阅读说明：本技术 一种基于大数据模型的按转子磁场定向矢量控制的方法 (Method for controlling vector according to rotor magnetic field orientation based on big data model ) 是由 潘安远 凌文锋 张今朝 于 2019-08-08 设计创作，主要内容包括：本发明公开一种基于大数据模型的按转子磁场定向矢量控制方法,基于电机参数的大数据历史值,建立离线转子磁链模型和空间角度模型,并根据新数据采集,多参数大数据更新,更新局部模型参数,得到新的转子磁链模型和空间角度模型,按照电机运行工况,向电机系统提供空间角度和转子磁链反馈量,实现多电机系统的控制；步骤包括多参考数据采集阶段、大数据处理阶段和局部调整阶段。本发明无需电机本体参数则可给出精确反馈值的大数据模型方法,最终实现对多电机系统的控制。(The invention discloses a control method of a vector oriented according to a rotor magnetic field based on a big data model, which is characterized in that an off-line rotor flux linkage model and a space angle model are established based on big data historical values of motor parameters, and according to new data acquisition, multi-parameter big data updating and local model parameter updating, new rotor flux linkage model and space angle model are obtained, and according to the motor operation condition, a space angle and rotor flux linkage feedback quantity are provided for a motor system, so that the control of a multi-motor system is realized; the method comprises a multi-reference data acquisition stage, a big data processing stage and a local adjustment stage. The method can provide a big data model method of accurate feedback value without motor body parameters, and finally realize the control of a multi-motor system.)

1. A control method of vector orientation according to rotor magnetic field based on big data model is characterized in that: establishing an offline rotor flux linkage model and a space angle model based on a big data historical value of a motor parameter, updating a local model parameter according to new data acquisition and multi-parameter big data, obtaining a new rotor flux linkage model and a new space angle model, and providing a space angle and rotor flux linkage feedback quantity for a motor system according to the operation condition of the motor to realize the control of the multi-motor system; the method comprises a multi-reference data acquisition stage, a big data processing stage and a local adjustment stage;

The multi-reference data acquisition stage is controlled according to a rotor magnetic field orientation method, a motor runs in a room without other mechanical or noise sources, synchronous acquisition is carried out under different working conditions, the sampling frequency is 3-5 times of the maximum frequency parameter of a detected object, a data sample sequence is obtained, and the data sample is subjected to noise elimination and normalization processing by a traditional method;

In the big data processing stage, an offline rotor flux linkage model and a spatial angle model are established by adopting a fuzzy clustering and local model network method according to a data sample sequence obtained in the multi-reference data acquisition stage;

The magnetic flux linkage model is used for modeling the magnetic flux linkage,

Wherein n is₁＝n₂＝n₃＝n₄Taking 3 as the regression vector order; n is the number of submodels, the number of submodels is 6, [ kappa ]₁ κ₂ L κ₆]For local model parameters, the local model parameters are implemented by a least squares algorithm,Is each sub-model;

A magnetic linkage space angle model is obtained,

The flux linkage model and the space angle model are weighted and summed by 6 submodels to form a total model;

in the local adjustment stage, under a certain working condition, the load, the rotating speed and the feedback current of the motor stator are uploaded to a main controller, corresponding rotor flux linkage and flux linkage space angle data are obtained through table look-up, and are fed back to a forward channel to complete control; and the big data processing center performs online adjustment on the offline model according to the continuously updated numerical value, and finally realizes the control of the multi-motor system.

2. The big data model-based vector control method oriented according to the rotor magnetic field according to claim 1, wherein the motor parameters are three-phase stator current, voltage, direct-current bus current, rotation speed and torque.

Technical Field

the invention relates to a control method, in particular to a method for controlling a vector oriented by a rotor magnetic field based on a big data model.

Background

for three-phase servo asynchronous motor, as the main power of drive in industrial production processing equipment, the control mode often adopts the typical method of directional control according to rotor magnetic field, and the basic idea is that firstlyconverting the three-phase current of the stator by 3 and 2, converting the coordinates of alpha beta, dq and MT, and superposing the coordinate M in the coordinate MT and the flux linkage of the rotor, so that the three-phase current of the stator is finally decomposed into i_st、i_smcomponent i_stComponent of the control torque exclusively, i_smOnly the flux linkage size is controlled because i_st、i_smThe two components are in an orthogonal relation, so that complete decoupling is achieved, and the motor driving system obtains better control performance. However, in the control method, the flux linkage psi is obtained through a rotor flux linkage calculation model during feedback_rAnd angle of space thereofThe current calculation of the flux linkage and the space angle is influenced by the motor parameters, two feedback parameters are calculated at the same time, the calculation amount is large, and when the inductance, the resistance and the operation temperature of the motor change, the calculation results of the flux linkage and the space direction angle possibly have large errors, so that the control effect of the motor is influenced.

disclosure of Invention

Based on the problems, the invention provides a rotor magnetic field oriented vector control method based on a big data model, which adopts multi-parameter big data historical values to establish an offline model of flux linkage and a space angle, updates local model parameters according to new data acquisition to obtain a new flux linkage and space angle model, establishes a big data model method capable of giving an accurate feedback value without motor body parameters, and finally realizes the control of a multi-motor system.

the adopted technical scheme is as follows: a control method based on a big data model and oriented vectors of a rotor magnetic field is characterized in that an off-line rotor flux linkage model and a space angle model are established based on big data historical values of motor parameters, new data collection is carried out, multi-parameter big data are updated, local model parameters are updated, a new rotor flux linkage model and a new space angle model are obtained, and space angles and rotor flux linkage feedback quantities are provided for a motor system according to the operation condition of the motor, so that the control of a multi-motor system is realized; the method comprises a multi-reference data acquisition stage, a big data processing stage and a local adjustment stage;

The multi-reference data acquisition stage is controlled by a rotor magnetic field orientation method, the motor operates in a room without other mechanical or noise sources, synchronous acquisition is carried out under different working conditions, the sampling frequency is 3-5 times of the maximum frequency parameter of a detected object, a data sample sequence is obtained, and the data sample is subjected to noise elimination and normalization processing by a traditional method;

A big data processing stage, wherein an offline rotor flux linkage model and a spatial angle model are established by adopting a fuzzy clustering and local model network method according to a data sample sequence obtained in the multi-reference data acquisition stage;

The magnetic flux linkage model is used for modeling the magnetic flux linkage,

Wherein n is₁＝n₂＝n₃＝n₄Taking 3 as the regression vector order; n is the number of submodels, the number of submodels is 6, [ kappa ]₁ κ₂L κ₆]For local model parameters, the local model parameters are implemented by a least squares algorithm,is each sub-model;

A magnetic linkage space angle model is obtained,

The flux linkage model and the space angle model are weighted and summed by 6 submodels to form a total model;

In the local adjustment stage, under a certain working condition, the load, the rotating speed and the motor stator feedback current are uploaded to a main controller, corresponding rotor flux linkage and flux linkage space angle data are obtained through table look-up, and are fed back to a forward channel to complete control; and the big data processing center performs online adjustment on the offline model according to the continuously updated numerical value, and finally realizes the control of the multi-motor system.

The motor parameters are three-phase stator current, voltage, direct current bus current, rotating speed and torque.

the method is based on the multi-parameter big data historical value, an off-line rotor flux linkage model and a space angle model are established, local model parameters are updated according to new data acquisition and multi-parameter big data updating, a new rotor flux linkage model and a new space angle model are obtained, and motor unit control is achieved.

Drawings

FIG. 1 is a block diagram of the control method of the present invention;

FIG. 2 is a block diagram of a control system according to the rotor field orientation method of the present invention;

FIG. 3 is a data sample sequence diagram in the present invention;

FIG. 4 is a block diagram of flux linkage and space angle obtained by a flux linkage calculation model during motor parameter feedback according to the present invention.

Detailed Description

The present invention will be described in further detail below by way of examples with reference to the accompanying drawings, which are illustrative of the present invention and are not to be construed as limiting the present invention.

referring to fig. 1 to 4, a big data model-based rotor magnetic field oriented vector control method establishes an offline rotor flux linkage model and a space angle model based on big data historical values of motor parameters such as three-phase stator current, voltage, direct current bus current, rotating speed and torque, updates local model parameters according to new data acquisition and multi-parameter big data, obtains a new rotor flux linkage model and a new space angle model, provides a space angle and rotor flux linkage feedback quantity for a motor system according to motor operation conditions, and realizes control of a multi-motor system; the method comprises a multi-reference data acquisition stage, a big data processing stage and a local adjustment stage; wherein, the multi-reference data acquisition stage is controlled by a rotor magnetic field orientation method, the motor runs in a room without other mechanical or noise sources, the synchronous acquisition is carried out under different working conditions, the sampling frequency is 3-5 times of the maximum frequency parameter of the detected object, and a data sample is obtainedthe sequence is to perform denoising and normalization processing on a data sample by a traditional method; as can be analyzed from fig. 2, the motor feedback link is mainly focused on the rotor flux linkage ψ_rComputing (including) In order to save the capacity of the processor to the maximum extent, when the motor runs, the data acquisition is synchronously acquired under different working conditions by combining the graph 2 and according to the graph 4, the data sample sequence is shown in the graph 3, and the acquired array is subjected to noise elimination and normalization processing by a traditional method.

a big data processing stage, wherein an offline rotor flux linkage model and a spatial angle model are established by adopting a fuzzy clustering and local model network method according to a data sample sequence obtained in the multi-reference data acquisition stage; according to the literature [ zhang toni, bang chinese, pantian red, multi-model identification [ J ] of multi-motor synchronous system, motor and control bulletin, 2009, 13 (1): 138- > 142 ] it has been deduced that as a hard coupling the N motor models fit that the jacobian matrix is not equal to 0 and the relative order is equal to the system order, the system is proved to be reversible. The single three-phase asynchronous motor model is also reversible, and a corresponding relation is determined by a multi-input multi-output system as shown in FIG. 4. It is feasible to use a multidimensional data sequence to establish the rotor magnetic system and the coordinate space angle off-line as shown in figure 3.

Based on the fuzzy clustering and local model network method adopted in the literature, the rotor flux linkage model and flux linkage space angle model can be obtained respectively

The magnetic flux linkage model is used for modeling the magnetic flux linkage,

wherein n is₁＝n₂＝n₃＝n₄Taking 3 as the regression vector order; n is the number of submodels, the number of submodels is 6, [ kappa ]₁ κ₂L κ₆]for local model parameters, officemodel identification of a PWA system based on fuzzy clustering [ J ] is realized by using a least square algorithm to realize the parameters of the partial model [ Pantianhong, Lishao-Yuan ]]automated journal 2007,33(3): 327 and 330.]；Is each sub-model;

A magnetic linkage space angle model is obtained,

the flux linkage model and the space angle model are weighted and summed by 6 submodels to form a total model; the regression vector order and the number of the submodels are determined according to the input quantity dimension and the historical value required by the model parameters on one hand, and on the other hand, the requirement that 3 is taken as the regression vector order and 6 is taken as the submodel number for the three-phase asynchronous motor model construction needs to be met in consideration of the storage and processing speed of a processor.

in the local adjustment stage, under a certain working condition, the load, the rotating speed and the motor stator feedback current are uploaded to a main controller, corresponding rotor flux linkage and flux linkage space angle data are obtained through table look-up, and are fed back to a forward channel to complete control; and the big data processing center performs online adjustment on the offline model according to the continuously updated numerical value, and finally realizes the control of the multi-motor system.