Semi-direct-drive large-scale wind driven generator based on permanent magnet blocks and optimization method

Abstract

The invention provides a semi-direct-drive large-scale wind driven generator based on permanent magnet blocks and an optimization method, and belongs to the technical field of semi-direct-drive large-scale permanent magnet synchronous wind driven generators, the semi-direct-drive large-scale wind driven generator comprises a rotor, the rotor comprises a plurality of permanent magnet units distributed in the circumferential direction, and each permanent magnet unit comprises a plurality of permanent magnet blocks; grooves are formed in the inner layer and the outer layer, distributed in the circumferential direction, of the permanent magnet block, the grooves are evenly distributed in the surface of the permanent magnet block in the radial direction, and the grooves formed in the inner layer and the outer layer are arranged alternately.

Description

Technical Field

[0001] The present invention belongs to the technical field of semi-direct-driven large-scale permanent magnet synchronous wind turbine generators, and in particular relates to a semi-direct-driven large-scale wind turbine generator based on permanent magnet segmentation and an optimization method. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] As the core component of the wind power system, the performance of large permanent magnet synchronous wind turbines has an important impact on the operating efficiency and reliability of wind turbines. In the existing technology, the semi-direct drive structure has the advantages of both direct drive and gear transmission systems. It reduces the transmission links and achieves efficient and reliable energy transmission. It is widely used in high-power wind turbines. However, with the continuous improvement of the power level and size of the generator, the loss problem during its operation has become more and more prominent, becoming an important bottleneck restricting the improvement of generator efficiency.

[0004] Permanent magnets are key components of permanent magnet synchronous wind turbines, and their design plays a decisive role in the electromagnetic performance and operating efficiency of the generator. The traditional monolithic permanent magnet design is prone to produce large permanent magnet eddy current losses, resulting in serious heating problems and increasing the risk of demagnetization of the permanent magnet. The block permanent magnet can effectively reduce the eddy current loss of the permanent magnet by dividing the permanent magnet into several small pieces, but since the permanent magnet needs to be insulated and re-tightened, this will increase the manufacturing cost and reduce the mechanical robustness of the permanent magnet.

[0005] For example, in the prior art, CN118783672A discloses an axial magnetic field permanent magnet motor with a block-type permanent magnet and an assembly method thereof, each permanent magnet structure includes a plurality of permanent magnet blocks, and each permanent magnet block is in contact with a magnetic conductive adapter plate. The above scheme reduces the magnetization and assembly difficulties of the motor with a block-type permanent magnet, but the following problems still exist: Due to incomplete block division, the loss is concentrated in some areas of the permanent magnet, which causes the permanent magnet to overheat under some conditions, resulting in local demagnetization, thereby changing the air gap flux density and reducing the motor efficiency. Summary of the invention

[0006] To overcome the above-mentioned deficiencies of the prior art, the present invention provides a semi-direct-drive large wind turbine generator based on permanent magnet block. Compared with the traditional integral permanent magnet, the eddy current loss of the permanent magnet of the generator is improved, thereby further improving the efficiency of the generator.

[0007] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: In a first aspect, a semi-direct-drive large-scale wind turbine based on permanent magnet block is disclosed, comprising: a rotor, the rotor comprising a plurality of permanent magnet units distributed along the circumferential direction, each of the permanent magnet units comprising a plurality of permanent magnet blocks; The inner layer and the outer layer of the permanent magnet block distributed along the circumferential direction are both provided with grooves, each groove is evenly distributed on the surface of the permanent magnet block in the radial direction, and the grooves provided in the inner layer and the outer layer are arranged alternately.

[0008] As a further technical solution, the wind turbine further comprises a rotating shaft, a rotor and a stator are coaxially arranged with the rotating shaft, and an axial air gap exists between the rotor and the adjacent stator; The stator is made of laminated silicon steel sheets, with evenly distributed slots punched on it, and three-phase symmetrical windings are placed in the slots, which are the stator windings; The slot type of the stator is an open slot, and a stator slot wedge is placed at the slot opening.

[0009] As a further technical solution, the stator winding is a double-layer winding, the rotor structure is built-in, and the permanent magnet unit is built into the rotor core.

[0010] As a further technical solution, the permanent magnet blocks are parallel to each other and perpendicular to the radial direction.

[0011] As a further technical solution, the number of slots and the number of poles of the stator are selected based on the greatest common divisor of the number of slots and the number of poles, and the least common multiple of the number of slots and the number of poles.

[0012] As a further technical solution, the built-in rotor, permanent magnets and stator core together constitute a magnetic circuit structure.

[0013] As a further technical solution, the three-phase double-layer integer-slot stacked winding is connected in a star shape, and the current flows out from the head ends of the three-phase double-layer integer-slot stacked winding respectively, and the ends of the three-phase double-layer integer-slot stacked winding are connected together.

[0014] In a second aspect, a method for assembling a semi-direct-drive large wind turbine based on permanent magnet segmentation is disclosed, comprising: The inner layer and the outer layer of the permanent magnet block distributed along the circumferential direction are both provided with grooves, each groove is evenly distributed on the surface of the permanent magnet block in the radial direction, and the grooves provided in the inner layer and the outer layer are arranged alternately.

[0015] One or more of the above technical solutions have the following beneficial effects: The stator part of the generator of the technical solution of the present invention is made of laminated silicon steel sheets to reduce the stator iron loss. It is punched with evenly distributed slots, and three-phase symmetrical windings are placed in the slots. The stator slot type adopts open slots, and stator slot wedges are placed at the ends of the slots. The stator winding is a double-layer stacked winding, and the rotor structure adopts an internal type. The surface of the permanent magnet is incompletely divided, and its division form adopts bilateral alternating division. Compared with the traditional whole permanent magnet, the eddy current loss of the permanent magnet of the generator is significantly improved, further improving the efficiency of the generator. The loss of this application is reduced due to the optimization of the structure, and the area where the loss is concentrated is reduced.

[0016] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0018] Figure 1 A partial three-dimensional structural diagram of a semi-direct-drive large permanent magnet synchronous wind turbine generator with bilateral alternating incompletely split permanent magnets according to Example 1 of the present invention; Figure 2 This is a structural diagram of a bilaterally alternating incompletely split permanent magnet according to Example 1 of the present invention; Figure 3 is the direction diagram of eddy current on permanent magnet; Figure 3 (a) is a schematic diagram of the simulation block method, Figure 3 (b) is a schematic diagram of the theoretical vortex direction; Figure 4 It is the distribution cloud diagram of eddy current density on the front side of the permanent magnet; Figure 5 It is the distribution cloud diagram of eddy current density on the opposite side of the permanent magnet; Figure 6 Assemble the flow chart for the motor; Figure 7 This is a flow chart of the overall optimization method for this embodiment; Figure 8 This is a schematic diagram of the particle swarm optimization algorithm of this embodiment; Explanation of the accompanying drawings: 1. stator; 2. winding coil; 3. permanent magnet unit; 4. rotor; 5. housing; 6. rotating shaft. DETAILED DESCRIPTION

[0019] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0020] It should be noted that the terms used herein are for describing specific embodiments only and are not intended to be limiting of exemplary embodiments according to the present invention.

[0021] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0022] Embodiment 1 like Figure 1 As shown, this embodiment discloses a semi-direct-drive large wind turbine based on permanent magnet segmentation, including: a stator 1, a winding coil 2, a permanent magnet unit 3, a rotor 4, and a rotating shaft 6, wherein the rotor 4 and the stator 1 are coaxially arranged with the rotating shaft, the winding coil is nested in the stator teeth, the casing 5 is connected to the outside of the stator, wraps the other parts of the motor, is the outermost structure of the motor, and has heat dissipation ribs on the outside of the casing to enhance the heat dissipation capacity of the motor.

[0023] In this implementation example, the concept of a semi-direct drive generator is a solution to the problems encountered in the development of direct drive generators and doubly fed generators. It not only reduces the transmission ratio of the gearbox on the basis of the doubly fed model, but also improves the speed and compactness of the structure on the basis of the direct drive transmission model, thereby improving the reliability and service life of the gearbox. On the one hand, the size and manufacturing cost of the generator are reduced, thereby reducing the overall weight of the cabin and reducing the pressure on the tower; on the other hand, the requirements for the speed increase mechanism are greatly reduced, making the gearbox with the highest failure rate more reliable and more stable.

[0024] There is an axial air gap between the rotor 4 and the adjacent stator 1. The rotor includes a plurality of permanent magnet units distributed along the circumferential direction. Each permanent magnet unit includes three permanent magnet blocks. The permanent magnet blocks are parallel to each other and radially perpendicular to the plane where the rotor is located. The rotor adopts a built-in type, that is, the permanent magnet unit is built into the rotor.

[0025] The built-in rotor, permanent magnet unit and stator together constitute a magnetic circuit structure. The magnetic flux lines are emitted by the permanent magnet unit, pass through the rotor to reach the air gap between the rotor and the stator, and further pass through the air gap to reach the stator. In the stator, a back electromotive force is generated by cutting the copper wire in the winding, and finally returns to the built-in permanent magnet unit in the rotor through the air gap.

[0026] The permanent magnet unit adopts incomplete segmentation with alternating bilateral sides on one side, that is, rectangular slots are adopted close to the edge of one surface of the permanent magnet, each slot is evenly distributed on the surface of the permanent magnet in the radial direction, and the slots are alternately made on both sides of the surface. Finally, the permanent magnet block is incompletely segmented, and the permanent magnet block is arranged according to the unit.

[0027] The three-phase double-layer integer-slot stacked windings are connected in a star shape, and the current flows out from the head ends of the three-phase double-layer integer-slot stacked windings respectively, and the ends of the three-phase double-layer integer-slot stacked windings are connected together.

[0028] The working principle of the above motor: As the rotor rotates, the permanent magnet unit emits magnetic lines of force, which pass through the rotor to reach the air gap between the rotor and the stator, and then pass through the air gap to reach the stator. In the stator, a back electromotive force is generated by cutting the copper wire in the winding, which is then output to the energy storage device. The magnetic lines of force finally pass through the air gap back to the permanent magnet unit built into the rotor.

[0029] A method for dividing a permanent magnet by alternating incomplete segments on both sides, i.e., a rectangular slot is used close to the edge of one surface of a permanent magnet unit, each slot is evenly distributed on the surface of the permanent magnet unit in the radial direction, and the slots are alternately made on both sides of the surface. The above-mentioned means in this embodiment are different from the conventional complete block method, and the incomplete block has great requirements on the precision and strength of the cutting equipment. This incomplete block makes the mechanical stability of the permanent magnet higher, significantly reduces the eddy current loss of the permanent magnet, and reduces the occurrence of motor accidents.

[0030] The stator part of the generator is made of laminated silicon steel sheets to reduce the stator iron loss. It is punched with evenly distributed slots, and three-phase symmetrical windings are placed in the slots. The stator slot type adopts open slots, and stator slot wedges are placed at the ends of the slots. The stator winding is a double-layer stacked winding, and the rotor structure adopts an internal type. The surface of the permanent magnet is incompletely divided, and its division form adopts bilateral alternating division. Compared with the traditional whole permanent magnet, the eddy current loss of the permanent magnet of the generator is significantly improved, further improving the efficiency of the generator.

[0031] A large permanent magnet synchronous wind turbine using a bilateral alternating incompletely divided permanent magnet segmentation method has a lower eddy current loss of the permanent magnet than traditional non-divided permanent magnets, and has the characteristic of maintaining the mechanical stability of the permanent magnets compared to completely divided permanent magnets.

[0032] The bilaterally alternating incompletely segmented permanent magnet segmentation solution provided in this embodiment can effectively reduce the eddy current loss of the permanent magnet and improve the efficiency of the generator.

[0033] The present embodiment provides a large permanent magnet synchronous wind turbine generator with a bilateral alternating incompletely segmented permanent magnet segmentation method, which adopts a three-phase integer slot double-layer winding structure to minimize the winding ends, reduce the end copper loss, reduce the heat generation of the stator winding, and to a certain extent reduce the overall volume of the motor and improve the integration.

[0034] Specifically, Figure 2 As shown, the bilaterally alternating incompletely segmented permanent magnet blocks have rectangular slots on one surface of the permanent magnet close to the edge, each slot is evenly distributed on the permanent magnet surface in the radial direction, and the two sides of the surface are slotted alternately.

[0035] like Figure 3 As shown, the bilaterally alternating incompletely segmented permanent magnet blocks can change the direction of the permanent magnet eddy current from the cutting point, extending the path of the eddy current loss. The current density after the change of direction will be greatly reduced, thereby achieving the purpose of weakening the permanent magnet eddy current loss and improving the motor efficiency.

[0036] like Figure 4 and Figure 5 As shown, Ohmic-Loss corresponds to Chinese: permanent magnet eddy current loss, W / m^3 is the permanent magnet eddy current loss density, representing how much loss there is per cubic meter; in the cloud map of permanent magnet eddy current loss in the motor simulation of bilateral alternating incompletely divided permanent magnet blocks, it can be seen that the permanent magnet eddy current loss is mainly at the edges around the permanent magnet. Since the division extends the path of the eddy current loss, the permanent magnet eddy current loss is weakened to a certain extent. At the incomplete division, due to the dense number of edges, the permanent magnet eddy current loss is relatively large here.

[0037] The stator part of the generator is made of laminated silicon steel sheets to reduce the stator iron loss. It is punched with evenly distributed slots, and three-phase symmetrical windings are placed in the slots. The stator slot type adopts open slots, and stator slot wedges are placed at the slots. The stator winding is a double-layer winding, and the rotor structure adopts an internal type. The surface of the permanent magnet is not completely divided, and its division form adopts bilateral alternating division, that is, rectangular slots are used on one surface of the permanent magnet close to the edge, and each slot is evenly distributed on the surface of the permanent magnet in the radial direction, and the two sides of the surface are alternately slotted. Compared with the traditional integral permanent magnet, the eddy current loss of the permanent magnet of the generator is improved, further improving the efficiency of the generator.

[0038] Embodiment 2 The purpose of this embodiment is to provide an assembly method for a semi-direct-drive large wind turbine based on permanent magnet segmentation. Different from other permanent magnet assembly methods, this method uses a fixed mold to fix the permanent magnet, so that the permanent magnet completes the steps of magnetization and incomplete segmentation.

[0039] The specific steps include: Figure 6, the assembly method mainly includes: S1: Install several permanent magnet units on the workbench.

[0040] Exemplarily, first, all permanent magnet blocks used to form the same permanent magnet unit are installed on a workbench. During installation, the permanent magnet blocks can be installed on the workbench in a nested external mold manner.

[0041] S2: Magnetize the magnetic pole unit.

[0042] Exemplarily, the assembled permanent magnet unit is placed on a magnetizer to be magnetized.

[0043] Furthermore, in this step, the permanent magnet unit is placed in a strong magnetic field environment. Subsequently, the magnetic domain structure inside the magnetic material is reoriented by adjusting the strength and direction of the magnetic field. The corresponding magnetic field strength and magnetization time are accurately selected according to the size, shape and required magnetic performance index of the magnetic material. During the magnetization process, the strength and direction of the magnetic field must be strictly controlled to avoid damage to the magnetic performance of the magnetic material.

[0044] S3: Incompletely splitting the permanent magnet blocks after magnetization.

[0045] Exemplarily, the permanent magnet block after magnetization is placed in a cutting device, and a cutter is used to cut the permanent magnet block according to the following steps: Figure 3 The model shown is a double-sided incomplete segmentation model that physically cuts the entire permanent magnet.

[0046] S4: Install the magnetic pole unit to the rotor core to form a rotor.

[0047] Exemplarily, the magnetic pole unit can be fixedly mounted on the rotor by matching the nested mold in the magnetized magnetic pole unit with the built-in rotor.

[0048] Embodiment 3 The purpose of this embodiment is to provide an optimization method for a semi-direct-drive large wind turbine based on permanent magnet segmentation. Normally, permanent magnet segmentation adopts an average segmentation method. This method has mechanical deviations in processing. In reality, the eddy current loss density of permanent magnets in some areas is higher than that in other parts, resulting in excessive regional temperature, causing demagnetization in some areas, and causing non-uniformity of the permanent magnet magnetic field. Different from the average segmentation permanent magnet segmentation optimization method, this method adopts a particle swarm optimization algorithm. This optimization algorithm can obtain the optimal segmentation method through continuous iterative calculation. Under this segmentation method, it can ensure that the eddy current loss of the permanent magnet reaches the optimal distribution when segmenting, reducing the risk of demagnetization in some areas due to excessive permanent magnet temperature. In the permanent magnet segmentation design of the particle swarm optimization algorithm, exhibitors Figure 8, transform the physical problem into a mathematical optimization model, clarify the design variables, objective function and constraints, and obtain the optimal solution after multiple iterations.

[0049] The specific steps include: Figure 7 , the particle swarm optimization algorithm mainly includes: S1: The length, width and thickness of each magnetic block are taken as geometric parameters, and the number of blocks N is taken as a discrete variable; each sub-block is regarded as a particle.

[0050] S2: Establish magnetic field simulation model: (1) Construct a three-dimensional permanent magnet motor model; (2) Add excitation, boundary conditions, and meshing; (3) Set the step size and final time for calculation.

[0051] S3: Configure the particle swarm optimization algorithm: Add geometric constraints to ensure that particles always stay within the domain and do not exceed the boundaries: Add physical constraints to determine the size of the permanent magnet and connect it to engineering practice: Particle encoding and initialization: The size, position, and magnetization direction of each sub-block are normalized to the interval [0, 1].

[0052] S4: Iterative Optimization: The particle swarm optimization algorithm is used to update the speed of each logical instance in each particle in the current particle swarm.

[0053] In this implementation example, specifically, S1: the length, width, thickness, etc. of each magnetic block are used as geometric parameters; the number of blocks is variable, and the number N is used as a discrete variable. Each sub-block is regarded as a particle, and the size of each sub-block is: length ,Width w i ,high h i ; Sub-block position coordinates: x i , y i , z i , with the center of the cuboid as the origin. The magnetization angle of each sub-block is: θ i , azimuth .

[0054] Divide the cuboid into N sub-blocks, and the parameters of each sub-block are: x n =[ , w 1, h1, x 1, y 1, z 1, θ 1, … ，w n ， h n ，x n ，y n ，z n ,θ n ， ].

[0055] In this implementation example, specifically, S2: establishing a magnetic field simulation model: (1) Construct a 3D permanent magnet motor model in Maxwell; (2) Add excitation, boundary conditions, and meshing; (3) Set the step size and final time for calculation.

[0056] In this implementation example, specifically, S3: configuring the particle swarm optimization algorithm: It is necessary to quantify the magnetic field uniformity and energy conversion efficiency and combine them into a single objective function through weighting: ; in, α、β is the weight coefficient, Uniformity ( x ) is the magnetic field uniformity function, Efficiency ( x ) is the energy conversion efficiency function, f ( x ) is the weighted objective function.

[0057] It should be noted that the weight coefficient α、β Each iteration is different, and the optimal coefficient is found in multiple iterations. α、β The value range is [0, 1].

[0058] Magnetic field homogeneity ( Uniformity ) is quantified using the magnetic field standard deviation or maximum deviation. For example, in the target area, M Sampling points, calculate the magnetic induction intensity B k Variance of: ; in, Uniformity ( x) is the magnetic field uniformity function; is the average magnetic induction intensity, B k is the air gap flux density, M =1, 2, 3….

[0059] The energy conversion efficiency is defined by the ratio of actual output power to theoretical maximum power ( Efficiency ): ; in, Efficiency ( x ) is the energy conversion efficiency function; B is the air gap magnetic density; B max is the maximum air gap flux density; H is the magnetic field strength; H max is the maximum magnetic field strength, V For volume.

[0060] Add geometric constraints to ensure that particles always stay within the domain and do not exceed the boundaries: Sub-blocks cannot overlap: for any two sub-blocks i , j ,satisfy . Total volume is constant: .

[0061] Among the above parameters, x i For the i Block sub-block; x j For the j Block sub-block; For the i The length of the block sub-block; For the j The length of the block sub-block; N= 1, 2, 3…; w i For the i The width of the block sub-block; h i For the i The height of the block sub-block; V total is the total volume of the permanent magnet.

[0062] Then add physical constraints to determine the size of the permanent magnet and link it with engineering practice: Minimum size of the magnet block: . For the i The length of the block sub-block; w i For the i The width of the block sub-block;h i For the i The height of the block sub-block; For the i Minimum length of a block sub-block; w min For the i The minimum width of a block subblock; h min For the i Minimum height of a block's sub-blocks.

[0063] The particles are then encoded and initialized: the size, position, and magnetization direction of each sub-block are normalized to the [0, 1] interval. x n =[ , w 1, h 1, x 1, y 1, z 1, θ 1, … ，w n ，h n ，x n ，y n ，z n ,θ n ， ].

[0064] The actual size above is , the actual location is ; Randomly generate particle positions to ensure that the sub-blocks do not overlap and the total volume is constant, and the initial velocity is set to v i =0 or randomly generated in [-0.1, 0.1]. For the i The length of the block sub-block; L max is the maximum length of the sub-block; L min is the minimum length of a sub-block; is the normalized length after normalization, and the formula is ; x i For the i The location of the block sub-blocks; L max is the maximum length of the permanent magnet; L min is the minimum length of the permanent magnet; is the normalized position after normalization, and the formula is ; x max is the maximum position; x min is the minimum position.

[0065] In this implementation example, specifically, S4: iterative optimization: Using the particle swarm optimization algorithm, the speed of each logical instance in each particle in the current particle swarm is updated. The speed update formula is: ; in ，t is the number of iterations; i =1, 2, ... are the particle numbers, j is the particle dimension; w is the inertia weight, which controls the global and local balance of the search steps; c 1, c 2 is the learning factor, c 1Control the influence of individual optimality, c 2. Control the impact of global optimality; v ij For particles i history j The speed of generation time; x ij For particles i history j The position of the generation; p best，ij For particles i The best position in history j Coordinates of the generation time; g best，j is the global best position; r 1. r 2 is a random number, which increases the randomness of the algorithm and avoids falling into the local optimum.

[0066] After the speed and position of each particle in the particle swarm are updated, the value of each particle is calculated again according to the updated speed and position using the weighted objective function, and the value of each particle is compared with the value corresponding to the updated and stored individual optimal value. If the value of the particle is larger, the value corresponding to the particle and the speed and position of all logical instances replace the updated and stored individual optimal value of the particle. After the individual optimal values ​​of all particles are updated, the values ​​corresponding to the individual optimal values ​​of all particles are compared with the updated and stored global optimal value. If the value corresponding to the individual optimal value of a particle in the particle swarm is larger, the updated and stored global optimal value is replaced, thereby realizing the update of the global optimal value of the particle swarm, until the termination condition is reached, and the logical instance block mode in the particle corresponding to the current global optimal value is used as the final permanent magnet block result.

[0067] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. A semi-direct-drive large wind turbine based on permanent magnet block, characterized in that: include: A rotor, wherein the rotor comprises a plurality of permanent magnet units distributed along the circumferential direction, and each of the permanent magnet units comprises a plurality of permanent magnet blocks; The inner layer and the outer layer of the permanent magnet block distributed along the circumferential direction are both provided with grooves, each groove is evenly distributed on the surface of the permanent magnet block in the radial direction, and the grooves provided in the inner layer and the outer layer are arranged alternately.

2. The semi-direct drive large wind turbine generator based on permanent magnet segmentation as claimed in claim 1 is characterized in that: The wind turbine generator further comprises a rotating shaft, a rotor and a stator are coaxially arranged with the rotating shaft, and an axial air gap exists between the rotor and the adjacent stator; The stator is made of laminated silicon steel sheets, and is punched with evenly distributed slots. Three-phase symmetrical windings are placed in the slots, which are stator windings.

3. The semi-direct-drive large wind turbine generator based on permanent magnet segmentation as claimed in claim 2 is characterized in that: The slot type of the stator is an open slot, and a stator slot wedge is placed at the slot opening; The stator winding is a double-layer winding, the rotor structure is built-in, and the permanent magnet unit is built-in the rotor; The permanent magnet blocks are parallel to each other and perpendicular to the radial direction.

4. The optimization method for a semi-direct-drive large wind turbine generator based on permanent magnet segmentation according to any one of claims 1 to 3, characterized in that: include: S1: The length, width and thickness of each magnetic block are taken as geometric parameters, and the number of blocks N is taken as a discrete variable; Treat each sub-block as a particle; S2: Establish a magnetic field simulation model; (1) Construct a three-dimensional permanent magnet motor model; (2) Add excitation, boundary conditions, and meshing; (3) Set the step size and final time for calculation; S3: Configure particle swarm optimization algorithm; Add geometric constraints to ensure that particles always stay within the domain and do not exceed the boundaries: Add physical constraints to determine the size of the permanent magnet and connect it to engineering practice: Particle encoding and initialization: Normalize the size, position, and magnetization direction of each sub-block to the interval [0, 1]; S4: Iterative optimization; Using the particle swarm optimization algorithm, the speed of each logical instance in each particle in the current particle swarm is updated; After the speed and position of each particle in the particle swarm are updated, the value of each particle is calculated again using the weighted objective function according to the updated speed and position, and the value of each particle is compared with the value corresponding to the updated and stored individual optimal value. If the value of the particle is larger, the value corresponding to the particle and the speed and position of all logical instances replace the updated and stored individual optimal value of the particle.

5. The optimization method of the semi-direct-drive large wind turbine generator based on permanent magnet segmentation as claimed in claim 4 is characterized in that: Add geometric constraints to ensure that particles always stay within the domain and do not exceed the boundaries: Sub-blocks cannot overlap; Add physical constraints to determine the size of the permanent magnet and connect it to engineering reality.

6. The optimization method of the semi-direct-drive large wind turbine generator based on permanent magnet segmentation as claimed in claim 4 is characterized in that: After the individual optimal values ​​of all particles are updated, the values ​​corresponding to the individual optimal values ​​of all particles are compared with the updated and stored global optimal value. If the value corresponding to the individual optimal value of a particle in the particle swarm is larger, the updated and stored global optimal value is replaced, thereby realizing the global optimal value update of the particle swarm until the termination condition is reached. The logical instance partitioning method in the particle corresponding to the current global optimal value is used as the final permanent magnet partitioning result.

7. The optimization method of the semi-direct drive large wind turbine generator based on permanent magnet segmentation as claimed in claim 4 is characterized in that the speed The update formula is: t is the number of iterations; i =1, 2, ... are the particle numbers, j is the particle dimension; w is the inertia weight, which controls the global and local balance of the search steps; c 1, c 2 is the learning factor, c 1Control the influence of individual optimality, c 2. Control the impact of global optimality; v ij For particles i history j The speed of generation time; x ij For particles i history j The position of the generation; p best，ij For particles i The best position in history j Coordinates of the generation time; g best，j is the global best position; r 1. r 2 is a random number, which increases the randomness of the algorithm and avoids falling into the local optimum.

8. The optimization method of the semi-direct drive large wind turbine generator based on permanent magnet segmentation as claimed in claim 4 is characterized in that: In the particle swarm optimization algorithm, the magnetic field uniformity and energy conversion efficiency are quantified and combined into a single objective function through weighting: α、β is the weight coefficient, Uniformity ( x ) is the magnetic field uniformity function, Efficiency ( x ) is the energy conversion efficiency function, f ( x ) is the weighted objective function.

9. The optimization method of the semi-direct-drive large-scale wind turbine generator based on permanent magnet segmentation as claimed in claim 8, characterized in that: Magnetic field uniformity function Uniformity ( x ) Use magnetic field standard deviation or maximum deviation quantification, set in the target area M Sampling points, calculate the magnetic induction intensity B k Variance of: in, Uniformity ( x ) is the magnetic field uniformity function; is the average magnetic induction intensity, B k is the air gap flux density, M =1, 2, 3….

10. The optimization method of a semi-direct-drive large-scale wind turbine generator based on permanent magnet segmentation as claimed in claim 8, characterized in that: The energy conversion efficiency function is defined by the ratio of actual output power to theoretical maximum power. Efficiency : in, Efficiency ( x ) is the energy conversion efficiency function; B is the air gap magnetic density; B max is the maximum air gap magnetic flux density; H is the magnetic field strength; H max is the maximum magnetic field strength, V For volume.

Images