Frequency domain characteristic-based aero-engine dual-channel sensor signal voting method

Abstract

The invention relates to an aero-engine dual-channel sensor signal voting method based on frequency domain characteristics, and the method comprises the steps: collecting a dual-channel sensor signal, carrying out the high-pass filtering and low-pass filtering of the dual-channel sensor signal, and obtaining a high-frequency signal and a low-frequency signal; setting an initial low-frequency signal health degree for the dual-channel low-frequency signal, and adjusting the low-frequency signal health degree through a preset low-frequency signal health judgment condition; setting an initial high-frequency signal health degree for the dual-channel high-frequency signal, and adjusting the high-frequency signal health degree through a preset high-frequency signal health judgment condition; and performing signal voting according to the low-frequency signal health degree and the high-frequency signal health degree after dual-channel adjustment, and selecting a trusted channel from the dual channels. Abnormalities such as sensor signal burrs, small-amplitude jumping and drifting caused by poor electrical contact in the working process of the engine can be accurately detected and isolated, and the running safety of the aero-engine is remarkably improved.

Description

Technical Field

[0001] The invention belongs to the technical field of aero-engines, and in particular relates to a voting method for aero-engine dual-channel sensor signals based on frequency domain characteristics. Background Art

[0002] Aircraft engine sensors mainly include speed sensors, temperature sensors, and pressure sensors, etc. The aircraft engine electronic controller uses the engine speed, temperature and pressure of each section and other state parameters collected by these sensors to implement control functions such as steady-state control, acceleration and deceleration control, and parameter limit protection of the aircraft engine, so as to achieve thrust regulation and safe operation of the full envelope of the aircraft engine.

[0003] Aircraft engine sensors generally adopt a dual-channel design, such as Figure 2 The sensor contains two redundancies, component A and component B, which pass through cables A and B respectively and enter the engine electronic controller for signal voting, which is used for control calculation and control output.

[0004] Failure modes such as electrical damage to the sensor's internal components, poor contact between the plug and socket, and cable damage can all lead to sensor signal distortion. Aircraft engine airworthiness regulations stipulate that the design and construction of the control system should ensure that an unsafe state will not occur due to a single failure or malfunction of the control system's electrical or electronic components.

[0005] When the engine is in use, the internal components of the sensor are electrically damaged, the cable is damaged and the connection or disconnection is virtual, and the plug and socket are loose or oil-soaked, resulting in poor contact. Under the influence of the high-frequency vibration of the engine itself, the circuit can be equivalent to the original sensor with an additional low-impedance circuit in series in a high-frequency on-off mode, making the sensor signal appear as signal glitches, small jumps and drifts. The existing dual-channel sensor signal voting method is to simply compare and select the dual-channel signals, which often cannot effectively detect and isolate abnormal signals. Abnormal signals may be used for control, resulting in engine control fluctuations, loss of thrust (LOTC) and even over-temperature, over-speed and other problems, affecting the flight safety of aircraft engines.

[0006] In summary, the existing dual-channel sensor signal voting method has insufficient coverage of fault modes and there is a problem that faults cannot be detected and isolated. It is hoped that an improved voting method for dual-channel aircraft engine sensor signals based on frequency domain characteristics is proposed without changing the existing engine hardware configuration, so as to accurately detect and isolate abnormal signals caused by poor signal contact during aircraft engine operation, thereby improving the safety of aircraft engine operation. Summary of the invention

[0007] The present invention provides a voting method for dual-channel sensor signals of an aircraft engine based on frequency domain characteristics, so as to detect and isolate abnormal sensor signals and select sensor signals with high credibility as control sources, thereby solving the technical problems existing in the background technology.

[0008] The technical solution of the present invention is as follows: comprising:

[0009] S10: collecting dual-channel sensor signals, performing high-pass filtering and low-pass filtering on the dual-channel sensor signals respectively, and obtaining high-frequency signals and low-frequency signals;

[0010] S20: setting an initial low-frequency signal health level for the low-frequency signals of the dual channels, and adjusting the low-frequency signal health level according to a preset low-frequency signal health judgment condition;

[0011] S30: setting an initial high-frequency signal health level for the high-frequency signals of the dual channels, and adjusting the high-frequency signal health level according to a preset high-frequency signal health judgment condition;

[0012] S40: performing signal voting according to the low-frequency signal health and the high-frequency signal health after adjustment of the dual channels, and selecting a trusted channel from the dual channels;

[0013] S50: Determine the final selected signal according to the signal of the trusted channel.

[0014] Furthermore, the method for obtaining the high-frequency signal and the low-frequency signal in step S10 includes:

[0015] The dual-channel sensor signal is low-pass filtered to form a low-frequency signal y(k);

[0016] The low-pass filtering method is:

[0017] a(0)·y(k)=b(0)·x(k)+b(1)·x(k-1)+……+b(n)·x(kn)

[0018] -a(1)·y(k-1)-……-a(n)·y(kn)

[0019] Wherein, y(k) represents the low-frequency signal of the current cycle, y(kn) represents the low-frequency signal of the previous n cycles, x(k) represents the collected value of the sensor signal of the current cycle, x(kn) represents the collected value of the sensor signal of the previous n cycles, n is the order of the low-pass filter, which is an integer greater than or equal to 1, and a(0), a(1), ... a(n) and b(0), b(1), ... b(n) are all filter coefficients;

[0020] The sensor signal is high-pass filtered to form a high-frequency signal yy(k);

[0021] The high-pass filtering method is:

[0022] aa(0)·yy(k)=bb(0)·x(k)+bb(1)·x(k-1)+……+bb(m)·x(km)

[0023] -aa(1)·yy(k-1)-……-aa(m)·yy(km)

[0024] Among them, yy(k) represents the high-frequency signal of the current period, yy(km) represents the high-frequency signal of the previous m periods, x(k) represents the collected value of the sensor signal of the current period, x(km) represents the collected value of the sensor signal of the previous m periods, m is the filter order, which is an integer greater than or equal to 1, aa(0), aa(1), ...aa(m) and bb(0), bb(1), ...bb(m) are all filter coefficients.

[0025] Furthermore, the low frequency signal health judgment condition in S20 includes:

[0026] When y(k)>Δ1 or y(k)<Δ2, the low-frequency signal health is set to -2;

[0027] When |y(k)-y(k-1)|>Δ3, the low-frequency signal health is set to -2;

[0028] In other cases, the health of the low-frequency signal in this cycle remains the health of the low-frequency signal in the previous cycle;

[0029] The initial value of the low-frequency signal health is 1, and Δ1, Δ2, and Δ3 are all judgment thresholds.

[0030] Furthermore, the high frequency signal health judgment condition in S30 includes:

[0031] If |yy(k)|>Δ4, the high-frequency signal health is reduced by 2;

[0032] If |yy(k)|>Δ4 is ​​not satisfied, the high-frequency signal healthiness increases by 1;

[0033] The initial value of the high-frequency signal health is 1, and Δ4 is ​​the judgment threshold.

[0034] Further, the S40 includes:

[0035] If the low-frequency signal health of the two channels is not equal, the channel with the larger low-frequency signal health value is selected as the trusted channel, and the number of trusted channels is 1;

[0036] If the health of the low-frequency signals of the two channels is equal and the health of the high-frequency signals of the two channels is not equal, the channel with the larger high-frequency signal health value is selected as the trusted channel, and the number of trusted channels is 1;

[0037] If the health of the low-frequency signals of the two channels is equal and the health of the high-frequency signals of the two channels is equal, both channels are used as trusted channels, and the number of trusted channels is 2.

[0038] Further, the S50 includes:

[0039] If the number of trusted channels is 1, the low-frequency signal of the trusted channel is selected as the final selected signal;

[0040] If the number of trusted channels is 2, the low-frequency signals of the two channels are averaged as the final selected signal.

[0041] Beneficial effects of the present invention: The present invention performs low-pass filtering and high-pass filtering on dual-channel sensor signals, obtains low-frequency signals within the engine response frequency band and high-frequency signals outside the engine response frequency band from the frequency domain perspective, and calculates the signal health of the low-frequency signals and high-frequency signals. A comprehensive signal vote is performed on the health of the low-frequency signals and the health of the high-frequency signals of the dual channels according to the signal frequency domain characteristics to determine the final selected signal. The present invention can accurately detect and isolate sensor signal glitches, small jumps, drifts and other anomalies caused by poor electrical contact during engine operation, significantly improving the safety of aircraft engine operation. The present invention is implemented using a software method, does not change the existing engine sensor and electronic controller hardware configuration, and saves manufacturing costs. At the same time, the method has low computational complexity and low time consumption, and is suitable for real-time operation in the engine electronic controller. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a flow chart of the present invention.

[0043] Figure 2 It is a schematic diagram of the use of dual-channel sensor signals in aircraft engine control systems.

[0044] Figure 3 is a flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. The described embodiment is only a part of the embodiment of the present invention, not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0046] Figure 2 This is the schematic diagram of the dual-channel sensor signal in the aircraft engine control system. Figure 2 The signal voting method, in an embodiment of the present invention, Figure 1 and Figure 3 The specific steps of the aircraft engine dual-channel sensor signal voting method based on frequency domain characteristics are given, such as Figure 1 and Figure 3 As shown, the present invention includes:

[0047] S10: collecting dual-channel sensor signals, performing high-pass filtering and low-pass filtering on the dual-channel sensor signals respectively, and obtaining high-frequency signals and low-frequency signals.

[0048] Among them, specifically Figure 2 As shown, the engine electronic controller collects dual-channel sensor signals respectively.

[0049] The sensor signal is low-pass filtered to form a low-frequency signal y(k).

[0050] The low-pass filtering method is:

[0051] a(0)·y(k)=b(0)·x(k)+b(1)·x(k-1)+……+b(n)·x(kn)

[0052] -a(1)·y(k-1)-……-a(n)·y(kn)

[0053] Among them, y(k) represents the low-frequency signal of the current period, y(kn) represents the low-frequency signal of the previous n periods, x(k) represents the collected value of the sensor signal of the current period, x(kn) represents the collected value of the sensor signal of the previous n periods, n is the order of the low-pass filter, which is an integer greater than or equal to 1, and a(0), a(1), ...a(n) and b(0), b(1), ...b(n) are all filter coefficients.

[0054] Regarding the low-pass filter design considerations, the low-pass filter can be designed based on a cutoff frequency of 1.5 to 2 times the actual response bandwidth of the engine, and the filter order n can be set to 2 to 6.

[0055] The sensor signal is high-pass filtered to form a high-frequency signal yy(k).

[0056] The high-pass filtering method is:

[0057] aa(0)·yy(k)=bb(0)·x(k)+bb(1)·x(k-1)+……+bb(m)·x(km)

[0058] -aa(1)·yy(k-1)-……-aa(m)·yy(km)

[0059] Among them, yy(k) represents the high-frequency signal of the current period, yy(km) represents the high-frequency signal of the previous m periods, x(k) represents the collected value of the sensor signal of the current period, x(km) represents the collected value of the sensor signal of the previous m periods, m is the filter order, which is an integer greater than or equal to 1, aa(0), aa(1), ... aa(m) and bb(0), bb(1), ... bb(m) are all filter coefficients.

[0060] Regarding high-pass filter design considerations, the high-pass filter can be designed based on a cutoff frequency of 2.5 to 3 times the engine response frequency, and the filter order m can be set to 2 to 6.

[0061] S20: setting an initial low-frequency signal health level for the low-frequency signals of the dual channels, and adjusting the low-frequency signal health level according to a preset low-frequency signal health judgment condition.

[0062] The specific conditions for judging the health of low-frequency signals are as follows:

[0063] If |y(k)-y(k-1)|>Δ1, the low-frequency signal health is set to -1;

[0064] If y(k)>Δ2 or y(k)<Δ3, the low-frequency signal health is set to -2;

[0065] In other cases, the health of the low-frequency signal in this cycle remains the health of the low-frequency signal in the previous cycle;

[0066] The initial value of the low-frequency signal health is 1, and Δ1, Δ2, and Δ3 are all judgment thresholds.

[0067] Regarding the design considerations of the judgment threshold, the maximum rate of change of the sensor signal under the engine working conditions, the maximum acquisition value of the sensor signal, and the minimum acquisition value of the sensor signal are used, while leaving a certain margin, as the values ​​of Δ1, Δ2, and Δ3 respectively.

[0068] S30: setting an initial high-frequency signal health level for the high-frequency signals of the dual channels, and adjusting the high-frequency signal health level according to a preset high-frequency signal health judgment condition.

[0069] The specific conditions for judging the health of high-frequency signals are as follows:

[0070] If |yy(k)|>Δ4, the high-frequency signal health is reduced by 2;

[0071] If |yy(k)|>Δ4 is ​​not satisfied, the high-frequency signal healthiness increases by 1;

[0072] If the high frequency signal health is less than -5, the high frequency signal health is set to -5; if the high frequency signal health is greater than 5, the high frequency signal health is set to 5.

[0073] The initial value of the high-frequency signal health is 1, and Δ4 is ​​the judgment threshold.

[0074] Regarding the threshold design considerations, a certain margin is reserved as the value of Δ4 based on the high-frequency noise characteristics of the sensor signal and the requirements of the engine's steady-state thrust fluctuation indicator for sensor acquisition accuracy.

[0075] S40: Perform signal voting according to the low-frequency signal health and the high-frequency signal health after adjustment of the dual channels, and select a trusted channel from the dual channels.

[0076] If the low-frequency signal health of the two channels is not equal, the channel with the larger low-frequency signal health value is selected as the trusted channel, and the number of trusted channels is 1;

[0077] If the health of the low-frequency signals of the two channels is equal and the health of the high-frequency signals of the two channels is not equal, the channel with the larger high-frequency signal health value is selected as the trusted channel, and the number of trusted channels is 1;

[0078] If the health of the low-frequency signals of the two channels is equal and the health of the high-frequency signals of the two channels is equal, both channels are used as trusted channels, and the number of trusted channels is 2.

[0079] S50: Calculate and obtain the final control selection signal according to the signal of the trust channel.

[0080] If the number of trusted channels is 1, the low-frequency signal of the trusted channel is selected as the final selected signal;

[0081] If the number of trusted channels is 2, the low-frequency signals of the two channels are averaged and used as the final selected signal.

[0082] Finally, it should be noted that the above specific implementation methods are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to examples, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

Claims

1. A voting method for aircraft engine dual-channel sensor signals based on frequency domain characteristics, characterized in that: include: S10: collecting dual-channel sensor signals, performing high-pass filtering and low-pass filtering on the dual-channel sensor signals respectively, and obtaining high-frequency signals and low-frequency signals; S20: setting an initial low-frequency signal health level for the low-frequency signals of the dual channels, and adjusting the low-frequency signal health level according to a preset low-frequency signal health judgment condition; S30: setting an initial high-frequency signal health level for the high-frequency signals of the dual channels, and adjusting the high-frequency signal health level according to a preset high-frequency signal health judgment condition; S40: performing signal voting according to the low-frequency signal health and the high-frequency signal health after adjustment of the dual channels, and selecting a trusted channel from the dual channels; S50: Determine the final selected signal according to the signal of the trusted channel.

2. The voting method for aircraft engine dual-channel sensor signals based on frequency domain characteristics according to claim 1, characterized in that: The method for obtaining the high frequency signal and the low frequency signal in step S10 includes: The dual-channel sensor signal is low-pass filtered to form a low-frequency signal y(k); The low-pass filtering method is: a(0)·y(k)=b(0)·x(k)+b(1)·x(k-1)+……+b(n)·x(kn) -a(1)·y(k-1)-……-a(n)·y(kn) Wherein, y(k) represents the low-frequency signal of the current cycle, y(kn) represents the low-frequency signal of the previous n cycles, x(k) represents the collected value of the sensor signal of the current cycle, x(kn) represents the collected value of the sensor signal of the previous n cycles, n is the order of the low-pass filter, which is an integer greater than or equal to 1, and a(0), a(1), ... a(n) and b(0), b(1), ... b(n) are all filter coefficients; The sensor signal is high-pass filtered to form a high-frequency signal yy(k); The high-pass filtering method is: aa(0)·yy(k)=bb(0)·x(k)+bb(1)·x(k-1)+……+bb(m)·x(km) -aa(1)·yy(k-1)-……-aa(m)·yy(km) Among them, yy(k) represents the high-frequency signal of the current period, yy(km) represents the high-frequency signal of the previous m periods, x(k) represents the collected value of the sensor signal of the current period, x(km) represents the collected value of the sensor signal of the previous m periods, m is the filter order, which is an integer greater than or equal to 1, aa(0), aa(1), ... aa(m) and bb(0), bb(1), ... bb(m) are all filter coefficients.

3. The voting method for aircraft engine dual-channel sensor signals based on frequency domain characteristics according to claim 2, characterized in that: The health judgment conditions of the low-frequency signal in S20 include: When y(k)>Δ1 or y(k)<Δ2, the low-frequency signal health is set to -2; When |y(k)-y(k-1)|>Δ3, the low-frequency signal health is set to -2; In other cases, the health of the low-frequency signal in this cycle remains the health of the low-frequency signal in the previous cycle; The initial value of the low-frequency signal health is 1, and Δ1, Δ2, and Δ3 are all judgment thresholds.

4. The voting method for aircraft engine dual-channel sensor signals based on frequency domain characteristics according to claim 2, characterized in that: The high frequency signal health judgment conditions in S30 include: If |yy(k)|>Δ4, the high-frequency signal health is reduced by 2; If |yy(k)|>Δ4 is ​​not satisfied, the high-frequency signal healthiness increases by 1; The initial value of the high-frequency signal health is 1, and Δ4 is ​​the judgment threshold.

5. The voting method for aircraft engine dual-channel sensor signals based on frequency domain characteristics according to claim 1, characterized in that: The S40 includes: If the low-frequency signal health of the two channels is not equal, the channel with the larger low-frequency signal health value is selected as the trusted channel, and the number of trusted channels is 1; If the health of the low-frequency signals of the two channels is equal and the health of the high-frequency signals of the two channels is not equal, the channel with the larger high-frequency signal health value is selected as the trusted channel, and the number of trusted channels is 1; If the health of the low-frequency signals of the two channels is equal and the health of the high-frequency signals of the two channels is equal, both channels are used as trusted channels, and the number of trusted channels is 2.

6. The aircraft engine dual-channel sensor signal voting method based on frequency domain characteristics according to claim 5, characterized in that: The S50 includes: If the number of trusted channels is 1, the low-frequency signal of the trusted channel is selected as the final selected signal; If the number of trusted channels is 2, the low-frequency signals of the two channels are averaged as the final selected signal.

Images