Permanent magnet synchronous motor temperature monitoring and demagnetization early warning method based on chaotic signal injection and multi-parameter fusion

Abstract

The invention discloses a permanent magnet synchronous motor temperature monitoring and demagnetization early warning method based on chaotic signal injection and multi-parameter fusion. Comprising the steps of 1, generating a wide-spectrum chaotic signal based on a Lorenz chaotic system, and injecting the wide-spectrum chaotic signal into a d-axis voltage channel of a d-q-axis mathematical model of the permanent magnet synchronous motor; 2, identifying stator resistance RS and d-axis inductance Ld by using a least square method; 3, a Kalman state space model with the winding temperature Tw and the magnetic steel temperature Tm as state variables is constructed, the resistance Rs and the inductance Ld which are measured in real time are fused, the observation noise covariance R is adjusted in a self-adaptive mode, and the winding temperature Tw and the magnetic steel temperature Tm are output in a decoupling mode; step 4, establishing a motor back electromotive force model to calculate a real-time flux linkage psi f, and calculating a motor temperature rise rate # imgabs0 # through a sliding window weighted average method; and step 5, performing demagnetization risk grading early warning according to the flux linkage and the temperature rise rate of the motor.

Description

Technical Field

[0001] The present invention belongs to the fields of permanent magnet synchronous motor control and fault diagnosis, and relates to a parameter identification method for a permanent magnet synchronous motor. Background Art

[0002] Permanent magnet synchronous motors (PMSMs) have the advantages of high efficiency, high power factor, measurable rotor parameters, and good control performance, and have received extensive attention and applications in the academic and industrial fields. They are widely used in electric vehicles, industrial servo systems, wind power generation, aerospace and other fields. However, during the long-term operation of permanent magnet synchronous motors, problems such as motor overheating and permanent magnet demagnetization will be faced. These problems will affect the performance of the motor, especially when the motor starts and stops frequently, stalls, or is under high load, it is easy to cause the motor temperature to rise and trigger demagnetization. In the prior art, the motor temperature monitoring method usually installs temperature sensors on the stator winding or the motor housing. The temperature changes in different regions such as the winding and magnetic steel of the motor are uneven. Limited by the space of the motor, the sensors can usually only cover part of the region and cannot comprehensively and accurately reflect the temperature change of the motor. When the motor temperature rises to the demagnetization critical threshold, it may trigger an alarm only after demagnetization due to sensor monitoring lag or position limitation. In high-performance application scenarios, real-time and accurate temperature rise monitoring and demagnetization early warning are crucial for the safe operation of the motor. In the prior art, there is also a sensorless monitoring method that indirectly measures the motor temperature by using the electrical parameters of the motor (such as resistance, inductance, etc.). This method has no lag in response compared with sensors, but its effect depends on the accurate identification of electrical parameters. In practical applications, the identification of resistance and inductance is often affected by various factors, which will lead to parameter identification errors. Moreover, inductance and resistance are usually coupled, and a single electrical parameter is difficult to accurately reflect the temperature of the motor. In addition, the demagnetization phenomenon is a key fault mode of permanent magnet synchronous motors. Timely monitoring of demagnetization and issuing an early warning are crucial for preventing motor damage. Some demagnetization monitoring methods usually judge whether the motor is demagnetized based on a fixed magnetic flux threshold, and infer the demagnetization state by monitoring the magnetic field of the motor and comparing it with the fixed magnetic flux threshold. When the motor operates under dynamic conditions such as load mutation or environmental temperature change, the temperature fluctuates greatly, and the fixed magnetic flux threshold often cannot accurately reflect the actual demagnetization risk. There are problems such as lag in the demagnetization early warning mechanism and lack of hierarchical early warning.

[0003] Therefore, there is an urgent need for a temperature monitoring method without lag and multiple regions, as well as a demagnetization early warning mechanism without lag and multiple levels. Summary of the Invention

[0004] To solve the above problems, the present invention provides a method for temperature monitoring and demagnetization warning of a permanent magnet synchronous motor based on chaotic signal injection and multi-parameter fusion, which can monitor the temperature of key areas such as the motor winding and permanent magnet accurately and in real time, and classify and warn the demagnetization risk of the motor. The motor adopted in the present invention is a surface-mounted permanent magnet synchronous motor. The technical solution provided by the present invention includes the following steps:

[0005] Step 1: Generate a chaotic signal based on the Lorenz chaotic system, modulate it and inject it into the d-axis voltage channel of the dq-axis mathematical model of the permanent magnet synchronous motor;

[0006] Step 2: Use the least squares method to extract the stator resistance R s and the d-axis inductance L d ;

[0007] Step 3: Construct a Kalman state space model with the winding temperature T w and the permanent magnet temperature T m as state variables, fuse the real-time measured resistance R s and inductance L d and adaptively adjust the observation noise covariance R. Decouple and output the winding temperature T w and the permanent magnet temperature T m ;

[0008] Step 4: Establish a motor back electromotive force model to calculate the real-time magnetic flux Ψ f , and calculate the temperature rise rate through the sliding window weighted average method

[0009] Step 5: Perform demagnetization risk classification and warning according to the magnetic flux and temperature rise rate of the motor.

[0010] In a possible implementation manner, the step 1 includes:

[0011] Generate a chaotic signal, which can cover both low-frequency and high-frequency ranges at the same time, can effectively improve the signal-to-noise ratio of the resistance and inductance parameters of the permanent magnet synchronous motor, thus contributing to the accurate identification of the two parameters of inductance and resistance, and having a certain robustness at the same time. First, use the fourth-order Runge-Kutta method (PK4) to solve the Lorenz differential equations to generate a signal with chaotic characteristics. Since the high-frequency components of the chaotic signal generated by the Lorenz system are less, the generated chaotic signal needs to be modulated with a high-frequency carrier to extend its spectrum to the high-frequency range. To retain the low-frequency components of the original signal, the chaotic signal is then superimposed with the modulated signal to ensure that the finally injected chaotic signal contains rich low-frequency and high-frequency components. Finally, inject it under the d-axis voltage of the permanent magnet synchronous motor FOC control.

[0012]

[0013] Wherein is the output voltage of the d-axis current loop PI controller, X inj (t1) is the generated chaotic signal, S inj (t2) is the modulated high-frequency chaotic signal, and k is the chaotic signal injection gain. is the d-axis voltage signal finally injected into the system.

[0014] The Lorenz differential equations are as follows:

[0015]

[0016] The selected classical chaotic parameters in the formula are σ = 10, ρ = 28, β = 8 / 3, where σ is the Prandtl number, ρ is the Rayleigh number, and the initial values of x, y, z are (x0 = 1, y0 = 1, z0 = 1). By numerically solving the above differential equations, the obtained solutions show a chaotic phenomenon distribution.

[0017] The modulated high-frequency carrier is:

[0018] s inj (t) = A·x(t)·cos(2πf c t)

[0019] Wherein x(t) is the chaotic signal generated by the Lorenz equation, A is the signal amplitude gain, and f c is the carrier frequency.

[0020] In a possible implementation manner, the step 2 includes:

[0021] Establish a dq-axis permanent magnet synchronous motor mathematical model.

[0022]

[0023] Wherein, L d and L q are the d-axis and q-axis inductances respectively, u d and u q are the stator d-axis and q-axis voltages respectively, i d and i q are the stator d-axis and q-axis currents respectively, R s is the stator resistance, ψ f is the permanent magnet flux linkage, and ω is the electrical angular velocity of the motor.

[0024] The operating state of the permanent magnet synchronous motor will change with the injection of the chaotic signal, there are multiple operating points, and it is easy to reach the full-rank condition. A new set of equations can be constructed for the adjacent two sampling results of the d-axis voltage. Use the least squares method to identify the real-time stator resistance R s and inductance L d。The motor used in this invention is a surface-mounted permanent magnet synchronous motor, which has the characteristic of L d = L q . Construct the following new equations:

[0025]

[0026] In the formula, the subscript "1" represents the nth sampling result, and the subscript "2" represents the (n + 1)th sampling result. After the chaotic signal is injected into the d-axis, the equation for identification is

[0027]

[0028] By the RLS algorithm, the identification values of R s and L d can be obtained in real time according to the following formula

[0029]

[0030] In the formula is the branch weight matrix, d is the expected response matrix, and u T is the branch input matrix.

[0031] In a possible implementation manner, the step 3 includes:

[0032] (1) After completing the step 2, let the winding temperature T w and the magnet temperature T m , as state variables, and the observed values are R s , L d . Construct the mapping model between temperature and observed values,

[0033]

[0034] In the formula, α is the winding temperature coefficient and β is the magnet temperature coefficient, T0 is the temperature calibrated by the motor at room temperature, and R s0 and L d0 are the resistance and inductance of the permanent magnet synchronous motor measured at room temperature respectively.

[0035] The state space of Kalman is defined as: x(k) = [T w (k), T m (k)] T , and the observation equation is: z(k) = Hx(k) + v(k)

[0036] In the above formula, z(k) is the observation vector, which contains the resistance and inductance information: H: Observation matrix, which describes the relationship between the state and the observed value: α is the sensitivity of the resistance to temperature, β is the sensitivity of the inductance to temperature, and v(k): Observation noise (covariance matrix R).

[0037] According to the inertial characteristics of temperature change, the following state equation is established:

[0038] x(k)=Ax(k-1)+Bu(k)+w(k)

[0039] Where A is the transfer matrix, Bu(k) is the external input term, and w(k) is the process noise (covariance matrix Q). Its value is a fixed value.

[0040] (2) The observed noise covariance R is determined by the identified resistance and inductance errors:

[0041]

[0042] Under dynamic conditions (such as variable speed and variable load), the resistance and inductance parameters of the motor are not constant, and the confidence of the resistance and inductance will fluctuate with the change of working conditions. If the fixed observation noise covariance matrix R is still used in the process of data fusion of motor parameters, the actual changes of resistance and inductance parameters will not be accurately reflected. Because the fixed observation noise does not take into account the identification error of the resistance and inductance parameters of the motor under dynamic conditions, this will affect the accuracy of data fusion of the motor under dynamic conditions, and thus affect the accuracy of motor temperature monitoring. In order to improve the accuracy of data fusion and the robustness of temperature monitoring, it is necessary to adaptively adjust the observation noise covariance R to adapt to the parameter identification error caused by uncertainty factors in the process of motor parameter identification, thereby causing temperature detection error. Current i q It can well characterize the degree of load disturbance, i q Significant changes are often accompanied by unstable growth of system noise and measurement errors. q The observation noise covariance R is dynamically adjusted according to the change of , thereby reducing the errors of resistance and inductance in the data fusion process and improving the accuracy of temperature detection.

[0043] When the load is high, the winding temperature rises quickly. s Significant change, need to reduce At low speed: Inductance L d Susceptible to magnetic saturation interference, increasing Therefore, the adaptive adjustment formula of the observation noise covariance R is:

[0044]

[0045] In the formula, The initial standard deviation of the resistance measurement noise, Initial standard deviation of inductance measurement noise, I q (k) represents the q-axis current at the kth moment. q(k) increases, i.e., when the motor is under high load, decreases, indicating an increase in the confidence level of the resistance measurement and an increase in the weight. I d I(k) represents the d-axis current at the k-th moment. When I d (k) increases, i.e., when the motor is under magnetic saturation interference, increases, indicating a decrease in the confidence level of the inductance measurement and a decrease in the weight.

[0046] In a possible implementation manner, the step 4 includes:

[0047] (1) After completing the step 3, establish a back electromotive force model to calculate the real-time magnetic flux. According to the resistance R identified in the step 2 s and the inductance L d the online calculation of the magnetic flux can be realized and the calculated magnetic flux can be ensured to have certain robustness and accuracy. The back electromotive force model is:

[0048]

[0049] In the formula: V q is the q-axis voltage, I q , I d are the d-q axis currents, ω e is the electrical angular velocity, Ψ f (k) is the calculated real-time magnetic flux.

[0050] Since the temperature change of the motor will cause a certain change in the critical magnetic flux when the permanent magnet synchronous motor demagnetizes. Therefore, a fixed magnetic flux threshold cannot be used as the critical standard for judging whether the motor demagnetizes. It is necessary to consider the influence of temperature on the magnetic flux threshold and judge whether the motor demagnetizes by dynamically adjusting the magnetic flux threshold. It can more accurately reflect whether the motor demagnetizes and avoid misjudgment of demagnetization caused by the change of the magnetic flux threshold due to temperature fluctuation.

[0051] Establish a relationship model between the critical magnetic flux and temperature when the motor demagnetizes, and determine the dynamic magnetic flux threshold:

[0052]

[0053] In the formula, γ is the magnetic flux temperature coefficient, T m , the real-time temperature of the permanent magnet, Ψ f0 is the magnetic flux reference value measured at room temperature T0, is the magnetic flux threshold.

[0054] (2) To make the demagnetization judgment more accurate and avoid misjudgment of demagnetization caused by uncertain factors. Therefore, the temperature rise rate in the permanent magnet region of the permanent magnet synchronous motor is combined to judge whether the motor is demagnetized. If the motor temperature rises sharply in a short period of time, it indicates that the motor is overloaded or the current is too large. It is easy for the local temperature of the motor to rise, which may lead to irreversible local demagnetization.

[0055] The method of using a sliding window is adopted to calculate the temperature rise rate of the motor:

[0056]

[0057] In the formula: N is the window length, w i is the weight coefficient, and Δt is the sampling period. According to the material characteristics of the motor, the temperature rise rate of the motor is set. Under special working conditions, the temperature rise rate of the motor is allowed to increase briefly.

[0058] In a possible implementation manner, the step 5 includes:

[0059] When the motor is facing the demagnetization trend, the magnetic flux of the motor will continue to decrease. Once it drops below the demagnetization critical value of the set motor magnetic flux, the motor will undergo irreversible demagnetization. Therefore, in the fourth step, when the magnetic flux drops below 0.95 times of the set magnetic flux, that is this condition, it is preliminarily judged that the permanent magnet synchronous motor has a demagnetization risk. The purpose of the magnetic flux being lower than 0.95 times of the set magnetic flux is to prevent false alarms in the motor demagnetization warning.

[0060] When the temperature rise rate reaches this condition, it is also preliminarily judged that the permanent magnet synchronous motor has a demagnetization risk. According to the above motor demagnetization conditions, it can be divided into two-level demagnetization warnings.

[0061] Level 1 warning,

[0062] When the temperature rise rate or the magnetic flux reaches the motor demagnetization condition, the system will first issue a warning signal, but will not immediately make a demagnetization judgment. The core of this approach is to avoid the risk of false alarms or missed alarms caused by premature judgment through early warning. Through this level 1 warning method, people can timely identify and respond to temporary faults or abnormal working conditions that occur during the operation of the motor. For example, when the motor is blocked or in a high-load state. The temperature rise rate rises sharply, but it does not mean that the motor is demagnetized at this time, and manual intervention measures can be taken to solve this problem.

[0063] Level 2 warning,

[0064] When the temperature rise rate and the magnetic flux value reach the motor demagnetization judgment conditions simultaneously, the system will trigger a demagnetization alarm and can automatically take corresponding protection measures. This secondary warning method effectively improves the accuracy of demagnetization identification. Compared with the above-mentioned primary warning method, through the cross-verification of the two key indicators of the temperature rise rate and the magnetic flux, it can not only reduce the occurrence of false alarms, but also issue a warning in time and start the corresponding protection measures to ensure the safe operation of the motor and the load it drives when the demagnetization risk actually occurs.

[0065] In summary, the solution provided by the present invention realizes the parameter identification of the resistance and inductance of the permanent magnet synchronous motor, monitors the temperature of key areas such as the motor winding and the permanent magnet in real time, and simultaneously warns of the demagnetization risk of the motor. Injecting chaotic signals improves the accuracy of parameter identification. This method can accurately capture the dynamic behavior of the motor and improve the robustness and accuracy of system identification in an unstable or noisy environment. Using Kalman filtering to fuse the inductance and resistance parameters and decouple the temperatures of the motor winding and the permanent magnet can avoid inaccurate monitoring of the motor temperature by a single parameter. By calculating the dynamic threshold of the magnetic flux and the temperature rise rate of the permanent magnet, the demagnetization risk of the motor can be identified in time and accurately, and hierarchical warnings can be issued. Thereby avoiding motor failures and improving the safety and reliability of the motor. Brief Description of the Drawings

[0066] Figure 1 is a schematic flow chart of a method for temperature monitoring and demagnetization warning of a permanent magnet synchronous motor based on chaotic signal injection and multi-parameter fusion of the present application

[0067] Figure 2 is a system control block diagram of a method for temperature monitoring and demagnetization warning of a permanent magnet synchronous motor based on chaotic signal injection and multi-parameter fusion of the present application Detailed Embodiment

[0068] In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the description herein is only used to explain the present invention and is not used to limit the present invention. In addition, the present invention covers any alternatives, modifications, equivalent methods and solutions made within the spirit and scope of the present invention defined by the claims.

[0069] As Figure 1 shown, a method for temperature monitoring and demagnetization warning of a permanent magnet synchronous motor based on chaotic signal injection and multi-parameter fusion according to an embodiment of the present invention includes:

[0070] Step 1: Generate chaotic signals based on the Lorenz chaotic system, and inject them into the d-axis voltage channel of the dq-axis mathematical model of the permanent magnet synchronous motor after modulation;

[0071] Step 2: Use the least squares method to extract the stator resistance Rs and d-axis inductance L d ;

[0072] Step 3: Construct a Kalman state space model with winding temperature T w and permanent magnet temperature T m as state variables, fuse the real-time measured resistance R s and inductance L d and adaptively adjust the observation noise covariance R. Decouple and output the winding temperature T w and permanent magnet temperature T m ;

[0073] Step 4: Establish a motor back electromotive force model to calculate the real-time magnetic flux Ψ f , and calculate the temperature rise rate through the sliding window weighted average method

[0074] Step 5: Perform demagnetization risk classification and early warning according to the magnetic flux and temperature rise rate of the motor.

[0075] In this embodiment, first, chaotic signal injection is used in the permanent magnet synchronous motor for parameter identification of resistance and inductance, aiming to ensure the accuracy and robustness of the identified parameters. Secondly, mapping models of resistance, inductance with the motor winding, and permanent magnet with temperature are established respectively. The Kalman filter is used to fuse the resistance and inductance to decouple the temperatures of areas such as the motor winding and permanent magnet, aiming to solve the lag effect of sensor-monitored temperature and the instability of single-parameter monitoring of motor temperature. Finally, the dynamic magnetic flux threshold of motor demagnetization and the temperature rise rate of the motor are calculated using the monitored permanent magnet temperature of the motor, and a multi-level demagnetization early warning mechanism is constructed through these two. Ensure the safety and reliability of motor operation.

[0076] As described above, only the preferred specific embodiments of the present invention are given, and the specific data given are only for reference. In actual applications, appropriate changes should be made according to different motors. However, the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. A temperature monitoring and demagnetization warning method for permanent magnet synchronous motors based on chaotic signal injection and multi-parameter fusion, characterized in that, It includes the following steps: Step 1: Generate a chaotic signal based on the Lorenz chaotic system, modulate it, and inject it into the d-axis voltage channel of the dq-axis mathematical model of the permanent magnet synchronous motor; Step 2: Extract the stator resistance R using the least squares method s and the d-axis inductance L d ; Step 3: Construct a Kalman state space model with the winding temperature T w and the permanent magnet temperature T m as state variables, fuse the resistance R s and the inductance L d measured in real time, and adaptively adjust the observation noise covariance R. Decouple and output the winding temperature T w and the permanent magnet temperature T m ; Step 4: Establish a motor back electromotive force model to calculate the real-time magnetic flux Ψ f , and calculate the temperature rise rate by the sliding window weighted average method Step 5: Perform demagnetization risk classification and early warning according to the magnetic flux and temperature rise rate of the motor.

2. Step 1 described in claim 1 includes: Use the fourth-order Runge-Kutta method (PK4) to solve the Lorenz differential equations to generate a signal with chaotic characteristics. Since the high-frequency components of the chaotic signal generated by the Lorenz system are insufficient, it is necessary to modulate it to a high-frequency carrier, and then superimpose the chaotic signal and the modulated signal to ensure that the finally injected chaotic signal contains rich low-frequency and high-frequency components. Finally, it is injected under the d-axis voltage of the permanent magnet synchronous motor under FOC control. The injected signal contains both low-frequency and high-frequency components, which can effectively improve the signal-to-noise ratio of the resistance and inductance parameters of the permanent magnet synchronous motor, realize accurate identification of the parameters and have a certain robustness. where is the output voltage of the d-axis current loop PI controller, X inj (t1) is the generated chaotic signal, S inj (t2) is the modulated high-frequency chaotic signal, k is the chaotic signal injection gain, is the d-axis voltage signal finally injected into the system. The Lorenz differential equations are as follows: In the formula, the classical chaotic parameters are selected as σ = 10, ρ = 28, β = 8 / 3, where σ is the Prandtl number, ρ is the Rayleigh number, and β is the wave number. The initial values of x, y, and z are all taken as 1. The modulated high-frequency carrier is: s inj (t) = A·x(t)·cos(2πf c t) where x(t) is the chaotic signal generated by the Lorenz equation, A is the signal amplitude gain, and f c is the carrier frequency.

3. Step 2 described in claim 1 includes: Improve the signal-to-noise ratio of the motor parameters by injecting chaotic signals, use the d-axis voltage mathematical model of the motor, and then identify the resistance and inductance of the motor by the least squares method. Establish the d-q axis permanent magnet synchronous motor mathematical model: Where, L d and L q are the d-axis and q-axis inductances respectively, u d and u q are the stator d-axis and q-axis voltages respectively, i d and i q are the stator d-axis and q-axis currents respectively, R s is the stator resistance, ψ f is the permanent magnet flux linkage, and ω is the electrical angular velocity of the motor. The operating state of the permanent magnet synchronous motor will change with the injection of chaotic signals, there are multiple operating points, and it is easy to reach the full-rank condition. A new set of equations can be constructed based on the results of two adjacent samplings of the d-axis voltage. The real-time stator resistance R of the permanent magnet synchronous motor is identified using the least squares method s and inductance L d . The motor used in the present invention is a surface-mounted permanent magnet synchronous motor, which has the characteristic of L d = L q . The new set of equations is constructed as follows: In the formula, the subscript "1" represents the nth sampling result, and the subscript "2" represents the (n + 1)th sampling result. After the chaotic signal is injected into the d-axis, the equation for identification is: With the RLS algorithm, the recognition values of R s and L d can be obtained in real time according to the following formula where is the branch weight matrix, d is the expected response matrix, and u T is the branch input matrix.

4. Step 3 described in claim 1 includes: Utilize the identified parameters such as resistance and inductance to monitor the motor temperature in different regions of the permanent magnet synchronous motor in real time, that is, the winding temperature and the magnet temperature of the motor. Use the state space model of Kalman filtering to define the winding temperature T w and the magnet temperature T m as state variables, fuse the real-time resistance R s and inductance L d parameters, and adaptively adjust the observation noise covariance R. Decouple the winding temperature and the magnet temperature. The observed values are R s , L d . (1) Construct a mapping model between temperature and observed values: where α is the winding temperature coefficient and β is the permanent magnet temperature coefficient, T0 is the temperature calibrated for the motor at room temperature, R s0 and L d0 are the resistance and inductance of the permanent magnet synchronous motor measured at room temperature, respectively. The state space of Kalman is defined as: x(k) = [T w (k), T m (k)] T , and the observation equation is: z(k) = Hx(k) + v(k) In the above equation, z(k) is the observation vector, which contains resistance and inductance information: H: Observation matrix, which describes the relationship between the state and the observed value: α is the sensitivity of resistance to temperature, β is the sensitivity of inductance to temperature, and v(k): Observation noise (covariance matrix R). Establish a state equation according to the inertial characteristics shown by the temperature change: x(k) = Ax(k - 1) + Bu(k) + w(k) where A is the transfer matrix; Bu(k) is the external input term; w(k) is the process noise (covariance matrix Q). Its value is a fixed value, (2) The observation noise covariance R is determined by the identified resistance and inductance errors Under dynamic operating conditions (such as variable speed and variable load), the confidence levels of resistance and inductance may change. Using a fixed observation noise covariance R will lead to errors in data fusion. Therefore, according to the load current i q dynamically adjust the observation noise covariance R. The adaptive adjustment formula is: In the formula, Initial standard deviation of the resistance measurement noise, Initial standard deviation of the inductance measurement noise. At high loads, the winding temperature rises rapidly and the resistance R s changes significantly and needs to be reduced At low speeds, the inductance L d is vulnerable to magnetic saturation interference and increases (Reduce the weight).

5. Step 4 described in claim 1 includes: Calculate the real-time magnetic flux Ψ using existing parameters f and the dynamic magnetic flux threshold, and then calculate the temperature rise rate of the motor permanent magnet by the sliding window weighted average method (1) Establish a back electromotive force model to calculate the real-time magnetic flux, based on the resistance R identified in step 2 of claim 1 s and inductance L d It is possible to realize the online calculation of the magnetic flux and ensure that the calculated magnetic flux has a certain robustness and accuracy. The back electromotive force model is as follows: Where: V q is the q-axis voltage, I q , I q are the d-q axis currents, and ω e is the electrical angular velocity. Since the temperature change of the motor will cause a certain change in the critical magnetic flux when the permanent magnet synchronous motor demagnetizes. Therefore, a fixed magnetic flux threshold cannot be used as the critical standard for judging whether the motor demagnetizes. It is necessary to consider the influence of temperature on the magnetic flux threshold and judge whether the motor demagnetizes by dynamically adjusting the magnetic flux threshold. It can more accurately reflect whether the motor demagnetizes and avoid misjudgment of demagnetization caused by the change of the magnetic flux threshold due to temperature fluctuation. Establish a relationship model between the critical magnetic flux and temperature when the motor demagnetizes, and determine the dynamic magnetic flux threshold: where γ is the temperature coefficient of magnetic flux linkage, T m is the real-time temperature of the permanent magnet, Ψ f0 is the reference value of the magnetic flux linkage measured at the normal temperature T0, is the magnetic flux linkage threshold value. (2) In order to make the demagnetization judgment more accurate and avoid misjudgment of the demagnetization described in the claim due to uncertain factors. Therefore, combined with the temperature rise rate in the magnet area of the permanent magnet synchronous motor to judge whether the motor demagnetizes. If the motor temperature rises sharply in a short time, it means that the motor is overloaded or the current is too large. It is easy to cause local temperature rise of the motor and then irreversible local demagnetization. Adopt the method of sliding window to calculate the temperature rise rate of the motor: where N is the window length, w i is the weighting coefficient, and Δt is the sampling period. According to the motor material characteristics, the motor temperature rise rate is set. Under special working conditions, the motor temperature rise rate is allowed to increase briefly.

6. Step 5 described in claim 1 includes: Based on the magnetic flux and temperature rise rate of the motor, a demagnetization risk grading warning for the permanent magnet synchronous motor is carried out. When the magnetic flux of the motor reaches this condition, it is initially judged that the permanent magnet synchronous motor has a demagnetization risk; when the temperature rise rate reaches this condition, it is also initially judged that the permanent magnet synchronous motor has a demagnetization risk. Level 1 Warning: When the temperature rise rate or magnetic flux reaches the demagnetization condition of the motor described in the claims, the system issues a warning signal but does not immediately judge it as demagnetization. This can provide an early warning, reduce the risk of false negatives, and also serve as a reminder during motor blockage or high load, enabling manual resolution of issues such as motor blockage. Secondary warning: When both the temperature rise rate and magnetic flux reach the demagnetization condition of the motor, the system triggers a demagnetization alarm, allowing protective measures to be taken, avoiding false positives while also ensuring no false negatives.

Images