KR101210825B1 - Apparatus and method for monitoring condition of permanent magnet synchronous motors, and recoding medium having program for executing the method - Google Patents

Abstract

PURPOSE: A state diagnosis device of permanent magnet synchronizing motor, a diagnosing method, and a recording medium to record the program for progressing the diagnosis method are provided to easily diagnose the demagnetization state of permanent magnet of permanent magnet synchronizing motor and the eccentricity state of rotor without disassembling the motor or using any expensive additional equipment. CONSTITUTION: A state diagnosis device of permanent magnet synchronizing motor comprises a magnetic field generator(1110), a stator current measuring part(1120), and a state diagnosis part(1130). A magnetic field generator generates multiple magnetic fields which were overlapped by the DC magnetic field and AC magnetic field toward the pole of permanent magnet when the rotor is in the stationary state. A stator current measuring part measures a flowing current in the stator winding that corresponds to a generated magnetic field respectively. A state diagnosis part diagnoses the trend of differential inductance regarding the polar direction constituent value of permanent magnet which flows the stator winding based on the demagnetized state of permanent magnet and the eccentric state of rotor. [Reference numerals] (1110) Magnetic field generator; (1120) Stator current measuring part; (1130) State diagnosis part; (AA) Stator; (BB) Rotor; (CC) Permanent magnet; (DD) Stator wiring; (EE) Motor

Description

Apparatus and method for monitoring condition of permanent magnet synchronous motors, and recoding medium having program for executing the method}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a recording medium on which a state diagnosis apparatus, a diagnostic method, and a program for executing the diagnostic method of a permanent magnet synchronous motor are recorded. An apparatus, a diagnostic method, and a recording medium having recorded thereon a program for executing the diagnostic method.

Permanent Magnet Synchronous Machines (PMSM) are widely used in electric vehicle wind generators, trams and propulsion systems because of their high efficiency, high power density and ease of control. Permanent Magnet (PM) rotors are one of the most important parts of PMSM that determine the performance, efficiency and reliability of motor drive systems. Recently, it has been found that the demagnetization of the permanent magnet, the eccentricity of the rotor, and the failure of the bearings are the main root causes of PMSM failure, and intensive studies have been conducted to detect such failures. These failures can be caused by manufacturing defects, stress during operation, and malfunctions, which can lead to serious degradation of motor performance and efficiency. Therefore, it is important to periodically check the condition of the rotor and the whole process of driving the motor for optimum performance of the motor.

Unrecoverable partial or total potatoes are a problem that can arise due to thermal, electrical, mechanical and environmental stresses on the permanent magnets. The main root cause of irreversible potatoes is the 'operating point effect', which is due to the effect of the transition of the performance (load) curve due to the temperature-dependent decrease in the potato curve and the reduction of the magneto-motive force (mmf). Combined results. If the operating point (intersection of the potato curve and the performance curve) falls below the inflection point of the potato curve, the magnet is permanently demagnetized. Changes in metal properties in magnetic materials due to increased temperature and / or corrosion / oxidation can also lead to irreversible potatoes. NdFeB magnets, known for their high energy density and reasonable price, are widely used to produce high performance, high energy density PMSMs. However, NdFeB magnets are irreversible due to their low temperature characteristics (low operating temperatures and high temperature coefficients), low corrosion resistance (under moisture or chlorine in the environment) and low mechanical stiffness (breaks due to corrosion, flakes or gold). Sensitive to potatoes

In addition, void eccentricity may occur due to a defect in the operation or manufacturing of the motor, and the original degree of eccentricity is required within 5-10% of the void. Static Eccentricity (SE) is a state where the position of the minimum radial void is fixed. This is caused by an ovoid shape of the stator core or an incorrect position of the stator core, which generally does not change over time. Dynamic Eccentricity (DE) is a state in which the minimum radial void rotates with the rotor. This is caused by old bearings, bent shafts, asymmetric thermal expansion of the rotor or high levels of static eccentricity. Eccentricity causes unbalanced magnetic attraction that results in vibration, creepy, bearing breakage and / or deflection of the rotor. Although the degree of SE does not change over time, the degree of DE increases, causing contact between the rotor and the stator, which can seriously damage permanent magnets, cores or windings. Therefore, much research has been made in the last 5-6 years to develop a technique for detecting potato and eccentric error of permanent magnets to improve the performance, reliability and efficiency of the PMSM drive system.

Motor Current Signature Analysis (MCSA) is the PMSM's rotor failure detection method that is most actively studied to provide continuous, motor-parameter- and temperature-independent on-line monitoring without sensors. The rotational speed frequency component of the rotor is expressed as Equation 1 and is presented as a measure for rotor anomalies including partial permanent magnet demagnetization, SE, DE, misalignment or rotor / load imbalance.

Here, f _rf and f _s are the rotor ideal and the fundamental frequency component, k is an integer, and p is the number of pole pairs.

Although the MCSA has many of the advantages mentioned above, there are many limitations in its implementation in practical use. In many PMSM applications, the motor operates under steady-state conditions where the speed and load vary continuously, making it difficult to obtain the exact result of Equation 1. Time-frequency analysis techniques have been studied for reliable detection. However, this also increases costs due to computational strength and hardware specifications, making it difficult to apply to low-cost drive products. Since there is an asymmetry of the magnetic flux pattern that induces the current component in Equation 1, the uniform or symmetrical potato that is usually present cannot be detected. A fatal limitation of the MCSA is that oscillating load torques (eg, reciprocating compressors) produce side lobes at the same frequency components as in equation (1). This is larger than the size due to the above, and there is no practical means to separate them.

Model-based estimation of rotor flux λ _dr or torque constant k _t proportional to mean PM intensity has also been studied for potato detection. However, the rotor flux λ _dr and the torque constant k _t also decrease with increasing temperature (especially for ceramic and NdFeB magnets) and change with the magnetic saturation of the core. In addition, the rotor flux λ _dr or torque constant k _t estimation is highly dependent on the accuracy and parameters of the model. Therefore, variations in the model due to nonlinear cross-magnetization effects, magnetic saturation, and rotor temperature variations should be considered, which is quite difficult, as rotor flux λ _dr or torque constant k _t is small or local defects in permanent magnets are considered. It becomes sensitive. As a result, rotor flux λ _dr or torque constant k _t estimation is difficult to provide as an indicator sensitive to void eccentricity.

Traditional off-line tests on the quality of permanent magnets or DE are visual inspection, gauss meter, dial test indicator measurement, etc. They can provide accurate results on rotor quality because they get measurement results directly on the surface of the rotor. However, they must dismantle the motor and require specific measurement equipment or settings. The average strength of a permanent magnet can also be indirectly estimated by measuring the electro-motive force (emf) voltage for a rotor rotating at a constant speed. This off-line test, too, is dependent on the temperature of the permanent magnet and is not convenient because it requires external means to rotate the rotor at a constant speed.

On the other hand, Japanese Laid-Open Patent Publication No. 2009-22091 discloses that when the induced voltage generated by the permanent magnet detected from the motor or the induced voltage calculated from the current detected by the motor in the no-load operation state is lower than the reducer limit curve for the normal permanent magnet. A technical configuration for determining that a permanent magnet is demagnetized is disclosed. In addition, Korean Patent Publication No. 2011-68446 discloses a ratio obtained for a normal permanent magnet after decomposing the stator voltage measured from the synchronous generator into a normal component voltage and a reverse phase component voltage and calculating a ratio of the reverse phase component voltage to the normal component voltage. Compared with the disclosed technology configuration for determining whether the permanent magnet demagnetization is disclosed. In addition, Korean Patent Publication No. 2010-45718 discloses a plurality of magnetic fields having different directions while passing through a rotating shaft of a rotor in a state where the motor is stopped. A technical arrangement is disclosed for diagnosing the demagnetization state of a permanent magnet by comparing the current measured at the stator windings. However, among these prior arts, the prior arts described in Korean Patent Laid-Open Publications disclose only a configuration for detecting only the potato state of a permanent magnet, and are different from the present invention in the technique of detecting potato state. In addition, the prior art described in the Japanese Laid-Open Patent Publication is measuring the demagnetization and eccentricity of the permanent magnet in the no-load operation state, not stationary state has the disadvantage that an external means for rotating the rotor at a constant speed is required.

The present invention has been made to solve the above-mentioned problems of the prior art, the technical problem to be achieved by the present invention is to easily diagnose the potato state and eccentric state of the permanent magnet synchronous motor without expensive separate equipment without disassembling the motor It is to provide a device and method that can be.

In addition, another technical problem to be achieved by the present invention is a recording medium for recording a program for executing a method in the computer to easily diagnose the potato state and the eccentric state of the permanent magnet synchronous motor without expensive separate equipment without disassembling the motor To provide.

In order to achieve the above technical problem, the apparatus for diagnosing a state of a permanent magnet synchronous motor according to the present invention includes generating a plurality of magnetic fields in which a direct current magnetic field and an alternating magnetic field are superposed in the pole direction of the permanent magnet when the rotor is in a stationary state. Magnetic field generating unit; A stator current measuring unit configured to measure a current flowing in the stator windings corresponding to each of the generated magnetic fields; And the potato state of the permanent magnet and the eccentric state of the rotor on the basis of the change of the differential inductance with respect to the polar component value of the permanent magnet of the current flowing in the stator windings measured corresponding to each of the generated magnetic fields. A diagnosing state diagnosing unit, wherein the differential inductance is measured from the alternating magnetic field component with respect to the amount of change in the polar component value of the permanent magnet of the current by the alternating magnetic field component included in the generated magnetic field. It is defined as the amount of change in the value of the polar component of the permanent magnet in flux linkage.

In order to achieve the above another technical problem, a method for diagnosing a state of a permanent magnet synchronous motor according to the present invention includes: Generating a magnetic field; (b) measuring a current flowing in the stator windings corresponding to each of the generated magnetic fields; And (c) the potato state of the permanent magnet and the rotor state of the rotor based on the change of the differential inductance with respect to the polar component value of the permanent magnet of the current flowing in the stator windings measured corresponding to each of the generated magnetic fields. Diagnosing an eccentric state, wherein the differential inductance is measured from the alternating magnetic field component with respect to the amount of change in the polar component value of the permanent magnet of the current by the alternating magnetic field component included in the generated magnetic field. It is defined as the amount of change in the value of the polar component of the permanent magnet in flux linkage.

According to the recording medium recording the condition diagnosis device, the diagnosis method, and the program for executing the diagnosis method of the permanent magnet synchronous motor according to the present invention, the permanent magnet in the permanent magnet synchronous motor can be easily installed without expensive equipment without disassembling the motor. The potato demagnetization of the magnet and the core of the rotor can be diagnosed. In addition, since the existing hardware resources such as the current sensor and controller of the inverter for driving the permanent magnet synchronous motor is used as it is, there is an advantage that there is no need to add additional hardware. In addition, it can quickly and easily diagnose the demagnetization and eccentricity of permanent magnets compared to conventional diagnostic methods that are time-consuming and expensive. In addition, compared to the conventional MCSA, model-based or counter electromotive force-based permanent magnet synchronous motor rotor detection method has a higher sensitivity and reliability, it costs less.

1 is a view for explaining a state diagnosis technique in the apparatus and method for diagnosing a state of a permanent magnet synchronous motor according to the present invention;
2 is a diagram showing a magnetic field density distribution of an unexcited permanent magnet synchronous motor in a normal state;
3 shows an equivalent magnetic circuit model for the magnetic field path in the d-axis direction, the stator windings in the d-axis direction, the voids and the stator, and the rotor core constituting the permanent magnet, FIG.
4 shows the magnetoresistance-magnetic force characteristics of a permanent magnet (potato curve of a permanent magnet) and a magnetic circuit outside the permanent magnet (investment curve), FIG.
5 is a diagram illustrating the magnetomotive force characteristics of a d-axis magnetic flux path for a stator winding and a permanent magnet modeled as a source of magnetomotive force;
6 is a diagram showing a relationship between a d-axis linkage flux and a current and a d-axis inductance and a differential inductance curve;
7 is a diagram showing a d-axis differential inductance curve according to the d-axis linkage flux and current relationship and the permanent magnet potato;
8 is a diagram showing the change of the DC current band d-axis differential inductance according to the permanent magnet potato and the eccentricity,
9 is a diagram showing a result of a DC sperm FE simulation of the magnetic flux density due to a permanent magnet without a stator current applied to an eccentric motor;
FIG. 10 shows the d-axis linkage flux and current curves and differential inductance curves for void eccentricity;
11 is a view showing the configuration of a preferred embodiment of a state diagnosis device of a permanent magnet synchronous motor according to the present invention;
12 is a flowchart illustrating an embodiment of a method for diagnosing a state of a permanent magnet synchronous motor according to the present invention;
13 is a photograph of the configuration of the motor rotor and the permanent magnet used in the experiment, and a permanent magnet that is broken to test a partial potato,
14 is a mechanical design of the axis and key arrangement to produce 50% dynamic eccentricity (sample G).
15 shows an experimental setup for producing a dynamic eccentric state;
16 shows experimental results for permanent magnet potatoes;
FIG. 17 shows the percent change in Xd 'and Xd' measured at Id = 5 A for demagnetized permanent magnet samples A-D;
18 is a diagram showing experimental results on void eccentricity, and
FIG. 19 shows the percent change in Xd 'and Xd' measured at Id = 4 A for eccentric rotor samples E-G.

Hereinafter, with reference to the accompanying drawings, a preferred embodiment of a recording medium on which a state diagnosis apparatus, a diagnostic method, and a program for executing the diagnostic method of a permanent magnet synchronous motor according to the present invention will be described in detail.

1 is a view for explaining a state diagnosis technique in the apparatus and method for diagnosing a state of a permanent magnet synchronous motor according to the present invention.

Referring to FIG. 1, a signal is applied to the permanent magnet synchronous motor 110 by using the inverter 100 whenever the motor is in a stopped state in order to detect and classify the potato state and the space eccentricity of the permanent magnet. The inverter 100 is operated to drive the permanent magnet synchronous motor 110 with a small alternating magnetic field added to a direct current magnetic field in the d-axis direction to obtain an equivalent differential inductance Ld '.

2 is a diagram showing a magnetic field density distribution of an unexcited permanent magnet synchronous motor in a normal state.

Referring to FIG. 2, DC static magnetic finite element simulation of a 4P internal permanent magnet synchronous motor was performed to understand magnetic properties of the d-axis. The magnetic field distribution of a permanent magnet synchronous motor without stator excitation appears at the rotor in a stationary state and is arranged in the path of least magnetoresistance. The magnetic field path in the d-axis direction, the stator windings in the d-axis direction, the voids, and the stator constituting the permanent magnet, and the rotor core may be represented by the equivalent magnetic circuit model shown in FIG. In Fig. 3, F, R, Φ and N represent the number of magnetic forces, magnetic resistance, magnetic flux and equivalent windings, respectively, and the subscripts pm, g, g and l mean permanent magnets, voids, cores and loads, respectively. Because magnetic materials have a specific permeability similar to air, permanent magnets are generally modeled as magnetic flux-dependent magnetic force with internal power magnetoresistance.

4 shows the magnetoresistance-magnetic force characteristics of the permanent magnet (potato curve of the permanent magnet) and the magnetic circuit outside the permanent magnet (investment curve).

Referring to Fig. 4, the operating point for the magnetoresistance and the magnetomotive force is determined at the intersection of the two curves, and the open circuit operating point at Φ _m and F _m in the unexcited state (i.e. i _d = 0) is shown. It is. As can be seen in FIG. 3, the d-axis stator current i _d ≠ 0 results in an additional magnetomotive force. This may be equivalent to the potato curve transition as shown in FIG. 4, changing the operating point. Meanwhile, as shown in FIG. 5, the analysis may be simplified by looking at the permanent magnet and the winding as a source of magnetomotive force and including R _pm as part of the total magnetic _permeability . The magnetic force curve forms a vertical line, the value of which is the sum of the permanent magnet and the winding magnetic force. The permeability curve includes the combined effect of the permeability of permanent magnets, voids and cores, and the slope of the permeability curve is proportional to the inductance. The operating point of the magnetic circuit in the d-axis direction is determined at the intersection of the two curves, similar to that shown in FIG. The positive current i _d > 0 generating magnetic flux in the d-axis direction from FIGS. 3 to 5 increases the magnetic force and the magnetic flux, and decreases the inductance (the slope of the permeability curve).

The magnetic force curve for the d-axis flux shown in FIG. 5 can be redrawn as a current i _d curve for flux lingage λ _d by multiplying or dividing N as shown in FIG. 6. In the linear region, the λ _d -i _d curve is equal to the inductance of the coil, and the rewritten curve is expressed by the magnetic flux equation in the d-axis direction of the permanent magnet synchronous motor as shown in Equation 2 below.

Here, λ _m is the linkage magnetic flux of the coil when i _d = 0 (NΦ _m ).

The lambda _d axis does not pass the origin of the lambda -i curve (0 in Fig. 6) because the magnetic force path is formed in the direction of the d axis due to the permanent magnet with i _d = 0. The λ _d axis intersects at the open circuit operating point a where λ _d is equal to λ _m when i _d = 0, and the equivalent i _d toward the origin corresponds to the dividing force of the permanent magnet divided by N. From Equation 2, the inductance L _d of the d-axis is defined as the ratio of the increase amount of λ _d and the d-axis current i _d with respect to the open circuit operating point a as in the following equation.

This can be seen in the L _d curve (point b ") shown on the right side of Fig. 6, where L _d decreases slightly due to saturation. The d-axis differential inductance L _d ', which is a value employed for abnormal monitoring in the present invention, is L _The differential inductance L _d 'is defined as the slope of the λ-i curve at each point on the λ-i curve, and is expressed by the following equation.

The reduction in L _d 'at saturation is more important compared to the reduction in L _d because it is a derivative of the λ-i curve.

The d-axis differential inductance of the operating point can be obtained by applying an alternating stator magnetic field in the d-axis direction and calculating the reactive component of the impedance from the fundamental components of the stator voltage and current. If the magnitude of the alternating magnetic field is not too large, the inductance can be considered equal to L _d 'at the operating point of the λ-i curve, and the operating point is generally designed to be located before the inflection point of the curve in the linear region for optimal performance. do. If a stator alternating magnetic field is added to the stator direct current magnetic field in the a-axis direction, the direct current magnetic field shifts the operating point, and at a corresponding point, the value of L _d 'can be obtained from the alternating current component. For example, if the applied DC current component is equal to I _d , the operating point of FIG. 6 moves from point a to point b, and the value of L _d 'obtained from the alternating current component is from L _{d , a} ' to L _d due to saturation. _{, b} '. The decrease in the differential inductance L _d 'value at point b' is very large compared to the decrease in the value corresponding to inductance L _d at point b ".

For the calculation of L _d 'as described above, the stator magnetic field applied at an angle of θ with respect to the d axis includes an alternating magnetic field having a peak pulsating between θ and θ + 180 ° and a fixed DC magnetic field at θ. The vector of the magnetic field, which is the sum of the alternating magnetic field and the direct current magnetic field at θ, can be obtained with sine-delta PWM with three phase voltage reference sets as in the following equation.

Where V _dc and V _ac are the voltage magnitudes of the dc and ac components, respectively, and ω is the excitation frequency.

The applied magnetic field does not cause torque in the device because the direct current field aligns the rotor at the rotor's current position, and the alternating magnetic field produces torque of the same magnitude in the opposite direction. The d-axis angle θ can be easily estimated by the initial rotor position estimation method based on rotor protrusion and / or saturation. The reference voltage and current vector in the d-axis direction [theta] can be calculated by the following equation from Equations 6 and i _as , i _bs and i _cs current measurements.

Where k is 2/3 and v ^* _abcs and i _abcs denote stator reference voltage and current matrix, respectively.

Equivalent L _d ′ may be extracted from the phase fundamental components of the reference voltage V _d and the current vector I _d by the following equation.

Partial or uniform permanent magnet demagnetization due to one of the reasons described above weakens the magnetomotive force (F _pm ) of the permanent magnet. This results in a weakening of Φ _m or λ _m , which can be expressed as shifting the λ _d axis of FIG. 6 to the left as shown in FIG. 7. The open circuit operating point is moved down in the linear region (point d to point d in FIG. 7) towards the origin with a decrease in magnetomotive force and λ _d . Since the operating points are in the linear region (the slope of the λ-i curve at points a and d is similar to L _{d , a} '≒ L _{d , d} '), the value of L _d 'does not change significantly with potatoes. Compared to normal motors, a high direct current magnetic field component is required to saturate the core of the motor with demagnetized permanent magnets. Also for the normal permanent magnet shown in Fig. 6, when a direct current magnetic field is measured which corresponds to the DC current I _d of the core it is to be saturated because L _{d, a} 'value of the L _{d, b'} will be reduced to a. However, L _d ′ is still the same (L _{d , c} '≒ L _{d , d} ') for the same direct current magnetic field excitation I _d for the case with the potato permanent magnet shown in FIG. Thus L _d ′ can be measured under the same excitation state (I _d > 0).

If, when the expression as a function of I _d, L _d as the core together with the increase of I _d, as shown in FIG. 8 starts to be saturated, L _d begins to decrease the. As the core reaches the supersaturation region, L _d ′ converges to a lower value. In a permanent magnet synchronous motor with a demagnetized rotor, the L _d '-I _d curve for a motor with a normal permanent magnet shifts to the right because a large I _d is needed to saturate the core. Thus, the value of L _d 'for a demagnetized permanent magnet increases to a constant range of applied DC magnetic field component or I _d . The optimum value of the direct current magnetic field for the maximum detection sensitivity depends on the design of the motor. These values, the absolute value of the change in | ΔL _d '| L _d may be determined from, which can be obtained from the FE analysis or experiment. This is in contrast to induction motors, and since permanent magnet synchronous motors are generally designed together for product applications, the characteristics of permanent magnet synchronous motors are well known by inverters and are not difficult to obtain. The DC magnetic field required for sensitive detection is a magnetic field corresponding to I _{d and dem} for the example shown in FIG. 8 since | ΔL _d '| is the maximum value. Since the alternating current component is used to extract L _d ', V _ac must be large enough to provide sufficient resolution for current measurements for consistent experimental results.

In order to observe the state in which the magnetic flux pattern changes with the eccentricity, DC static magnetic FE analysis was performed on an eccentric unexcited permanent magnet synchronous motor. The rotor is in a stationary state, and as shown in FIG. 9, the eccentric direction is assumed to be toward the center of the permanent magnet with the maximum magnetic attraction. 9 and 3, the magnetic flux in the stator core of the upper magnetic circuit increases because the magnetic resistance decreases, while the magnetic flux in the lower magnetic circuit decreases due to the increase of the magnetic resistance. As shown in FIG. 10, this can be expressed as the shift of the permeability curve for a magnetic circuit in the opposite direction from the permeability curve of a normal motor for a multipole pair (p) motor (p = 2 in this embodiment). (Operating points at points e, up and e, low). If the eccentricity is present, investment also because the slope of the curve decreased with large pores of the lower magnetic circuit L _d 'value of the L _{d, e, low'} is reduced. For the upper magnetic circuit, the slope of the permeability curve increases with the reduced void, but the operating point enters the saturation region due to the increase of the magnetic flux. This results in a reduction of L _d '(L _{d , e, up} ') as shown in FIG. 10. Since λ is equal to the underlying supersaturation, the permeability curves for the two magnetic circuits converge to similar values with large direct current components. L _{d, e, low} 'and L _{d, e, up'} because it reduces both the motor with eccentric L _d 'values L _{d, e'} is reduced from L _d 'values L _{d, a'} concentric motor. Therefore, if L _d ′ is measured in the same excited state, the reduction of L _d ′ can be used as an indicator of eccentricity.

If, when creating the graph of I _d on, I _d = from 0 L _{d, e,} L _{d, e} to the eccentric motor is lower than L _{d, a} ', as I _d increases sharply than concentric motor Increases. Fig from 10 L _{d, e} can be estimated to be due as the value decreases, the rapid decrease of the L _{d, e, up} as that of the upper magnetic core saturation circuit. Looking at the graph of | ΔL _d '|, it can be seen that there is an optimal DC magnetic field for maximum detection sensitivity to motor design-dependent eccentricity as in the case of potatoes. In the case of the example shown in Fig. 8, the DC magnetic field required for the detection sensitivity corresponds to I _{d , ecc} which is | ΔL _d '| rk max. The value of L _d 'at I _d = 0 can be measured with alternating magnetic field excitation on the d axis (V _dc = 0). The sensitivity of the eccentric detection can be improved by the present invention, which moves the operating point as shown in FIG.

In order to identify the demagnetism and eccentricity of the permanent magnet synchronous motor, the measured change in L _d 'for a predetermined level of direct current magnetic field or I _d is compared with the value of L _{d , a} ' for a normal permanent magnet. If the permanent magnet is demagnetized, the value of L _d 'is increased, and if the rotor is eccentric, the value of L _d ' is decreased, which enables clear detection and classification of the failure of the permanent magnet synchronous motor rotor.

Permanent magnets also suffer from demagnetization, which can be recovered as the permanent magnet temperature T _PM fluctuates. The temperature coefficient (in% change per ° C increase) for Φ _r depending on the permanent magnet material is between 0.0 2% / ° C and -0.02% / ° C. Therefore, all the techniques for detecting the failure of the rotor depending on the magnitude of the magnetic flux are affected by the change of T _PM , which should be considered for reliable measurement. In the present invention, the change of L _d 'due to the irreversible potato and the potato dependent on T _PM cannot be distinguished from each other because of the movement of the lambda _d axis and the movement of the operating point in the surface. One method of indirectly measuring T _PM is by estimation of stator resistance R _s , which is a good indicator of stator temperature T _s . Estimation of R _s can be easily obtained from the direct current component of voltage and current after applying a direct current component to a motor at standstill using an inverter. If the test is performed before starting the motor after the motor has cooled down to ambient temperature, the R _s estimate (T _s ) represents T _PM . Since the measured value of L _d 'from similar T _PMs is a stationary test, it can be a change in the state estimate of a reliable permanent magnet rotor that is not temperature sensitive. Another approach for eccentric detection is to observe the shape of the L _d 'curve shown in FIG. 8. The curve can be shifted left or right according to T _PM , but the shape does not change, allowing eccentricity detection independent of T _PM .

Normal values of L _{d , a} ′ may vary for motors of the same design due to individual variations such as permanent magnets and / or core materials or manufacturing defects or inconsistencies. Therefore, it is desirable to measure and store the normal values of L _{d , a} ′ for individual motors and to manage them over time for reliable measurements.

11 is a diagram illustrating a configuration of a state diagnosis device of a permanent magnet synchronous motor according to the present invention.

Referring to FIG. 11, the state diagnosis apparatus 1100 of the permanent magnet synchronous motor according to the present invention includes a magnetic field generating unit 1110, a stator current measuring unit 1120, and a state diagnosis unit 1130.

The magnetic field generator 1110 generates a magnetic field in which a DC magnetic field and an alternating magnetic field are superposed in the pole direction of the permanent magnet (that is, from the S pole to the N pole) when the rotor of the permanent magnet synchronous motor is in a stopped state. The magnetic field generator 1110 may be implemented by the inverter 100 as shown in FIG. 1. At this time, the magnetic field generated by the currents i _a , i _b , i _c applied by the magnetic field generating unit 1110 to the permanent magnet synchronous motor does not coincide with the pole direction of the permanent magnet and is a constant angle (ie, in FIG. 1). The angle θ) between the d axis, which is the pole direction of the rotor, and the generating direction of the magnetic field occurs. This d-axis angle θ can be easily estimated by an initial rotor position estimation method well known in the art based on rotor protrusion and / or saturation. And the magnetic field component in the d-axis direction can be obtained by the equation (7). At this time, the magnetic field generating unit 1110 changes the magnitude of the DC magnetic field within a certain range, thereby changing the degree of saturation of the core. For example, when a DC magnetic field is increased by flowing a DC current in a pole direction of the permanent magnet to the permanent magnet synchronous motor, the magnetic field of the permanent magnet and the magnetic field generated in the stator winding are overlapped to saturate the core, thereby reducing the differential inductance. If the permanent magnet of the permanent magnet synchronous motor is demagnetized, the core will begin to saturate only when a direct current magnetic field is applied to the permanent magnet. Therefore, for demagnetized permanent magnets, a larger direct current must be applied to the core to saturate and the differential inductance begins to decrease.

The stator current measuring unit 1120 measures current flowing through the stator windings in correspondence with each of the generated magnetic fields. Accordingly, the stator current measuring unit 1120 measures and outputs current flowing through the stator windings for each magnitude of the DC magnetic field generated by the magnetic field generating unit 1110 and applied to the permanent magnet synchronous motor.

Status diagnosis unit 1130 polarity of the permanent magnet, measured from the pole direction component and the alternating current magnetic field component of the permanent magnets of the current value of the stator winding measured with respect to the generated magnetic field, each polarity current i _d and the permanent magnet flux linkage Based on the change of the differential inductance L _d ′ calculated by Equation 4 on the basis of λ _d , the potato state and the eccentric state of the permanent magnet are diagnosed. At this time, the state diagnosis unit 1130 first calculates the differential inductance based on Equation 4 based on the current and the chain flux measured from the alternating magnetic field component to diagnose the potato state of the permanent magnet. Next, the state diagnosis unit 1130 diagnoses the potato state and the eccentric state of the permanent magnet by comparing the calculated differential inductance with a reference inductance previously measured and stored for a normal permanent magnet.

If the DC current values of different magnitudes flowing through the stator windings measured by the stator current measuring unit 1120 and the differential inductances calculated corresponding thereto are larger than the reference differential inductances calculated under the same conditions for the normal permanent magnet, the state diagnosis is performed. The unit 1130 determines that the permanent magnet is demagnetized. In addition, if the DC current values of different magnitudes applied to the permanent magnet synchronous motor and the differential inductance calculated corresponding thereto are smaller than the reference differential inductance calculated in the same situation for the normal permanent magnet, the state diagnosis unit 1130 is eccentric. It is considered to have occurred. At this time, the DC current values applied to the permanent magnet synchronous motor are preferably set to values that can most sensitively diagnose potato and eccentricity. Therefore, as shown in FIG. 8, it is preferable to determine I _{d , dem} , I _{d , and ecc} as the direct current value at which the difference in inductance is greatest when diagnosing a potato state and advancing an eccentric state, respectively.

Meanwhile, the state diagnosis unit 1130 may diagnose the eccentric state based on the slope of the DC current versus the differential inductance graph. That is, as shown in Fig. 8, the change ratio of the differential inductance calculated correspondingly with respect to the DC current values of different magnitudes measured from the stator windings of the permanent magnet synchronous motor (i.e., the slope of the graph of the DC current vs. the differential inductance graph). ) Is different from the change ratio of the reference differential inductance calculated in the same situation with respect to the normal permanent magnet, the state diagnosis unit 1130 determines that the rotor is eccentric.

12 is a flowchart illustrating an embodiment of a method for diagnosing a state of a permanent magnet synchronous motor according to the present invention.

Referring to FIG. 12, when the rotor of the permanent magnet synchronous motor is in a stationary state, the magnetic field generating unit 1110 includes a plurality of DC magnetic fields and AC magnetic fields superposed in the pole direction of the permanent magnet (that is, from the S pole to the N pole). Generate a magnetic field of (S1200). In this case, the plurality of magnetic fields generated by the magnetic field generating unit 1110 are set differently within the range in which the magnitude of the DC magnetic field is set in advance. Next, the stator current measuring unit 1120 measures the current flowing in the stator windings in correspondence with each of the generated magnetic fields (S1210). Next, the potato state diagnosis unit 1130 calculates the differential calculated by Equation 4 based on the current values of the stator windings measured for each generated magnetic field and the current i _d measured from the alternating magnetic field component and the linkage flux λ _d . The potato state and the eccentric state of the permanent magnet are diagnosed based on the change in inductance L _d ′ (S1220).

In order to verify the effect of the present invention, an experiment was performed on a permanent magnet synchronous motor having a 4-pole 7.5 kW NdFeB magnet as shown in FIG. A commercial DSP was used to control the IGBT inverter to drive the motor as a magnetic field combined with direct current and alternating magnetic fields in the d-axis direction. The value of V _dc is set so that Id varies between -1 A and 7 A (or 10 A), and I _{d is} used to check the change of detection sensitivity with respect to the differential inductance L _d 'according to the magnitude of the DC magnetic field. The size of was increased in units of 1 A. The AC frequency was set to 50 Hz to calculate L _d 'from the voltage v and current i measured from the alternating magnetic field, and V _ac was set so that the magnitude of the alternating current had a peak-to-peak of 2 A for each I _d magnitude. Set. The consistency of L _d ′ does not improve at alternating current magnitudes greater than the peak-to-peak of 2 A. Tests were performed with a motor sample with strict homogeneous potatoes (Sample B) and a partially broken permanent magnet (Samples C, D) and compared with a normal rotor to assess the detection sensitivity of the present invention under uniform and partial potatoes. . According to the results of the back EMF voltage measurement at 30000 RPM for Samples A and B, the back EMF constant of Sample B was reduced by 31.2% due to uniform potatoes. In order to obtain a partially demagnetized sample, 7.2% and 10.4% of the chisel magnets were intentionally broken as shown in FIGS. 13B and 13C.

In order to obtain 0%, 25% and 50% DE states for 0.65 mm voids, a special shaft assembly suitable for the rotor core as shown in FIG. 13 (a) was fabricated. The outer dimensions of the original shaft were reduced and four grooves were made so that the spacing between the shaft and the rotor core could be adjusted to control the eccentricity using keys of different heights. 14 shows the mechanical configuration for the case with 50% DE upwards. 15 shows a picture of the manufactured shaft, three keys for eccentricity of 0%, 25% and 50%, and the rotor core assembly. As can be seen from the FE simulation shown in FIG. 9, experimental measurements were performed with the direction of DE aligned to the center of the magnet having the largest magnetic attraction force.

In FIG. 16A, measurement results for X _d ′ of healthy and potato permanent magnet samples A to D are shown as a function of I _d . If the direct current magnetic field component is in a certain range, it can be clearly seen that X _d 'increases with the degree of potato for samples B-D. In the case of partially demagnetized samples C and D, the X _d 'curve shifts to the right because high magnetism is required to saturate the core. In the case of sample B, only the linear region of the curve is observed within the range of the direct current magnetic field applied to the d axis due to the strict uniform potato. In these samples a reduction in X _d 'due to core saturation can only be observed at very large I _d values. Because the operating point is far below the inflection point in the linear region, the X _d 'value of sample B is greater than samples A, C and D at I _d = 0.

In FIG. 16B, the difference ΔX _d ′ between the partially demagnetized sample and the X _d ′ values of the normal sample is shown to determine an optimal excitation condition for high sensitivity. Referring to FIG. 16B, ΔX _d ′ initially increases with I _d for samples C and D, but begins to decrease as the core reaches supersaturation because the curve converges to similar values. The value of ΔX _d ′ is greatest when I _d = 5 A, which means that the sensitivity to potato detection is maximum. The demagnetization of the permanent magnets, which is important for sample B, can be easily detected by model-based detection techniques or by back electromotive force-based detection techniques. However, small potatoes, such as those of samples C and D, are difficult to detect. If 5 A direct current excitation is used, X _d 'is increased by 6.2% and 12.6% for 1.8% and 2.6% partial potatoes, respectively, as shown in FIG. From this, it can be seen that the diagnostic apparatus and method according to the present invention are very sensitive to small amounts of partial potatoes because of the magnetic saturation in which the increase of X _d 'is amplified due to the nonlinear saturation effect in the potato permanent magnet. . This is a significant improvement in sensitivity compared to conventional techniques where the failure indicator is proportional to the strength of the magnet.

Of Figure 18 (a) and (b), each X sample _d of the eccentric E ~ G 'measured values of ΔX and _d' is shown as a function of I _d. Referring to FIG. 18, it can be seen that X _d ′ increases as the size of DE (samples F and G) increases. The X _d 'curve moves downwards over a certain range of I _d , mainly because the eccentricity saturates the portion of the core where the voids are reduced, reducing X _d ' in that portion of the core. It should be noted that the X _d 'curves for Samples A and E are slightly different due to individual variations in materials and / or non-ideal conditions of manufacture. Therefore, the change in X _d 'should be checked for the normal values of the individual motors under the same environment and preset excitation conditions. It can be seen from (b) of FIG. 18 that ΔX _d ′ initially increases and decreases as it saturates and converges to zero. The highest detection sensitivity for eccentricity in the study is obtained at I _d = 4 A (I _{d , ecc} ) with ΔX _d 'largest. The decrease in ΔX _d ′ at I _d = 4 A is much higher than when I _d = 0 A, whereby the diagnostic device and method according to the invention has a higher sensitivity than conventional techniques. Figure 19 shows the increase in Xd 'measured at I _{d , ecc} = 4 A for samples F and G. The% reduction is as high as 5.8% and 14.5% for 25% and 50% eccentricity, respectively. As with the potato case, the diagnostic device and method according to the invention is very sensitive to eccentricity because the nonlinear magnetic saturation amplifies the reduction of X _d 'for the eccentric rotor. This is a substantial improvement over the conventional MCSA, where the percent magnitude of the failure component relative to the base component is less than -60 dB, which is less than 0.1% of the base component.

The present invention can also be embodied as computer-readable codes on a computer-readable recording medium. A computer-readable recording medium includes all kinds of recording apparatuses in which data that can be read by a computer system is stored. Examples of the computer-readable recording medium include a ROM, a RAM, a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like, and may be implemented in the form of a carrier wave (for example, transmission via the Internet) . The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is clearly understood that the same is by way of illustration and example only and is not to be taken by way of limitation in the embodiment in which said invention is directed. It will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the scope of the appended claims.

Claims (5)

A magnetic field generating unit generating a plurality of magnetic fields in which a direct current magnetic field and an alternating magnetic field overlap in a pole direction of the permanent magnet when the rotor is in a stationary state;
A stator current measuring unit configured to measure a current flowing in the stator windings corresponding to each of the generated magnetic fields; And
Diagnosing the demagnetized state of the permanent magnet and the eccentric state of the rotor on the basis of the change of the differential inductance with respect to the polar component value of the permanent magnet of the current flowing in the stator windings corresponding to each of the generated magnetic fields It includes a; state diagnosis unit
The differential inductance is the pole of the permanent magnet of flux linkage measured from the alternating magnetic field component with respect to the amount of change in the polar component value of the permanent magnet of the current by the alternating magnetic field component contained in the generated magnetic field. A state diagnosis apparatus for a permanent magnet synchronous motor, characterized in that it is defined as a change amount of a direction component value. The method of claim 1,
The state diagnosis unit diagnoses that the permanent magnet is demagnetized when the differential inductance calculated based on the generated magnetic field is larger than the reference differential inductance calculated and retained in the state where the permanent magnet is normal, and based on the generated magnetic field. And diagnosing that the rotor is eccentric if the calculated differential inductance is less than the reference differential inductance calculated and retained in the state where the permanent magnet is normal. (a) generating a plurality of magnetic fields in which the direct current magnetic field and the alternating magnetic field overlap in the pole direction of the permanent magnet when the rotor is in a stationary state;
(b) measuring a current flowing in the stator windings corresponding to each of the generated magnetic fields; And
(c) the potato state of the permanent magnet and the eccentricity of the rotor on the basis of the change of the differential inductance with respect to the polar component value of the permanent magnet of the current flowing in the stator windings measured corresponding to each of the generated magnetic fields. Diagnosing the condition;
The differential inductance is the pole of the permanent magnet of flux linkage measured from the alternating magnetic field component with respect to the amount of change in the polar component value of the permanent magnet of the current by the alternating magnetic field component contained in the generated magnetic field. A method for diagnosing a condition of a permanent magnet synchronous motor, characterized in that it is defined by the amount of change in the direction component value. The method of claim 3,
In the step (c), if the differential inductance calculated based on the generated magnetic field is larger than the reference differential inductance calculated and retained in the state in which the permanent magnet is normal, the permanent magnet is diagnosed as demagnetized, and the generated magnetic field And diagnosing that the rotor is eccentric if the differential inductance calculated on the basis of the differential inductance is less than the reference differential inductance calculated and retained in the normal state of the permanent magnet. A computer-readable recording medium having recorded thereon a program for causing a computer to execute the method for diagnosing a condition of a permanent magnet synchronous motor according to claim 3 or 4.

Images