A Dual-Battery Charging Control Method for Electric Vehicles Based on Multi-Vector Space Decoupling

Abstract

The present invention discloses a dual-battery charging control method for electric vehicles based on multi-vector space decoupling. For the dual-battery drive topology of electric vehicles based on a three-phase open-winding permanent magnet synchronous motor, by switching the switch to the charging mode, the open-winding permanent magnet motor winding is reused as a rectifier and filter inductor, the midpoint of each phase winding is connected to the power grid, and at the same time, the negative electrodes of the dual batteries are connected; by using the method of multi-space vector modulation, dq1, dq2, and dq3 spaces are respectively established, the basic vectors with zero mutual projection in each space are selected to synthesize the reference vector, and at the same time, the action time of the zero vector is adjusted to decouple and control the electromagnetic torque, the total power of the dual batteries, and the power difference, so as to realize a dual-battery charging scheme with unity power factor on the grid side and solve the problem of the electromagnetic torque generated by the motor winding under the unbalanced charging power of the two groups of batteries.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle charging integrated topology, and particularly to a dual-battery charging control method for electric vehicles based on multi-vector space decoupling. Background Art

[0002] Limited by the in-vehicle battery capacity, electric vehicles have the problem of insufficient endurance. Charging piles are a solution, but they have a large footprint, high construction costs, and poor portability; on-vehicle chargers do not rely on external devices and are convenient for charging, but they increase the weight of the vehicle and occupy limited space inside the vehicle, which is not conducive to the endurance of electric vehicles; integrated chargers reuse the motor and driver as a rectifier, saving volume and weight, so the integrated charger solution has gradually received attention. Currently, for the integrated charging system of an electric vehicle with an open-winding motor and dual batteries, when the charging powers of the two batteries are inconsistent, there is a problem of electromagnetic torque generated in the motor windings, which has not been well studied and solved. Summary of the Invention

[0003] Object of the Invention: The present invention provides a dual-battery charging control method for electric vehicles based on multi-vector space decoupling, which overcomes the problem of electromagnetic torque generated in the motor during the unbalanced charging process of dual batteries without adding additional devices.

[0004] Technical Solution: A dual-battery charging control method for electric vehicles based on multi-vector space decoupling according to the present invention includes the following steps:

[0005] Step 1: Close the charging switch K, connect the neutral point of the three-phase windings of the open-winding permanent magnet motor to the three-phase power grid, switch the system to the charging mode, and sample the current signals i1, i2, i3, i4, i5, i6 at the two ports of the open-winding permanent magnet motor, the charging currents i dc1 , i dc2 of the dual batteries, the phase angle θ g of the three-phase power grid, and the rotor electrical angle θ r ;

[0006] Step 2: Taking the initial position θ init of the rotor of the open-winding permanent magnet motor as a reference value, combining with the rotor electrical angle θ r , obtain the reference currents i q1_ref , i d1_ref by the position loop P and the speed loop PI, and take them as 0;

[0007] Step 3: Combining the reference values i dc1_ref , i dc2_ref of the charging currents of the two batteries and the actual current feedback values i dc1 , i dc2 , obtain i d2_ref , i q2_ref through the PI regulator.Take 0 to ensure unity power factor control on the grid side, and multiply the difference between i dc2_ref and i dc1_ref by the proportionality coefficient 1 / 6 / a to obtain i α3_ref ;

[0008] Step 4: Combine the dual-port currents i1, i2, i3, i4, i5, i6 of the open-winding permanent magnet motor with the motor rotor position angle θ r and the grid phase angle θ g for analytical calculation to obtain the feedback currents i d1 , i q1 , i d2 , i q2 , i α3 , and respectively obtain the command voltages u d1 , u q1 , u d2 , u q2 , u α3 through the PI regulator according to the reference currents obtained in Steps 2 and 3;

[0009] Step 5: Synthesize the five command voltages u d1 , u q1 , u d2 , u q2 , u α3 with the motor rotor position θ r and the grid phase angle θ g to obtain six-way switching signals according to the multi-space vector demodulation method for driving the dual-three-phase bridge converter;

[0010] Step 6: Repeat Steps 1 to 5 to achieve unity power factor on the grid side and charging control of the dual batteries in the charging mode, and suppress the torque ripple of the open-winding permanent magnet motor with a reused motor winding.

[0011] Furthermore, in Step 4, the mathematical analysis method for the feedback currents i d1 , i q1 , i d2 , i q2 , i α3 is to left-multiply the current matrix I = [i1 i2 i3 i4 i5 i6] T by the transformation matrix T dq and then take the first five elements. The corresponding T dq expression is T dq = T park T αβ , where the T αβ expression is:

[0012]

[0013] T parkThe expression is:

[0014]

[0015] Further, in step 5, the five command voltages u d1 , u q1 , u d2 , u q2 , u α3 are combined with the motor rotor position θ r and the grid phase angle θ g . The six-way switching signals are obtained according to the multi-space vector demodulation method, which specifically includes the following steps:

[0016] Step 51: Establish a multi-vector space and interpret its physical meaning;

[0017] Step 52: Select appropriate basic voltage vectors so that the projection of each space vector on each other is zero to achieve decoupling control;

[0018] Step 53: Divide sectors in each space according to the selected vectors, calculate the action time of the vectors in each sector, and the action time of the required all-1 zero vectors;

[0019] Step 54: Generate the actual pulse width modulation waveform according to the action time of the selected vectors and the corresponding switching states.

[0020] Further, in step 51, the establishment process of the multi-vector space is as follows: The switching state variable s n (n = 1, 2,..., 6) of each phase bridge arm takes 1 when the upper tube is turned on and 0 when the lower tube is turned on. Then all switching states can be expressed as the binary number (s1s2s3s4s5s6)2 and correspond to the voltage vector v = V dc [s1 s2 s3 s4 s5 s6] T , where V dc is the DC side battery voltage;

[0021] The voltage vector matrix v = V dc [s1 s2 s3 s4 s5 s6] T is left-multiplied by T αβ , and the voltage vectors V1, V2, and V3 in the dq1, dq2, and dq3 spaces are obtained by taking the first five rows, and written in vector form. The expressions are as follows:

[0022]

[0023] Based on this, a multi-vector space can be established.

[0024] Further, in step 51, the interpretation of the physical meaning of the multi-vector space is:

[0025] The dq1 space vector is the actual magnetic field vector and is related to torque;

[0026] The dq2 space is interpreted as follows: the voltage applied by the power grid on each winding is left-multiplied by T αβ to obtain the projection of the power grid voltage in multiple spaces, as follows:

[0027]

[0028] It can be obtained therefrom that the power grid voltage is only projected as a rotating vector in the dq2 space. Controlling the synchronous rotation of the current vector in the dq2 space can control the charging power, that is, the dq2 space is related to the charging power;

[0029] The dq3 space is interpreted as follows, Obviously, it is related to the zero-sequence current of the three-phase winding, and the zero-sequence current can be related to the energy flow between the two inverters' upper tubes and the dual batteries, that is, i α3 is in a certain proportional relationship with the power difference of the dual batteries; in addition, the upper tube conduction mode will affect this ratio, and certain all-1 zero vectors need to be inserted to compensate this ratio to a constant, that is, the dq3 space is related to the power difference.

[0030] Furthermore, in step 52, appropriate basic voltage vectors are selected so that the projections of each space vector on each other are zero. Select 110001, 111000, 011100, 001110, 000111, 100011 in the dq1 space, 100100, 110110, 010010, 011011, 001001, 101101 in the dq2 space, and 101010, 010101, 000000 in the dq3 space; the vectors selected in the dq2 and dq3 spaces are projected as zero vectors in other spaces, and the vectors selected in the dq1 space are projected as short vectors with triple frequency in the dq3 space; since the motor does not rotate, the reference vector in the dq1 space is almost zero, so the triple-frequency vector projected by the dq1 space vector in the dq3 space can be ignored, thereby realizing the decoupling of the three spaces.

[0031] Furthermore, in step 53, the sector division process is as follows: both the dq1 and dq2 spaces start from the α-axis and are evenly divided into six sectors at intervals of 60°, and the sectors are numbered from sector 1 to sector 6 in the counterclockwise direction; u d1 、u q1 and θ r are transformed to the αβ coordinate system in the dq1 space through the inverse Park transformation to obtain u α1 、u β1 , and the sector N1 of the torque command voltage is judged according to the formed angle; similarly, u d2 、u q2 and θ gAfter the inverse Park transformation to the αβ coordinate system in the dq2 space, u is obtained α2 , u β2 , and the sector N2 of the suspension command voltage is judged according to the formed angle; the dq3 space is a one-dimensional space, and it can be divided into two sectors according to the positive and negative of u α3 .

[0032] Furthermore, in step 53, calculating the action time of the selected vectors in each sector includes: calculating the action time of the selected basic vectors in different sectors in the dq1 space and the dq2 space according to the general SVPWM algorithm; using the unipolar vector synthesis method to calculate the action time of the basic vectors in the dq3 space; the total action time is t m , and the switching period is T s . When t m ≤T s , no processing is done; when t m >T s , overmodulation processing is performed. Considering that the motor does not rotate and the input and output voltages are fixed during the charging process, the situation of overmodulation will not occur

[0033] Furthermore, in step 53, the action time of all-1 zero vectors in each switching period is aT s -1 / 3*t 21 -2 / 3*t 22 , t 21 is the action time of the vector with two upper switches on among the two basic voltage vectors selected when synthesizing the reference vector in the dq2 space, and t 22 is the action time of another vector with four upper switches on; at this time, the minimum value of the on-time of each phase upper switch is aT s -1 / 3*t 21 -2 / 3*t 22 , and the maximum value is aT s +2 / 3*t 21 +1 / 3*t 22 , and the on-time range is [0, T s . From this, the value range of a is calculated as [V g / V dc , 1 - V g / V dc , where V g is the maximum value of the grid voltage, and V dc is the DC side battery voltage, that is, the relationship compensation between the double battery current difference and i α3 is linear, and the expression is: i α3 =(i dc2 -i dc1 ) / 6 / a

[0034] Further, in step 54, the generation process of the actual pulse width modulation waveform is as follows: within one switching period, the time of each vector acting on each phase winding is summed to obtain the duty cycle of each phase, and the PWM waveform is modulated directly with the carrier according to the obtained duty cycle; finally, the volt-second product of each phase voltage remains unchanged and can be regarded as equivalent.

[0035] Beneficial effects: Compared with the prior art, the present invention has the following remarkable advantages: The present invention regards the open-winding three-phase motor as a six-phase motor, introduces the multi-vector space concept of the multi-phase motor, and establishes a physical connection between each space and the electromagnetic torque and charging power of the dual-battery charging topology to achieve decoupling control, overcoming the problem of generating electromagnetic torque during unbalanced charging of the dual batteries; in addition, the method has strong versatility and can be extended to other charging topologies or other types of motors. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is the topological structure diagram of the present invention.

[0037] Figure 2 is the dq1 vector space diagram of the present invention.

[0038] Figure 3 is the dq2 vector space diagram of the present invention.

[0039] Figure 4 is the dq3 vector space diagram of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0040] As Figure 1 shown, a dual-battery charging control method for electric vehicles based on multi-vector space decoupling includes the following steps:

[0041] Step 1: Close the charging switch K, connect the midpoint of the three-phase winding of the open-winding permanent magnet motor to the three-phase power grid, switch the system to the charging mode, and sample the current signals i1, i2, i3, i4, i5, i6 at the two ports of the open-winding permanent magnet motor, the charging currents i dc1 , i dc2 of the dual batteries, the phase angle θ g of the three-phase power grid, and the rotor electrical angle θ r ;

[0042] Step 2: Using the initial position θ init of the rotor of the open-winding permanent magnet motor as a reference value, combined with the rotor electrical angle θ r , the reference currents i q1_ref , i d1_ref are obtained as 0 through the position loop P and the speed loop PI;

[0043] Step 3: Combining the reference values i dc1_ref , i dc2_refand the actual current feedback value i dc1 、i dc2 , and through the PI regulator, i d2_ref is obtained. i q2_ref takes 0 to ensure unity power factor control on the grid side. Multiply the difference between i dc2_ref and i dc1_ref by the proportionality coefficient 1 / 6 / a to obtain i α3_ref ;

[0044] Step 4: Combine the open-winding permanent magnet motor's dual-port currents i1, i2, i3, i4, i5, i6 with the motor rotor position angle θ r and the grid phase angle θ g for analytical calculation to obtain the feedback currents i d1 , i q1 , i d2 , i q2 , i α3 , and respectively obtain the command voltages u d1 , u q1 , u d2 , u q2 , u α3 through the PI regulator according to the reference currents obtained in Steps 2 and 3;

[0045] Step 5: Synthesize the five command voltages u d1 , u q1 , u d2 , u q2 , u α3 with the motor rotor position θ r and the grid phase angle θ g , and obtain six switching signals according to the multi-space vector demodulation method to drive the dual three-phase bridge converter;

[0046] Step 6: Repeat Steps 1 to 5 to achieve unity power factor control on the grid side and charging control of the dual batteries in the charging mode, and suppress the torque ripple of the open-winding permanent magnet motor with the reused motor windings.

[0047] In the charging topology, the three-phase motor is reconstructed into a symmetric six-phase motor. The dual batteries are respectively connected to six ports through two three-phase half-bridges. The current is taken as positive when flowing into the motor. The A, -A, B, -B, C, -C phases correspond to the numbers 1, 4, 3, 6, 5, 2 respectively; the battery charging current is taken as positive when flowing into the battery; the voltages of the two batteries are at the same level, that is, V dc1 = V dc2 ; the rotor electrical angle θ r is obtained through the relevant angle sensor; the three-phase grid phase angle θ g is obtained through the three-phase voltages V ga , V gb , V gcObtained by PLL calculation; V ga V gb V gc Obtained through a voltage Hall sensor, with the maximum value being V g ; All currents are obtained through current Hall sensors.

[0048] In step 4, the mathematical analysis method for the feedback currents i d1 i q1 i d2 i q2 i α3 is to left-multiply the current matrix I = [i1 i2 i3 i4 i5 i6] T by the transformation matrix T dq and then take the first five elements. The corresponding T dq expression is T dq = T park T αβ where the T αβ expression is:

[0049]

[0050] The T park expression is:

[0051]

[0052] The specific modulation algorithm in step 5 includes the following sub-steps:

[0053] Step 51: Establish a multi-vector space and interpret its physical meaning;

[0054] Step 52: Select appropriate basic voltage vectors to make the projection of each space vector zero to achieve decoupling control;

[0055] Step 53: Divide sectors in each space according to the selected vectors, and calculate the action time of the vectors in each sector and the action time of the required all-1 zero vectors;

[0056] Step 54: Generate the actual pulse width modulation waveform according to the action time of the selected vectors and the corresponding switching states.

[0057] In step 51, the establishment process of the multi-vector space is as follows: For the switching state variables s n (n = 1, 2,..., 6) of each phase bridge arm, take 1 when the upper transistor is turned on and 0 when the lower transistor is turned on. Then all switching states can be expressed as the binary number of (s1 s2 s3 s4 s5 s6)2, and correspond to the voltage vector v = V dc [s1 s2 s3 s4 s5 s6] T where V dc is the DC side battery voltage;

[0058] The voltage vector matrix v = V dc [s1 s2 s3 s4 s5 s6] T is left - multiplied by T αβ and then the first five rows are taken to obtain the voltage vectors V1, V2, and V3 in the dq1, dq2, and dq3 spaces. Writing them in vector form, the expressions are as follows:

[0059]

[0060] Based on this, a multi - vector space can be established. The expanded vector space is as shown in Figure 2 、 3 、4. The dq1 and dq2 spaces are the same. The magnitude of the large vector is 2*V dc / 3, the magnitude of the medium vector is V dc / sqrt(3), and the magnitude of the small vector is V dc / 3; in the dq3 space, the magnitude of the large vector is V dc , the magnitude of the medium vector is 2*V dc / 3, and the magnitude of the small vector is V dc / 3;

[0061] In step 51, the physical meaning interpretation of the multi - vector space is as follows:

[0062] The vector in the dq1 space is the actual magnetic field vector and is related to torque;

[0063] For the dq2 space, the voltage applied by the power grid on each winding is left - multiplied by T αβ to obtain the projection of the power grid voltage in the multi - space, as follows:

[0064]

[0065] From this, it can be obtained that the power grid voltage is only projected as a rotating vector in the dq2 space. Controlling the synchronous rotation of the current vector in the dq2 space can control the charging power, that is, the dq2 space is related to the charging power;

[0066] For the dq3 space, the interpretation is as follows Obviously, it is related to the zero - sequence current of the three - phase windings, and the zero - sequence current can establish a connection with the energy flow between the two batteries through the conduction of the upper switches of the two inverters, that is, i α3 is in a certain proportional relationship with the power difference between the two batteries; in addition, the conduction mode of the upper switches will affect this ratio, and a certain all - 1 zero vector needs to be inserted to compensate this ratio to a constant, that is, the dq3 space is related to the power difference.

[0067] In step 52, appropriate basic voltage vectors are selected to make the projections of all space vectors on each other zero. In the dq1 space, vectors 110001, 111000, 011100, 001110, 000111, 100011 are selected; in the dq2 space, vectors 100100, 110110, 010010, 011011, 001001, 101101 are selected; in the dq3 space, vectors 101010, 010101, 000000 are selected. The vectors selected in the dq2 and dq3 spaces project to zero vectors in other spaces, and the vectors selected in the dq1 space project to short vectors with triple frequency in the dq3 space. Since the motor is not rotating, the reference vector in the dq1 space is almost zero. Therefore, the triple-frequency vectors projected by the vectors in the dq1 space in the dq3 space can be ignored, thus achieving decoupling in the three spaces.

[0068] In step 53, the sector division process is as follows: In both the dq1 and dq2 spaces, starting from the α-axis, each is evenly divided into six sectors with an interval of 60°, and the sectors are numbered from sector 1 to sector 6 in the counterclockwise direction; u d1 and u q1 and θ r are transformed to the αβ coordinate system in the dq1 space through the inverse Park transformation to obtain u α1 and u β1 , and the sector N1 of the torque command voltage is judged according to the formed angle. Similarly, u d2 and u q2 and θ g are transformed to the αβ coordinate system in the dq2 space through the inverse Park transformation to obtain u α2 and u β2 , and the sector N2 of the suspension command voltage is judged according to the formed angle. The dq3 space is a one-dimensional space, and it can be divided into two sectors according to the positive and negative of u α3 .

[0069] In step 53, calculating the action time of the selected vectors in each sector includes: In the dq1 and dq2 spaces, the action time of the selected basic vectors in their respective different sectors is calculated according to the general SVPWM algorithm; in the dq3 space, the single-polarity vector synthesis method is used to calculate the action time of the basic vectors; the total action time is t m , the switching period is T s , when t m ≤T s , no processing is done; when t m >T s , overmodulation processing is performed. Considering that the motor does not rotate and the input and output voltages are fixed during the charging process, the situation of overmodulation will not occur.

[0070] Let

[0071]

[0072] Then t 11 and t 12 The sizes under each sector are shown in Table 1:

[0073] Table 1 t 11 and t 12 The size table under each sector

[0074] <![CDATA[Sector N1]]> 1 2 3 4 5 6 <![CDATA[t 11 > -Z Z X -X -Y Y <![CDATA[t 12 > X Y -Y Z -Z -X

[0075] t 21 and t 22 The sizes under each sector are shown in Table 2:

[0076] Table 2 t 21 and t 22 The size table under each sector

[0077]

[0078]

[0079] In the dq3 space, the unipolar vector synthesis method is used to calculate the action time of the basic vectors. The action time of vector 101010 is denoted as t 31 , and the action time of vector 010101 is denoted as t 32 ; t 31 and t 32 The sizes under each sector are shown in Table 3:

[0080] Table 3 t 31 and t 32 The size table under each sector

[0081] <![CDATA[Sector N3]]> 1 2 <![CDATA[t 31 > <![CDATA[u α3 T s / V dc > 0 <![CDATA[t 32 > 0 <![CDATA[u α3 T s / V dc >

[0082] The total action time is t m = t 11 + t 12 + t 21 + t 22 + t 31 + t 32 , and the switching period is T s , when t m ≤ T s , no processing is done; when t m > T s , overmodulation processing is performed, and it is scaled proportionally to t m = T s . Considering that the motor does not rotate during the charging process and the input and output voltages are fixed, overmodulation generally does not occur.

[0083] The action time of the all-1 zero vector in each switching period is aT s-1 / 3*t 21 -2 / 3*t 22 , at this time, the minimum value of the conduction time of each phase upper tube is aT s -1 / 3*t 21 -2 / 3*t 22 , the maximum value is aT s +2 / 3*t 21 +1 / 3*t 22 , and the conduction time range is [0, T s , from which the value range of a is calculated as [V g / V dc , 1 - V g / V dc , where V g is the maximum value of the grid voltage, V dc is the battery voltage on the DC side, that is, the relationship compensation between the double - battery current difference and i α3 is linear, and the expression is: i α3 =(i dc2 - i dc1 ) / 6 / a.

[0084] The generation process of the actual pulse - width modulation waveform in step 54 is as follows: within one switching period, sum the time when each selected vector acts on each phase winding to obtain the duty cycle of each phase, and directly modulate the PWM waveform with the carrier according to the obtained duty cycle; finally, the volt - second product of each phase voltage remains unchanged and can be regarded as equivalent. Taking the reference vectors in sector 1 of dq1 space, sector 1 of dq2 space, and sector 1 of dq3 space as examples, the duty cycle of each phase is:

[0085]

Claims

1. A dual-battery charging control method for electric vehicles based on multi-vector spatial decoupling, characterized in that: The steps include: Step 1: Close the charging switch K, connect the three-phase grid to the midpoint of the three-phase winding of the open-winding permanent magnet motor, switch the system to the charging mode, and sample the current signals i1, i2, i3, i4, i5, and i6 at the dual ports of the open-winding permanent magnet motor. The charging current i dc1 、i dc2 , three-phase grid phase angle θ g and the rotor electrical angle θ r ; Step 2: Take the initial rotor position θ of the open-winding permanent magnet motor init As a reference value, combined with the rotor electrical angle θ r , the reference current i is obtained through the position loop P and the speed loop PI q1_ref ,i d1_ref Take 0; Step 3: Combine the reference values ​​of the charging current of the two batteries dc1_ref 、i dc2_ref and the actual current feedback value i dc1 、i dc2 , through the PI regulator to get i d2_ref ,i q2_ref Take 0 to ensure the unity power factor control on the grid side, and set i dc2_ref and i dc1_ref The difference is multiplied by the proportionality factor 1 / 6 / a to get i α3_ref ; Step 4: Combine the open-winding permanent magnet motor dual-port currents i1, i2, i3, i4, i5, and i6 with the motor rotor position angle θ r and the grid phase angle θ g The feedback current i is obtained by analytical calculation d1 、i q1 、i d2 、i q2 、i α3 , and according to the reference current obtained in steps 2 and 3, the command voltage u is obtained through the PI regulator respectively. d1 、u q1 、u d2 、u q2 、u α3 ; Step 5: Set the five command voltages u d1 、u q1 、u d2 、u q2 、u α3 and the motor rotor position θ r and the grid phase angle θ g Phase synthesis, according to the multi-space vector demodulation method, obtains six-way switching signals to drive the dual three-phase bridge converter; Step 6: Repeat steps 1 to 5 to achieve grid-side unity power factor and dual battery charging control in charging mode, and suppress torque pulsation of the open-winding permanent magnet motor of the reused motor winding.

2. The electric vehicle dual battery charging control method based on multi-vector space decoupling as claimed in claim 1 is characterized in that: In step 4, the feedback current i d1 、i q1 、i d2 、i q2 、i α3 The mathematical analysis method is to convert the current matrix I = [i1 i2i3 i4i5 i6] T Left multiply the transformation matrix T dq Then take the first five elements, corresponding to T dq The expression is T dq =T park T αβ , where T αβ The expression is: T park The expression is:

3. The electric vehicle dual battery charging control method based on multi-vector space decoupling as claimed in claim 1 is characterized in that: In step 5, the five command voltages u d1 、u q1 、u d2 、u q2 、u α3 and the motor rotor position θ r and the grid phase angle θ g The six-way switch signals are obtained by multi-space vector demodulation method, which specifically includes the following steps: Step 51, establish a multi-vector space and interpret the physical meaning; Step 52, selecting a suitable basic voltage vector so that the space vectors are mutually projected to zero, so as to realize decoupling control; Step 53, divide the space into sectors according to the selected vector, and calculate the action time of the vector in each sector, as well as the action time of the required all-1 zero vector; Step 54: Generate an actual pulse width modulation waveform according to the action time of the selected vector and the corresponding switch state.

4. The electric vehicle dual battery charging control method based on multi-vector space decoupling as claimed in claim 3 is characterized in that: In step 51, the process of establishing the multi-vector space is as follows: the switch state variable s of each phase bridge arm n (n=1,2,…,6), when the upper tube is turned on, it takes 1, and when the lower tube is turned on, it takes 0. Then all switch states are represented by a binary number of (s1s2s3s4s5s6)2, and the corresponding voltage vector v=V dc [s1 s2 s3 s4 s5 s6] T , where V dc is the DC side battery voltage; Voltage vector matrix v = V dc [s1 s2 s3 s4 s5 s6] T Left multiply T αβ Then take the first five rows to get the voltage vectors V1, V2, and V3 in the dq1, dq2, and dq3 spaces, and write them in vector form. The expression is as follows: Based on this, a multi-vector space is established.

5. The electric vehicle dual battery charging control method based on multi-vector space decoupling as claimed in claim 3 is characterized in that: In step 51, the physical meaning of the multi-vector space is interpreted as: The dq1 space vector is the actual magnetic field vector, which is related to the torque; The dq2 space is interpreted as follows: the voltage of the power grid on each winding is multiplied by T on the left. αβ The projection of the grid voltage in multiple spaces is obtained as follows: The grid voltage is only projected as a rotating vector in the dq2 space. The charging power can be controlled by controlling the synchronous rotation of the current vector in the dq2 space, that is, the dq2 space is related to the charging power. The dq3 space is interpreted as follows, It is related to the zero-sequence current of the three-phase winding, and the zero-sequence current is connected with the energy flow between the dual batteries through the conduction of the upper tubes of the inverters on both sides, that is, i α3 It is in a certain proportional relationship with the power difference of the dual batteries; in addition, the conduction mode of the upper tube will affect the ratio, and a certain all-1 zero vector needs to be inserted to compensate the ratio to a constant, that is, the dq3 space is related to the power difference.

6. The electric vehicle dual battery charging control method based on multi-vector space decoupling as claimed in claim 3 is characterized in that: In step 52, a suitable basic voltage vector is selected so that each space vector is projected to zero with respect to each other. 110001, 111000, 011100, 001110, 000111, 100011 are selected in the dq1 space; 100100, 110110, 010010, 011011, 001001, 101101 are selected in the dq2 space; 101010, 010101, 000000 are selected in the dq3 space; the vectors selected in the dq2 and dq3 spaces are projected as zero vectors in other spaces, and the vector selected in the dq1 space is projected as a short vector with three times the frequency in the dq3 space; since the motor does not rotate, the reference vector in the dq1 space is almost zero, and therefore the triple frequency vector projected by the dq1 space vector in the dq3 space can be ignored, thereby achieving decoupling of the three spaces.

7. The electric vehicle dual battery charging control method based on multi-vector space decoupling as claimed in claim 3 is characterized in that: In step 53, the sector division process is as follows: both the dq1 and dq2 spaces are divided into six sectors with the α axis as the starting point and 60° as the interval, and are numbered from sector 1 to sector 6 in a counterclockwise direction; d1 、u q1 With θ r After the inverse Park transformation, we can get u α1 、u β1 , determine the sector N1 of the torque command voltage according to the angle; similarly, u d2 、u q2 With θ g After the inverse Park transformation, it is transformed into the dq2 space αβ coordinate system to obtain u α2 、u β2 , according to the angle, determine the sector N2 of the suspension command voltage; dq3 space is a one-dimensional space, according to u α3 The positive and negative are divided into two sectors.

8. The electric vehicle dual battery charging control method based on multi-vector space decoupling as claimed in claim 3 is characterized in that: In step 53, calculating the action time of the selected vector for each sector includes: calculating the action time of the selected basic vector in each sector in dq1 space and dq2 space according to the SVPWM algorithm; using the unipolar vector synthesis method to calculate the action time of the basic vector in dq3 space; the total action time is t m , the switching period is T s , when t m ≤T s No processing is done when t m >T s Overmodulation is performed during the charging process. Considering that the motor does not rotate during the charging process and the input and output voltages are fixed, overmodulation will not occur.

9. The electric vehicle dual battery charging control method based on multi-vector space decoupling as claimed in claim 3 is characterized in that: In step 53, the action time of the full 1 zero vector in each switching cycle is aT s -1 / 3*t 21 -2 / 3*t 22 , t 21 The action time of two upper tubes in the two basic voltage vectors selected when synthesizing the reference vector for dq2 space, t 22 is the action time of another vector with four upper tubes turned on; at this time, the minimum opening time of each phase upper tube is aT s -1 / 3*t 21 -2 / 3*t 22 , the maximum value is aT s +2 / 3*t 21 +1 / 3*t 22 , and the opening time range [0,T s ], and thus the value range of a is calculated [V g / V dc ,1-V g / V dc ], where V g is the maximum grid voltage, V dc is the DC side battery voltage, that is, the current difference between the two batteries and i α3 The relationship compensation is linear, and the expression is: α3 =(i dc2 -i dc1 ) / 6 / a.

10. The electric vehicle dual battery charging control method based on multi-vector space decoupling as claimed in claim 3, characterized in that: In step 54, the actual pulse width modulation waveform generation process is: within a switching cycle, the time that each vector acts on each phase winding is summed to obtain the duty cycle of each phase, and the PWM waveform is directly modulated with the carrier according to the obtained duty cycle; finally, the volt-second product of each phase voltage remains unchanged and is regarded as equivalent.

Images