CN112511053A - Load inertia identification method based on motion model - Google Patents
Load inertia identification method based on motion model Download PDFInfo
- Publication number
- CN112511053A CN112511053A CN202011165658.XA CN202011165658A CN112511053A CN 112511053 A CN112511053 A CN 112511053A CN 202011165658 A CN202011165658 A CN 202011165658A CN 112511053 A CN112511053 A CN 112511053A
- Authority
- CN
- China
- Prior art keywords
- motor
- torque
- motion model
- load inertia
- actual
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 28
- 230000001133 acceleration Effects 0.000 claims abstract description 50
- 238000004364 calculation method Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 abstract description 5
- 230000002411 adverse Effects 0.000 abstract description 4
- 230000000694 effects Effects 0.000 abstract description 4
- 238000001914 filtration Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
- H02P21/00—Arrangements or methods for the control of electric machines by vector control, e.g. by control of field orientation
- H02P21/0003—Control strategies in general, e.g. linear type, e.g. P, PI, PID, using robust control
- H02P21/0017—Model reference adaptation, e.g. MRAS or MRAC, useful for control or parameter estimation
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
- H02P21/00—Arrangements or methods for the control of electric machines by vector control, e.g. by control of field orientation
- H02P21/14—Estimation or adaptation of machine parameters, e.g. flux, current or voltage
- H02P21/143—Inertia or moment of inertia estimation
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
- H02P21/00—Arrangements or methods for the control of electric machines by vector control, e.g. by control of field orientation
- H02P21/14—Estimation or adaptation of machine parameters, e.g. flux, current or voltage
- H02P21/20—Estimation of torque
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Control Of Electric Motors In General (AREA)
Abstract
The invention relates to the technical field of alternating current servo systems, in particular to a load inertia identification method based on a motion model. Through continuous iteration correction, the load inertia can be finally and accurately calculated, and the more the iteration times, the more accurate the load inertia. The acceleration of the motor does not need to be collected in the whole process, and the adverse effect of acceleration collection errors on load inertia identification is effectively avoided. In the invention, whether offline inertia learning or online inertia learning is adopted, the load inertia can be accurately identified by adjusting the iteration period.
Description
Technical Field
The invention relates to the technical field of alternating current servo systems, in particular to a load inertia identification method based on a motion model.
Background
The servo motor is used as a driving execution mechanism, and has the advantages of high response speed, high control precision and small volume. The general servo motor in the market is a permanent magnet synchronous motor, and the control mode adopts a digital vector control mode. Since the servo motor can be operated in a torque control mode, a speed control mode or a position control mode. Therefore, a three-loop control structure with nested position loops, speed loops and torque loops is generally adopted in control. Wherein the performance of the speed loop is greatly affected by the load inertia. If the servo motor is operated in a position control mode or a speed control mode, the load inertia parameter needs to be set correctly. As a general servo product, the load inertia of different devices is different and is unknown in most occasions. Therefore, the equipment provided with the servo motor needs a long time to calculate or measure the load inertia before trial operation.
The current servo load inertia learning methods mainly include the following two methods: (1) and a direct calculation method is adopted, so that the motor carries out acceleration and deceleration movement, and the load inertia is directly obtained by dividing the torque output by the motor by the actual acceleration. (2) And the least square method is used for enabling the motor to perform variable acceleration movement, acquiring a plurality of different torque values and a plurality of different acceleration values, and calculating the optimal inertia through the sum of squares of minimized errors.
In the above manner, the direct calculation method needs a higher acceleration to calculate an accurate inertia, the calculated inertia is greatly affected by load disturbance, different inertias may be calculated at different accelerations, and convergence is not easy. The least square method is complex in calculation, accurate in estimation only by different accelerated speeds, and low in practicability. No matter the direct calculation method or the least square method is adopted, the acceleration of the motor needs to be obtained, and when the acceleration is small, the inertia error obtained through calculation is extremely large.
Disclosure of Invention
The invention provides a load inertia identification method based on a motion model aiming at the problems in the prior art, wherein the virtual motor motion model is set, and the load inertia can be accurately self-learned by combining an iterative algorithm.
In order to solve the technical problems, the invention adopts the following technical scheme: a load inertia identification method based on a motion model comprises the following steps:
A. establishing a virtual motor motion model in the servo motor;
B. in a motion iteration period, carrying out algorithm iteration comparison on the output motor torque of the virtual motor motion model and the actual output torque to obtain the load inertia of the next iteration period;
C. and D, repeating the step B, and obtaining the load inertia of the servo motor after enough iteration times.
Preferably, the comparing step of the output motor torque of the virtual motor motion model in step B with the actual output torque includes:
B1. calculating the rotating speed v of the virtual motor motion modelfbAt the same time, the actual motor speed v is obtainedref；
B2. Calculating the rotating speed v of the virtual motor motion modelfbWith the actual motor speed vrefError between Tmpre；
B3. According to the error TmpreCalculating motor torque T of virtual motor motion modelm；
B4. Let the actual motor output torque be TrAccording to the actual motor output torque TrCalculating the actual acceleration torque Tacc；
B5. Load torque TlEqual to the actual motor output torque TrMinus the actual acceleration torque TaccThen the model acceleration torque T of the virtual motor motion model can be calculatedmaccMotor torque T equal to virtual motor motion modelmMinus the load torque Tl；
B6. Actual acceleration torque T to be obtainedaccAnd model acceleration torque TmaccAnd carrying out algorithm iteration so as to obtain the required load inertia.
Preferably, the rotation speed v of the virtual motor motion modelfbThe calculation method comprises the following steps:
firstly, setting initial load inertia and calculating the reciprocal of the initial load inertia to be Kj(k)；
Will Kj(k) Model acceleration torque T from the previous cyclemaccMultiplying and then calculating an integral to obtain the rotating speed v of the model motorfb。
Preferably, said error TmpreThe calculation method comprises the following steps:
setting a PI regulator in the virtual motor motion model to obtain the actual motor rotation speed vrefAnd the model motor speed vfbWhen the error is inputted into the PI regulator, the error can be calculated
Wherein, KpAnd KiProportional gain and integral gain, respectively, and T is discrete time.
Preferably, a low-pass filter is arranged in the virtual motor motion model, and the obtained error T ismpreAfter passing through a low-pass filter, model motor torque can be obtained
Tm(k+1)＝aTm(k)+(1-a)Tmpre(k)，
Where a is a low pass filter parameter.
The low-pass filter parameter a is identical to the actual torque filter parameter.
Preferably, a high-pass filter is arranged in the virtual motor motion model, and the actual motor output torque is TrAfter passing through a high-pass filter, the actual acceleration torque can be obtained
Tacc(k+1)＝Tr(k)-Tr(k-1)+Tacc(k)b，
Wherein b is a high-pass filter parameter, and the size of the high-pass filter parameter b depends on the load type of the motor.
Preferably, the actual acceleration torque T is measured in an iteration cycleaccAnd model acceleration torque TmaccRespectively summing to obtain sigma TaccSum Σ TmaccThen, every other iteration cycle, the load inertia of the next iteration can be updated as follows:
the invention has the beneficial effects that:
the load inertia identification method based on the motion model provided by the invention has the advantages that the virtual motor motion model is established in the servo, and the motor torque output by the virtual motor model is compared with the actual output torque, so that the load inertia value is further corrected. Through continuous iteration correction, the load inertia can be finally and accurately calculated, and the more the iteration times, the more accurate the load inertia. The acceleration of the motor does not need to be collected in the whole process, and the adverse effect of acceleration collection errors on load inertia identification is effectively avoided. In the invention, whether offline inertia learning or online inertia learning is adopted, the load inertia can be accurately identified by adjusting the iteration period.
Drawings
FIG. 1 is a flow chart of the steps of the present invention.
Fig. 2 is a schematic structural diagram of a virtual motor motion model according to the present invention.
Detailed Description
In order to facilitate understanding of those skilled in the art, the present invention will be further described with reference to the following examples and drawings, which are not intended to limit the present invention. The present invention is described in detail below with reference to the attached drawings.
The method for identifying load inertia based on a motion model provided by the embodiment, as shown in fig. 1, includes the following steps:
A. establishing a virtual motor motion model in the servo motor;
B. in a motion iteration period, carrying out algorithm iteration comparison on the output motor torque of the virtual motor motion model and the actual output torque to obtain the load inertia of the next iteration period;
C. and D, repeating the step B, and obtaining the load inertia of the servo motor after enough iteration times.
Specifically, a virtual motor motion model is established in the servo, and the motor torque output by the virtual motor model is compared with the actual output torque through an iterative algorithm, so that the load inertia value is further corrected. Through continuous iteration correction, the load inertia can be finally and accurately calculated, and the more the iteration times, the more accurate the load inertia. The acceleration of the motor does not need to be collected in the whole process, and the adverse effect of acceleration collection errors on load inertia identification is effectively avoided. In the invention, whether offline inertia learning or online inertia learning is adopted, the load inertia can be accurately identified by adjusting the iteration period.
In the load inertia identification method based on the motion model provided in this embodiment, the step of comparing the output motor torque of the virtual motor motion model with the actual output torque includes:
B1. calculating model motor rotating speed v of virtual motor motion modelfbAt the same time, the actual motor speed v is obtainedref；
B2. Calculating model motor rotation speed vfbWith the actual motor speed vrefError between Tmpre；
B3. According to the error TmpreCalculating motor torque T of virtual motor motion modelm；
B4. Let the actual motor output torque be TrAccording to the actual motor output torque TrCalculating the actual acceleration torque Tacc；
B5. Load torque TlEqual to the actual motor output torque TrMinus the actual acceleration torque TaccThen the model acceleration torque T of the virtual motor motion model can be calculatedmaccMotor torque T equal to virtual motor motion modelmMinus the load torque Tl；
B6. Actual acceleration torque T to be obtainedaccAnd model acceleration torque TmaccAnd carrying out algorithm iteration so as to obtain the required load inertia.
As shown in fig. 2, the virtual motor motion model of this embodiment includes a PI regulator, a low-pass filter, and a high-pass filter. The calculation process of this embodiment is: firstly, it is necessary to set an initial load inertia and calculate the reciprocal thereof as Kj(k) Then, K is addedj(k) Model acceleration torque T from the previous cyclemaccAfter multiplication, integral is calculated, so that the rotating speed v of the model motor can be obtainedfb。
Then, the actual motor speed v is obtainedrefAnd the obtained actual motor rotating speed vrefAnd the model motor speed vfbWhen the error is inputted into the PI regulator, the error can be calculated
KpAnd KiProportional gain and integral gain, respectively, and T is discrete time. Then, the error is processed by a low-pass filter to obtain the model motor torque Tm(k+1)＝aTm(k)+(1-a)Tmpre(k) Where a is a low pass filter parameter, the low pass filter parameter a being consistent with the actual torque filter parameter.
Then, it is necessary to start calculating the actual acceleration torque TaccAnd model acceleration torque TmaccThe main parameters of two iterative algorithms are that the actual motor output torque is set as TrWill output the actual motor torque TrAfter passing through the high-pass filter, the actual acceleration torque T can be obtainedacc(k+1)＝Tr(k)-Tr(k-1)+Tacc(k) b, wherein b is a high-pass filtering parameter, and the size of the high-pass filtering parameter b depends on the load type of the motor. Then, the actual motor output torque T is usedrMinus the actual acceleration torque TaccEqual to the load torque TlThen using the motor torque T of the virtual motor motion modelmMinus the load torque TlEqual to model acceleration torque Tmacc. Obtaining the actual acceleration torque TaccAnd model acceleration torque TmaccAfter two parameters, an iterative algorithm can be carried out, and the actual acceleration torque T can be processed in an iterative periodaccAnd model acceleration torque TmaccRespectively summing to obtain sigma TaccSum Σ TmaccThen, every other iteration cycle, the load inertia of the next iteration can be updated as follows:
this exampleIn the above-mentioned method, the model acceleration torque T is obtained by the virtual motor motion modelmaccAcceleration torque T with actual motoraccIterative operation is carried out, load inertia can be accurately learned by self, so that accurate load parameters are provided for servo parameter self-tuning, and debugging of automatic equipment is facilitated. In addition, in the whole self-learning process, the acceleration of the motor does not need to be acquired, the adverse effect of acceleration acquisition errors on load inertia identification is effectively avoided,
although the present invention has been described with reference to the above preferred embodiments, it should be understood that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (8)
1. A load inertia identification method based on a motion model is characterized by comprising the following steps:
A. establishing a virtual motor motion model in the servo motor;
B. in a motion iteration period, carrying out algorithm iteration comparison on the output motor torque of the virtual motor motion model and the output torque of an actual motor to obtain the load inertia of the next iteration period;
C. and D, repeating the step B, and obtaining the load inertia of the servo motor after enough iteration times.
2. The method for identifying load inertia based on motion model as claimed in claim 1, wherein the step of comparing the output motor torque of the virtual motor motion model and the actual output torque of step B comprises:
B1. calculating model motor rotating speed v of virtual motor motion modelfbAt the same time, the actual motor speed v is obtainedref；
B2. Calculating model motor rotation speed vfbWith the actual motor speed vrefError between Tmpre；
B3. According to the error TmpreCalculating motor torque T of virtual motor motion modelm；
B4. Let the actual motor output torque be TrAccording to the actual motor output torque TrCalculating the actual acceleration torque Tacc；
B5. Load torque TlEqual to the actual motor output torque TrMinus the actual acceleration torque TaccThen the model acceleration torque T of the virtual motor motion model can be calculatedmaccMotor torque T equal to virtual motor motion modelmMinus the load torque Tl；
B6. Actual acceleration torque T to be obtainedaccAnd model acceleration torque TmaccAnd carrying out algorithm iteration so as to obtain the required load inertia.
3. The method for identifying load inertia based on motion model as claimed in claim 2, wherein the rotating speed v of the motion model of the virtual motorfbThe calculation method comprises the following steps:
firstly, setting initial load inertia and calculating the reciprocal of the initial load inertia to be Kj(k)；
Will Kj(k) Model acceleration torque T from the previous cyclemaccMultiplying and then calculating an integral to obtain the rotating speed v of the model motorfb。
4. The method for identifying load inertia based on motion model as claimed in claim 2, wherein the error T ismpreThe calculation method comprises the following steps:
setting a PI regulator in the virtual motor motion model to obtain the actual motor rotation speed vrefAnd the model motor speed vfbWhen the error is inputted into the PI regulator, the error can be calculated
Wherein, KpAnd KiProportional gain and integral gain, respectively, and T is discrete time.
5. The method for identifying load inertia based on motion model as claimed in claim 2, wherein: setting a low-pass filter in the virtual motor motion model, and obtaining an obtained error TmpreAfter passing through a low-pass filter, model motor torque can be obtained
Tm(k+1)＝aTm(k)+(1-a)Tmpre(k)，
Where a is a low pass filter parameter.
6. The method for identifying load inertia based on motion model as claimed in claim 5, wherein: the low-pass filter parameter a coincides with the actual torque filter parameter.
7. The method for identifying load inertia based on motion model as claimed in claim 2, wherein: setting a high-pass filter in the virtual motor motion model, and setting the actual motor output torque as TrAfter passing through a high-pass filter, the actual acceleration torque can be obtained
Tacc(k+1)＝Tr(k)-Tr(k-1)+Tacc(k)b，
Wherein b is a high-pass filter parameter, and the size of the high-pass filter parameter b depends on the load type of the motor.
8. The method for identifying load inertia based on motion model as claimed in claim 2, wherein: during an iteration period, the actual acceleration torque T is measuredaccAnd model acceleration torque TmaccRespectively summing to obtain sigma TaccSum Σ TmaccThen, every other iteration cycle, the load inertia of the next iteration can be updated as follows:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011165658.XA CN112511053B (en) | 2020-10-27 | 2020-10-27 | Load inertia identification method based on motion model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011165658.XA CN112511053B (en) | 2020-10-27 | 2020-10-27 | Load inertia identification method based on motion model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112511053A true CN112511053A (en) | 2021-03-16 |
| CN112511053B CN112511053B (en) | 2023-03-10 |
Family
ID=74955959
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011165658.XA Active CN112511053B (en) | 2020-10-27 | 2020-10-27 | Load inertia identification method based on motion model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112511053B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6037736A (en) * | 1995-05-17 | 2000-03-14 | Kabushiki Kaisha Yaskawa Denki | Apparatus for determination of control constant |
| JP2008148402A (en) * | 2006-12-07 | 2008-06-26 | Toshiba Mitsubishi-Electric Industrial System Corp | Power conversion device |
| CN101292418A (en) * | 2005-10-18 | 2008-10-22 | 株式会社安川电机 | Motor control device, method and program storage medium |
| CN103338003A (en) * | 2013-06-28 | 2013-10-02 | 西安交通大学 | Online simultaneous identifying method for load torque and inertia of motor |
| CN103684188A (en) * | 2013-12-31 | 2014-03-26 | 上海英威腾工业技术有限公司 | Method and system for identifying rotational inertia of motion control system |
| CN105915121A (en) * | 2016-05-31 | 2016-08-31 | 西安交通大学 | Servo system inertia identification method adopting genetic algorithm for optimization |
| CN109347391A (en) * | 2018-10-29 | 2019-02-15 | 江苏大学 | A kind of adaptive method for identification of rotational inertia of Landau considering system noise |
-
2020
- 2020-10-27 CN CN202011165658.XA patent/CN112511053B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6037736A (en) * | 1995-05-17 | 2000-03-14 | Kabushiki Kaisha Yaskawa Denki | Apparatus for determination of control constant |
| CN101292418A (en) * | 2005-10-18 | 2008-10-22 | 株式会社安川电机 | Motor control device, method and program storage medium |
| JP2008148402A (en) * | 2006-12-07 | 2008-06-26 | Toshiba Mitsubishi-Electric Industrial System Corp | Power conversion device |
| CN103338003A (en) * | 2013-06-28 | 2013-10-02 | 西安交通大学 | Online simultaneous identifying method for load torque and inertia of motor |
| CN103684188A (en) * | 2013-12-31 | 2014-03-26 | 上海英威腾工业技术有限公司 | Method and system for identifying rotational inertia of motion control system |
| CN105915121A (en) * | 2016-05-31 | 2016-08-31 | 西安交通大学 | Servo system inertia identification method adopting genetic algorithm for optimization |
| CN109347391A (en) * | 2018-10-29 | 2019-02-15 | 江苏大学 | A kind of adaptive method for identification of rotational inertia of Landau considering system noise |
Non-Patent Citations (1)
| Title |
|---|
| 徐勇: "交流伺服系统转动惯量在线辨识及负载转矩观测", 《微电机》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112511053B (en) | 2023-03-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107070315B (en) | Multi-switch reluctance motor rotating speed synchronous control device and control method | |
| CN106487297B (en) | A kind of PMSM parameter identification method based on Unscented kalman filtering algorithm | |
| CN108092567B (en) | Permanent magnet synchronous motor rotating speed control system and method | |
| CN104378038B (en) | Permanent magnet synchronous motor parameter identification method based on artificial neural network | |
| CN111736471A (en) | Iterative feedback setting control and robust optimization method of rotary inverted pendulum | |
| CN110880895B (en) | Method for determining q-axis current compensation value of permanent magnet synchronous motor and method for suppressing pulsation | |
| CN105048896B (en) | A kind of brshless DC motor Direct Torque adaptive fuzzy control method | |
| CN104167973A (en) | Inertia estimating method and inertia estimating appartus of position control apparatus | |
| CN108365787A (en) | A kind of Permanent-magnet Synchronous-motor Speed Servo System and its design method based on internal model control | |
| CN109586645B (en) | Permanent magnet synchronous motor inertia identification method and equipment | |
| CN109491251B (en) | AC servo system model identification method and equipment considering data disturbance compensation | |
| CN109194225A (en) | A kind of double feedback electric engine parameter on-line identification method | |
| CN112564557A (en) | Control method, device and equipment of permanent magnet synchronous motor and storage medium | |
| CN112511053B (en) | Load inertia identification method based on motion model | |
| CN109743002B (en) | Servo system controller, feedforward control signal determination method and inertia identification method | |
| CN112986977B (en) | Method for overcoming radar extended Kalman track filtering divergence | |
| CN117040339B (en) | Vector decoupling method, device and equipment for permanent magnet synchronous motor and storage medium | |
| CN115566954A (en) | Embedded motor speed regulation control compensation method and system | |
| WO2023124921A1 (en) | Torque estimation method and apparatus for permanent magnet synchronous motor, and device and storage medium | |
| CN114301352B (en) | Motor speed measuring method and device and system | |
| Wang et al. | Speed and load torque estimation of SPMSM based on Kalman filter | |
| CN110048653B (en) | Sliding mode variable structure MRAS rotating speed identification method of bearingless asynchronous motor | |
| CN117639596B (en) | High-speed permanent magnet synchronous motor rotating speed calculation method based on FPGA | |
| CN109150026A (en) | A kind of no brush and no position sensing motor commutation method | |
| CN114977920B (en) | High-precision permanent magnet synchronous motor parameter identification method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240112 Address after: 518000 No. 502, building 13, Chuangke Town, University Town, Taoyuan Street, Nanshan District, Shenzhen, Guangdong Patentee after: SHENZHEN VECTOR SCIENCE CO.,LTD. Address before: 523000 a, building 12, Zhongji Zhigu, No.1, Nanshan Road, Songshanhu high tech Industrial Development Zone, Dongguan City, Guangdong Province Patentee before: VECTOR (DONGGUAN) INTELLIGENT CONTROL CO.,LTD. |