阅读说明：本技术 一种基于tnpc逆变器的永磁同步电机无传感器控制方法 (A kind of permanent magnet synchronous motor sensorless strategy method based on TNPC inverter ) 是由 李文娟 李玖云 马亮亮 于 2019-08-29 设计创作，主要内容包括：一种基于TNPC逆变器的永磁同步电机无传感器控制方法属于高精度电机控制领域：传统的电机控制系统对电机控制精度较低,传感器增大电机转动惯量影响电机性能、两电平逆变器输出电压精度较低等缺点,不适合高精度工业领域。本发明在双闭环矢量控制系统基础上进行了改进,采用三电平的TNPC型逆变器代替传统两电平逆变器,减小了输出电流谐波、增大了输出功率,提高了对电机控制的精度,扩展卡尔曼滤波观测器取代了传统位置传感器,减少了电机的转动惯量和摩擦,提高了电机运行的稳定性,同时也提高了对转速的估计精度,增强了系统的鲁棒性。(A kind of permanent magnet synchronous motor sensorless strategy method based on TNPC inverter belongs to high-precision motor control field: traditional electric machine control system is lower to motor control precision, sensor increases the disadvantages of motor rotary inertia influences motor performance, two-level inverter output voltage precision is lower, is not suitable for high-precision industrial circle.The present invention is improved on the basis of two close cycles vector control system, traditional two-level inverter is replaced using the TNPC type inverter of three level, it reduces output current harmonics, increase output power, improve the precision to motor control, Extended Kalman filter observer reduces rotary inertia and the friction of motor, improves the stability of motor operation instead of conventional position sensor, the estimated accuracy to revolving speed is also improved simultaneously, enhances the robustness of system.)

1. a kind of permanent magnet synchronous motor sensorless strategy method based on TNPC type inverter, which is characterized in that including direct current Power module (1), TNPC inverter (2), drive circuit module (6), permanent magnet synchronous motor (3), main control module (5), sensor Module (4), wherein sensor module include voltage detection module, current detection module, main control module include PARK coordinate transform, CLARK coordinate transform, PARK coordinate inverse transformation, speed control, current controller, space voltage vector screening, and extension Kalman Filter observer.

2. a kind of permanent magnet synchronous motor sensorless strategy method based on TNPC inverter, it is characterised in that: including following step It is rapid:

(1) permanent magnet synchronous motor input current is detected by current detection module；

(2) electric current that will test is sent into Kalman filter module, and it is same to observe permanent magnetism by Extended Kalman filter observer Walk motor speedω；

(3) revolving speed that will be observedωWith given rotating speedω ^* It is compared, the difference after comparing is sent into speed control and obtains q axis The reference value of electric currenti ^*, willi ^*Compared with the practical q shaft current of permanent magnet synchronous motor, difference obtains q axis through overcurrent controller and joins Examine voltage；

(4) the d shaft current that will test compared with 0, be sent into current controller and obtain q axis reference voltage by the difference after comparing；

(5) d, q axis reference voltage are obtained into voltage under α, β coordinate system through coordinate transform, voltage under obtained α, β coordinate system is sent Enter space voltage vector screening module, two optimal vectors are chosen in 27 kinds of voltage vectors of TNPC inverter and are used to synthesize Required voltage vector, the voltage reference value for exporting TNPC inverter realize the control to permanent magnet synchronous motor；

(6) enter next sampling instant, repeat step (1)~(4).

3. a kind of permanent magnet synchronous motor sensorless strategy method based on TNPC inverter according to claim 1, special Sign is: TNPC type converter main circuit described in step (3) is made of 12 IGBT interconnections altogether, A, B, C three-phase electricity The IGBT that road is mutually arranged in T-type by four respectively is formed, wherein T_a1Collector and derided capacitors C₁Anode is connected, T_a1Transmitting Pole and T_a3Collector be connected, T_a3Emitter and T_a4Emitter is connected, T_a4Collector and derided capacitors C₁Cathode and derided capacitors C₂Anode is connected, T_a2Collector and T_a1Emitter and T_a3Collector is connected, T_a2Emitter and derided capacitors C₂Cathode is connected, B, C Two-phase IGBT connection is identical as A phase, and A, B, C three-phase IGBT constitute TNPC type converter main circuit.

4. a kind of permanent magnet synchronous motor sensorless strategy method based on TNPC inverter according to claim 1, special Sign is: Extended Kalman filter observer as described in step (1) observes revolving speed specifically:

State vector is predicted, i.e., by inputtingu(k) and last time optimal state estimationTo predict future time point Value,

Wherein, T _sFor the sampling period, " ^ " indicates optimal State Estimation, and " ~ " indicates prediction Value；

The corresponding output of this premeasuring is calculated,

Error co-variance matrix is calculated,

Wherein,F(k) it is Jacobian matrix,

The gain matrix of expanded Kalman filtration algorithm is calculated,

To predicted state vectorFeedback compensation is carried out, the state estimation of optimization is obtained,

Evaluated error covariance matrix,

Into next sampling instant, above procedure is repeated, until estimation motor speed reaches given value.

A kind of permanent magnet synchronous motor sensorless strategy method based on TNPC inverter according to claim 1, feature It is: the screening of space voltage vector described in step (5) specifically:

The every bridge arm of 5.TNPC inverter exports 3 kinds of voltages, and three-phase bridge arm exports 27 kinds of voltage vectors altogether, and 27 kinds of space voltages are sweared Amount is divided into 6 big sectors, and each big sector is divided into 6 small sectors, and there are 3 kinds of basic voltage vectors in each small sector, sentence first Which big sector the given reference voltage vector that breaks belongs to, and then further judgement belongs to which small sector under the big sector, really The basic voltage vectors for determining synthesized reference voltage calculate the action time of each basic vector by voltage-second balance principle, volt Second equilibrium principle formula is as follows:

Wherein,V _refFor reference voltage value, V ₁、V ₂、V ₃For basic voltage vectors,T _sFor the sampling period,T ₁、T ₂、T ₃For fundamental voltage Vector action time

It switch state and the switch time that resulting basic voltage vectors and its action time are converted to every phase bridge arm, will switch State and switch time give inverter in the form of PWM wave, make inverter output reference voltage.

Technical field

The invention belongs to high-precision permanent magnet synchronous motor control field more particularly to a kind of permanent magnetism based on TNPC inverter Synchronous motor sensorless strategy method.

Background technique

The research of high-performance electric machine control system is to promote the key technology of industrial level.With the hair of power electronic technique Exhibition, Alternating Current Governor System have been increasingly becoming mainstream speed-regulating system, and DC speed regulation acquisition is even surmounted in many fields Wider industrial application.This is because comparing direct current generator, alternating current generator maintenance cost is low, and torque mass ratio is big, reliably Property is higher.

Permanent magnet synchronous motor have high power density, superior dynamic response performance, due to rotor use permanent magnet excitation, So that Permanent-magnet Synchronous-motor Speed Servo System structure is more much simpler than asynchronous induction machine, permanent magnet material is often closeer than metal material Spend small, this makes the rotary inertia of permanent magnet synchronous motor lower, therefore permanent magnet synchronous motor is especially suitable in servo-system, especially It is the fields such as the space flight for having strict demand to volume and quality or electric car, and permanent magnet synchronous motor has obtained widely answering With nowadays under the tide of industry 4.0, high-precision industrial circle such as industrial robot, mechanical arm, Novel numerical control machine are also The ample scope for abilities that PMSM servo-system is well worth doing, it may be said that it is China's scientific research instantly and industrial circle that promptly opportunity, which immerses oneself in development, Theme.And the development of above-mentioned industrial circle all be unable to do without the development of permanent magnet synchronous electric motor servo-controlled system.How further The control precision of control system for permanent-magnet synchronous motor is improved, steady-state error is reduced, promotes dynamic property and speed adjusting performance, further Lifting system efficiency and reliability, reducing cost has been various countries Electrified Transmission field, automation field, field of power electronics The project that expert and scholar bring into schedule.

Summary of the invention

The shortcomings that the present invention overcomes traditional high-precision control system for permanent-magnet synchronous motor, it is inverse using the TNPC type of three level Become device and replace traditional two-level inverter, reduces output current harmonics, increases output power, improve to motor control Precision, Extended Kalman filter observer (EKF) reduce the rotary inertia of motor and rub instead of conventional position sensor It wipes, improves the stability of motor operation, while also improving the estimated accuracy to revolving speed, enhance the robustness of system.

A kind of technical solution of the present invention: permanent magnet synchronous motor sensorless strategy side based on TNPC type inverter Method, comprising the following steps:

Step 1, the three-phase current i by the permanent magnet synchronous motor measured_a、i_b、i_cWith input voltage u_a、u_b、u_c, through CLARK Transformation obtains electric current i under two-phase stationary coordinate system_α、i_βWith input voltage u_α、u_β, then by i_α、i_βAnd u_α、u_βIt is sent to extension karr In graceful filtering observer；

Step 2, according to given voltage u_α、u_βWith electric current i_α、i_βMotor is estimated at this time by expanded Kalman filtration algorithm Carve rotational speed omega_e；

Step 3, the rotational speed omega for obtaining estimation_eWith given rotating speedPass through speed control compared to the difference obtained more afterwards Q current phasor reference value is obtained, q current phasor reference value is compared with the q axis actual current measured, obtained difference passes through again Overcurrent controller 1 obtains q shaft voltage vector reference value；

Step 4 obtains d shaft current after being coordinately transformed the three-phase current measured, compared with given current value 0, Obtained difference obtains d axis reference voltage value through overcurrent controller；

D axis, q axis reference voltage value are sent into space voltage vector screening module and are screened by step 5, inverse from TNPC 27 kinds of output voltage vectors for becoming device are chosen optimal voltage vector output and are controlled permanent magnet synchronous motor；

Step 6 repeats step 1-5, until permanent magnet synchronous motor revolving speed reaches given value.

Further, expanded Kalman filtration algorithm in step 2 specifically:

State vector is predicted,

Output is calculated,

Error co-variance matrix is calculated,

Wherein, F (k) is Jacobian matrix,

The gain matrix of expanded Kalman filtration algorithm is calculated,

To predicted state vectorFeedback compensation is carried out, the state estimation of optimization is obtained

Evaluated error covariance matrix,

Further, coordinate transform formula is in step 1

Further, coordinate transform formula is in step 2

Detailed description of the invention

Fig. 1 is one figure of structure of the invention

Fig. 2 is the bright control principle drawing of this law

Fig. 3 is expanded Kalman filtration algorithm structural block diagram

Fig. 4 is the voltage vector-diagram that TNPC inverter generates

Fig. 5 is TNPC inverter principle figure

Specific embodiment

The present invention is according to permanent magnet synchronous motor speed and current double closed loop vector control strategy, in conjunction with TNPC inverter and extension Kalman filtering algorithm realizes the high-precision control of permanent magnet synchronous motor.Estimated by given tachometer value with Kalman filter The actual speed value of calculation compares, and obtained difference passes through after rotational speed governor compared with the q axis actual current value measured To electric current q shaft current reference value, actually measured d shaft current value and 0 is compared to obtain d shaft current reference value, gained d, q axis Current reference value obtains reference voltage through overcurrent controller, and PWM wave control TNPC inverter output is exported after SVPWM is modulated Required voltage.TNPC multi-electrical level inverter exports 27 kinds of voltage vectors, improves voltage output precision, and Extended Kalman filter is calculated Method realizes permanent magnet synchronous motor sensorless strategy, simplifies control system, improves the control precision of permanent magnet synchronous motor.

Specific embodiments of the present invention are illustrated with reference to the accompanying drawing.

As shown in Fig. 1, a kind of permanent magnet synchronous motor sensorless strategy method based on TNPC inverter, including one kind Permanent magnet synchronous motor sensorless strategy method based on TNPC type inverter, which is characterized in that including DC power supplier (1), TNPC inverter (2), drive circuit module (6), permanent magnet synchronous motor (3), main control module (5), sensor module (4), It includes that PARK coordinate transform, CLACK are sat that wherein sensor module, which includes voltage detection module, current detection module, main control module, Mark transformation, the screening of PARK coordinate inverse transformation, speed control, current controller, space voltage vector and spreading kalman filter Wave observer；

The sensor module includes voltage detection module, the equal space wave module CHV-25P Hall voltage of current detection module Sensor, the TNPC inverters IGBT switching tube select the IGBT of model CM200DY-34A, told driving circuit The chip used IR2110 driving chip for the production of U.S. IR company of module, the governor circuit use chip for DSP, model The TMS320F28335 of TI company production.

Working principle

As shown in Fig. 2, Hall sensor detects permanent magnet synchronous motor three-phase current, the electric current warp that will test Electric current under α, β coordinate system is obtained after CLARK transformation, estimates practical turn of permanent magnet synchronous motor by expanded Kalman filtration algorithm Speed, compared with given rotating speed relatively after, difference is obtained to current reference value after speed control and is obtained through overcurrent controller Reference voltage value controls inverter output voltage, controls motor.

Permanent magnet synchronous motor mathematical model under two-phase stationary coordinate system is,

Wherein, R is stator resistance, and L is stator inductance, ω_eFor rotor angular rate, ψ_fFor rotor flux, θ_eFor rotor electricity Angle

Above formula is transformed to current equation, can be obtained

In view of relational expression shown in following formula,

Following state equation can be obtained,

Y=Cx

Wherein, x is state variable, and u is input quantity, and y is output quantity,

The discrete mathematical model of state equation is,

X (k+1)=f [x (k)]+B (k) u (k)+V (k)

Y (k)=C (k) x (k)+W (k)

Wherein, V (k) is system noise, and W (k) is measurement noise, and V (k) and W (k) are zero-mean white noise, is met Relationship shown in following formula,

E { V (k) }=0, E { W (k) }=0

As shown in Fig. 3, extended Kalman filter principle is embodied as,

Using the discrete mathematical model of above-mentioned permanent magnet synchronous motor state equation, state vector is predicted,

Output is calculated,

Error co-variance matrix is calculated,

Wherein, F (k) is Jacobian matrix,

The gain matrix of expanded Kalman filtration algorithm is calculated,

To predicted state vectorFeedback compensation is carried out, the state estimation of optimization is obtained

Evaluated error covariance matrix,

TNPC type three-level inverter working principle is as shown in Fig. 5, and the TNPC type converter main circuit is altogether by 12 A IGBT, which is interconnected, to be constituted, and the IGBT that A, B, C three-phase circuit are mutually arranged in T-type by four respectively is formed, wherein T_a1Current collection Pole and derided capacitors C₁Anode is connected, T_a1Emitter and T_a3Collector be connected, T_a3Emitter and T_a4Emitter is connected, T_a4Collection Electrode and derided capacitors C₁Cathode and derided capacitors C₂Anode is connected, T_a2Collector and T_a1Emitter and T_a3Collector is connected, T_a2 Emitter and derided capacitors C₂Cathode is connected, and the IGBT connection of B, C two-phase is identical as A phase, and it is inverse that A, B, C three-phase IGBT constitute TNPC type Become device main circuit, wherein derided capacitors C₁=C₂, resistance R_a=R_b=R_c, reactor L_a=L_b=L_c。

Further, the TNPC type three-level inverter working principle, working condition I (u_a> 0, i_a< 0) it is opened for power Close pipe T_a2Normal off, T_a4Normal open, T_a1、T_a3Alternate conduction makees PWM switching.Work as T_a1Conducting, T_a3When shutdown, electric current i_aFlow through T_a1It is anti- Parallel diode flows to capacitor C₁Anode, u_aThe U being positive at this time_d/2.Work as T_a1Shutdown, T_a3When conducting, electric current i_aFlow through T_a3With T_a4Anti-paralleled diode, flow to neutral point, at this time u_aIt is at this time 0.

Further, the TNPC type three-level inverter working principle, working condition II (u_a> 0, i_aIt > 0) is power Switch transistor T_a2Normal off, T_a4Normal open, T_a1、T_a3Alternate conduction makees PWM switching.Work as T_a1Conducting, T_a3When shutdown, electric current i_aFlow through T_a1, Flow direction link reactor, u_aThe U being positive at this time_d/2.Work as T_a1Shutdown, T_a3When conducting, electric current i_aFlow through T_a4And T_a3Inverse parallel two Pole pipe, flow direction link reactor, at this time u_aIt is at this time 0.

Further, the TNPC type three-level inverter working principle, working condition III (u_a< 0, i_aIt > 0) is power Switch transistor T_a1Normal off, T_a3Normal open, T_a2、T_a4Alternate conduction makees PWM switching.Work as T_a2Conducting, T_a4When shutdown, electric current i_aFlow through T_a2's Anti-paralleled diode, flow direction link reactor, u_aThe U being negative at this time_d/2.Work as T_a2Shutdown, T_a4When conducting, electric current i_aFlow through T_a4With T_a3Anti-paralleled diode, flow direction link reactor, u_aIt is at this time 0.

Further, the TNPC type three-level inverter working principle, working condition IV (u_a< 0, i_aIt < 0) is power Switch transistor T_a1Normal off, T_a3Normal open, T_a2、T_a4Alternate conduction makees PWM switching.Work as T_a2Conducting, T_a4When shutdown, electric current i_aFlow through T_a2, Flow to capacitor C₂Cathode, u_aThe U being negative at this time_d/2.Work as T_a2Shutdown, T_a4When conducting, electric current i_aFlow through T_a3And T_a4Inverse parallel Diode flows to neutral point, at this time u_aIt is 0.

Further, the TNPC type three-level inverter working principle, there are three types of output states altogether for every phase bridge arm, use Switch function indicates output state are as follows:

Wherein, S_a、S_b、S_c1,0, -1 is respectively represented, tri- level points of P, O, N are corresponded to, it can according to different switch states To be combined into 27 voltage output situations, i.e. 27 kinds of basic voltage vectors.

Definition three-phase circuit output voltage vector is U₀,

Different voltage vectors is can be obtained into the substitution of not switch state of the same race, 27 switch states, which are brought into, can obtain pair The output basic voltage vectors answered, attached drawing 4 are that TNPC inverter exports basic voltage vectors figure.

As shown in Fig. 4,27 kinds of space voltage vectors are divided into 6 big sectors, each big sector is divided into 6 small fans There are 3 kinds of basic voltage vectors in area, each small sector, first determine whether which big sector given reference voltage vector belongs to, then Further judgement belongs to which small sector under the big sector, determines the basic voltage vectors of synthesized reference voltage, flat by weber Weighing apparatus principle calculates the action time of each basic vector, and voltage-second balance principle formula is as follows:

Wherein, V_refFor reference voltage value, V₁、V₂、V₃For basic voltage vectors, T_sFor sampling period, T₁、T₂、T₃It is basic Voltage vector action time

Resulting basic voltage vectors and its action time are converted to switch state and the switch time of every phase bridge arm, it will Switch state and switch time give inverter in the form of PWM wave, make inverter output reference voltage.