On-line identification method and system for initial position of rotor of alternating current excitation system

Abstract

The invention discloses an AC excitation system rotor initial position on-line identification method and system, and the method comprises the steps: collecting a stator three-phase voltage signal when a stator is no-load, carrying out the Clarke conversion of the signal, obtaining a voltage component under a static coordinate system, and carrying out the low-pass filtering processing; calculating a stator flux linkage based on the stator current and the processed voltage component, and calculating a rotor flux linkage based on the rotor current; the rotor initial position offset is continuously accumulated through closed-loop adjustment by calculating the stator and rotor flux linkage angle and the phase difference until the phase difference approaches zero, and the accumulated rotor initial position offset is output; and combining the mechanical angle output by the encoder with the accumulated rotor initial position offset to obtain a corrected rotor position angle, and updating the corrected rotor position angle to an excitation system. According to the method, the in-phase characteristics of stator and rotor flux linkages are utilized, and the optimally designed low-pass filter is combined, so that the identification precision of the initial position of the rotor is remarkably improved, and the zero offset of the encoder is effectively compensated.

Description

Technical Field

[0001] This invention relates to the field of motor control system technology, and specifically to an online method for identifying the initial rotor position of an AC excitation system. Background Technology

[0002] Variable speed pumped storage systems, with their advantages of flexible power adjustment and efficient energy conversion, are important frequency regulation and peak shaving devices in modern power systems. The doubly-fed induction machine (DFIM), as its core drive motor, requires precise excitation through a rotor-side converter (RSC), and accurately obtaining the initial rotor position is a crucial prerequisite for achieving coordinate transformation.

[0003] Traditional rotor position identification methods mainly rely on offline testing, such as obtaining the initial angle through mechanical positioning devices or the induced electromotive force method with rotor open circuit. These methods have the following limitations:

[0004] 1) Reliance on complex wiring: Additional sensors need to be installed or the system needs to be disconnected for testing, which increases implementation costs and operational complexity;

[0005] 2) Limited speed range: Most methods are only applicable to low speed or stationary conditions, which makes it difficult to meet the online identification requirements under wide speed conditions;

[0006] 3) Insufficient immunity: It is susceptible to changes in motor parameters, voltage harmonics and integral drift, which leads to a decrease in identification accuracy.

[0007] Furthermore, existing encoder zero-position offset compensation technologies often neglect filter phase shift and integral stage errors, further reducing the dynamic performance of the excitation. Therefore, there is an urgent need for a wiring-free, high-precision, and highly disturbance-resistant online identification method across a wide speed range to improve the robustness and grid connection efficiency of dual-feedback asynchronous motor control systems. Summary of the Invention

[0008] To overcome the technical shortcomings of existing technologies where encoder zero-position offset compensation often ignores the errors of filter phase shift and integration, this invention provides an online method and system for identifying the initial position of the rotor in an AC excitation system.

[0009] To solve the above problems, the present invention is implemented according to the following technical solution:

[0010] In a first aspect, the present invention provides an online method for identifying the initial position of a rotor in an AC excitation system, comprising the following steps: Step S1: Under stator no-load conditions, the stator three-phase voltage signal is acquired and obtained by Clarke transformation to obtain the stationary position. Step S2: Calculate the stator flux linkage using the voltage integral method based on the stator current and the processed voltage component signal; calculate the rotor flux linkage using the current model based on the rotor current and rotor time constant; Step S3: Calculate the stator flux linkage angle and rotor flux linkage angle respectively based on the stator flux linkage and rotor flux linkage, and calculate their phase difference. Accumulate the rotor initial position offset through closed-loop adjustment until the phase difference approaches zero, at which point the accumulated rotor initial position offset is output; Step S4: Combine the mechanical angle output by the encoder with the accumulated rotor initial position offset to obtain the corrected rotor position angle, and update it to the excitation system.

[0011] In conjunction with the first aspect, the present invention provides a first specific implementation of the first aspect, wherein the low-pass filter processing of the voltage component specifically involves: using an amplitude-phase shift compensation module to dynamically compensate the filtered voltage component in order to correct the amplitude attenuation and phase delay introduced by the low-pass filter.

[0012] In conjunction with the first aspect, the present invention provides a second specific implementation of the first aspect. Specifically, the correction of amplitude attenuation and phase delay introduced by the low-pass filter is as follows: The formula for correcting amplitude attenuation introduced by the low-pass filter is: ;in, This is the amplitude compensation amount. The amplitude-frequency characteristic is that of a pure integral element. The amplitude-frequency response of the low-pass filter stage. It is the pure integral angular frequency. The cutoff frequency of the filter; the formula for correcting the phase delay introduced by the low-pass filter is: ;in, This is the phase compensation amount. The phase-frequency characteristic is that of a pure integral element. Phase frequency characteristics of the low-pass filter stage.

[0013] In conjunction with the first aspect, the present invention provides a third specific implementation of the first aspect, wherein the corrected rotor position angle is obtained by superimposing the mechanical angle output by the encoder with the accumulated rotor initial position offset.

[0014] Secondly, the present invention also provides an online identification system for the initial rotor position of an AC excitation system, comprising: a signal acquisition module for acquiring three-phase stator voltage signals under stator no-load conditions; and a signal processing module for performing Clarke transform on the acquired three-phase voltage signals to obtain a stationary position. The system includes a voltage component in a coordinate system, which is then low-pass filtered. A flux linkage calculation module calculates the stator flux linkage based on the stator current and the processed voltage component, and the rotor flux linkage based on the rotor current and rotor time constant. An angle processing module calculates the stator flux linkage angle and rotor flux linkage angle based on the stator and rotor flux linkages, respectively. A closed-loop adjustment module calculates the phase difference between the stator and rotor flux linkage angles, and continuously accumulates the rotor initial position offset through closed-loop adjustment until the phase difference approaches zero, outputting the accumulated rotor initial position offset. A position correction module combines the encoder-output mechanical angle with the accumulated rotor initial position offset to obtain the corrected rotor position angle. A control output module updates the corrected rotor position angle to the excitation system.

[0015] In conjunction with the second aspect, the present invention provides a first specific implementation of the second aspect. Specifically, the signal processing module includes: a dynamic compensation unit, used to dynamically compensate the filtered voltage component using an amplitude-phase shift compensation module to correct the amplitude attenuation and phase delay introduced by the low-pass filter.

[0016] In conjunction with the second aspect, the present invention provides a second specific implementation of the second aspect. Specifically, the flux linkage calculation module includes: a stator flux linkage calculation unit, which calculates the stator flux linkage using the voltage integration method; and a rotor flux linkage calculation unit, which calculates the rotor flux linkage based on the rotor current and the rotor time constant using a current model.

[0017] In conjunction with the second aspect, the present invention provides a third specific implementation of the second aspect. Specifically, the angle processing module includes an angle calculation unit, used to perform arctangent calculation on the stator flux linkage and the rotor flux linkage to obtain the stator flux linkage angle and the rotor flux linkage angle, respectively.

[0018] In conjunction with the second aspect, the present invention provides a fourth specific implementation of the second aspect. Specifically, the closed-loop adjustment module includes: a phase difference calculation unit for calculating the phase difference between the stator flux linkage angle and the rotor flux linkage angle; and a PI adjustment unit for continuously accumulating the rotor initial position offset through closed-loop adjustment until the phase difference approaches zero, and outputting the accumulated rotor initial position offset.

[0019] In conjunction with the second aspect, the present invention provides a fifth specific implementation of the second aspect. Specifically, the position correction module includes an angle superposition unit, used to superimpose the mechanical angle output by the encoder with the accumulated rotor initial position offset to obtain the corrected rotor position angle.

[0020] Compared with the prior art, the beneficial effects of the present invention are:

[0021] This invention calculates stator and rotor flux linkage based on stator voltage integration and rotor current under stator no-load conditions, achieving high-precision zero-position correction without offline testing. This simplifies system implementation and reduces maintenance costs. Feedback adjustment corrects the coordinate transformation angle to synchronize stator and rotor flux linkage phases, providing wide speed range applicability and enabling identification over a wide speed range. This supports rapid grid connection and stable operation of DFIM over a wide speed range. A low-pass filter replaces the pure integrator, and an amplitude-phase shift compensation module is designed to eliminate signal processing errors, resulting in strong anti-interference capabilities and effective resistance to motor parameter variations and voltage harmonics. This invention addresses the impact of integral drift; by combining encoder output and identification results to correct the rotor position angle in excitation in real time, it significantly improves the accuracy of rotor initial position identification and enhances the dynamic performance of excitation; by utilizing the in-phase characteristics of stator and rotor flux linkages and combining the optimized design of low-pass filters, it significantly improves the accuracy of rotor initial position identification, effectively compensates for encoder zero-position offset, and enhances the dynamic performance of excitation; this invention achieves high-precision zero-position correction without offline testing, and ensures the accuracy of rotor position angle through real-time calculation and feedback, providing a reliable foundation for the wide-speed operation of DFIM. Attached Figure Description

[0022] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings, wherein:

[0023] Figure 1 This is the DFIM system before no-load startup of the present invention.

[0024] Figure 2 This is a schematic diagram showing the difference between the absolute zero position of the encoder and the zero position of the DFIM rotor winding in this invention.

[0025] Figure 3 This is a schematic diagram of the online rotor initial position identification algorithm of the present invention. Detailed Implementation

[0026] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0027] like Figures 1-3 As shown, this invention provides a method and system for online identification of the initial rotor position in an AC excitation system.

[0028] In existing technologies, variable-speed pumped-storage systems rely on doubly-fed induction motors for efficient energy conversion and power regulation, while excitation accuracy is highly dependent on the accurate acquisition of the rotor's initial position. Traditional methods typically achieve position identification through offline testing, such as using mechanical positioning devices or induced electromotive force methods to measure the initial angle after disconnecting the system. However, these methods suffer from operational complexity and low testing efficiency, and are difficult to dynamically adjust position information during motor operation, resulting in limited control accuracy under actual operating conditions. Furthermore, insufficient phase shift and integral error compensation in existing technologies further impacts the system's dynamic response performance.

[0029] To address the aforementioned issues, the applicant observed that traditional offline testing cannot meet the requirements of wide-speed-range online identification and neglects dynamic error compensation in the signal processing stage. By analyzing the voltage and current signal characteristics under stator no-load conditions, it was found that the phase difference between the stator flux linkage and the rotor flux linkage can reflect the initial position offset of the rotor. Based on this, a closed-loop adjustment mechanism is proposed to dynamically correct the phase difference, while combining it with encoder output to achieve position correction. This approach overcomes the dependence of offline testing on system connectivity and solves the compensation error problem caused by filter phase shift.

[0030] Example 1

[0031] like Figure 1 As shown, the stator-side grid-connected circuit breaker is in the open-circuit state, and the grid-side converter (GSC) is directly connected to the power grid to provide a stable DC bus voltage for the rotor-side converter (RSC); the prime mover (hydro turbine) governor starts, controlling the speed at... A specific speed need not be specified here, as this identification can be performed over a wide speed range; only a fixed speed is required. The stator-side grid-connected circuit breaker should be kept open, ensuring no current flows through the stator. The RSC controls the rotor current based on the excitation principle; the rotor current command can be arbitrarily selected, as long as a constant terminal voltage is induced on the stator side. The rotor-side converter adjusts the output voltage of the rotor-side AC excitation system to achieve precise control of the stator terminal voltage, thereby completing the no-load start-up and grid-connected operation of the doubly-fed induction motor (DFIM). When implementing the excitation strategy, the rotor position angle of the DFIM needs to be accurately obtained to achieve coordinate system transformation; therefore, it is necessary to determine the zero-position angle offset between the encoder measurement value and the actual rotor position angle.

[0032] A method for online identification of the initial position of a rotor in an AC excitation system includes the following steps:

[0033] Step S1: Acquire the three-phase voltage signal of the stator under the stator no-load condition, obtain the voltage component in the stationary coordinate system through Clarke transformation, and perform low-pass filtering on the voltage component.

[0034] Step S2: Calculate the stator flux linkage based on the processed stator voltage components and stator current, and calculate the rotor flux linkage based on the rotor current;

[0035] Step S3: Calculate the stator flux linkage angle and rotor flux linkage angle based on the stator flux linkage and rotor flux linkage respectively, and calculate the phase difference between them. Through closed-loop adjustment, the rotor initial position offset is continuously accumulated until the phase difference approaches zero, and the accumulated rotor initial position offset is output.

[0036] Step S4: The mechanical angle output by the encoder is combined with the accumulated rotor initial position offset to obtain the corrected rotor position angle, which is then updated to the excitation system.

[0037] Clarke transform refers to the mathematical transformation that converts three-phase AC quantities into two-phase stationary coordinate system components. This can be implemented using matrix operations to eliminate the coupling characteristics of three-phase voltage signals. Signal processing includes filtering and dynamic compensation of voltage components. For example, a low-pass filter suppresses high-frequency noise, and an amplitude-phase shift compensation module corrects signal distortion caused by filtering, thus ensuring the accuracy of flux linkage calculation. Stator flux linkage calculation uses the voltage integration method, obtaining the flux linkage value by integrating the difference between stator voltage and current, reflecting the actual flux linkage state. Rotor flux linkage calculation is based on a current model, deriving the flux linkage using the product of rotor current and time constant, avoiding the influence of voltage signal fluctuations. Phase difference closed-loop adjustment dynamically adjusts the offset using a PI controller, for example, by adjusting the proportional coefficient and integral time constant to control the convergence speed, synchronizing the stator and rotor flux linkage angles. Position correction is achieved by superimposing the encoder output value and the offset, for example, by using an adder to synthesize mechanical and electrical angles, thereby eliminating initial installation errors.

[0038] Specifically, the three-phase voltage signal is collected under the stator no-load condition. By eliminating interphase coupling through Clarke transform, the voltage components in the two-phase stationary coordinate system are obtained. and The voltage component is then low-pass filtered to suppress high-frequency interference, and the signal amplitude and phase are restored through a dynamic compensation module. The processed voltage signal and stator current are integrated, and the stator flux linkage is obtained by combining this with stator resistance voltage drop compensation. Simultaneously, the rotor flux linkage is calculated based on the rotor current and the known rotor time constant. Arctangent calculations are performed on the two flux linkages to obtain the stator flux linkage angle and rotor flux linkage angle respectively. The phase difference between the two is calculated and input to the PI controller, which outputs the accumulated initial rotor position offset and dynamically adjusts the coordinate transformation angle until the phase difference approaches zero. Finally, the mechanical angle output by the encoder is superimposed with the offset to generate the corrected rotor position angle, which is then input to the excitation system. In other words, the stator and rotor flux linkage angles are passed through a comparator, and the rotor initial position offset is continuously accumulated through integration until the angle difference between the stator and rotor flux linkages converges to zero. At this point, the integrator no longer corrects the initial rotor position offset, the two flux linkages are synchronized, and the accumulated initial rotor position offset is output.

[0039] To address the technical shortcomings of existing compensation algorithms that do not adequately consider filter phase offset, this invention employs a dynamic compensation module to simultaneously correct signal amplitude attenuation and phase delay, significantly improving the accuracy of flux linkage angle calculation. It overcomes the technical limitations of traditional methods under low-speed conditions by using a collaborative calculation mechanism of the no-load voltage signal and rotor current model to achieve stable identification across the entire speed range from zero speed to rated speed. The system adopts an online initial position identification scheme, enabling position calibration without the need for additional sensors or system disconnection, greatly reducing the complexity of system operation. The innovative dynamic compensation mechanism effectively suppresses phase delay and amplitude attenuation introduced by filters in the signal processing stage, significantly improving the accuracy of flux linkage angle calculation. The optimized closed-loop adjustment algorithm enables the phase difference between stator and rotor flux linkages to converge quickly to zero, ensuring the accuracy and reliability of the initial position correction. The corrected rotor position signal generated through the angle superposition output mechanism improves the dynamic response time of the excitation system and enhances system operational stability.

[0040] In a preferred embodiment, for The voltage component in the coordinate system is subjected to low-pass filtering, and the amplitude-phase shift compensation module is used to dynamically compensate the filtered voltage component to correct the amplitude attenuation and phase delay introduced by the low-pass filtering.

[0041] Low-pass filtering refers to using a filter with a cutoff frequency between the fundamental frequency and the switching frequency to suppress high-frequency noise interference. This is achieved by setting a reasonable cutoff frequency and order to filter out switching harmonics and sampling noise. The amplitude-phase shift compensation module is a dynamic compensation stage designed based on the amplitude-frequency and phase-frequency characteristics of the filter's transfer function. This can be implemented using a lead compensation network or a digital filter. By inversely compensating for the amplitude attenuation and phase lag of the filter, the original dynamic characteristics of the voltage signal are restored. Specifically, after the stator three-phase voltage signal is transformed to a stationary coordinate system, the voltage components may contain harmonic components near the switching frequency and high-frequency noise. Directly calculating the flux linkage will lead to integration errors. Filtering the voltage components with a low-pass filter can effectively suppress high-frequency interference, but it introduces amplitude attenuation and phase delay. The dynamic compensation module, based on the filter's frequency response characteristics, adjusts the amplitude gain and performs phase lead compensation on the filtered voltage signal within the fundamental frequency range, eliminating the distortion effect of the filter on the effective signal and ensuring the accuracy of subsequent flux linkage calculations. By introducing a dynamic compensation module, the advantages of low-pass filtering are retained while the amplitude and phase deviation of the filtering stage on the effective signal is corrected, thereby improving both the signal-to-noise ratio and phase accuracy of the voltage signal. This solves the problem of inaccurate flux linkage observation caused by traditional filtering methods, improves the signal-to-noise ratio and phase consistency of the stator voltage signal, and provides an accurate signal basis for subsequent flux linkage calculation and rotor position identification, thus ensuring the dynamic response accuracy of the excitation system.

[0042] More specifically, the stator voltage, under no-load conditions, is in the form of high-frequency chopping. It needs to be filtered by a low-pass filter (LPF) to remove ripple and aliasing before being used for actual sampling control. The phase shift angle of the LPF at the fundamental frequency needs to be considered. The cutoff frequency of this LPF must be between the fundamental frequency and the switching frequency of the rotor-side converter. Let the phase shift at the fundamental frequency be denoted as... Secondly, pure integration of AC signals is highly susceptible to additional errors caused by the cumulative effects of sampling DC bias or drift. Therefore, an improved method is proposed that replace pure integration with low-pass filtering. The method also considers the compensation for amplitude and phase shift at the fundamental frequency by low-pass filtering compared to pure integration. Specifically:

[0043] Pure integration: Low-pass filter:

[0044] →

[0045] in, This is the amplitude compensation amount. The amplitude-frequency characteristic is that of a pure integral element. The amplitude-frequency response of the low-pass filter stage. It is the pure integral angular frequency. The cutoff frequency of the filter;

[0046] →

[0047] in, This is the phase compensation amount. The phase-frequency characteristic is that of a pure integral element. Phase frequency characteristics of the low-pass filter stage.

[0048] Taking a DFIM fundamental frequency of 50Hz as an example, and a low-pass filter with a cutoff frequency of 10Hz replacing pure integration, the compensation amounts for amplitude and phase shift are:

[0049]

[0050]

[0051] like Figure 3 As shown, stator voltage It is not pure integral but integral The low-pass filter, and multiplied The coefficient is used as the stator flux linkage. , and In the rotation transformation matrix converted to the dq coordinate system, the actual transformation angle is: By considering voltage sampling low-pass filtering and low-pass filter compensation that replaces pure integration, a more accurate identification angle can be obtained, thereby enabling DFIM to achieve better dq decoupling excitation performance.

[0052] In a preferred embodiment, the stator flux linkage is calculated using the voltage integration method, while the rotor flux linkage is calculated based on the rotor current and rotor time constant through a current model. The voltage integration method involves obtaining the flux linkage value by integrating the stator voltage components. Specifically, this can be achieved by performing time integration on the stator voltage in a stationary coordinate system. High-pass filtering or compensation can be introduced during the integration process to suppress DC bias errors. The current model refers to establishing a flux linkage observation equation based on the relationship between the rotor current and the rotor time constant. Specifically, this can be achieved by real-time acquisition of the rotor current signal and combining it with preset rotor inductance parameters for flux linkage estimation.

[0053] Specifically, under the no-load condition on the stator side, the stator current isd = isq = 0, therefore the DFIM flux linkage equation can be written as:

[0054] (1)

[0055] in: , , , These are the d-axis and q-axis components of the stator and rotor flux linkages, respectively. , Let L and L be the d-axis and q-axis components of the rotor current, respectively; Lm is the mutual inductance between the coaxial equivalent windings of the stator and rotor in the dq coordinate system; and Lr is the self-inductance between the equivalent two-phase windings of the rotor in the dq coordinate system. Equation (1) shows that the magnetic flux linkages of the stator and rotor are in phase at this time; at the same time, the steady-state stator magnetic flux linkage can be obtained by integrating the stator voltage in the stationary two-phase αβ coordinate system:

[0056] (2)

[0057] Based on the characteristic that the stator and rotor flux linkages are in phase when unloaded, the stator and rotor flux linkages are calculated separately using the stator terminal voltage and rotor current. By correcting the angle of coordinate transformation with the help of feedback, the initial position of the rotor can be identified online.

[0058] Specifically, under stator no-load conditions, the stator windings are not connected to load current, and the stator voltage signal includes a component induced by the rotor magnetic field. By continuously integrating the stator voltage using the voltage integration method, the trajectory of stator flux linkage changes can be directly reflected, avoiding the dependence of traditional current models on motor parameters. For rotor flux linkage calculation, a current model is used to directly correlate the physical relationship between rotor current and flux linkage. The phase delay in the current dynamic response process is corrected by the rotor time constant, thereby ensuring the real-time performance and accuracy of rotor flux linkage observation. The synchronous calculation of stator and rotor flux linkage forms a dual-channel observation mechanism, providing independent and complementary data sources for subsequent angle phase difference closed-loop adjustment. Through the synergistic application of the voltage integration method and the current model, high-precision integration is achieved on the stator side by utilizing the purity of the voltage signal under no-load conditions, while on the rotor side, the robustness of parameters is ensured by utilizing the current closed-loop control characteristics, forming a complementary and error-correcting flux linkage observation system. It effectively overcomes the integral drift and parameter sensitivity problems of the single flux linkage observation method. The stator flux linkage calculation does not depend on the rotor parameters, and the rotor flux linkage calculation avoids voltage harmonic interference. The dual-model collaborative operation improves the flux linkage observation accuracy over a wide speed range, provides reliable input for subsequent angle closed-loop adjustment, and ensures the stability and anti-interference of the initial position identification.

[0059] In a preferred embodiment, the method of calculating the phase difference based on the stator flux linkage and the rotor flux linkage and outputting the initial position offset through closed-loop adjustment includes performing arctangent calculation on the stator flux linkage and the rotor flux linkage to obtain the stator flux linkage angle and the rotor flux linkage angle, calculating the phase difference between them, continuously accumulating the initial rotor position offset through closed-loop adjustment until the phase difference approaches zero, and outputting the accumulated initial rotor position offset.

[0060] Specifically, arctangent calculation refers to converting the Cartesian coordinate components of the flux linkage into polar coordinate angle values ​​using mathematical functions. This can be achieved using the CORDIC algorithm or a lookup table method, and is used to extract the angle information of the flux linkage space vector from the α and β components of the stator and rotor flux linkages. Phase difference is the difference between the stator and rotor flux linkage angles, achieved through angle subtraction, and is used to characterize the spatial position deviation between the stator and rotor flux linkages. The PI controller, or proportional-integral controller, can be implemented using a discretization algorithm and is used to generate control signals based on the phase difference to eliminate steady-state errors. Dynamic adjustment of the coordinate transformation angle refers to real-time correction of the reference angle for coordinate transformation based on the initial position offset. This can be achieved through angle superposition operations, and is used to ensure the synchronization of the excitation system during phase difference closed-loop adjustment.

[0061] Specifically, the stator flux linkage angle and rotor flux linkage angle are extracted from the α and β axis components of the flux linkage using an arctangent calculation module. These two angles are then input into a subtractor to obtain a phase difference signal. This phase difference signal is sent to a PI controller for closed-loop regulation, outputting an initial position offset. This initial position offset is fed back to the coordinate transformation module, which corrects the coordinate transformation reference angle in real time by superimposing the encoder's mechanical angle, thereby forcing the phase difference to gradually decrease to zero, i.e., the stator and rotor flux linkages achieve synchronization. During this process, the proportional gain and integral time constant of the PI controller can be adjusted according to the system's dynamic characteristics, for example, by using the critical proportional method or trial-and-error method for parameter tuning, to ensure rapid convergence of the phase difference without overshoot.

[0062] In a preferred embodiment, the corrected rotor position angle is obtained by superimposing the mechanical angle output by the encoder with the accumulated rotor initial position offset.

[0063] The mechanical angle refers to the actual rotation angle of the rotor directly measured by the encoder, which serves as the fundamental angle signal to reflect the rotor's physical position. The initial position offset refers to the phase difference correction amount output through closed-loop regulation, specifically generated using error integral or proportional-integral control algorithms, used to compensate for encoder installation errors or zero-position drift during dynamic operation. Superposition refers to the algebraic addition of the mechanical angle and the initial position offset, which can be achieved through an adder or digital signal processing module, thereby directly applying the dynamic compensation amount to the fundamental angle signal to eliminate accumulated errors.

[0064] Specifically, during system operation, the mechanical angle output by the encoder in real time serves as the basic reference value for the rotor position, while the initial position offset is dynamically generated by the closed-loop control module based on the phase difference between the stator and rotor flux linkages. The two are then superimposed to obtain the corrected rotor position angle. This process does not require disconnecting the system or installing additional sensors; it directly uses the offset output by the closed-loop control to correct the encoder signal online, thereby eliminating zero-position errors caused by filter phase shift, integral drift, or mechanical installation deviations. The execution frequency of the superposition operation can be synchronized with the main loop of the control system, for example, updating once every millisecond or microsecond, ensuring that the corrected rotor position angle reflects the rotor's true electrical angle position in real time.

[0065] Example 2

[0066] An online rotor initial position identification system for an AC excitation system includes: a signal acquisition module for acquiring three-phase stator voltage signals under stator no-load conditions; a signal processing module for performing Clarke transform on the acquired three-phase voltage signals to obtain voltage components in a stationary coordinate system, and performing low-pass filtering on the voltage components; a flux linkage calculation module for calculating stator flux linkage based on the processed voltage components and stator current, and calculating rotor flux linkage based on rotor current and rotor time constant; an angle processing module for calculating stator flux linkage angle and rotor flux linkage angle respectively based on the stator flux linkage and rotor flux linkage; a closed-loop adjustment module for calculating the phase difference between the stator flux linkage angle and the rotor flux linkage angle, and continuously accumulating the rotor initial position offset through closed-loop adjustment until the phase difference approaches zero, and outputting the accumulated rotor initial position offset; a position correction module for combining the mechanical angle output by the encoder with the accumulated rotor initial position offset to obtain the corrected rotor position angle; and a control output module for updating the corrected rotor position angle to the excitation system.

[0067] The signal acquisition module is a device that acquires the three-phase stator voltage signal under stator no-load conditions. Specifically, it can be implemented using a voltage transformer in conjunction with an AD sampling circuit to acquire the raw voltage data during motor operation in real time. The signal processing module includes a Clarke transform unit and a filtering compensation unit. The Clarke transform unit converts the three-phase voltage into two-phase stationary coordinate system components using a coordinate transformation algorithm. The flux linkage calculation module includes a voltage integration unit and a current model solution unit. The voltage integration unit uses an integrator with a limiting circuit to process the stator voltage components, while the current model solution unit calculates the flux linkage based on rotor current sensor data and a preset rotor time constant, used to synchronously acquire stator and rotor flux linkage. The angle processing module includes a dual-channel arctangent calculator that uses the CORDIC algorithm to solve for the stator and rotor flux linkage angles separately. The phase difference calculation unit uses a digital subtractor to measure the instantaneous difference between the two angles. The closed-loop control module uses a proportional-integral controller to dynamically adjust the phase difference, and the coordinate transformation adjustment unit at the output end corrects the coordinate transformation parameters in real time using an incremental angle superposition algorithm. The position correction module includes an angle superposition unit to achieve arithmetic superposition of the encoder's raw data and the offset.

[0068] Specifically, the signal acquisition module acquires three-phase voltage signals when the motor is running under no-load conditions. After coordinate transformation by the signal processing module, the α-β axis voltage components are obtained. The filtering and compensation unit performs low-pass filtering on the voltage signals and uses a complex gain compensation module to correct for amplitude attenuation and phase lag. In the flux linkage calculation module, the stator flux linkage is calculated by integrating the voltage components and multiplying them by the stator current, while the rotor flux linkage is calculated by multiplying the rotor current and the rotor time constant. The angle processing module inputs the stator and rotor flux linkage components into a dual-channel arctangent arithmetic unit and outputs the angle difference to the PI controller to generate the position offset. The position correction module superimposes the mechanical angle output by the encoder with the offset and updates the coordinate transformation unit of the excitation system through the control output module.

[0069] In a preferred embodiment, the signal processing module includes a low-pass filtering unit and a dynamic compensation unit. The low-pass filtering unit is used to perform low-pass filtering on the voltage components in the coordinate system. The dynamic compensation unit is used to dynamically compensate the filtered voltage components using an amplitude-phase shift compensation module to correct the amplitude attenuation and phase delay introduced by the low-pass filtering.

[0070] Low-pass filtering refers to eliminating high-frequency noise interference by using a filter with a set cutoff frequency. This can be achieved using Butterworth or Chebyshev filters to suppress switching harmonics and random noise in the acquired signal. Dynamic compensation refers to establishing a reverse compensation model based on the filter transfer function. This can be achieved using frequency response-based amplitude gain adjustment and phase lead correction algorithms to counteract signal distortion caused by the low-pass filtering stage.

[0071] Specifically, after the stator voltage signal is transformed by Clarke to obtain the stationary coordinate system component, a second-order low-pass filter is first used to suppress high-frequency noise in the voltage component. The cutoff frequency of the filter is between the fundamental frequency and the switching frequency of the RSC. Subsequently, based on the deviation between the actual phase lag of the filter and the design parameters, a compensation coefficient is calculated in real time by a digital signal processor to proportionally amplify and compensate the amplitude of the filtered voltage. At the same time, a phase lead angle correction is superimposed, so that the final output voltage signal retains the effective fundamental component and restores the original phase characteristics.

[0072] In a preferred embodiment, the flux linkage calculation module includes a stator flux linkage calculation unit and a rotor flux linkage calculation unit. The stator flux linkage calculation unit calculates the stator flux linkage using the voltage integration method, while the rotor flux linkage calculation unit calculates the rotor flux linkage based on the rotor current and rotor time constant using a current model. The voltage integration method refers to solving for the flux linkage by integrating the stator voltage components with the stator current. Specifically, this can be achieved by subtracting the resistance voltage drop from the stator terminal voltage and then performing time integration. This method can correct for DC offset and noise interference during the integration process. The rotor time constant is a parameter reflecting the electromagnetic inertia of the rotor windings. It can be obtained from the motor nameplate parameters or offline testing. This parameter is used to construct the current model to eliminate the influence of rotor resistance changes on flux linkage observation.

[0073] Specifically, the stator flux linkage calculation unit acquires the stator voltage components in a stationary coordinate system, compensates for the voltage drop by combining the stator winding resistance parameters, and then calculates the stator flux linkage through an integrator. The rotor flux linkage calculation unit calculates the rotor flux linkage in real time within a current model based on the actual measured value of the rotor current and a pre-calibrated rotor time constant. The calculation results of both flux linkages are synchronously transmitted to the angle processing module for subsequent phase difference calculation and position correction.

[0074] In a preferred embodiment, the angle processing module and the closed-loop adjustment module include a phase difference calculation unit for calculating the phase difference between the stator flux linkage angle and the rotor flux linkage angle; and a PI adjustment unit for continuously accumulating the rotor initial position offset through closed-loop adjustment until the phase difference approaches zero, and outputting the accumulated rotor initial position offset.

[0075] Among them, arctangent calculation refers to the calculation of magnetic flux linkage. The axis components are converted into polar coordinate angle values, which can be implemented using the CORDIC algorithm or lookup table method, and used to extract the electrical angle from the flux linkage component. Phase difference calculation involves performing algebraic subtraction on two angle signals, which can be implemented using the arithmetic logic unit of a digital signal processor, and is used to quantify the deviation of the stator and rotor flux linkage angles. The PI control unit refers to a proportional-integral controller, which can be implemented using discretized difference equations, and is used to convert the phase difference error signal into a position offset correction value. The coordinate transformation adjustment unit dynamically corrects the angle parameters of the Park transform or inverse transform, which can be implemented through online updating of the rotation matrix coefficients, and is used to eliminate coordinate transformation errors caused by initial position offsets.

[0076] Specifically, under stator no-load operation, the phase difference calculation unit continuously acquires the stator flux linkage and rotor flux linkage. The axial components are used to obtain the stator flux linkage angles through arctangent calculation. and rotor flux angle The phase difference calculation unit performs real-time calculations. The difference is then input into the PI control unit. The PI control unit adjusts the PI control according to the preset adjustment parameters. Perform integral and proportional calculations to output the initial position offset. The coordinate transformation adjustment unit will... Superimposed on the current coordinate transformation angle This forms a closed-loop control circuit, forcing It approaches zero infinitely, thus achieving automatic calibration of the rotor's initial position.

[0077] In a preferred embodiment, the position correction module includes an angle superposition unit, which is used to superimpose the mechanical angle output by the encoder. The corrected rotor position angle is obtained by superimposing the accumulated initial rotor position offset. .

[0078] Among them, the mechanical angle refers to the actual physical angle of the rotor directly measured by the rotary encoder, which can be acquired in real time by a photoelectric encoder or magnetic encoder. This parameter reflects the actual mechanical rotation position of the rotor. The initial position offset refers to the angle compensation amount generated by closed-loop adjustment, which can be achieved by using the angle error integral output of the PI controller. Its function is to eliminate the coordinate transformation error caused by the magnetic flux phase difference. Angle superposition refers to the algebraic addition of two angle values, which can be implemented by the adder unit of the digital signal processor. This operation can fuse the encoder measurement value and the closed-loop compensation value into the final rotor position angle for control.

[0079] Specifically, during the rotor position angle correction process, the mechanical angle signal output by the encoder is input to the angle superposition unit, while the initial position offset output by the closed-loop adjustment module is also input synchronously. The superposition operation uses the mechanical angle as the reference value, and after superimposing the offset, a new rotor position angle signal is generated. This process continues dynamically, and when the phase difference converges to zero, the superimposed position angle is the accurate rotor position information that eliminates the zero-position deviation and integral error.

[0080] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Therefore, any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for online identification of the initial position of a rotor in an AC excitation system, characterized in that, Includes the following steps: Step S1: Under stator no-load conditions, acquire the stator three-phase voltage signal, and obtain the static voltage signal through Clarke transformation. The voltage component in the coordinate system is then subjected to low-pass filtering. Step S2: Calculate the stator flux linkage using the voltage integration method based on the stator current and the processed voltage component signal; calculate the rotor flux linkage using the current model based on the rotor current and rotor time constant. Step S3: Calculate the stator flux linkage angle and rotor flux linkage angle based on the stator flux linkage and rotor flux linkage respectively, and calculate the phase difference between them. Through closed-loop adjustment, the rotor initial position offset is continuously accumulated until the phase difference approaches zero, and the accumulated rotor initial position offset is output. Step S4: The mechanical angle output by the encoder is combined with the accumulated rotor initial position offset to obtain the corrected rotor position angle, which is then updated to the excitation system.

2. The method for online identification of the initial rotor position of an AC excitation system according to claim 1, characterized in that, The low-pass filtering process for the voltage component further includes: An amplitude-phase shift compensation module is used to dynamically compensate the filtered voltage components in order to correct the amplitude attenuation and phase delay introduced by the low-pass filter.

3. The method for online identification of the initial rotor position of an AC excitation system according to claim 2, characterized in that, The amplitude attenuation and phase delay introduced by the modified low-pass filter are specifically as follows: The amplitude attenuation formula introduced by the modified low-pass filter is as follows: ； in, This is the amplitude compensation amount. The amplitude-frequency characteristic is that of a pure integral element. The amplitude-frequency response of the low-pass filter stage. It is the pure integral angular frequency. The cutoff frequency of the filter; The formula for correcting the phase delay introduced by the low-pass filter is as follows: ； in, This is the phase compensation amount. The phase-frequency characteristic is that of a pure integral element. This represents the phase frequency characteristic of the low-pass filter stage.

4. The method for online identification of the initial position of an AC excitation system rotor according to claim 1, characterized in that: The corrected rotor position angle is obtained by superimposing the mechanical angle output by the encoder with the accumulated rotor initial position offset.

5. An online rotor initial position identification system for an AC excitation system, characterized in that, include: The signal acquisition module is used to acquire the three-phase voltage signals of the stator under no-load conditions. The signal processing module is used to perform Clarke transform on the acquired three-phase voltage signals to obtain the static signal. The voltage component in the coordinate system is then subjected to low-pass filtering. The flux linkage calculation module is used to calculate the rotor flux linkage based on the stator current and the processed voltage component signal, and based on the rotor current and the rotor time constant. An angle processing module is used to calculate the stator flux linkage angle and the rotor flux linkage angle based on the stator flux linkage and the rotor flux linkage, respectively. The closed-loop adjustment module is used to calculate the phase difference between the stator flux linkage angle and the rotor flux linkage angle. Through closed-loop adjustment, the rotor initial position offset is continuously accumulated until the phase difference approaches zero, and the accumulated rotor initial position offset is output. The position correction module is used to combine the mechanical angle output by the encoder with the accumulated initial rotor position offset to obtain the corrected rotor position angle; The control output module is used to update the corrected rotor position angle to the excitation system.

6. The online rotor initial position identification system for an AC excitation system according to claim 5, characterized in that, The signal processing module includes: The dynamic compensation unit is used to dynamically compensate the filtered voltage components using the amplitude-phase shift compensation module, in order to correct the amplitude attenuation and phase delay introduced by the low-pass filter.

7. The online rotor initial position identification system for an AC excitation system according to claim 5, characterized in that, The magnetic flux calculation module includes: The stator flux linkage calculation unit uses the voltage integration method to calculate the stator flux linkage; The rotor flux linkage calculation unit calculates the rotor flux linkage based on the rotor current and rotor time constant using a current model.

8. The online rotor initial position identification system for an AC excitation system according to claim 5, characterized in that, The angle processing module includes: An angle calculation unit is used to perform arctangent calculations on the stator flux linkage and rotor flux linkage to obtain the stator flux linkage angle and rotor flux linkage angle, respectively.

9. The online rotor initial position identification system for an AC excitation system according to claim 5, characterized in that, The closed-loop control module includes: The phase difference calculation unit is used to calculate the phase difference between the stator flux linkage angle and the rotor flux linkage angle; The PI control unit is used to continuously accumulate the rotor initial position offset through closed-loop control until the phase difference approaches zero, and then outputs the accumulated rotor initial position offset.

10. The online rotor initial position identification system for an AC excitation system according to claim 5, characterized in that, The position correction module includes: An angle superposition unit is used to superimpose the mechanical angle output by the encoder with the accumulated rotor initial position offset to obtain the corrected rotor position angle.

Images