Aviation actuator PMSM position sensor adaptive fault-tolerant control method

Abstract

The invention provides a self-adaptive fault-tolerant control method for a PMSM (permanent magnet synchronous motor) position sensor of an aviation actuator, which relates to the technical field of aviation electromechanical actuating systems and comprises the following steps of: receiving an original sensor signal of the aviation electromechanical actuator, and preprocessing the signal; carrying out adaptive state reconstruction and fault diagnosis on the preprocessed original sensor signal; carrying out fault recording and classification according to the diagnosis result to generate a maintenance decision; development and verification are carried out based on the maintenance decision, and a final signal meeting the aviation safety standard is output; the method can adapt to an aviation extreme environment, realizes high-precision state reconstruction, has a multi-fault mode diagnosis capability, can realize seamless smooth fault-tolerant switching, improves the autonomous survival capability of an aviation electromechanical actuation system, and reduces the whole life cycle cost.

Description

Technical Field

[0001] This invention relates to the field of aircraft electromechanical actuation system technology, and more specifically, to an adaptive fault-tolerant control method for an aircraft actuator PMSM position sensor. Background Technology

[0002] Aerospace electromechanical actuation systems are a critical component of flight safety, and their performance directly impacts aircraft handling qualities and mission reliability. With the advancement of more-electric / all-electric aircraft technology, permanent magnet synchronous motors (PMSMs), due to their high efficiency, high power density, and excellent dynamic response characteristics, have been widely applied in key flight control actuators such as flaps, rudders, and elevators, forming the power core of these electromechanical actuators. In these high-integrity systems, accurate measurement of the motor rotor position is a prerequisite for achieving high-performance field-oriented control, while the reliability of the position sensor becomes the cornerstone of the entire system's safe operation.

[0003] However, the aviation application environment poses extremely stringent challenges to position sensors. When an aircraft cruises at an altitude of 10,000 meters, the external temperature of the cabin can drop as low as -55°C, while the temperature around the engine nacelle can reach over 125°C. This intense temperature cycling can easily lead to drift in the internal material properties of the sensor, solder fatigue, and seal failure. Simultaneously, the wideband, high-intensity vibrations caused by the aircraft engine and aerodynamic turbulence can easily cause loosening of the sensor's mechanical structure, code disk contamination, or signal interruption. Furthermore, the complex electromagnetic environment, including the switching of high-power airborne equipment, lightning strikes, and high-intensity radiation fields, can severely interfere with sensitive analog or digital signal chains, causing instantaneous jumps, shifts, or complete loss of position information. Among electrical faults in aviation electromechanical actuation systems, position sensor anomalies induced by environmental factors account for over 30%, making it one of the main risk sources affecting system readiness and flight safety.

[0004] To ensure continuous system operation in the event of sensor failure, the traditional aviation field mainly relies on a "hardware redundancy" architecture. This involves configuring dual or even triple sensors for critical channels and achieving fault tolerance through majority voting or fault switching logic. While this approach provides high fault coverage, it also brings significant side effects: it not only greatly increases system weight, size, and wiring complexity but also significantly increases procurement costs and maintenance difficulty, contradicting the design trends of modern aircraft towards lightweight and highly integrated designs. Another approach is the "analytical redundancy" method based on motor mathematical models and observers, which generates redundancy through software algorithms. Virtual sensor signals serve as backups; however, existing observer technologies generally suffer from the following bottlenecks when applied to complex aviation conditions: First, the algorithm is extremely sensitive to the drift of motor parameters, and the winding modeling error can reach more than 20% at high temperatures, leading to a sharp deterioration in state reconstruction accuracy; Second, under strong vibration and EMI noise, conventional filtering methods are difficult to effectively extract the true rotor position, and the observer is prone to divergence or hysteresis; Third, there is a lack of refined diagnosis and rapid, non-destructive switching capabilities for typical aviation fault modes, making it difficult to meet the stringent requirements of flight control for millisecond-level actuator response and seamless switching. Summary of the Invention

[0005] The purpose of this invention is to provide an adaptive fault-tolerant control method for PMSM position sensors of aircraft actuators, which can adapt to extreme aviation environments, achieve high-precision state reconstruction, have multi-fault mode diagnosis capabilities, and achieve seamless and smooth fault-tolerant switching, thereby improving the autonomous survivability of aircraft electromechanical actuator systems and reducing the total life cycle cost.

[0006] The technical solution of this invention is as follows:

[0007] In a first aspect, this application provides an adaptive fault-tolerant control method for a PMSM position sensor of an aircraft actuator, which includes the following steps:

[0008] S1. Receive raw sensor signals from aircraft electromechanical actuators and perform signal preprocessing;

[0009] S2. Adaptive state reconstruction and fault diagnosis are performed on the preprocessed raw sensor signals;

[0010] S3. Record and classify faults based on diagnostic results to generate maintenance decisions;

[0011] S4. Develop and validate based on maintenance decisions, and output a final signal that meets aviation safety standards.

[0012] Furthermore, in step S1, the signal preprocessing includes temperature compensation, multilayer EMI filtering, and signal isolation and shielding; the temperature compensation is achieved by establishing a temperature compensation model, the model formula of which is:

[0013]

[0014]

[0015]

[0016] In the formula, This refers to the magnetic flux linkage of a permanent magnet at a winding temperature of T. , , This is the nominal value at 25℃. , , All are temperature coefficients. This is the current winding temperature. For reference temperature, This represents the resistance of the permanent magnet at a winding temperature of T. Let T be the inductance of the permanent magnet at a winding temperature of T.

[0017] Furthermore, the above-mentioned multilayer EMI filtering process employs a finite impulse response filter and uses a window function design method to control the filter's frequency response characteristics; the expressions for the filter transfer function and the window function are as follows:

[0018]

[0019]

[0020] In the formula, Let h[k] be the transfer function of the filter (represented in the frequency domain or z-domain), M be the filter order, h[k] be the k-th coefficient of the FIR filter, i.e., the impulse response sequence, k be the discrete-time index from 0 to M-1, and z be the complex variable in the z-transform, usually representing the unit delay in the discrete-time system. For Kaiser window functions, For zero-order modified Bessel functions, This is the discrete-time index of the window function. The point is the center of symmetry of the window function. The shape parameters of the Kaiser window are defined. Through multi-layer LCπ-type filters, TVS transient suppression devices, and precision shielding and grounding design, it achieves source suppression of high-intensity radiation fields and conducted interference at the nanosecond level.

[0021] Furthermore, in step S2, the calculation process for the above-mentioned adaptive state reconstruction includes:

[0022] A state-space model of the aviation PMSM is established to perform real-time correction on the preprocessed raw sensor signals, providing high-quality input for subsequent algorithms.

[0023]

[0024]

[0025]

[0026]

[0027]

[0028]

[0029]

[0030]

[0031] The second-order Runge-Kutta method is used for discretization to ensure the high precision requirements of aerospace applications.

[0032]

[0033]

[0034]

[0035] State prediction, measurement prediction, and state update are performed using the capacitive Kalman filter algorithm.

[0036] The calculation process for state prediction is as follows:

[0037]

[0038]

[0039]

[0040]

[0041]

[0042] The calculation process for measurement prediction is as follows:

[0043]

[0044]

[0045]

[0046]

[0047]

[0048]

[0049] The calculation process for state updates is as follows:

[0050]

[0051]

[0052]

[0053] Online adaptive noise covariance estimation:

[0054]

[0055]

[0056]

[0057]

[0058]

[0059]

[0060]

[0061] Error reconstruction through dynamic error system modeling:

[0062]

[0063]

[0064]

[0065]

[0066]

[0067] In the formula, x is the state vector, Let t be the input vector and t be the time. , The state matrix, For process noise, For the observation vector, To observe the noise, For the α-axis current, For β-axis current, The rotor angular velocity, For rotor electrical angle, The voltage along the α-axis. For β-axis voltage, This represents the resistance of the permanent magnet at a winding temperature of T. Let T be the inductance of the permanent magnet at a winding temperature of T. This refers to the magnetic flux linkage of a permanent magnet at a winding temperature of T. for, Let T be the moment of inertia, and let T denote the transpose matrix. Let be the state vector at time k+1. Let k be the state vector at time k. Sampling time, The first-order slope of the Runge-Kutta slope. The second-order slope of the Runge-Kutta slope. The process noise at time k, Let K be the motor temperature at time k. Let k be the input vector at time k. The Cholesky factor is the covariance matrix. For Cholesky decomposition, The state estimation covariance matrix at time k-1 is given. For volume point, This is the state estimate at time k-1; For the i-th basic volume point, The volume point after propagation. Let the first-order slope of Runge-Kutta be the volume point i. Let the first-order slope of Runge-Kutta be the volume point i. This is the predicted value of the system state. The covariance matrix of the state prediction values. The process noise covariance matrix; This is the covariance matrix of the state predictions after Cholesky decomposition. The parameters for the Cholesky decomposition form are chosen to represent the decomposition as a product of a lower triangular matrix and its transpose. This is the standard configuration for generating volume points using the capacitive Kalman filter. This is the predicted state value. This is the first measurement point of volume. For the observation matrix, The volume points are generated for the estimation of the current time step and the covariance matrix generated from the previous time step after decomposition. For the system's measurement prediction values, The covariance matrix of the measured predicted values, The noise covariance matrix is... This is the cross-covariance matrix between the state predictions and the measurement predictions. For volumetric Kalman gain, This is the optimal estimate of the state. The actual observed value at time k. Update the covariance matrix; For the new information sequence, For theoretical covariance, For mathematical expectation, The update law for the noise covariance matrix, For the time index within the sliding window, For adaptively adjustable gain, It is a sliding window. This is the estimated actual covariance of the new sequence. For the information at time j, This is the estimated actual covariance of the new sequence. It is a residual sequence. The observed values ​​are remeasured after the state is updated at time k. For covariance estimation, Let j be the residual. Let k be the process noise covariance matrix at time k; For reconstruction error, For dynamic error, Here is the state transition matrix. For higher-order linearization error, It is composite noise. It is an identity matrix with the same dimensions as the state vector. Let f be the Jacobian matrix, and let f be the nonlinear continuous-time state transition function of the permanent magnet synchronous motor. The midpoint of the remainder term in the Taylor expansion. To connect line segments, connect and line segments, Let be the observation noise at time k.

[0068] Furthermore, the above fault diagnosis process includes:

[0069] The electrical angle output from the position sensor is acquired in real time to calculate the position residual and the rate of change of the signal. The calculation formulas include:

[0070]

[0071]

[0072] In the formula, For positional residuals, For electrical angle, This is the reconstructed electrical angle value from the position sensor. For time, Let k be the error of the position sensor output value. The position sensor output value error at time k-1, Sampling time;

[0073] Fault diagnosis is performed based on position residuals and signal change rates.

[0074] Furthermore, fault diagnosis includes signal disconnection fault diagnosis, signal stagnation fault diagnosis, signal offset fault diagnosis, and complex fault diagnosis.

[0075] The criteria for diagnosing the above-mentioned signal disconnection fault are as follows:

[0076] ΔT≥T threshold

[0077] If all conditions are met simultaneously, it is determined to be a disconnection fault, and the fault flag Flag=1;

[0078] The criteria for diagnosing the above-mentioned signal stagnation fault are as follows:

[0079] , ΔT≥T threshold

[0080] If all conditions are met simultaneously, it is determined to be a disconnection fault, and the fault flag Flag=2;

[0081] The criteria for diagnosing the above signal offset fault are as follows:

[0082] Δθ≥θ threshold ΔT≥T threshold

[0083] If all conditions are met simultaneously, it is determined to be a disconnection fault, and the fault flag Flag=3;

[0084] In the formula, Let ΔT be the electrical angle, and ΔT be the duration. threshold This is the time threshold for abnormal position sensor signals. For time, The output value at a certain moment. For the position residual, θ threshold The threshold value for the residual of the position sensor signal;

[0085] The above-mentioned complex fault diagnosis includes intermittent disconnection and disconnection-offset coupling faults. The diagnostic criteria are as follows: intermittent disconnection is manifested by the periodic switching of Flag between 1 and 0; disconnection-offset coupling is triggered by first triggering Flag=1, and then immediately triggering Flag=3 after the fault is recovered.

[0086] Secondly, this application provides an electronic device, comprising:

[0087] Memory, used to store one or more programs;

[0088] processor;

[0089] When one or more of the above programs are executed by the above processor, an adaptive fault-tolerant control method for an aircraft actuator PMSM position sensor as described in any of the first aspects above is implemented.

[0090] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements an adaptive fault-tolerant control method for an aircraft actuator PMSM position sensor as described in any of the first aspects above.

[0091] Compared with the prior art, the present invention has at least the following advantages or beneficial effects:

[0092] (1) The present invention provides an adaptive fault-tolerant control method for PMSM position sensors of aircraft actuators. In view of the technical bottleneck of the decrease in state estimation accuracy caused by parameter drift of permanent magnet synchronous motor in extreme aviation environment, an improved commensurate Kalman filter (ACKF) algorithm that integrates online temperature compensation and noise adaptive estimation is proposed. The core is to directly embed the real-time data of temperature sensor into the state prediction equation by establishing an augmented state model of the motor containing temperature variables. At the same time, based on the dual sliding window adaptive mechanism of innovation and residual sequence, the process and measurement noise covariance matrix is ​​estimated and updated online in real time. This fundamentally solves the dual problems of model mismatch and unknown noise characteristics in aviation environment, and provides a high-precision and robust state backup signal for fault-tolerant control.

[0093] (2) In order to overcome the defects of false alarms and missed alarms of single threshold diagnosis under complex aviation interference, this invention designs a dual-parameter dynamic threshold diagnosis mechanism based on position residual and abnormal duration. The mechanism sets position residual threshold and abnormal time threshold. Fault judgment must meet the dual conditions of amplitude and duration at the same time. Parallel refined diagnosis logic is established for signal disconnection, stagnation, offset and compound faults. By creating a fault feature space of "amplitude-time" joint judgment, the accurate identification and classification of multiple fault modes are realized.

[0094] (3) In order to solve the risk of actuator jump caused by control signal switching after sensor failure, the present invention constructs a deterministic smooth fault-tolerant switching architecture based on a preset switching function. The cosine square smooth transition function is used to control the mixed weights to ensure that the signal source switching process is completed within a set time and the switching trajectory remains continuous in both position and velocity. The core of this architecture is to use the switching time as a design variable and ensure determinism through an analytical function, which meets the highest safety requirements of aviation systems for predictable and shock-free fault-tolerant processes.

[0095] (4) In order to achieve seamless integration of fault tolerance function and aircraft maintenance system, this invention defines a complete fault event encapsulation, recording and reporting mechanism and hierarchical degradation strategy; the scheme standardizes the fault record format including fields such as timestamp, fault type, environmental snapshot, ATA / MEL project number, etc., and can automatically generate maintenance suggestions; at the same time, it implements four-level degradation control from enhanced monitoring to safety maintenance according to the severity of the fault, and can realize the data flow and business chain from airborne real-time fault tolerance control to ground intelligent maintenance decision-making;

[0096] (5) To ensure the reliable deployment of the proposed fault-tolerant algorithm in aviation safety-critical systems, this invention provides a complete implementation scheme covering algorithm optimization, resource allocation, and verification testing. This scheme performs in-depth algorithm optimization and static resource allocation for aviation DSPs, ensures timing determinism through worst-case execution time analysis, and generates a complete compliance evidence package that meets the requirements of DO-178CDAL-B level, thus achieving high-security deployment under limited resources. Attached Figure Description

[0097] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0098] Figure 1 This is a diagram illustrating the overall architecture of an adaptive fault-tolerant control method for a PMSM position sensor of an aircraft actuator according to the present invention.

[0099] Figure 2 This is a diagram of a three-level EMI suppression system architecture.

[0100] Figure 3 The flowchart of the Adaptive Capacitive Kalman Filter (ACKF) algorithm is shown below.

[0101] Figure 4 Here is a flowchart of the adaptive noise estimation process;

[0102] Figure 5 Fault classification and diagnosis logic diagram;

[0103] Figure 6 This is a timing diagram for fault-tolerant smooth switching;

[0104] Figure 7 This is a diagram of a fault-tolerant control system architecture.

[0105] Figure 8 A schematic block diagram of an electronic device.

[0106] Icons: 101, memory; 102, processor; 103, communication interface. Detailed Implementation

[0107] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0108] Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0109] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0110] It should be noted that, in this document, the term "comprising" or any other variation thereof is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0111] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the various embodiments and features described below can be combined with each other.

[0112] Example 1

[0113] This invention provides an adaptive fault-tolerant control method for a PMSM position sensor of an aircraft actuator, which includes the following steps:

[0114] S1. Receive raw sensor signals from aircraft electromechanical actuators and perform signal preprocessing;

[0115] S2. Adaptive state reconstruction and fault diagnosis are performed on the preprocessed raw sensor signals;

[0116] S3. Record and classify faults based on diagnostic results to generate maintenance decisions;

[0117] S4. Develop and validate based on maintenance decisions, and output a final signal that meets aviation safety standards.

[0118] Please see Figure 1 , Figure 1 The diagram shown is an overall architecture diagram of an adaptive fault-tolerant control method for a PMSM position sensor of an aircraft actuator provided in an embodiment of this application.

[0119] Figure 1 The system demonstrates the four-layer architecture of this invention patent. The system receives raw sensor signals from aircraft electromechanical actuators, performs preprocessing (temperature compensation, vibration filtering, EMI suppression) in the environmental adaptability layer, and then sends them to the core algorithm layer for state reconstruction and fault diagnosis. The diagnosis results trigger a smooth switch, and simultaneously enter the safety management layer to generate maintenance decisions. The entire architecture is developed and verified under the guidance of the airworthiness compliance layer to ensure compliance with aviation safety standards. The data flow between each layer is shown by arrows, reflecting a complete closed loop from signal acquisition to control output.

[0120] Environmental adaptability layer: Designed for extreme aviation environments, the temperature compensation module adjusts control parameters in real time based on NTC sensors, the vibration filtering adopts a high-order finite impulse response (FIR) filter, and the EMI suppression module is designed with thresholds according to the DO-160G standard.

[0121] Core algorithm layer: Adaptive CKF achieves high-precision state estimation, dual threshold diagnosis mechanism (position threshold and time threshold) ensures high fault detection coverage, and smooth switching controller guarantees no overshoot switching within 10ms.

[0122] Safety management layer: Complies with ATA (Aircraft Maintenance Authority) standards, automatically identifies MEL (Mechanical Equipment List) items, and implements a four-level degradation strategy based on the severity of the fault.

[0123] Airworthiness Compliance Layer: Ensures the entire system meets the three major aviation standards: DO-178C (software), DO-160G (environment), and ARP4761 (safety).

[0124] In a preferred implementation, step S1 includes signal preprocessing, which includes temperature compensation, multilayer EMI filtering, and signal isolation and shielding. Temperature compensation is achieved by establishing a temperature compensation model. Since the operating temperature range of aerospace electromechanical actuators is wide, key parameters such as permanent magnet flux linkage and winding resistance change significantly with temperature. Therefore, a temperature compensation model is established.

[0125]

[0126]

[0127]

[0128] In the formula, This refers to the magnetic flux linkage of a permanent magnet at a winding temperature of T. , , This is the nominal value at 25℃. , , All are temperature coefficients. This is the current winding temperature. For reference temperature, This represents the resistance of the permanent magnet at a winding temperature of T. Let T be the inductance of the permanent magnet at a winding temperature of T.

[0129] As a preferred implementation, considering the unique wideband vibration environment of aviation, the multilayer EMI filtering process employs a finite impulse response (FIR) filter, and uses a window function design method to control the filter's frequency response characteristics; the expressions for the filter transfer function and the window function are as follows:

[0130]

[0131]

[0132] In the formula, Let h[k] be the transfer function of the filter (represented in the frequency domain or z-domain), M be the filter order, h[k] be the k-th coefficient of the FIR filter, i.e., the impulse response sequence, k be the discrete-time index from 0 to M-1, and z be the complex variable in the z-transform, usually representing the unit delay in the discrete-time system. For Kaiser window functions, For zero-order modified Bessel functions, This is the discrete-time index of the window function. The point is the center of symmetry of the window function. The shape parameters of the Kaiser window are defined. Through multi-layer LCπ-type filters, TVS transient suppression devices, and precision shielding and grounding design, it achieves source suppression of high-intensity radiation fields and conducted interference at the nanosecond level.

[0133] like Figure 2As shown, this invention constructs a complete electromagnetic compatibility solution from physical protection to intelligent recovery: the first-level hardware physical protection layer, through multi-layer LCπ-type filters, TVS transient suppression devices, and precision shielding and grounding design, achieves source suppression of high-intensity radiated fields and conducted interference at the nanosecond level; the second-level software intelligent detection layer, using sliding window statistical analysis and real-time feature extraction algorithms, accurately identifies and classifies typical EMI events such as transient pulses and periodic interference at the microsecond level; the third-level algorithm adaptive recovery layer, comprehensively utilizing predictive correction reconstruction, wavelet threshold denoising, and extended Kalman filter enhancement techniques, completes the repair and reconstruction of contaminated signals at the millisecond level; the three-level collaborative EMI suppression architecture, through deep collaborative design of hardware and software, achieves reliable protection of sensor signals in complex electromagnetic environments while meeting stringent aviation standards, providing a key guarantee for the stable operation of the entire electromechanical actuator fault-tolerant control system.

[0134] In a preferred embodiment, step S2, the adaptive state reconstruction calculation process includes:

[0135] A state-space model of the aviation PMSM is established to perform real-time correction on the preprocessed raw sensor signals, providing high-quality input for subsequent algorithms.

[0136] State equations for a permanent magnet synchronous motor considering temperature compensation:

[0137]

[0138]

[0139] The state vector and control vector are respectively:

[0140]

[0141]

[0142]

[0143] The state matrix is:

[0144]

[0145]

[0146]

[0147] Discretization using the second-order Runge-Kutta method ensures the high precision requirements of aerospace applications.

[0148]

[0149]

[0150]

[0151] State prediction, measurement prediction, and state update are performed using the Adaptive Cumulative Kalman Filter (ACKF) algorithm.

[0152] It should be noted that the Capacitive Kalman Filter (CKF) algorithm is based on the third-order spherical radial volume criterion. It uses a set of equally weighted volume points (2n=8 points for a 4-dimensional system) to approximate the state distribution. It mainly includes three processes: state prediction, measurement prediction, and state update. The main algorithm flow is as follows: Figure 3 As shown.

[0153] The calculation process for state prediction is as follows:

[0154] (1) Covariance matrix decomposition:

[0155]

[0156] (2) The volume points generated based on the previous time-instance estimate and the decomposed covariance matrix are:

[0157]

[0158] (3) Propagation of volume points based on the nonlinear model:

[0159]

[0160] (4) Calculate the predicted system state:

[0161]

[0162] (5) Calculate the covariance matrix of the predicted state values:

[0163]

[0164] The calculation process for measurement prediction is as follows:

[0165] (1) Apply Cholesky decomposition to the state covariance matrix:

[0166]

[0167] (2) The volume points generated based on the current time-time estimate and the covariance matrix generated from the previous time-time after decomposition are:

[0168]

[0169] (3) Propagate the obtained volume points according to the nonlinear model:

[0170]

[0171] (4) Calculate the predicted measurement values ​​of the system:

[0172]

[0173] (5) Calculate the covariance matrix of the predicted measurement values:

[0174]

[0175] (6) Calculate the cross-covariance matrix between the state prediction and the measurement prediction:

[0176]

[0177] The calculation process for state updates is as follows:

[0178] (1) Calculate the volumetric Kalman gain based on the measurement covariance matrix and the cross-covariance matrix:

[0179]

[0180] (2) The optimal estimate of the state is:

[0181]

[0182] (3) The covariance update matrix is:

[0183]

[0184] Given that the fixed noise covariance matrices Q and R may degrade filtering performance under varying operating conditions, an adaptive scheme (ACKF) based on online parameter adjustment using innovation and residual sequences is designed. Utilizing the statistical properties of the innovation and residual sequences, Q and R are estimated and updated online using the covariance matching principle. The adaptive process is as follows: Figure 4 As shown, the online estimation process for adaptive noise covariance includes:

[0185] Define the new information sequence and its theoretical covariance They are respectively:

[0186]

[0187]

[0188] The actual covariance estimate of the innovation sequence is calculated using a sliding window of length N:

[0189]

[0190] Wherein: The value of the sliding window N is usually taken in the range of 10-50, and can be dynamically adjusted according to the system;

[0191] Based on the covariance matching principle, by ensuring that the real-time estimated value approximates the theoretical value, the update law of the measurement noise covariance matrix R can be derived as follows:

[0192]

[0193] Similarly, define the residual sequence And calculate its covariance estimate. They are respectively:

[0194]

[0195]

[0196] Process noise covariance matrix The update law can be obtained through Kalman gain. calculate:

[0197]

[0198] To ensure numerical stability and positive definiteness, in practice only the diagonal elements of the covariance matrix are adaptively adjusted, and their lower bound is constrained:

[0199]

[0200] in: and To obtain the preset minimum covariance matrix, the above adaptive steps are added to the standard CKF algorithm to update it in real time. and Available for use in the next cycle;

[0201] Error reconstruction through dynamic error system modeling:

[0202] The reconstruction error is defined as:

[0203]

[0204] The error dynamic equation is:

[0205]

[0206] The state transition matrix is:

[0207]

[0208] The higher-order linearization error is:

[0209]

[0210] The composite noise is:

[0211]

[0212] In the formula, x is the state vector, Let t be the input vector and t be the time. , The state matrix, For process noise, For the observation vector, To observe the noise, For the α-axis current, For β-axis current, The rotor angular velocity, For rotor electrical angle, The voltage along the α-axis. For β-axis voltage, This represents the resistance of the permanent magnet at a winding temperature of T. Let T be the inductance of the permanent magnet at a winding temperature of T. This refers to the magnetic flux linkage of a permanent magnet at a winding temperature of T. for, Let T be the moment of inertia, and let T denote the transpose matrix. Let be the state vector at time k+1. Let k be the state vector at time k. Sampling time, The first-order slope of the Runge-Kutta slope. The second-order slope of the Runge-Kutta slope. The process noise at time k, Let K be the motor temperature at time k. Let k be the input vector at time k. The Cholesky factor is the covariance matrix. For Cholesky decomposition, The state estimation covariance matrix at time k-1 is given. For volume point, This is the state estimate at time k-1; For the i-th basic volume point, The volume point after propagation. Let the first-order slope of Runge-Kutta be the volume point i. Let the first-order slope of Runge-Kutta be the volume point i. This is the predicted value of the system state. The covariance matrix of the state prediction values. The process noise covariance matrix; This is the covariance matrix of the state predictions after Cholesky decomposition. The parameters for the Cholesky decomposition form are chosen to represent the decomposition as a product of a lower triangular matrix and its transpose. This is the standard configuration for generating volume points using the capacitive Kalman filter. This is the predicted state value. This is the first measurement point of volume. For the observation matrix, The volume points are generated for the estimation of the current time step and the covariance matrix generated from the previous time step after decomposition. For the system's measurement prediction values, The covariance matrix of the measured predicted values, The noise covariance matrix is... This is the cross-covariance matrix between the state predictions and the measurement predictions. For volumetric Kalman gain, This is the optimal estimate of the state. The actual observed value at time k. Update the covariance matrix; For the new information sequence, For theoretical covariance, For mathematical expectation, The update law for the noise covariance matrix, For the time index within the sliding window, For adaptively adjustable gain, It is a sliding window. This is the estimated actual covariance of the new sequence. For the information at time j, This is the estimated actual covariance of the new sequence. It is a residual sequence. The observed values ​​are remeasured after the state is updated at time k. For covariance estimation, Let j be the residual. Let k be the process noise covariance matrix at time k; For reconstruction error, For dynamic error, Here is the state transition matrix. For higher-order linearization error, It is composite noise. It is an identity matrix with the same dimensions as the state vector. Let f be the Jacobian matrix, and let f be the nonlinear continuous-time state transition function of the permanent magnet synchronous motor. The midpoint of the remainder term in the Taylor expansion. To connect line segments, connect and line segments, Let be the observation noise at time k.

[0213] As a preferred implementation, the fault diagnosis process includes:

[0214] The electrical angle output from the position sensor is acquired in real time to calculate the position residual and the rate of change of the signal. The calculation formulas include:

[0215]

[0216]

[0217] In the formula, For positional residuals, For electrical angle, This is the reconstructed electrical angle value from the position sensor. For time, Let k be the error of the position sensor output value. The position sensor output value error at time k-1, Sampling time;

[0218] Fault diagnosis is performed based on position residuals and signal change rates.

[0219] As a preferred implementation, fault diagnosis includes signal disconnection fault diagnosis, signal stagnation fault diagnosis, signal offset fault diagnosis, and complex fault diagnosis.

[0220] The criteria for diagnosing signal disconnection faults are as follows:

[0221] ΔT≥T threshold

[0222] If all conditions are met simultaneously, it is determined to be a disconnection fault, and the fault flag Flag=1;

[0223] The criteria for diagnosing signal stagnation faults are as follows:

[0224] , ΔT≥T threshold

[0225] If all conditions are met simultaneously, it is determined to be a disconnection fault, and the fault flag Flag=2;

[0226] The criteria for diagnosing signal offset faults are as follows:

[0227] Δθ≥θ threshold ΔT≥T threshold

[0228] If all conditions are met simultaneously, it is determined to be a disconnection fault, and the fault flag Flag=3;

[0229] In the formula, Let ΔT be the electrical angle, and ΔT be the duration. threshold This is the time threshold for abnormal position sensor signals. For time, The output value at a certain moment. For the position residual, θ threshold The threshold value for the residual of the position sensor signal;

[0230] Complex fault diagnosis includes intermittent disconnection and disconnection-offset coupling faults. The diagnostic criteria are as follows: intermittent disconnection is manifested by the periodic switching of Flag between 1 and 0; disconnection-offset coupling is first triggered by Flag=1, and then immediately triggered by Flag=3 after the fault is recovered.

[0231] In the fault diagnosis process, the diagnostic threshold is determined by the performance indicators of the aircraft actuator. Table 1 shows the threshold parameter setting table:

[0232] Table 1

[0233] parameter symbol numerical values in accordance with Location residual threshold θ_th 2° 80% of the actuator's positional accuracy Abnormal time threshold T_th 20ms 40% of the actuator response time Location change rate threshold dθ_th 100° / s 120% of the maximum angular velocity

[0234] like Figure 5 The diagram shown is a fault classification and diagnosis logic diagram. It monitors position sensor signals in real time and introduces a dual-threshold diagnosis mechanism based on position residual and signal abnormality duration threshold. This effectively solves the problem of misjudgment or missed judgment caused by noise and parameter perturbation associated with a single reference value in traditional methods, and greatly improves the robustness of the diagnosis strategy.

[0235] As a preferred implementation, considering the principles that the signal switching process must be smooth to avoid actuator jumps; the switching time must be determined and bounded; and the switching process must be monitorable and verifiable, a cosine square function switching function is designed as shown in the following formula:

[0236]

[0237]

[0238]

[0239] in, For the final position sensor value, To control the weights, For sensor state reconstruction values, For time, For the time of failure, The maximum allowable switching time for aircraft actuators is 10ms;

[0240] like Figure 6 The diagram shown is a timing diagram for fault-tolerant smooth switching. Figure 6The entire process from fault occurrence to fault-tolerant switching completion is demonstrated. During the fault occurrence and initial detection period of 5-20ms, the system detects the anomaly but awaits confirmation; during the switching period of 20-30ms, a cosine square function is used to smoothly transition the control weight α(t); after 30ms, the system enters a stable fault-tolerant operating mode. The entire process ensures the continuity of actuator position and control smoothness.

[0241] As a preferred implementation method, the overall structure of the fault-tolerant control system is as follows: Figure 7 As shown, in fault-tolerant mode, the system enters a sensorless operation state. To further improve robustness, the speed loop controller adopts an improved sliding mode control (ISMC) strategy. ISMC effectively suppresses the chattering problem inherent in traditional sliding mode control methods by introducing adaptive laws or boundary layer optimization. Its control law can be designed as follows:

[0242]

[0243]

[0244]

[0245] In the formula, For sliding surface, This is the slope coefficient of the sliding surface. For speed tracking error, The rate of change of error, For state vectors, For adaptively adjustable gain, For boundary layer thickness, For linear reaching law, It is a saturation function. It is a symbolic function.

[0246] As a preferred implementation, in step S3, when recording faults based on the diagnostic results, the fault record format must conform to the ATA specification, including information such as timestamp, fault type, environmental parameters, actuator status, and maintenance recommendations; at the same time, maintenance information such as MEL items, maintenance actions, and whether release conditions are met are automatically generated based on the fault type to guide maintenance personnel to perform additional inspections and maintenance on the actuators.

[0247] As shown in Table 2, different graded degradation strategies are implemented based on the severity of the fault, constructing an intelligent health management architecture deeply integrated with the Aircraft Health Management System (HUMS), realizing closed-loop management from fault tolerance to predictive maintenance. This architecture encapsulates real-time fault diagnosis information into standardized ATA specification data packages, automatically associates them with Maintenance Manual (MEL) items, and generates guiding maintenance recommendations, which can effectively reduce mean time to repair.

[0248] Table 2

[0249] Fault Level Downgrade strategy Performance limitations Remark Grade A (Mild) Normal mode, continuous monitoring Unrestricted Level B (Intermediate) Reduce bandwidth, limit rate The rate dropped to 80% Level C (Severe) Switch to fault-tolerant mode Only maintain basic functions Level D (Dangerous) Enter safe mode The actuator maintains its current position.

[0250] Simultaneously, based on the severity of the fault and environmental conditions, the system implements a four-level dynamic degradation strategy, from enhanced monitoring to safety maintenance, maximizing system availability while ensuring flight safety. All health data is synchronized to the ground maintenance system via the airborne data link, supporting full lifecycle performance trend analysis and spare parts prediction, forming an integrated health management system of "airborne real-time fault tolerance - ground intelligent decision-making," which significantly improves fleet operation and maintenance efficiency and flight safety margin.

[0251] Example 2

[0252] Please see Figure 8 , Figure 8 This is a schematic structural block diagram of an electronic device provided in Embodiment 2 of this application.

[0253] An electronic device includes a memory 101, a processor 102, and a communication interface 103. The memory 101, processor 102, and communication interface 103 are electrically connected directly or indirectly to enable data transmission or interaction. For example, these components can be electrically connected to each other via one or more communication buses or signal lines. The memory 101 can be used to store software programs and modules. The processor 102 executes the software programs and modules stored in the memory 101 to perform various functional applications and data processing. The communication interface 103 can be used for signaling or data communication with other node devices.

[0254] The memory 101 may be, but is not limited to, random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), etc.

[0255] The processor 102 can be an integrated circuit chip with signal processing capabilities. The processor 102 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0256] It is understood that the structure shown in the figure is for illustrative purposes only. An adaptive fault-tolerant control method for an aircraft actuator PMSM position sensor may include more or fewer components than shown in the figure, or have a different configuration. The components shown in the figure may be implemented in hardware, software, or a combination thereof.

[0257] In the embodiments provided in this application, it should be understood that the disclosed methods can also be implemented in other ways. The embodiments described above are merely illustrative. For example, the flowcharts or block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of methods and computer program products according to various embodiments of this application. In this regard, each block in the flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0258] In addition, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0259] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0260] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0261] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from the spirit or essential characteristics of this application. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of this application is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within this application. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. An adaptive fault-tolerant control method for a PMSM position sensor of an aircraft actuator, characterized in that, Includes the following steps: S1. Receive raw sensor signals from aircraft electromechanical actuators and perform signal preprocessing; S2. Adaptive state reconstruction and fault diagnosis are performed on the preprocessed raw sensor signals; S3. Record and classify faults based on diagnostic results to generate maintenance decisions; S4. Develop and validate based on maintenance decisions, and output a final signal that meets aviation safety standards.

2. The adaptive fault-tolerant control method for a PMSM position sensor of an aircraft actuator as described in claim 1, characterized in that, In step S1, the signal preprocessing includes temperature compensation, multilayer EMI filtering, and signal isolation and shielding; the temperature compensation is achieved by establishing a temperature compensation model, the model formula of which is: ， ， ， In the formula, This refers to the magnetic flux linkage of a permanent magnet at a winding temperature of T. , , This is the nominal value at 25℃. , , All are temperature coefficients. This is the current winding temperature. For reference temperature, This represents the resistance of the permanent magnet at a winding temperature of T. Let T be the inductance of the permanent magnet at a winding temperature of T.

3. The adaptive fault-tolerant control method for a PMSM position sensor of an aircraft actuator as described in claim 2, characterized in that, The multilayer EMI filtering process employs a finite impulse response filter and uses a window function design method to control the filter's frequency response characteristics; the expressions for the filter transfer function and the window function are as follows: ， ， In the formula, Let h[k] be the transfer function of the filter (represented in the frequency domain or z-domain), M be the filter order, h[k] be the k-th coefficient of the FIR filter, i.e., the impulse response sequence, k be the discrete-time index from 0 to M-1, and z be the complex variable in the z-transform, usually representing the unit delay in the discrete-time system. For Kaiser window functions, For zero-order modified Bessel functions, This is the discrete-time index of the window function. The point is the center of symmetry of the window function. These are the shape parameters of the Kaiser window.

4. The adaptive fault-tolerant control method for a PMSM position sensor of an aircraft actuator as described in claim 1, characterized in that, In step S2, the calculation process for the adaptive state reconstruction includes: Establishing a state-space model for aviation PMSM: ， ， ， ， ， ， ， ， Discretization is performed using the second-order Runge-Kutta method: ， ， ， State prediction, measurement prediction, and state update are performed using the capacitive Kalman filter algorithm. The calculation process for state prediction is as follows: ， ， ， ， ， The calculation process for measurement prediction is as follows: ， ， ， ， ， ， The calculation process for state updates is as follows: ， ， ， Online adaptive noise covariance estimation: ， ， ， ， ， ， ， Error reconstruction through dynamic error system modeling: ， ， ， ， ， In the formula, x is the state vector, Let t be the input vector and t be the time. , The state matrix, For process noise, For the observation vector, To observe the noise, For the α-axis current, For β-axis current, The rotor angular velocity, For rotor electrical angle, The voltage along the α-axis. For β-axis voltage, This represents the resistance of the permanent magnet at a winding temperature of T. Let T be the inductance of the permanent magnet at a winding temperature of T. This refers to the magnetic flux linkage of a permanent magnet at a winding temperature of T. for, Let T be the moment of inertia, and let T denote the transpose matrix. Let be the state vector at time k+1. Let k be the state vector at time k. Sampling time, The first-order slope of the Runge-Kutta slope. The second-order slope of the Runge-Kutta slope. The process noise at time k, Let K be the motor temperature at time k. Let k be the input vector at time k. The Cholesky factor is the covariance matrix. For Cholesky decomposition, The state estimation covariance matrix at time k-1 is given. For volume point, This is the state estimate at time k-1; For the i-th basic volume point, For the volume point after propagation, Let the first-order slope of Runge-Kutta be the volume point i. Let the first-order slope of Runge-Kutta be the volume point i. This is the predicted value of the system state. The covariance matrix of the state prediction values. The process noise covariance matrix; This is the covariance matrix of the state predictions after Cholesky decomposition. The parameters for the Cholesky decomposition form are chosen to represent the decomposition as a product of a lower triangular matrix and its transpose. This is the standard configuration for generating volume points using the capacitive Kalman filter. This is the predicted state value. This is the first measurement point of volume. For the observation matrix, The volume points generated for the current time step estimate and the covariance matrix generated from the previous time step after decomposition. For the system's measurement prediction values, The covariance matrix of the measured predicted values, The noise covariance matrix is... This is the cross-covariance matrix between the state predictions and the measurement predictions. For volumetric Kalman gain, This is the optimal estimate of the state. The actual observed value at time k. Update the covariance matrix; For the new information sequence, For theoretical covariance, For mathematical expectation, The update law for the noise covariance matrix, For the time index within the sliding window, For adaptively adjustable gain, It is a sliding window. This is the estimated actual covariance of the new sequence. For the information at time j, This is the estimated actual covariance of the new sequence. It is a residual sequence. The observed values ​​are remeasured after the state is updated at time k. For covariance estimation, Let j be the residual. Let k be the process noise covariance matrix at time k; For reconstruction error, For dynamic error, Here is the state transition matrix. For higher-order linearization error, It is composite noise. It is an identity matrix with the same dimensions as the state vector. Let f be the Jacobian matrix, and let f be the nonlinear continuous-time state transition function of the permanent magnet synchronous motor. The midpoint of the remainder term in the Taylor expansion. To connect line segments, connect and line segments, Let be the observation noise at time k.

5. The adaptive fault-tolerant control method for a PMSM position sensor of an aircraft actuator as described in claim 4, characterized in that, The fault diagnosis process includes: The electrical angle output from the position sensor is acquired in real time to calculate the position residual and the rate of change of the signal. The calculation formulas include: ， ， In the formula, For positional residuals, For electrical angle, This is the reconstructed electrical angle value from the position sensor. For time, Let k be the error of the position sensor output value. The error of the position sensor output value at time k-1, Sampling time; Fault diagnosis is performed based on position residuals and signal change rates.

6. The adaptive fault-tolerant control method for a PMSM position sensor of an aircraft actuator as described in claim 5, characterized in that, The fault diagnosis includes signal disconnection fault diagnosis, signal stagnation fault diagnosis, signal offset fault diagnosis, and complex fault diagnosis. The determination criteria for the signal disconnection fault diagnosis are as follows: ,ΔT≥T threshold， If all conditions are met simultaneously, it is determined to be a disconnection fault, and the fault flag Flag=1; The criteria for determining the signal stagnation fault diagnosis are as follows: , ,ΔT≥T threshold， If both conditions are met, it is determined to be a disconnection fault, and the fault flag Flag=2; The criteria for determining the signal offset fault are as follows: Δθ≥θ threshold ,ΔT≥T threshold， If all conditions are met simultaneously, it is determined to be a disconnection fault, and the fault flag Flag=3; In the formula, Let ΔT be the electrical angle, and ΔT be the duration. threshold This is the time threshold for abnormal position sensor signals. For time, The output value at a certain moment. For the position residual, θ threshold The threshold value for the residual of the position sensor signal; The complex fault diagnosis includes intermittent disconnection and disconnection-offset coupling faults. The diagnostic criteria are as follows: intermittent disconnection is manifested by the periodic switching of Flag between 1 and 0; disconnection-offset coupling is first triggered by Flag=1, and then immediately triggered by Flag=3 after the fault is recovered.

7. An electronic device, characterized in that, include: Memory, used to store one or more programs; processor; When the processor executes the one or more programs, it implements an adaptive fault-tolerant control method for an aircraft actuator PMSM position sensor as described in any one of claims 1-6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements an adaptive fault-tolerant control method for an aircraft actuator PMSM position sensor as described in any one of claims 1-6.

Images