Control method and system applied to current source type inverter permanent magnet electric drive system

Abstract

The application discloses a control method and system applied to a current source type inverter permanent magnet electric drive system, and the method comprises the following steps: obtaining currents and voltages in a dq space coordinate system in a current control period; establishing a state space model of the current source type electric drive system; obtaining a state variable prediction equation of the system; simultaneously controlling filter capacitor voltages and stator currents based on the state variable prediction equation of the system and considering all state variables of the system, and designing an evaluation function; calculating optimized control inputs, obtaining a simplified evaluation function based on the optimized control inputs; taking a current vector closest to the optimized control inputs; and converting the selected current vector into a symmetrical pulse sequence to be applied to the current source type inverter; the application has the advantages that resonance is avoided, calculation is reduced, and operation power consumption and cost are greatly reduced.

Description

Technical Field

[0001] This invention relates to the field of three-phase AC motor control system design and manufacturing, and more specifically to a control method and system applied to a current source inverter permanent magnet electric drive system. Background Technology

[0002] Compared with traditional multi-level voltage source inverters, using current source inverters to drive and control load motors has the following advantages: 1) Current source inverters have a simple topology, fewer components, and DC side short-circuit overcurrent protection; 2) Current source inverters are topologies with boost capability; 3) The output filter voltage in a current source inverter can effectively reduce the instantaneous rate of change of the motor's stator voltage, making the resulting stator voltage waveform closer to a sine wave, thereby reducing high-frequency noise and improving system efficiency, which is highly beneficial to extending the motor's service life.

[0003] Although current-source inverters can achieve a more sinusoidal motor stator voltage waveform compared to traditional voltage-source inverters, current-source electric drive systems are essentially second-order systems composed of filter capacitors and motor stator inductance. LC resonance in high-order systems can easily induce instability, resulting in significant oscillations in the motor's voltage and current waveforms. Therefore, suppressing resonance is a problem that needs to be addressed in the application of current-source electric drive systems. Traditional active damping schemes are generally based on additional state variable feedback, introducing appropriate virtual damping into the closed-loop transfer function of the system to suppress resonance phenomena. For example, the real-time feedback active damping method for improving the robustness of grid-connected inverters disclosed in "Pan Donghua, Ruan Xinbo, Wang Xuehua, et al. A Capacitor-current real-time feedback active damping method for improving robustness of the LCL-type grid-connected inverter[J]. Proceedings of the CSEE, 2013, 33(18): 1-10(in Chinese)" is an example of such a method.

[0004] Essentially, traditional control schemes decouple multi-input multi-output systems, employing multiple single-input single-output loops for control. Therefore, traditional schemes are complex, utilize multi-loop control mechanisms, and involve cumbersome parameter adjustments that fail to meet all motor operating point requirements. Furthermore, traditional control methods suffer from limited control bandwidth and slow motor dynamic response due to the cascaded control structure. Compared to traditional space-field-oriented vector control algorithms, Finite-Control-Set Model Predictive Control (FCS-MPC) is a more advanced control algorithm proposed in recent years. The most significant feature of FCS-MPC is its integration of the inverter's switching sequence decision-making process into the controller. It's a strategy that directly selects the optimal control switching state online, eliminating the need for an additional modulator. Due to its intuitive design, ease of handling constraints and control objectives with nonlinear characteristics, and flexible multivariable control capabilities, FCS-MPC has become an important research branch in the field of power electronics and electric drive control algorithms in recent years.

[0005] Therefore, leveraging the advantages of the FCS-MPC scheme to solve the resonance problem in current-source inverter drive systems can eliminate the multi-loop control structure in traditional algorithms, thereby simplifying parameter design and improving system bandwidth and response speed. The typical design steps of FCS-MPC are: first, determine the available finite set of switching sequences based on the converter type; then, design an evaluation function to constrain motor performance; next, calculate the corresponding evaluation function value for each current vector in the control set; and finally, select the switching state sequence with the minimum evaluation function value and directly apply it to the inverter to drive the motor. However, in the design of the evaluation function, it is necessary to consider how to solve the resonance problem caused by the filter capacitor and the motor stator inductance. Simultaneously, to minimize the computational load of the FCS-MPC algorithm in the digital controller, the steps of the traditional FCS-MPC algorithm need to be improved to avoid a large number of repetitive prediction calculations, thereby reducing the computational load during digital implementation and lowering operating power consumption and cost. Summary of the Invention

[0006] The technical problem to be solved by this invention is that the existing finite set model predictive control method for permanent magnet electric drive systems has a resonance problem caused by the filter capacitor and the stator inductance of the motor. At the same time, the algorithm has a large amount of computation, high power consumption and high cost during digital implementation.

[0007] This invention solves the above-mentioned technical problems through the following technical means: a control method applied to a permanent magnet electric drive system of a current source inverter, the method comprising:

[0008] Step 1: Obtain the current and voltage in the dq space coordinate system of the current control cycle;

[0009] Step 2: Select the filter capacitor voltage as the state variable and establish the state-space model of the current source type electric drive system;

[0010] Step 3: Discretize the state-space equations in the continuous time domain to obtain the state variable prediction equations of the system;

[0011] Step 4: Based on the system's state variable prediction equations and considering all system state variables, simultaneously control the filter capacitor voltage and stator current, and design an evaluation function;

[0012] Step 5: Calculate the reference value of the filter capacitor in the evaluation function;

[0013] Step 6: Calculate the optimized control input, and based on the optimized control input, obtain the simplified evaluation function;

[0014] Step 7: Calculate the current vector of the current source inverter in the stationary αβ coordinate system, and select the current vector that is closest to the optimized control input;

[0015] Step 8: Convert the selected current vector into a symmetrical pulse sequence and apply it to the current source inverter.

[0016] Beneficial effects: This invention uses an evaluation function to directly constrain all state variables of the system. By simultaneously controlling the filter capacitor voltage and stator current, it can be ensured that they can track the reference setpoint well, thereby avoiding resonance. An optimized control input is calculated in advance. Based on this optimized control input, a simplified evaluation function is obtained, and this simplified evaluation function is used to select the final given current vector. This reduces repetitive prediction calculations and the computational load during digital implementation, significantly reducing operating power consumption and cost.

[0017] Further, step 1 includes:

[0018] The stator currents of phase a and phase b of the motor in the current control cycle are sampled, and the current in the dq space coordinate system is obtained through Parker transformation. The specific calculation process is as follows:

[0019] i sc =-i sa -i sb (1)

[0020]

[0021] Among them, i sa i sb and i scThese represent the three-phase stator currents a, b, and c, respectively; i sd and i sq Let θ represent the stator current in the dq coordinate system, respectively. e Indicates angle;

[0022] The stator terminal voltages of phase a and phase b of the motor in the current control cycle are sampled and then transformed by Parker transformation to obtain the voltages in the dq spatial coordinate system. The specific calculation process is as follows:

[0023] v sc =-v sa -v sb (3)

[0024]

[0025] Among them, v sa ,v sb and v sc These represent the three-phase stator voltages a, b, and c, respectively; v sd and v sq These represent the stator voltages in the dq spatial coordinate system, respectively.

[0026] Furthermore, step 2 includes:

[0027] In the synchronously rotating dq coordinate system, the stator voltage state-space equation of the permanent magnet synchronous motor is established as follows:

[0028]

[0029] Among them, R s ω represents the stator resistance of the motor; e ψ represents the angular velocity of the motor stator side; sd and ψ sq Let denot dq represent the stator flux linkage in the dq spatial coordinate system, and their calculation equations are as follows:

[0030]

[0031] L s This represents the stator-side inductance of the motor. For surface-mounted permanent magnet synchronous motors, the d-axis and q-axis inductances are equal; ψ f Indicates the size of the flux linkage of the rotor permanent magnet;

[0032] For a current-source inverter, the filter capacitor voltage is selected as the state variable, and its state-space equations are established according to Kirchhoff's current law as follows:

[0033]

[0034] Among them, i wd and i wqC represents the current on the inverter side of the current source in the dq spatial coordinate system; f The size of the filter capacitor is represented by equations (5) and (7). The continuous-time state-space model of the current-source inverter drive system is expressed as follows:

[0035]

[0036] Wherein, the state space vector x = [i sd i sq v sd v sq ] T , control input vector u = [i wd i wq ] T The specific expressions for the system matrix A, input matrix B, and interference vector E are as follows:

[0037]

[0038] Furthermore, step 3 includes:

[0039] Discretizing the state-space equations in the continuous-time domain yields the following prediction equations for the system's state variables:

[0040] x(k+1)=Gx(k)+Hu(k)+F (10)

[0041] Where G = e ATs e represents the natural index, T s H represents the sampling control period interval; H = (GI)A -1 B,F=(GI)A -1 E, I i Is with G i Identity matrices with the same dimensions;

[0042] The calculation process for the delay compensation of one beat ahead for the current sampled variable is as follows:

[0043] x c (k+1)=Gx(k)+Hu i +F (11)

[0044] Where x c (k+1) represents the state variable after delay compensation at time k+1, u i This represents the control input of the current source inverter at the current moment. Therefore, based on the compensated state variables, controlling the predicted value at time k+2 can overcome the one-step delay phenomenon of the digital system. The predicted value at time k+2 is calculated as follows:

[0045] x(k+2)=Gxc (k+1)+Hu+F (12).

[0046] Furthermore, step 4 includes:

[0047] The evaluation function designed for current-source inverter electric drive systems is as follows:

[0048] J = [x * (k+2)-x(k+2)] T W[x * (k+2)-x(k+2)] (13)

[0049] Where x * (k+2)=[i * sd (k+2)i * sq (k+2)v * sd (k+2)v * sq (k+2)] T For reference value vectors; x(k+2)=[i sd (k+2)i sq (k+2)v sd (k+2)v sq (k+2)] T Let W be the prediction vector of the state variables; W is the weight coefficient matrix, and its specific expression is as follows:

[0050]

[0051] Furthermore, step 5 includes:

[0052] For permanent magnet synchronous motors, the q-axis current is referenced to the given value i. * sq Derived from the outer speed loop, the d-axis current references the given value i. * sd To achieve maximum torque-to-current ratio control, the reference setpoint of the filter capacitor is calculated based on the stator current reference value of the dq axis and the system's state-space model, as follows:

[0053]

[0054] Furthermore, step 6 includes:

[0055] Substituting equation (11) into (13), and differentiating (13) with respect to the control input variable, setting the derivative to zero, i.e.

[0056]

[0057] Then, the optimized control input is obtained, and its expression is as follows:

[0058] u * =(H T WH) -1 ·H T W·[x * -Gx c -F] (17)

[0059] Based on this optimized control input, the simplified evaluation function is designed as follows:

[0060] J 简化 =[u * (k+2)-u(k+2)] T [u * (k+2)-u(k+2)] (18).

[0061] Furthermore, step 7 includes:

[0062] A current-source inverter includes six active current vectors corresponding to switching states and three zero current vectors corresponding to switching states. The current vector of the current-source inverter is defined as i based on the switching states. wi =[S 1i S 3i S 5i S 4i S 6i S 2i ], '1' indicates that the corresponding switch state is on, and '0' indicates that the corresponding switch state is off. The current vector is represented in the stationary αβ coordinate system as follows:

[0063]

[0064] Where χ=exp(j2π / 3),i wia =(S 1i -S 4i )i dc i wib =(S 3i -S 6i )i dc i wic =(S 5i -S 2i )i dc The spatial coordinate system is divided into 6 sectors and the angle of each sector is π / 3. Each possible discrete control input variable u is substituted into the evaluation function (18) and the current vector that is closest to the control input of the distance optimization is selected.

[0065] This invention also provides a control system for a permanent magnet electric drive system for a current source inverter, the system comprising:

[0066] The current and voltage acquisition module is used to acquire the current and voltage in the dq space coordinate system of the current control cycle;

[0067] The model building module is used to select the filter capacitor voltage as the state variable and establish the state space model of the current source type electric drive system.

[0068] The prediction equation construction module is used to discretize the state-space equations in the continuous time domain to obtain the prediction equations of the system's state variables.

[0069] The evaluation function construction module is used to design evaluation functions by simultaneously controlling the filter capacitor voltage and stator current based on the system's state variable prediction equation and considering all system state variables.

[0070] The reference given value calculation module is used to calculate the reference given value of the filter capacitor in the evaluation function;

[0071] The function simplification module is used to calculate the optimized control input, and based on the optimized control input, a simplified evaluation function is obtained;

[0072] The vector optimization module is used to calculate the current vector of the current source inverter in the stationary αβ coordinate system and select the current vector that is closest to the optimized control input.

[0073] The result output module is used to convert the selected current vector into a symmetrical pulse sequence and apply it to the current source inverter.

[0074] Furthermore, the current and voltage acquisition module is also used for:

[0075] The stator currents of phase a and phase b of the motor in the current control cycle are sampled, and the current in the dq space coordinate system is obtained through Parker transformation. The specific calculation process is as follows:

[0076] i sc =-i sa -i sb (1)

[0077]

[0078] Among them, i sa i sb and i sc These represent the three-phase stator currents a, b, and c, respectively; i sd and i sq Let θ represent the stator current in the dq coordinate system, respectively. e Indicates angle;

[0079] The stator terminal voltages of phase a and phase b of the motor in the current control cycle are sampled and then transformed by Parker transformation to obtain the voltages in the dq spatial coordinate system. The specific calculation process is as follows:

[0080] v sc =-v sa -v sb (3)

[0081]

[0082] Among them, v sa ,v sb and v sc These represent the three-phase stator voltages a, b, and c, respectively; v sd and v sq These represent the stator voltages in the dq spatial coordinate system, respectively.

[0083] Furthermore, the model building module is also used for:

[0084] In the synchronously rotating dq coordinate system, the stator voltage state-space equation of the permanent magnet synchronous motor is established as follows:

[0085]

[0086] Among them, R s ω represents the stator resistance of the motor; e ψ represents the angular velocity of the motor stator side; sd and ψ sq Let denot dq represent the stator flux linkage in the dq spatial coordinate system, and their calculation equations are as follows:

[0087]

[0088] L s This represents the stator-side inductance of the motor. For surface-mounted permanent magnet synchronous motors, the d-axis and q-axis inductances are equal; ψ f Indicates the size of the flux linkage of the rotor permanent magnet;

[0089] For a current-source inverter, the filter capacitor voltage is selected as the state variable, and its state-space equations are established according to Kirchhoff's current law as follows:

[0090]

[0091] Among them, i wd and i wq C represents the current on the inverter side of the current source in the dq spatial coordinate system; f The size of the filter capacitor is represented by equations (5) and (7). The continuous-time state-space model of the current-source inverter drive system is expressed as follows:

[0092]

[0093] Wherein, the state space vector x = [i sd i sq v sd v sq ] T , control input vector u = [i wd i wq ] T The specific expressions for the system matrix A, input matrix B, and interference vector E are as follows:

[0094]

[0095] Furthermore, the prediction equation construction module is also used for:

[0096] Discretizing the state-space equations in the continuous-time domain yields the following prediction equations for the system's state variables:

[0097] x(k+1)=Gx(k)+Hu(k)+F (10)

[0098] Where G = e ATs e represents the natural index, T s H represents the sampling control period interval; H = (GI)A -1 B,F=(GI)A -1 E, I i Is with G i Identity matrices with the same dimensions;

[0099] The calculation process for the delay compensation of one beat ahead for the current sampled variable is as follows:

[0100] x c (k+1)=Gx(k)+Hu i +F (11)

[0101] Where x c (k+1) represents the state variable after delay compensation at time k+1, u i This represents the control input of the current source inverter at the current moment. Therefore, based on the compensated state variables, controlling the predicted value at time k+2 can overcome the one-step delay phenomenon of the digital system. The predicted value at time k+2 is calculated as follows:

[0102] x(k+2)=Gx c (k+1)+Hu+F (12).

[0103] Furthermore, the evaluation function construction module is also used for:

[0104] The evaluation function designed for current-source inverter electric drive systems is as follows:

[0105] J = [x * (k+2)-x(k+2)] T W[x * (k+2)-x(k+2)] (13)

[0106] Where x * (k+2)=[i * sd (k+2)i * sq (k+2)v * sd (k+2)v * sq (k+2)] T For reference value vectors; x(k+2)=[i sd (k+2)i sq (k+2)v sd (k+2)v sq (k+2)] T Let W be the prediction vector of the state variables; W is the weight coefficient matrix, and its specific expression is as follows:

[0107]

[0108] Furthermore, the reference given value calculation module is also used for:

[0109] For permanent magnet synchronous motors, the q-axis current is referenced to the given value i. * sq Derived from the outer speed loop, the d-axis current references the given value i. * sd To achieve maximum torque-to-current ratio control, the reference setpoint of the filter capacitor is calculated based on the stator current reference value of the dq axis and the system's state-space model, as follows:

[0110]

[0111] Furthermore, the function simplification module is also used for:

[0112] Substituting equation (11) into (13), and differentiating (13) with respect to the control input variable, setting the derivative to zero, i.e.

[0113]

[0114] Then, the optimized control input is obtained, and its expression is as follows:

[0115] u * =(H TWH) -1 ·H T W·[x * -Gx c -F] (17)

[0116] Based on this optimized control input, the simplified evaluation function is designed as follows:

[0117] J 简化 =[u * (k+2)-u(k+2)] T [u * (k+2)-u(k+2)] (18).

[0118] Furthermore, the vector optimization module is also used for:

[0119] A current-source inverter includes six active current vectors corresponding to switching states and three zero current vectors corresponding to switching states. The current vector of the current-source inverter is defined as i based on the switching states. wi =[S 1i S 3i S 5i S 4i S 6i S 2i ], '1' indicates that the corresponding switch state is on, and '0' indicates that the corresponding switch state is off. The current vector is represented in the stationary αβ coordinate system as follows:

[0120]

[0121] Where χ=exp(j2π / 3),i wia =(S 1i -S 4i )i dc i wib =(S 3i -S 6i )i dc i wic =(S 5i -S 2i )i dc The spatial coordinate system is divided into 6 sectors and the angle of each sector is π / 3. Each possible discrete control input variable u is substituted into the evaluation function (18) and the current vector that is closest to the control input of the distance optimization is selected.

[0122] The advantages of this invention are:

[0123] (1) This invention uses an evaluation function to directly constrain all state variables of the system. By simultaneously controlling the filter capacitor voltage and stator current, it can be ensured that they can track the reference setpoint well, thereby avoiding the occurrence of resonance. The optimized control input is calculated in advance. Based on the optimized control input, a simplified evaluation function is obtained. The simplified evaluation function is used to select the final given current vector, reducing the repetitive prediction calculation process and the amount of computation during digital implementation, and greatly reducing the operating power consumption and cost.

[0124] (2) The present invention directly uses the evaluation function to control multiple spatial state variables, which is simple in design concept and highly flexible.

[0125] (3) The present invention eliminates the multi-loop control structure in the traditional control scheme, thereby improving the system bandwidth and dynamic response speed.

[0126] (4) Compared with the traditional active damping control scheme, the method proposed in this invention requires fewer parameters to be designed and the adjustment process is simple.

[0127] (5) The present invention is based on the analysis of a current source type permanent magnet motor drive system. The proposed algorithm can also be extended to other motor applications with high-order output filters. Attached Figure Description

[0128] Figure 1 This is a system function partitioning block diagram of the control method for a current-source inverter permanent magnet electric drive system provided in Embodiment 1 of the present invention;

[0129] Figure 2 This is a schematic block diagram of the composition of the current source inverter permanent magnet electric drive system in the control method applied to the current source inverter permanent magnet electric drive system provided in Embodiment 1 of the present invention.

[0130] Figure 3 This is a spatial current vector distribution diagram of the current source inverter in the control method for a current source inverter permanent magnet electric drive system provided in Embodiment 1 of the present invention;

[0131] Figure 4 This is a block diagram for calculating the coordinate transformation of the filter capacitor voltage and the motor stator current in the control method for a current source inverter permanent magnet electric drive system provided in Embodiment 1 of the present invention.

[0132] Figure 5 This is a block diagram of the calculation of filter capacitor voltage and motor stator current delay compensation in the control method for a current source inverter permanent magnet electric drive system provided in Embodiment 1 of the present invention;

[0133] Figure 6This is a schematic block diagram of the outer loop of motor speed control in the control method for a current source inverter permanent magnet electric drive system provided in Embodiment 1 of the present invention;

[0134] Figure 7 This is a block diagram for calculating the reference setpoint of the state variable in the control method for a current-source inverter permanent magnet electric drive system provided in Embodiment 1 of the present invention;

[0135] Figure 8 This is a block diagram of optimized control input variable calculation in the control method for a current-source inverter permanent magnet electric drive system provided in Embodiment 1 of the present invention;

[0136] Figure 9 The flowchart of the control method for a current-source inverter permanent magnet electric drive system provided in Embodiment 1 of the present invention, which selects the control variable of the current-source inverter corresponding to the minimum value according to a simplified evaluation function;

[0137] Figure 10 The steady-state waveforms of motor current, filter capacitor voltage, and speed in the control method for a current-source inverter permanent magnet electric drive system provided in Embodiment 1 of the present invention are shown.

[0138] Figure 11 The motor current, filter capacitor voltage, and speed waveforms under dynamic conditions are shown in the control method for a current-source inverter permanent magnet electric drive system provided in Embodiment 1 of the present invention.

[0139] Figure 12 The waveforms of motor current, filter capacitor voltage, and speed under no-load start-up conditions are provided in the control method for a current-source inverter permanent magnet electric drive system provided in Embodiment 1 of the present invention. Detailed Implementation

[0140] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0141] Example 1

[0142] like Figure 1 As shown, the present invention provides a control method for a permanent magnet electric drive system applied to a current source inverter, the method comprising:

[0143] Step 1: Obtain the current and voltage in the dq space coordinate system of the current control cycle; the specific process is as follows:

[0144] The stator currents of phase a and phase b of the motor in the current control cycle are sampled, and the current in the dq space coordinate system is obtained through Parker transformation. The specific calculation process is as follows:

[0145] i sc =-i sa -i sb (1)

[0146]

[0147] Among them, i sa i sb and i sc These represent the three-phase stator currents a, b, and c, respectively; i sd and i sq Let θ represent the stator current in the dq coordinate system, respectively. e Indicates angle;

[0148] Similarly, the stator terminal voltages of phase a and phase b of the motor in the current control cycle are sampled, and the voltages in the dq space coordinate system are obtained through Parker transformation. The specific calculation process is as follows:

[0149] v sc =-v sa -v sb (3)

[0150]

[0151] Among them, v sa ,v sb and v sc These represent the three-phase stator voltages a, b, and c, respectively; v sd and v sq These represent the stator voltages in the dq spatial coordinate system, respectively.

[0152] Step 2: Select the filter capacitor voltage as the state variable and establish the state-space model of the current source type electric drive system; the specific process is as follows:

[0153] In the synchronously rotating dq coordinate system, the stator voltage state-space equation of the permanent magnet synchronous motor is established as follows:

[0154]

[0155] Among them, R s ω represents the stator resistance of the motor; e ψ represents the angular velocity of the motor stator side; sd and ψ sq Let denot dq represent the stator flux linkage in the dq spatial coordinate system, and their calculation equations are as follows:

[0156]

[0157] L s This represents the stator-side inductance of the motor. For surface-mounted permanent magnet synchronous motors, the d-axis and q-axis inductances are equal; ψ f Indicates the size of the flux linkage of the rotor permanent magnet;

[0158] For a current-source inverter, the filter capacitor voltage is selected as the state variable, and its state-space equations are established according to Kirchhoff's current law as follows:

[0159]

[0160] Among them, i wd and i wq C represents the current on the inverter side of the current source in the dq spatial coordinate system; f The size of the filter capacitor is represented by equations (5) and (7). The continuous-time state-space model of the current-source inverter drive system is expressed as follows:

[0161]

[0162] Wherein, the state space vector x = [i sd i sq v sd v sq ] T , control input vector u = [i wd i wq ] T The specific expressions for the system matrix A, input matrix B, and interference vector E are as follows:

[0163]

[0164] Step 3: Discretize the state-space equations in the continuous-time domain to obtain the system's state variable prediction equations; the specific process is as follows:

[0165] Discretizing the state-space equations in the continuous-time domain yields the following prediction equations for the system's state variables:

[0166] x(k+1)=Gx(k)+Hu(k)+F (10)

[0167] Where G = e ATs e represents the natural index, T s H represents the sampling control period interval; H = (GI)A -1 B,F=(GI)A -1 E, I i Is with G iThe identity matrix has the same dimension; therefore, based on (10), the next state variable of the system can be predicted. Typically, due to the one-beat delay phenomenon of the digital controller, it is necessary to compensate for the delay of the current sampled variable by one beat. The calculation process is as follows:

[0168] x c (k+1)=Gx(k)+Hu i +F (11)

[0169] Where x c (k+1) represents the state variable after delay compensation at time k+1, u i This represents the control input of the current source inverter at the current moment. Therefore, based on the compensated state variables, controlling the predicted value at time k+2 can overcome the one-step delay phenomenon of the digital system. The predicted value at time k+2 is calculated as follows:

[0170] x(k+2)=Gx c (k+1)+Hu+F (12).

[0171] Step 4: Based on the system's state variable prediction equations and considering all system state variables, simultaneously control the filter capacitor voltage and stator current, and design an evaluation function; the specific process is as follows:

[0172] The core decision-making process of FCS-MPC is achieved through the design of an evaluation function. For current-source inverters, due to the second-order characteristics introduced by the filter capacitor and the stator inductance of the motor, controlling the stator current alone cannot suppress the system's resonance problem. Therefore, an evaluation function is needed to directly constrain all state variables of the system. By simultaneously controlling the filter capacitor voltage and the stator current, it can be ensured that they both track the reference setpoint well, thereby avoiding resonance. One of the advantages of FCS-MPC is its ability to easily design and control multivariable problems. Therefore, the evaluation function design for a current-source inverter electric drive system is as follows:

[0173] J = [x * (k+2)-x(k+2)] T W[x * (k+2)-x(k+2)] (13)

[0174] Where x * (k+2)=[i * sd (k+2)i * sq (k+2)v * sd (k+2)v * sq (k+2)]T For reference value vectors; x(k+2)=[i sd (k+2)i sq (k+2)v sd (k+2)v sq (k+2)] T Let W be the prediction vector of the state variables; W is the weighting coefficient matrix, which is also a 4th-order diagonal matrix. Typically, the weighting coefficient for the stator current of the motor can be set to 1, allowing the weighting coefficient for the filter capacitor voltage to be selected as a design parameter to adjust the system performance. Based on this, the specific expression for W is as follows:

[0175]

[0176] Step 5: Calculate the reference value of the filter capacitor in the evaluation function; the specific process is as follows:

[0177] For permanent magnet synchronous motors, the q-axis current is referenced to the given value i. * sq Derived from the outer speed loop, the d-axis current references the given value i. * sd To achieve maximum torque-to-current ratio control, the reference setpoint of the filter capacitor is calculated based on the stator current reference value of the dq axis and the system's state-space model, as follows:

[0178]

[0179] Step 6: Calculate the optimized control input, and based on this optimized control input, obtain the simplified evaluation function; the specific process is as follows:

[0180] In the implementation steps of FCS-MPC, it is common practice to repeatedly use Equation (12) for each possible control input condition to perform repeated prediction calculations, and then substitute each possible prediction quantity into the evaluation function (13), selecting the control input corresponding to the minimum evaluation function value to apply to the inverter. However, this process of repeatedly performing prediction calculations using Equation (12) involves a large number of arithmetic operations, which will significantly increase the computational burden during digital implementation. Therefore, in order to simplify the digital implementation process of the proposed FCS-MPC, this invention adopts the method of pre-calculating the optimized control input, then comparing the squared error between each possible discrete control input and the optimized control input, and finally selecting the input quantity closest to the optimized control input to apply to the inverter. Based on this, Equation (11) is substituted into (13), and the derivative of (13) with respect to the control input variable is taken, and the derivative is set to zero, i.e.

[0181]

[0182] Then, the optimized control input is obtained, and its expression is as follows:

[0183] u * =(H T WH) -1 ·H T W·[x * -Gx c -F] (17)

[0184] Based on this optimized control input, the simplified evaluation function is designed as follows:

[0185] J 简化 =[u * (k+2)-u(k+2)] T [u * (k+2)-u(k+2)] (18).

[0186] Step 7: Calculate the current vector of the current source inverter in the stationary αβ coordinate system, and select the current vector that is closest to the optimized control input; the specific process is as follows:

[0187] like Figure 2 As shown, in a current-source inverter, to prevent open-circuit conditions, at least one switch in each of the upper three bridge arms and the lower three bridge arms must be on simultaneously. Therefore, a current-source inverter can support nine switching states, including six switching states corresponding to effective current vectors and three switching states corresponding to zero current vectors. The current vector of a current-source inverter is defined as i based on the switching state. wi =[S 1i S 3i S 5i S 4i S 6i S 2i ], '1' indicates that the corresponding switch state is on, and '0' indicates that the corresponding switch state is off. The current vector is represented in the stationary αβ coordinate system as follows:

[0188]

[0189] Where χ=exp(j2π / 3),i wia =(S 1i -S 4i )i dc i wib =(S 3i -S 6i )i dc i wic =(S 5i -S 2i )i dcThe spatial coordinate system is divided into 6 sectors and the angle of each sector is π / 3. Each possible discrete control input variable u is substituted into the evaluation function (18) and the current vector that is closest to the control input of the distance optimization is selected.

[0190] Step 8: Convert the selected current vector into a symmetrical pulse sequence and apply it to the current source inverter.

[0191] Figure 1 The functional partitioning block diagram of the FCS-MPC algorithm control system for suppressing current source inverter resonance provided by this invention is shown below. The entire system is divided into seven parts: coordinate transformation of stator current and filter capacitor voltage (01), one-step delay compensation for state variables (02), speed outer loop (03), calculation of state variable reference values ​​(04), calculation of optimized current control variables (05), determination of finite control set (06), and simplified objective function calculation and selection of the current vector corresponding to the minimum value (07). These seven parts are described in detail below:

[0192] 1) Coordinate transformation of stator current and filter capacitor voltage 01

[0193] Figure 4 The process involves sampling the stator currents of phases a and b of the motor, then performing a Parker transformation to obtain the current in the dq spatial coordinate system, and sampling the filter capacitor voltages of phases a and b, then performing a Parker transformation to obtain the voltage in the dq spatial coordinate system. This process only requires collecting the state information of phases a and b; the state information of phase c can be calculated using the three-phase balancing principle, thus reducing the number of sensors required.

[0194] 2) Perform one-time delay compensation on the state variables 02

[0195] Because digital controllers have a one-beat delay, it is necessary to compensate for the delay of the current sampled variable by one beat. Figure 5 A calculation diagram of the compensation process is provided. The compensated variables will be used to optimize the calculation of the control input, thereby effectively overcoming the effects of digital delay.

[0196] 3) Rotational speed outer ring 03

[0197] The motor speed control is still achieved through proportional-integral control. Figure 6 A block diagram of this outer loop control is provided. The outer loop provides the q-axis stator current reference setpoint for the motor speed. The d-axis stator current reference setpoint is set to zero to achieve maximum torque-to-current ratio control.

[0198] 4) Calculation of reference values ​​for state variables 04

[0199] Since the evaluation function in the proposed FCS-MPC method controls both the filter capacitor voltage and the motor stator current, it is necessary to calculate the reference setpoint value for the filter capacitor voltage. Figure 7 The process of calculating the reference setpoint of the filter capacitor voltage based on the reference setpoint of the stator current and the system model is given.

[0200] 5) Calculation of optimized current control variables 05

[0201] To reduce the repetitive prediction calculations and computational load during digital implementation, it is necessary to pre-calculate the optimized current control variable inputs and use a simplified evaluation function to select the final given current vector. Figure 8 A schematic diagram is given showing how to calculate the optimal control input based on the reference values ​​of the state variables and the system model.

[0202] 6) Determining the finite control set 06

[0203] Figure 3 Nine current vector space distribution diagrams provided by the current source inverter are given, which are used as optional control sets and evaluated one by one in a simplified evaluation function.

[0204] 7) Simplified objective function calculation and selection of the current vector corresponding to the minimum value 07

[0205] The nine current vectors generated by the current source inverter are first converted and calculated in the dq coordinate system, and then substituted into the simplified evaluation function to obtain the corresponding objective function values. Finally, the current vector corresponding to the minimum objective function is selected as the output. Figure 9 A flowchart for selecting the current vector is provided.

[0206] Figure 10 These are the motor current, filter capacitor voltage, and speed waveform under steady-state conditions (n ​​= 1000 r / min, TL = 8 N·m). Figure 11 It is the motor current, filter capacitor voltage and speed waveform under dynamic conditions (n=1000r / min, TL=0N·m jumps to 8N·m). Figure 12 It shows the motor current, filter capacitor voltage, and speed waveform under no-load starting conditions (n ​​= 0 r / min jumps to 1000 r / min).

[0207] Through the above technical solutions, this invention leverages the advantages of the FCS-MPC method to flexibly control multivariable problems, eliminating multi-loop feedback control in traditional control structures, simplifying algorithm design and parameter adjustment, and improving the system's response bandwidth and dynamic response speed. This invention suppresses LC filter resonance while avoiding complex repetitive calculations, facilitating digital processor implementation and providing valuable reference for engineering applications. The control method proposed in this invention can also be extended to other motor drive scenarios with output filters.

[0208] Example 2

[0209] Based on Embodiment 1, Embodiment 2 of the present invention also provides a control system for a permanent magnet electric drive system of a current source inverter, the system comprising:

[0210] The current and voltage acquisition module is used to acquire the current and voltage in the dq space coordinate system of the current control cycle;

[0211] The model building module is used to select the filter capacitor voltage as the state variable and establish the state space model of the current source type electric drive system.

[0212] The prediction equation construction module is used to discretize the state-space equations in the continuous time domain to obtain the prediction equations of the system's state variables.

[0213] The evaluation function construction module is used to design evaluation functions by simultaneously controlling the filter capacitor voltage and stator current based on the system's state variable prediction equation and considering all system state variables.

[0214] The reference given value calculation module is used to calculate the reference given value of the filter capacitor in the evaluation function;

[0215] The function simplification module is used to calculate the optimized control input, and based on the optimized control input, a simplified evaluation function is obtained;

[0216] The vector optimization module is used to calculate the current vector of the current source inverter in the stationary αβ coordinate system and select the current vector that is closest to the optimized control input.

[0217] The result output module is used to convert the selected current vector into a symmetrical pulse sequence and apply it to the current source inverter.

[0218] Specifically, the current and voltage acquisition module is also used for:

[0219] The stator currents of phase a and phase b of the motor in the current control cycle are sampled, and the current in the dq space coordinate system is obtained through Parker transformation. The specific calculation process is as follows:

[0220] isc =-i sa -i sb (1)

[0221]

[0222] Among them, i sa i sb and i sc These represent the three-phase stator currents a, b, and c, respectively; i sd and i sq Let θ represent the stator current in the dq coordinate system, respectively. e Indicates angle;

[0223] The stator terminal voltages of phase a and phase b of the motor in the current control cycle are sampled and then transformed by Parker transformation to obtain the voltages in the dq spatial coordinate system. The specific calculation process is as follows:

[0224] v sc =-v sa -v sb (3)

[0225]

[0226] Among them, v sa ,v sb and v sc These represent the three-phase stator voltages a, b, and c, respectively; v sd and v sq These represent the stator voltages in the dq spatial coordinate system, respectively.

[0227] More specifically, the model building module is also used for:

[0228] In the synchronously rotating dq coordinate system, the stator voltage state-space equation of the permanent magnet synchronous motor is established as follows:

[0229]

[0230] Among them, R s ω represents the stator resistance of the motor; e ψ represents the angular velocity of the motor stator side; sd and ψ sq Let denot dq represent the stator flux linkage in the dq spatial coordinate system, and their calculation equations are as follows:

[0231]

[0232] L s This represents the stator-side inductance of the motor. For surface-mounted permanent magnet synchronous motors, the d-axis and q-axis inductances are equal; ψ f Indicates the size of the flux linkage of the rotor permanent magnet;

[0233] For a current-source inverter, the filter capacitor voltage is selected as the state variable, and its state-space equations are established according to Kirchhoff's current law as follows:

[0234]

[0235] Among them, i wd and i wq C represents the current on the inverter side of the current source in the dq spatial coordinate system; f The size of the filter capacitor is represented by equations (5) and (7). The continuous-time state-space model of the current-source inverter drive system is expressed as follows:

[0236]

[0237] Wherein, the state space vector x = [i sd i sq v sd v sq ] T , control input vector u = [i wd i wq ] T The specific expressions for the system matrix A, input matrix B, and interference vector E are as follows:

[0238]

[0239] More specifically, the prediction equation construction module is also used for:

[0240] Discretizing the state-space equations in the continuous-time domain yields the following prediction equations for the system's state variables:

[0241] x(k+1)=Gx(k)+Hu(k)+F (10)

[0242] Where G = e ATs e represents the natural index, T s H represents the sampling control period interval; H = (GI)A -1 B,F=(GI)A -1 E, I i Is with G i Identity matrices with the same dimensions;

[0243] The calculation process for the delay compensation of one beat ahead for the current sampled variable is as follows:

[0244] x c (k+1)=Gx(k)+Hu i +F (11)

[0245] Where x c(k+1) represents the state variable after delay compensation at time k+1, u i This represents the control input of the current source inverter at the current moment. Therefore, based on the compensated state variables, controlling the predicted value at time k+2 can overcome the one-step delay phenomenon of the digital system. The predicted value at time k+2 is calculated as follows:

[0246] x(k+2)=Gx c (k+1)+Hu+F (12).

[0247] More specifically, the evaluation function construction module is also used for:

[0248] The evaluation function designed for current-source inverter electric drive systems is as follows:

[0249] J = [x * (k+2)-x(k+2)] T W[x * (k+2)-x(k+2)] (13)

[0250] Where x * (k+2)=[i * sd (k+2)i * sq (k+2)v * sd (k+2)v * sq (k+2)] T For reference value vectors; x(k+2)=[i sd (k+2)i sq (k+2)v sd (k+2)v sq (k+2)] T Let W be the prediction vector of the state variables; W is the weight coefficient matrix, and its specific expression is as follows:

[0251]

[0252] More specifically, the reference given value calculation module is also used for:

[0253] For permanent magnet synchronous motors, the q-axis current is referenced to the given value i. * sq Derived from the outer speed loop, the d-axis current references the given value i. * sd To achieve maximum torque-to-current ratio control, the reference setpoint of the filter capacitor is calculated based on the stator current reference value of the dq axis and the system's state-space model, as follows:

[0254]

[0255] More specifically, the function simplification module is also used for:

[0256] Substituting equation (11) into (13), and differentiating (13) with respect to the control input variable, setting the derivative to zero, i.e.

[0257]

[0258] Then, the optimized control input is obtained, and its expression is as follows:

[0259] u * =(H T WH) -1 ·H T W·[x * -Gx c -F] (17)

[0260] Based on this optimized control input, the simplified evaluation function is designed as follows:

[0261] J 简化 =[u * (k+2)-u(k+2)] T [u * (k+2)-u(k+2)] (18).

[0262] More specifically, the vector optimization module is also used for:

[0263] A current-source inverter includes six active current vectors corresponding to switching states and three zero current vectors corresponding to switching states. The current vector of the current-source inverter is defined as i based on the switching states. wi =[S 1i S 3i S 5i S 4i S 6i S 2i ], '1' indicates that the corresponding switch state is on, and '0' indicates that the corresponding switch state is off. The current vector is represented in the stationary αβ coordinate system as follows:

[0264]

[0265] Where χ=exp(j2π / 3),i wia =(S 1i -S 4i )i dc i wib =(S 3i -S 6i )i dc i wic =(S 5i-S 2i )i dc The spatial coordinate system is divided into 6 sectors and the angle of each sector is π / 3. Each possible discrete control input variable u is substituted into the evaluation function (18) and the current vector that is closest to the control input of the distance optimization is selected.

[0266] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A control method applied to a permanent magnet electric drive system of a current source inverter, characterized in that, The method includes: Step 1: Obtain the current and voltage in the dq space coordinate system of the current control cycle; Step 2: Select the filter capacitor voltage as the state variable and establish the state-space model of the current source type electric drive system; In the synchronously rotating dq coordinate system, the stator voltage state-space equation of the permanent magnet synchronous motor is established as follows: (5) in, R s This indicates the stator-side resistance of the motor; ω e Indicates the angular velocity of the motor stator side; ψ sd and ψ sq Let denot dq represent the stator flux linkage in the dq spatial coordinate system, and their calculation equations are as follows: (6) L s This represents the stator-side inductance of the motor. For surface-mounted permanent magnet synchronous motors, the d-axis and q-axis inductances are equal. ψ f Indicates the size of the flux linkage of the rotor permanent magnet; For a current-source inverter, the filter capacitor voltage is selected as the state variable, and its state-space equations are established according to Kirchhoff's current law as follows: (7) in, i wd and i wq These represent the currents on the inverter side of the current source in the dq spatial coordinate system, respectively. C f The size of the filter capacitor is represented by equations (5) and (7). The continuous-time state-space model of the current-source inverter drive system is expressed as follows: (8) Among them, the state space vector x =[ i sd i sq v sd v sq ] T , control input vector u =[ i wd i wq ] T System matrix A Input matrix B and interference vector E The specific expression is as follows: ， ， (9)； Step 3: Discretize the state-space equations in the continuous time domain to obtain the state variable prediction equations of the system; Step 4: Based on the system's state variable prediction equations and considering all system state variables, simultaneously control the filter capacitor voltage and stator current, and design an evaluation function; The evaluation function designed for current-source inverter electric drive systems is as follows: (13) in A reference value vector; Let W be the prediction vector of the state variables; W is the weight coefficient matrix, and its specific expression is as follows: (14)； Step 5: Calculate the reference value of the filter capacitor in the evaluation function; Step 6: Calculate the optimized control input, and based on the optimized control input, obtain the simplified evaluation function; Step 7: Calculate the current vector of the current source inverter in the stationary αβ coordinate system, and select the current vector that is closest to the optimized control input; Step 8: Convert the selected current vector into a symmetrical pulse sequence and apply it to the current source inverter.

2. The control method for a permanent magnet electric drive system applied to a current source inverter according to claim 1, characterized in that, Step 1 includes: The current control cycle is obtained by sampling. Harmony The stator current of the phase motor is obtained in the dq space coordinate system through Parker transformation. The specific calculation process is as follows: (1) (2) in, i sa , i sb and i sc They represent Three-phase stator current; i sd and i sq Let represent the stator current in the dq spatial coordinate system, respectively. Indicates angle; The current control cycle is obtained by sampling. Harmony The stator terminal voltage of the phase motor is obtained in the dq space coordinate system through Parker transformation. The specific calculation process is as follows: (3) (4) in, v sa , v sb and v sc These represent the three-phase stator voltages a, b, and c, respectively. v sd and v sq These represent the stator voltages in the dq spatial coordinate system, respectively.

3. The control method for a permanent magnet electric drive system applied to a current source inverter according to claim 2, characterized in that, Step 3 includes: Discretizing the state-space equations in the continuous-time domain yields the following prediction equations for the system's state variables: (10) Where G= e ATs , e Represents the natural index. T s H represents the sampling control period interval; H=(GI)A -1 B, F=(GI)A -1 E, I i Is with G i Identity matrices with the same dimensions; The calculation process for the delay compensation of one beat ahead for the current sampled variable is as follows: (11) in express The state variables after time delay compensation u i This represents the control input of the current source inverter at the current moment. Therefore, based on the compensated state variables, for k Controlling the prediction at time +2 can overcome the one-beat delay phenomenon in digital systems. k The prediction at time +2 is calculated as follows: (12)。 4. The control method for a permanent magnet electric drive system of a current source inverter according to claim 1, characterized in that, Step 5 includes: For permanent magnet synchronous motors, the q-axis current is referenced to a given value. Derived from the outer speed loop, d-axis current reference value To achieve maximum torque-to-current ratio control, the reference setpoint of the filter capacitor is calculated based on the stator current reference value of the dq axis and the system's state-space model, as follows: (15)。 5. The control method for a permanent magnet electric drive system of a current source inverter according to claim 4, characterized in that, Step 6 includes: Substituting equation (11) into (13), and differentiating (13) with respect to the control input variable, setting the derivative to zero, i.e. (16) Then, the optimized control input is obtained, and its expression is as follows: (17) Based on this optimized control input, the simplified evaluation function is designed as follows: (18)。 6. The control method for a permanent magnet electric drive system of a current source inverter according to claim 5, characterized in that, Step 7 includes: A current-source inverter includes six active current vectors corresponding to switching states and three zero current vectors corresponding to switching states. The current vectors of a current-source inverter are defined according to the switching states. '1' indicates that the corresponding switch state is on, and '0' indicates that the corresponding switch state is off. The current vector is represented in the stationary αβ coordinate system as follows: (19) in The spatial coordinate system is divided into 6 sectors, and the angle of each sector is π / 3. Each possible discrete control input variable... u Substitute it into the evaluation function (18) and select the current vector that is closest to the control input of the distance optimization.

7. A control system applied to a permanent magnet electric drive system for a current source inverter, characterized in that, The system performs the method according to any one of claims 1-6, the system comprising: The current and voltage acquisition module is used to acquire the current and voltage in the dq space coordinate system of the current control cycle; The model building module is used to select the filter capacitor voltage as the state variable and establish the state space model of the current source type electric drive system. The prediction equation construction module is used to discretize the state-space equations in the continuous time domain to obtain the prediction equations of the system's state variables. The evaluation function construction module is used to design evaluation functions by simultaneously controlling the filter capacitor voltage and stator current based on the system's state variable prediction equation and considering all system state variables. The reference given value calculation module is used to calculate the reference given value of the filter capacitor in the evaluation function; The function simplification module is used to calculate the optimized control input, and based on the optimized control input, a simplified evaluation function is obtained; The vector optimization module is used to calculate the current vector of the current source inverter in the stationary αβ coordinate system and select the current vector that is closest to the optimized control input. The result output module is used to convert the selected current vector into a symmetrical pulse sequence and apply it to the current source inverter.

8. The control system for a current-source inverter permanent magnet electric drive system according to claim 7, characterized in that, The current and voltage acquisition module is also used for: The current control cycle is obtained by sampling. Harmony The stator current of the phase motor is obtained in the dq space coordinate system through Parker transformation. The specific calculation process is as follows: (1) (2) in, i sa , i sb and i sc They represent Three-phase stator current; i sd and i sq Let represent the stator current in the dq spatial coordinate system, respectively. Indicates angle; The current control cycle is obtained by sampling. Harmony The stator terminal voltage of the phase motor is obtained in the dq space coordinate system through Parker transformation. The specific calculation process is as follows: (3) (4) in, v sa , v sb and v sc These represent the three-phase stator voltages a, b, and c, respectively. v sd and v sq These represent the stator voltages in the dq spatial coordinate system, respectively.

Images