A hub motor temperature rise analysis method based on a co-node magnetic heat bidirectional coupling network model

Abstract

The application discloses a kind of based on common node magnetic heat two-way coupling network model's wheel hub motor temperature rise analysis method, comprising the following steps: step 1, establish local equivalent magnetic network model;Step 2, establish global transient equivalent magnetic network model;Step 3, solve global equivalent magnetic network model, solve each node magnetic potential;Step 4, establish and solve global equivalent thermal network model;Step 5, coupling global equivalent magnetic network model and global equivalent thermal network model;The application fully considers the characteristics of wheel hub motor highly integrated integration design, when calculating temperature field distribution, not only consider the internal heat source such as iron core loss and winding copper loss, but also consider many external heat sources such as the heat generated by braking and the heat generated by tire friction, improve the precision of permanent magnet wheel hub motor magnetic-thermal coupling analysis model under considering multiple heat sources.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of electromagnetic field calculation of electric machines, and relates to a wheel hub motor temperature rise analysis method based on a common node magnetic-thermal bidirectional coupling network model. BACKGROUND

[0002] In the field of new energy vehicles, permanent magnet wheel hub motors are widely used due to their high transmission efficiency, flexible control, compact structure and other advantages. Generally, permanent magnet wheel hub motors generally adopt high-integration integrated design, which not only needs to place electrical components such as motor stator and rotor, cooling system and controller, but also has conventional brake components such as brake disc and brake caliper. This results in multiple heat sources in the narrow wheel space of the permanent magnet wheel hub motor, and the heat transfer mechanism is complex. In addition, the working environment of the wheel hub motor is complex, and the operating conditions are variable, which will also have a significant impact on the performance of the wheel hub motor. Therefore, in the design process of high-performance wheel hub motors, it is necessary to accurately analyze the temperature rise distribution of the motor under the influence of multiple heat sources and the bidirectional coupling effect of the motor electromagnetic performance.

[0003] The current temperature rise analysis of the motor usually adopts a one-way calculation method of motor loss-temperature rise, that is, after the distribution characteristics of the motor core loss and permanent magnet eddy current loss are calculated by finite element or thermal circuit method, the temperature distribution characteristics of the motor are obtained. For example, the document with Chinese publication number CN110414074A discloses a motor temperature rise method based on an equivalent thermal network model, which can quickly calculate the temperature distribution in the motor. However, since this method ignores the influence of motor temperature rise on motor core loss, it cannot accurately analyze the coupling effect of motor thermal field and magnetic field, and the calculation accuracy needs to be further improved. The document with Chinese patent publication number CN106446364A discloses a motor thermal analysis method based on direct coupling of temperature field-thermal circuit based on finite element, which analyzes the temperature rise distribution of the motor based on finite element and thermal circuit method. However, since this method simplifies the external heat source by using thermal circuit method, it is difficult to effectively analyze the complex influence of the external heat source, and does not consider the actual complex heat source of the wheel hub motor, so it cannot be directly applied to the magnetic-thermal design of the wheel hub motor. SUMMARY

[0004] The purpose of the present application is to propose a wheel hub motor temperature rise analysis method based on a common node magnetic-thermal bidirectional coupling network model in view of the characteristics of complex heat generation mechanism and variable operating conditions of the wheel hub motor. This method can consider the magnetic-thermal coupling effect of the wheel hub motor under complex heat sources, and quickly and accurately obtain the temperature rise distribution of the motor under different operating conditions.

[0005] To achieve the above-mentioned purpose, the technical scheme adopted by the present application is as follows: a wheel hub motor temperature rise analysis method based on a common node magnetic-thermal bidirectional coupling network model, comprising the following steps:

[0006] Step 1, establish a local equivalent magnetic network model: based on the motor topology, the entire motor is divided into stator yoke, stator tooth, stator tooth shoe, air gap, permanent magnet and rotor, and each region is meshed; the overall structure of the motor is a circle, and the shape of each mesh element is a sector during layering; based on the magnetic field distribution of each component, the appropriate meshing accuracy is selected to shorten the calculation time while ensuring the calculation accuracy;

[0007] Step 2, establish a global transient equivalent magnetic network model: based on the local equivalent magnetic network model, each component is connected through nodes. Based on the meshing accuracy of each component, the adjacent two parts are connected in a one-to-one or one-to-many relationship between nodes, and the global steady-state equivalent magnetic network model is established; when the motor rotates, the connection between the stator tooth shoe and the air gap node is changed, that is, every time the motor rotates one degree, the air gap node is disconnected from the current node and connected to the node adjacent to the tooth shoe node in the rotation direction, and the global transient equivalent magnetic network model is established;

[0008] Step 3, solve the global equivalent magnetic network model to solve the magnetic potential of each node;

[0009] Step 4, establish and solve the global equivalent thermal network model;

[0010] Step 5, couple the global equivalent magnetic network model and the global equivalent thermal network model: first, use the equivalent magnetic network model to calculate the core loss at the initial temperature, and use this loss as one of the initial heat sources of the equivalent thermal network model; on this basis, add the winding copper loss and external heat source to calculate the temperature of each node, then recalculate the core loss at this temperature, repeat the above steps until the temperature of the permanent magnet meets the set convergence accuracy, that is, the temperature is considered to be the target temperature, and finally the electromagnetic performance and temperature distribution of the motor at this temperature are calculated.

[0011] Further, the motor is a 36-slot / 34-pole three-phase surface-mounted permanent magnet wheel hub motor, which includes a stator, a rotor, a permanent magnet and an armature winding, the stator and rotor cores are silicon steel sheets, the permanent magnet is a NdFeB ferrite-boron permanent magnet, and the armature winding adopts a concentrated winding method.

[0012] Further, in step 1, the stator tooth shoe is connected to the air gap, and the meshing accuracy of this part also has a great influence on the accuracy of the final calculation result, so the meshing accuracy of this part is consistent with that of the air gap part, the number of tangential meshing is 1224, and the radial is also divided into 2 layers; in addition to the stator tooth shoe, the tooth slot is also included in this part, and each slot of the stator corresponds to 34 tangential mesh elements, of which 4 are tooth slot mesh elements and 30 are tooth shoe mesh elements.

[0013] The permanent magnet is connected with the air gap, and the calculation accuracy has less influence on the final calculation result than the air gap part. In consideration of the calculation accuracy and the time consumption, the number of tangential grid division of the permanent magnet is selected as 306, the corresponding relationship between the number of grid of the permanent magnet and the air gap part is 1:4, the consistency of the grid can be realized, the thickness of the permanent magnet is 4mm, and therefore the permanent magnet is divided into 2 layers in the radial direction, and the sector-shaped magnetic conductance unit is approximately a square;

[0014] The rotor is connected with the permanent magnet. According to the modeling experience, the number of tangential grid division of the rotor part is the same as that of the permanent magnet, and is also selected as 306. The calculation accuracy can be maintained while the programming complexity is greatly reduced. The thickness of the rotor is 11mm, and therefore the rotor is divided into 3 layers in the radial direction. The sector-shaped magnetic conductance unit is also approximately a square.

[0015] Further, in step 2, when the transient equivalent magnetic network model is established, the node connection between the air gap and the toothed shoe is automatically reset every time the motor rotates an angle. The number of tangential grid division between the air gap and the toothed shoe is set to 1224, that is, the motor rotates 0.294 degrees each time, and rotates 1224 times to complete one mechanical cycle. When the rotor rotates an angle, the node of the air gap is disconnected from the current toothed shoe node, and is connected to the adjacent node in the same direction of the current connected toothed shoe node. The magnetic conductance calculation mode between the nodes does not need to be changed.

[0016] Further, the specific process of solving the magnetic potential of each node in step 3 is as follows: first, based on the Kirchhoff's flux law and the Kirchhoff's magnetic pressure law, a nonlinear equation set is established. When the node magnetic flux matrix is established, the coercive force and the magnetic permeability of the permanent magnet at the initial temperature need to be calculated, so as to consider the influence of temperature on the performance of the motor. Second, the equation set is solved to obtain the non-iterative node magnetic potential matrix.

[0017] Further, the specific process of step 3 is as follows: first, based on the Kirchhoff's flux law and the Kirchhoff's magnetic pressure law, a nonlinear equation set is established. When the node magnetic flux matrix is established, the coercive force and the magnetic permeability of the permanent magnet at the initial temperature need to be calculated, so as to consider the influence of temperature on the performance of the motor. Second, the equation set is solved to obtain the non-iterative node magnetic potential matrix. Finally, the multi-variable damping method is used to iterate the nonlinear magnetic permeability of the silicon steel sheet to obtain accurate node magnetic potentials. Based on the node magnetic potentials, the air gap magnetic density, the magnetic chain, the back electromotive force, the torque and the loss and other performances can be calculated.

[0018] Further, in step 4, the bearing, the casing, the winding, the tire, the brake and the air node are directly added to the equivalent magnetic network model to generate a global equivalent thermal network model, so that the common node model can calculate the electromagnetic performance and the temperature field distribution. It is not necessary to re-establish the equivalent thermal network model, and the time and the operation complexity of modeling and calculation are greatly reduced.

[0019] Further, in step 4, when solving the global equivalent thermal network model, not only the internal heat sources such as core loss and winding copper loss are considered, but also many external heat sources such as heat generated by braking and heat generated by tire friction are considered, so that the precision of the magnetic-thermal coupling analysis model of the permanent magnet wheel motor under the consideration of multiple heat sources is improved.

[0020] Further, the specific process of establishing and solving the global equivalent thermal network model in step 4 is as follows: based on the global equivalent magnetic network model, bearing, casing, winding and air nodes are added to generate a common node model; secondly, according to the basic theory of heat transfer, the internal heat conduction of the motor is divided into conduction heat conduction and convection heat conduction, and the heat conduction values are calculated according to the formulas; then, the core loss and winding copper loss and external heat sources calculated by the magnetic network are converted into heat generation rates; finally, the heat conduction matrix and heat source matrix are established, and the steady-state thermal balance equation is established, so as to solve the temperature rise of each node.

[0021] Further, in step 5, the mutual influence of the magnetic field and the temperature field in the motor is fully considered, and the magnetic field and the temperature field are bidirectionally coupled when solving, and the loss calculation result is introduced as a heat source into the equivalent thermal network model; then, based on the equivalent thermal network model, the temperature field is calculated, the calculated permanent magnet temperature is fed back to the equivalent magnetic network model, the corresponding electromagnetic material and temperature related property parameters are changed, and the loss is solved again to calculate a more accurate steady-state temperature distribution.

[0022] The beneficial effects of the technical scheme adopted by the present application are as follows:

[0023] 1. The present application fully considers the characteristics of the highly integrated and integrated design of the wheel motor, and when calculating the temperature field distribution, not only the internal heat sources such as core loss and winding copper loss are considered, but also many external heat sources such as heat generated by braking and heat generated by tire friction are considered, so that the precision of the magnetic-thermal coupling analysis model of the permanent magnet wheel motor under the consideration of multiple heat sources is improved.

[0024] 2. The present application fully considers the characteristics of the complex working environment and the variable operating conditions of the wheel motor, can consider the changes of electromagnetic properties and temperature properties of the wheel motor during starting, accelerating and decelerating, and the influence of different cooling methods on the internal temperature field distribution of the motor, and realizes the rapid calculation of electromagnetic performance and temperature distribution under multiple working conditions.

[0025] 3. The present application adopts a bidirectional coupling calculation method of magnetic field and temperature field, fully considers the mutual coupling influence of the magnetic field and the temperature field of the wheel motor during operation, i.e. the changes of the properties of the permanent magnet material caused by high temperature, the changes of the core loss and electromagnetic performance caused by the changes, and the influence of the changes of the core loss on the internal temperature distribution of the motor, so that more accurate electromagnetic performance and temperature distribution calculation is realized.

[0026] 4. The application adopts a co-node modeling idea, an equivalent thermal network model directly adds bearing, casing, winding, tire, brake and air node on the basis of an equivalent magnetic network model to generate a global equivalent thermal network model, so that the co-node model can calculate electromagnetic performance and temperature field distribution, and greatly reduces the time and operation complexity of modeling and calculation. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 It is a three-dimensional topological structure diagram of the embodiment of the application.

[0028] Figure 2 It is a parametric modeling structure diagram of the embodiment of the application.

[0029] Figure 3 It is a co-node magnetic and thermal bidirectional coupling network model schematic diagram of the embodiment of the application.

[0030] Figure 4 It is a hub motor temperature rise analysis method flow chart based on the co-node magnetic and thermal bidirectional coupling network model of the embodiment of the application.

[0031] Figure 5 It is a global equivalent magnetic network model schematic diagram of the embodiment of the application.

[0032] Figure 6 It is a magnetic network and finite element calculation magnetic density cloud map comparison: (a) magnetic network; (b) finite element.

[0033] Figure 7 It is a magnetic network and finite element calculation air gap magnetic density and output torque comparison chart before and after coupling: (a) air gap magnetic density; (b) output torque.

[0034] Figure 8 It is a thermal network and finite element calculation temperature cloud map comparison: (a) thermal network; (b) finite element. DETAILED DESCRIPTION

[0035] The technical solutions in the embodiments of the application will be described clearly and perfectly in combination with the drawings in the embodiments of the application.

[0036] In order to more clearly illustrate the specific implementation steps and beneficial effects of the application, a surface-mounted permanent magnet hub motor will be specifically described as follows: Figure 1 It is a three-dimensional topological structure diagram of the motor, wherein 1 is a rotor, 2 is a stator, 3 is a ferrite-boron permanent magnet, and 4 is an armature winding. The motor is a 36-slot / 34-pole outer rotor surface-mounted hub motor. The stator and rotor cores are silicon steel sheets, the permanent magnet is a ferrite-boron permanent magnet, and the armature winding adopts a concentrated winding mode.

[0037] As shown in FIG. 1, the equivalent magnetic network model of the motor is established by using the finite element software ANSYS, and the equivalent magnetic network model is shown in FIG. 2. Figure 4The flow chart shown, a total of the following steps to achieve:

[0038] Step 1, the establishment of local equivalent magnetic network model: as Figure 3 Shown, first of all, the motor topology as the basis, the entire motor is divided into stator yoke, stator teeth, stator tooth shoes, air gap, permanent magnet and rotor, etc. Several areas, respectively, on each area grid subdivision. The overall structure of the motor is a circle, in the hierarchical subdivision, each grid element shape is a sector, therefore, in order to ensure the accuracy of the calculation, each grid element according to the sector to calculate the permeance, no approximation. The radial permeance and tangential permeance of the calculation formula is different, the radial permeance G r And tangential permeance G t The formula is:

[0039]

[0040] In the formula, μ is the material permeability of the sector permeance unit, w1 and w2 are the upper and lower width of the sector permeance unit in the radial direction, l a Is the motor shaft length, h is the height of the sector permeance unit in the radial direction, R2 and R1 are the outer diameter and inner diameter of the sector permeance unit in the tangential direction, θ is the central angle of the sector permeance unit in the tangential direction.

[0041] Next, the establishment process of equivalent magnetic network model of stator yoke, stator teeth, stator tooth shoes, air gap, permanent magnet and rotor is described in detail.

[0042] Further, the magnetic flux density gradient of the stator yoke and the stator teeth is small, and the subdivision accuracy of this part has little effect on the accuracy of the calculation result, so the subdivision of this part does not have to be too fine. The stator yoke is divided into 36 sector permeance units in the tangential direction, which is consistent with the number of stator teeth, and can be regarded as the connecting permeance between the stator teeth. While ensuring the calculation accuracy, the calculation complexity is reduced. The stator teeth are divided into 36 rectangular permeance units in the radial direction, that is, a stator tooth is regarded as a magnetic permeance unit, and the calculation formula of rectangular magnetic permeance is:

[0043]

[0044] In the formula, μ is the material permeability of the rectangular permeance unit, l a Is the motor shaft length, w is the width of the rectangular cross section perpendicular to the direction of magnetic flux, h is the length of the rectangular section in the same direction as the magnetic flux.

[0045] Furthermore, the air gap is the site of energy conversion in the entire motor, and its magnetic field distribution is complex. The accuracy of the magnetic field calculation results in the air gap has a significant impact on the accuracy of the overall motor magnetic field calculation results. Therefore, this part needs to be meticulously meshed. Given that the motor has 36 slots / 34 poles, considering the consistency of the mesh—that is, the air gap mesh should be uniformly connected to both the stator gear mesh and the permanent magnet mesh—the number of tangential mesh subdivisions in the air gap is set to the least common multiple of 36 and 34, i.e., 1224. Radially, the air gap width is 1mm; based on experience, the number of radial mesh subdivisions in the air gap is set to 2.

[0046] Furthermore, the stator tooth shoe is connected to the air gap, and the meshing accuracy of this part also has a significant impact on the accuracy of the final calculation results. Therefore, the meshing accuracy of this part is consistent with that of the air gap part, with 1224 meshes in the tangential direction and two layers in the radial direction. Besides the stator tooth shoe, the tooth slots are also included in this part. Each slot in the stator corresponds to 34 tangential meshes, with 4 meshes for the tooth slots and 30 meshes for the tooth shoes.

[0047] Furthermore, since the permanent magnet is connected to the air gap, its computational accuracy has a smaller impact on the final calculation result than that of the air gap. Considering both computational accuracy and time consumption, a tangential mesh size of 306 was chosen for the permanent magnet, corresponding to a 1:4 ratio with the air gap mesh size, ensuring consistent mesh connectivity. The permanent magnet is 4mm thick, therefore it is divided into two radial layers, with its fan-shaped magnetic permeability units approximating rectangles.

[0048] Furthermore, the rotor is connected to the permanent magnet. Based on modeling experience, the number of tangential mesh subdivisions for the rotor is the same as that for the permanent magnet, also set to 306. This can greatly reduce programming complexity while maintaining computational accuracy. The rotor thickness is 11mm, so it is divided into 3 layers radially, and each sector magnetic permeability unit is approximately rectangular.

[0049] In summary, equivalent magnetic network models were established for the stator yoke, stator teeth, stator tooth shoes, air gap, permanent magnets, and rotor, respectively. Each magnetic permeation unit is either sector-shaped or rectangular, and the size of each magnetic permeation unit is dynamically determined by the motor dimensions. Figure 2 As shown, when establishing the equivalent magnetic network model, the rotor thickness, permanent magnet thickness, stator tooth height, tooth space width, stator tooth length, stator tooth width, and stator yoke width are set as variable parameters, whose values ​​can be dynamically set. The size of each sector magnetic permeability unit can be dynamically changed according to these parameters. For example, as the rotor thickness increases, the width of the rotor mesh automatically increases. Using the above method, an equivalent magnetic network model of any accuracy can be established for any motor.

[0050] Step 2, establishing a global parameterized equivalent magnetic network model: from step 1, the parameterized equivalent magnetic network model of each component is obtained independently, but the magnetic flux flows through all components in the motor, so it is necessary to connect each component in the magnetic network model to ensure the integrity of the magnetic network. In order from inside to outside, the stator yoke, the stator tooth, the stator tooth shoe, the air gap, the permanent magnet and the rotor are connected through dynamic grids in a one-to-one or one-to-many relationship between nodes. Finally, the nodes are numbered in order, which is the basis for establishing the magnetic permeability matrix.

[0051] The stator yoke node connects the two adjacent stator teeth in the tangential direction. When the stator yoke is divided, the two adjacent stator teeth are divided into a sector-shaped magnetic permeability unit, so the number of divisions is consistent with the stator teeth. The outer stator tooth is connected to the tooth shoe, and the stator tooth has one node and the tooth shoe has 30 nodes. The magnetic flux flows from the tooth shoe into the stator tooth, so the 30 nodes of the tooth shoe are connected to the one node of the stator tooth.

[0052] The grid division density of the tooth shoe and the air gap is the same, so the connection between the tooth shoe and the air gap is one node to one node. When the width of the tooth slot changes, the width of the tooth shoe and tooth slot grid will change accordingly, and the width of the air gap part grid connected thereto will also change automatically.

[0053] The corresponding relationship between the air gap and the permanent magnet is 4 nodes to 1 node. The permanent magnet is composed of 34 pieces, so one piece of permanent magnet is divided into 9 nodes in the tangential direction, corresponding to 36 nodes of the air gap, which can realize the consistency of grid connection.

[0054] The grid division density of the permanent magnet and the rotor is consistent, and the corresponding relationship is one node to one node. The size of the tangential grid will not change.

[0055] In summary, a global static equivalent magnetic network model can be established. Next, a transient equivalent magnetic network model is established, that is, when the motor is running, the node connection between the air gap and the tooth shoe is reset every time the motor rotates an angle. The number of tangential grid divisions between the air gap and the tooth shoe in the present invention is set to 1224, that is, the motor rotates 0.294 degrees each time, and rotates 1224 times to complete one mechanical period. When the rotor rotates an angle, the node connection between the air gap and the current tooth shoe node is disconnected, and the node connection between the air gap and the adjacent node in the same direction of the current connected tooth shoe node is connected, so that the transient equivalent magnetic network model at the current time can be established.

[0056] Step 3, solving the parameterized global equivalent magnetic network model: first, based on the Kirchhoff's current law and the Kirchhoff's voltage law, a nonlinear equation set is established to solve the magnetic potential of each node in the magnetic circuit, and the target nonlinear equation set is as follows:

[0057] [G][F]=[Φ] (3)

[0058] In the formula, G is the branch magnetic permeability matrix, F is the node magnetic potential matrix, and Φ is the node magnetic flux matrix. The non-iterative node magnetic potential is obtained by solving the equation set, wherein the magnetic flux matrix is calculated according to the demagnetization curve of the permanent magnet at the initial temperature to calculate the coercive force. Then, the multi-variable damping method is used to iterate the silicon steel sheet permeability until convergence, so as to obtain accurate node magnetic potential as the data source for subsequent calculation of motor performance.

[0059] The establishment steps of the magnetic permeability matrix and the magnetic flux matrix, and the process of using the multi-variable damping method to iterate and converge the silicon steel sheet nonlinear permeability are described below. Formula (3) is listed in the matrix form as follows:

[0060]

[0061] In the step 1, the equivalent magnetic network model has a total of 6498 nodes, that is, the magnetic permeability matrix G is a 6498*6498 square matrix. First, the magnetic permeability matrix is initialized as a zero matrix, and in the step 1 and the step 2, the connection between nodes is established, and the branch magnetic permeability between nodes is calculated according to the formula (1) (2) and filled into the magnetic permeability matrix according to the node number.

[0062] The magnetic flux matrix Φ includes the magnetic flux generated by the permanent magnet and the magnetic flux generated by the winding. When the winding is at no load, the magnetic flux generated by the winding is 0, and when the winding is at load, the magnetic flux generated by the winding is calculated according to the following formula:

[0063] Φ winding =NiG tooth (5)

[0064] In the formula, N is the number of turns of the winding, i is the current of the winding, G tooth is the stator tooth magnetic permeability, wherein i = I max sin(2πnpt / 60), I max is the maximum value of the winding current, n is the rotating speed, and p is the number of permanent magnet pole pairs. When calculating, the maximum current and the rotating speed are changed, so that the electromagnetic performance and the loss distribution under different working conditions are obtained, thereby the internal temperature rise distribution characteristics of the wheel hub motor under different working conditions are calculated.

[0065] The calculation formula of the permanent magnet magnetic flux is as follows:

[0066] Φ PM =H c h PM G PM (6)

[0067] In the formula, H c is the coercive force of the permanent magnet at the initial temperature, h PM is the thickness of the magnetization direction of the permanent magnet, and G PMThe permeance of the magnetization direction of the permanent magnet. Finally, based on the Python open source library SciPy, the equation set is solved to obtain the non-iterative node magnetic potential matrix.

[0068] According to the above-mentioned node magnetic potential, the branch magnetic density is calculated, and the calculation formula is as follows:

[0069]

[0070] In the formula, F(i) is the node i magnetic potential, F(j) is the node j magnetic potential, G (i,j) is the node i and node j branch permeance, S (i,j) is the cross-sectional area of the branch between node i and node j perpendicular to the magnetic flux flow direction. According to the obtained branch magnetic density, the magnetic hysteresis curve of the silicon steel sheet of this type is queried, and the permeability of the branch magnetic density is calculated. The multivariable damping method is used to update the branch permeability, and the calculation formula of the multivariable damping method is as follows:

[0071]

[0072]

[0073]

[0074] In the formula, is the calculated damping coefficient, c d is the damping factor, which is set to 0.6, is the branch permeability at the kth iteration, is the branch permeability at the k-1th iteration, when formula (10) is satisfied, it is considered that the branch permeability converges, and ε is the solution accuracy, which can be set according to the situation.

[0075] According to the above, the node magnetic potential can be obtained. Based on the node magnetic potential, the motor air gap magnetic density, three-phase magnetic flux, three-phase counter electromotive force, core loss and instantaneous torque can be further obtained. The magnetic flux calculation formula is as follows:

[0076]

[0077] In the formula, i is the ABC three-phase, N is the number of turns, is the i-phase jth stator tooth magnetic flux density, is the i-phase jth stator tooth cross-sectional area.

[0078] The counter electromotive force calculation formula is as follows:

[0079]

[0080] That is, the derivative of the magnetic flux with respect to time is the counter electromotive force of a certain phase.

[0081] The core loss calculation formula is as follows:

[0082]

[0083] According to the mechanism of iron loss, the core loss can be divided into hysteresis loss, eddy current loss and abnormal loss by Bertotti three-component iron loss separation model. In the formula, f is the frequency of alternating magnetic field, B m is the magnetic flux density amplitude, P h is the hysteresis loss, k h is the silicon steel hysteresis loss coefficient; P c is the eddy current loss, k c is the silicon steel eddy current loss coefficient; P e is the abnormal loss, k e is the silicon steel abnormal loss coefficient.

[0084] The instantaneous torque calculation formula is as follows:

[0085]

[0086] In the formula, l a is the axial length of the motor, u0 is the vacuum permeability, r airgap is the average radius of the motor air gap part, B t and B r are the tangential magnetic flux density and radial magnetic flux density of the air gap part.

[0087] The calculation results are shown in Figure 6 and Figure 7 , Figure 6 the magnetic flux density cloud diagram obtained by the magnetic network and the finite element is compared, Figure 7 (a) is a comparison diagram of no-load magnetic network and finite element calculation air gap magnetic flux density.

[0088] Step 4, establish a common node equivalent magnetic thermal network model: as shown in Figure 3 , first, the global common node model is divided into common nodes and thermal network nodes, the nodes of the magnetic network model obtained in step 2 become common nodes in the common node model, and the bearing, casing, winding and air nodes are added to the magnetic network model to become the common node thermal network nodes. Secondly, based on the basic theory of heat transfer, the internal thermal conductance of the motor is divided into conduction thermal conductance and convection thermal conductance, and their calculation formulas are as follows:

[0089]

[0090] G convection = hS dwm (16)

[0091] In the formula, G conduction is the conduction thermal conductance, G convectionwhere λ is the thermal conductivity of the material, L is the thermal conduction distance, h is the heat transfer coefficient of convection, S is the area of the isothermal surface perpendicular to the heat flow direction. dwm where λ is the thermal conductivity of the material, L is the thermal conduction distance, h is the heat transfer coefficient of convection, S is the area of the isothermal surface perpendicular to the heat flow direction.

[0092] where λ is the thermal conductivity of the material, L is the thermal conduction distance, h is the heat transfer coefficient of convection, S is the area of the isothermal surface perpendicular to the heat flow direction.

[0093]

[0094]

[0095] where v is the air speed inside the machine housing, T0 is the temperature of the environment outside the machine housing, v is the linear speed of the outer circle of the rotor. rt According to the material properties of the motor and the calculation formula, the thermal conductivities and heat dissipation coefficients of each part of the motor are shown in Table 1 and Table 2:

[0096] Table 1

[0097] Material Thermal conductivity (W / (m·℃) Specific heat capacity (KJ / (kg·℃) Density (kg / m3 3 )]]> M19_29G (core) 28 0.46 7800 NdFeB (permanent magnet) 7.6 0.46 7500 Copper (winding) 401 0.385 8933 Aluminum (casing, end cap) 180 0.963 2950 Stainless steel (bearing) 15 0.477 7900

[0098] Table 2

[0099] Surface position Heat dissipation coefficient (W / (m 2 • °C) Casing 15.33 End 37.93

[0100] Finally, the core loss calculated by the magnetic network, as well as the winding copper loss and external heat sources, are converted into heat generation rates, where external heat sources such as brakes and tires are added to the machine housing and bearing nodes, and a steady-state heat balance equation is established to solve the temperature of each node. The steady-state heat balance equation is as follows:

[0101] [G][T] = [P] (19)

[0102] where G is the thermal conduction matrix between the nodes of the thermal network, which can be obtained by combining Table 1, Table 2 and formula (15) (16), T is the temperature rise matrix of the nodes of the thermal network, and P is the heat source matrix in the thermal network, which can be obtained according to the converted heat generation rate. Solving the equation set, the temperature rise matrix T of each node can be obtained. The calculation results are shown in Table 3. Figure 8 The thermal network calculation shows that the average temperature of the stator tooth is 120.8°C, the average temperature of the stator yoke is 113.5°C, the average temperature of the permanent magnet is 96.0°C, and the average temperature of the rotor is 94.7°C; the finite element calculation shows that the average temperature of the stator tooth is 118.4°C, the average temperature of the stator yoke is 114.1°C, the average temperature of the permanent magnet is 96.3°C, and the average temperature of the rotor is 94.4°C.

[0103] Step 5, Two-way coupling of the global equivalent magnetic network model and the global equivalent thermal network model: Based on the above steps, unidirectional coupling from the magnetic field to the temperature field and unidirectional coupling from the temperature field to the magnetic field can be achieved. The losses calculated by the magnetic network are imported into the thermal network, and the temperatures calculated by the thermal network are imported into the magnetic network, respectively. This step is iterated until the temperature converges. The specific steps are as follows:

[0104] I. Import the initial temperature t0 into the magnetic network model, and calculate the core loss P based on the established parameterized equivalent magnetic network model. i ;

[0105] II. Calculate the core loss P in step I. i Using the winding copper loss and external heat sources as the heat sources of the thermal network, the permanent magnet temperature t is solved based on the global equivalent thermal network model established in step 4. i ;

[0106] III. Replace the initial temperature in step I with the temperature t obtained in step II. i ;

[0107] IV. Iterate through steps I, II, and III until the following formula is satisfied, at which point the iteration is considered converged:

[0108]

[0109] In the formula, t i+1 Let t be the temperature of the permanent magnet calculated in the (i+1)th iteration. i ε represents the permanent magnet temperature calculated in the i-th iteration, and ε is the iteration accuracy, which can be set according to the calculation accuracy requirements.

[0110] V. Output the motor temperature field distribution when the iteration condition is met, and then output the permanent magnet temperature t that meets the iteration condition. i Import the equivalent magnetic network model and calculate the air gap magnetic flux density, three-phase magnetic flux linkage, three-phase back electromotive force, core loss and output torque respectively.

[0111] The calculation results are as follows Figure 7 As shown in (b), the average output torque before coupling is 86.85 Nm calculated by the finite element method and 86.81 Nm calculated by the magnetothermal coupling network model; the average output torque after coupling is 80.03 Nm calculated by the finite element method and 79.70 Nm calculated by the magnetothermal coupling network model.

[0112] In summary, the application first proposes a wheel motor temperature rise analysis method based on a co-node magnetic heat two-way coupling network model. First, a local parameterized equivalent magnetic network model is established for each component according to the motor structure, and each component is connected through adjacent nodes to establish a global parameterized equivalent magnetic network model. Based on the Kirchhoff's current law and the Kirchhoff's voltage law, the equivalent magnetic network model is solved to obtain the magnetic potential of each node. When establishing the node magnetic flux matrix, the speed and current can be changed according to the working condition to calculate the electromagnetic performance and temperature distribution under different operating conditions. Secondly, based on the established global parameterized equivalent magnetic network model, the bearing, casing, winding, tire, brake and air nodes are added to establish the motor steady-state heat balance equation to solve the temperature of each node. Then, the global magnetic network model and the global thermal network model are two-way coupled using an iterative algorithm to calculate the accurate steady-state temperature, and the electromagnetic performance of the motor is calculated based on the temperature using the magnetic network model. The application first considers the multi-heat source multi-condition magnetic heat two-way coupling analysis of the surface-mounted wheel motor, and the provided scheme can provide a reference for the magnetic field and temperature field analysis of this type of wheel motor.

[0113] The above series of detailed descriptions are only specific descriptions of the feasible embodiments of the application, and are not used to limit the protection scope of the application. Any equivalent means or changes without departing from the technology of the application should be included in the protection scope of the application.

Claims

1. A wheel motor temperature rise analysis method based on a co-node magneto-thermal bidirectional coupling network model, characterized in that, Comprise the following steps: Step 1, establish local equivalent magnetic network model: on the basis of motor topology, the whole motor is divided into stator yoke, stator tooth, stator tooth shoe, air gap, permanent magnet and rotor several areas, respectively to each area for grid subdivision;The overall structure of the motor is a circle, when layering subdivision, the shape of each grid unit is a sector, when grid subdivision, on the basis of the magnetic field distribution of each component, select the subdivision accuracy, shorten the calculation time on the premise of ensuring the calculation accuracy; Step 2, establish global transient equivalent magnetic network model: based on the local equivalent magnetic network model, each component is connected through the node, on the basis of the subdivision accuracy of each component, the adjacent two parts are connected with one-to-one or one-to-many relationship between nodes, that is, the global steady-state equivalent magnetic network model can be established;When the motor rotates, the connection relationship between the stator tooth shoe and the air gap node is changed, that is, every time the motor rotates an angle, the air gap node disconnects with the current node, and connects with the node adjacent to the rotation direction of the shoe node, so that the global transient equivalent magnetic network model can be established; Step 3, solve the global equivalent magnetic network model, solve the magnetic potential of each node; Step 4, establish and solve the global equivalent thermal network model; The specific process of establishing and solving the global equivalent thermal network model in step 4 is: based on the global equivalent magnetic network model, add bearing, casing, winding and air node to generate a common node model;Secondly, according to the basic theory of heat transfer, the internal thermal conductivity of the motor is divided into conduction thermal conductivity and convection thermal conductivity, and the thermal conductivity value is calculated according to the formula respectively;Then, according to the core loss and winding copper loss calculated by the magnetic network and the external heat source, the heat generation rate is converted;Finally, the thermal conductivity matrix and the heat source matrix are established respectively, and the steady-state thermal balance equation is established, so as to solve the temperature rise of each node; Step 5, coupling global equivalent magnetic network model and global equivalent thermal network model: first, use the equivalent magnetic network model to calculate the core loss at the initial temperature, and take this loss as one of the initial heat sources of the equivalent thermal network model, on this basis, add the winding copper loss and the external heat source to calculate the temperature of each node, then recalculate the core loss at this temperature, repeat the above steps until the temperature of the permanent magnet meets the set convergence accuracy, that is, the temperature is considered as the target temperature, and finally the electromagnetic performance and temperature distribution of the motor at this temperature can be calculated.

2. The hub motor temperature rise analysis method based on the co-node magnetic heat bidirectional coupling network model according to claim 1, characterized in that, The motor is a 36-slot / 34-pole three-phase surface-mounted permanent magnet wheel hub motor, which comprises a stator, a rotor, a permanent magnet and an armature winding, the stator and rotor cores are silicon steel sheets, the permanent magnet is NdFeB ferrite boron permanent magnet, and the armature winding adopts concentrated winding mode.

3. The hub motor temperature rise analysis method based on the co-node magneto-thermal bidirectional coupling network model according to claim 1, characterized in that, In step 1, the stator tooth shoe is connected with the air gap, the subdivision accuracy of this part is consistent with that of the air gap part, the number of tangential grid subdivision is 1224, and the radial is also divided into 2 layers, in addition to the stator tooth shoe, the tooth slot is also included in this part, and the number of tangential grid corresponding to each slot of the stator is 34, wherein the number of grid corresponding to the tooth slot is 4, and the number of grid corresponding to the tooth shoe is 30. The permanent magnet is connected with the air gap, and the number of tangential grid division of the permanent magnet is 306, and the corresponding relationship between the number of grid of the air gap and the number of grid of the permanent magnet is 1:4, so that the consistency of the grid can be realized, the thickness of the permanent magnet is 4mm, so that the permanent magnet is divided into 2 layers in the radial direction, and the sector-shaped magnetic conductive unit is approximately a square; The rotor is connected with the permanent magnet, the number of tangential grid division of the rotor is the same as that of the permanent magnet, and is also set to 306, so that the calculation accuracy can be maintained while the programming complexity is greatly reduced, the thickness of the rotor is 11mm, so that the rotor is divided into 3 layers in the radial direction, and the sector-shaped magnetic conductive unit is also approximately a square.

4. The hub motor temperature rise analysis method based on the co-node magneto-thermal bidirectional coupling network model according to claim 1, characterized in that, In step 2, when the transient equivalent magnetic network model is established, the connection between the nodes of the air gap and the toothed shoe is automatically reset every time the motor rotates an angle, the number of tangential grid division between the air gap and the toothed shoe is set to 1224, that is, the motor rotates 0.294 degrees each time, and rotates 1224 times to complete one mechanical cycle, when the rotor rotates an angle, the node of the air gap is disconnected from the current toothed shoe node, and is connected with the adjacent node in the same direction of the current connected toothed shoe node, without changing the magnetic conductive calculation mode between the nodes.

5. The hub motor temperature rise analysis method based on the co-node magneto-thermal bidirectional coupling network model according to claim 1, characterized in that, In step 3, the specific process for solving the magnetic potential of each node is as follows: firstly, based on the Kirchhoff's flux law and the Kirchhoff's voltage law, a nonlinear equation set is established, wherein the coercive force and the magnetic permeability of the permanent magnet at the initial temperature need to be calculated when the node magnetic flux matrix is established, so as to consider the influence of temperature on the performance of the motor; secondly, the equation set is solved to obtain the non-iterative node magnetic potential matrix.

6. The hub motor temperature rise analysis method based on the co-node magneto-thermal bidirectional coupling network model according to claim 1, characterized in that, In step 3, the specific process for solving the magnetic potential of each node is as follows: firstly, based on the Kirchhoff's flux law and the Kirchhoff's voltage law, a nonlinear equation set is established, wherein the coercive force and the magnetic permeability of the permanent magnet at the initial temperature need to be calculated when the node magnetic flux matrix is established, so as to consider the influence of temperature on the performance of the motor; secondly, the equation set is solved to obtain the non-iterative node magnetic potential matrix; finally, the multi-variable damping method is used for iteration of the nonlinear magnetic permeability of the silicon steel sheet, to obtain accurate node magnetic potentials; based on the node magnetic potentials, the air gap magnetic flux density, the magnetic flux linkage, the back electromotive force, the torque and the loss performance can be calculated.

7. The hub motor temperature rise analysis method based on the co-node magneto-thermal bidirectional coupling network model according to claim 1, characterized in that, In step 4, the bearing, the casing, the winding, the tire, the brake and the air node are directly added to the equivalent magnetic network model to generate a global equivalent thermal network model, so that the common node model can calculate the electromagnetic performance and the temperature field distribution.

8. The hub motor temperature rise analysis method based on the co-node magneto-thermal bidirectional coupling network model according to claim 1, characterized in that, In step 4, when the global equivalent thermal network model is solved, not only the core loss and the winding copper loss are considered, but also the heat generated by braking and the heat generated by tire friction are considered.

9. The hub motor temperature rise analysis method based on the co-node magneto-thermal bidirectional coupling network model according to claim 1, characterized in that, In step 5, the mutual influence of the magnetic field and the temperature field in the motor is fully considered, the loss calculation result is introduced into the equivalent thermal network model as a heat source when the magnetic field and the temperature field are bidirectionally coupled; then the temperature field is calculated based on the equivalent thermal network model, the calculated permanent magnet temperature is fed back to the equivalent magnetic network model, the corresponding electromagnetic material and temperature related property parameters are changed, and the loss is solved again to calculate a more accurate steady-state temperature distribution.

Images