Method for determining the irreversible demagnetization of the magnets of a rotor of an electric machine

Abstract

A method (1) for determining the state of health of the permanent magnets (10) in the rotor of an electric machine on the basis of a voltage parameter of the electric machine, which is based on a control effect of a control signal as a measure of the induced counter electromotive force.

Description

[0001]Method for determining irreversible demagnetization of the magnets of a rotor of an electrical machine The invention relates to a method for determining irreversible demagnetization of the magnets of a rotor of an electrical machine based on a voltage parameter of the electrical machine. State of the art The central approach of the forecast and health management system is the use of data on usage and condition from the operation of an individual system in order to detect malfunctions and failures at an early stage and to define suitable countermeasures. Decisions can thus be made taking overall economic interests into account. For example, costs can be reduced, risks minimized, usable operating hours maximized, and availability improved simultaneously. Electric vehicles use a purely electric drive system, known as an e-drive.The drive system of an electric vehicle includes an electric motor, a mechanical reduction gear, an inverter, and a motor controller. The inverter inverts DC power from batteries into AC power, which is used to drive the motor. The motor supplies torque to a vehicle drivetrain. The controller controls the power delivered to the motor by the inverter. Compared to internal combustion engines, EM electric motors typically require less but more careful maintenance over their lifetime. Therefore, it is critical to have a clear understanding of the historical, current, and projected health of the electric motor. 2024P00009 WO Irreversible demagnetization of a magnet is the permanent loss of a magnet's magnetic properties.This phenomenon occurs when a magnet is exposed to external factors such as extreme temperatures, high mechanical stress, or opposing magnetic fields that exceed its performance capability. Once a magnet has been irreversibly demagnetized, it cannot regain its original magnetic strength without remagnetization. Permanent magnet synchronous motors (PMSM) rely on the permanent magnetic flux of the rotor to couple with the rotating electromagnetic field of the stator, resulting in the conversion of electrical power into mechanical power as rotary motion. The magnetic flux of the rotor changes depending on the operating point and temperature conditions of the machine. Different methods are used to detect and quantify irreversible demagnetization.Model-based solutions consist of a demagnetization model of the magnet, which is given by the BH characteristic curve and the dependence of the magnetic properties on temperature, see "Temperature-Dependent Demagnetization Model of Permanent Magnets for Finite Element Analysis", P. Zhou et al., February 2012, IEEE Transactions on Magnetics 48(2):1031-1034. Here, B represents the magnetic flux density in Tesla and H represents the magnetic field strength in amperes per meter (A / m), respectively. The operating point is continuously monitored, observing its position relative to the knee point on the demagnetization curve. The knee point is the dividing point between the reversible and irreversible demagnetization of the magnet. The result is the relative change in the magnetic remanence. 2024P00009 WO A data-driven approach consists in generating synthetic data based either on simulations or on experiments.Synthetic datasets contain measurements of current, voltage, torque, and speed for various demagnetization levels. The dataset is then used for supervised learning. This approach also includes fitting regression models or training artificial neural networks, which represent the statistical relationship between the target of determining demagnetization and the inputs, the measured electrical and mechanical parameters. Another approach uses signal injection. This is based on injecting a series of signals through the inverter into the motor under standstill conditions. An analysis of the current flowing through the motor windings is performed. A baseline is first established before the motor is deployed in the field.Once the user starts up the motor, the same indicator is monitored over time and compared with the initial values ​​before use, which corresponds to the magnet with the optimal remanence. DE 10243219 A1 discloses a device for detecting faults due to weakening permanent magnets in a motor in a vehicle. The device includes a voltage monitor that detects a permanent magnet-induced voltage in the motor at a predetermined speed and in the no-load state. The voltage monitor is coupled to a processor that records the permanent magnet-induced voltage as measured at the predetermined speed and compares it with a reference voltage that reflects the permanent magnet-induced voltage for the motor with a fully magnetized permanent magnet.The processor determines a difference between the detected permanent magnet-induced voltage and the reference voltage. The difference is analyzed to determine if a component is faulty. Specifically, the permanent magnet-induced voltage is a function of the relative positions and locations of the permanent magnets in the motor. This relationship is used to detect a faulty magnet. Specifically, the permanent magnets are designed so that a change in magnetic resistance or magnet strength is used to identify the faulty magnet. A diagnostic code is set to alert others of the location of the faulty magnet for replacement or other corrective action.The object of the invention is to propose a method for determining the state of health of an electrical machine through an irreversible degree of demagnetization, which method utilizes control signals, temperature and speed measurements, and / or estimates of speed and temperature to enable data-driven condition monitoring for the rotor of a permanent magnet synchronous motor. Description of the invention The object is achieved with a method for determining the state of health of the permanent magnets in the rotor of an electrical machine based on a voltage parameter of the electrical machine, which is based on a control effect of a control signal as a measure of the induced counter-electromotive force. The proposed method is based on the control effect of the control signal and is not an actual measurement of the induced counter-electromotive force at the terminals of the machine.The voltage parameter depends on the rotor temperature and the rotor speed. A model- and data-driven method is used. 2024P00009 WO The method collects baseline data in a baselining step, which is terminated by a major demagnetization event. After the baselining step, the continuous monitoring step begins. The deviation of the voltage parameter indicates the degree of irreversible demagnetization of the permanent magnets. The information is also used to evaluate the effects of driving after an ASC event, i.e., in the so-called "limp-home mode." By quantifying the irreversible demagnetization after an active short circuit, the possibility arises of reducing heavy rare earths in the electrical machine. The method represents part of an overall procedure for monitoring the health of an electrical machine.The collected data and the degree of irreversible demagnetization of the permanent magnets are transmitted to the driver and / or a remote monitoring system. Description of the figures Figure 1 shows a history of the state of health of an electrical machine, Figure 2 shows a process diagram. The proposed solution aims at on-board condition monitoring of the rotor of a permanent magnet synchronous motor. Descriptive analysis methods are used to quantify the condition. Assuming that each component of the machine has either an optimistic or a conservative deterioration characteristic 2024P00009 WO under real operating conditions, the end of service life for each component is expected to deviate from the design specifications. Figure 1 shows a schematic of the history of the state of health of the electrical machine, starting from its commissioning at time 0 until the end of its service life EoL.The declining trend in health status is evident in the middle line of the diagram. Two additional trends are shown in the diagram. The EoLc trend follows a very conservative estimate, meaning life expectancy ends earlier. The EoL trend is an optimistic estimate. othe end of life is delayed. The diagram shows a very schematic illustration of the influence of various parameters in estimating the health of the electrical machine. This application proposes a model- and data-driven approach for monitoring the condition of the permanent magnets. This method involves quantifying the extent of irreversible demagnetization in the event of a significant demagnetization event, such as an active short circuit, and is illustrated in the block diagram of Figure 2. The idea is based on monitoring the back electromotive force (BEMF) at appropriate driving intervals, which correspond to the operating points in which the motor is almost not loaded and the rotor is rotating. At the appropriate driving interval, ^^^^ ~0 with ^^^^ - direct component of the current, ^^^^ ~0 with ^^^^ - quadratic component of the current, ^^ > 0 with ^^ - rotor speed.The concept consists of two stages. 2024P00009 WO Stage 1: Collecting basic data In Figure 2, the process begins with step 2 with the collection of measurement data and control signals. The rotor temperature TR, the rotor speed ^^ and the current data ^^^^, ^^^^, ^^^^ are collected. An appropriate driving interval 3 is determined with the parameters as mentioned above. The first stage is determining the basic data, a so-called baselining 4, in which a considerable number of BEMF, Back Electromotive Force data points 6 are collected for various operating points. The following equations are used in the rotor-fixed reference system and describe the dependencies of voltage u, current i and flux ^^ for a motor with a specific parameter set. ^^. d = ^^ s ^^ d + ^^^^ ^^ / ^^t-^^^^ q (1) ^^ q = ^^ s ^^ q + ^^^^ q / ^^t +^^^^ d (2) ^^ ^^ = ^^^^^^^^ + ^^ ^^^^ (3) ^^^^ = ^^^^^^^^ (4) with ^^ ^^ - DC component of the voltage, ^^ ^^ - quadratic component of the voltage ^^^^ - direct component of the current ^^^^ - quadratic component of the current ^^^^ - direct component of the flux ^^^^ - quadratic component of the flux 2024P00009 WO ^^^^ ^^ - flux of the permanent magnet ^^^^ - stator resistance ^^^^ - stator inductance in the d-axis ^^^^ - stator inductance in the q-axis ^^ - rotor speed. In an idle scenario (^^ ^^ ~0 and ^^ ^^ ~0) with a rotor speed greater than 0 (^^ > 0), equations 2 and 3 represent the following relationship between ^^ ^^ , ^^ and ^^ ^^^^ here: ^^ ^^ = ^^^^ ^^^^ (^^ ^^), (5) where ^^r is the estimated / measured rotor temperature. Since ^^^^, as a quadratic component of the voltage, is directly proportional to the flux of the permanent magnet ^^^^^^, it is used as an indicator of the remanent flux density at a specific rotor temperature TR and speed ^^. It should be noted that ^^^^' in step 7 is a control function, a control signal, and not an actual measurement of the induced BEMF at the terminals of the machine windings. This control signal comes from a proportional-integral controller. The integral component is essential for the function, as it ensures that the desired setpoints are reached even in the case of parameter uncertainties or demagnetization. The baselining phase ends when a major demagnetization event occurs, for example, an active short circuit ASC.Alternatively, a statistical approach can be used to determine the actual termination of the baseline simulation when a sufficiently representative range of operating points is covered. 2024P00009 WO Stage 2: Continuous Monitoring Phase The end of baseline simulation phase 4 marks the beginning of continuous monitoring phase 5. Baseline simulation phase 4 contains data points from ^^. ^^ ' with respect to the measured rotor speed ^^ and temperature T Rand thus represents one or more completely healthy magnets that have not been subjected to irreversible demagnetization. In the monitoring phase, the current ^^^^ data points are compared with the corresponding ^^^^' data from baselining phase 4. One of the challenges is to identify the identical operating points in both phases, baselining and the continuous monitoring phase. To solve this problem, a regression model is proposed that maps the values ​​of ^^^^ with the rotor speed ^^ and the rotor temperature TR during the baseline process. In this way, it is possible to approximate the baseline value of the ^^^^ for each data point of interest in the continuous monitoring phase in step 9. The relative difference between the ^^^^ values ​​dr = (^^^^' - ^^^^^ / ^^^^' indicates the degree of irreversible demagnetization of the permanent magnets.In step 10, the data is displayed or reused as the permanent magnet health status. The permanent magnet health indicator can be used as part of a comprehensive concept for monitoring the health status of the electric motor or the entire system, which takes into account the principle of multifactorial aging. The permanent magnet health indicator can be used by original equipment manufacturers to gain insights into fleet management and reduce the occurrence of early failures and detect overengineering. 2024P00009 WO

Claims

Claims 1. Method (1) for determining the state of health of the permanent magnets (10) in the rotor of an electrical machine based on a voltage parameter (^^^^ ^ of the electrical machine, which is based on a control effect of a control signal as a measure of the induced counter-electromotive force.

2. Method according to claim 1, characterized in that the voltage parameter (^^^^ ^ depends on the rotor temperature (^^r ^ and the rotor speed (^^).

3. Method according to claim 1 or 2, characterized in that a model and data-driven method is used.

4. Method according to one of the preceding claims, characterized in that the method collects baseline data in a baselining step, which is terminated by a major demagnetization event.

5. Method according to claim 4, characterized in that the continuous monitoring step begins after the baselining step.Method according to one of the preceding claims, characterized in that the deviation of the voltage parameter (^^^^ ^ indicates the degree of irreversible demagnetization of the permanent magnets.

7. Method according to one of the preceding claims, wherein the method represents part of an overall method for monitoring the health of an electrical machine.

8. Method according to one of the preceding claims, wherein the collected data and the degree of irreversible demagnetization of the permanent magnets are transmitted to the driver and / or a remote monitoring device. 2024P00009 WO.

Images