CN103940392B - The rotor-position of a kind of magnetic suspension switched reluctance motor/displacement self-sensing method - Google Patents
The rotor-position of a kind of magnetic suspension switched reluctance motor/displacement self-sensing method Download PDFInfo
- Publication number
- CN103940392B CN103940392B CN201410154098.6A CN201410154098A CN103940392B CN 103940392 B CN103940392 B CN 103940392B CN 201410154098 A CN201410154098 A CN 201410154098A CN 103940392 B CN103940392 B CN 103940392B
- Authority
- CN
- China
- Prior art keywords
- displacement
- rotor
- magnetic suspension
- learning machine
- reluctance motor
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000006073 displacement reaction Methods 0.000 title claims abstract description 53
- 239000000725 suspension Substances 0.000 title claims abstract description 35
- 238000000034 method Methods 0.000 title abstract description 20
- 238000012549 training Methods 0.000 claims abstract description 27
- 238000001514 detection method Methods 0.000 claims abstract description 20
- 238000013461 design Methods 0.000 claims abstract description 11
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000005070 sampling Methods 0.000 claims description 8
- 230000005284 excitation Effects 0.000 claims description 3
- 230000006870 function Effects 0.000 claims description 3
- 238000004806 packaging method and process Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 230000002194 synthesizing effect Effects 0.000 claims description 3
- 238000004804 winding Methods 0.000 claims description 3
- 239000000203 mixture Substances 0.000 claims description 2
- 238000013528 artificial neural network Methods 0.000 abstract description 7
- 238000009434 installation Methods 0.000 abstract description 4
- 238000002347 injection Methods 0.000 description 3
- 239000007924 injection Substances 0.000 description 3
- 238000013178 mathematical model Methods 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000004323 axial length Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000005339 levitation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Landscapes
- Control Of Electric Motors In General (AREA)
Abstract
The invention discloses the rotor-position/displacement self-sensing method of a kind of magnetic suspension switched reluctance motor, including step: obtain training sample set；Extreme learning machine model training；Rotor-position/displacement observation device design.Present invention achieves the sensorless strategy of magnetic suspension switched reluctance motor, it is to avoid sensor uses the installation brought to safeguard the problems such as inconvenience, poor reliability, increase motor radial-axial size；Position/displacement observation the device of the INTELLIGENT IDENTIFICATION method design rotor of limit of utilization learning machine, realize the estimation of magnetic suspension switched reluctance motor rotor-position/displacement, effectively prevent conventional observation device method to accurate model and the dependence of operational factor, improve detection robustness and capacity of resisting disturbance；The learning algorithm design magnetic suspension switched reluctance motor rotor-position/displacement observation device of operating limit learning machine, it is to avoid the problems such as traditional neural network training sample amount is big, pace of learning is slow, it is achieved that quick, the accurate modeling of Small Sample Database.
Description
Technical Field
The invention belongs to the field of magnetic suspension switched reluctance motors, and is used for estimating the position/displacement of a rotor of a magnetic suspension switched reluctance motor.
Background
The control of the magnetic suspension switched reluctance motor comprises the following steps: the suspension control and the rotation control of the rotor are realized on the basis of accurate detection of the radial displacement and the angular position of the rotor. At present, the radial displacement and angular position of a magnetic suspension switch reluctance motor are usually detected by adopting an eddy current type sensor, a Hall type sensor or a photoelectric type sensor, the operation of the magnetic suspension switch reluctance motor inevitably causes strong electromagnetic interference to the sensor, in addition, the installation of the sensor needs extra axial length of the motor, the critical rotating speed of the motor is reduced, and meanwhile, the radial length of the motor possibly increased by the installation of the sensor enables the volume of the motor to be enlarged, and the use of the magnetic suspension switch reluctance motor in certain occasions with specific requirements on the volume of the motor is limited. In view of the above problems of the sensor, a position/displacement self-detection technology of the magnetic suspension switched reluctance motor has appeared.
The position/displacement self-detection method of the magnetic suspension switch reluctance motor comprises the following steps: the method comprises a high-frequency signal injection method, a salient pole tracking method, a direct calculation method, an observer design method and the like, wherein the high-frequency signal injection method and the salient pole tracking method need continuous loading of high-frequency signals and can be realized only by special hardware circuits, and the detected position/displacement signals have certain hysteresis, so that the high-frequency signal injection method and the salient pole tracking method are difficult to be suitable for the condition of high-speed operation of a motor; the direct calculation method is insensitive to the change of the motor operation parameters and lacks of anti-interference capability; the observer design method meets the requirements of real-time performance and accuracy of position/displacement detection under the condition of high-speed operation of the motor, but has strong dependence on a mathematical model and is insensitive to the change of motor operation parameters, so that the position/displacement observer is constructed by learning algorithms such as a neural network and a support vector machine, the dependence of the traditional observer design method on an accurate mathematical model and the motor operation parameters is avoided, and the detection robustness and the anti-interference capability are improved. However, the grid training of the traditional neural network (such as BP neural network, RBF neural network) requires a large amount of data, which inevitably reduces the learning speed, and meanwhile, the acquisition of a large amount of training data increases the experimental workload; the support vector machine is suitable for model training of small sample data, but the accuracy of a model obtained by training is general under the condition that parameters of the support vector machine are optimized by an intelligent algorithm.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a rotor position/displacement self-detection method of a magnetic suspension switched reluctance motor, which applies the learning algorithm of an extreme learning machine to the design of a magnetic suspension switched reluctance motor position/displacement observer as the learning algorithm which is also suitable for small sample data, wherein the extreme learning machine is a single-hidden-layer feedforward neural network, the learning process does not need iteration, the method has the characteristic of extreme rapidness, the model obtained by training has high precision, and the real-time and accurate detection of the motor rotor position/displacement signal is realized.
The technical scheme of the invention is as follows:
a rotor position/displacement self-detection method of a magnetic suspension switch reluctance motor comprises the following steps:
(1) obtaining a training sample set: random signals in the actual working range of the motor are selected as given values of exciting current, displacement and rotating speed, the output response of the magnetic suspension switched reluctance motor is sampled at high speed, and the sampling signals comprise three-phase winding currentRadial displacement of the magnetAngular position of rotorTo the sampled signalSignal processing is carried out to obtain respective second derivativeAnd synthesizing to obtain samples for training extreme learning machine；
(2) Training an extreme learning machine model: for a sample composition ofThe training sample set is obtained by training with an extreme learning machine asOutput isThe extreme learning machine model of (1);
(3) rotor position/displacement observer design: packaging the extreme learning machine model, and realizing signals inside the packagePrediction, double integral calculation andis updated to form an inputAnd an outputThe magnetic suspension switched reluctance motor rotor position/displacement observer realizes the radial displacement of the motor rotorAngular positionSelf-detection of (3).
Furthermore, in the learning algorithm setting of the extreme learning machine, the number of nodes of the hidden layer is usedNot greater than the number of training samplesPrinciple of (1) determinationNumerical, excitation function selection、Or。
Further, the extreme learning machine model output signalObtaining real-time rotor radial displacement through double integral link calculationAngular positionAnd fed back and inputted to the extreme learning machine model, and the input signal of the extreme learning machine model is updated。
The invention has the advantages that:
1. the sensorless control of the magnetic suspension switched reluctance motor is realized, and the problems of inconvenience in installation and maintenance, poor reliability, increase in the radial and axial size of the motor and the like caused by the use of a sensor are avoided;
2. the learning algorithm is used for training and modeling the data sample set, so that the dependence of a traditional observer design method on an accurate mathematical model and motor operating parameters is avoided, and the detection robustness and the anti-interference capability are improved;
3. the magnetic suspension switched reluctance motor rotor position/displacement observer is designed by using a learning algorithm of an extreme learning machine, the problems of large sample size, low learning speed and the like of the traditional neural network training are solved, and the rapid and accurate modeling of small sample data is realized.
Drawings
FIG. 1 is a schematic diagram of a magnetic levitation switched reluctance motor rotor position/displacement observer;
fig. 2 is an equivalent block diagram of fig. 1.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings.
When the extreme learning mechanism is adopted to build the position/displacement observer of the magnetic suspension switched reluctance motor, the sample data acquisition process of learning training is as follows:
in order to obtain representative and ergodic sample data, selecting a random signal in the actual working range of the motor as given values of exciting current, displacement and rotating speed, and sampling the output response of the magnetic suspension switched reluctance motor at a high speed, wherein the sampling signal comprises three-phase winding currentRadial displacement of the magnetAngular position of rotorTo the sampled signalSignal processing is carried out to obtain respective second derivativeAnd synthesizing to obtain sample data for extreme learning machine training. In the extreme learning machine learning algorithm setting: by implying the number of layer nodesNot greater than the number of training samplesReasonably determines the principle ofValues, excitation functions being selectable "”、“'or'"etc., for the above obtained training sample set, training is established as inputOutput isExtreme learning machine model, output signal of extreme learning machine modelObtaining real-time rotor radial displacement through double integral link calculationAngular positionAnd fed back and inputted to the extreme learning machine model, and the input signal of the extreme learning machine model is updated. Packaging the extreme learning machine model, as shown in FIG. 1, and implementing signals inside the packagePrediction, double integral calculation andthe feedback is updated, so that the input can be designed by using the extreme learning machineAnd an outputThe magnetic suspension switched reluctance motor rotor position/displacement observer realizes the radial displacement of the motor rotorAngular positionReal-time and accurate self-detection.
In order to facilitate the detailed description of the position/displacement self-detection principle of the rotor position/displacement observer of the magnetic suspension switched reluctance motor, the time is defined in time sequence、、…. At the moment when the magnetic suspension switched reluctance motor is started from the static state of the rotorUsing sensors to obtain initial position/displacement signalsWhile, at the same time, sampling to obtain a signal，Andsimultaneously inputting the extreme learning machine model to predict and obtain the momentIs as followsCalculated by the formula (1)Position/displacement signal of time of day：
(1)
Similarly, the calculation can be obtained by the iterative calculation of the formula (2)Position/displacement signal of time of day：
(2)
Wherein,andsimultaneously inputting extreme learning machine model and predicting output. From the above analysis, it can be known from the time of dayOnly need to sampleObtaining a current signalThe observer 2 can automatically output any time by the rotor position/displacement observer of the magnetic suspension switched reluctance motorPosition/displacement signal of. Finally, the preconditions for the establishment of equations (1), (2) are discussed:at any time during the time periodAre all equal toOf time of day(ii) a When the current signalAt high speed sampling, the sampling rate, accordingly,the time interval of (a) is small, and using knowledge of the limits:at any time during the time periodAre all approximately equal toOf time of dayEquations (1) and (2) hold.
The invention relates to a rotor position/displacement self-detection method of a magnetic suspension switched reluctance motor, which designs a position/displacement observer of a rotor by utilizing an intelligent identification method of an extreme learning machine, realizes the estimation of the position/displacement of the rotor of the magnetic suspension switched reluctance motor, effectively avoids the dependence of the traditional observer method on an accurate model and operation parameters, and improves the detection robustness and the anti-interference capability.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the present invention. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (3)
1. A rotor position and displacement self-detection method of a magnetic suspension switch reluctance motor comprises the following steps:
(1) obtaining a training sample set: random signals in the actual working range of the motor are selected as given values of exciting current, displacement and rotating speed, the output response of the magnetic suspension switched reluctance motor is sampled at high speed, and the sampling signals comprise three-phase winding current (i)A,iB,iC) Radial displacement (x, y) and rotor angular position theta, and processing the sampling signals (x, y, theta) to obtain respective second derivativesAnd synthesizing to obtain sample for training extreme learning machine
(2) Training an extreme learning machine model: for a sample composition ofTraining the training sample set with an extreme learning machine to obtain the input of (i)A,iB,iCX, y, θ) output isThe extreme learning machine model of (1);
(3) rotor position and displacement observer design: packaging the extreme learning machine model, and realizing signals inside the packageThe double integral calculation and the feedback update of (x, y, theta) to form the input (i)A,iB,iC) And the magnetic suspension switched reluctance motor rotor position and displacement observer outputs (x, y, theta) to realize the self-detection of the radial displacement (x, y) and the angular position theta of the motor rotor.
2. The self-detection method for the rotor position and the displacement of the magnetic suspension switched reluctance motor as claimed in claim 1, wherein: in the learning algorithm setting of the extreme learning machine, the value of L is determined on the basis that the number L of hidden layer nodes is not more than the number N of training samples, and the excitation function selects Sigmoid, Sine or RBF.
3. The self-detection method for the rotor position and the displacement of the magnetic suspension switched reluctance motor according to claim 1 or 2, characterized in that: the extreme learning machine model output signalAnd calculating by a double integral link to obtain real-time radial displacement (x, y) and angular position theta of the rotor, feeding back and inputting the real-time radial displacement (x, y) and angular position theta to the extreme learning machine model, and updating input signals (x, y, theta) of the extreme learning machine model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410154098.6A CN103940392B (en) | 2014-04-17 | 2014-04-17 | The rotor-position of a kind of magnetic suspension switched reluctance motor/displacement self-sensing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410154098.6A CN103940392B (en) | 2014-04-17 | 2014-04-17 | The rotor-position of a kind of magnetic suspension switched reluctance motor/displacement self-sensing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103940392A CN103940392A (en) | 2014-07-23 |
| CN103940392B true CN103940392B (en) | 2016-08-17 |
Family
ID=51188163
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410154098.6A Active CN103940392B (en) | 2014-04-17 | 2014-04-17 | The rotor-position of a kind of magnetic suspension switched reluctance motor/displacement self-sensing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103940392B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR3068465A1 (en) * | 2017-06-30 | 2019-01-04 | Safran Electronics & Defense | MEASURING METHOD USING AN INDUCTIVE DISPLACEMENT SENSOR |
| CN110007605B (en) * | 2019-05-20 | 2020-03-24 | 长沙学院 | Robust prediction control method of repelling magnetic suspension device |
| CN118111652B (en) * | 2024-04-23 | 2024-07-16 | 西南石油大学 | Distributed reluctance torsion vibration reduction test bed and test method thereof |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1551484A (en) * | 2003-04-24 | 2004-12-01 | 开关磁阻驱动有限公司 | Rotor position determination in a switched reluctance machine |
| CN1638259A (en) * | 2004-01-09 | 2005-07-13 | 开关磁阻驱动有限公司 | Rotor position detection of an electrical machine |
| CN1655438A (en) * | 2005-03-11 | 2005-08-17 | 江苏大学 | Magnetic levitation switch reluctance motor radial neural network reversed decoupling controller and method for constructing same |
| CN101958685A (en) * | 2010-03-04 | 2011-01-26 | 江苏大学 | Nonlinear inverse decoupling controller for bearingless synchronous reluctance motor and construction method thereof |
| CN103473598A (en) * | 2013-09-17 | 2013-12-25 | 山东大学 | Extreme learning machine based on length-changing particle swarm optimization algorithm |
| CN103647359A (en) * | 2013-12-13 | 2014-03-19 | 江苏大学 | Magnetic suspension switch magnetic resistance motor |
-
2014
- 2014-04-17 CN CN201410154098.6A patent/CN103940392B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1551484A (en) * | 2003-04-24 | 2004-12-01 | 开关磁阻驱动有限公司 | Rotor position determination in a switched reluctance machine |
| CN1638259A (en) * | 2004-01-09 | 2005-07-13 | 开关磁阻驱动有限公司 | Rotor position detection of an electrical machine |
| CN1655438A (en) * | 2005-03-11 | 2005-08-17 | 江苏大学 | Magnetic levitation switch reluctance motor radial neural network reversed decoupling controller and method for constructing same |
| CN101958685A (en) * | 2010-03-04 | 2011-01-26 | 江苏大学 | Nonlinear inverse decoupling controller for bearingless synchronous reluctance motor and construction method thereof |
| CN103473598A (en) * | 2013-09-17 | 2013-12-25 | 山东大学 | Extreme learning machine based on length-changing particle swarm optimization algorithm |
| CN103647359A (en) * | 2013-12-13 | 2014-03-19 | 江苏大学 | Magnetic suspension switch magnetic resistance motor |
Non-Patent Citations (4)
| Title |
|---|
| Extreme learning machine based phase angle control for stator-doubly-fed doubly salient motor for electric vehicles;Kong X，Cheng M，Shu Y;《Vehicle Power and Propulsion Conference》;20081231;1-5 * |
| 单绕组磁悬浮开关磁阻电机无径向位移传感器控制;项倩雯，孙玉坤，嵇小辅，张新华;《电工技术学报》;20130831;第28卷(第8期);259-267 * |
| 极限学习机的快速留一交叉验证算法;刘学艺，李平，郜传厚;《上海交通大学学报》;20110831;第45卷(第8期);1140-1145 * |
| 磁悬浮开关磁阻电机转子位移/位置观测器设计;朱志莹，孙玉坤，嵇小辅，黄永红;《中国电机工程学报》;20120425;第32卷(第12期);83-89 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103940392A (en) | 2014-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111600523B (en) | Model prediction current control method of permanent magnet synchronous motor | |
| CN107482977B (en) | A kind of permanent-magnet synchronous motor rotor position and Rotating speed measring method | |
| CN101938246B (en) | Fuzzy fusion identification method of rotating speed of sensorless motor | |
| CN110829904B (en) | Grey wolf optimization-based parameter optimization method for brushless direct current motor controller | |
| CN103940392B (en) | The rotor-position of a kind of magnetic suspension switched reluctance motor/displacement self-sensing method | |
| CN103116281B (en) | Axial mixed magnetic bearing MFA control system and control method thereof | |
| CN105680746A (en) | Method for designing current of permanent-magnet synchronous motor and parameter of speed controller PI by using online particle swarm optimization algorithm | |
| CN108390597A (en) | Permanent magnet synchronous motor nonlinear predictive controller design with disturbance observer | |
| CN109742999B (en) | Dynamic neural network adaptive inverse SRM torque control method and system | |
| CN109412488A (en) | A kind of permanent magnet synchronous motor dynamic matrix control method | |
| CN112564557A (en) | Control method, device and equipment of permanent magnet synchronous motor and storage medium | |
| CN110829934B (en) | Permanent magnet alternating current servo intelligent control system based on definite learning and mode control | |
| CN111342729B (en) | Self-adaptive reverse thrust control method of permanent magnet synchronous motor based on gray wolf optimization | |
| Chaouch et al. | Optimized torque control via backstepping using genetic algorithm of induction motor | |
| Bayoumi | Stator resistance estimator for direct torque control of permanent magnet synchronous motor drive systems using multi-resolution analysis wavelet | |
| Messaoudi et al. | MRAS and Luenberger Observer Based Sensorless Indirect | |
| Jin-Feng et al. | Control of switched reluctance motors based on improved BP neural networks | |
| CN108551286B (en) | AC servo motor field efficiency detection method and system | |
| CN104539200B (en) | The rotational speed governor of motor, the method for controlling number of revolution of motor | |
| CN111605406B (en) | Control method and control device for obtaining temperature of motor rotor and vehicle | |
| CN110601622B (en) | Method and device for detecting position of motor rotor and computer storage medium | |
| Huangfu et al. | Permanent magnet synchronous motor fault detection and isolation using second order sliding mode observer | |
| Comanescu et al. | A sliding mode adaptive MRAS speed estimator for induction motors | |
| Quang et al. | A sensorless approach for tracking control problem of tubular linear synchronous motor | |
| Comanescu et al. | Full order EMF observer for PMSM—design, analysis and performance under improper speed signal |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230731 Address after: 518000 building 7, fashion brand industrial park, EBU Town, Shenzhen Shantou Special Cooperation Zone, Guangdong Province Patentee after: SHENZHEN SAMKOON TECHNOLOGY Corp.,Ltd. Address before: 212013 No. 301, Xuefu Road, Zhenjiang, Jiangsu Patentee before: JIANGSU University |