A decoupled tracking control method and system for fundamental and multi-order harmonic currents

By using a control architecture that connects a low-pass internal mode and a resonant-leading branch in parallel under a single rotating orthogonal coordinate system, and by precisely tuning the lead correction parameters, the complexity and stability issues of the control structure in the control of fundamental and multi-order harmonic currents are solved, and high-precision decoupled tracking in a wide frequency band is achieved.

CN123122380APending Publication Date: 2026-11-17HEFEI UNIV OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611636912.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-22
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing fundamental and multi-harmonic current control technologies struggle to achieve high-precision decoupled tracking and stable operation across a wide frequency band while simplifying the control structure. In particular, high-frequency current tracking suffers from amplitude attenuation and phase lag issues, and parameter tuning is highly complex.

Method used

A parallel structure of low-pass internal mode branch and resonant-leading branch under a single set of synchronous rotating orthogonal coordinate system is adopted. Combined with the precise closed-loop tuning of the lead correction parameter, the inter-axis cross-coupling and high-frequency inductor roll-off effect are offset by decoupling the inverse model, so as to realize independent decoupling tracking control of fundamental and multi-order harmonic currents.

Benefits of technology

The control structure is simplified, the system computation and parameter debugging complexity are reduced, independent decoupled tracking in a wide frequency band is achieved, the system stability and tracking accuracy are improved, and it is suitable for high-performance scenarios such as motor drive and power quality management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123122380A_ABST
    Figure CN123122380A_ABST
Patent Text Reader

Abstract

This invention belongs to the field of power electronics and AC current control technology, specifically relating to a decoupled tracking control method and system for fundamental and multi-order harmonic currents. This method is based on a single rotating orthogonal coordinate system synchronized with the fundamental wave. It employs a control architecture of a decoupled inverse model, consisting of a low-pass internal mode branch and multiple resonant-leading branches connected in parallel and then in series. The overall transfer function of the parallel branches is constrained to have an amplitude of 1 and a phase of 0 at the target harmonic frequency. The gain coefficient, time constant, and indexing coefficient of the lead correction branch are precisely calculated and tuned using closed-loop calculations, and are complemented by low-frequency segmented adaptation and frequency pre-correction discretization schemes. This method simplifies the multi-order harmonic control structure, avoids multi-coordinate system coupling interference, and cancels inter-axis cross-coupling and high-frequency inductor roll-off effects. It achieves independent decoupled control of the fundamental wave and each harmonic with zero amplitude and phase deviation across a wide frequency band, while simultaneously improving the overall operational stability and digital implementation accuracy.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of power electronics and AC current control technology, specifically to a decoupled tracking control method for fundamental and multi-order harmonic currents, and a decoupled tracking control system for fundamental and multi-order harmonic currents. Background Technology

[0002] With the rapid development of power electronics technology and the fields of motor drive and power quality management, high-precision tracking control of fundamental and multi-order harmonic currents has become a core research direction in the industry. In applications such as electromagnetic noise suppression by harmonic current injection in permanent magnet synchronous motors, multi-order harmonic compensation in active power filters, and ripple suppression in power factor correction converters, it is often necessary to independently decouple and control the fundamental and multi-order harmonic currents simultaneously, and to achieve good tracking accuracy and dynamic response performance over a wide frequency range.

[0003] Currently, the control of fundamental and multi-order harmonic currents mainly adopts the following two types of technical solutions: The first type is the multi-synchronous rotating coordinate system scheme, which establishes a corresponding synchronous rotating coordinate system for each harmonic, transforms each harmonic component into a DC component in the corresponding coordinate system, and then controls them separately through a PI controller. This scheme can achieve independent control of each harmonic, but it also faces some challenges in practical applications: each additional harmonic requires an additional set of coordinate transformations and controllers; the computational load and complexity of the control structure increase linearly with the harmonic order; and there is coupling interference between multiple coordinate systems, making parameter tuning difficult. Its application is limited in scenarios with many harmonic orders and high real-time requirements.

[0004] The second type is the proportional resonant control scheme in a single coordinate system. This scheme uses multiple proportional resonant controllers connected in parallel within a single stationary coordinate system or a fundamental rotating coordinate system to control the fundamental wave and each harmonic separately. While this scheme is relatively simple and does not require multiple coordinate transformations, there is still room for improvement in control accuracy and stability: First, the controlled object in the rotating coordinate system has inter-axis cross-coupling terms, and the inductor impedance exhibits a roll-off characteristic as frequency increases. Conventional proportional resonant control does not have a dedicated decoupling compensation stage, making amplitude attenuation and phase lag prone to occur in high-frequency current tracking, which affects tracking accuracy as frequency increases. Second, the parallel control branches lack precise parameter tuning methods, making it difficult to ensure accurate matching of the total transfer function amplitude and phase at the target frequency, and amplitude-phase deviations are difficult to completely eliminate. Third, the low-frequency resonant branch and the fundamental branch are prone to coupled oscillations, requiring improved system stability. Furthermore, the amplitude-frequency and phase-frequency characteristics at the target frequency are prone to shift after discretization, which can affect control accuracy.

[0005] In summary, existing technical solutions cannot simultaneously meet the requirements of control structure simplicity, multi-order decoupling independence, wide-band tracking accuracy and system stability, and cannot fully satisfy the decoupling tracking control requirements of fundamental and multi-order harmonic currents in high-performance scenarios. Summary of the Invention

[0006] To address the challenge of achieving high-precision decoupling tracking and stable operation of fundamental and multi-harmonic currents over a wide frequency band while simplifying the control structure, existing fundamental and multi-harmonic current control technologies provide a decoupling tracking control method and system. Based on a single rotating orthogonal coordinate system synchronized with the fundamental, this invention utilizes a control architecture that decouples the inverse model through parallel and series connections of a low-pass internal mode branch and a resonant-lead branch. Combined with precise closed-loop tuning of the lead correction parameter, this simplifies the multi-order control structure while offsetting inter-axis cross-coupling and high-frequency inductor roll-off effects, achieving independent decoupling tracking control of the fundamental and multi-harmonic currents with zero amplitude and phase deviation over a wide frequency band.

[0007] To achieve the above objectives, the present invention provides the following technical solution: A decoupled tracking control method for fundamental and multi-order harmonic currents includes the following steps: In a rotating orthogonal coordinate system synchronized with the fundamental wave, the controlled object is modeled as a transfer function matrix containing cross-coupling terms, and the decoupling inverse model of the transfer function matrix is ​​obtained. The low-pass internal mode branch is connected in parallel with at least one resonant-lead branch, and then the inverse model is decoupled in series to form a current controller; wherein, the resonant-lead branch is formed by connecting the resonant branch and the lead correction branch in series. The real-time controlled current of the controlled object is collected and the difference between the current and the current reference value is used to obtain the current error signal. In the rotating orthogonal coordinate system, the current error signal is decomposed into the fundamental error component and at least one target subharmonic error component. The current error signal is input to the current controller. Each parallel branch performs independent calculations on the corresponding error components, and after compensation calculations of the decoupled inverse model, the output voltage control quantity is applied to the controlled object to form a negative feedback closed loop, thereby realizing decoupled tracking control of the fundamental wave and multi-order harmonic current.

[0008] As a further improvement to the above scheme: the resonant branch resonates at the target subharmonic frequency w0. k is the target harmonic order, ω base The fundamental angular frequency; the transfer function of the lead compensation branch. for: ; ; ; ; In the formula, s is the Laplace operator; K l b is the gain coefficient; T is the time constant; b is the lead angle resolution coefficient; a is the design bandwidth of the low-pass internal mode branch; With the resonant-lead branch and the low-pass internal mode branch connected in parallel, and the total transfer function at w0 having an amplitude of 1 and a phase of 0 as constraints, calculate K. l 、T、b.

[0009] The core purpose of setting "amplitude of 1 and phase of 0" is to offset the amplitude attenuation and phase lag of the low-pass internal mode branch at the target subharmonic frequency, so that the error signal at the target frequency can pass through the control branch "without loss and without phase shift". Finally, in conjunction with the decoupled inverse model, the entire forward channel is equivalent to unity gain at the target frequency point, so as to achieve zero steady-state error tracking of the closed-loop current.

[0010] As a further improvement to the above scheme: the low-pass internal mode branch is used to calculate the fundamental error component in the current error signal, thereby realizing closed-loop control of the fundamental current component. Its transfer function is: ; In the formula, a is the design bandwidth of the low-pass internal mode branch.

[0011] As a further improvement to the above scheme: the resonant branch is used to form a resonant frequency selection characteristic for the target subharmonic error component in the current error signal at the target subharmonic frequency w0, and its transfer function is: ; Among them, K r For the resonant gain, w c For the resonant bandwidth, satisfying w c =w0 / Q, where Q is the quality factor. By setting K... r =1, so that the gain of the resonant branch at the target subharmonic frequency w0 is . ( The phase angle coincides; This means that at the target subharmonic frequency, the amplitude gain of the transfer function is 1 (the output amplitude is exactly equal to the input, with no attenuation) and the phase deviation is 0° (the output phase is perfectly aligned with the input phase, with no lag / lead), which is the ideal zero-deviation amplitude-phase tracking.

[0012] As a further improvement to the above scheme: when the target subharmonic frequency or fundamental angular frequency is lower than the preset threshold, all resonant-leader branches are shut down, and only the low-pass inner mode branch is put into closed-loop operation; or the design bandwidth 'a' of the low-pass inner mode branch is increased to the target bandwidth first, and then all resonant-leader branches are enabled.

[0013] As a further improvement to the above scheme: when the current reference value is the preset target value of the target subharmonic current component, the target subharmonic current component is actively injected; when the current reference value is zero or the reverse compensation value, the target subharmonic current component is actively suppressed.

[0014] As a further improvement to the above scheme: the current controller is discretized using a frequency pre-correction discretization method, so that the amplitude and phase frequency characteristics of the discretized current controller at the target subharmonic frequency w0 are consistent with the theoretical model of the current controller in the continuous time domain; the series resonant branch and the lead correction branch are combined and discretized into a third-order digital filter; the low-pass internal mode branch and the decoupled inverse model are combined and discretized into a recursive calculation formula containing an integral term; the third-order digital filter and the recursive calculation formula update the output value in each control cycle.

[0015] As a further improvement to the above scheme: the discretization method for frequency pre-correction is a pre-shearing bilinear transform, and its transformation relationship is as follows: ; In the formula, dT is the control period, z is the discrete-time Z-transform operator; K t is the pre-shearing coefficient; tan(·) is the tangent function.

[0016] As a further improvement to the above scheme: the coefficients of the third-order digital filter and the coefficients of the recursive calculation formula are updated online in each control cycle or according to the set task cycle as the target subharmonic frequency w0 changes, or updated in advance by looking up a table after offline calculation according to the speed / frequency gear.

[0017] As a further improvement to the above scheme: the controlled object is any one of permanent magnet synchronous motor, induction motor, multiphase motor, power factor correction converter, active power filter, grid-connected inverter, motor simulator or grid simulator.

[0018] As a further improvement to the above scheme: when the controlled object is a multiphase motor, the transfer function matrix and the decoupling inverse model are extended to a block matrix under a multidimensional orthogonal plane, and the same set of parallel control architecture of low-pass internal mode branch and resonant-leading branch is reused in each orthogonal plane.

[0019] As a further improvement to the above scheme: when simultaneously tracking and controlling multiple target subharmonic current components, an independent resonant-lead branch is constructed for each target subharmonic current component; after all resonant-lead branches are connected in parallel, they are connected in series with the low-pass internal mode branch to the same decoupled inverse model; the gain coefficient K of each resonant-lead branch... lThe time constant T and the lead angle division coefficient b are substituted into their respective target subharmonic frequencies and calculated independently to ensure that the phase lag and amplitude deviation at each target subharmonic frequency are both zero.

[0020] A decoupled tracking control system for fundamental and multi-order harmonic currents, used to execute a decoupled tracking control method for fundamental and multi-order harmonic currents, comprising: The decoupling inverse model module is used to model the controlled object as a transfer function matrix containing cross-coupling terms in a rotating orthogonal coordinate system synchronized with the fundamental wave, and to obtain the decoupling inverse model of the transfer function matrix. The low-pass internal mode branch module is used for internal mode control of the fundamental error component; The resonant branch module is used to form a resonant frequency selection characteristic for the target subharmonic error component in the current error signal at the target subharmonic frequency. The lead correction branch module has a transfer function, and the parameters of the transfer function are calculated by the corresponding calculation formula; The parallel and compensation module is used to connect the resonant branch module and the lead compensation branch module in series, then connect them in parallel with the low-pass internal mode branch module, and then connect them in series with the decoupled inverse model module. Through parameter compensation of the lead compensation branch module, the phase lag and amplitude deviation generated by the parallel superposition are reduced to zero at the target subharmonic frequency. The control output module is used to output control quantities to the controlled object with the current reference value of the target subharmonic current component as a reference signal, so as to realize closed-loop tracking control of the target subharmonic current component.

[0021] Compared with the prior art, the beneficial effects of the present invention are: 1. In the decoupled tracking control method of this invention, compared with the traditional multi-synchronous rotating coordinate system scheme, there is no need to build a separate coordinate transformation and control link for each harmonic. Independent decoupled control of the fundamental wave and multiple harmonics can be achieved simply by connecting a low-pass internal mode branch and multiple resonant-leading branches in parallel. This significantly simplifies the control structure, reduces the system's computational load and parameter tuning complexity, and fundamentally avoids the coupling interference problem between multiple coordinate systems. By decoupling the inverse model in series in the controller path, on the one hand, the inter-axis cross-coupling terms of the controlled object in the rotating coordinate system are canceled, achieving completely independent control of the two-phase axes and eliminating the impact of inter-axis interference on control accuracy. On the other hand, the roll-off characteristic of the controlled object's inductive impedance as the frequency increases is compensated, effectively solving the problems of amplitude attenuation and phase lag in high-frequency current tracking, and significantly broadening the range of... The effective operating frequency band of the control is controlled by connecting a lead correction branch in series after the resonant branch and completing precise closed-loop tuning of parameters with the total transfer function of the parallel branch having an amplitude of 1 and a phase of 0 at the target harmonic frequency. This can accurately offset the phase lag and amplitude deviation of the low-pass internal mode branch at the target frequency, ensuring that the amplitude and phase errors at each target frequency point are precisely zero, thus guaranteeing the tracking accuracy of each harmonic. At the same time, it can be combined with the branch switching and bandwidth adjustment strategy in the low-frequency band to alleviate the coupling oscillation problem between the resonant branch and the fundamental branch in the low-frequency band, and improve the operating stability of the system across the entire frequency band. The overall solution can maintain the simplicity of the control structure while taking into account the independence of multi-order decoupling, wide-band tracking accuracy and system operating stability. It can be adapted to the fundamental and multi-order harmonic current control requirements of various high-performance scenarios such as motor drive and power quality management.

[0022] 2. By adopting a segmented adaptation strategy of switching low-frequency resonant-leader branch or increasing the bandwidth of low-pass internal mode branch, low-frequency coupling oscillation can be effectively suppressed without changing the high-frequency control parameters, thereby improving the system's operational stability under low-speed conditions and achieving performance balance across the entire operating frequency band.

[0023] 3. By employing a frequency pre-correction discretization method using pre-shearing bilinear transform, the amplitude and phase characteristics of the discretized digital controller at the target subharmonic frequency point are guaranteed to be completely consistent with the continuous domain theoretical model, avoiding discretization deviations that reduce control accuracy. Simultaneously, by merging the resonant branch and the lead correction branch into a third-order digital filter, and merging the fundamental branch and the decoupled inverse model into a recursive calculation formula containing integral terms, the computational load can be significantly reduced, and control real-time performance improved. Combined with the flexible configuration of online real-time calculation or offline table lookup updates for coefficients, it can be adapted to digital processors of different computing power levels, improving engineering feasibility.

[0024] 4. Based on a general control architecture that can be adapted to multiple types of controlled objects, different scenarios can be deployed simply by adjusting the specific parameters and target harmonic order of the decoupled inverse model. It has good scenario versatility and engineering reusability, and can cover the harmonic current control needs of multiple fields such as motor drive, power quality management, and power simulation.

[0025] 5. By adopting a block matrix extension structure under multidimensional orthogonal planes, the transfer function matrix and decoupling inverse model are extended to multiphase motor scenarios. Each orthogonal plane can reuse the same parallel control architecture, eliminating the need to redesign the control link for new orthogonal planes. While ensuring the independent decoupling control accuracy of each plane, the development complexity and architecture redundancy of the multiphase motor harmonic control system are significantly reduced, and the development cycle of the multiphase system is shortened.

[0026] 6. The configuration method of independent resonance-leader branch and independent parameter tuning for each harmonic is adopted. The control process of each harmonic does not interfere with each other. The increase or decrease of the harmonic order does not require adjustment of the overall control architecture and fundamental branch parameters. It can flexibly adapt to the application requirements of different harmonic orders, has good scheme scalability, and at the same time ensures that each harmonic can achieve precise control with zero amplitude and phase deviation. Attached Figure Description

[0027] Figure 1 The flowchart shows the decoupled tracking control method.

[0028] Figure 2 This is a block diagram of the controller structure.

[0029] Figure 3 This is a flowchart illustrating the discrete deployment execution process in Embodiment 4 of the present invention.

[0030] Figure 4 The image shows the amplified waveform of the harmonic current step transient response under the condition of electrode rotation speed of 50 rpm.

[0031] Figure 5 This is an amplified waveform of the steady-state AC response of the harmonic current under the condition that the motor speed is 50 rpm.

[0032] Figure 6 This is an amplified waveform of the harmonic current step transient response under the condition of motor speed of 2000 rpm.

[0033] Figure 7 This is an amplified waveform of the steady-state AC response of the harmonic current under the condition of motor speed of 2000 rpm.

[0034] Figure 8 This is an amplified waveform of the harmonic current step transient response under the condition of motor speed of 10000rpm.

[0035] Figure 9This is an amplified waveform of the steady-state AC response of the harmonic current under the condition of motor speed of 10000rpm. Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and 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.

[0037] I. Decoupling Tracking Control Method for Fundamental Wave and Multi-order Harmonic Currents

[0038] like Figure 1 As shown, the decoupled tracking control method for fundamental and multi-order harmonic currents in this embodiment achieves independent decoupled control of the fundamental and multi-order harmonics using a single controller architecture. It compensates for high-frequency inductor roll-off through decoupling inverse modeling and achieves precise zeroing of the target point's phase and amplitude through accurate closed-loop parameters. This method is adaptable to various scenarios such as motor drives, power conversion, and grid compensation. The method is executed as follows: (I) Coordinate system establishment, current decomposition and object modeling First, a rotating orthogonal coordinate system synchronized with the fundamental wave is established, with the rotation angle of the rotating orthogonal coordinate system synchronized with the fundamental wave electrical angle. Depending on the type of controlled object, the fundamental wave angle and the fundamental wave angular frequency ω... base There are two types of ways to obtain them: When the controlled object is a motor or a motor simulator, the fundamental angular frequency is obtained synchronously from the motor's speed / position signal, and the electrical angle of the rotating orthogonal coordinate system is synchronized with the rotor electrical angle. When the controlled object is a converter or grid simulator connected to the power grid, the fundamental angle information and fundamental angular frequency are extracted through a phase-locked loop (PLL), quadrature signal generator or voltage / current sampling, and the rotating orthogonal coordinate system rotates synchronously according to this angle information.

[0039] In this rotating orthogonal coordinate system, the controlled current of the controlled object is decomposed into a fundamental current component and at least one target subharmonic current component. The fundamental current component rotates synchronously with the rotating orthogonal coordinate system and is represented as a DC quantity; the frequency of the target subharmonic current component is an integer multiple of the fundamental frequency, and it is represented as an AC quantity relative to the rotating orthogonal coordinate system. Its frequency w0 is calculated using the following formula: ; In the formula, k is the target harmonic order, which is dimensionless; ω baseThe fundamental angular frequency is expressed in rad / s. For example, the 5th negative harmonic and the 7th positive harmonic of a three-phase motor are both AC quantities that are 6 times the fundamental frequency in a synchronous rotating orthogonal coordinate system, corresponding to k=6.

[0040] The controlled object is then modeled as a transfer function matrix containing cross-coupling terms in a rotated orthogonal coordinate system. The voltage-current transfer function matrix G(s) of the controlled object in the dq rotated orthogonal coordinate system is as follows: ; Where s is the Laplace operator, and the unit is s. -1 (per second). The -1 in the upper right corner is for matrix inversion. This is the d-axis self-impedance transfer function. This is the dq-axis cross-coupling impedance. This is the qd-axis cross-coupling impedance. This is the q-axis self-impedance transfer function, with units of Ω.

[0041] Inverting the transfer function matrix G(s) yields the decoupling inverse model. : ; This decoupling inverse model is connected in series at the controller output. On the one hand, it cancels the cross-coupling terms of the controlled object itself through off-diagonal terms, thus achieving complete decoupling between the d-axis and the q-axis. On the other hand, it compensates for the roll-off effect of the controlled object's inductance impedance as the frequency increases through the inductance impedance terms in the diagonal terms, thereby fundamentally solving the problem of current tracking amplitude collapse in the ultra-high frequency band.

[0042] (ii) Constructing a current controller

[0043] like Figure 2 As shown, the low-pass internal mode branch is connected in parallel with at least one resonant-lead branch, and then the inverse model is decoupled in series to form a current controller. The resonant-lead branch is formed by connecting the resonant branch and the lead correction branch in series.

[0044] 1. Low-pass internal mold branch

[0045] The low-pass internal mode branch is used to calculate the fundamental error component in the current error signal, thereby achieving closed-loop control of the fundamental current component. Its transfer function is: ; In the formula, 'a' represents the design bandwidth of the low-pass inner mode branch, in rad / s.

[0046] The low-pass internal mode branch is a first-order internal mode control structure, possessing integral characteristics for DC currents and enabling zero steady-state error tracking of the fundamental current. When the relative deviation between the mathematical model parameters and the actual parameters of the controlled object does not exceed 5%, the fundamental closed-loop transfer function formed by the low-pass internal mode branch is equivalent to a first-order low-pass circuit. The dynamic response is stable, with no overshoot or oscillation.

[0047] 2. Resonant branch

[0048] The resonant branch is used to form a resonant frequency selection characteristic for the target subharmonic error component in the current error signal at the target subharmonic frequency w0. Its transfer function is: ; Among them, K r The resonant gain is dimensionless. c The resonant bandwidth is expressed in rad / s and satisfies w c =w0 / Q, where Q is the quality factor, which is dimensionless. By setting K... r =1, so that the gain of the resonant branch at the target subharmonic frequency w0 is . (This indicates that the output signal has a gain of 1 relative to the input signal and a phase deviation of 0). The resonant branch only produces high gain for AC signals at the target frequency, while the gain for the fundamental and other harmonic signals is extremely low, thus enabling independent control of multiple harmonics.

[0049] 3. Precise closed-loop calculation of lead correction branches and parameters

[0050] The lead compensation branch is connected in series after the resonant branch, and its transfer function is... for: ; In the formula, K l is the gain coefficient, used to compensate for amplitude attenuation, dimensionless; T is the time constant, in seconds, used for alignment compensation of the target frequency; b is the lead angle division coefficient, used to provide phase lead, dimensionless.

[0051] The three parameters b, T, and K of the lead compensation branch l The total transfer function at the target subharmonic frequency w0, after being superimposed in parallel with the resonant-leading branch and the low-pass internal mode branch, has an amplitude of 1 and a phase of 0, is determined by the following precise closed-form calculation: ; ; ; Three parameters K lThe calculation logic for T and b is as follows: First, calculate the lead angle division coefficient b based on the internal mode bandwidth a and the target subharmonic frequency w0, so that the lead phase provided by the lead compensation network exactly cancels the phase lag of the low-pass internal mode branch at w0; then calculate the time constant T based on w0 and b, so that the maximum lead phase point of the lead compensation network exactly falls at the target subharmonic frequency w0; finally, calculate the gain coefficient K. l This offsets the amplitude attenuation of the low-pass internal mode branch at w0, making the total amplitude after parallel connection exactly 1.

[0052] K in the lead correction branch l Compensation for T and b, after the resonant-lead branch and the low-pass internal mode branch are superimposed in parallel, the phase lag and amplitude deviation at w0 are precisely reduced to zero, and the total transfer function F satisfies the condition at the target subharmonic frequency. j is the imaginary unit, which is dimensionless.

[0053] (III) Multi-order harmonic spread control

[0054] When simultaneously tracking and controlling multiple target subharmonic current components, an independent resonant-leader branch is constructed for each target subharmonic current component; after all resonant-leader branches are connected in parallel, they are connected in series with the low-pass inner mode branch to the same decoupled inverse model.

[0055] Gain coefficient K of each resonant-leader branch l The time constant T and the lead angle division coefficient b are independently calculated and determined by substituting them into their respective target subharmonic frequencies, so that the phase lag and amplitude deviation at each target subharmonic frequency are both reduced to zero. This method does not require repeatedly establishing a rotating coordinate system for each order; independent control of multiple harmonics can be achieved simply through parallel branches.

[0056] (iv) Full-band segmented adaptation control

[0057] To address the issues of oscillation and large step ringing caused by low-frequency resonant branches, a full-frequency segmented adaptation strategy is adopted: When the target subharmonic frequency or fundamental angular frequency is lower than the preset threshold, two adaptation methods can be used: Method 1: Shut down all resonant-leader branches, and only keep the low-pass internal mode branch in closed-loop operation to avoid coupling oscillation between the low-frequency resonant branch and the internal mode branch, thereby improving system stability; Method 2: First, increase the design bandwidth 'a' of the low-pass internal mode branch to the target bandwidth to improve the low-frequency margin and response speed of the system. Then, enable all resonant-leader branches to balance the control accuracy and stability of the low-frequency band.

[0058] (v) Closed-loop tracking control

[0059] Using the current reference value of the target subharmonic current component as a reference signal, the difference between this reference signal and the actual controlled current component obtained from feedback is used to obtain the current error signal. The current error signal is input to the current controller, and after parallel calculations by the low-pass internal mode branch and each resonant-leading branch, and then after compensation calculations by the decoupled inverse model, the dq-axis voltage control quantity is output. The dq-axis voltage control quantity is converted into a multiphase voltage command in the stationary coordinate system through inverse coordinate transformation, and then a drive pulse is generated through pulse width modulation and applied to the power conversion circuit of the controlled object. This drives the power switching devices to turn on and off according to a preset timing sequence, thereby generating the fundamental current and each subharmonic current with corresponding amplitude and phase in the main circuit of the controlled object, and finally realizing the closed-loop tracking control of the target subharmonic current component.

[0060] The target subharmonic frequency w0 varies with the fundamental angular frequency ω base Synchronous changes are achieved by updating the parameters of each branch of the current controller accordingly with the target subharmonic frequency w0, ensuring tracking accuracy across the entire frequency band, as detailed below: Target subharmonic frequency w0 and fundamental angular frequency ω base Strictly meet The target subharmonic frequency w0 is directly proportional to the fundamental angular frequency ω. base The fundamental angular frequency changes synchronously and linearly with the change in motor speed. When the controlled object is a motor, the fundamental angular frequency is synchronized with the motor's electrical angular velocity and changes continuously from low speed to high speed. The corresponding target harmonic frequencies also increase or decrease synchronously in proportion. When the controlled object is grid-side equipment, the fundamental angular frequency fluctuates slightly with the grid power frequency, and the target harmonic frequencies also fluctuate synchronously.

[0061] To address the synchronous change in frequency, the parameters of each branch of the current controller are updated accordingly with the target subharmonic frequency w0, specifically including: (1) Resonant branch parameter update: The center resonant frequency of the resonant branch is directly set to the target subharmonic frequency w0, and the resonant bandwidth ω c =w0 / Q is adjusted synchronously with w0 to ensure that the quality factor Q remains unchanged at the design value, so that the resonant frequency selection characteristic is always accurately aligned with the current target subharmonic frequency, avoiding frequency deviation and detuning.

[0062] (2) Parameter update of lead compensation branch: the lead angle division factor b, time constant T, and gain factor K of the lead compensation branch. l The calculation is updated in real time by substituting the current target subharmonic frequency w0 into the precise closed-form formula. This ensures that at any operating frequency, the total transfer function of the resonant-leader branch and the low-pass internal mode branch connected in parallel always satisfies the ideal characteristics of amplitude 1 and phase 0 at the target subharmonic frequency, thus ensuring that the accuracy of phase compensation and amplitude compensation does not decrease with frequency changes.

[0063] (3) Decoupling inverse model parameter update: The cross-coupling term in the decoupling inverse model is related to the fundamental angular frequency and changes with ω. base Synchronous updates ensure that the cross-coupling between the d-axis and q-axis is always completely canceled, maintaining the independent decoupling control effect of the two phase axes.

[0064] Parameter updates can be implemented in two ways: first, online real-time calculation within each control cycle to adapt to operating conditions with rapid and continuous frequency changes; second, pre-calculation of parameter tables offline according to speed / frequency levels, with updates called by table lookup during runtime, adapting to scenarios with limited computing resources for digital processors. Through the mechanism of frequency synchronization tracking and corresponding parameter updates, the controller maintains accurate resonant frequency selection, complete amplitude and phase compensation, and stable decoupling effect across the entire operating frequency band from low to high frequencies, thereby ensuring harmonic current tracking accuracy across the entire frequency band.

[0065] (vi) Operating condition switching control

[0066] By simply adjusting the current reference value, the two operating modes of active injection and active suppression can be switched within the same controller architecture: When the current reference value is the preset target value of the target subharmonic current component, the active injection of the target subharmonic current component is realized, which is suitable for scenarios such as motor electromagnetic noise suppression that require the injection of specific harmonics. When the current reference value is zero or the reverse compensation value is reversed, the target subharmonic current component is actively suppressed. It is suitable for scenarios that require the elimination of harmonics, such as active power filtering, power factor correction ripple suppression, etc.

[0067] II. Verification of Implementation Examples

[0068] (I) Example 1: Active injection of multi-order harmonic current in motor drive (for electromagnetic noise suppression)

[0069] This embodiment uses a vehicle permanent magnet synchronous motor as the controlled object, with the goal of actively suppressing 24th-order electromagnetic noise. The generated SVPWM drive pulse is applied to the motor's three-phase inverter, which drives the power switch tubes of the drive bridge arm to turn on and off according to the command, outputting the corresponding voltage to the stator winding of the permanent magnet synchronous motor. The 5th and 7th harmonic currents with the target amplitude and phase are superimposed on the stator fundamental current to cancel the harmonic magnetic field components in the motor air gap, reduce the amplitude of the radial electromagnetic wave, and achieve active suppression of electromagnetic noise.

[0070] In the first-order dq synchronous rotating coordinate system, the fundamental current is a DC quantity; the fifth-order negative-sequence harmonic current and the seventh-order positive-sequence harmonic current in this coordinate system are respectively negative-sequence and positive-sequence ±6θ AC quantities, and the corresponding target harmonic frequencies satisfy w0=6ω. r , where ω r It is the electric angular velocity of the motor (i.e., the fundamental angular frequency, in rad / s).

[0071] At this point, the decoupling inverse model of the permanent magnet synchronous motor is: ; In the formula, L d L q These are the d-axis and q-axis inductances, respectively, in H; R s This is the stator resistance of the motor, measured in Ω.

[0072] The off-diagonal terms in the matrix are cross-coupling terms, reflecting the mutual coupling between the d-axis and the q-axis; the diagonal terms include sL d sL q This is the inductive impedance term, and its amplitude increases with frequency, which is the root cause of the current tracking amplitude collapse in the high-frequency band.

[0073] The design bandwidth of the low-pass internal mode branch is set to a = 800 rad / s, and the control period is set to dT = 1 × 10⁻⁴. 4 s, the quality factor of the resonant branch resonant bandwidth w c =w0 / Q, resonant gain K r =1. The lead correction parameters calculated using precise closed-loop calculation at different speeds are shown in Table 1.

[0074] Table 1 Lead Compensation Parameters

[0075] The corresponding harmonic current time-domain tracking waveform is as follows: Figures 4-9 As shown in the figures, the design bandwidth of the low-pass internal mode branch is uniformly set to a = 800 rad / s in all figures. The tracking performance of three control methods—a conventional proportional resonant scheme without a lead branch, a continuous domain lead compensation scheme, and a discrete domain lead compensation scheme—is compared across the entire speed domain. In the figures, the black curve represents the reference current, the red dashed line represents the conventional proportional resonant scheme without a lead branch, the blue solid line represents the continuous domain lead compensation scheme, and the green dashed line represents the discrete domain lead compensation scheme.

[0076] 1. Low-speed operation (50rpm, target subharmonic frequency 20Hz)

[0077] like Figure 4 The figure shows the amplified waveform of the harmonic current step transient response at 50 rpm (time interval 0.450s~0.750s). It can be seen from the figure that the scheme without a lead branch exhibits significant and continuous oscillations after the current step, with a large overshoot and slow convergence speed; while the output waveforms with lead compensation in both the continuous and discrete domains have stable step responses, small oscillation amplitudes, and can quickly converge to the reference current value.

[0078] like Figure 5The figure shows the amplified steady-state AC response waveform of the harmonic current under 50 rpm operating conditions (excluding the 100A DC bias, time interval 0.900s~1.000s). It can be seen from the figure that the scheme without a lead branch exhibits significant amplitude deviation and phase shift; while the output waveforms in both the continuous and discrete domains with lead correction highly coincide with the reference current, and the amplitude and phase errors are negligible.

[0079] 2. Medium-speed operation (2000rpm, target subharmonic frequency 800Hz)

[0080] like Figure 6 The figure shows the amplified waveform of the harmonic current step transient response at 2000 rpm (time interval 0.490s~0.560s). It can be seen from the figure that the scheme without a lead branch has begun to show amplitude decay and a slower response speed; the continuous and discrete schemes of this invention can still quickly track the step reference value, exhibiting excellent dynamic response performance.

[0081] like Figure 7 The figure shows the amplified steady-state AC response waveform of the harmonic current at 2000 rpm (excluding the 100A DC bias, time interval 0.945s~0.950s). As can be seen from the figure, the scheme without a lead branch exhibits significant amplitude attenuation and phase lag; the continuous domain scheme of this invention almost completely overlaps with the reference current, while the discrete domain scheme has only a very small discretization error, and its tracking accuracy is far superior to the scheme without a lead branch.

[0082] 3. High-speed operation (10000rpm, target subharmonic frequency 4000Hz)

[0083] like Figure 8 The figure shows the amplified waveform of the harmonic current step transient response at 10000 rpm (time interval 0.498s~0.510s). It can be seen from the figure that the scheme without a lead branch suffers severe amplitude attenuation and a significant decrease in response speed, making it unable to properly track step changes; while the continuous domain scheme of this invention has a fast step response, and its amplitude is close to the reference value; the discrete domain scheme also maintains good tracking performance.

[0084] like Figure 9 The figure shows the amplified steady-state AC response waveform of the harmonic current at 10000 rpm (excluding the 100A DC bias, time interval 0.949s~0.950s). It can be seen from the figure that the scheme without a lead branch suffers severe amplitude collapse and extreme phase lag, resulting in near-complete tracking failure. The continuous domain lead compensation scheme of this invention almost completely overlaps with the reference current, while the discrete domain scheme exhibits only negligible discretization error, maintaining high-precision tracking.

[0085] The above Figures 4-9The waveform comparison directly verifies that: the decoupled inverse model can effectively compensate for the high-frequency inductance roll-off effect of the controlled object; the precisely closed-loop design of the lead compensation branch can completely eliminate the phase lag and amplitude deviation after parallel connection; and after frequency pre-correction, the discretization scheme can maintain tracking accuracy close to that of the continuous domain scheme, achieving ideal performance across the entire operating frequency band from low speed to high speed. Tracking features.

[0086] (II) Example 2: Subharmonic Current Suppression of Power Factor Correction (PFC) Converter

[0087] This embodiment uses a power factor correction converter as the controlled object to suppress specific harmonic currents in the input current.

[0088] In an orthogonal coordinate system synchronized with the power grid frequency, the fundamental current of the PFC converter is a DC quantity, while the target subharmonic current to be suppressed is an AC quantity, corresponding to a target subharmonic frequency that is sufficient for... , where ω grid This is the grid angular frequency (i.e., the fundamental angular frequency, in rad / s).

[0089] Using the same controller structure as in Example 1, the decoupling inverse model is replaced with the decoupling inverse model of the PFC object in a synchronous orthogonal coordinate system. The current reference value is set to zero or an inverse compensation value, thus achieving active suppression of the harmonic current. If multiple target harmonics need to be suppressed simultaneously, multiple parallel resonant-leading branches are set up, with each branch corresponding to a target harmonic frequency.

[0090] (III) Example 3: Multi-order harmonic suppression of active power filter (APF)

[0091] This embodiment uses the grid-connected inverter of the active power filter as the controlled object to simultaneously compensate for multi-order harmonic currents in the power grid.

[0092] In an orthogonal coordinate system synchronized with the fundamental frequency of the power grid, for each target harmonic (e.g., 5th, 7th, 11th, and 13th), at their respective target harmonic frequencies... (w) 0,i Let k be the frequency of the i-th target harmonic, with the subscript i used to distinguish harmonics of different orders; i Construct a resonant-leader branch at each of the harmonic order coefficients corresponding to the i-th target harmonic; after multiple resonant-leader branches are connected in parallel, they share the same decoupling inverse model, and the current reference value is taken as the inverse compensation value, so that multiple harmonic currents can be suppressed at the same time.

[0093] (iv) Example 4: Discretization Implementation

[0094] To deploy the current controller within a digital controller, a frequency pre-correction discretization method is used to discretize the controller, ensuring that the discrete system is precisely consistent with the continuous system at the target subharmonic frequency w0.

[0095] The frequency pre-calibration discretization method employs a pre-sheared bilinear transform, with the transform relationship as follows: ; Where dT is the control period, z is the discrete-time Z-transform operator, which is dimensionless; K t is the pre-shear coefficient, which is dimensionless.

[0096] During the discretization process, the series resonant branch and the lead compensation branch are merged and discretized into a third-order digital filter. The combined term of the low-pass inner mode branch and the decoupled inverse model is discretized into a recursive formula containing integrals. Each coefficient is calculated online or updated offline by looking up a table according to the target subharmonic frequency w0.

[0097] The execution process of discrete deployment is as follows: Figure 3 As shown, the harmonic branch and the fundamental branch are calculated in parallel and independently. After completion, the total voltage is synthesized and then output for modulation. The discrete execution flow within each control cycle is as follows: 1. Parameter update: Calculate the target subharmonic frequencies based on the current fundamental angular frequency, and then calculate the pre-shearing coefficient, the coefficients of the third-order digital filter, and the coefficients of the recursive calculation formula; 2. Harmonic branch calculation: Input the current error signal into each third-order digital filter to calculate the voltage control quantity corresponding to each harmonic; 3. Calculation of fundamental branch: Input the current error signal into the recursive calculation formula containing the integral term to calculate the voltage control quantity corresponding to the fundamental branch; 4. Voltage Synthesis: The fundamental voltage control quantity and the voltage control quantities of each harmonic are synthesized to obtain the total voltage control quantity; 5. Output modulation: After the total voltage control quantity is transformed by inverse coordinates, it is modulated by SVPWM to generate drive pulse output.

[0098] (V) Example 5: Decoupling and tracking of fundamental and multi-order harmonics of a multiphase motor

[0099] This embodiment uses a five-phase permanent magnet synchronous motor as the controlled object, but it can also be applied to six-phase, double three-phase, and any motor with more than or equal to three phases.

[0100] The voltage-current relationship of a multiphase motor is decoupled and mapped onto multiple orthogonal planes (such as d1-q1, d2-q2, etc.) that rotate synchronously with the rotor. d1-q1 is the first set of orthogonal planes, typically corresponding to the principal plane containing the fundamental component; d2-q2 is the second set of orthogonal planes, corresponding to the sub-planes mapped to higher harmonic components. The fundamental component corresponds to the DC quantity in the d1-q1 plane, which is tracked and controlled by a low-pass internal mode branch. Each target harmonic component is represented as an AC quantity in its corresponding plane, and is tracked and controlled at the corresponding target harmonic frequency w using a decoupled tracking control method between the fundamental and higher harmonic currents. 0,i =k i ·ω base Each branch constructs an independent resonant-leader branch; after multiple resonant-leader branches are connected in parallel, they share the same decoupling inverse model, which is generalized to a block matrix form.

[0101] The block-form decoupled inverse model is isomorphic to the three-phase case in each orthogonal plane, and contains sL terms and cross-coupling terms. Therefore, the conclusion of this method that "a single controller structure can uniformly cover any number of phases and multiple harmonics" is directly valid, without the need to repeatedly build multiple synchronous rotating coordinate systems for each plane.

[0102] (vi) Example 6: Three-phase motor simulator / three-phase power grid simulator

[0103] (1) Three-phase motor simulator

[0104] Using a three-phase power amplifier as the actuator, the three-phase current (including the fundamental component and the target subharmonic components) of a simulated motor is reproduced. The electrical angle θ (in rad) of the dq rotating orthogonal coordinate system is synchronously obtained from the speed / position signal of the simulated motor, with no angle estimation error. The fundamental current is a DC quantity in the dq rotating orthogonal coordinate system and is tracked and controlled by the low-pass internal mode branch. The target subharmonic currents, such as the 5th and 7th, are folded into ±6θ AC quantities and tracked and controlled by the corresponding resonant-lead branch, realizing simultaneous closed-loop control of the three-phase current.

[0105] (2) Three-phase power grid simulator / three-phase PFC / three-phase APF

[0106] This type of controlled object differs from a motor; it does not provide speed signals or fundamental angle information θ. base The fundamental angular frequency ω is obtained through phase-locked loop (PLL) or voltage / current sampling extraction (such as dq-PLL, SOGI orthogonalization). base The power frequency is used. The rest of the controller structure and precise closed-loop lead parameters are completely consistent with those in Example 1, except that the decoupling inverse model is taken as the decoupling inverse model of the power grid side object in the synchronous orthogonal coordinate system.

[0107] It should be noted that the power grid frequency is approximately constant but not precisely known. The steady-state angle error of the phase-locked loop will enter the resonant point in the form of frequency shift. Therefore, in Examples 1 and 3, "ω" r Due to its advantages of "speed synchronization and no angle estimation error", the locking accuracy of the phase-locked loop is required to ensure its performance in the grid-side scenario.

[0108] (vii) Verification of implementation results

[0109] To verify the technical effectiveness of this invention, a simulation model of the controlled object was built in the MATLAB / Simulink environment, and different control schemes were compared and verified. The following two schemes were compared in detail: Ordinary proportional resonance control scheme (ordinary PR for short): resonant branches are decoupled in series inverse model (diagonal resonance), with no model prior compensation; The internal mode resonant decoupling control scheme of the present invention (hereinafter referred to as the scheme IMCR of the present invention): the resonant branch is connected in series with the lead compensation branch and then in parallel with the low-pass internal mode branch, and then in series with the inverse model for decoupling. The lead compensation parameters are determined according to the precise closed-form formula.

[0110] Table 2 shows the comparison of the amplitude ratio and phase lag of the 6th harmonic current closed-loop tracking at different speeds (sampling point alignment criterion): Table 2. Comparison of amplitude ratio and phase lag in closed-loop tracking of the 6th harmonic current.

[0111] Effect Analysis: Improved Tracking Performance in the Ultra-High Frequency Band: Compared to the ordinary PR scheme, this invention improves the tracking amplitude ratio from 0.266 to 1.000 in the ultra-high frequency band (10000 rpm), and reduces the phase lag from -74.2° to zero. Time-domain tracking waveform comparison shows that the outputs of both the continuous-band lead branch and the discrete-band lead branch precisely coincide with the reference harmonic current, while the scheme without a lead branch exhibits significant amplitude deviation and phase lag across the entire frequency band. This comparison verifies the effectiveness of the "resonant branch series decoupling inverse model" in compensating for the high-frequency inductance roll-off of the controlled object, and the effectiveness of lead correction in eliminating parallel phase lag—the ordinary PR scheme suffers amplitude collapse due to the loop gain decreasing with speed according to 1 / (w0L). This invention offsets this roll-off effect through the sL term in the decoupling inverse model, achieving full-band performance. Ideal tracking characteristics.

[0112] Precise elimination of phase and amplitude deviation: By using precisely closed-loop determined lead compensation parameters, the phase lag and amplitude deviation at the target subharmonic frequency, caused by the parallel superposition of the low-pass internal mode branch and the resonant branch, are completely eliminated. The continuous band lead branch satisfies the requirements at all three speeds. .

[0113] Low-speed stability optimization: After adopting a segmented control strategy that increases the internal mode bandwidth (a=800rad / s) or shuts down the resonant branch, the tracking performance under low-speed (50 / 100 / 200rpm) conditions is restored to a stable state, and the step ringing is significantly reduced.

[0114] III. Decoupling Tracking Control System for Fundamental Wave and Multi-order Harmonic Currents

[0115] The decoupling tracking control system for fundamental and multi-order harmonic currents in this embodiment is used to execute the above-described decoupling tracking control method for fundamental and multi-order harmonic currents. The system includes a decoupling inverse model module, a low-pass internal model branch module, a resonant branch module, a lead compensation branch module, a parallel and compensation module, and a control output module.

[0116] (a) Decoupling the inverse model module

[0117] The decoupling inverse model module is used to model the controlled object as a transfer function matrix containing cross-coupling terms in a rotating orthogonal coordinate system synchronized with the fundamental wave, and to obtain the decoupling inverse model of the transfer function matrix.

[0118] This module receives the corresponding fundamental angle and frequency signals based on the type of the controlled object (motor, converter, simulator, etc.) and updates the parameters of the decoupling inverse model in real time. For multiphase motors, this module outputs a decoupling inverse model in the form of a block matrix under a multidimensional orthogonal plane. The decoupling inverse model is used to cancel the cross-coupling between two phase shafts and to compensate for the high-frequency roll-off effect of the controlled object's inductive impedance.

[0119] (ii) Low-pass internal module branch module

[0120] The low-pass internal mode branch module is used to perform internal mode calculations on the fundamental error component in the current error signal, thereby realizing closed-loop control of the fundamental current component in the controlled current.

[0121] This module adopts a first-order low-pass internal mode structure, takes the current error signal as input, and outputs the fundamental voltage control quantity to achieve zero steady-state error tracking of the fundamental current.

[0122] The core parameter of the low-pass internal mode branch module is the design bandwidth 'a', which can be adjusted according to operating conditions to adapt to the segmented control strategy of the low-frequency band.

[0123] (III) Resonant Branch Module

[0124] The resonant branch module is used to form a resonant frequency selection characteristic at the target subharmonic error frequency for the target subharmonic error component in the current error signal.

[0125] Each target harmonic corresponds to a resonant branch module, which adopts a proportional resonant control structure. It generates high gain only for the AC component of the target frequency in the error signal, and has extremely low gain for other frequency components, thus realizing independent frequency selection control of multi-order harmonics.

[0126] The resonant frequency of the resonant branch module is updated synchronously with the target subharmonic frequency.

[0127] (iv) Lead correction branch module

[0128] The lead correction branch module is connected in series after the resonant branch module, and the parameters of its transfer function are calculated by the corresponding calculation formula.

[0129] Each resonant branch module corresponds to a lead compensation branch module. The module has three core parameters (lead angle division factor b, time constant T, and gain factor K). l The current target subharmonic frequency and internal mode bandwidth are calculated in real time to compensate for the phase lag and amplitude deviation at the target frequency after parallel connection.

[0130] (v) Parallel and Compensation Module

[0131] The parallel and compensation module is used to connect the resonant branch module and the lead correction branch module in series, then connect them in parallel with the low-pass internal mode branch module, and finally connect them in series with the decoupled inverse model module.

[0132] This module combines the signals from each branch, performs compensation calculations using a decoupled inverse model, and outputs the final voltage control value. Through parameter compensation of the lead correction branch module, the phase lag and amplitude deviation generated by parallel superposition are precisely reduced to zero at the target subharmonic frequency.

[0133] When controlling multiple target subharmonics simultaneously, the system includes multiple sets of resonant branch modules and lead compensation branch modules, each corresponding to multiple target subharmonic current components. Each set of modules determines the lead compensation parameters according to its own target subharmonic frequency. After multiple sets are connected in parallel, they are connected in series with the low-pass internal mode branch module to form the same decoupled inverse model module.

[0134] (vi) Control output module

[0135] The control output module is used to output control quantities to the controlled object with the current reference value of the target subharmonic current component as a reference signal, so as to realize closed-loop tracking control of the target subharmonic current component.

[0136] This module receives the current reference value and the actual current value fed back, calculates the current error signal, and sends it to each branch module in the preceding stage; at the same time, it receives the voltage control quantity output by the parallel and compensation modules, and after inverse coordinate transformation and pulse width modulation, generates a drive pulse to be applied to the controlled object to complete the closed-loop control.

[0137] (vii) Discretization of the system

[0138] When the system is deployed on a digital processor, the resonant branch module and the lead compensation branch module are combined to implement a third-order digital filter module; the low-pass internal mode branch module and the decoupled inverse model module are combined to implement a recursive calculation module with an integral term. The coefficients of the third-order digital filter and the coefficients of the recursive calculation formula are updated online in each control cycle or according to the set task cycle, depending on the change of the target subharmonic frequency, or updated in advance by looking up a table after offline calculation according to the speed / frequency gear.

[0139] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A decoupled tracking control method for fundamental and multi-order harmonic currents, characterized in that, Includes the following steps: In a rotating orthogonal coordinate system synchronized with the fundamental wave, the controlled object is modeled as a transfer function matrix containing cross-coupling terms, and the decoupling inverse model of the transfer function matrix is ​​obtained. The low-pass internal mode branch is connected in parallel with at least one resonant-lead branch, and then the inverse model is decoupled in series to form a current controller; wherein, the resonant-lead branch is formed by connecting the resonant branch and the lead correction branch in series. The real-time controlled current of the controlled object is collected and the difference between the current and the current reference value is used to obtain the current error signal. In the rotating orthogonal coordinate system, the current error signal is decomposed into the fundamental error component and at least one target subharmonic error component. The current error signal is input to the current controller. Each parallel branch performs independent calculations on the corresponding error components, and after compensation calculations of the decoupled inverse model, the output voltage control quantity is applied to the controlled object to form a negative feedback closed loop, thereby realizing decoupled tracking control of the fundamental wave and multi-order harmonic current.

2. The decoupling tracking control method for fundamental and multi-order harmonic currents according to claim 1, characterized in that, The resonant branch resonates at the target subharmonic frequency w0. k is the target harmonic order, ω base The fundamental angular frequency; the transfer function of the lead compensation branch. for: ; ; ; ; In the formula, s is the Laplace operator; K l b is the gain coefficient; T is the time constant; b is the lead angle resolution coefficient; a is the design bandwidth of the low-pass internal mode branch; With the resonant-lead branch and the low-pass internal mode branch connected in parallel, and the total transfer function at w0 having an amplitude of 1 and a phase of 0 as constraints, calculate K. l 、T、b.

3. The decoupling tracking control method for fundamental and multi-order harmonic currents according to claim 1, characterized in that, When the current reference value is the preset target value of the target subharmonic current component, the target subharmonic current component is actively injected; when the current reference value is zero or the reverse compensation value, the target subharmonic current component is actively suppressed.

4. The decoupling tracking control method for fundamental and multi-order harmonic currents according to claim 2, characterized in that, A frequency pre-correction discretization method is used to discretize the current controller, ensuring that the amplitude and phase frequency characteristics of the discretized current controller at the target subharmonic frequency w0 are consistent with the theoretical model of the current controller in the continuous time domain. The series resonant branch and the lead correction branch are combined and discretized into a third-order digital filter. The low-pass internal mode branch and the decoupled inverse model are combined and discretized into a recursive calculation formula containing an integral term. The third-order digital filter and the recursive calculation formula update the output value in each control cycle.

5. The decoupling tracking control method for the fundamental and multi-order harmonic currents according to claim 4, characterized in that, The discretization method for frequency pre-correction is a pre-sheared bilinear transform, and its transformation relationship is as follows: ; In the formula, dT is the control period, z is the discrete-time Z-transform operator, and K t is the pre-shearing coefficient; tan(·) is the tangent function.

6. The decoupling tracking control method for fundamental and multi-order harmonic currents according to claim 4, characterized in that, The coefficients of the third-order digital filter and the coefficients of the recursive calculation formula are updated online in each control cycle or according to the set task cycle, depending on the change of the target subharmonic frequency w0, or updated by looking up a table after offline calculation according to the speed / frequency gear.

7. The decoupling tracking control method for fundamental and multi-order harmonic currents according to claim 1, characterized in that, The controlled object can be any one of the following: permanent magnet synchronous motor, induction motor, multiphase motor, power factor correction converter, active power filter, grid-connected inverter, motor simulator, or grid simulator.

8. The decoupling tracking control method for fundamental and multi-order harmonic currents according to claim 1 or 7, characterized in that, When the controlled object is a multiphase motor, the transfer function matrix and the decoupling inverse model are generalized to a block matrix under a multidimensional orthogonal plane. The same set of parallel control architecture of low-pass internal mode branch and resonant-leading branch is reused in each orthogonal plane.

9. The decoupling tracking control method for fundamental and multi-order harmonic currents according to claim 1, characterized in that, When simultaneously tracking and controlling multiple target subharmonic current components, an independent resonant-lead branch is constructed for each target subharmonic current component; all resonant-lead branches are connected in parallel and then connected in series with the low-pass inner mode branch to the same decoupled inverse model; the gain coefficient K of each resonant-lead branch... l The time constant T and the lead angle division coefficient b are substituted into their respective target subharmonic frequencies and calculated independently to ensure that the phase lag and amplitude deviation at each target subharmonic frequency are both zero.

10. A decoupled tracking control system for fundamental and multi-order harmonic currents, characterized in that, The method for decoupling tracking control of fundamental and multi-order harmonic currents according to any one of claims 1-9 includes: The decoupling inverse model module is used to model the controlled object as a transfer function matrix containing cross-coupling terms in a rotating orthogonal coordinate system synchronized with the fundamental wave, and to obtain the decoupling inverse model of the transfer function matrix. The low-pass internal mode branch module is used for internal mode control of the fundamental error component; The resonant branch module is used to form a resonant frequency selection characteristic for the target subharmonic error component in the current error signal at the target subharmonic frequency. The lead correction branch module has a transfer function, and the parameters of the transfer function are calculated by the corresponding calculation formula; The parallel and compensation module is used to connect the resonant branch module and the lead compensation branch module in series, then connect them in parallel with the low-pass internal mode branch module, and then connect them in series with the decoupled inverse model module. Through parameter compensation of the lead compensation branch module, the phase lag and amplitude deviation generated by the parallel superposition are reduced to zero at the target subharmonic frequency. The control output module is used to output control quantities to the controlled object with the current reference value of the target subharmonic current component as a reference signal, so as to realize closed-loop tracking control of the target subharmonic current component.