阅读说明：本技术 时间最优无轴承磁通切换电机转矩及悬浮力预测控制方法 (Time-optimal bearingless magnetic flux switching motor torque and suspension force prediction control method ) 是由 周扬忠 刘汪彤 钟天云 屈艾文 杨公德 于 2021-09-18 设计创作，主要内容包括：本发明涉及一种时间最优无轴承磁通切换电机转矩及悬浮力预测控制方法。首先,通过预选的18个零序电压为0的基本电压矢量,结合采样电流,计算出下一周期内将产生的反电动势。然后,将预选电压矢量产生的反电动势代入代价函数中,筛选出最优的电压矢量编号n,并记录代价函数计算数值。代价函数通过进一步的计算得到矢量的最优作用时间。最后,使用计算出的电压矢量和作用时间在下一周期时控制逆变器。本发明能够解决六相逆变器供电情况下,电机的电磁转矩和悬浮力的快速而精确控制的难题。(The invention relates to a time-optimal bearingless flux switching motor torque and suspension force prediction control method. Firstly, through the pre-selected 18 basic voltage vectors with zero sequence voltage of 0, and combining with the sampling current, the counter electromotive force to be generated in the next period is calculated. And then substituting the counter electromotive force generated by the preselected voltage vector into a cost function, screening out the optimal voltage vector number n, and recording the calculation value of the cost function. The cost function obtains the optimal action time of the vector through further calculation. Finally, the inverter is controlled at the next cycle using the calculated voltage vector and the application time. The invention can solve the problem of rapid and accurate control of the electromagnetic torque and the levitation force of the motor under the condition of power supply of the six-phase inverter.)

1. A time optimal bearingless magnetic flux switching motor torque and suspension force prediction control method is characterized in that firstly, through 18 preselected basic voltage vectors with zero-sequence voltage of 0, the counter electromotive force to be generated in the next period is calculated by combining with sampling current; then substituting the back electromotive force generated by the preselected voltage vector into a cost function, screening out the optimal voltage vector number n, recording the calculation value of the cost function, and further calculating by the cost function to obtain the optimal action time of the vector; finally, the inverter is controlled at the next cycle using the calculated voltage vector and the application time.

2. The time-optimal bearingless magnetic flux switching motor torque and levitation force prediction control method according to claim 1, is specifically realized by the following steps: sampling six-phase current i at time K_A～i_FBus voltage U_DCRadial rotor displacements x and y, rotor tangential position angle theta_eAnd calculating the rotor speed omega_r(ii) a Converting the six-phase current sampling value to a static coordinate system to obtain a torque plane current alpha_TAxis beta_TAxial component i_αT、i_βTAnd a floating plane current alpha_sAxis beta_sAxial component i_αS、i_βS(ii) a Obtaining the torque plane stator flux linkage alpha of the motor according to the mathematical model of the motor_TAxis beta_TAxial component psi_αT、ψ_βTAnd a suspended planar stator flux linkage alpha_sAxis beta_sAxial component psi_αS、ψ_βS(ii) a Obtaining torque plane stator flux linkage alpha through direct torque control and direct suspension force control method_TAxis beta_TAxial component increment Δ ψ_αT、Δψ_βTAnd a suspended planar stator flux linkage alpha_sAxis beta_sAxial component increment Δ ψ_αS、Δψ_βS(ii) a Based on Delta psi_αT、Δψ_βT、Δψ_αS、Δψ_βSCalculating the back electromotive force alpha corresponding to the preselected voltage vector_Tβ_Tα_Sβ_SShafting components are calculated, corresponding cost functions are obtained through calculation and substituted into the cost functions, and finally the optimal voltage vector for controlling the inverter and the optimal acting time of the optimal voltage vector are obtained to control the inverter in the next period.

3. The time-optimal bearingless magnetic flux switching motor torque and levitation force predictive control method according to claim 2, which is characterized by comprising the following concrete implementation steps:

step S1, detecting six-phase winding current i by using current detection channel_A～i_FDetecting the DC bus of the inverter by using the voltage detection channelVoltage U_DCDetecting a rotor tangential position angle theta using a rotor tangential position sensor_eDetecting the radial displacement x and y of the rotor by using a radial displacement sensor of the rotor;

step S2, six-phase winding current i_A～i_FObtaining the torque plane stator current alpha through T6 transformation_TAxis beta_TAxial component i_αT、i_βTAnd a floating planar stator current alpha_sAxis beta_sAxial component i_αs、i_βs，i_z1、i_z2All are zero-sequence currents, and the T6 transformation formula is as follows:

step S3, according to i_αT、i_βT、i_αS、i_βSRadial displacement x and y of rotor, tangential position angle theta of rotor_eCalculating the torque plane stator flux linkage alpha_TAxis beta_TAxis psi_αT、ψ_βTAnd a suspended planar stator flux linkage alpha_sAxis beta_sAxis psi_αs、ψ_βs：

Wherein L is_TIs a torque plane inductance, L_SFor suspended planar inductance, | ψ_fTI is the torque plane permanent magnetic flux linkage amplitude, K is the suspension force coefficient,is the suspension force phase difference;

step S4, setting according to the torqueAnd electromagnetic torque T_eCalculating a torque error Δ T_e：

Step S5, based on the motor torque plane mathematical model, according to the torque error delta T_eTorque plane stator flux linkage amplitude settingTorque plane stator flux linkage psi_αTAnd psi_βTTorque angle increment delta, calculating torque plane stator flux linkage alpha_TAxis beta_TAxial component increment Δ ψ_αT、Δψ_βT：

Wherein, ω is_rFor the electrical angular velocity, T, of the rotor rotation_sIs a digital control period;

step S6, according to the suspension force given in the x and y directionsAndx and y direction suspension force F_xAnd F_yCalculating the suspension force error delta F in the x and y directions_xAnd Δ F_y：

Step S7, based on the motor suspension plane mathematical model, according to the suspension force error delta F in the x and y directions_xAnd Δ F_yCalculating the magnetic linkage alpha of the suspended planar stator_sAxis beta_sAxial increment Δ ψ_αS、Δψ_βS：

Step S8, selecting one voltage vector from No. 2-19 voltage vectors in sequenceAccording to its on-off state S_A(i)～S_F(i)DC bus voltage U_DCCalculating alpha_Tβ_Tα_Sβ_SThe shafting components are as follows:

wherein, i is 2, 3.., 19;

step S9, according to the ith voltage vector alpha_Tβ_Tα_Sβ_SAxial component u_αT(i)、u_βT(i)、u_αS(i)、u_βS(i)And the winding current alpha_Tβ_Tα_Sβ_SAxial component i_αT、i_βT、i_αS、i_βSCalculating the back electromotive force alpha corresponding to the ith voltage vector_Tβ_Tα_Sβ_SShafting component:

step S10, the counter electromotive force α corresponding to the ith voltage vector_Tβ_Tα_Sβ_SAxial component and torque plane stator flux linkage increment delta psi_αT、Δψ_βTMagnetic linkage increment delta psi of suspension plane stator_αS、Δψ_βSCalculating cost function cost corresponding to ith voltage vector_(i)And optimal workBy time t_s(i)：

cost_(i)＝Δψ_αTE_αT(i)+Δψ_βTE_βT(i)+Δψ_αSE_αS(i)+Δψ_βSE_βS(i)

Step S11, corresponding to 18 cost functions cost according to 18 voltage vectors_(i)(i 2, 3.., 19) from which the minimum cost is found_(n)Finding out the corresponding voltage vector as the optimal voltage vector, and the corresponding inverter switching state is S_A(n)～S_F(n)Optimum acting time t of the corresponding optimum voltage vector_s(n)：

Step S12, optimum acting time t to optimum voltage vector_s(n)Perform clipping if t_s(n)If < 0, then t_s(n)0; if t_s(n)＞T_sThen t is_s(n)＝T_s；

Step S13, obtaining the optimal action time t of the optimal voltage vector according to the step S12_s(n)Calculating the zero voltage vector action time t₀；

t₀＝T_s-t_s(n)

Step S14, the inverter switch state corresponding to the optimal voltage vector is S_A(n)～S_F(n)Optimal action time t of optimal voltage vector_s(n)Zero voltage vector action time t₀By means of a PWM method, with a six-phase inverter output with an action time t_s(n)The optimal voltage vector realizes the accurate control of the magnetic linkage of the torque plane and the suspension plane, finally achieves the accurate control of the electromagnetic torque and the suspension force, and reduces the steady-state pulsation of the electromagnetic torque and the suspension force.

4. The time-optimal bearingless flux switching motor torque and levitation force prediction control method according to claim 3, wherein in step S3, the levitation force coefficient K and the levitation force phase differenceThe obtaining method comprises the following steps:

step S31, using rotor tangential position angle theta_eTorque plane current i_αT、i_βTRotation transformation to d_TqT shafting

Step S32, according to i_dT、i_qTCalculating the suspension force coefficient K and the suspension force phase differenceAs follows

Wherein k is_PM、k_dT、k_qTRespectively being permanent magnets, unit d_Tq_TThe fundamental amplitude of the suspension force generated by the interaction of the shaft current and the unit suspension force current is obtained by finite element simulation software.

5. The method for predictive control of torque and levitation force in a time-optimal bearingless flux switching motor of claim 3, wherein in step S4, torque is givenObtainable by means of a speed closed-loop controller, electromagnetic torque T_eThe magnetic flux can be obtained from the cross product of the torque plane flux linkage and the current:

T_e＝n_p(ψ_αTi_βT-ψ_βTi_αT)

wherein n is_pIs the number of pole pairs of the motor.

6. The method for predictive control of torque and levitation force of a time-optimal bearingless flux switching motor according to claim 3, wherein in step S5, the method for obtaining the torque angle increment Δ δ comprises: will torque error Δ T_eObtaining a torque angle delta for a PI regulator

Δδ＝K_pTΔT_e+K_iT∫ΔT_edt

Wherein, K_pT、K_iTProportional and integral coefficients, respectively.

7. The method for predictive control of torque and levitation force in a time-optimized bearingless flux switching motor of claim 3, wherein levitation forces in x and y directions are given in step S6Obtainable by means of an x, y radial displacement closed loop controller.

8. The method for predictive control of torque and levitation force in a time-optimized bearingless flux switching motor of claim 7, wherein in step S6, levitation forces F in x and y directions_xF_yThe calculation can be carried out by means of a suspension plane mathematical model to obtain:

Technical Field

The invention relates to a time-optimal bearingless flux switching motor torque and suspension force prediction control method.

Background

The six-phase single-winding bearingless flux switching motor is powered by a six-phase inverter, and voltage vectors output by the inverter need to be simultaneously controlled on a torque plane, a suspension plane and a zero-sequence current plane so as to realize the rotating operation of a rotor in a magnetic suspension state. The six-phase single-winding bearingless magnetic flux switching motor driving system controlled by direct torque and suspension force has the advantage of rapid dynamic response of the torque and the suspension force, but has the defect that (1) because a six-phase inverter can output 64 voltage vectors, an optimal switching table is difficult to establish, and the optimal state is difficult to achieve; (2) the voltage vector selected from the optimal switching vector table is applied to the whole digital control period, so that the control of the torque and the suspension force is in an overshoot state or an under-control state, and the magnetic suspension rotation performance of the rotor is influenced due to the large torque and suspension force pulsation. (3) In the traditional prediction control, the calculation of the cost function value occupies a large amount of calculation resources, but the cost function value is only used for screening, and the value is not further mined. One of the key scientific problems for reducing torque ripple and suspension force ripple is to optimize an optimal voltage vector and determine the optimal acting time according to the torque plane error, the suspension plane error and the zero sequence plane error.

Therefore, the invention provides a torque and levitation force prediction control method based on time optimization aiming at a six-phase single-winding bearingless magnetic flux switching motor.

Disclosure of Invention

The invention aims to provide a time-optimal bearingless magnetic flux switching motor torque and levitation force prediction control method to solve the problem of rapid and accurate control of electromagnetic torque and levitation force of a motor under the condition of power supply of a six-phase inverter.

In order to achieve the purpose, the technical scheme of the invention is as follows: a time optimal bearingless magnetic flux switching motor torque and suspension force prediction control method comprises the steps of firstly, calculating back electromotive force to be generated in the next period through 18 preselected basic voltage vectors with zero-sequence voltage of 0 and combining sampling current; then substituting the back electromotive force generated by the preselected voltage vector into a cost function, screening out the optimal voltage vector number n, recording the calculation value of the cost function, and further calculating by the cost function to obtain the optimal action time of the vector; finally, the inverter is controlled at the next cycle using the calculated voltage vector and the application time.

In an embodiment of the present invention, the method is specifically implemented as follows: sampling six-phase current i at time K_A～i_FBus voltage U_DCRadial rotor displacements x and y, rotor tangential position angle theta_eAnd calculating the rotor speed omega_r(ii) a Converting the six-phase current sampling value to a static coordinate system to obtain a torque plane current alpha_TAxis beta_TAxial component i_αT、i_βTAnd a floating plane current alpha_sAxis beta_sAxial component i_αS、i_βS(ii) a Obtaining the torque plane stator flux linkage alpha of the motor according to the mathematical model of the motor_TAxis beta_TAxial component psi_αT、ψ_βTAnd a suspended planar stator flux linkage alpha_sAxis beta_sAxial component psi_αS、ψ_βS(ii) a Obtaining torque plane stator flux linkage alpha through direct torque control and direct suspension force control method_TAxis beta_TAxial component increment Δ ψ_αT、Δψ_βTAnd a suspended planar stator flux linkage alpha_sAxis beta_sAxial component increment Δ ψ_αS、Δψ_βS(ii) a Based on Delta psi_αT、Δψ_βT、Δψ_αS、Δψ_βSCalculating a preselected voltageCounter electromotive force alpha corresponding to vector_Tβ_Tα_Sβ_SShafting components are calculated, corresponding cost functions are obtained through calculation and substituted into the cost functions, and finally the optimal voltage vector for controlling the inverter and the optimal acting time of the optimal voltage vector are obtained to control the inverter in the next period.

In an embodiment of the present invention, the method specifically includes the following steps:

step S1, detecting six-phase winding current i by using current detection channel_A～i_FDetecting the DC bus voltage U of the inverter by using the voltage detection channel_DCDetecting a rotor tangential position angle theta using a rotor tangential position sensor_eDetecting the radial displacement x and y of the rotor by using a radial displacement sensor of the rotor;

step S2, six-phase winding current i_A～i_FObtaining the torque plane stator current alpha through T6 transformation_TAxis beta_TAxial component i_αT、i_βTAnd a floating planar stator current alpha_sAxis beta_sAxial component i_αs、i_βs，i_z1、i_z2All are zero-sequence currents, and the T6 transformation formula is as follows:

step S3, according to i_αT、i_βT、i_αS、i_βSRadial displacement x and y of rotor, tangential position angle theta of rotor_eCalculating the torque plane stator flux linkage alpha_TAxis beta_TAxis psi_αT、ψ_βTAnd a suspended planar stator flux linkage alpha_sAxis beta_sAxis psi_αs、ψ_βs：

Wherein L is_TIs a torque plane inductance, L_SFor suspended planar inductance, | ψ_fTI is the torque plane permanent magnet flux linkage amplitude, K isThe coefficient of the suspension force is,is the suspension force phase difference;

step S4, setting according to the torqueAnd electromagnetic torque T_eCalculating a torque error Δ T_e：

Step S5, based on the motor torque plane mathematical model, according to the torque error delta T_eTorque plane stator flux linkage amplitude settingTorque plane stator flux linkage psi_αTAnd psi_βTTorque angle increment delta, calculating torque plane stator flux linkage alpha_TAxis beta_TAxial component increment Δ ψ_αT、Δψ_βT：

Wherein, ω is_rFor the electrical angular velocity, T, of the rotor rotation_sIs a digital control period;

step S6, according to the suspension force given in the x and y directionsAndx and y direction suspension force F_xAnd F_yCalculating the suspension force error delta F in the x and y directions_xAnd Δ F_y：

Step S7, based on the motor suspension plane mathematical model, according to the suspension force error delta F in the x and y directions_xAnd Δ F_yCalculating the magnetic linkage alpha of the suspended planar stator_sAxis beta_sAxial increment Δ ψ_αS、Δψ_βS：

Step S8, selecting one voltage vector from No. 2-19 voltage vectors in sequenceAccording to its on-off state S_A(i)～S_F(i)DC bus voltage U_DCCalculating alpha_Tβ_Tα_Sβ_SThe shafting components are as follows:

wherein, i is 2, 3.., 19;

step S9, according to the ith voltage vector alpha_Tβ_Tα_Sβ_SAxial component u_αT(i)、u_βT(i)、u_αS(i)、u_βS(i)And the winding current alpha_Tβ_Tα_Sβ_SAxial component i_αT、i_βT、i_αS、i_βSCalculating the back electromotive force alpha corresponding to the ith voltage vector_Tβ_Tα_Sβ_SShafting component:

step S10, the counter electromotive force α corresponding to the ith voltage vector_Tβ_Tα_Sβ_SAxial component and torque plane stator flux linkage increment delta psi_αT、Δψ_βTMagnetic linkage increment delta psi of suspension plane stator_αS、Δψ_βSCalculating cost function cost corresponding to ith voltage vector_(i)And optimum action time t_s(i)：

cost_(i)＝Δψ_αTE_αT(i)+Δψ_βTE_βT(i)+Δψ_αSE_αS(i)+Δψ_βSE_βS(i)

Step S11, corresponding to 18 cost functions cost according to 18 voltage vectors_(i)(i 2, 3.., 19) from which the minimum cost is found_(n)Finding out the corresponding voltage vector as the optimal voltage vector, and the corresponding inverter switching state is S_A(n)～S_F(n)Optimum acting time t of the corresponding optimum voltage vector_s(n)：

Step S12, optimum acting time t to optimum voltage vector_s(n)Perform clipping if t_s(n)If < 0, then t_s(n)0; if t_s(n)＞T_sThen t is_s(n)＝T_s；

Step S13, obtaining the optimal action time t of the optimal voltage vector according to the step S12_s(n)Calculating the zero voltage vector action time t₀；

t₀＝T_s-t_s(n)

Step S14, the inverter switch state corresponding to the optimal voltage vector is S_A(n)～S_F(n)Optimal action time of optimal voltage vectort_s(n)Zero voltage vector action time t₀By means of a PWM method, with a six-phase inverter output with an action time t_s(n)The optimal voltage vector realizes the accurate control of the magnetic linkage of the torque plane and the suspension plane, finally achieves the accurate control of the electromagnetic torque and the suspension force, and reduces the steady-state pulsation of the electromagnetic torque and the suspension force.

In an embodiment of the invention, in step S3, the levitation force coefficient K and the levitation force phase differenceThe obtaining method comprises the following steps:

step S31, using rotor tangential position angle theta_eTorque plane current i_αT、i_βTRotation transformation to d_Tq_TShaft system

Step S32, according to i_dT、i_qTCalculating the suspension force coefficient K and the suspension force phase differenceAs follows

Wherein k is_PM、k_dT、k_qTRespectively being permanent magnets, unit d_Tq_TThe fundamental amplitude of the suspension force generated by the interaction of the shaft current and the unit suspension force current is obtained by finite element simulation software.

In one embodiment of the present invention, in step S4, the torque is givenObtainable by means of a speed closed-loop controller, electromagnetic torque T_eThe magnetic flux can be obtained from the cross product of the torque plane flux linkage and the current:

T_e＝n_p(ψ_αTi_βT-ψ_βTi_αT)

wherein n is_pIs the number of pole pairs of the motor.

In an embodiment of the present invention, in step S5, the torque angle increment Δ δ is obtained by: will torque error Δ T_eObtaining a torque angle delta for a PI regulator

Δδ＝K_pTΔT_e+K_iT∫ΔT_edt

Wherein, K_pT、K_iTProportional and integral coefficients, respectively.

In one embodiment of the present invention, in step S6, the levitation forces in the x and y directions are givenObtainable by means of an x, y radial displacement closed loop controller.

In step S6, according to one embodiment of the present invention, the levitation forces F in the x and y directions_xF_yThe calculation can be carried out by means of a suspension plane mathematical model to obtain:

compared with the prior art, the invention has the following beneficial effects: compared with the existing six-phase single-winding bearingless flux switching motor torque and suspension force control method, the method of the invention has the following advantages:

(1) the inverter is used for outputting the voltage vector with the optimal action time, so that the accurate control of the torque and the suspension force is realized, the stability of the magnetic suspension operation of the rotor is enhanced, the pulsation of the torque and the suspension force is reduced, and particularly the high-frequency pulsation of the motor is reduced;

(2) as the zero sequence voltage corresponding to the alternative voltage vector in the prediction is equal to zero, the zero sequence current flowing through the motor winding is greatly reduced, and the loss of the motor is reduced.

(3) The inverter output voltage vector is used for directly controlling the electromagnetic torque and the suspension force, so that the electromagnetic torque and the suspension force are quickly controlled, and the dynamic response speed of the system is improved.

Drawings

Fig. 1 is a block diagram of a structure for predicting and controlling torque and levitation force of a bearingless flux switching motor based on time optimization.

FIG. 2 is a schematic diagram of a predictive control algorithm of the present invention.

Fig. 3 is a six-phase single-winding bearingless flux switching motor structure.

Fig. 4 shows a hardware structure of a driving system according to an embodiment of the present invention.

FIG. 5 is a schematic view of a torque plane vector distribution.

FIG. 6 is a schematic view of the vector distribution of the suspension plane.

Fig. 7 is a schematic diagram of inverter voltage vector distribution.

FIG. 8 shows cost _ old (t)_s)～t_sSchematic diagram of quadratic function.

FIG. 9 is a schematic diagram of voltage vector action.

Detailed Description

The technical scheme of the invention is specifically explained below with reference to the accompanying drawings.

The invention discloses a prediction control method for torque and levitation force of a time-optimal bearingless magnetic flux switching motor, which can solve the problem of rapid and accurate control of electromagnetic torque and levitation force of the motor under the condition of power supply of a six-phase inverter. Preselecting 19 basic voltage vectors according to the principle that zero sequence voltage is zero; calculating a corresponding back electromotive force vector according to each basic voltage vector; calculating a cost function and voltage vector action time according to the torque plane flux linkage increment, the suspension plane flux linkage increment and the back electromotive force vector; determining the maximum value of the cost function values of the 18 voltage vectors except the zero voltage vector, thereby finding out that the corresponding basic voltage vector is the optimal voltage vector and the corresponding action time is the optimal action time; the predicted optimal voltage vector and the predicted optimal acting time act on the motor through the six-phase inverter, so that the electromagnetic torque and the levitation force can be accurately and quickly controlled. The specific explanation is as follows.

The time-optimal bearingless magnetic flux switching motor torque and suspension force prediction control structure block diagram is shown in figure 1, and the prediction control algorithm principle is shown in figure 2. As in FIG. 1, six-phase current i is sampled at time K_A～i_FBus voltage U_DCX and y radial displacements, rotor position electrical angle θ_eAnd calculating the rotor speed omega_r. Converting the six-phase current sampling value to a static coordinate system to obtain a torque plane current alpha_TAxis beta_TAxial component i_αT、i_βTAnd a floating plane current alpha_sAxis beta_sAxial component i_αS、i_βS. Obtaining the torque plane stator flux linkage alpha of the motor according to the mathematical model of the motor_TAxis beta_TAxial component psi_αT、ψ_βTAnd a suspended planar stator flux linkage alpha_sAxis beta_sAxial component psi_αS、ψ_βS. Obtaining torque plane stator flux linkage alpha through direct torque control and direct suspension force control method_TAxis beta_TAxial component increment Δ ψ_αT、Δψ_βTAnd a suspended planar stator flux linkage alpha_sAxis beta_sAxial component increment Δ ψ_αS、Δψ_βS. To make delta psi_αT、Δψ_βT、Δψ_αS、Δψ_βSAnd sending the voltage vector to a predictive control algorithm to finally obtain the optimal voltage vector for controlling the inverter and the optimal acting time of the optimal voltage vector.

As shown in fig. 2, first, through the preselected 18 basic voltage vectors with zero sequence voltage of 0, the counter electromotive force to be generated in the next cycle is calculated in combination with the sampled current. And then substituting the counter electromotive force generated by the preselected voltage vector into a cost function, screening out the optimal voltage vector number n, and recording the calculation value of the cost function. The cost function obtains the optimal action time of the vector through further calculation. Finally, the inverter is controlled at the next cycle using the calculated voltage vector and the application time.

The motor structure is shown in FIG. 3, theta_MIs the included angle of the axis direction A of the motor in the x direction. 12U-shaped iron cores of the motor are clamped between every two U-shaped iron cores and filled along the tangential directionThe magnetizing directions of the magnetic permanent magnets are alternately opposite, and the rotor has 10 teeth. And each phase of winding of the stator is wound on the stator teeth which are vertical to each other in space in series to form 6 symmetrical windings. The winding space of the A phase and the D phase is symmetrical, the winding space of the B phase and the E phase is symmetrical, the winding space of the C phase and the F phase is symmetrical, and the six-phase winding has a mechanical angle of 60 degrees in space. And if the rotating speed of the motor needs to be adjusted, the rotating speed closed-loop control is utilized to output a given torque value.

The invention provides a prediction control method for torque and suspension force of a time-optimal bearingless flux switching motor, aiming at the problem of accurate and rapid control of the torque and suspension force of a six-phase single-winding bearingless flux switching motor, which comprises the following specific implementation steps:

(1) detection of six-phase winding current i using current detection channel_A～i_FDetecting the DC bus voltage U of the inverter by using the voltage detection channel_DCDetecting a rotor tangential position angle theta using a rotor tangential position sensor_eAnd detecting the radial displacement x and y of the rotor by using a rotor radial displacement sensor.

(2) Six-phase winding current i_A～i_FObtaining the torque plane stator current alpha through T6 transformation_TAxis beta_TAxial component i_αT、i_βTAnd a floating planar stator current alpha_sAxis beta_sAxial component i_αs、i_βs。i_z1、i_z2Are all zero sequence currents.

(3) According to i_αT、i_βT、i_αS、i_βSRadial displacement x and y of rotor, tangential position angle theta of rotor_eCalculating the torque plane stator flux linkage alpha_TAxis beta_TAxis psi_αT、ψ_βTAnd a suspended planar stator flux linkage alpha_sAxis beta_sAxis psi_αs、ψ_βs。

Wherein L is_TIs a torque plane inductance, L_SFor suspended planar inductance, | ψ_fTAnd | is the torque plane permanent magnet flux linkage amplitude. K is the coefficient of the suspension force,is the suspension force phase difference.

(4) According to torque givenAnd electromagnetic torque T_eCalculating a torque error Δ T_e

(5) Based on a motor torque plane mathematical model and according to a torque error delta T_eTorque plane stator flux linkage amplitude settingTorque plane stator flux linkage psi_αTψ_βTTorque angle increment delta, calculating torque plane stator flux linkage alpha_TAxis beta_TAxial component increment Δ ψ_αT、Δψ_βT

Wherein, ω is_rFor the electrical angular velocity, T, of the rotor rotation_sIs a digital control period.

(6) Given according to the suspension force in the x and y directionsAndx and y direction suspension force F_xAnd F_yCalculating the suspension force error delta F in the x and y directions_xAnd Δ F_y

(7) Based on a motor suspension plane mathematical model, according to suspension force errors delta F in x and y directions_xAnd Δ F_yCalculating the magnetic linkage alpha of the suspended planar stator_sAxis beta_sAxial increment Δ ψ_αS、Δψ_βS

(8) Selecting one voltage vector from No. 2-19 voltage vectors in sequenceAccording to its on-off state S_A(i)～S_F(i)DC bus voltage U_DCCalculating alpha_Tβ_Tα_Sβ_SThe shafting components are as follows:

wherein, i is 2,3

(9) According to ith voltage vector alpha_Tβ_Tα_Sβ_SAxial component u_αT(i)、u_βT(i)、u_αS(i)、u_βS(i)And the winding current alpha_Tβ_Tα_Sβ_SAxial component i_αT、i_βT、i_αS、i_βSCalculating the back electromotive force alpha corresponding to the ith voltage vector_Tβ_Tα_Sβ_SComponent of axis system

(10) Back electromotive force alpha corresponding to ith voltage vector_Tβ_Tα_Sβ_SAxial component and torque plane stator flux linkage increment delta psi_αT、Δψ_βTMagnetic linkage increment delta psi of suspension plane stator_αS、Δψ_βSCalculating cost function cost corresponding to ith voltage vector_(i)And optimum action time t_s(i)

cost_(i)＝Δψ_αTE_αT(i)+Δψ_βTE_βT(i)+Δψ_αSE_αS(i)+Δψ_βSE_βS(i)

(11) 18 cost functions cost calculated according to the above_(i)(i 2, 3.., 19) from which the minimum cost is found_(n)Thereby finding out the corresponding voltage vector as the optimal voltage vector, and the corresponding inverter switching state is S_A(n)～S_F(n)Optimum acting time t of the corresponding optimum voltage vector_s(n)

(12) Optimal action time t on optimal voltage vector_s(n)Perform clipping if t_s(n)If < 0, then t_s(n)0; if t_s(n)＞T_sThen t is_s(n)＝T_s

(13) According to the optimal action time t of the optimal voltage vector obtained in the step (12)_s(n)Calculating the zero voltage vector action time t₀

t₀＝T_s-t_s(n)

(14) Switching state S corresponding to the optimal voltage vector_A(i)～S_F(i)Optimal action time t of optimal voltage vector_s(n)Zero voltage vector action time t₀By means of a PWM method, with a six-phase inverter output with an action time t_s(n)The optimal voltage vector realizes the accurate control of the magnetic linkage of the torque plane and the suspension plane, finally achieves the accurate control of the electromagnetic torque and the suspension force, and reduces the steady-state pulsation of the electromagnetic torque and the suspension force.

The suspension force coefficient K and the suspension force phase difference used in the step (3)The obtaining method comprises the following steps:

step (3.1) utilizing rotor tangential position angle theta_eTorque plane current i_αT、i_βTRotation transformation to d_Tq_TShaft system

Step (3.2) according to i_dT、i_qTCalculating the suspension force coefficient K and the suspension force phase differenceAs follows

Wherein k is_PM、k_dT、k_qTRespectively being permanent magnets, unit d_Tq_TThe fundamental amplitude of the suspension force generated by the interaction of the shaft current and the unit suspension force current can be obtained by finite element simulation software.

Torque given T in step (4)_e ^*Obtainable by means of a speed closed-loop controller, electromagnetic torque T_eThe magnetic flux can be obtained from the cross product of the torque plane flux linkage and the current:

T_e＝n_p(ψ_αTi_βT-ψ_βTi_αT)

wherein n is_pIs the number of pole pairs of the motor.

The method for obtaining the torque angle increment delta in the step (5) comprises the following steps: the torque error Delta T can be adjusted_eObtaining a torque angle delta for a PI regulator

Δδ＝K_pTΔT_e+K_iT∫ΔT_edt

Wherein, K_pT、K_iTProportional and integral coefficients, respectively.

Setting suspension force in x and y directions in step (6)Can be obtained by means of an x, y radial displacement closed-loop controller, for example, the xy radial displacement errors Deltax, Deltay can be obtained by a PI controller respectively

In the step (6), the suspension force F in the x and y directions_xF_yThe calculation can be carried out by means of a mathematical model of the suspension plane to obtain:

the following are specific examples of the present invention.

The hardware structure of the driving system for implementing the invention is shown in FIG. 4.

The whole control system comprises: the device comprises an alternating current power supply, a rectifying circuit, a filter circuit, a direct current bus part, a voltage acquisition circuit, a six-phase inverter, a six-phase bearingless motor, a rotary encoder, a rotor radial displacement acquisition circuit, a six-phase winding current acquisition circuit, a controller, an isolation drive part, a human-computer interaction part and the like.

Wherein the six-phase inverter dc bus voltage may also be provided using a suitable dc power supply. The power tube in the inverter adopts IGBT or MOSFET with parallel diode, and the controller adopts DSP or single chip microcomputer. The winding current acquisition circuit is formed by combining a Hall current sensor and an operational amplifier, and can also be formed by combining a winding series power resistor and a differential operational amplifier. The Hall scheme can effectively realize the electrical isolation of the control loop and the main loop, and the winding series power resistance scheme can reduce the cost of the driving system. The direct current bus voltage acquisition circuit is formed by combining a Hall voltage sensor and an operational amplifier, and can also be formed by combining a voltage follower formed by an operational amplifier after voltage division of a parallel resistor. The rotor position angle detection circuit can be formed by connecting a rotary encoder with a level conversion circuit and can also be formed by connecting a rotary transformer with a decoding circuit, wherein the cost of the former is lower, but the position angle sampling precision is limited by the number of lines of the encoder, and the cost of the latter is higher, but the position angle sampling precision is higher. The rotor radial xy offset acquisition circuit is formed by combining an eddy current sensor and a subsequent operational amplifier, and can also be formed by combining a linear optical coupler and a subsequent operational amplifier. Weak current signals output by the current detection circuit, the voltage sampling circuit and the rotor radial displacement acquisition circuit are sent to an A/D conversion module of the controller, and pulse signals output by the position angle detection circuit are sent to a QEP module of the controller. According to the obtained signal and the prediction control method of the present invention, an inverter arm switching signal is output, and the switching operation of the power switching tube in the inverter is controlled via the isolation driver.

The basic principle is described as follows:

in the static coordinate system, the torque plane vector distribution of the motor is shown in figure 5.

Wherein is defined as_TThe axis is consistent with the direction of the A phase winding,in order to provide a magnetic linkage of the rotor,in order to sample the resulting stator torque flux linkage,to turn to the targetThe magnetic flux linkage of the moment magnetic flux is adopted,for a target incremental flux linkage, θ_eIs alpha_TElectrical angle of the shaft to the flux linkage of the motor rotor, delta being the torque angle, delta₁Is an increment of torque angle change.

In a static coordinate system, the suspension plane vectors of the motor are distributed as shown in FIG. 6.

Wherein is defined as_SThe axis is consistent with the positive direction of the A-phase levitation current, and a virtual levitation flux linkage is definedAnd d_SThe axes are in the same direction, and the axial direction is the same,is alpha_SAxis to d_SElectrical angle of axis, gamma being virtual levitation flux linkageTo controllable suspension flux linkageElectrical angle of (γ)^*For virtual levitation flux linkageMagnetic linkage to target suspensionThe electrical angle of (a) of (b),the target levitated flux linkage increment is obtained.

The six-phase inverter can output 2⁶The voltage vectors can be projected on a torque plane, a levitation plane, and an o1 zero sequence plane, as shown in fig. 7.

Wherein, bridge arms A-F respectively use switch state variables S_A～S_FShowing the switching condition of the power tube. When the upper pipe of the bridge arm is opened and the lower pipe is closed,the switching signal is S_i1(i ═ a to F); when the lower tube of the bridge arm is switched on and the upper tube is switched off, the switching signal is S_i0(i ═ a to F); thus, a group of switching vectors is a six-bit binary number S_AS_BS_CS_DS_ES_F. The number in the figure is S_AS_BS_CS_DS_ES_FConverted to a decimal voltage vector number. In order to reduce the loss in the motor control, the zero sequence current is controlled to be zero, and 20 voltage vectors with zero sequence voltage of 0 can be selected from fig. 7 (c). Again, since the actions numbered 0 and 63 are repeated, the final selected voltage vector is 19 voltage vectors in bold italics in fig. 7. The numbering of the 19 voltage vectors is shown in table 1.

Table 119 voltage vector number correspondence tables

19 voltage vector numbering	In 26Number in voltage vector	SASBSCSDSESFNumbering
			1	0	000000
2	3	000011
			3	6	000110
4	9	001001
			5	12	001100
6	15	001111
			7	18	010010
8	24	011000
			9	27	011011
10	30	011110
			11	33	100001
12	36	100100
			13	39	100111
14	45	101101
			15	48	110000
16	51	110011
			17	54	110110
18	57	111001
			19	60	111100

In the preselected voltage vectors, since the zero vector (numbered 1 in 19 voltage vectors) has no influence on the motor flux linkage, the voltage vectors No. 2 to No. 19 are converted into the counter electromotive force vectors.

Wherein R is_sIs the resistance of the stator phase winding under the static coordinate system,for a pre-selected voltage vector in a stationary coordinate system,the stator current vector is sampled for a stationary frame.

After data is sampled, the target torque plane stator flux linkage increment can be obtained through direct torque control and direct suspension force controlAnd suspended planar stator flux linkage incrementIn conventional predictive control, the target flux linkage and the preselected back emf are substituted into a cost function cost _ old:

cost_old＝(Δψ_αT-E_{α T (Pre-selected)}t_s)²+(Δψ_βT-E_{Beta T (Pre-selected)}t_s)²+(Δψ_αS-E_{AlphaS (Pre-selected)}t_s)²+(Δψ_βS-E_{Beta S (Pre-selection)}t_s)² (2)

Then let t_sFor a control period T_sAnd the voltage vector corresponding to the minimum cost function cost _ old value is the optimal voltage vector. However, when the voltage vector is used for controlling the motor, the torque and the levitation force are either in an overshoot state or an under-control state, and even if the voltage vector is the optimal voltage vector, the following of the target magnetic linkage has large pulsation, so that the large torque and the levitation force pulsation occur, and the steady-state performance of the rotor levitation rotation operation is influenced.

Aiming at the problems, the invention improves the traditional cost function and constructs a new cost function.

T in order_sFor the argument, it can be seen that the cost function cost _ old is one with respect to t_sThe original cost function is sorted by the quadratic function of (1):

due to the fact that

Wherein, C₁、C₂Is a constant.

Therefore, substituting the voltage vectors with serial numbers 2-19 will result in a set of quadratic curves with substantially the same curvature and upward openings. Suppose that when flux linkage increment is given a certain timing, the trial substitution is carried outGoing to cost _ old, we get the diagram as shown in FIG. 8. Wherein the circle points are extreme points.

A certain preselected voltage vectorCalculating back EMF vector from substitutionSubstituting again to obtain the action time t corresponding to the minimum cost _ old_sThe following were used:

as can be seen from FIG. 7, at t_sWhen the minimum value of the curve is positive, t is obtained_sThe larger. Since the denominators are basically the same in the substitution of different voltage vectors, in order to reduce the calculation amount of predictive control and simplify, a new cost function is obtained:

cost＝Δψ_αTE_{α T (Pre-selected)}+Δψ_βTE_{Beta T (Pre-selected)}+Δψ_αSE_{AlphaS (Pre-selected)}+Δψ_βSE_{Beta S (Pre-selection)} (6)

In this way, the counter electromotive force vectors are calculated by sequentially substituting the voltage vectors No. 2 to 19, and the voltage vector n corresponding to the time when the cost is maximized is obtained as the optimum voltage vector.

Obtaining the cost function cost corresponding to the optimal voltage vector n by using the formula_(n)And then dividing by the denominator of the formula to obtain the optimal action time.

When t is_sOver controller period T_sThe action time is the controller period. When t is_sAnd when the control period is shorter than the control period, filling the control period by using a zero vector. The vector diagram is shown in fig. 9. Wherein the optimal voltage vectorActing only for the whole control period T_sT in (1)_sPart, remainder using zero vectorAnd filling, so that the precise control of the torque and the suspension plane flux linkage is realized, and the steady-state pulsation of the torque and the suspension force is further reduced.

The above are preferred embodiments of the present invention, and all changes made according to the technical scheme of the present invention that produce functional effects do not exceed the scope of the technical scheme of the present invention belong to the protection scope of the present invention.