory chips to persistently store the underlying machine instructions and underlying parameter matrices of the electromagnetic simulation model, the ladder section allocation algorithm model, and the end topology reconstruction algorithm model.

[0074] The processor employs an industrial-grade programmable gate array processor with floating-point arithmetic and multi-threaded parallel processing capabilities. By calling and executing computer programs stored in memory, it transforms abstract stator structure parameters and leakage magnetic field partial differential equations into winding wiring process data readable by automated equipment. By utilizing the high-speed computing performance of integrated circuits, it overcomes the data iterative calculation obstacles faced by analyzing the multi-layer electromagnetic physical fields inside complex motors and optimizing the nonlinearity of end pins. It provides the necessary computing resources and hardware support for implementing customized solutions for cross-layer non-equal thickness winding configurations, ensuring that the theoretically optimal ladder section allocation scheme is accurately and without omission mapped to real automated production line control instructions.

[0075] The motor stator assembly manufactured using the winding wiring process data output by the designed method is a hardware execution carrier that transforms the aforementioned electromagnetic optimization configuration and geometric topology reconstruction parameters into a three-dimensional solid electromagnetic energy conversion unit. The core material basis of the motor stator assembly consists of a stator core and multi-layer flat wire windings passing through the stator slots. The stator core is made of non-oriented silicon steel sheet material, which is stamped and stacked to further suppress eddy current losses induced by the high-frequency rotating magnetic field inside the core.

[0076] The multilayer flat wire winding uses high-conductivity rectangular cross-section oxygen-free copper wire coated with a high-temperature insulating varnish film of polyimide. Inside this component, the thickness of the flat wire cross-section near the slot opening on the inner diameter side of the stator core is less than that near the slot bottom on the outer diameter side of the stator core. The principle of this non-uniform cross-section distribution structure comes from the spatial decoupling of the leakage flux density in the stator slot. This allows conductors with smaller cross-sectional areas to be arranged in the strong leakage flux region to cut off the alternating electromagnetic induction loop of the skin effect and proximity effect, while conductors with larger cross-sectional areas are retained in the weak leakage flux region to maintain a lower system DC resistance. This completely eliminates the technical contradiction between the surge in local AC losses and the decrease in the overall slot fill factor of the stator under high-frequency operating conditions from the physical hardware level, giving the motor stator assembly comprehensive electromagnetic performance that balances low-frequency high efficiency and high-frequency high-power output.

[0077] In the end structure of the motor stator assembly, the top solid form of all flat wire winding units is in a cross-layer non-uniform thickness state, while the two pins at their end welding connection are in a cross-layer uniform thickness docking state. This structural feature is an absolute physical hardware reproduction of the preceding end topology reconstruction algorithm model. The cross-layer non-uniform thickness state represents that the same continuous integrally formed variable cross-section U-shaped flat copper wire has different cross-sectional thicknesses on both sides of its straight slot section according to different stator slot radial depths, resulting in an asymmetrical structure with thickness differences at both ends of the bending transition section extending out of the slot at the top. The cross-layer uniform thickness docking state represents that by breaking the standard fixed pitch regular pin arrangement, two pins that should have different thicknesses due to layered arrangement are made to have precisely equal cross-sectional physical dimensions at the final convergence guide or series welding node through asymmetrical spatial deflection and re-addressing. The realization of this physical connection state completely eliminates process defects such as focal misalignment, non-uniform collapse of the molten pool, and cold welding defects caused by laser alignment welding of thick and thin pins. While retaining the electromagnetic loss reduction characteristics of the variable cross-section in the slot, it achieves a high degree of process adaptability of the irregular winding end connection structure to the existing standardized automated laser welding production line.

[0078] The asymmetric high-frequency oil-cooled microchannels inside the stator slots constitute a physical structure for the active high-frequency heat dissipation of the motor stator assembly in the dimension of fluid thermodynamics. Between the thin-section flat wire sidewalls at the stator slot openings and the inner wall of the stator slots, a continuous asymmetric gap is formed by a unilateral physical offset compression action during mechanical assembly, with one side tightly attached to the aramid composite insulating paper material and the other side left open. This asymmetric gap runs through the stator core axially, creating a fluid scouring cavity directly facing the high-frequency heat source inside the stator slots. The principle behind this built-in microchannel is based on the lateral redundancy within the slots derived from the forced thinning of the slot opening conductors to suppress eddy currents in the stepped cross-section distribution logic. This allows for the introduction of automatic transmission-specific cooling oil or insulating coolant as the heat exchange medium without expanding the original slot area of ​​the stator laminations. The contribution of this microchannel structure lies in directly exposing the conductor in the slot region of the stator assembly, where heat generation is most intense due to the skin effect and leakage flux distortion, to the cooling medium. By shortening the heat conduction path from the insulation layer to the fluid boundary and expanding the direct convection heat transfer area, a high degree of reuse and integration of the electromagnetic loss reduction region and the fluid heat dissipation channel is achieved within the same confined physical space. This improvement in physical structure further reduces the extreme value of thermal equilibrium temperature rise inside the motor, significantly extending the continuous operating time and safety margin of the motor stator assembly under extreme high-frequency and high-speed conditions.

[0079] Based on the preferred embodiments of the present invention described above, those skilled in the art can make various changes and modifications without departing from the inventive concept. The technical scope of this invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

Claims

1. A method for designing flat wire windings for motors with high slot fill factor and low AC loss, characterized in that, Includes the following steps: S1: Obtain the stator parameters and high-frequency operating condition parameters of the target motor, and construct an electromagnetic simulation model of the stator core containing multiple stator slots; S2: Analyze the leakage magnetic field density gradient curve distributed radially in the stator slot. Based on the leakage magnetic field density gradient curve, generate the configuration sequence of the thickness of the stepped cross-section of the multi-layer flat wire in the stator slot through the stepped cross-section allocation algorithm model. The configuration sequence satisfies the following condition: the thickness of the flat wire cross-section near the groove opening layer is less than the thickness of the flat wire cross-section near the bottom of the groove. S3: Generate a U-shaped winding unit with a first span and a second span. The two ends of the winding unit are respectively positioned in different wiring layers determined based on the first span and the second span, and the two ends respectively have a target cross-sectional thickness corresponding to the wiring layer in the configuration sequence. S4: Calculate the deflection matrix of each layer of flat wire pins at the motor end using the equal cross-section end topology reconstruction algorithm model, so that any two pins that need to be connected by welding at the motor end can meet the equal cross-section matching condition of equal wiring layer thickness. S5: Outputs winding unit spatial configuration and deflection matrix winding wiring process data.

2. The method for designing a flat wire winding for a motor with high slot fill factor and low AC loss according to claim 1, characterized in that: In step S2, the step of executing the tiered section allocation algorithm model to generate the configuration sequence specifically includes: Construct a total loss assessment objective function model that includes DC copper loss and high-frequency AC loss; The stator slots are divided into strong leakage magnetic field region and weak leakage magnetic field region. Under high frequency operating conditions, the skin effect and proximity effect loss coefficients induced by the leakage magnetic field density gradient curve in each layer are calculated. Under the constraint of maintaining the total stator slot fill factor not lower than a preset threshold, the optimal cross-sectional thickness of each wiring layer is dynamically solved by minimizing the total loss evaluation objective function model, and a configuration sequence is generated. The configuration sequence exhibits a thickness gradient distribution characteristic where the thickness of the wiring layer closest to the stator slot opening increases layer by layer or remains partially unchanged from the wiring layer closest to the slot bottom.

3. The method for designing a flat wire winding for a motor with high slot fill factor and low AC loss according to claim 1, characterized in that: In step S3, the step of generating a winding unit with a first span and a second span specifically includes: The first span and the second span are configured with different span values ​​to form a non-full pitch winding arrangement with alternating long and short pitches, thereby using the short pitch torque effect to suppress low-frequency circulating currents between parallel branches. Based on the configuration sequence, continuously formed variable cross-section conductors are generated as winding units; In the top U-shaped bend area of ​​the winding unit, a geometric model of the cross-section gradient transition segment is generated, which allows the cross-section to smoothly transition from the target cross-section thickness at one end to the target cross-section thickness at the other end.

4. The method for designing a flat wire winding for a motor with high slot fill factor and low AC loss according to claim 1, characterized in that: In step S4, the step of executing the end topology reconstruction algorithm model based on the equal cross section specifically includes: Construct an initial connection adjacency matrix that represents the mapping relationship between the interconnections of all winding pins at the motor end; In the traversal optimization of the initial connection adjacency matrix, an equal cross-section matching constraint model is incorporated: it is determined whether the cross-sectional thickness of the wiring layer where the two pins to be connected in the initial mapping are located is equal. If the determination result is that there is a difference in cross-sectional thickness, the variable step size torsion adjustment logic algorithm is triggered to break the fixed pitch of the first span or the second span, generate a non-linear pin deflection angle fine adjustment amount to reconstruct the connection topology path, until a pin that meets the equal cross-sectional matching condition is found to connect the guide section or series section. The set of all pin deflection angles that have been reconstructed and verified is output as the deflection matrix.

5. The method for designing a flat wire winding for a motor with high slot fill factor and low AC loss according to claim 1, characterized in that: The design method further includes S5: generating asymmetric high-frequency oil-cooled microchannels, specifically: When each wiring layer uses a configuration sequence, calculate the redundancy of the lateral gaps relative to the maximum available slot width of the stator slot; The algorithm model of spatial rearrangement in the stator slot is used to shift the redundancy generated near the slot opening layer to the same side of the stator slot. In the thin-section flat wire region of the slot, the redundancy after offset is used to generate a three-dimensional model of the high-frequency oil-cooled microchannel with an asymmetrical distribution, and the three-dimensional model data is attached to the winding wiring process data.

6. The method for designing a flat wire winding for a motor with high slot fill factor and low AC loss according to claim 1, characterized in that: In S5, the output winding wiring process data is directly input to the control system of the automated robotic arm and laser welding equipment to guide the non-destructive insertion positioning and equal cross-section alignment welding process of the irregular cross-section flat wire.

7. A design system for flat wire windings of a motor with high slot fill factor and low AC loss, based on the design method described in any one of claims 1-6, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the steps of the design method as described in any one of claims 1 to 6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the design method as described in any one of claims 1 to 6.

9. A motor stator assembly, characterized in that, The motor stator assembly is manufactured using the winding wiring process data output by the design method described in any one of claims 1-6; The motor stator assembly includes a stator core and multiple layers of flat wire windings passing through the stator slots. The cross-sectional thickness of the flat wires near the slot opening is less than that near the bottom of the slot. The top of all flat wire winding units is in a non-equal thickness state across layers, and the two pins at their end welded connections are in a cross-layer equal thickness butt joint state.

10. The motor stator assembly according to claim 9, characterized in that, Between the thin-section flat wire sidewall of the stator slot opening and the inner wall of the stator slot, there is an asymmetrical gap with one side tightly attached to the insulating paper and the other side left open. The asymmetrical gap extends through the axial direction of the stator core, forming a built-in high-frequency oil cooling microchannel.

Images