A method and system for predicting the demagnetization rate of a neodymium-iron-boron permanent magnet

Abstract

The application provides a method and system for predicting the demagnetization rate of a neodymium-iron-boron permanent magnet, and relates to the technical field of predicting the demagnetization rate of a neodymium-iron-boron permanent magnet, comprising the following steps: adopting a composite Simpson numerical integration method on a transient damage rate sequence to calculate the equivalent cumulative damage of the magnet in a single sampling period; calculating the equivalent cumulative damage of a single period of each experimental sample; constructing a corrected mapping function model, combining the magnetic flux loss rate of each group of magnets and the total experimental damage, fitting the mapping function model, and establishing the corresponding relationship between the total experimental damage and the magnetic flux loss rate; obtaining a magnet to be detected, repeating step S1, and obtaining the predicted magnetic flux loss rate of the magnet to be detected. Through a dynamic load spectrum experiment with ladder changes, a plurality of magnetic flux loss rate data are obtained, the corrected mapping function model is fitted by using a Levenberg-Marquardt algorithm, and the fitting precision and the generalization ability of the prediction model are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of predicting the demagnetization rate of NdFeB permanent magnets, specifically to a method and system for predicting the demagnetization rate of NdFeB permanent magnets. Background Technology

[0002] Neodymium iron boron (NdFeB) permanent magnets are widely used in new energy vehicle drive motors, wind turbines, aerospace actuators, and precision medical devices due to their excellent energy product and coercivity. However, during actual service, especially in complex environments with high temperature and reverse magnetic field coupling, NdFeB permanent magnets are prone to irreversible demagnetization, leading to a gradual loss of magnetic flux, i.e., an increase in demagnetization rate, which in turn affects the efficiency, reliability, and even the safe lifespan of the entire equipment. In the prior art, the document with publication number CN121031158A proposes a method for predicting the demagnetization rate of NdFeB permanent magnets based on physical field simulation. However, this method is based on static magnetic field distribution simulation and simplified dynamic attenuation assumptions. It lacks a physical description of the instantaneous damage evolution process of the magnet under the coupling effect of time-varying temperature and time-varying reverse magnetic field. It does not establish a cross-scale damage accumulation model from microscopic demagnetization mechanism to macroscopic magnetic flux loss. It also lacks intermediate parameters with clear physical meaning, such as instantaneous damage rate and equivalent cumulative damage. Therefore, it is impossible to quantitatively compare the cumulative damage effect under dynamic load spectra of different severity. Therefore, there is an urgent need for a method and system for predicting the demagnetization rate of NdFeB permanent magnets.

[0003] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for predicting the demagnetization rate of neodymium iron boron permanent magnets, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for predicting the demagnetization rate of NdFeB permanent magnets, comprising the following steps: S1: Set the sampling period, collect temperature data and magnetic field strength data of the magnet at equal time intervals within the sampling period, and construct the instantaneous damage rate model of the magnet based on the temperature data and magnetic field strength data. Calculate the instantaneous damage rate of the magnet at any sampling moment, and construct the instantaneous damage rate sequence of the magnet accordingly. Use the composite Simpson numerical integration method on the instantaneous damage rate sequence to calculate the equivalent cumulative damage of the magnet within a single sampling period. S2: Obtain multiple sets of magnets to form experimental samples, apply a dynamic load spectrum to the experimental samples with temperature and reverse magnetic field strength varying stepwise over time, set the total number of cycles of the dynamic load spectrum, collect the residual magnetic flux of each set of magnets before and after the experiment, and calculate the magnetic flux loss rate. Repeat step S1 for the applied dynamic load spectrum, calculate the equivalent cumulative damage of the magnets in a single sampling period, and calculate the total experimental damage of each set of magnets based on the equivalent cumulative damage. S3: Construct a modified mapping function model, pair the flux loss rate of each group of magnets with the corresponding total experimental damage, and fit the mapping function model based on the data pairs to establish the correspondence between the total experimental damage and the flux loss rate. S4: Calculate the equivalent cumulative damage of the magnet under test in a single sampling period according to step S1, determine the total number of cycles expected to be used by the magnet under test, multiply the equivalent cumulative damage by the total number of cycles to obtain the total experimental damage, and obtain the predicted magnetic flux loss rate of the magnet under test according to the correspondence between the total experimental damage and the magnetic flux loss rate.

[0006] Furthermore, a model for the instantaneous damage rate of the magnet is constructed, specifically as follows: Based on temperature and magnetic field strength data, an instantaneous damage rate model for the magnet is constructed, and the instantaneous damage rate of the magnet at any sampling time is calculated: in, Indicates the pre-factor, The kinetic index is obtained by fitting multiple linear regression data from accelerated aging experiments using an isothermal constant magnetic field. This represents the magnetocrystalline anisotropy field of the magnet at the reference temperature; This represents the temperature coefficient of an anisotropic field; Indicates the average activation energy; Indicates the permeability of free space; Indicates saturation magnetization; Represents the Boltzmann constant; Indicates reference temperature; Indicates the sampling time index; Indicates the first Instantaneous damage rate of the magnet at each sampling moment; Indicates in The strength of the reverse magnetic field used for the magnet at any given time; Indicates in The absolute temperature of the magnet at any given moment; Indicates in The temperature softening function corresponding to the absolute temperature of the magnet at any given moment; The isothermal constant magnetic field accelerated aging experiment specifically involves: using multiple neodymium iron boron permanent magnet samples, maintaining the temperature of each sample at a constant set value during the experiment, and simultaneously applying a constant reverse magnetic field for a period of time; measuring the residual magnetic flux of the neodymium iron boron permanent magnet samples before and after the experiment to calculate the magnetic flux loss rate under the given temperature and magnetic field conditions; then, repeating the experiment by changing different constant temperature levels and different constant reverse magnetic field strength levels to obtain a series of temperature-magnetic field-magnetic flux loss rate data combinations. Since both temperature and magnetic field remain constant over time, the instantaneous damage rate in the instantaneous damage rate model is a constant. Therefore, the cumulative damage equals the instantaneous damage rate multiplied by the experimental time. Taking the logarithm of both sides of the instantaneous damage rate model, with the logarithm of the cumulative damage measured in the isothermal constant magnetic field accelerated aging experiment as the dependent variable and the reciprocal of temperature and the magnetic field correlation function as independent variables (the magnetic field correlation function refers to the natural logarithm of the reverse magnetic field strength and the ratio of the reverse magnetic field strength to the absolute temperature), a multiple linear regression is used for fitting, thereby simultaneously determining the antecedent factors. and dynamic index The value.

[0007] Furthermore, the equivalent cumulative damage of the magnet within a single sampling period is calculated as follows: Temperature data of the magnet at each sampling time and magnetic field strength data Substituting these values ​​into the instantaneous damage rate model yields the complete instantaneous damage rate sequence. ,in, This represents the number of sampling points within a single sampling period. An index indicating the sampling time; Indicates the first Instantaneous damage rate at each sampling time; The equivalent cumulative damage to the magnet within a single sampling period is calculated using the composite Simpson numerical integration method on the instantaneous damage rate sequence. in, Indicates the sampling time interval; Indicates equivalent cumulative damage: This represents the instantaneous damage rate of the magnet at the first sampling time. Indicates the first Instantaneous damage rate of the magnet at each sampling moment; Indicates the index of even-numbered sampling times; Indicates the index of the odd-numbered sampling time; Indicates the first Instantaneous damage rate of the magnet at each sampling moment; Indicates the first Instantaneous damage rate of the magnet at each sampling time.

[0008] Furthermore, the calculation of the magnetic flux loss rate specifically involves subtracting the residual magnetic flux after the experiment from the residual magnetic flux before the experiment for each group of magnets, and then dividing the difference by the residual magnetic flux before the experiment to obtain the true magnetic flux loss rate.

[0009] Furthermore, the total experimental damage of each group of magnets was calculated based on the equivalent cumulative damage, specifically as follows: For the dynamic load spectrum where the applied temperature and magnetic field strength change stepwise over time, repeat step S1 to calculate the equivalent cumulative damage for a single sampling period of each experimental sample. Multiply the equivalent cumulative damage of each group of magnets in a single sampling period by the total number of cycles of each group of magnets to obtain the total experimental damage of each group of magnets. The length of the sampling period is set to be equal to the length of a complete cycle in the dynamic load spectrum.

[0010] Furthermore, a modified mapping function model is constructed, specifically as follows: The flux loss rate and total experimental damage of each group of magnets were combined to form data pairs. Based on the data, the mapping function model was fitted to establish the correspondence between the total experimental damage and the flux loss rate. in, Indicates the maximum irreversible loss. This represents the damage efficiency coefficient. This represents the nonlinear shape factor, determined by fitting using the nonlinear least squares method. Indicates the total damage in the experiment Flux loss rate at the following conditions; This indicates the total damage in the experiment.

[0011] Furthermore, the relationship between total experimental damage and flux loss rate is established, specifically as follows: The mapping function model is fitted using a nonlinear least squares method, with the Levenberg-Marquardt algorithm used for iterative solution until the change in the sum of squared residuals between the predicted flux loss rate and the experimentally measured actual flux loss rate is less than a preset convergence threshold.

[0012] Furthermore, the predicted flux loss rate of the magnet to be tested is obtained, specifically: Obtain the total number of cycles of the magnet to be tested and its expected service life. Repeat step S1 to calculate the equivalent cumulative damage within a single sampling period. Multiply the equivalent cumulative damage within a single sampling period by the total number of cycles of expected service life to obtain the total experimental damage. Based on the correspondence between the total experimental damage and the flux loss rate, obtain the predicted flux loss rate of the magnet to be tested. in, This represents the predicted flux loss rate of the magnet under test; This represents the total experimental damage to the magnet being tested.

[0013] The present invention also provides a system for predicting the demagnetization rate of NdFeB permanent magnets. This system is used to perform the aforementioned method for predicting the demagnetization rate of NdFeB permanent magnets, comprising: Data acquisition module: Used to set the sampling period, collect temperature and magnetic field strength data of the magnet at equal time intervals within the sampling period, construct an instantaneous damage rate model of the magnet based on the temperature and magnetic field strength data, calculate the instantaneous damage rate of the magnet at any sampling moment, and construct an instantaneous damage rate sequence of the magnet accordingly; use the composite Simpson numerical integration method on the instantaneous damage rate sequence to calculate the equivalent cumulative damage of the magnet within a single sampling period; Damage calculation module: used to acquire multiple groups of magnets to form experimental samples, apply a dynamic load spectrum to the experimental samples with temperature and reverse magnetic field strength varying stepwise over time, set the total number of cycles of the dynamic load spectrum, collect the residual magnetic flux of each group of magnets before and after the experiment, calculate the magnetic flux loss rate, repeat step S1 for the applied dynamic load spectrum, calculate the equivalent cumulative damage of the magnets in a single sampling period, and calculate the total experimental damage of each group of magnets based on the equivalent cumulative damage. Function mapping module: used to construct the modified mapping function model, to form a data pair between the magnetic flux loss rate of each group of magnets and the corresponding total experimental damage, and to fit the mapping function model based on the data pair to establish the correspondence between the total experimental damage and the magnetic flux loss rate; The prediction module is used to calculate the equivalent cumulative damage of the magnet under test in a single sampling period according to step S1, determine the total number of cycles expected to be used by the magnet under test, multiply the equivalent cumulative damage by the total number of cycles to obtain the total experimental damage, and obtain the predicted magnetic flux loss rate of the magnet under test according to the correspondence between the total experimental damage and the magnetic flux loss rate.

[0014] Compared with the prior art, the beneficial effects of the present invention are: This scheme employs a composite Simpson numerical integral method to integrate the instantaneous damage rate sequence to obtain the equivalent cumulative damage within a single sampling period. Then, multiple sets of flux loss rate data are obtained through a step-changing dynamic load spectrum experiment. The Levenberg-Marquardt algorithm is used to perform nonlinear least-squares fitting on the modified mapping function model, establishing a high-precision correspondence between the total experimental damage and the flux loss rate with clear physical meaning, significantly improving the fitting accuracy and generalization ability of the prediction model. Through isothermal constant magnetic field accelerated aging experiments combined with multiple linear regression, the prefactor and kinetic exponent in the instantaneous damage rate model are independently obtained, ensuring the identifiability and physical interpretability of the model parameters. For the magnet under test, the equivalent cumulative damage in a single sampling period is multiplied by the expected total number of service cycles to obtain the total experimental damage, which is then substituted into the mapping function to quickly predict the flux loss rate without the need for destructive experiments, significantly reducing testing costs and time. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the overall method flow of the present invention; Figure 2 A graph showing the relationship between the flux loss rate and the total experimental damage to the magnet under test; Figure 3 This is a schematic diagram of the overall system structure of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0017] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0018] Example: Please see Figures 1-2 The present invention provides a technical solution: A method for predicting the demagnetization rate of NdFeB permanent magnets, comprising the following steps: S1: Set the sampling period, collect temperature data and magnetic field strength data of the magnet at equal time intervals within the sampling period, and construct the instantaneous damage rate model of the magnet based on the temperature data and magnetic field strength data. Calculate the instantaneous damage rate of the magnet at any sampling moment, and construct the instantaneous damage rate sequence of the magnet accordingly. Use the composite Simpson numerical integration method on the instantaneous damage rate sequence to calculate the equivalent cumulative damage of the magnet within a single sampling period. In the above process, the actual working cycle of the magnet is first analyzed. The time length of a complete cycle of temperature and reverse magnetic field strength change is set as the sampling period, such as a complete working condition cycle of the motor from start-up, rated operation, overload to shutdown. The magnet is pre-tested to calculate the rate of change of temperature and the rate of change of reverse magnetic field strength over time. The physical quantity that changes faster is selected as the dominant factor, and the time interval is set to one-fifth to one-tenth of the characteristic change period of the dominant factor to ensure that there are at least 5 to 10 sampling points in each transient fluctuation period. At the same time, referring to the Nyquist sampling theorem, it is ensured that the time interval is less than half of the period corresponding to the highest frequency component in the acquired signal.

[0019] In this embodiment, the instantaneous damage rate model of the magnet is constructed as follows: Based on temperature and magnetic field strength data, an instantaneous damage rate model for the magnet is constructed, and the instantaneous damage rate of the magnet at any sampling time is calculated: in, Indicates the pre-factor, The kinetic index is obtained by fitting multiple linear regression data from accelerated aging experiments using an isothermal constant magnetic field. This represents the magnetocrystalline anisotropy field of the magnet at the reference temperature; This represents the temperature coefficient of an anisotropic field; Indicates the average activation energy; Indicates the permeability of free space; Indicates saturation magnetization; Represents the Boltzmann constant; Indicates reference temperature; Indicates the sampling time index; Indicates the first Instantaneous damage rate of the magnet at each sampling moment; Indicates the first The reverse magnetic field strength at each sampling time; Indicates the first The absolute temperature of the magnet at each sampling moment; Indicates in The temperature softening function corresponding to the absolute temperature of the magnet at any given moment; In the above process, the dependent variable Indicates the first The instantaneous damage rate of a magnet at each sampling moment has the physical meaning of the degree of irreversible demagnetization damage caused by the coupling effect of high temperature and reverse magnetic field on the magnet per unit time. The larger the value, the faster the demagnetization rate of the magnet at the current moment and the more significant the tendency of micro-magnetic structure instability. Conversely, it indicates that the magnet is in a relatively stable magnetization state. The influence of the independent variable on the dependent variable is set based on the actual demagnetization physical mechanism of neodymium iron boron permanent magnets: reverse magnetic field strength. Through the magnetocrystalline anisotropic field and temperature softening function After normalization, then through the kinetic index The magnification reflects the effect of the reverse magnetic field weakening the magnet's resistance to demagnetization; absolute temperature Through Boltzmann factor The introduction demonstrates the physical process by which the thermal activation effect assists the reverse magnetic field in overcoming the magnetocrystalline anisotropy barrier, wherein the reverse magnetic field... Its presence in the molecule of the exponential term indicates that an external magnetic field can reduce the average activation energy required for demagnetization. ; From the perspective of positive and negative correlation, instantaneous damage rate With reverse magnetic field strength It shows a positive correlation with the temperature softening function. A negative correlation means that the increase in temperature causes Decrease thus increase the fractional value, relative to absolute temperature The exponential term exhibits a complex positive correlation; that is, as temperature increases, the absolute value of the negative exponential term decreases, thereby increasing the value of the exponential function overall. and The larger the instantaneous damage rate The higher the value, the more consistent it is with the physical law that neodymium iron boron permanent magnets demagnetize and accelerate under high temperature and strong reverse magnetic field. Magnetocrystalline anisotropic field This parameter indicates the reference temperature. The minimum reverse magnetic field strength required to change the spontaneous magnetization direction of a magnet from the easy magnetization axis to the difficult magnetization axis reflects the magnet's intrinsic ability to resist demagnetization. In practice, a vibrating sample magnetometer or torque magnetometer is typically used to measure the magnetization curves of the magnet in the orientation and pressing directions. By analyzing the magnetization curve in the first quadrant and extending the intersection of the two curves, the magnetic field strength value corresponding to this intersection on the abscissa is the anisotropic field. The anisotropic field measured at different temperatures is extrapolated to a reference temperature. You can get ; Temperature coefficient of anisotropic field This parameter describes the sensitivity of the magnetocrystalline anisotropic field to temperature variations. To determine this value, the anisotropic field of the magnet needs to be measured at multiple different temperature points, such as 20°C, 60°C, 100°C, and 140°C. Then, using the reference temperature Below Based on the linear relationship By fitting the data and performing least squares regression, the temperature coefficient can be obtained. The value; Average activation energy This parameter represents the average energy barrier that a magnet needs to overcome to achieve irreversible reversal of the magnetic domains over the magnetocrystalline anisotropy barrier when no external reverse magnetic field is applied. It is typically determined using the isothermal remanence curve method or viscous magnetization experiments: At a constant temperature, different magnitudes of reverse magnetic fields are applied to the magnet and held for a fixed time; the rate of decay of the remanence over time is measured; and the decay rate is then determined according to the Arrhenius relation of thermal activation theory. By fitting the data in a proportional manner, the average activation energy can be derived. The value; Reference temperature This is a reference temperature point used for normalization processing. It is usually selected based on the actual application conditions of NdFeB permanent magnets. A common practice is to set room temperature, i.e., 25°C or 20°C, as the magnetic performance data is easy to measure and has good repeatability at this temperature.

[0020] The isothermal constant magnetic field accelerated aging experiment specifically involves: using multiple neodymium iron boron permanent magnet samples, maintaining the temperature of each sample at a constant set value during the experiment, and simultaneously applying a constant reverse magnetic field for a period of time; measuring the residual magnetic flux of the neodymium iron boron permanent magnet samples before and after the experiment to calculate the magnetic flux loss rate under the given temperature and magnetic field conditions; then, repeating the experiment by changing different constant temperature levels and different constant reverse magnetic field strength levels to obtain a series of temperature-magnetic field-magnetic flux loss rate data combinations. Since both temperature and magnetic field remain constant over time, the instantaneous damage rate in the instantaneous damage rate model is a constant. Therefore, the cumulative damage equals the instantaneous damage rate multiplied by the experimental time. Taking the logarithm of both sides of the instantaneous damage rate model, with the logarithm of the cumulative damage measured in the isothermal constant magnetic field accelerated aging experiment as the dependent variable and the reciprocal of temperature and the magnetic field correlation function as independent variables (the magnetic field correlation function refers to the natural logarithm of the reverse magnetic field strength and the ratio of the reverse magnetic field strength to the absolute temperature), a multiple linear regression is used for fitting, thereby simultaneously determining the antecedent factors. and dynamic index The value.

[0021] In the above process, by keeping the temperature and reverse magnetic field strength constant, the instantaneous damage rate under the experimental conditions is kept constant. This simplifies the cumulative damage to the product of the instantaneous damage rate and the experimental time, allowing the logarithm of both sides of the instantaneous damage rate model to be transformed into a linear form. Changing the gradient to different constant temperature levels and different constant reverse magnetic field strength levels allows for the acquisition of cumulative damage data covering multiple temperature-magnetic field combinations. Temperature changes are used to introduce the effects of the reciprocal of the temperature term and the temperature softening function, while changes in the reverse magnetic field strength are used to introduce the effects of the natural logarithm of the reverse magnetic field strength and the ratio of the reverse magnetic field strength to the absolute temperature. Since these independent variables are mathematically linearly separable, multiple linear regression can be used to simultaneously fit the antecedent factors. and dynamic index The values ​​are used to ensure that the parameters of the two models have global consistency and physical interpretability, and to avoid parameter coupling deviation caused by fitting a single working condition. Multiple linear regression fitting can be accomplished using various specialized software or programming languages, such as Python, R, MATLAB, SPSS, Origin, and data analysis toolkits in Excel. Taking Python as an example, researchers can use the OLS function in the statsmodels library or the LinearRegression class in the scikit-learn library to achieve fitting. The specific steps are as follows: First, organize the data obtained from the isothermal constant magnetic field accelerated aging experiment into a spreadsheet, where the dependent variable is set as the natural logarithm of cumulative damage, and the independent variables are set as the reciprocal of temperature, the natural logarithm of the reverse magnetic field strength, and the ratio of the reverse magnetic field strength to the absolute temperature. Then, use software to read the data file, call the corresponding regression function to establish a multiple linear regression model. After the model runs, it will automatically output the regression coefficients of each independent variable. Based on the correspondence between the output regression coefficients and the parameters in the original instantaneous damage rate model, the values ​​of the initial factor A and the kinetic exponent m can be solved simultaneously.

[0022] The equivalent cumulative damage of the magnet during a single sampling period is calculated as follows: Temperature data of the magnet at each sampling time and magnetic field strength data Substituting these values ​​into the instantaneous damage rate model yields the complete instantaneous damage rate sequence. ,in, This represents the number of sampling points within a single sampling period. An index indicating the sampling time; Indicates the first Instantaneous damage rate at each sampling time; The equivalent cumulative damage to the magnet within a single sampling period is calculated using the composite Simpson numerical integration method on the instantaneous damage rate sequence. in, Indicates the sampling time interval; Indicates equivalent cumulative damage: This represents the instantaneous damage rate of the magnet at the first sampling time. Indicates the first Instantaneous damage rate of the magnet at each sampling moment; Indicates the index of even-numbered sampling times; Indicates the index of the odd-numbered sampling time; Indicates the first Instantaneous damage rate of the magnet at each sampling moment; Indicates the first Instantaneous damage rate of the magnet at each sampling time.

[0023] In the above process, the dependent variable This represents the equivalent cumulative damage of a magnet within a single sampling period. Its physical meaning is the instantaneous damage rate that changes continuously with time within the sampling period. The equivalent total damage obtained by time-weighted accumulation; The larger the value, the more severe the irreversible demagnetization accumulation caused by the coupling effect of temperature and reverse magnetic field on the magnet during the sampling period; conversely, the smaller the value, the less severe the magnet degradation during the period. The influence of the independent variable on the dependent variable is represented by a weighted integral sum: the instantaneous damage rate at each sampling time. These are the basic input quantities, which are generated by the instantaneous damage rate model based on the temperature data at each time point. and reverse magnetic field strength data The calculation was obtained; and the sampling time interval was... The different weighting coefficients corresponding to odd and even sampling points—odd index points with a weight of 4, even index points with a weight of 2, and endpoints with a weight of 1—determine the contribution of each instantaneous damage rate to the final equivalent cumulative damage. Assigning a weight of 4 to the middle sampling point of the odd index, a weight of 2 to the internal sampling point of the even index, and a weight of 1 to the two endpoints enables the integral formula formed by these weighted sums to accurately fit the integral value of the cubic polynomial function, thus achieving fourth-order algebraic accuracy. Since the instantaneous damage rate of NdFeB permanent magnets under dynamic load spectrum usually exhibits a smooth and continuous characteristic over time, this weighting method can more accurately approximate the true area under the instantaneous damage rate curve, i.e., the equivalent cumulative damage. Compared with the trapezoidal method (endpoint weight 0.5), the internal point weight 1 can effectively reduce the numerical integration error caused by discrete sampling, and obtain a higher accuracy equivalent cumulative damage calculation result while ensuring a limited number of sampling points. From the perspective of positive and negative correlation, equivalent cumulative damage Instantaneous damage rate at each sampling time Both are positively correlated, meaning that an increase in the instantaneous damage rate at any given moment will lead to Increase accordingly; at the same time With sampling time interval It also shows a positive correlation, meaning that under the same instantaneous damage rate sequence, the larger the sampling interval, the larger the estimated cumulative damage value. The rationality and logical design of this formula are reflected in the fact that the composite Simpson numerical integration method is the optimal solution for numerically integrating the instantaneous damage rate sequence. Compared to the rectangular or trapezoidal method, it assigns different weighting coefficients to intermediate sampling points, making the cumulative damage estimate larger with the same number of sampling points. It can achieve higher algebraic accuracy and has smaller integral error for smooth and continuously changing functions. Since the instantaneous damage rate of a magnet under dynamic load spectrum usually changes smoothly and continuously with time, the composite Simpson method can obtain a more accurate equivalent cumulative damage estimate under limited equal time interval sampling conditions, thus providing more accurate input data for establishing the mapping relationship between total experimental damage and magnetic flux loss rate.

[0024] S2: Obtain multiple sets of magnets to form experimental samples, apply a dynamic load spectrum to the experimental samples with temperature and reverse magnetic field strength varying stepwise over time, set the total number of cycles of the dynamic load spectrum, collect the residual magnetic flux of each set of magnets before and after the experiment, and calculate the magnetic flux loss rate. Repeat step S1 for the applied dynamic load spectrum, calculate the equivalent cumulative damage of the magnets in a single sampling period, and calculate the total experimental damage of each set of magnets based on the equivalent cumulative damage. In the above process, the dynamic load spectrum applied to the experimental sample, with temperature and reverse magnetic field strength varying stepwise over time, is typically achieved through a programmable temperature test chamber and electromagnet combined control. Specifically, the stepwise variation refers to dividing a complete cycle of the dynamic load spectrum into several consecutive time periods, i.e., steps. Within each time period, the temperature and reverse magnetic field strength remain constant at their respective set values. However, between different time periods, the temperature and reverse magnetic field strength gradually increase or decrease according to preset step sizes, thus forming a stepwise waveform. For example, the temperature can be set to increase stepwise from 60°C to 80°C, 100°C, and 120°C, remaining at each temperature step for the same amount of time; simultaneously, the reverse magnetic field strength can be set synchronously or asynchronously to increase stepwise from 0.1T to 0.3T and 0.5T, remaining at each magnetic field step for the same or proportional time as the temperature step.

[0025] In this embodiment, the calculation of the magnetic flux loss rate is specifically as follows: subtract the residual magnetic flux after the experiment from the residual magnetic flux before the experiment for each group of magnets, and divide the difference by the residual magnetic flux before the experiment to obtain the true magnetic flux loss rate.

[0026] The flux loss rate is calculated in this way because the formula directly defines the relative attenuation ratio of the magnet's residual magnetic flux after undergoing dynamic load spectrum cycling. Here, the residual magnetic flux before the experiment... As a reference value, it reflects the initial magnetization state of the magnet when it is not under load; the residual magnetic flux after the experiment This reflects the magnetization state of the magnet after it has undergone a set total number of cycles of dynamic load spectrum; the difference between the two is... This is the absolute value of the magnetic flux lost by the magnet due to irreversible demagnetization during the entire experiment. Divide this difference by the remaining magnetic flux before the experiment. This can eliminate the influence of the difference in the absolute value of the initial magnetic flux between different magnets, thus obtaining a dimensionless relative loss rate, which is convenient for horizontal comparison between different magnets and different experimental batches; the larger the value, the more severe the demagnetization of the magnet under the action of the dynamic load spectrum.

[0027] The total experimental damage for each group of magnets was calculated based on the equivalent cumulative damage, specifically as follows: For the dynamic load spectrum where the applied temperature and magnetic field strength change stepwise over time, repeat step S1 to calculate the equivalent cumulative damage for a single sampling period of each experimental sample. Multiply the equivalent cumulative damage of each group of magnets in a single sampling period by the total number of cycles of each group of magnets to obtain the total experimental damage of each group of magnets. The length of the sampling period is set to be equal to the length of a complete cycle in the dynamic load spectrum.

[0028] The above process is handled in this way because the dynamic load spectrum consists of multiple repeating cycles, and the temperature and reverse magnetic field strength change with time in exactly the same way in each cycle. The equivalent cumulative damage per sampling cycle calculated by the instantaneous damage rate model and the composite Simpson numerical integration method is... The total demagnetization damage suffered by the magnet in a complete cycle has been accurately quantified. Since the damage accumulation is linearly additived, that is, the damage in different time periods can be arithmetically added, the total damage suffered by the magnet in the whole experiment can be obtained by multiplying the equivalent cumulative damage of each cycle by the total number of cycles. There is no need to repeatedly integrate the sampled data in each cycle, which greatly simplifies the process of solving the total experimental damage. The sampling period length is set to be equal to the length of a complete cycle in the dynamic load spectrum to ensure that the composite Simpson numerical integration method can precisely cover a complete variation mode in the dynamic load spectrum, thus ensuring the calculated equivalent cumulative damage. This can accurately represent the true damage contribution to the magnet in each cycle. If the sampling period is less than a complete cycle, the instantaneous damage rate sequence within a single sampling period cannot fully capture all the temperature and reverse magnetic field strength variations in that cycle, leading to inaccurate calculations. The sampling period cannot represent the cumulative damage of a single complete cycle, thus making it impossible to obtain the accurate total experimental damage by multiplying it by the total number of cycles. If the sampling period is longer than one complete cycle, the damage of some cycles will be repeatedly included in the integration or unnecessary redundant sampling will be introduced, which will also lead to errors in the cumulative damage calculation. Therefore, setting the sampling period length to be equal to the length of one complete cycle in the dynamic load spectrum is a prerequisite for accurately quantifying the equivalent cumulative damage and efficiently extrapolating it to the total number of cycles.

[0029] S3: Construct a modified mapping function model, pair the flux loss rate of each group of magnets with the corresponding total experimental damage, and fit the mapping function model based on the data pairs to establish the correspondence between the total experimental damage and the flux loss rate. In this embodiment, a modified mapping function model is constructed as follows: The flux loss rate and total experimental damage of each group of magnets were combined to form data pairs. Based on the data, the mapping function model was fitted to establish the correspondence between the total experimental damage and the flux loss rate. in, Indicates the maximum irreversible loss. This represents the damage efficiency coefficient. This represents the nonlinear shape factor, determined by fitting using the nonlinear least squares method. Indicates the total damage in the experiment Flux loss rate at the following conditions; This indicates the total damage in the experiment.

[0030] In the above process, the correspondence between the total experimental damage and the flux loss rate established through the fitting process is actually the specific mathematical expression of the modified mapping function model; specifically, this modified mapping function model... It itself describes the total damage in the experiment. With flux loss rate The mathematical expression for the nonlinear saturation relationship between them, where the maximum irreversible loss is... Damage efficiency coefficient and nonlinear shape factor The specific parameter values ​​are determined by fitting using the nonlinear least squares method; therefore, establishing the correspondence between the total experimental damage and the magnetic flux loss rate is equivalent to completing the construction and parameter calibration of the modified mapping function model, so that the model can be used to describe the mapping law from any total experimental damage input to the corresponding magnetic flux loss rate output. Dependent variable Indicates the total damage in the experiment The flux loss rate under the given conditions, in its physical sense, is the degree of irreversible flux attenuation that occurs after a magnet has accumulated a certain amount of total experimental damage. A larger value indicates a more severe demagnetization of the magnet at that cumulative damage level, while a smaller value indicates a stronger resistance to demagnetization and better flux retention. (Independent variable) It represents the total experimental damage. It quantifies the temperature and reverse magnetic field intensity history in the dynamic load spectrum into a cumulative physical quantity through the instantaneous damage rate model and the composite Simpson numerical integration method. It increases linearly with the increase of the total number of cycles. Independent variable For dependent variable The influence relationship is represented by a saturation exponential function: when the total experimental damage is small, The value is small, the exponential term When the flux loss rate is close to 1, it is close to 1. The value is approximately 0, reflecting the physical phenomenon that the magnet experiences almost no magnetic flux loss in the early stages of damage accumulation (i.e., the damage threshold effect); as the total experimental damage... As the exponential term gradually increases, it gradually decreases and approaches 0, resulting in a higher flux loss rate. Gradually approaching the maximum value This reflects the physical law that after the damage to a magnet accumulates to a certain extent, the reversible demagnetization is exhausted, leaving only the irreversible magnetic flux loss gradually saturating. From the perspective of positive and negative correlation and The positive correlation indicates that the greater the total experimental damage, the higher the magnetic flux loss rate. Moreover, this positive correlation exhibits an S-shaped characteristic of "first rising slowly, then rising rapidly, and then approaching saturation," which is consistent with the actual evolution law of the demagnetization process of neodymium iron boron permanent magnets under thermomagnetic coupling stress. The rationality of the formula's form and the logic of its design are reflected in: the adoption of... This modified exponential saturation function form allows three parameters to control the asymptotic maximum value of the damage curve (maximum irreversible loss). Damage accumulation efficiency (damage efficiency coefficient) ) and the nonlinear bending shape of the damage curve (nonlinear shape factor) Compared with linear fitting or single exponential models, it has stronger expressive power and can accurately fit the nonlinear saturation relationship between the total experimental damage and magnetic flux loss rate of NdFeB permanent magnets of different grades and geometries under dynamic load spectrum, thus providing a reliable mapping basis for predicting the long-term demagnetization rate of the magnet under test based on monitoring data in step S4.

[0031] The relationship between total experimental damage and flux loss rate was established as follows: The mapping function model is fitted using a nonlinear least squares method, with the Levenberg-Marquardt algorithm used for iterative solution until the change in the sum of squared residuals between the predicted flux loss rate and the experimentally measured actual flux loss rate is less than a preset convergence threshold.

[0032] In the above process, the maximum irreversible loss is determined by fitting using the nonlinear least squares method. Damage efficiency coefficient and nonlinear shape factor The specific process is as follows: First, the multiple sets of experimental data obtained in step S2, i.e., the total experimental damage of each magnet, are... and its corresponding true flux loss rate (in , (where the number of experimental groups is used as the input data pairs); then, the objective function of the sum of squared residuals is constructed. Its physical meaning is the sum of squares of the deviations between the model-predicted flux loss rate and the experimentally measured actual flux loss rate; the goal of fitting is to find a set of optimal parameter combinations. This makes the sum of squares of the residuals... The objective function reaches its minimum value. This is because the objective function has respect to the parameters... It is nonlinear and cannot be solved analytically by direct differentiation. Therefore, a numerical optimization algorithm is used for iterative solution, specifically the Levenberg-Marquardt algorithm. This algorithm first... Set an initial value (usually 1), and in each iteration, calculate the residual vector based on the current parameters. Its weight is ) and Jacobi matrix That is, each residual component pairs The partial derivatives are then obtained by introducing a damping factor. Adaptively update parameter values: decrease the damping factor as the prediction error decreases. To accelerate convergence; increase the damping factor as the prediction error increases. Then recalculate a more conservative parameter adjustment; repeat this process iteratively until the sum of squared residuals from two consecutive iterations is obtained. The change is less than the preset convergence threshold, for example... At this point, we get This is the optimal fitting result; Set the convergence threshold to This means that when the relative change in the sum of squared residuals between two consecutive iterations is less than one part per million, the deviation between the model-predicted flux loss rate and the experimentally measured actual flux loss rate is considered to have basically stabilized, and further iterations can be performed on the parameters. , , The correction amount is negligible, and its impact on the final fitting result is insignificant. This order of magnitude avoids the problem of an excessively large threshold. This avoids the problems of premature algorithm termination and insufficient fitting accuracy, while also preventing issues caused by excessively small thresholds. This leads to excessive iteration in the algorithm, consuming unnecessary computational resources, and also affects the numerical precision of double-precision floating-point operations (approximately...). to Maintaining an appropriate distance ensures that the convergence criterion is not affected by rounding errors. It is a reasonable convergence threshold widely used in this type of nonlinear least squares fitting problem.

[0033] S4: Calculate the equivalent cumulative damage of the magnet under test in a single sampling period according to step S1, determine the total number of cycles expected to be used by the magnet under test, multiply the equivalent cumulative damage by the total number of cycles to obtain the total experimental damage, and obtain the predicted magnetic flux loss rate of the magnet under test according to the correspondence between the total experimental damage and the magnetic flux loss rate.

[0034] In this embodiment, the predicted flux loss rate of the magnet to be tested is obtained as follows: Obtain the total number of cycles of the magnet to be tested and its expected service life. Repeat step S1 to calculate the equivalent cumulative damage within a single sampling period. Multiply the equivalent cumulative damage within a single sampling period by the total number of cycles of expected service life to obtain the total experimental damage. Based on the correspondence between the total experimental damage and the flux loss rate, obtain the predicted flux loss rate of the magnet to be tested. in, This represents the predicted flux loss rate of the magnet under test; This represents the total experimental damage to the magnet being tested.

[0035] In the above process, in the prediction formula of this step, the dependent variable It represents the predicted flux loss rate of the magnet under test, and its physical meaning is the degree of irreversible flux attenuation that occurs in the magnet during its expected total number of service cycles due to the coupling effect of dynamic temperature and reverse magnetic field. The larger the value, the more severe the demagnetization of the magnet at the end of its service life; conversely, the smaller the value, the better the magnet can maintain its magnetic flux. Independent variable The total experimental damage to the magnet under test is represented by multiplying the equivalent cumulative damage within a single sampling period by the expected total number of service cycles, thus quantifying the total cumulative damage the magnet suffers over its entire service life. Independent variable For dependent variable The influence relationship is determined by the modified mapping function model that was fitted in step S3. To reflect: the maximum irreversible loss in this model Damage efficiency coefficient and nonlinear shape factor The intrinsic mapping law between the total experimental damage and flux loss rate of this type of neodymium iron boron permanent magnet under thermomagnetic coupling stress has been determined by fitting multiple sets of experimental samples using the nonlinear least squares method. From the perspective of positive and negative correlation and A positive correlation exists—the longer the total number of expected service cycles of the magnet under test, or the greater the equivalent cumulative damage within a single sampling cycle, the greater the total experimental damage. The larger the value, the higher the corresponding predicted flux loss rate. The higher the value, the better; and this positive correlation shows an S-shaped characteristic of "first rising slowly, then rising rapidly, and then approaching saturation", which is consistent with the physical law that magnetic flux loss is not obvious in the early stage of damage accumulation, demagnetization is accelerated in the middle stage, and irreversible loss gradually saturates in the later stage. The rationality and logical design of this formula are reflected in the fact that it reuses the high-precision mapping relationship established in step S3, eliminating the need for destructive experiments on the magnet under test. It only requires real-time acquisition of temperature and reverse magnetic field strength data within a single sampling period to calculate the equivalent cumulative damage within that period, which is then multiplied by the expected total number of service cycles to obtain the total experimental damage. This allows for the rapid prediction of the flux loss rate at the end of long-term service by substituting the data into the saturation exponential function, thus achieving low-cost and high-efficiency prediction of the long-term demagnetization rate from short-term monitoring data.

[0036] In the above embodiments, 20 sets of data on the total experimental damage of the magnet under test and the corresponding predicted flux loss rate are given to reflect the change of the predicted flux loss rate with the change of the total experimental damage of the magnet under test, as shown in Table 1: Table 1: Relationship between total experimental damage to the magnet under test and the corresponding predicted flux loss rate Table 1 above gives the maximum irreversible loss. This means that the magnet loses a maximum of 35% of its magnetic flux under the complete demagnetization limit, which is consistent with the irreversible demagnetization range of actual NdFeB permanent magnets under the action of a high-temperature reverse magnetic field, and the damage efficiency coefficient is... Nonlinear shape factor As can be seen from the data in the table above, the predicted flux loss rate... With the total experimental damage of the magnet to be tested The value increases monotonically with increasing intensity, exhibiting an S-shaped characteristic of "first rising slowly, then rising rapidly, and then approaching saturation": when When increasing from 0.05 to 0.25, The increase from 0.00 to 0.05 was relatively slow; when When it increases from 0.25 to 0.60, The value rapidly increased from 0.05 to 0.22, indicating an accelerated demagnetization phase; when After exceeding 0.70, The growth rate is gradually slowing down, and is gradually approaching the maximum irreversible loss. This trend aligns with the actual physical laws governing the accumulation of damage and flux loss in neodymium iron boron permanent magnets under thermomagnetic coupling stress.

[0037] Please see Figure 3 The present invention also provides a system for predicting the demagnetization rate of a neodymium iron boron permanent magnet, wherein the system for predicting the demagnetization rate of a neodymium iron boron permanent magnet is used to perform the above-described method for predicting the demagnetization rate of a neodymium iron boron permanent magnet, comprising: Data acquisition module: Used to set the sampling period, collect temperature and magnetic field strength data of the magnet at equal time intervals within the sampling period, construct an instantaneous damage rate model of the magnet based on the temperature and magnetic field strength data, calculate the instantaneous damage rate of the magnet at any sampling moment, and construct an instantaneous damage rate sequence of the magnet accordingly; use the composite Simpson numerical integration method on the instantaneous damage rate sequence to calculate the equivalent cumulative damage of the magnet within a single sampling period; Damage calculation module: used to acquire multiple groups of magnets to form experimental samples, apply a dynamic load spectrum to the experimental samples with temperature and reverse magnetic field strength varying stepwise over time, set the total number of cycles of the dynamic load spectrum, collect the residual magnetic flux of each group of magnets before and after the experiment, calculate the magnetic flux loss rate, repeat step S1 for the applied dynamic load spectrum, calculate the equivalent cumulative damage of the magnets in a single sampling period, and calculate the total experimental damage of each group of magnets based on the equivalent cumulative damage. Function mapping module: used to construct the modified mapping function model, to form a data pair between the magnetic flux loss rate of each group of magnets and the corresponding total experimental damage, and to fit the mapping function model based on the data pair to establish the correspondence between the total experimental damage and the magnetic flux loss rate; The prediction module is used to calculate the equivalent cumulative damage of the magnet under test in a single sampling period according to step S1, determine the total number of cycles expected to be used by the magnet under test, multiply the equivalent cumulative damage by the total number of cycles to obtain the total experimental damage, and obtain the predicted magnetic flux loss rate of the magnet under test according to the correspondence between the total experimental damage and the magnetic flux loss rate.

[0038] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0039] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0040] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for predicting the demagnetization rate of a neodymium iron boron permanent magnet, characterized in that, include: S1: Set the sampling period, collect temperature data and magnetic field strength data of the magnet at equal time intervals within the sampling period, and construct the instantaneous damage rate model of the magnet based on the temperature data and magnetic field strength data. Calculate the instantaneous damage rate of the magnet at any sampling moment, and construct the instantaneous damage rate sequence of the magnet accordingly. Use the composite Simpson numerical integration method on the instantaneous damage rate sequence to calculate the equivalent cumulative damage of the magnet within a single sampling period. S2: Obtain multiple sets of magnets to form experimental samples, apply a dynamic load spectrum to the experimental samples with temperature and reverse magnetic field strength varying stepwise over time, set the total number of cycles of the dynamic load spectrum, collect the residual magnetic flux of each set of magnets before and after the experiment, and calculate the magnetic flux loss rate. Repeat step S1 for the applied dynamic load spectrum, calculate the equivalent cumulative damage of the magnets in a single sampling period, and calculate the total experimental damage of each set of magnets based on the equivalent cumulative damage. S3: Construct a modified mapping function model, pair the flux loss rate of each group of magnets with the corresponding total experimental damage, and fit the mapping function model based on the data pairs to establish the correspondence between the total experimental damage and the flux loss rate. S4: Calculate the equivalent cumulative damage of the magnet under test in a single sampling period according to step S1, determine the total number of cycles the magnet under test is expected to serve, multiply the equivalent cumulative damage by the total number of cycles to obtain the total experimental damage, and obtain the predicted flux loss rate of the magnet under test based on the correspondence between the total experimental damage and the flux loss rate; Construct the revised mapping function model, specifically as follows: The flux loss rate and total experimental damage of each group of magnets were combined to form data pairs. Based on the data, the mapping function model was fitted to establish the correspondence between the total experimental damage and the flux loss rate. in, Indicates the maximum irreversible loss. This represents the damage efficiency coefficient. This represents the nonlinear shape factor, determined by fitting using the nonlinear least squares method. Indicates the total damage in the experiment Flux loss rate at the following conditions; This indicates the total damage in the experiment.

2. The method for predicting the demagnetization rate of a neodymium iron boron permanent magnet according to claim 1, characterized in that, The instantaneous damage rate model of the magnet is constructed as follows: Based on temperature and magnetic field strength data, an instantaneous damage rate model for the magnet is constructed, and the instantaneous damage rate of the magnet at any sampling time is calculated: in, Indicates the pre-factor, The kinetic index is obtained by fitting multiple linear regression data from accelerated aging experiments using an isothermal constant magnetic field. This represents the magnetocrystalline anisotropy field of the magnet at the reference temperature; This represents the temperature coefficient of an anisotropic field; Indicates the average activation energy; Indicates the permeability of free space; Indicates saturation magnetization; Represents the Boltzmann constant; Indicates reference temperature; Indicates the sampling time index; Indicates the first Instantaneous damage rate of the magnet at each sampling moment; Indicates in The strength of the reverse magnetic field used for the magnet at any given time; Indicates in The absolute temperature of the magnet at any given moment; Indicates in The temperature softening function corresponding to the absolute temperature of the magnet at any given moment; The isothermal constant magnetic field accelerated aging experiment specifically involves: using multiple neodymium iron boron permanent magnet samples, maintaining the temperature of each sample at a constant set value during the experiment, and simultaneously applying a constant reverse magnetic field for a period of time; measuring the residual magnetic flux of the neodymium iron boron permanent magnet samples before and after the experiment to calculate the magnetic flux loss rate under the given temperature and magnetic field conditions; then, repeating the experiment by changing different constant temperature levels and different constant reverse magnetic field strength levels to obtain a series of temperature-magnetic field-magnetic flux loss rate data combinations. Since both temperature and magnetic field remain constant over time, the instantaneous damage rate in the instantaneous damage rate model is a constant. Therefore, the cumulative damage equals the instantaneous damage rate multiplied by the experimental time. Taking the logarithm of both sides of the instantaneous damage rate model, with the logarithm of the cumulative damage measured in the isothermal constant magnetic field accelerated aging experiment as the dependent variable and the reciprocal of temperature and the magnetic field correlation function as independent variables (the magnetic field correlation function refers to the natural logarithm of the reverse magnetic field strength and the ratio of the reverse magnetic field strength to the absolute temperature), a multiple linear regression is used for fitting, thereby simultaneously determining the antecedent factors. and dynamic index The value.

3. The method for predicting the demagnetization rate of a neodymium iron boron permanent magnet according to claim 2, characterized in that, The equivalent cumulative damage of the magnet during a single sampling period is calculated as follows: Temperature data of the magnet at each sampling time and magnetic field strength data Substituting these values ​​into the instantaneous damage rate model yields the complete instantaneous damage rate sequence. ,in, This represents the number of sampling points within a single sampling period. An index indicating the sampling time; Indicates the first Instantaneous damage rate at each sampling time; The equivalent cumulative damage to the magnet within a single sampling period is calculated using the composite Simpson numerical integration method on the instantaneous damage rate sequence. in, Indicates the sampling time interval; Indicates equivalent cumulative damage: This represents the instantaneous damage rate of the magnet at the first sampling time. Indicates the first Instantaneous damage rate of the magnet at each sampling moment; Indicates the index of even-numbered sampling times; Indicates the index of the odd-numbered sampling time; Indicates the first Instantaneous damage rate of the magnet at each sampling moment; Indicates the first Instantaneous damage rate of the magnet at each sampling time.

4. The method for predicting the demagnetization rate of a neodymium iron boron permanent magnet according to claim 1, characterized in that, The calculation of the magnetic flux loss rate is specifically as follows: subtract the residual magnetic flux after the experiment from the residual magnetic flux before the experiment for each group of magnets, and divide the difference by the residual magnetic flux before the experiment to obtain the true magnetic flux loss rate.

5. The method for predicting the demagnetization rate of a neodymium iron boron permanent magnet according to claim 3, characterized in that, The total experimental damage for each group of magnets was calculated based on the equivalent cumulative damage, specifically as follows: For the dynamic load spectrum where the applied temperature and magnetic field strength change stepwise over time, repeat step S1 to calculate the equivalent cumulative damage for a single sampling period of each experimental sample. Multiply the equivalent cumulative damage of each group of magnets in a single sampling period by the total number of cycles of each group of magnets to obtain the total experimental damage of each group of magnets. The length of the sampling period is set to be equal to the length of a complete cycle in the dynamic load spectrum.

6. The method for predicting the demagnetization rate of a neodymium iron boron permanent magnet according to claim 1, characterized in that, The relationship between total experimental damage and flux loss rate was established as follows: The mapping function model is fitted using a nonlinear least squares method, with the Levenberg-Marquardt algorithm used for iterative solution until the change in the sum of squared residuals between the predicted flux loss rate and the experimentally measured actual flux loss rate is less than a preset convergence threshold.

7. The method for predicting the demagnetization rate of a neodymium iron boron permanent magnet according to claim 1, characterized in that, The predicted flux loss rate of the magnet to be tested is obtained as follows: Obtain the total number of cycles of the magnet to be tested and its expected service life. Repeat step S1 to calculate the equivalent cumulative damage within a single sampling period. Multiply the equivalent cumulative damage within a single sampling period by the total number of cycles of expected service life to obtain the total experimental damage. Based on the correspondence between the total experimental damage and the flux loss rate, obtain the predicted flux loss rate of the magnet to be tested. in, This represents the predicted flux loss rate of the magnet under test. This represents the total experimental damage to the magnet being tested.

8. A system for predicting the demagnetization rate of a NdFeB permanent magnet, said system for implementing the method for predicting the demagnetization rate of a NdFeB permanent magnet according to any one of claims 1-7, characterized in that, include: Data acquisition module: Used to set the sampling period, collect temperature and magnetic field strength data of the magnet at equal time intervals within the sampling period, construct an instantaneous damage rate model of the magnet based on the temperature and magnetic field strength data, calculate the instantaneous damage rate of the magnet at any sampling moment, and construct an instantaneous damage rate sequence of the magnet accordingly; use the composite Simpson numerical integration method on the instantaneous damage rate sequence to calculate the equivalent cumulative damage of the magnet within a single sampling period; Damage calculation module: used to acquire multiple groups of magnets to form experimental samples, apply a dynamic load spectrum to the experimental samples with temperature and reverse magnetic field strength varying stepwise over time, set the total number of cycles of the dynamic load spectrum, collect the residual magnetic flux of each group of magnets before and after the experiment, calculate the magnetic flux loss rate, repeat step S1 for the applied dynamic load spectrum, calculate the equivalent cumulative damage of the magnets in a single sampling period, and calculate the total experimental damage of each group of magnets based on the equivalent cumulative damage. Function mapping module: used to construct the modified mapping function model, to form a data pair between the magnetic flux loss rate of each group of magnets and the corresponding total experimental damage, and to fit the mapping function model based on the data pair to establish the correspondence between the total experimental damage and the magnetic flux loss rate; The prediction module is used to calculate the equivalent cumulative damage of the magnet under test in a single sampling period according to step S1, determine the total number of cycles expected to be used by the magnet under test, multiply the equivalent cumulative damage by the total number of cycles to obtain the total experimental damage, and obtain the predicted magnetic flux loss rate of the magnet under test according to the correspondence between the total experimental damage and the magnetic flux loss rate.

Images