阅读说明：本技术 基于电流预测的适用于永磁电机控制的死区补偿方法 (Dead zone compensation method suitable for permanent magnet motor control based on current prediction ) 是由 邹会杰 张涛 张宇龙 张吉斌 张瑞峰 詹哲军 于 2020-06-18 设计创作，主要内容包括：本发明涉及电机控制的死区补偿方法,具体为基于电流预测的适用于永磁电机控制的死区补偿方法。解决现有技术死区补偿效果较差的问题。基于电流预测的适用于永磁电机控制的死区补偿方法,根据电机运行的频率分为低速区死区补偿策略和高速区死区补偿策略；本发明通过预测电流进行死区补偿,解决了由于数字控制器存在一定延迟,本拍计算的结果到下一拍才进行更新,导致过零点附件死区补偿效果较差的问题。(The invention relates to a dead zone compensation method for motor control, in particular to a dead zone compensation method suitable for permanent magnet motor control based on current prediction. The problem of prior art blind spot compensation effect relatively poor is solved. The dead zone compensation method based on current prediction and suitable for permanent magnet motor control is divided into a low-speed zone dead zone compensation strategy and a high-speed zone dead zone compensation strategy according to the running frequency of a motor; the dead zone compensation is carried out through the prediction current, and the problem that due to the fact that a digital controller has certain delay, the calculated result of the beat is updated until the next beat, and the dead zone compensation effect of the zero-crossing point accessory is poor is solved.)

1. A dead zone compensation method suitable for permanent magnet motor control based on current prediction is characterized in that a low-speed zone dead zone compensation strategy and a high-speed zone dead zone compensation strategy are divided according to the running frequency of a motor;

when the motor runs at a low speed stage, firstly, the three-phase current i of the motor is converted into the three-phase current_A、i_BAnd i_CI is obtained through a change formula of converting a three-phase static coordinate system into a two-phase static coordinate system_αAnd i_βThen i is_αAnd i_βThe dq axis current i is obtained through a change formula of converting a two-phase static coordinate system into a two-phase rotating coordinate system_dAnd i_q(ii) a Then calculating the obtained i_dAnd i_qThe predicted αβ axis current i is obtained by a change formula of converting a two-phase rotating coordinate system into a two-phase stationary coordinate system_{α_pre}And i_{β_pre}Where phi used for coordinate transformation passes phi + w_sT_sCalculated, wherein phi is the synchronous rotation angle of the next beat, theta is the synchronous rotation angle of the motor of the current beat, and w_sIs the synchronous angular frequency, T_sIs the sampling interval time; and obtaining the predicted three-phase current i of the motor by a change formula of converting the two-phase static coordinate system into the three-phase static coordinate system_{A_pre}、i_{B_pre}And i_{C_pre}(ii) a I obtained by calculation_{A_pre}、i_{B_pre}And i_{C_pre}Performing dead zone compensation;

when the motor runs at a high speed stage, firstly, the three-phase current i of the motor is converted into the three-phase current_A、i_BAnd i_CI is obtained through a change formula of converting a three-phase static coordinate system into a two-phase static coordinate system_αAnd i_βThen i is_αAnd i_βI is obtained through a change formula of converting a two-phase static coordinate system into a two-phase rotating coordinate_dAnd i_q(ii) a Then calculating the obtained i_dAnd i_qObtaining predicted dq axis current i through stator voltage equation calculation_{d_pre}And i_{q_pre}，i_{d_pre}And i_{q_pre}The predicted αβ axis current i is obtained by a change formula of converting a two-phase rotating coordinate system into a two-phase stationary coordinate system_{α_pre}And i_{β_pre}Where phi used for coordinate transformation passes phi + w_sT_sCalculated, where phi is the next beat synchronous rotation angleDegree, theta is the synchronous rotation angle of the motor of the racket, w_sIs the synchronous angular frequency, T_sIs the sampling interval time; and obtaining the predicted three-phase current i of the motor by a change formula of converting the two-phase static coordinate system into the three-phase static coordinate system_{A_pre}、i_{B_pre}And i_{C_pre}(ii) a I obtained by calculation_{A_pre}、i_{B_pre}And i_{C_pre}Dead zone compensation is performed.

2. The current prediction based dead-zone compensation method for permanent magnet motor control according to claim 1,

1) low-speed zone dead zone compensation strategy

Obtaining a current predicted value through coordinate transformation at a low-speed stage to perform dead zone compensation control, and specifically comprising the following steps:

3/2 transformation formula for transforming motor current from three-phase stationary coordinate system to two-phase stationary coordinate system:

in the formula i_A、i_BAnd i_CRespectively representing three-phase currents of the motor; i.e. i_α、i_βRespectively representing the currents of two static αβ coordinate axes;

2/2 transformation formula for transformation of motor current from two-phase stationary coordinate system to two-phase rotating coordinate:

in the formula i_α、i_βRespectively representing αβ coordinate axis currents i_d、i_qRespectively representing two-phase rotating dq coordinate axis currents; theta is the synchronous rotation angle of the motor;

the synchronous rotation angle of the next beat is changed into phi, phi is obtained through calculation of the synchronous rotation angle theta of the motor of the next beat, and a calculation formula is as follows:

φ＝θ+w_sT_s(3)

in the formula, w_sIs the synchronous angular frequency; t is_sIs the sampling interval time; phi is the next beat synchronous rotation angle;

then the αβ coordinate axis prediction current i is obtained by a change formula of converting a two-phase rotating coordinate system into a two-phase stationary coordinate_{α_pre}And i_{β_pre}The 2/2 transformation formula for transforming the two-phase rotating coordinate system to the two-phase stationary coordinate system is:

in the formula i_{α_pre}、i_{β_pre}Respectively representing αβ coordinate axis predicted current;

finally, obtaining three-phase static coordinate axis prediction motor three-phase current i through a change formula of converting a two-phase static coordinate system into a three-phase static coordinate system_{A_pre}、i_{B_pre}And i_{C_pre}2/3 transformation formula for transformation of motor current from two-phase stationary frame to three-phase stationary frame:

in the formula i_{A_pre}、i_{B_pre}And i_{C_pre}Respectively representing three-phase predicted currents of the motor;

2) high-speed zone dead zone compensation strategy

According to the equivalent circuit of the permanent magnet motor, the flux linkage equation of the motor is as follows:

in the formula (I), the compound is shown in the specification,

the voltage equation for the motor is as follows:

in the formula of U_d、U_qRespectively representing dq coordinate axis voltages; r_sRepresenting the motor stator resistance; p represents the differential; w represents the stator angular frequency;

three-phase current i of the motor_A、i_BAnd i_CI is obtained through a change formula of converting a three-phase static coordinate system into a two-phase static coordinate system_αAnd i_βThen i is_αAnd i_βI is obtained through a change formula of converting a two-phase static coordinate system into a two-phase rotating coordinate_dAnd i_q；

Exciting current change rate is measured by the exciting current i of this beat_dAnd a predicted value i_{d_pre}Obtaining as shown in formula (10); similarly, the torque current change rate calculation formula is shown in (11);

in the formula i_{d_pre}、i_{q_pre}Respectively representing dq coordinate axis prediction currents; t is_sIs the sampling interval time;

bringing formulae (10) and (11) into formulae (8) and (9) gives:

obtaining an excitation current predicted value i through the formula_{d_pre}And torque current predicted value i_{q_pre}；

The synchronous rotation angle of the next beat is changed into phi, phi is obtained through calculation of the synchronous rotation angle theta of the motor of the next beat, and a calculation formula is as follows:

φ＝θ+w_sT_s(14)

in the formula, w_sIs the synchronous angular frequency; theta is the synchronous rotation angle of the motor; phi is the next beat synchronous rotation angle;

then the αβ coordinate axis prediction current i is obtained by a change formula of converting a two-phase rotating coordinate system into a two-phase stationary coordinate_{α_pre}And i_{β_pre}The transformation formula for transforming the two-phase rotating coordinate system to the two-phase stationary coordinate is as follows:

in the formula i_{d_pre}、i_{q_pre}Respectively representing dq coordinate axis prediction currents; i.e. i_{α_pre}、i_{β_pre}Respectively representing αβ coordinate axis predicted current;

finally, obtaining the three-phase stationary coordinate axis prediction current i through a change formula of converting the two-phase stationary coordinate system to the three-phase stationary coordinate system_{A_pre}、i_{B_pre}And i_{C_pre}The change formula of the motor current transformed from the two-phase static coordinate system to the three-phase static coordinate system is as follows:

in the formula i_{A_pre}、i_{B_pre}And i_{C_pre}Respectively representing the predicted current of the three phases of the motor.

Technical Field

The invention relates to a dead zone compensation method for motor control, in particular to a dead zone compensation method suitable for permanent magnet motor control based on current prediction.

Background

The inverter main circuit topology of the electric locomotive generally adopts a bridge circuit structure, the switching devices of bridge arms adopt high-voltage-grade IGBTs, and because the IGBTs are not ideal devices and have turn-on and turn-off delay, certain dead time needs to be added into upper and lower IGBT driving pulses of the same bridge arm to ensure the reliable work of the switching devices; the turn-on and turn-off delay of the high-voltage level IGBT is more serious, so in order to ensure the reliable work of devices, longer dead time needs to be added to the driving pulse of the upper and lower tubes, the added dead time can cause the problem that the actual output voltage waveform is inconsistent with the theoretical voltage waveform, so that a dead time effect is caused, the dead time effect can generate harmonic voltage and current with different frequencies, the operation of a motor is influenced, particularly, the dead time effect is worse under the working condition of low-speed light load of a variable-frequency speed control system, and therefore the dead time needs to be compensated.

Disclosure of Invention

The invention solves the problem of poor dead zone compensation effect in the prior art, and provides a dead zone compensation method suitable for permanent magnet motor control based on current prediction for the dead zone compensation control of a permanent magnet motor. The method predicts the motor current by combining coordinate transformation and a permanent magnet motor equivalent model, and performs dead zone compensation by predicting the motor current.

The invention is realized by adopting the following technical scheme: the dead zone compensation method based on current prediction and suitable for permanent magnet motor control is divided into a low-speed zone dead zone compensation strategy (below rated frequency) and a high-speed zone dead zone compensation strategy (above rated frequency) according to the running frequency of a motor;

when the motor runs at a low speed stage (below a rated frequency), firstly, three-phase currents i of the motor are supplied_A、i_BAnd i_CI is obtained through a change formula of converting a three-phase static coordinate system into a two-phase static coordinate system_αAnd i_βThen i is_αAnd i_βThe dq axis current i is obtained through a change formula of converting a two-phase static coordinate system into a two-phase rotating coordinate system_dAnd i_q(ii) a Then calculating the obtained i_dAnd i_qThe predicted αβ axis current i is obtained by a change formula of converting a two-phase rotating coordinate system into a two-phase stationary coordinate system_{α_pre}And i_{β_pre}Where phi used for coordinate transformation passes phi + w_sT_sCalculated, wherein phi is the synchronous rotation angle of the next beat, theta is the synchronous rotation angle of the motor of the current beat, and w_sIs the synchronous angular frequency, T_sIs the sampling interval time; and obtaining the predicted three-phase current i of the motor by a change formula of converting the two-phase static coordinate system into the three-phase static coordinate system_{A_pre}、i_{B_pre}And i_{C_pre}(ii) a I obtained by calculation_{A_pre}、i_{B_pre}And i_{C_pre}Performing dead zone compensation;

when the motor runs at a high speed stage (above rated frequency), firstly, the three-phase current i of the motor is converted into the three-phase current_A、i_BAnd i_CI is obtained through a change formula of converting a three-phase static coordinate system into a two-phase static coordinate system_αAnd i_βThen i is_αAnd i_βI is obtained through a change formula of converting a two-phase static coordinate system into a two-phase rotating coordinate_dAnd i_q(ii) a Then calculating the obtained i_dAnd i_qObtaining predicted dq axis current i through stator voltage equation calculation_{d_pre}And i_{q_pre}，i_{d_pre}And i_{q_pre}The predicted αβ axis current i is obtained by a change formula of converting a two-phase rotating coordinate system into a two-phase stationary coordinate system_{α_pre}And i_{β_pre}Where phi used for coordinate transformation passes phi + w_sT_sCalculated, where φ is the nextThe synchronous rotation angle of the racket is theta, the synchronous rotation angle of the motor of the racket is w_sIs the synchronous angular frequency, T_sIs the sampling interval time; and obtaining the predicted three-phase current i of the motor by a change formula of converting the two-phase static coordinate system into the three-phase static coordinate system_{A_pre}、i_{B_pre}And i_{C_pre}(ii) a I obtained by calculation_{A_pre}、i_{B_pre}And i_{C_pre}Dead zone compensation is performed.

The dead zone compensation is carried out through the prediction current, and the problem that due to the fact that a digital controller has certain delay, the calculated result of the beat is updated until the next beat, and the dead zone compensation effect of the zero-crossing point accessory is poor is solved.

Drawings

FIG. 1 is a schematic diagram of current closed loop control during a low speed phase;

FIG. 2 is a diagram of a main circuit topology employed in the present invention;

FIG. 3 is i_{A_pre}0 dead zone compensation schematic diagram;

FIG. 4 shows i_{A_pre}< 0 dead zone compensation schematic diagram;

fig. 5 is a schematic diagram of the current open loop control during the high speed phase.

Detailed Description

The dead zone compensation method based on current prediction and suitable for permanent magnet motor control is divided into a low-speed zone dead zone compensation strategy (below rated frequency) and a high-speed zone dead zone compensation strategy (above rated frequency) according to the running frequency of a motor;

1) low speed zone dead zone compensation strategy (below rated frequency)

In the low-speed stage, the control strategy adopts current closed-loop control, and the control strategy is shown in figure 1. The motor current value is predicted by coordinate transformation.

Obtaining a current predicted value through coordinate transformation at a low-speed stage to perform dead zone compensation control, and specifically comprising the following steps:

3/2 transformation formula for transforming motor current from three-phase stationary coordinate system to two-phase stationary coordinate system:

in the formula i_A、i_BAnd i_CRespectively representing three-phase currents of the motor; i.e. i_α、i_βRespectively representing two-phase stationary αβ axis currents.

2/2 transformation formula for transformation of motor current from two-phase stationary coordinate system to two-phase rotating coordinate:

in the formula i_α、i_βRespectively representing αβ coordinate axis currents i_d、i_qRespectively representing two-phase rotating dq coordinate axis currents; theta is the synchronous rotation angle of the motor of the beat.

The synchronous rotation angle of the next beat is changed into phi, phi is obtained through calculation of the synchronous rotation angle theta of the motor of the next beat, and a calculation formula is as follows:

φ＝θ+w_sT_s(3)

in the formula, w_sIs the synchronous angular frequency; t is_sIs the sampling interval time; phi is the next beat synchronous rotation angle.

Then the αβ coordinate axis prediction current i is obtained by a change formula of converting a two-phase rotating coordinate system into a two-phase stationary coordinate_{α_pre}And i_{β_pre}The 2/2 transformation formula for transforming the two-phase rotating coordinate system to the two-phase stationary coordinate system is:

in the formula i_{α_pre}、i_{β_pre}Respectively representing αβ coordinate axes predicted current.

Finally, obtaining three-phase static coordinate axis prediction motor three-phase current i through a change formula of converting a two-phase static coordinate system into a three-phase static coordinate system_{A_pre}、i_{B_pre}And i_{C_pre}2/3 transformation formula for transformation of motor current from two-phase stationary frame to three-phase stationary frame:

in the formula i_{A_pre}、i_{B_pre}And i_{C_pre}Respectively representing the predicted current of the three phases of the motor.

I obtained by calculation of dead zone compensation module_{A_pre}、i_{B_pre}And i_{C_pre}Dead zone compensation is performed. The specific compensation process is as follows:

the main circuit topology adopted by the invention is a three-phase voltage type inverter as shown in fig. 2, wherein A, B and C respectively represent three bridge arms of the inverter. The dead zone effect is analyzed by taking the arm A as an example, and V1 and V2 correspond to the upper and lower tubes of the arm A.

Phase A is passed through judgment i_{A_pre}The polarity of the voltage is subjected to dead zone compensation, the voltage V1 and the voltage V2 correspond to two IGBT devices of an A-phase bridge arm, and when i is equal to the voltage I, the voltage I is analyzed through the driving pulse and the output voltage waveform of the voltage V1 and the voltage V2_{A_pre}The principle of dead zone compensation is shown in FIG. 3, where V is_AOThe voltage representing point a for O is a theoretical voltage waveform without added dead zone; v1_ pulse and V2_ pulse are drive pulses of V1 and V2, respectively.

When i is_{A_pre}When the voltage is more than 0, the V2 turn-off process is carried out by adding V1 after dead zone compensation, as shown in b) in FIG. 3, and the pulse of V1 and V_AOKeeping consistent, the pulse of V2 advances the dead time T _ dead to turn off; the process of turning off V2 by V1 is shown as c) in FIG. 3, and the added dead zone compensation pulse is the pulse of V1 and the pulse of V2_AOIn agreement, the pulse of V2 is delayed by the dead time T dead on.

When i is_{A_pre}If < 0, the compensation principle is as shown in FIG. 4.

When i is_{A_pre}When the voltage is less than 0, V1 is turned on after dead zone compensation, V2 is turned off as shown in b) in FIG. 4, and the pulse of V2 and V_AOKeeping the complementation, and turning on the pulse delay dead time T _ dead of V1; the process of turning off V2 by V1 is shown as c) in FIG. 4, and the added dead zone compensation pulse is the pulse of V2 and the pulse of V2_AORemaining complementary, the pulse of V1 is turned off early by the dead time T _ dead.

Phase B passing judgment i_{B_pre}Polarity dead zoneCompensation when i_{B_pre}When the value is more than 0, the compensation principle is shown in FIG. 3; when i is_{B_pre}If < 0, the compensation principle is as shown in FIG. 4. C phase passing judgment i_{C_pre}Is dead zone compensated when i_{C_pre}When the value is more than 0, the compensation principle is shown in FIG. 3; when i is_{C_pre}If < 0, the compensation principle is as shown in FIG. 4.

2) Dead zone compensation strategy of high speed zone (above rated frequency)

In the high-speed stage, the control strategy adopts current open-loop control, the motor current value is predicted through a permanent magnet motor equivalent model, and a schematic diagram is shown in fig. 5.

According to the equivalent circuit of the permanent magnet motor, the flux linkage equation of the motor is as follows:

in the formula (I), the compound is shown in the specification,respectively represent dq magnetic chains; l is_d、L_qAre dq-axis inductances, respectively; i.e. i_d、i_qAre dq-axis currents, respectively;

representing a permanent magnet flux linkage.

The voltage equation for the motor is as follows:

in the formula of U_d、U_qRespectively representing dq coordinate axis voltages; r_sRepresenting the motor stator resistance; p represents the differential; w represents stator angular frequency。

Three-phase current i of the motor_A、i_BAnd i_CI is obtained through a change formula of converting a three-phase static coordinate system into a two-phase static coordinate system_αAnd i_βThen i is_αAnd i_βI is obtained through a change formula of converting a two-phase static coordinate system into a two-phase rotating coordinate_dAnd i_q。

Exciting current change rate is measured by the exciting current i of this beat_dAnd a predicted value i_{d_pre}Obtaining as shown in formula (10); similarly, the torque current change rate calculation formula is shown in (11).

In the formula i_{d_pre}、i_{q_pre}Respectively representing dq coordinate axis prediction currents; t is_sIs the sampling interval time.

Bringing formulae (10) and (11) into formulae (8) and (9) gives:

obtaining an excitation current predicted value i through the formula_{d_pre}And torque current predicted value i_{q_pre}And obtaining the current value of the motor after coordinate transformation.

The synchronous rotation angle of the next beat is changed into phi, phi is obtained through calculation of the synchronous rotation angle theta of the motor of the next beat, and a calculation formula is as follows:

φ＝θ+w_sT_s(14)

in the formula, w_sIs the synchronous angular frequency; theta is the synchronous rotation angle of the motor; phi is the next beat synchronous rotation angle.

Then the αβ coordinate axis prediction current i is obtained by a change formula of converting a two-phase rotating coordinate system into a two-phase stationary coordinate_{α_pre}And i_{β_pre}The transformation formula for transforming the two-phase rotating coordinate system to the two-phase stationary coordinate is as follows:

in the formula i_{d_pre}、i_{q_pre}Respectively representing dq coordinate axis prediction currents; i.e. i_{α_pre}、i_{β_pre}Respectively representing αβ coordinate axes predicted current.

Finally, obtaining the three-phase stationary coordinate axis prediction current i through a change formula of converting the two-phase stationary coordinate system to the three-phase stationary coordinate system_{A_pre}、i_{B_pre}And i_{C_pre}The change formula of the motor current transformed from the two-phase static coordinate system to the three-phase static coordinate system is as follows:

in the formula i_{A_pre}、i_{B_pre}And i_{C_pre}Respectively representing the predicted current of the three phases of the motor.

I obtained by calculation of dead zone compensation module_{A_pre}、i_{B_pre}And i_{C_pre}Dead zone compensation is performed. Phase A is passed through judgment i_{A_pre}Is dead zone compensated when i_{A_pre}When the value is more than 0, the compensation principle is shown in FIG. 3; when i is_{A_pre}If < 0, the compensation principle is as shown in FIG. 4. Phase B passing judgment i_{B_pre}Is dead zone compensated when i_{B_pre}When the value is more than 0, the compensation principle is shown in FIG. 3; when i is_{B_pre}If < 0, the compensation principle is as shown in FIG. 4. C phase passing judgment i_{C_pre}Is dead zone compensated when i_{C_pre}When the value is more than 0, the compensation principle is shown in FIG. 3; when i is_{C_pre}If < 0, the compensation principle is as shown in FIG. 4.