Nonlinear flux linkage reactanceless control method of dual three-phase PWM rectifier

Abstract

The application discloses a kind of nonlinear flux linkage non-inductive control methods of double three-phase PWM rectifier, the method is introduced based on the mathematical model of permanent magnet synchronous motor flux linkage SMENFO, under the condition of no position sensor, the anti-interference ability and convergence performance of observer under the condition of external disturbance and parameter uncertainty are improved;On the basis of flux linkage observer, CESCO is constructed, respectively to the disturbance of double three-phase PWM rectifier control system and the internal disturbance of observer is estimated and compensated in real time, to realize the adaptive inhibition of external disturbance, and dynamically compensate the influence caused by system parameter variation, improve the overall control precision and robustness of system;ADRC and ESO PLL are combined, construct DC-SPR-ESO-PLL, further improve the continuity and accuracy of the accuracy of alternating current side synchronization and rotor position estimation.The application can keep stable and reliable non-inductive control performance under winding disturbance or local abnormal condition.

Description

Technical Field

[0001] This invention belongs to the field of power electronic conversion and motor drive control technology, and relates to a sensorless control method for a dual three-phase PWM rectifier. Background Technology

[0002] Dual three-phase PWM rectifiers are widely used in new energy power generation grid-connected systems, aviation electric propulsion systems, electric aircraft, electric drones, and high-reliability industrial drives due to their high power density, high efficiency, good dynamic performance, and strong fault tolerance.

[0003] In the operation of dual three-phase PWM rectifiers, precise control of the rectifier input current, grid-side power factor, and DC bus voltage is required to achieve high-quality power conversion. Traditional control methods mainly include vector control, direct power control, and decoupled control strategies based on space vector modulation. In these methods, rotor position information or synchronous reference frame angle typically needs to be detected in real time by position sensors. However, in practical applications, the installation of traditional position sensors is often limited by space constraints, especially in high-power-density applications such as aerospace electric propulsion and electric aircraft, where sensor placement is difficult. Furthermore, position sensors are susceptible to vibration, temperature changes, and electromagnetic interference during long-term operation, resulting in high failure rates and insufficient reliability, which seriously affects the safe operation of the rectifier. Therefore, to reduce system costs and improve reliability, sensorless control technology has gradually become a research hotspot. Sensorless control can estimate rotor position information by observing back EMF or flux linkage. Compared to back EMF observation, flux linkage observation maintains more stable signal characteristics under low-speed and dynamic conditions, and is less affected by noise and parameter changes, thus improving position estimation accuracy and system robustness, making it more suitable for the high-performance control requirements of dual-three-phase PWM rectifiers. Dual-three-phase PWM rectifiers operate under complex conditions, necessitating the development of a nonlinear flux linkage observer with stronger robustness and dynamic performance. Achieving high-precision estimation of flux linkage and rotor position through nonlinear observation algorithms is crucial for improving the sensorless control performance of dual-three-phase PWM rectifiers. However, although existing sensorless control methods for dual-three-phase PWM rectifiers have achieved flux linkage estimation and position recognition to a certain extent and obtained good control results, the current technology still has the following shortcomings in complex operating conditions and high-dynamic application scenarios:

[0004] 1. Insufficient robustness and disturbance rejection capability of sensorless systems

[0005] Traditional sensorless flux control methods typically rely on the accuracy of the system's mathematical model and assume relatively stable grid parameters, load conditions, and operating environment. However, in actual operation, dual three-phase PWM rectifiers are often affected by external disturbances, such as grid voltage fluctuations, load abrupt changes, harmonic interference, and switching noise. These disturbances increase flux observation errors, leading to deviations in sensorless estimation results. Existing methods lack adaptive adjustment capabilities when facing complex disturbances, easily resulting in decreased control accuracy, deteriorated dynamic response, and even system oscillations or instability, making it difficult to meet the requirements of high-reliability applications.

[0006] 2. Insufficient rotor position recognition accuracy

[0007] In sensorless control systems, the accuracy of rotor position recognition directly determines the effectiveness of current decoupling control and the rectifier's operating performance. Existing flux linkage observers or back EMF estimation methods are easily affected by factors such as small signal amplitude and strong noise interference under low- and medium-speed conditions, resulting in insufficient position recognition accuracy. Simultaneously, under rapidly changing dynamic conditions, rotor position estimation often exhibits phase lag or error accumulation, leading to current control deviations, power factor decline, and output voltage fluctuations. Especially in high-dynamic scenarios such as aerospace electric propulsion and electric aircraft, the requirements for position recognition accuracy are even higher, and traditional methods struggle to meet the demands for real-time performance and precision.

[0008] 3. Insufficient utilization of redundancy characteristics in dual- and three-phase systems

[0009] Dual three-phase PWM rectifiers possess a naturally redundant multiphase structure, providing more voltage and current information compared to traditional three-phase systems. This offers richer observational data for rotor position estimation in sensorless control. Since the dual three-phase system comprises two independent yet coupled three-phase windings, their voltage and current information can be used for rotor position identification and flux linkage estimation respectively. When one set of observations is disturbed or fails, the control system can still utilize the electrical quantities of the other healthy winding to achieve effective rotor position estimation, thereby improving the reliability of position identification and system fault tolerance. Existing control strategies still do not fully utilize this redundant information and fail to fully leverage the advantages of the dual three-phase structure in improving position estimation accuracy and enabling continuous operation under fault conditions. Therefore, there is still room for further optimization and improvement. Summary of the Invention

[0010] This invention aims to provide a nonlinear flux sensorless control method for dual three-phase PWM rectifiers to improve the control accuracy and robustness of the system under external disturbances, parameter fluctuations, and high dynamic load changes. Especially in high-dynamic, high-precision applications such as low-altitude economic operation, this method can effectively accelerate the response speed of dual three-phase PWM rectifiers, enhance operational stability, and meet the requirements for high-quality power conversion control under complex operating conditions.

[0011] The objective of this invention is achieved through the following technical solution:

[0012] A nonlinear flux sensorless control method for a dual three-phase PWM rectifier includes the following steps:

[0013] Step 1: Provide a mathematical model of the flux linkage of a dual-winding permanent magnet synchronous motor and a mathematical model of a dual three-phase PWM rectifier, and establish a reference model for flux linkage state observation.

[0014] Step 2: Based on the mathematical model of flux linkage of a dual-winding permanent magnet synchronous motor, the sliding mode approach law is introduced to design a sliding-mode enhanced nonlinear flux linkage observer (SMENFO). The high-precision sensorless estimation of flux linkage state and rotor position information is achieved through the observer output.

[0015] Step 3: Construct a Cascaded Extended State Observer (CESO) based on SMENFO to estimate external disturbances and internal disturbances of the dual three-phase PWM rectifier control system in real time, and perform dynamic compensation based on the estimation results, thereby realizing adaptive disturbance suppression and parameter uncertainty compensation.

[0016] Step 4: Combining Active Disturbance Rejection Control (ADRC) with Extended State Observer Based Phase Locked Loop (ESO-PLL), establish a Disturbance Compensation and Secondary Position Refinement Extended State Observer Based Phase Locked Loop (DC-SPR-ESO-PLL) structure to achieve high-precision tracking of AC-side synchronization angle and frequency, and improve the system's dynamic response capability.

[0017] Step 5: Utilize the multiphase redundancy structure characteristics of the dual three-phase PWM rectifier to collect voltage and current information of two sets of three-phase windings for flux estimation and rotor position identification. When the observation information of one set of windings is interfered with or fails, automatically switch or merge the electrical quantities of the other healthy winding.

[0018] A novel nonlinear flux sensorless control system for a dual three-phase PWM rectifier implementing the above method includes a flux reference model generation module, a sliding mode enhanced nonlinear flux observation module SMENFO, a cascaded expansion state observation module CESO, a two-stage disturbance compensation and secondary expansion state observation ESO-PLL position refinement structure, and a multiphase redundancy module for the dual three-phase PWM rectifier, wherein:

[0019] The flux linkage reference model generation module is used to describe the desired model of flux linkage state and rotor position information of a dual-winding permanent magnet synchronous motor.

[0020] The sliding mode enhanced nonlinear flux linkage observation module SMENFO is used to achieve high-precision estimation of flux linkage and rotor position without position sensors.

[0021] The cascaded extended state observation module CESO is used to estimate external disturbances and internal disturbances of the observer, and generate real-time compensation control quantities.

[0022] The ESO-PLL position refinement structure with two-stage disturbance compensation and secondary expansion state observation is used to improve synchronization accuracy and enhance the system's dynamic response capability.

[0023] The multiphase redundancy module of the dual three-phase PWM rectifier is used to ensure the fault-tolerant operation capability of the system under disturbance or local fault conditions.

[0024] Compared with the prior art, the present invention has the following advantages:

[0025] This invention significantly improves the anti-interference capability and convergence performance of the observer under external disturbances and parameter uncertainties by introducing SMENFO based on the mathematical model of the flux linkage of a permanent magnet synchronous motor, without position sensors. Furthermore, a CESO is constructed based on the flux linkage observer to perform real-time estimation and compensation for disturbances in the control system of the dual three-phase PWM rectifier and internal disturbances in the observer, thereby achieving adaptive suppression of external disturbances and dynamic compensation for the impact of system parameter changes, improving the overall control accuracy and robustness of the system. In addition, this invention combines ADRC and ESO PLL to construct a DC-SPR-ESO-PLL, further improving the AC side synchronization accuracy and the continuity and accuracy of rotor position estimation. Combined with the inherent multiphase redundancy structure of the dual three-phase PWM rectifier, this invention can maintain stable and reliable sensorless control performance under winding disturbances or local anomalies. Attached Figure Description

[0026] Figure 1 Schematic diagram of a nonlinear flux sensorless control method for a dual three-phase PWM rectifier;

[0027] Figure 2This is a block diagram of a sensorless control system based on nonlinear magnetic flux.

[0028] Figure 3 This is a block diagram of sensorless control based on SMENFO;

[0029] Figure 4 This is a structural diagram of the first-stage extended state observer;

[0030] Figure 5 Here is a structural diagram of the resonant module;

[0031] Figure 6 This is a diagram of the second-stage extended state observer structure;

[0032] Figure 7 Here is a structural diagram of CESO;

[0033] Figure 8 This is a classic ESO-PLL structure diagram;

[0034] Figure 9 This is a diagram of the ESO-PLL structure for single-compensation.

[0035] Figure 10 This is a diagram of the DC-SPR-ESO-PLL structure.

[0036] Figure 11 This is a block diagram of a sensorless control system based on SMENFO, combining CESO and DC-SPR-ESO-PLL.

[0037] Figure 12 Experimental results for ESO-PLL rotational speed estimation;

[0038] Figure 13 The experimental results for DC-SPR-ESO-PLL rotational speed estimation;

[0039] Figure 14 Experimental results for rotor position estimation using a traditional nonlinear flux linkage observer;

[0040] Figure 15 Experimental results for SMENFO rotor position estimation;

[0041] Figure 16 The results are from experiments on DC output without load surges in the CESO system.

[0042] Figure 17 To incorporate the experimental results of DC output under sudden load changes in the CESO system;

[0043] Figure 18 The results of the DC output experiment for the prime mover speed change without the CESO system were obtained.

[0044] Figure 19The results of the DC output experiment for the prime mover speed change when the CESO system was added. Detailed Implementation

[0045] The technical solution of the present invention will be further described below with reference to the accompanying drawings, but it is not limited thereto. Any modifications or equivalent substitutions to the technical solution of the present invention that do not depart from the spirit and scope of the technical solution of the present invention should be covered within the protection scope of the present invention.

[0046] This invention provides a nonlinear flux linkage sensorless control method for a dual three-phase PWM rectifier. Based on a mathematical model of the flux linkage of a permanent magnet synchronous motor, the method introduces the sliding mode reaching law and designs a SMENFO, enabling high-precision estimation of flux linkage state and rotor position information without position sensors, and effectively improving the observer's anti-disturbance capability and convergence performance. Furthermore, based on SMENFO, a CESO is constructed to estimate and compensate for disturbances in the dual three-phase PWM rectifier control system and internal disturbances within SMENFO in real time, thereby achieving adaptive suppression of external disturbances and dynamic compensation for the effects of system parameter uncertainties and changes. In addition, this invention combines ADRC and ESO-PLL to construct a DC-SPR-ESO-PLL to further improve AC side synchronization accuracy and system dynamic response capability. The specific steps are as follows:

[0047] Step 1: Provide a mathematical model of the flux linkage of a dual-winding permanent magnet synchronous motor and a mathematical model of a dual three-phase PWM rectifier, and establish a reference model for flux linkage state observation.

[0048] In this step, the flux linkage state observation reference model is used to describe the expected estimation characteristics of the flux linkage state and rotor position information of the permanent magnet synchronous motor, and is dynamically updated according to the motor operating state.

[0049] Step 2: Based on the mathematical model of the flux linkage of a permanent magnet synchronous motor, the sliding mode reaching law is introduced, and SMENFO is designed to achieve high-precision, sensorless estimation of flux linkage state and rotor position information through the observer output. The specific steps are as follows:

[0050] Step 2-1: Establish a nonlinear flux observer based on the flux linkage state equation of the permanent magnet synchronous motor.

[0051] Step 2-2: Introduce the sliding mode reaching law, design the sliding surface based on the characteristics of the nonlinear flux observer, and add a sliding mode switching term to the dynamic equation of the observation error to improve the convergence speed of the observer. The specific steps are as follows:

[0052] (a) Obtain the observation error between the estimated flux and the reference flux output by the nonlinear flux observer, and construct the error state vector;

[0053] (b) Design a sliding surface based on the dynamic characteristics of flux linkage observation error. The sliding surface is composed of a linear combination of flux linkage error and its rate of change, or a coupling term of flux linkage error and current error, in order to match the nonlinear characteristics of flux linkage observation of permanent magnet synchronous motor.

[0054] (c) Select a sliding mode reaching law form, wherein the reaching law includes continuous reaching terms and switching reaching terms. Continuous reaching terms are used to ensure that the system quickly approaches the sliding surface, and switching reaching terms are used to enhance the robustness to disturbances and uncertainties.

[0055] (d) Inject the sliding mode switching approach law into the dynamic equation of the observation error or the observer correction term to form the sliding mode enhanced nonlinear flux observation update law, so that the observation error converges within a finite time or a preset time.

[0056] Steps 2-3: Suppress external disturbances during flux estimation using a sliding mode enhancement mechanism to improve the observer's anti-interference capability. The specific steps are as follows:

[0057] (a) Apply boundary layer or saturation function processing to the sliding mode switching term to suppress high-frequency chattering and reduce the impact of noise on flux estimation, while maintaining the observer convergence speed;

[0058] (b) The approach law parameters and sliding mode gain are adaptively adjusted according to the magnitude of the observation error and the operating conditions, so that the system can maintain stable convergence and high-precision flux estimation under low speed, sudden load change and external disturbance conditions.

[0059] Steps 2-4: Output the estimated flux linkage value and further calculate the rotor position information to achieve high-precision flux linkage observation under sensorless conditions.

[0060] Step 3: Construct CESO based on SMENFO to estimate external disturbances and internal disturbances of the dual three-phase PWM rectifier control system in real time, and perform dynamic compensation based on the estimation results, thereby achieving adaptive disturbance suppression and parameter uncertainty compensation. The specific steps are as follows:

[0061] Step 3-1: Based on the stator current model of the permanent magnet synchronous motor, construct the first-stage extended state observer to estimate external disturbances to the dual three-phase PWM rectifier control system in real time, including load fluctuations and grid disturbances. The estimated disturbances are then input to the second-stage extended state observer. The specific steps are as follows:

[0062] (a) Collect information such as output voltage, current, and electrical angle frequency of the dual three-phase PWM rectifier;

[0063] (b) Calculate the perturbation error between the actual output and the expected output, and estimate the perturbation as an expanded state;

[0064] (c) The output disturbance compensation is compensated to the second-stage extended state observer in the form of voltage disturbance.

[0065] Step 3-2: Based on the permanent magnet synchronous motor flux linkage model, construct a second-stage extended state observer to estimate the estimation deviation caused by internal disturbances and parameter changes in SMENFO in real time. The specific steps are as follows:

[0066] (a) Collect output voltage, current and system parameter information of dual three-phase PWM rectifier;

[0067] (b) Calculate the perturbation error between the actual output and the expected output, and estimate the perturbation as an expanded state;

[0068] (c) Utilize the estimated disturbance to dynamically adjust the observer and reduce the impact of system parameter uncertainty on control performance.

[0069] Step 3-3: Generate compensation control quantity based on the two-stage disturbance estimation results, dynamically correct the flux observation output and rectifier control command, and improve the system robustness.

[0070] Step 4: Combine ADRC and ESO-PLL to establish DC-SPR-ESO-PLL, achieving high-precision tracking of AC side synchronization angle and frequency, and improving the system's dynamic response capability. Specific steps are as follows:

[0071] Step 4-1: Based on the derivative relationship between phase angle, angular velocity, and angular acceleration, establish the ESO PLL. The specific steps are as follows:

[0072] (a) Establish a state model of the derivative relationship between phase angle, angular velocity and angular acceleration, wherein the AC side synchronization angle is used as the first state variable, the synchronization angular velocity is used as the second state variable and the synchronization angular acceleration is used as the third state variable, and a chain integral structure of "angle-rotation speed-acceleration" is constructed.

[0073] (b) Using phase error or equivalent observation as the measurement input of ESO-PLL, construct extended state equations to unify grid disturbances, harmonic effects and parameter uncertainties into extended disturbance states, so as to facilitate centralized estimation and compensation.

[0074] (c) Based on the synchronization angle, synchronization angular velocity and synchronization angular acceleration output by ESO, a correction amount for the phase-locked loop is generated. The correction amount is used to suppress phase error and compensate for the impact of disturbance on phase tracking, thereby improving synchronization accuracy and dynamic response.

[0075] Step 4-2: Establish a two-stage disturbance compensation and secondary expansion state observation ESO-PLL position refinement structure for real-time compensation of phase-locked loop errors and disturbances. The specific steps are as follows:

[0076] (a) Obtain the rotor position related error e in the phase-locked loop and use this error as the input for active disturbance rejection regulation;

[0077] (b) Based on ESO, the rotational speed dynamics are extended and modeled. In the observer, grid fluctuations, load changes, parameter uncertainties and model errors are uniformly equivalent to "angular acceleration disturbance term" and the estimated value of the disturbance is output.

[0078] (c) Under normal operating conditions, taking advantage of the prior property that “the derivative of the rotational speed approaches 0 when running at a constant speed”, the estimated quantity of the disturbance is regarded as the main disturbance component, and the disturbance component is compensated to the system under normal operating conditions to generate a speed estimate after disturbance compensation.

[0079] (d) Based on the second-order extended state observation and position refinement of the fine rotational speed, the fine rotational speed estimate obtained by the previous disturbance compensation is used as the known system state variable to construct the kinematic relationship model between the rotor position and rotational speed. Based on the kinematic relationship model between the rotor position and rotational speed, the extended state is introduced. The unmodeled dynamics caused by model mismatch, high-frequency disturbance, parameter changes and noise are uniformly equivalent to the residual disturbance state. The residual disturbance estimate is introduced into the position estimation update process to correct the rotational speed integral or position prediction process online. Based on the corrected state variable, the fine estimate of the rotor position is output, thereby realizing the second-order refined estimation of the rotor position.

[0080] Step 4-3: Output synchronization angle information to achieve high-precision control of the dual three-phase PWM rectifier.

[0081] Step 5: Utilizing the multiphase redundancy structure of the dual three-phase PWM rectifier, voltage and current information from two sets of three-phase windings are collected for flux linkage estimation and rotor position identification. If the observation information from one set of windings is interfered with or fails, the electrical quantities of the other healthy winding are automatically switched or merged. The specific steps are as follows:

[0082] Step 5-1: Construct disturbance observers or extended state observers based on two sets of three-phase windings respectively, and estimate the equivalent disturbance components in the current of each winding in real time.

[0083] Step 5-2: Based on the amplitude, rate of change, or stability characteristics of the disturbance estimation results, determine the reliability of the corresponding winding speed estimation. When the current disturbance estimation value corresponding to one set of windings increases or the fluctuation intensifies, it is determined that the reliability of the winding speed estimation decreases.

[0084] Step 5-3: The control system prioritizes the speed estimation result corresponding to the other winding with smaller and more stable disturbance estimation, or performs weighted fusion of the two speed estimation results to improve the accuracy of rotor position estimation and the fault-tolerant operation capability of the system under disturbance or local fault conditions.

[0085] Example:

[0086] This embodiment provides a nonlinear flux sensorless control method for a dual three-phase PWM rectifier, such as... Figure 1 As shown, the method includes the following steps:

[0087] Step S1: Provide the mathematical model of the flux linkage of the permanent magnet synchronous motor and the mathematical model of the three-phase PWM rectifier, and establish a flux linkage state observation reference model. This reference model is used to describe the expected estimation characteristics of the flux linkage state and rotor position information of the permanent magnet synchronous motor. The sensorless control block diagram based on nonlinear flux linkage is shown below. Figure 2 As shown.

[0088] To facilitate the construction of the flux linkage state observation model, the stator stationary coordinate system is adopted. Formula. For surface-mounted permanent magnet synchronous motors (i.e. Under the sinusoidal steady-state assumption, the voltage equation can be expressed as:

[0089] (1)

[0090] in:

[0091] (2)

[0092] For surface-mount PMSMs, the stator flux linkage is obtained by superimposing the inductor flux linkage and the permanent magnet flux linkage:

[0093] (3)

[0094] in, These represent the winding voltage, stator current, and stator flux linkage, respectively. The amount. For stator resistance, For stator equivalent inductance, This refers to the flux linkage amplitude of the permanent magnet. This represents the electrical angle rotor position.

[0095] Substituting the magnetic flux expression into the voltage equation yields the dynamic form of the current:

[0096] (4)

[0097] in, ω is the electric angular velocity.

[0098] To construct a nonlinear flux linkage / position observer, a new state variable is defined. For "Extended flux linkage state":

[0099] (5)

[0100] Define auxiliary quantity (Can be obtained from voltage and current measurements, excluding unknowns):

[0101] (6)

[0102] From the voltage equation and The state equation can be obtained directly:

[0103] (7)

[0104] Step S2: Based on the mathematical model of the flux linkage of the permanent magnet synchronous motor, the sliding mode approach law is introduced, and SMENFO is designed to realize sensorless high-precision estimation of flux linkage state and rotor position information through the observer output.

[0105] Let the flux linkage state estimate be Then the basic observation structure can be constructed:

[0106] (8)

[0107] in, The observer correction term is taken as the sliding mode enhancement correction input in the SMENFO described in this invention. , is used to compensate for the flux linkage state estimation based on the sliding surface error, and its expression is shown in equation (12).

[0108] Equation for estimating permanent magnet flux linkage:

[0109] (9)

[0110] Permanent magnet flux amplitude Since the magnitude is a constant, a nonlinear sliding surface can be constructed using the "amplitude constraint error":

[0111] (10)

[0112] (11)

[0113] when Sometimes, The amplitude was pulled back ,thereby It converges to the true permanent magnet flux trajectory; its direction corresponds to the rotor position information.

[0114] To ensure fast approximation and robustness, a composite reaching law is chosen:

[0115] (12)

[0116] Saturation functions are defined by components:

[0117] (13)

[0118] in, To improve convergence speed and smoothness, the term is a continuous approaching term. To switch the approaching term, enhance robustness to parameter uncertainties, disturbances, and modeling errors, and through boundary layers... Suppress chattering.

[0119] Equation (12) is used as the observer correction term. Injection formula (8) yields SMENFO:

[0120] (14)

[0121] in, Take equation (10).

[0122] To accommodate low-speed, sudden load changes, and noisy operating conditions, the sliding mode gain is adaptively adjusted:

[0123] (15)

[0124] in, These are the base term of the sliding mode gain and the adjustment coefficient of the sliding mode gain, respectively.

[0125] The permanent magnet flux linkage is estimated as follows:

[0126] (16)

[0127] The block diagram of sensorless control based on SMENFO is as follows: Figure 3 As shown.

[0128] Prove convergence:

[0129] Selecting Lyapunov functions:

[0130] (17)

[0131] Differentiating, we get:

[0132] (18)

[0133] because and For the observer state, It can be written as:

[0134] (19)

[0135] in For Jacobian matrices, A summary of the data from Bounded terms arising from dynamics, parameter errors, etc. We take the following bounded assumption:

[0136] (20)

[0137] Substituting the observer update law (14) into (20) and then into (19), we get:

[0138] (twenty one)

[0139] when hour, and In the same direction, and satisfying:

[0140] (twenty two)

[0141] Substituting into (21), we get:

[0142] (twenty three)

[0143] Therefore, as long as the sliding mode gain satisfies:

[0144] (twenty four)

[0145] This can guarantee (Strict descent outside the boundary layer), thus Entering the boundary layer within a finite time .

[0146] when hour, ,but:

[0147] (25)

[0148] Substituting into (21), we get:

[0149] (26)

[0150] By the standard comparison lemma, we can obtain Ultimately, it converges to a small, adjustable neighborhood:

[0151] (27)

[0152] Therefore, by increasing , or reduce This can reduce steady-state boundary layer errors and improve estimation accuracy.

[0153] Select In the bounded perturbation Under the condition that the sliding mode gain satisfies ,but Entering the boundary layer within a finite time And remain ultimately bounded;

[0154] The upper bound of the final error within the boundary layer is Therefore, by adjusting the parameters, the flux estimation error can be compressed to a preset accuracy;

[0155] This plan is based on Coordinate system flux linkage state equation Based on, through Construct a permanent magnet flux linkage estimate and employ a constant amplitude constrained sliding mode surface. By injecting continuous and switching terms into the observer update law, fast convergence and strong perturbation resistance of flux linkage estimation are achieved, ultimately resulting in... To obtain high-precision sensorless rotor position estimation.

[0156] Step S3: Construct CESO based on SMENFO, and estimate the external disturbances of the dual three-phase PWM rectifier control system and the internal disturbances of SMENFO in real time, and perform dynamic compensation based on the estimation results.

[0157] A first-stage extended state observer is constructed to estimate external disturbances to the dual three-phase PWM rectifier control system in real time, including load fluctuations and grid disturbances. The estimated disturbances are then input to the second-stage extended state observer. The structure of the first-stage extended state observer is as follows: Figure 4 As shown, although the traditional Linear Extended State Observer (LESO) can estimate the disturbance, it is essentially a second-order low-pass structure and cannot track rapidly changing sinusoidal quantities well. Therefore, a resonant module needs to be added to identify disturbances at a certain frequency and design the resonant transfer function.

[0158] (28)

[0159] Wherein, the input vector is: Output vector: ; : Resonant gain; Controller bandwidth; : Resonant frequency.

[0160] Since the disturbance frequency of the PMSG is always consistent with the electric angular velocity of the motor, it can be achieved by setting: This enables the resonant module to achieve near-hysteresis-free tracking of the main frequency band of the disturbance.

[0161] When the stator is stationary In the coordinate system, the controlled object (with the current loop as the observed object) is written as:

[0162] (29)

[0163] in, Stator voltage, Stator current, Nominal stator resistance and inductance : Back electromotive force component Unmodeled disturbance.

[0164] Construct a second-order ESO on each axis, estimate the current in the first state, and estimate the total extended state in the second state. :

[0165] (30)

[0166] in, :right The observation status, For the observed state of the unmodeled disturbance term, Back electromotive force, : Observation error injection amount The axis formula is similar. This is a typical LESO structure, upon which a resonant module is embedded. The structure of the resonant module is as follows: Figure 5 As shown, the expression for the resonant expansion state observer can be obtained:

[0167] (31)

[0168] in, Estimate the disturbance for the oscillator output in the current system. The first-level perturbation estimate is used to compensate the second-level extended state observer. The structure of the second-level extended state observer is as follows: Figure 6 As shown, a second-level extended state observer is established based on equation (14):

[0169] (32)

[0170] A nonlinear function is defined as:

[0171] (33)

[0172] Where β1,β2 >0: observer gain, : Nonlinear order : Threshold of the linear region of the FAL function : Symbolic function, As an estimate of the unmodeled perturbation term in the extended state, it is ultimately compensated for by the nonlinear flux observer in the form of a voltage input. The complete CESO structure is as follows: Figure 7 As shown.

[0173] Step S4: Combine ADRC and ESO-PLL to construct DC-SESO-PR-PLL.

[0174] Using the synchronization angle of the AC side as The synchronous angular velocity is The synchronous angular acceleration is The "unknown disturbance / unmodeled dynamics" are uniformly incorporated into the acceleration derivative term. This yields the chain integral model:

[0175] (34)

[0176] Using estimated angles generate Construct a phase error with the equivalent orthogonal flux linkage observation of the motor. :

[0177] (35)

[0178] Let the three states of ESO correspond to the angle, rotational speed, and acceleration estimates, respectively: ESO adopts a typical structure:

[0179] (36)

[0180] in It is the ESO gain.

[0181] This is the classic ESO-PLL structure, and its structure diagram is as follows: Figure 8 As shown.

[0182] Construct DC-SESO-PR-PLL, Level 1 compensation: ... Treating it as the primary disturbance, the integral yields velocity compensation. Level 2 Refinement: Based on With the model already in place, a second ESO is introduced to estimate the residual perturbation, resulting in a more refined angle output.

[0183] In the chain model If the system is in a steady state or quasi-steady state with uniform velocity, then:

[0184] (37)

[0185] at this time The true value is mainly composed of equivalent terms caused by power grid fluctuations, load abrupt changes, harmonics, parameter uncertainties, measurement noise, etc., therefore the value given by ESO is... It can be used as an estimate of "total disturbance / equivalent angular acceleration disturbance", therefore it will Compensation to The system obtains the speed compensation amount. :

[0186] (38)

[0187] The structural diagram after one compensation is as follows: Figure 9 As shown.

[0188] A more precise speed estimate is obtained after compensation. Based on fine rotation speed A quadratic ESO is established, and the residual perturbation is estimated to obtain angle refinement.

[0189] Angle satisfied:

[0190] (39)

[0191] Use fine rotation speed As known inputs, the angle prediction model is obtained:

[0192] (40)

[0193] in, This indicates the residual items that have not been eliminated even after compensation.

[0194] Let the state of the second ESO be: , .

[0195] And the measurement error is still the same phase error. Constructing a second-order ESO:

[0196] (41)

[0197] in: For fine angle estimation output, Output for residual perturbation estimation. This is the gain of the second-order ESO.

[0198] DC-SESO-PR-PLL structure as follows Figure 10 As shown.

[0199] In summary The rotor position related error of the phase-locked loop is constructed from the quadrature signal and the estimated angle, and is used to drive the correction of the ESO-PLL and the secondary ESO.

[0200] Third-order ESO-PLL output As an estimate of synchronous angular acceleration (expansion state); under uniform or quasi-steady-state conditions, Therefore It mainly reflects the equivalent terms of external disturbances and model uncertainties.

[0201] Through the Integral generation speed compensation amount A fine speed estimate is obtained after disturbance compensation. It is used to improve the anti-disturbance performance and smoothness of speed estimation.

[0202] exist Based on the existing model, a secondary extended state observer is introduced to uniformly represent and estimate the residual disturbances that still exist after compensation as extended states. ,Will An angle update is injected to correct for speed integral error, thereby outputting a fine angle. This enables a secondary, refined estimate of the rotor position.

[0203] The block diagram of the sensorless control based on SMENFO, combined with CESO and DC-SESO-PR-PLL is as follows: Figure 11 As shown.

[0204] Step S5, Redundancy switching or fusion process: Establish two sets of winding q-axis disturbance estimates based on the dq-axis model of the permanent magnet synchronous motor.

[0205] (42)

[0206] Based on the estimated disturbance By selecting the winding information with smaller disturbance estimation as the input of the sensorless observer, the information from the two sets of windings can be used reasonably to improve the redundancy of the system.

[0207] Figure 12 , 13 The figures show the experimental results of ESO-PLL and DC-SPR-ESO-PLL for speed estimation. The experimental results show that the ESO-PLL output has obvious noise, while the estimation result obtained by DC-SPR-ESO-PLL is smoother and has less fluctuation, which verifies its advantage in extracting synchronization angle in actual systems.

[0208] Figure 14 , 15 The images show the rotor angle observation waveforms of the traditional nonlinear flux linkage observer and the improved system, respectively. Both algorithms can identify the rotor position well, and the estimated position curve obtained by the improved method is more consistent with the actual position.

[0209] Figure 16 , 17The experimental results are shown for the rectifier system with and without CESO when the load suddenly increases from 0 W to 10 W. The experimental results show that under the condition of sudden load change, the transient response speed of the system is significantly improved and the settling time is significantly shortened after adding CESO. At the same time, the voltage fluctuation is effectively suppressed when the load is small, and the system recovery capability is further enhanced when the load is large. This shows that CESO can effectively improve the dynamic regulation performance and disturbance rejection capability of the rectifier output voltage.

[0210] Figure 18 , 19 The experimental results are shown for the prime mover speed suddenly increasing from 800 r / min to 825 r / min with and without CESO. The experimental results show that under the condition of sudden speed change, the amplitude of the system output voltage fluctuation is reduced overall after adding CESO, the recovery process is more stable, and the anti-disturbance capability is enhanced.

[0211] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A nonlinear flux sensorless control method for a dual three-phase PWM rectifier, characterized in that... The method includes the following steps: Step 1: Provide a mathematical model of the flux linkage of a dual-winding permanent magnet synchronous motor and a mathematical model of a dual three-phase PWM rectifier, and establish a reference model for flux linkage state observation. Step 2: Based on the mathematical model of flux linkage of a dual-winding permanent magnet synchronous motor, the sliding mode approach law is introduced, and a sliding mode enhanced nonlinear flux linkage observer SMENFO is designed. The high-precision sensorless estimation of flux linkage state and rotor position information is achieved through the observer output. Step 3: Construct a cascaded extended state observer CESO based on SMENFO to estimate the external disturbances of the dual three-phase PWM rectifier control system and the internal disturbances of SMENFO in real time, and perform dynamic compensation based on the estimation results, thereby realizing adaptive disturbance suppression and parameter uncertainty compensation. Step 4: Combine Active Disturbance Rejection Control (ADRC) with Extended State Observer-based Phase-Locked Loop (ESO-PLL) to establish an Extended State Observer-based Phase-Locked Loop structure DC-SPR-ESO-PLL for disturbance compensation and secondary position refinement, thereby achieving high-precision tracking of AC-side synchronization angle and frequency and improving the system's dynamic response capability. Step 5: Utilize the multiphase redundancy structure characteristics of the dual three-phase PWM rectifier to collect voltage and current information of two sets of three-phase windings for flux estimation and rotor position identification. When the observation information of one set of windings is interfered with or fails, automatically switch or merge the electrical quantities of the other healthy winding.

2. The nonlinear flux sensorless control method for a dual three-phase PWM rectifier according to claim 1, characterized in that... The specific steps of step 2 are as follows: Step 2-1: Establish a nonlinear flux linkage observer based on the flux linkage state equation of the permanent magnet synchronous motor; Step 2-2: Introduce the sliding mode reaching law, design the sliding surface in combination with the characteristics of the nonlinear flux observer, and add a sliding mode switching term to the dynamic equation of the observation error to improve the convergence speed of the observer; Steps 2-3: Suppress the influence of external disturbances in the flux estimation process through the sliding mode enhancement mechanism to enhance the anti-interference capability of the observer; Steps 2-4: Output the estimated flux linkage value and further calculate the rotor position information to achieve high-precision flux linkage observation under sensorless conditions.

3. The nonlinear flux sensorless control method for a dual three-phase PWM rectifier according to claim 2, characterized in that... The specific steps of step 2-2 are as follows: (a) Obtain the observation error between the estimated flux and the reference flux output by the nonlinear flux observer, and construct the error state vector; (b) Design a sliding surface based on the dynamic characteristics of flux linkage observation error. The sliding surface is composed of a linear combination of flux linkage error and its rate of change, or a coupling term of flux linkage error and current error, in order to match the nonlinear characteristics of flux linkage observation of permanent magnet synchronous motor. (c) Select a sliding mode reaching law form, wherein the reaching law includes continuous reaching terms and switching reaching terms. Continuous reaching terms are used to ensure that the system quickly approaches the sliding surface, and switching reaching terms are used to enhance the robustness to disturbances and uncertainties. (d) Inject the sliding mode switching approach law into the dynamic equation of the observation error or the observer correction term to form the sliding mode enhanced nonlinear flux observation update law, so that the observation error converges within a finite time or a preset time.

4. The nonlinear flux sensorless control method for a dual three-phase PWM rectifier according to claim 2, characterized in that... The specific steps of steps 2-3 are as follows: (a) Apply boundary layer or saturation function processing to the sliding mode switching term to suppress high-frequency chattering and reduce the impact of noise on flux estimation, while maintaining the observer convergence speed; (b) The approach law parameters and sliding mode gain are adaptively adjusted according to the magnitude of the observation error and the operating conditions, so that the system can maintain stable convergence and high-precision flux estimation under low speed, sudden load change and external disturbance conditions.

5. The nonlinear flux sensorless control method for a dual three-phase PWM rectifier according to claim 1, characterized in that... The specific steps of step 3 are as follows: Step 3-1: Based on the stator current model of the permanent magnet synchronous motor, construct the first-stage extended state observer to estimate the external disturbances of the dual three-phase PWM rectifier control system in real time, and input the estimated disturbances to the second-stage extended state observer. Step 3-2: Based on the flux linkage model of the permanent magnet synchronous motor, construct the second-stage extended state observer to estimate the estimation deviation caused by internal disturbances and parameter changes in SMENFO in real time. Step 3-3: Generate compensation control quantity based on the two-stage disturbance estimation results, dynamically correct the flux observation output and rectifier control command, and improve the system robustness.

6. The nonlinear flux sensorless control method for a dual three-phase PWM rectifier according to claim 5, characterized in that... The specific steps of step 3-1 are as follows: (a) Collect the output voltage, current, and electrical angle frequency information of the dual three-phase PWM rectifier; (b) Calculate the perturbation error between the actual output and the expected output, and estimate the perturbation as an expanded state; (c) The output disturbance compensation is compensated to the second-stage extended state observer in the form of voltage disturbance; The specific steps of step 3-2 are as follows: (a) Collect output voltage, current and system parameter information of dual three-phase PWM rectifier; (b) Calculate the perturbation error between the actual output and the expected output, and estimate the perturbation as an expanded state; (c) Utilize the estimated disturbance to dynamically adjust the observer and reduce the impact of system parameter uncertainty on control performance.

7. The nonlinear flux sensorless control method for a dual three-phase PWM rectifier according to claim 1, characterized in that... The specific steps of step 4 are as follows: Step 4-1: Based on the derivative relationship between phase angle, angular velocity and angular acceleration, establish the ESO PLL; Step 4-2: Establish an ESO-PLL position refinement structure with two-level disturbance compensation and secondary expansion state observation, which is used to compensate for phase-locked errors and disturbances in real time. Step 4-3: Output synchronization angle information to achieve high-precision control of the dual three-phase PWM rectifier.

8. The nonlinear flux sensorless control method for a dual three-phase PWM rectifier according to claim 7, characterized in that... The specific steps of step 4-1 are as follows: (a) Establish a state model of the derivative relationship between phase angle, angular velocity and angular acceleration, wherein the AC side synchronization angle is used as the first state variable, the synchronization angular velocity is used as the second state variable and the synchronization angular acceleration is used as the third state variable, and a chain integral structure of "angle-rotation speed-acceleration" is constructed. (b) Using phase error or equivalent observation as the measurement input of ESO-PLL, construct extended state equations to unify grid disturbances, harmonic effects and parameter uncertainties into extended disturbance states, so as to facilitate centralized estimation and compensation. (c) Based on the synchronization angle, synchronization angular velocity and synchronization angular acceleration output by ESO, a correction amount for the phase-locked loop is generated. The correction amount is used to suppress phase error and compensate for the impact of disturbance on phase tracking, thereby improving synchronization accuracy and dynamic response.

9. The nonlinear flux sensorless control method for a dual three-phase PWM rectifier according to claim 7, characterized in that... The specific steps of step 4-2 are as follows: (a) Obtain the rotor position related error e in the phase-locked loop and use this error as the input for active disturbance rejection regulation; (b) Based on ESO, the rotational speed dynamics are extended and modeled. In the observer, grid fluctuations, load changes, parameter uncertainties and model errors are uniformly equivalent to "angular acceleration disturbance term" and the estimate of the disturbance is output. (c) Under normal operating conditions, taking advantage of the prior property that "the derivative of the rotational speed approaches 0 when running at a constant speed", the estimated quantity of the disturbance is regarded as the main disturbance component, and the disturbance component is compensated to the system under normal operating conditions to generate a speed estimate after disturbance compensation. (d) Based on the second-order extended state observation and position refinement of the fine rotational speed, the fine rotational speed estimate obtained by the previous disturbance compensation is used as the known system state variable to construct the kinematic relationship model between rotor position and rotational speed. Based on the kinematic relationship model between rotor position and rotational speed, the extended state is introduced. The unmodeled dynamics caused by model mismatch, high-frequency disturbance, parameter changes and noise are uniformly equivalent to the residual disturbance state. The residual disturbance estimate is introduced into the position estimation update process to correct the rotational speed integral or position prediction process online. Based on the corrected state variable, the fine estimate of rotor position is output, thereby realizing the second-order refined estimation of rotor position.

10. The nonlinear flux sensorless control method for a dual three-phase PWM rectifier according to claim 1, characterized in that... The specific steps of step 5 are as follows: Step 5-1: Construct disturbance observers or extended state observers based on two sets of three-phase windings respectively, and estimate the equivalent disturbance components in the current of each winding in real time. Step 5-2: Based on the amplitude, rate of change, or stability characteristics of the disturbance estimation results, determine the reliability of the corresponding winding speed estimation. When the current disturbance estimation value corresponding to one set of windings increases or the fluctuation intensifies, it is determined that the reliability of the winding speed estimation decreases. Step 5-3: The control system prioritizes the speed estimation result corresponding to the other winding with smaller and more stable disturbance estimation, or performs weighted fusion of the two speed estimation results to improve the accuracy of rotor position estimation and the fault-tolerant operation capability of the system under disturbance or local fault conditions.

Images