A temperature prediction method for permanent magnet synchronous motor under turn-to-turn short circuit fault

Abstract

The present invention provides a method for predicting the temperature of a permanent magnet synchronous motor under a turn-to-turn short circuit fault, and relates to the field of temperature monitoring of key positions of a motor. The method for predicting the temperature of a permanent magnet synchronous motor under a turn-to-turn short circuit fault comprises: based on the physical principle of the permanent magnet synchronous motor system, obtaining an equation describing the relationship between the stator winding copper loss and the phase current change, an equation describing the relationship between the motor iron loss and its frequency and synthetic back electromotive force change, and an equation describing the relationship between the mechanical loss and the motor speed change; according to the physical principle of the permanent magnet synchronous motor system under normal and faulty operating states, respectively establishing a mathematical model characterizing the motor loss under normal operating conditions and a mathematical model of the motor additional copper loss under faulty operating conditions, further, based on the physical principle of heat transfer of the permanent magnet synchronous motor system, establishing a four-node low-order thermal network model of the motor under faulty operating conditions, and online predicting the motor temperature under turn-to-turn short circuit faults by measuring the motor operating information in real time.

Description

Technical Field

[0001] The invention relates to the technical field of temperature monitoring at key positions of motors, and in particular to a method for predicting the temperature of a permanent magnet synchronous motor under a turn-to-turn short circuit fault. Background Art

[0002] Motors will inevitably fail when they work for a long time under harsh conditions such as humidity, pollution, and mechanical vibration. In some important fields, the failure of a component may cause the motor to unexpectedly stop, ultimately causing huge economic losses. Turn-to-turn short circuit failure is one of the most common motor failures. The motor's induced back electromotive force will generate a short-circuit current far higher than the rated value in the short-circuit loop of the winding, which will not only reduce the output torque, generate vibration and noise, but also cause the temperature of the faulty part to rise sharply, destroying the motor insulation system. It is the initial cause of many major motor failures.

[0003] For motor fault-tolerant control after a turn-to-turn short-circuit fault occurs, existing studies often eliminate the torque fluctuation of the faulty motor by current injection, but ignore the thermal effect caused by the short-circuit current and cannot solve the problem of fault propagation caused by local high temperature. Therefore, obtaining the temperature information of the motor under turn-to-turn short-circuit fault and then implementing fault-tolerant control is of great significance to improving the reliability of the motor.

[0004] For the temperature monitoring of key parts of motors, using sensors to measure temperature is the most commonly used method in industrial applications. However, the location of the turn-to-turn short-circuit fault is uncertain. Accurately measuring the temperature of the fault point requires embedding sensors in each coil, which is obviously very uneconomical. At the same time, the sensors embedded in the motor cannot be repaired or replaced. Once damaged, they lose the ability to monitor temperature. In the non-sensor temperature monitoring method, some studies have used the thermal properties of copper resistors to estimate the temperature of the motor windings, but this result is only the average temperature of the entire winding and cannot accurately reflect the hot spot temperature of the faulty coil. Existing studies use complex thermal network models to describe the heat transfer process of the motor, which can accurately predict the motor temperature, but requires a lot of computing resources and can only be applied to offline simulation. On this basis, the low-order thermal network method only retains the main thermal circuit of the motor, combines experimental data training to identify the thermal circuit parameters, has small calculation amount and high accuracy, and is very suitable for online motor temperature prediction, but this method has not yet been applied to the temperature prediction of the motor under turn-to-turn short circuit. Summary of the invention

[0005] 1. Technical issues to be resolved

[0006] In view of the shortcomings of the prior art, the present invention provides a method for predicting the temperature of a permanent magnet synchronous motor under a turn-to-turn short-circuit fault. According to the physical principles of the permanent magnet synchronous motor system under normal and faulty operating states, a mathematical model for characterizing the motor loss under normal operating conditions and a mathematical model for characterizing the additional copper loss of the motor under faulty operating conditions are established respectively. Furthermore, based on the physical principles of heat transfer in the permanent magnet synchronous motor system, a four-node low-order thermal network model of the motor under faulty operating conditions is established. By measuring the motor operating information in real time, the motor temperature under a turn-to-turn short-circuit fault is predicted online.

[0007] (II) Technical solution

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0009] A method for predicting the temperature of a permanent magnet synchronous motor under a turn-to-turn short circuit fault, comprising:

[0010] Based on the physical principle of permanent magnet synchronous motor system, the loss model of the motor under normal operation is established, and the equation describing the relationship between the stator winding copper loss and the phase current change, the equation describing the relationship between the motor iron loss and its frequency and synthetic back electromotive force change, and the equation describing the relationship between the mechanical loss and the motor speed change are obtained;

[0011] The parameters in the motor iron loss and mechanical loss equations are unified as undetermined loss parameters. The current and loss data of the motor under normal operation are collected through experiments. The loss parameters are identified offline using the particle swarm optimization algorithm, and an online calculation model for motor loss is established.

[0012] Based on the physical principle of permanent magnet synchronous motor system under turn-to-turn short-circuit fault, a mathematical model of permanent magnet synchronous motor under fault operation state is established, and the additional copper loss model generated by short-circuit current is derived and established, and the equation describing the relationship between the copper loss of the phase where the fault occurs and the change of its phase voltage is obtained.

[0013] Based on the physical principle of heat transfer in permanent magnet synchronous motor system under inter-turn short circuit, a four-node low-order thermal network model of the motor under faulty operation is established, and the first-order transient equation describing the relationship between the temperature and loss change of the motor faulty coil, healthy coil, stator core and rotor is obtained.

[0014] The thermal resistance, thermal capacity and stator-rotor iron loss ratio in the thermal network model are regarded as undetermined parameters. The temperature and loss data of the motor under inter-turn short-circuit fault operation are collected experimentally, and the parameters are identified offline by combining the particle swarm optimization algorithm.

[0015] The current, voltage and speed information of the motor under faulty operation are measured in real time, the motor loss is calculated online, the room temperature is measured in real time, and the temperature of the motor faulty coil, healthy coil, stator core and rotor part is predicted online through the four-node thermal network model under faulty operation.

[0016] Preferably, establishing a loss model of the motor under normal operating conditions specifically includes:

[0017] Iron loss model affected by iron loss coefficient,

[0018]

[0019] Where, f represents the motor frequency; B m Indicates the maximum magnetic flux density;

[0020] The hysteresis coefficient k h and eddy current coefficient k e They are respectively expressed as cubic polynomials related to magnetic flux density, and the model describing the relationship between motor iron loss and its frequency and back EMF is obtained.

[0021]

[0022] Among them, k h0 , k h1 , k h2 , k h3 and k e0 , k e1 , k e2 , k e3 Respectively represent the constant coefficients of the equation; ω represents the electrical angular frequency; V m The equation that represents the motor's synthetic back EMF and describes its relationship with the change of dq axis current is:

[0023]

[0024] Among them, ψ PM Indicates the size of the permanent magnet flux of the motor; L d Indicates the d-axis inductance; L q represents the q-axis inductance.

[0025] Preferably, the establishment of an online calculation model for motor losses specifically includes:

[0026] Ignore the eddy current loss of permanent magnets and take the total motor loss P loss Minus copper loss P cu Then the remaining loss P is obtained si , by the motor iron loss P Fe and mechanical loss P me composition:

[0027]

[0028] Among them, k fri represents the friction loss coefficient; k windIndicates the wind friction loss coefficient. The motor friction loss is related to the speed, and the wind friction loss is related to the cube of the speed.

[0029] Preferably, the mathematical model of the permanent magnet synchronous motor in the faulty operating state specifically includes:

[0030] According to the physical principle of permanent magnet synchronous motor under turn-to-turn short circuit fault, the first-order transient equation describing the relationship between the voltage, current and flux change of the motor under turn-to-turn short circuit fault is obtained:

[0031]

[0032] In which, it is assumed that phase A is the phase with inter-turn short circuit fault, u sf 、i sf , sf They represent the motor stator phase voltage, phase current and winding flux matrix under the fault condition including the short circuit loop; R sf represents the stator resistance matrix under fault; L sf Represents the stator inductance matrix under fault.

[0033] Preferably, the first-order transient equation describing the relationship between the voltage, current and flux change of the motor during a turn-to-turn short circuit fault is specifically:

[0034] i sf =[i a i b i c i af ] T ;u sf =[u a u b u c 0] T

[0035]

[0036]

[0037]

[0038] Considering the relationship between the symmetry of the motor structure and the fault phase inductance, the stator inductance matrix under fault is further:

[0039]

[0040]

[0041] Among them, R s Indicates stator resistance; L l Indicates the stator winding self-leakage inductance; L ms Indicates the main self-inductance of the stator winding; Lδ Represents the magnetoresistance inductance caused by the salient pole effect; M 1 Represents the mutual inductance between the three-phase windings; R f represents the short-circuit resistance; μ represents the ratio of the number of turns of the short-circuit winding to the total winding, which is called the short-circuit turns ratio; the parameter m is determined by the inductance of a single coil and is related to the motor structure.

[0042] Preferably, the establishment of a four-node low-order thermal network model of the motor in a faulty operating state specifically includes:

[0043] According to the symmetry of the motor structure, after a turn-to-turn short-circuit fault occurs at any position, the winding can be divided into a faulty coil and a healthy coil. The faulty coil refers to the coil with a short-circuit loop in the stator winding, and the healthy coil refers to the other coils without short-circuit fault. The state space expression of the motor four-node thermal network model under turn-to-turn short-circuit fault is established:

[0044]

[0045] Among them, T f Indicates the motor temperature state vector under fault; u f Represents the input vector; A f represents the system matrix; B f Represents the input matrix.

[0046] Preferably, the specific form of each variable in the state space expression of the four-node thermal network model of the motor under the inter-turn short circuit fault is:

[0047]

[0048] T f =[T s T wh T r T wf ] T

[0049] u f =[P s P wh P r P wf T e ] T

[0050]

[0051]

[0052] Among them, T s , T r , T wh , T wf , Te Respectively represent the stator core, rotor, healthy coil, faulty coil temperature and room temperature; P s , P r , P wh , P wf Respectively represent the stator core, rotor, healthy coil and faulty coil losses; C s , C r , C wh , C wf Respectively represent the heat capacity of stator core, rotor, healthy coil and faulty coil; R swh , R swf , R se Respectively represent the thermal resistance between the motor stator core and the healthy coil, the faulty coil and the room temperature; R sr , R re They represent the connection thermal resistance between the motor rotor and stator core and room temperature respectively.

[0053] Preferably, the input variables of the four-node thermal network model of the motor under the inter-turn short circuit fault are calculated as follows:

[0054]

[0055] Among them, k rs Indicates the proportion of stator iron loss to the total iron loss of the motor; P cu_M Indicates the healthy copper loss of the Mth phase winding; P cu_ah Indicates the copper loss of the healthy coil in the fault phase; P cu_af Represents the copper loss of the faulty coil in the faulty phase. The calculation equation based on the faulty phase voltage is as follows:

[0056]

[0057]

[0058] Among them, n 1 Indicates the number of coils contained in a single-phase winding; R sh Indicates the stator resistance value at healthy coil temperature; R sf It represents the stator resistance value at the fault coil temperature, and the calculation equation is as follows:

[0059]

[0060] Among them, T 0 Indicates the basic temperature value; R s0 Indicates temperature as T 0 The stator resistance value at α 0 Represents the temperature coefficient of resistance of copper.

[0061] (III) Beneficial effects

[0062] The present invention discloses a method for predicting the temperature of a permanent magnet synchronous motor under a turn-to-turn short-circuit fault. According to the physical principles of the permanent magnet synchronous motor system under normal and faulty operating states, a mathematical model for characterizing the motor loss under normal operating state and a mathematical model for characterizing the additional copper loss of the motor under faulty operating state are respectively established. Furthermore, based on the physical principles of heat transfer of the permanent magnet synchronous motor system, a four-node low-order thermal network model of the motor under faulty operating state is established. By measuring the motor operating information in real time, the motor temperature under a turn-to-turn short-circuit fault is predicted online. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 It is a flow chart of the present invention;

[0064] Figure 2 The experimental results of online calculation of copper loss of fault phase of the present invention are as follows;

[0065] Figure 3 The online calculation experimental result of the total motor loss under fault conditions of the present invention;

[0066] Figure 4 It is a schematic diagram of a four-node thermal network model of a motor with a turn-to-turn short-circuit fault according to the present invention;

[0067] Figure 5 This is a schematic diagram of online prediction of the temperature of a motor with a turn-to-turn short-circuit fault according to the present invention;

[0068] Figure 6 The motor operating condition changes under the fault of the present invention;

[0069] Figure 7 The temperature prediction experiment results of the four-node thermal network model under fault conditions of the present invention are shown. DETAILED DESCRIPTION

[0070] The following will be combined with the accompanying drawings of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0071] Example

[0072] A method for predicting the temperature of a permanent magnet synchronous motor under a turn-to-turn short circuit fault. The flow chart of the method is as follows: Figure 1 , specifically including the following steps:

[0073] S1, based on the physical principle of permanent magnet synchronous motor system, establish the loss model of the motor under normal operation, and obtain the equation describing the relationship between stator winding copper loss and phase current change, the equation describing the relationship between motor iron loss and its frequency and synthetic back electromotive force change, and the equation describing the relationship between mechanical loss and motor speed change;

[0074] S2, the parameters in the motor iron loss and mechanical loss equations are unified as undetermined loss parameters, the current and loss data of the motor under normal operation are collected through experiments, the loss parameters are identified offline in combination with the particle swarm optimization algorithm, and an online calculation model for motor loss is established;

[0075] S3, based on the physical principle of the permanent magnet synchronous motor system under turn-to-turn short-circuit fault, a mathematical model of the permanent magnet synchronous motor under fault operation is established, and then the additional copper loss model generated by the short-circuit current is derived to obtain an equation describing the relationship between the copper loss of the phase where the fault is located and the change in its phase voltage;

[0076] S4, based on the physical principle of heat transfer of permanent magnet synchronous motor system under inter-turn short circuit, a four-node low-order thermal network model of the motor under fault operation state is established, and the first-order transient equation describing the relationship between the temperature and loss change of the motor fault coil, healthy coil, stator core and rotor is obtained;

[0077] S5, taking the thermal resistance, thermal capacity and stator-rotor iron loss ratio in the thermal network model as undetermined parameters, collecting the temperature and loss data under the operation of the motor turn-to-turn short-circuit fault through experiments, and combining the particle swarm optimization algorithm to perform offline identification of the parameters;

[0078] S6, real-time measurement of the current, voltage and speed information of the motor under faulty operation state, online calculation of motor loss. Then, real-time measurement of room temperature, online prediction of the temperature of the motor faulty coil, healthy coil, stator core and rotor part through the four-node thermal network model under fault.

[0079] The simplified Steinmetz formula in step S1 is as follows:

[0080]

[0081] Where, f represents the motor frequency; B m Indicates the maximum magnetic flux density. Considering the influence of magnetic flux density on the iron loss coefficient in the simplified Steinmetz formula, the hysteresis coefficient k h and eddy current coefficient k e They are respectively expressed as cubic polynomials related to magnetic flux density, and the model describing the relationship between motor iron loss and its frequency and back EMF is obtained:

[0082]

[0083] Among them, k h0 , k h1 , k h2 , k h3 and k e0 , k e1 , k e2 , k e3 Respectively represent the constant coefficients of the equation; ω represents the electrical angular frequency; V m The equation that represents the motor's synthetic back EMF and describes its relationship with the dq axis current change is as follows:

[0084]

[0085] Among them, ψ PM Indicates the size of the permanent magnet flux of the motor; L d Indicates the d-axis inductance; L q represents the q-axis inductance.

[0086] In step S2, the permanent magnet eddy current loss is ignored and the total motor loss P is loss Minus copper loss P cu Then the remaining loss P is obtained si , by the motor iron loss P Fe and mechanical loss P me composition:

[0087]

[0088] Among them, k fri represents the friction loss coefficient; k wind Represents the wind-friction loss coefficient. The motor friction loss is related to the speed, and the wind-friction loss is related to the cube of the speed. The above formula reflects the relationship between the motor's residual loss and the synthetic back EMF and speed. The residual loss can be calculated by measuring the input power, output power and winding copper loss under normal operation of the motor. Therefore, by experimentally collecting multiple sets of data of the motor under different working conditions, and combining the particle swarm optimization algorithm to identify the loss parameters offline, a more accurate motor loss model can be established and applied to online loss calculation.

[0089] In step S3, according to the physical principle of the permanent magnet synchronous motor under the inter-turn short circuit fault, a first-order transient equation describing the relationship between the voltage, current and flux change of the motor under the inter-turn short circuit fault is obtained:

[0090]

[0091] In which, it is assumed that phase A is the phase with inter-turn short circuit fault, u sf 、i sf , sf They represent the motor stator phase voltage, phase current and winding flux matrix under the fault condition including the short circuit loop; Rsf represents the stator resistance matrix under fault; L sf Represents the stator inductance matrix under fault. The detailed expansion of each matrix is ​​as follows:

[0092] i sf =[i a i b i c i af ] T ;u sf =[u a u b u c 0] T ;

[0093]

[0094]

[0095] Considering the symmetry of the motor structure and the relationship between the fault phase inductance, the stator inductance matrix under fault can be further written as:

[0096]

[0097] Among them, R s Indicates stator resistance; L l Indicates the stator winding self-leakage inductance; L ms Indicates the main self-inductance of the stator winding; L δ Represents the magnetoresistance inductance caused by the salient pole effect; M 1 Represents the mutual inductance between the three-phase windings; R f Represents the short-circuit resistance; μ represents the ratio of the number of turns of the short-circuit winding to the total winding, which is called the short-circuit turns ratio. The above formula can be rearranged to obtain the equation describing the relationship between the fault phase voltage and the short-circuit current:

[0098]

[0099] The traditional calculation formula for the copper loss of the phase winding where the fault is located is known to be:

[0100]

[0101] It can be seen from this formula that the traditional copper loss calculation of the fault phase requires three fault parameters: short-circuit current, short-circuit resistance and short-circuit turns ratio. However, since the inter-turn short-circuit fault occurs inside the motor, it is impossible to measure the short-circuit current through sensors, nor is it possible to measure the short-circuit resistance and short-circuit turns ratio online or offline.

[0102] Combining the above two equations, we can get the equation describing the relationship between the copper loss of the fault phase and the change of its phase voltage:

[0103]

[0104] The variables required to calculate the copper loss of the fault phase through the above formula are: fault severity F, satisfying F = μi af , which can be obtained through existing fault diagnosis technology; the fault phase voltage u a , for motors with neutral point taps, they can be directly measured; motor self-inductance L, reluctance inductance L δ , which can be obtained through finite element simulation or experiment.

[0105] Set up different experimental fault conditions and motor operating conditions, measure the actual value of copper loss, and compare it with the calculated value based on the fault phase voltage. It is worth noting that since the inverter uses PWM modulation, the phase voltage and line voltage of the motor are high-frequency chopped and cannot be measured directly. Therefore, an LC low-pass filter with a cutoff frequency of 700Hz is used to filter the fault phase voltage, and the magnitude of the filtered phase voltage is measured. Assuming the fault severity F (F = μi af ) is completely accurate, such as Figure 2 The figure shows the comparison of the actual value of the copper loss of the fault phase under different fault conditions and the experimental results of the model calculation value. The circles represent the actual values, which are calculated by the traditional copper loss formula; the cross lines represent the loss model calculation values, which are calculated by the copper loss formula based on phase voltage. It can be seen from the results that the copper loss calculation method based on the fault phase voltage has a higher accuracy, and the maximum error is about 2W.

[0106] Furthermore, the total motor loss model under fault conditions is experimentally verified. The actual total motor loss under fault conditions is measured experimentally and compared with the model calculated value. Figure 3 As shown in the figure, the circles represent the experimental measurement values, which are obtained by subtracting the output power from the input power; the cross lines represent the calculated values ​​of the motor loss model, where the iron loss and mechanical loss are calculated using the model under normal motor operation, the copper loss of the healthy phase is calculated using the traditional copper loss, and the copper loss of the fault phase is calculated using the copper loss based on the phase voltage. It can be seen from the experimental results that the calculated values ​​of the motor loss model under fault established in this paper are basically consistent with the experimental measurement values, indicating that the iron loss and mechanical loss model under normal operation can still be applied to fault conditions, thereby obtaining more accurate online loss calculation values ​​and providing effective loss information for the thermal network model under fault.

[0107] In step S4, according to the symmetry of the motor structure, after a turn-to-turn short circuit fault occurs at any position, the winding can be divided into a faulty coil and a healthy coil. Figure 4 As shown in the figure, the faulty coil refers to the coil with a short circuit in the stator winding, and the healthy coil refers to the other coils without short circuit faults. The state space expression of the motor four-node thermal network model under turn-to-turn short circuit fault is as follows:

[0108]

[0109] Among them, T f Indicates the motor temperature state vector under fault; u f Represents the input vector; A f represents the system matrix; B f Represents the input matrix. The specific forms of each variable are as follows:

[0110] T f =[T s T wh T r T wf ] T ; u f =[P s P wh P r P wf T e ] T ;

[0111]

[0112]

[0113] Among them, T s , T r , T wh , T wf , T e Respectively represent the stator core, rotor, healthy coil, faulty coil temperature and room temperature; P s , P r , P wh , P wf Respectively represent the stator core, rotor, healthy coil and faulty coil losses; C s , C r , C wh , C wf Respectively represent the heat capacity of stator core, rotor, healthy coil and faulty coil; R swh , R swf , R se Respectively represent the thermal resistance between the motor stator core and the healthy coil, the faulty coil and the room temperature; R sr , R re They represent the connection thermal resistance between the motor rotor and stator core and room temperature respectively. Due to the heat dissipation method of heat convection in these two thermal circuits, the above two thermal resistances will change with the motor speed and show a negative correlation.

[0114] The input variables of the model are calculated as follows:

[0115]

[0116] Among them, k rs Indicates the proportion of stator iron loss to the total iron loss of the motor; P cu_M Indicates the healthy copper loss of the Mth phase winding; P cu_ah Indicates the copper loss of the healthy coil in the fault phase; P cu_af Represents the copper loss of the faulty coil in the faulty phase. The calculation equation based on the faulty phase voltage is as follows:

[0117]

[0118]

[0119] Among them, R sh Indicates the stator resistance value at healthy coil temperature; R sf It represents the stator resistance value at the fault coil temperature, and the calculation equation is as follows:

[0120]

[0121] Among them, T 0 Indicates the basic temperature value; R s0 Indicates temperature as T 0 The stator resistance value at α 0 Represents the temperature coefficient of resistance of copper.

[0122] In the step S5, a single coil of the test motor is short-circuited, a short-circuit turn ratio of 0.25 and a short-circuit resistance of 0.059Ω are set, and a temperature rise experiment is performed to simulate a turn-to-turn short-circuit fault condition. The temperature and loss data of the motor running steadily for 30 minutes at different speeds are recorded, including the temperature of the stator core, rotor, normal and faulty coils, the copper loss of normal coils and faulty coils, and the iron loss and mechanical loss of the stator and rotor. The data sampling frequency is taken as 1Hz, and the thermal convection resistance is proportional to the inverse of the motor speed. The thermal capacity, thermal resistance and iron loss ratio parameters are identified offline by the particle swarm optimization algorithm. After the thermal network parameters are known, the temperature online prediction method is as follows Figure 5 As shown in the figure, the fault diagnosis technology is used to obtain the severity of the motor fault, and the fault phase voltage and other information are measured, which are then brought into the loss model for online calculation to obtain the motor loss value. Then, the loss value is brought into the fault thermal network model, combined with the room temperature and speed measurement values, and the temperature of each part of the motor is predicted online.

[0123] The temperature prediction of the four-node thermal network model under turn-to-turn short-circuit fault is experimentally verified. By continuously adjusting the reference speed and load torque, the motor temperature rise is changed, and the waveform of the working condition change is as follows: Figure 6(a) Motor speed change, (b) Motor torque change. At the same time, the fault parameters are changed to verify the accuracy of temperature prediction under different fault conditions. The working condition changes are shown in Figure 6 (c) short-circuit resistance change, (d) short-circuit turns ratio change. Figure 7 The figure shows the comparison between the measured and predicted values ​​of the temperature changes of various parts of the motor. The red line on the left shows the predicted temperature value, and the blue line shows the actual measured value; the figure on the right shows the error between the measured and predicted values. It can be seen from the experimental results that during the operation of the motor, the temperature prediction values ​​of each part can guarantee a high degree of accuracy, with the maximum error not exceeding 1.5°C, which proves the accuracy of the model.

[0124] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or equipment. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or equipment including the elements.

Claims

1. A temperature prediction method for permanent magnet synchronous motors under inter-turn short-circuit faults, characterized in that, it includes: Based on the physical principles of the permanent magnet synchronous motor system, establish a loss model of the motor under normal operating conditions, and obtain equations describing the relationship between stator winding copper loss and phase current change, equations describing the relationship between motor iron loss and its frequency and synthetic back electromotive force change, and equations describing the relationship between mechanical loss and motor speed change; Unify the parameters in the motor iron loss and mechanical loss equations as undetermined loss parameters, collect current and loss data of the motor under normal operation through experiments, and combine the particle swarm optimization algorithm to perform offline identification of the loss parameters to establish an online calculation model of motor loss; Based on the physical principles of the permanent magnet synchronous motor system under inter-turn short-circuit faults, establish a mathematical model of the permanent magnet synchronous motor under fault operating conditions, deduce and establish an additional copper loss model generated by the short-circuit current, and obtain an equation describing the relationship between the copper loss of the faulty phase and its phase voltage change; Based on the physical principle of heat transfer of the permanent magnet synchronous motor system under inter-turn short-circuit faults, establish a four-node low-order thermal network model of the motor under fault operating conditions, and obtain a first-order transient equation describing the relationship between the temperatures of the motor's faulty coil, healthy coil, stator core, and rotor and loss changes; Regard the thermal resistance, heat capacity, and stator-rotor iron loss ratio in the thermal network model as undetermined parameters, collect temperature and loss data of the motor during inter-turn short-circuit fault operation through experiments, and combine the particle swarm optimization algorithm to perform offline identification of the parameters; Measure the current, voltage, and speed information of the motor in real time under fault operating conditions, calculate the motor loss online, measure the room temperature in real time, and perform online prediction of the temperatures of the motor's faulty coil, healthy coil, stator core, and rotor parts through the four-node thermal network model under faults; The establishment of the loss model of the motor under normal operating conditions specifically includes: The iron loss model affected by the iron loss coefficient, Where, f represents the motor frequency; B m Indicates the maximum magnetic flux density; The hysteresis coefficient k h and eddy current coefficient k e They are respectively expressed as cubic polynomials related to magnetic flux density, and the model describing the relationship between motor iron loss and its frequency and back EMF is obtained. Among them, k h0 , k h1 , k h2 , k h3 and k e0 , k e1 , k e2 , k e3 They represent the constant coefficients of the equation respectively; ω represents the electrical angular frequency; V m The equation that represents the motor's synthetic back EMF and describes its relationship with the change of dq axis current is: Among them, ψ PM Indicates the size of the permanent magnet flux of the motor; L d Indicates the d-axis inductance; L q represents the q-axis inductance; The establishment of the online calculation model of motor loss specifically includes: Ignore the eddy current loss of permanent magnets and take the total motor loss P loss Minus copper loss P cu Then the remaining loss P is obtained si , by the motor iron loss P Fe and mechanical loss P me composition: Among them, k fri represents the friction loss coefficient; k wind Indicates the wind friction loss coefficient; the motor friction loss is related to the speed, and the wind friction loss is related to the cube of the speed; The mathematical model of the permanent magnet synchronous motor under fault operating conditions specifically includes: According to the physical principles of the permanent magnet synchronous motor under inter-turn short-circuit faults, obtain a first-order transient equation describing the relationship between the voltage, current, and magnetic flux of the motor with inter-turn short-circuit faults: In which, it is assumed that phase A is the phase with inter-turn short circuit fault, u sf 、i sf , sf They represent the motor stator phase voltage, phase current and winding flux matrix under the fault condition including the short circuit loop; R sf represents the stator resistance matrix under fault; L sf represents the stator inductance matrix under fault; The specific form of the first-order transient equation describing the relationship between the voltage, current, and magnetic flux of the motor with inter-turn short-circuit faults is: and sf =[and a and b and c and af ] T ;in sf =[in a in b in c 0] T Considering the symmetry of the motor structure and the relationship between the inductance of the faulty phase, the stator inductance matrix under faults is further: Among them, R s Indicates stator resistance; L l Indicates the stator winding self-leakage inductance; L ms Indicates the main self-inductance of the stator winding; L δ Represents the magnetoresistance inductance caused by the salient pole effect; M 1 Represents the mutual inductance between the three-phase windings; R f represents the short-circuit resistance; μ represents the ratio of the number of turns of the short-circuit winding to the total winding, which is called the short-circuit turns ratio; the parameter m is determined by the inductance of a single coil and is related to the motor structure; The establishment of the four-node low-order thermal network model of the motor under fault operating conditions specifically includes: According to the symmetry of the motor structure, after an inter-turn short-circuit fault occurs at any position, the winding can be divided into two parts: the faulty coil and the healthy coil. The faulty coil is the coil in the stator winding with a short-circuit loop, and the healthy coil is the coil without a short-circuit fault. Establish the state space expression of the four-node thermal network model of the motor under inter-turn short-circuit faults: Among them, T f Indicates the motor temperature state vector under fault; u f Represents the input vector; A f represents the system matrix; B f represents the input matrix; The specific forms of the variables in the state space expression of the four-node thermal network model of the motor under inter-turn short-circuit faults are: T f =[T s T wh T r T wf ] T u f =[P s P wh P r P wf T e ] T Among them, T s , T r , T wh , T wf , T e Respectively represent the stator core, rotor, healthy coil, faulty coil temperature and room temperature; P s , P r , P wh , P wf Respectively represent the stator core, rotor, healthy coil and faulty coil losses; C s , C r , C wh , C wf Respectively represent the heat capacity of stator core, rotor, healthy coil and faulty coil; R swh , R swf , R se Respectively represent the thermal resistance between the motor stator core and the healthy coil, the faulty coil and the room temperature; R sr , R re They represent the connection thermal resistance between the motor rotor and the stator core and room temperature respectively; The input variables of the four-node thermal network model of the motor under inter-turn short-circuit faults are calculated as follows: Among them, k rs Indicates the proportion of stator iron loss to the total iron loss of the motor; P cu_M Indicates the healthy copper loss of the Mth phase winding; P cu_ah Indicates the copper loss of the healthy coil in the fault phase; P cu_af Represents the copper loss of the faulty coil in the faulty phase. The calculation equation based on the faulty phase voltage is as follows: Among them, n 1 Indicates the number of coils contained in a single-phase winding; R sh Indicates the stator resistance value at healthy coil temperature; R sf It represents the stator resistance value at the fault coil temperature, and the calculation equation is as follows: R sh =R s0 [1+α(T wh -T 0 )] R sf =R s0 [1+α(T wf -T 0 )] Among them, T 0 Indicates the basic temperature value; R s0 Indicates temperature as T 0 The stator resistance value at α 0 Represents the temperature coefficient of resistance of copper.

Images