Sensor signal active fault tolerant control system for high electromagnetic pulse environments

Abstract

The application discloses a kind of sensor signal active fault-tolerant control system for strong electromagnetic pulse environment, belong to digital control fault-tolerant technical field.System includes the pre-processing module, signal identification module, pseudo-cluster separation module and signal reconstruction module connected in turn.Pretreatment module carries out FIR low-pass filtering to original sensor signal;Signal identification module generates abnormal flag by sliding window difference and energy mutation monitoring;Pseudo-cluster separation module carries out pseudo-cluster to continuous abnormal flag by finite state machine, and outputs interference section mask;Signal reconstruction module reconstructs signal to interference section based on mask, using local autoregressive prediction model.The scheme can identify and reconstruct sensor signal distorted due to strong electromagnetic pulse interference in real time and efficiently, significantly improve the reliability and robustness of digital control system such as aero-engine in complex electromagnetic environment, and is particularly suitable for FPGA platform implementation.

Description

Technical Field

[0001] This invention relates to the field of digital control fault-tolerant technology, and in particular to an active fault-tolerant control system for sensor signals in strong electromagnetic pulse environments. Background Technology

[0002] Digital controllers of critical infrastructure such as aero-engines and power systems are highly susceptible to coupling interference from strong electromagnetic pulses when operating in complex electromagnetic environments. This can cause severe distortion of signals collected by sensors, leading to control command errors and threatening the safe and stable operation of the entire system.

[0003] Traditional fault-tolerance techniques, such as hardware redundancy, software verification, and conventional filtering, have significant limitations when facing strong electromagnetic pulse interference: hardware redundancy is costly; software verification has poor real-time performance; and conventional filtering algorithms (such as low-pass filtering) cannot completely filter out residual high-energy pulse interference that overlaps with the effective signal frequency band. Furthermore, some software-based data-driven algorithms (such as K-means clustering and global autoregressive models) have high computational complexity, making them difficult to implement in real-time on resource-constrained embedded hardware platforms (such as FPGAs).

[0004] Therefore, there is an urgent need in this field for an active fault-tolerant technology that can balance high real-time performance, low resource consumption, and high accuracy in order to address the threat posed by strong electromagnetic pulses to sensor signals. Summary of the Invention

[0005] This invention provides an active fault-tolerant control system for sensor signals in strong electromagnetic pulse environments. It can identify, separate and reconstruct sensor signals interfered with by strong electromagnetic pulses in real time and accurately, while ensuring high system reliability and low processing latency, and is suitable for FPGA hardware platforms.

[0006] A first aspect of the present invention provides an active fault-tolerant control system for sensor signals in a strong electromagnetic pulse environment, comprising:

[0007] The preprocessing module is used to perform finite impulse response (FIR) low-pass filtering on the input raw sensor signal and output the filtered signal.

[0008] A signal recognition module, connected to the preprocessing module, is used to perform abrupt change detection on the filtered signal based on the dual criteria of sliding window difference and energy monitoring, and output an anomaly flag signal;

[0009] The pseudo-clustering separation module is connected to the signal recognition module. It uses a finite state machine to perform time-series logic analysis on the abnormal flag signal, merges the abnormal points that appear continuously in time into interference segments, and outputs an interference mask signal.

[0010] The signal reconstruction module, connected to the preprocessing module and the pseudo-clustering separation module, uses a local autoregressive prediction and state machine control compensation mechanism to interpolate and reconstruct the interference segment based on the interference mask signal and the filtered signal, and outputs a fault-tolerant signal.

[0011] Optionally, in one embodiment of the present invention, the FIR low-pass filter used for filtering is designed using the window function method, and the cutoff frequency is set according to the frequency characteristics of the sensor signal and the electromagnetic pulse, with the cutoff frequency set to 10kHz.

[0012] Optionally, in one embodiment of the present invention, the signal recognition module includes: a sliding window mean calculator for calculating the difference between the current window value and the historical mean of the window; an energy accumulator for calculating the sum of squares of the signals within the window; and a threshold comparator for comparing the difference and the sum of squares with a corresponding threshold, and outputting an abnormal flag when either the difference or the sum of squares exceeds the corresponding threshold.

[0013] Optionally, in one embodiment of the present invention, the finite state machine in the pseudo-clustering separation module includes three states: IDLE, ACTIVE, and GAP_WAIT, which are used to mark the boundaries of the interference segments and output the corresponding mask signals.

[0014] Optionally, in one embodiment of the present invention, the signal reconstruction module includes: NORMAL state: directly transmit the filtered signal; RECON state: generate compensation values ​​using a local autoregressive prediction model within the interference segment; FILL state: maintain the previous valid value output when the prediction window is insufficient; RESET state: resume transmission after the interference segment ends.

[0015] Optionally, in one embodiment of the present invention, the local autoregressive prediction model uses N valid samples before the interference to construct a weighted fitting model to predict the current interference point value, the mathematical expression of which is:

[0016]

[0017] Where p represents the length of the sliding window, a i Indicates empirical weights, This represents the estimated current distorted sample value, and x[ni] represents the historical valid sample value.

[0018] Optionally, in one embodiment of the present invention, the preprocessing module, the signal recognition module, the pseudo-clustering separation module, and the signal reconstruction module are all implemented on an FPGA using a hardware description language and employ a parallel pipeline structure for data processing.

[0019] A second aspect of the present invention provides an active fault-tolerant control method for sensor signals in a strong electromagnetic pulse environment, comprising the following steps:

[0020] Step 1: Perform finite impulse response (FIR) low-pass filtering on the original sensor signal and output the filtered signal;

[0021] Step 2: Perform a real-time sliding window scan on the filtered signal, and use the sliding window differential and energy monitoring dual criteria to detect abrupt changes in the filtered signal and generate an anomaly flag signal;

[0022] Step 3: Perform time-series logic analysis on the anomaly flag signal based on a finite state machine, merge the anomaly points that occur continuously in time into interference segments, and output an interference mask signal.

[0023] Step 4: Based on the interference mask signal and the filtered signal, the interference section is interpolated and reconstructed using a local autoregressive prediction and state machine control compensation mechanism, the signal in the normal section is passed through, and a fault-tolerant signal is output.

[0024] A third aspect of the present invention provides a digital electronic controller that integrates the active fault-tolerant control system for sensor signals in strong electromagnetic pulse environments described in the above embodiments.

[0025] A fourth aspect of the present invention provides an aircraft engine control system, including the digital electronic controller described in the above embodiments.

[0026] The active fault-tolerant control system for sensor signals in strong electromagnetic pulse environments according to embodiments of the present invention has the following beneficial effects:

[0027] 1. Strong real-time performance: Modular hardware design, low algorithm complexity, and parallel pipelined processing of the entire process on FPGA meet microsecond-level real-time requirements.

[0028] 2. Low resource consumption: The use of state machines to replace complex clustering algorithms and local prediction to replace global models greatly saves logical resources and storage overhead.

[0029] 3. High fault tolerance and accuracy: Dual criteria recognition combined with state machine accurately calibrates interference boundaries, and local autoregressive prediction effectively recovers signal waveforms with small reconstruction errors.

[0030] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0031] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0032] Figure 1This is a schematic diagram of an active fault-tolerant control system for sensor signals in a strong electromagnetic pulse environment, provided according to an embodiment of the present invention.

[0033] Figure 2 This is a schematic diagram of the FPGA implementation structure of the preprocessing module in an embodiment of the present invention.

[0034] Figure 3 This is a structural diagram of the FPGA implementation of the signal recognition module according to an embodiment of the present invention;

[0035] Figure 4 This is a state machine transition diagram of the pseudo-clustering separation module in an embodiment of the present invention;

[0036] Figure 5 This is a schematic diagram of the FPGA implementation of the signal reconstruction module according to an embodiment of the present invention.

[0037] Figure 6 This is a schematic flowchart of an active fault-tolerant control method for sensor signals in a strong electromagnetic pulse environment according to an embodiment of the present invention. Detailed Implementation

[0038] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0039] Figure 1 This is a schematic diagram of an active fault-tolerant control system for sensor signals in a strong electromagnetic pulse environment, provided by an embodiment of the present invention.

[0040] like Figure 1 As shown, the active fault-tolerant control system for sensor signals in strong electromagnetic pulse environments includes:

[0041] The preprocessing module is used to perform finite impulse response (FIR) low-pass filtering on the input raw sensor signal and output the filtered signal.

[0042] The signal recognition module, connected to the preprocessing module, is used to perform abrupt change detection on the filtered signal based on the dual criteria of sliding window differential and energy monitoring, and output anomaly flag signal;

[0043] The pseudo-clustering separation module is connected to the signal recognition module. It uses a finite state machine to perform time-series logic analysis on the abnormal flag signals, merges the abnormal points that appear continuously in time into interference segments, and outputs an interference mask signal.

[0044] The signal reconstruction module, connected to the preprocessing module and the pseudo-clustering separation module, uses a local autoregressive prediction and state machine control compensation mechanism to interpolate and reconstruct the interference section based on the interference mask signal and the filtered signal, and outputs a fault-tolerant signal.

[0045] In embodiments of the present invention, the FIR low-pass filter used for filtering is designed using the window function method to filter out high-frequency noise and some high-frequency interference in the sensor signal, providing a clean signal source for subsequent processing. The cutoff frequency is set according to the frequency characteristics of the sensor signal and electromagnetic pulse, preferably 10kHz, to ensure linear phase characteristics.

[0046] In an embodiment of the present invention, the signal recognition module includes: a sliding window mean calculator for calculating the difference between the current window value and the historical mean of the window; an energy accumulator for calculating the sum of squares of the signals within the window; and a threshold comparator for comparing the difference and the sum of squares with a corresponding threshold, and outputting an abnormal flag when either the difference or the sum of squares exceeds the corresponding threshold.

[0047] Understandably, the signal recognition module receives the preprocessed signal and employs a sliding window mechanism. Two detection strategies are used in parallel: one is the sliding window difference method, which calculates the difference between the current value and the historical mean of the window and compares it with a first threshold; the other is the energy mutation monitoring method, which calculates the sum of squares of the signals within the window and compares it with a second threshold. If either condition exceeds the threshold, it is determined to be an anomaly, and an anomaly flag signal is output. This dual-criteria mechanism improves the sensitivity and robustness of the detection.

[0048] In an embodiment of the present invention, the finite state machine in the pseudo-clustering separation module includes three states: IDLE, ACTIVE, and GAP_WAIT, which are used to mark the boundaries of the interference segments and output the corresponding mask signals.

[0049] Understandably, the pseudo-clustering separation module is one of the key innovations of this invention. It abandons the computationally complex traditional clustering algorithm and creatively adopts a "pseudo-clustering" strategy based on a finite state machine (FSM). This state machine performs temporal logic analysis on the input anomaly flag signal through three states: IDLE, ACTIVE, and GAP_WAIT. It groups temporally consecutive anomalies into the same "interference cluster segment" and outputs a corresponding binary mask signal to accurately mark the start and end boundaries of the interference.

[0050] In an embodiment of the present invention, the signal reconstruction module includes: NORMAL state: directly transmits the filtered signal; RECON state: generates compensation values ​​using a local autoregressive prediction model within the interference segment; FILL state: maintains the previous valid value output when the prediction window is insufficient; RESET state: resumes transmission after the interference segment ends.

[0051] Understandably, the signal reconstruction module receives the preprocessed signal and the mask signal. It determines the current point's state based on the mask signal: if it's a normal point, it outputs it directly; if it's an interference point, it activates the Local Autoregressive Prediction (LAR) model. This model uses a segment of normal historical data preceding the interference segment to predict the estimated value of the current point through weighted fitting, thereby achieving seamless signal reconstruction and ensuring the continuity and smoothness of the output signal.

[0052] In an embodiment of the present invention, the local autoregressive prediction model uses N valid samples before the interference to construct a weighted fitting model to predict the current interference point value. The mathematical expression is:

[0053]

[0054] Where p represents the length of the sliding window, a i Indicates empirical weights, This represents the estimated current distorted sample value, and x[ni] represents the historical valid sample value.

[0055] In the embodiments of the present invention, the preprocessing module, signal recognition module, pseudo-clustering separation module and signal reconstruction module are all implemented on FPGA using a hardware description language, and data processing is performed using a parallel pipeline structure, which greatly improves data throughput and processing real-time performance.

[0056] The active fault-tolerant control system for sensor signals in strong electromagnetic pulse environments according to the present invention will be described in detail below through a specific embodiment.

[0057] This embodiment uses the sensor signal fault tolerance of a certain type of aero-engine speed control system as an application scenario. The system input is the original sensor digital signal f[n] after sampling by an ADC, and the ADC sampling frequency f s The frequency is 10MHz and the quantization bit width is 14 bits. The final output of the system is a fault-tolerant signal y[n], which is used by the downstream controller.

[0058] The specific implementation of the preprocessing module;

[0059] like Figure 2 As shown, the core of the preprocessing module is a 64th-order FIR low-pass filter.

[0060] Filter design: The Hamming window method is used for design, with a cutoff frequency f. c Set to 10kHz. The filter coefficients h[0:63] are generated by calling the fir1(63,10000 / (10000000 / 2)) function in MATLAB.

[0061] Coefficient quantization: To ensure the accuracy of FPGA implementation and save resources, the generated floating-point coefficients are quantized into 16-bit signed integers and stored in the FPGA's ROM.

[0062] FPGA implementation structure: Implemented using a transposed direct form architecture. This structure allows multipliers to operate in parallel and naturally forms a pipeline, offering the advantage of high throughput.

[0063] Input data f[n] enters a shift register array with a depth of 64. A 64-multiplier array computes f[nk]*h[k] in parallel (k = 0, 1, ..., 63). This embodiment uses the DSP48E1 hard core unit inside the FPGA to perform the multiplication operation, improving efficiency and reducing logic resource consumption. A multi-stage pipelined addition tree adds the 64 product results. The addition tree is divided into 6 stages, each inserting a pipelined register to achieve a high clock frequency. The final output is a filtered signal x[n], whose bit width is controlled to 18 bits after convergence processing to prevent data overflow.

[0064] The specific implementation of the signal recognition module;

[0065] like Figure 3 As shown, the signal recognition module executes the sliding window difference method and the energy mutation monitoring method in parallel.

[0066] Sliding window settings: The length W of the sliding window is fixed at 100 sampling points.

[0067] Amplitude difference method: Uses a sliding accumulator to calculate the sum of the first 99 historical samples (x[n-99] to x[n-1]) within the window in real time. hist Each clock cycle, sum hist Updated to: sum hist =sum hist -x[n-100]+x[n-1]. Mean μ=sum hist / 99. To conserve divider resources, this division is achieved by multiplying by a pre-calculated reciprocal approximation and shifting to the right. Calculate the absolute value difference. diff [n] = |x[n] - μ|. Set the first threshold T. diff This threshold can be dynamically configured according to the system operating conditions; in this embodiment, it is set to 0.5V. Given that the ADC range is ±5V (14-bit quantization, 1LSB≈0.61mV), the corresponding digital threshold is 0.5 / 0.00061≈819LSB.

[0068] The comparator determines if... diff [n]>T diff Then the amplitude difference condition is triggered.

[0069] Energy mutation monitoring method:

[0070] Calculate the approximate energy values ​​for 100 points within the current window. Similarly, using sliding motion calculation: E[n] = E[n-1] - x[n-100] 2 +x[n] 2 The squaring operation is implemented using a dedicated multiplier IP core.

[0071] Set the second threshold T energy In this embodiment, the average energy E of the normal signal is set. avg Five times that. avg It is calculated by the system during its self-learning phase after startup.

[0072] Comparator judgment: If E[n] > T energy Then the energy condition is triggered.

[0073] Logic Synthesis: The two conditions above are synthesized through an OR gate. If either condition is true, the output exception flag signal flag[n] is 1; otherwise, it is 0. flag[n] is a 1-bit wide signal.

[0074] The specific implementation of the pseudo-clustering separation module;

[0075] like Figure 4 As shown, the pseudo-clustering separation module implements its core logic using a finite state machine (FSM).

[0076] State Definition: FSM contains three states: IDLE, ACTIVE, and GAP_WAIT.

[0077] State transition logic: IDLE→ACTIVE: When the IDLE state detects that flag[n] = 1, it enters the ACTIVE state and assigns a unique identifier to the new interference segment.

[0078] ACTIVE→GAP_WAIT: In ACTIVE state, if flag[n] = 0 is detected, a gap counter (gapc) is started. ounter ), and enter the GAP_WAIT state.

[0079] GAP_WAIT→ACTIVE: In the GAP_WAIT state, if gap counter If flag[n] = 1 is detected before the limit is exceeded, then the gap is reset. counter It then returns to the ACTIVE state and assigns the current point to the previous interference segment.

[0080] GAP_WAIT→IDLE: If gap... counter The count exceeded the preset gap threshold T. idle(In this embodiment, it is set to 10 consecutive clock cycles), then it is considered that the current interference segment has ended, and the state machine returns to IDLE.

[0081] Output logic: In ACTIVE state, the output segment mask mask[n] = 1; in IDLE and GAP_WAIT states, the output segment mask mask[n] = 0.

[0082] The specific implementation of the signal reconstruction module;

[0083] like Figure 5 As shown in Table 1, the signal reconstruction module operates according to the value of mask[n].

[0084] Table 1. State transition table in the FPGA implementation structure diagram of the signal reconstruction module.

[0085]

[0086]

[0087] Prediction model: A Local Autoregressive (LAR) prediction model is used, with a prediction window length N. pred Set it to 20.

[0088] Weighting coefficients: The weighting coefficients ω[1], ω[2], ..., ω

[20] are a set of fixed values, which are obtained by collecting a large amount of historical data from normal sensors and performing offline training fitting using the least squares method, and are pre-programmed into the ROM of the FPGA. This method avoids the complexity of online calculation and greatly saves FPGA resources.

[0089] Implementation structure:

[0090] The module maintains a circular buffer (FIFO) of depth 20, continuously storing the latest 20 valid sample values ​​x when mask[n] = 0. hist [1],x hist [2],…x hist

[20]

[0091] When mask[n] changes from 0 to 1, the module enters the RECON state. At this time, forward prediction is performed using 20 historical values ​​in the buffer:

[0092] y[n]=ω[1]*x hist [1]+ω[2]*x hist [2]+…+ω

[20] *x hist

[20]

[0093] The multiplication and accumulation operation is completed within 4 clock cycles via a parallel MAC chain. If interference occurs at the beginning of the signal and there are fewer than 20 valid historical points, the module enters the fill state, repeatedly outputting the last valid value until there is enough data. When mask[n] returns to 0, the module switches back to the normal state, resumes direct pass-through of x[n], and updates the historical buffer.

[0094] All modules are implemented on a Xilinx Kintex-7 series XC7K325T FPGA, with a system master clock frequency of 100MHz. The modules interact via an AXI-Stream interface to achieve efficient pipelining. Key threshold parameters (such as T...) diff ,T energy ,T idle It can be dynamically adjusted through the configuration register connected to the APB bus to adapt to different application scenarios.

[0095] Next, referring to the accompanying drawings, we describe the active fault-tolerant control method for sensor signals in a strong electromagnetic pulse environment proposed according to an embodiment of the present invention.

[0096] Figure 6 This is a schematic flowchart of an active fault-tolerant control method for sensor signals in a strong electromagnetic pulse environment according to an embodiment of the present invention.

[0097] Step 1: Perform finite impulse response (FIR) low-pass filtering on the original sensor signal and output the filtered signal;

[0098] Step 2: Perform real-time scanning of the filtered signal using a sliding window, and use the sliding window differential and energy monitoring dual criteria to detect abrupt changes in the filtered signal and generate an anomaly flag signal;

[0099] Step 3: Perform timing logic analysis on the anomaly flag signal based on the finite state machine, merge the anomaly points that occur continuously in time into interference segments, and output the interference mask signal.

[0100] Step 4: Based on the interference mask signal and the filtered signal, a local autoregressive prediction and state machine control compensation mechanism is used to interpolate and reconstruct the interference section, and the signal of the normal section is passed through to output a fault-tolerant signal.

[0101] The present invention discloses an active fault-tolerant control system and method for sensor signals in strong electromagnetic pulse environments. The preprocessing module performs FIR low-pass filtering on the original sensor signal; the signal identification module generates anomaly flags through sliding window differential and energy mutation monitoring; the pseudo-clustering separation module performs pseudo-clustering on continuous anomaly flags using a finite state machine, outputting an interference segment mask; and the signal reconstruction module reconstructs the signal of the interference segment based on the mask using a local autoregressive prediction model. This scheme can identify and reconstruct sensor signals distorted by strong electromagnetic pulse interference in real time and efficiently, significantly improving the reliability and robustness of digital control systems such as those for aero-engines in complex electromagnetic environments, and is particularly suitable for implementation on FPGA platforms.

[0102] It should be noted that the foregoing explanation of the active fault-tolerant control system embodiment for sensor signals in strong electromagnetic pulse environments also applies to the active fault-tolerant control method for sensor signals in strong electromagnetic pulse environments in this embodiment, and will not be repeated here.

[0103] A third aspect of the present invention provides a digital electronic controller that integrates the active fault-tolerant control system for sensor signals in strong electromagnetic pulse environments described in the above embodiments.

[0104] A fourth aspect of the present invention provides an aircraft engine control system, including the digital electronic controller described in the above embodiments.

[0105] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0106] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0107] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

Claims

1. An active fault-tolerant control system for sensor signals in a strong electromagnetic pulse environment, characterized in that, include: The preprocessing module is used to perform finite impulse response (FIR) low-pass filtering on the input raw sensor signal and output the filtered signal. A signal recognition module, connected to the preprocessing module, is used to perform abrupt change detection on the filtered signal based on the dual criteria of sliding window difference and energy monitoring, and output an anomaly flag signal; The signal recognition module includes: a sliding window mean calculator for calculating the difference between the current window value and the historical mean of the window; an energy accumulator for calculating the sum of squares of the signals within the window; and a threshold comparator for comparing the difference and the sum of squares with corresponding thresholds, and outputting an abnormal flag when either the difference or the sum of squares exceeds the corresponding threshold. The pseudo-clustering separation module is connected to the signal recognition module. It uses a finite state machine to perform time-series logic analysis on the abnormal flag signal, merges the abnormal points that appear continuously in time into interference segments, and outputs an interference mask signal. The signal reconstruction module, connected to the preprocessing module and the pseudo-clustering separation module, performs interpolation reconstruction of the interference segment based on the interference mask signal and the filtered signal, using a local autoregressive prediction and state machine control compensation mechanism, and outputs a fault-tolerant signal. The signal reconstruction module includes: NORMAL state: directly transmits the filtered signal; RECON state: generates compensation values ​​in the interference segment using a local autoregressive prediction model; FILL state: maintains the previous valid value output when the prediction window is insufficient; RESET state: resumes transmission after the interference segment ends. The local autoregressive prediction model uses N valid samples before the interference to construct a weighted fitting model to predict the current interference point value. The mathematical expression is: in, Indicates the length of the sliding window. Indicates empirical weights, This indicates an estimate of the current distorted sample value. This represents the historical valid sample value.

2. The active fault-tolerant control system for sensor signals according to claim 1, characterized in that, The FIR low-pass filter used for filtering is designed using the window function method. The cutoff frequency is set according to the frequency characteristics of the sensor signal and the electromagnetic pulse, and the cutoff frequency is set to 10kHz.

3. The active fault-tolerant control system for sensor signals according to claim 1, characterized in that, The finite state machine in the pseudo-clustering separation module includes three states: IDLE, ACTIVE, and GAP_WAIT, which are used to mark the boundaries of the interference segments and output the corresponding interference mask signal.

4. The active fault-tolerant control system for sensor signals according to claim 1, characterized in that, The preprocessing module, the signal recognition module, the pseudo-clustering separation module, and the signal reconstruction module are all implemented on the FPGA using a hardware description language and employ a parallel pipeline structure for data processing.

5. A method for active fault-tolerant control of sensor signals in a strong electromagnetic pulse environment, characterized in that, The sensor signal active fault-tolerant control system as described in any one of claims 1-4, wherein the sensor signal active fault-tolerant control method includes the following steps: Step 1: Perform finite impulse response (FIR) low-pass filtering on the original sensor signal and output the filtered signal; Step 2: Perform a real-time sliding window scan on the filtered signal, and use the sliding window differential and energy monitoring dual criteria to detect abrupt changes in the filtered signal and generate an anomaly flag signal; Step 3: Perform time-series logic analysis on the anomaly flag signal based on a finite state machine, merge the anomaly points that occur continuously in time into interference segments, and output an interference mask signal. Step 4: Based on the interference mask signal and the filtered signal, the interference section is interpolated and reconstructed using a local autoregressive prediction and state machine control compensation mechanism, the signal in the normal section is passed through, and a fault-tolerant signal is output.

6. A digital electronic controller, characterized in that, It integrates an active fault-tolerant control system for sensor signals in strong electromagnetic pulse environments as described in any one of claims 1-4.

7. An aircraft engine control system, characterized in that, Includes the digital electronic controller as described in claim 6.

Images