Dynamic diagnostic methods, systems, media and equipment for sensor faults

Abstract

A method, system, medium, and device for dynamic diagnosis of sensor faults in an engine control system are disclosed. In this method, an engine component-level model calculates predicted values ​​of air path parameters based on input parameters; calculates the residual signal between the sensor measurements of each air path parameter and the corresponding predicted value; filters the residual signal; calculates the root mean square error between the predicted value of the engine low-pressure rotor speed and the measurement value of the low-pressure rotor speed sensor, and records this as the prediction error of the low-pressure rotor speed; uses the prediction error of the low-pressure rotor speed as a benchmark to calculate a dynamic threshold for fault diagnosis of other air path sensors; sets a fault indicator; if the residual between the sensor measurement value and the model prediction value is greater than the dynamic threshold, the indicator value increases; otherwise, the indicator value decreases; when the fault indicator value exceeds a set counting boundary, a sensor fault is reported.

Description

Technical Field

[0001] This invention relates to the field of engine sensor fault diagnosis technology, and in particular to a dynamic fault diagnosis method, system, medium and equipment for engine control system fault tolerance. Background Technology

[0002] As a crucial signal source in engine control systems, the reliability of sensor measurement signals directly impacts engine safety. As vital components operating in harsh environments such as high temperature, high pressure, and strong vibration, sensors are prone to failure, which in turn affects the calculation results of the engine controller. Fault detection, isolation, and handling of sensors are essential aspects of the fault-tolerant design of engine control systems.

[0003] Sensor fault diagnosis based on the residual signal between sensor measurements and engine model predictions is an important diagnostic method. The magnitude of the residual is compared with a fault threshold to determine the sensor's fault condition. Due to the complexity of engine service conditions and the variability of operating states, the magnitude of the prediction residual varies with the engine's thermal state. Furthermore, a large fixed fault threshold may lead to a large number of missed faults, while a small fixed fault threshold may cause false alarms, affecting the effectiveness of subsequent control actions of the control system. Therefore, it is necessary to design a dynamic fault threshold that is related to the engine's thermal state to accurately and reliably diagnose sensor faults.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of the present invention, and therefore may contain information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] This invention provides a method, system, medium, and device for dynamic diagnosis of sensor faults in engine control systems. Based on the prediction error of parameters related to engine thermal state, such as low-pressure speed, dynamic thresholds corresponding to other sensors besides low-pressure speed are calculated for sensor fault diagnosis.

[0006] Dynamic diagnostic methods for sensor faults aimed at improving the fault tolerance of engine control systems include:

[0007] Step 100: Detect and record the sensor measurements of the engine, and calculate the predicted values ​​of the air path parameters of the engine component-level model based on the engine component-level model corresponding to the engine.

[0008] Step 200: Calculate the residual signal between the measured values ​​of each gas path parameter sensor and the predicted values ​​of the corresponding gas path parameter, and filter the residual signal;

[0009] Step 300: Calculate the root mean square error of the predicted low-pressure rotor speed, denoted as the prediction error of the low-pressure rotor speed, and use the prediction error as a benchmark to calculate the dynamic threshold corresponding to the fault diagnosis of other gas path sensors.

[0010] Step 400: Set the sensor fault detection counter and calculate the value of the fault detection counter by comparing the filtered residual signal with the dynamic threshold.

[0011] Step 500: Compare the counter value with the counter setting value. If the counter value exceeds the setting value, report a fault.

[0012] In the aforementioned dynamic diagnostic method for sensor faults aimed at the fault tolerance of engine control systems, the engine includes a low-pressure rotor, a high-pressure rotor, a low-pressure compressor, a high-pressure compressor, a high-pressure turbine, and a low-pressure turbine.

[0013] In the aforementioned dynamic diagnostic method for sensor faults aimed at improving the fault tolerance of engine control systems, the sensor measurements include engine input parameters and airflow parameters, and the input parameter I at time t... E (t)={T t0 (t), P t0 (t), W f (t),A8(t)}, where T t0 (t) represents the total intake temperature, P t0 (t) represents the total intake pressure, W f (t) represents the fuel flow rate, A8(t) represents the nozzle area, and the gas path parameters O at time t are... E (t)={N L,E (t), N H,E (t), P t25,E (t), P t31,E (t), T t6,E (t)}, where N L,E (t) represents the measured value of the low-pressure rotor speed, N H,E (t) represents the measured value of the high-pressure rotor speed, P t25,E (t) represents the measured pressure at the outlet of the low-pressure compressor, P t31,E (t) represents the measured pressure at the outlet of the high-pressure compressor, T. t6,E (t) represents the measured temperature at the outlet of the low-pressure turbine.

[0014] In the aforementioned dynamic diagnostic method for sensor faults aimed at improving the fault tolerance of engine control systems, based on the engine input parameter I... E (t) Calculate the predicted gas path parameters O at time t. M (t)={N L,M (t), N H,M (t), Pt25,M (t), P t31,M (t), T t6,M (t)}, where N L,M (t) represents the predicted low-pressure rotor speed calculated from the engine component-level model, N. H,M (t) represents the predicted high-pressure rotor speed calculated from the engine component-level model, P t25,M (t) represents the predicted pressure at the low-pressure compressor outlet calculated from the engine component-level model, P. t31,M (t) represents the predicted pressure at the high-pressure compressor outlet calculated from the engine component-level model, T. t6,M (t) represents the predicted temperature at the low-pressure turbine outlet calculated from the engine component-level model.

[0015] In the aforementioned dynamic diagnostic method for sensor faults in engine control systems, in step 200, the method for calculating the predicted residual is as follows: [The method involves] calculating the sensor N... H P t25 P t31 and T t6 The numbers are sequentially recorded as 1 to 4, and the measurement value of the i-th sensor at time t is recorded as O. E,i (t), let O be the predicted value of the i-th sensor at time t. M,i (t), calculate the prediction residual of the i-th sensor at time t. .

[0016] In the aforementioned dynamic diagnostic method for sensor faults in engine control systems, in step 200, the formula for calculating the filtering of the residual signal is:

[0017]

[0018] in, is the smoothed residual at time t, and θ is the smoothing factor between 0 and 1.

[0019] In the aforementioned dynamic diagnostic method for sensor faults aimed at the fault tolerance of engine control systems, in step 300, the predicted value N of the low-pressure rotor speed at time t is calculated. L,M The root mean square error R(t) corresponding to (t) is used as a dynamic threshold for fault diagnosis of the high-pressure rotor speed sensor, the low-pressure compressor outlet pressure sensor, the high-pressure compressor outlet pressure sensor, and the low-pressure turbine outlet temperature sensor. The iterative calculation formula for the root mean square error R(t) is as follows:

[0020] , where L R This calculates the window length, R(t-1) is the root mean square error at time t-1, and N... L,M (tL R) for tL R Predicted low-pressure rotor speed at time N L,E (tL R ) for tL R The low-pressure rotor speed sensor reading at any given time.

[0021] A dynamic fault diagnosis system for sensors aimed at improving the fault tolerance of engine control systems includes:

[0022] The measurement unit detects and records the sensor measurements of the engine's air path parameters, and calculates the predicted values ​​of the air path parameters of the engine component-level model based on the engine component-level model corresponding to the engine.

[0023] The residual unit is used to: calculate the residual signal between the measured value of each gas path parameter sensor and the predicted value of the corresponding gas path parameter, and filter the residual signal;

[0024] The threshold calculation unit is used to: calculate the root mean square error of the predicted low-pressure rotor speed, denoted as the prediction error of the low-pressure rotor speed, and use the prediction error as a benchmark to calculate the dynamic threshold corresponding to the fault diagnosis of other gas path sensors.

[0025] The fault detection counter is used to calculate the value of the fault detection counter by comparing the filtered residual signal with the dynamic threshold.

[0026] The comparison unit is used to compare the counter value with the counter set value. If the counter value exceeds the set value, a fault is reported.

[0027] A computer storage medium including computer instructions that, when run on a computer, cause the computer to perform the method.

[0028] An electronic device, the electronic device comprising:

[0029] Memory, processor, and computer programs stored in memory and executable on the processor, wherein,

[0030] The processor implements the method when executing the program.

[0031] Compared with existing technologies, the beneficial effects of this disclosure are as follows: The low-pressure rotor speed is an important parameter representing the engine's thermal state, and the measurement reliability and accuracy of the speed signal are both higher. The prediction error of the low-pressure rotor speed is used as the design benchmark for the dynamic threshold of other sensor fault diagnosis. By combining sensor signal filtering, dynamic thresholding, and fault counters, the high false alarm rate and high false alarm rate problems of sensor fault diagnosis methods based on fixed thresholds are solved. It can promptly diagnose sensor bias and drift faults and features a simple algorithm that is easy to implement. Attached Figure Description

[0032] Various other advantages and benefits of the present invention will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiments below. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. It is obvious that the drawings described below are merely some embodiments of the invention, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort. Furthermore, the same reference numerals denote the same parts throughout the drawings.

[0033] In the attached diagram:

[0034] Figure 1 This is an architecture diagram of a sensor fault dynamic diagnosis method for fault tolerance in engine control systems provided in one embodiment of this disclosure;

[0035] Figure 2(a) shows a sensor P provided in an embodiment of this disclosure. t25 Schematic diagram of bias fault, Figure 2(b) is a sensor P provided in an embodiment of this disclosure. t25 Schematic diagram of drift fault;

[0036] Figure 3(a) shows a sensor P provided in an embodiment of this disclosure. t25 A schematic diagram of the bias fault diagnosis results, Figure 3(b) shows a sensor P provided in an embodiment of this disclosure. t25 A schematic diagram of the drift fault diagnosis results.

[0037] The present invention will be further explained below with reference to the accompanying drawings and embodiments. Detailed Implementation

[0038] Specific embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While specific embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0039] It should be noted that certain terms are used in the specification and claims to refer to specific components. Those skilled in the art will understand that different terms may be used to refer to the same component. This specification and claims do not distinguish components based on differences in terminology, but rather on differences in function. The terms "comprising" or "including" used throughout the specification and claims are open-ended and should be interpreted as "comprising but not limited to." The following descriptions are preferred embodiments for carrying out the invention; however, these descriptions are for the purpose of understanding the general principles of the specification and are not intended to limit the scope of the invention. The scope of protection of this invention is determined by the appended claims.

[0040] To facilitate understanding of the embodiments of the present invention, further explanations and descriptions will be provided below with reference to the accompanying drawings and specific embodiments. The accompanying drawings do not constitute a limitation on the embodiments of the present invention.

[0041] like Figures 1 to 3(b) As shown, the dynamic diagnostic method for sensor faults aimed at improving the fault tolerance of engine control systems includes:

[0042] Step 100: Detect and record the sensor measurements of the engine, and calculate the predicted values ​​of the air path parameters of the engine component-level model based on the engine component-level model corresponding to the engine.

[0043] Step 200: Calculate the residual signal between the measured values ​​of each gas path parameter sensor and the predicted values ​​of the corresponding gas path parameter, and filter the residual signal;

[0044] Step 300: Calculate the root mean square error of the predicted low-pressure rotor speed, denoted as the prediction error of the low-pressure rotor speed, and use the prediction error as a benchmark to calculate the dynamic threshold corresponding to the fault diagnosis of other gas path sensors.

[0045] Step 400: Set the sensor fault detection counter and calculate the value of the fault detection counter by comparing the filtered residual signal with the dynamic threshold.

[0046] Step 500: Compare the counter value with the counter setting value. If the counter value exceeds the setting value, report a fault.

[0047] In a preferred embodiment of the sensor fault dynamic diagnosis method for fault tolerance of engine control system, the engine includes a low-pressure rotor, a high-pressure rotor, a low-pressure compressor, a high-pressure compressor, a high-pressure turbine, and a low-pressure turbine.

[0048] In a preferred embodiment of the sensor fault dynamic diagnosis method for engine control system fault tolerance, the sensor measurements include engine input parameters and air path parameters, and the input parameter I at time t...E (t)={T t0 (t),P t0 (t), W f (t), A8(t)}, where T t0 (t) represents the total intake temperature, P t0 (t) represents the total intake pressure, W f (t) represents the fuel flow rate, A8(t) represents the nozzle area, and the gas path parameters O at time t are... E (t)={N L,E (t), N H,E (t), P t25,E (t), P t31,E (t), T t6,E (t)}, where N L,E (t) represents the measured value of the low-pressure rotor speed, N H,E (t) represents the measured value of the high-pressure rotor speed, P t25,E (t) represents the measured pressure at the outlet of the low-pressure compressor, P t31,E (t) represents the measured pressure at the outlet of the high-pressure compressor, T. t6,E (t) represents the measured temperature at the outlet of the low-pressure turbine.

[0049] In a preferred embodiment of the sensor fault dynamic diagnosis method for the fault tolerance of the engine control system, the method is based on the engine input parameter I. E (t) Calculate the predicted gas path parameters O at time t. M (t)={N L,M (t), N H,M (t),P t25,M (t), P t31,M (t), T t6,M (t)}, where N L,M (t) represents the predicted low-pressure rotor speed calculated from the engine component-level model, N. H,M (t) represents the predicted high-pressure rotor speed calculated from the engine component-level model, P t25,M (t) represents the predicted pressure at the low-pressure compressor outlet calculated from the engine component-level model, P. t31,M (t) represents the predicted pressure at the high-pressure compressor outlet calculated from the engine component-level model, T. t6,M (t) represents the predicted temperature at the low-pressure turbine outlet calculated from the engine component-level model.

[0050] In a preferred embodiment of the sensor fault dynamic diagnosis method for engine control system fault tolerance, in step 200, the prediction residual calculation method is as follows: [The text abruptly ends here, likely due to an incomplete sentence or a formatting error.] H P t25 P t31 and Tt6 The numbers are sequentially recorded as 1 to 4, and the measurement value of the i-th sensor at time t is recorded as O. E,i (t), let O be the predicted value of the i-th sensor at time t. M,i (t), calculate the prediction residual of the i-th sensor at time t. .

[0051] In a preferred embodiment of the sensor fault dynamic diagnosis method for engine control system fault tolerance, the filtering calculation formula for the residual signal is as follows:

[0052]

[0053] in, is the smoothed residual at time t, and θ is the smoothing factor between 0 and 1.

[0054] In a preferred embodiment of the sensor fault dynamic diagnosis method for engine control system fault tolerance, in step 300, the predicted value N of the low-pressure rotor speed at time t is calculated. L,M The root mean square error R(t) of (t) is used to calculate the predicted low-pressure rotor speed N. L,M The dynamic threshold for fault diagnosis of the high-pressure rotor speed sensor, the low-pressure compressor outlet pressure sensor, the high-pressure compressor outlet pressure sensor, and the low-pressure turbine outlet temperature sensor other than (t), and the root mean square error R(t) are calculated iteratively using the following formula:

[0055] , where L R This calculates the window length, R(t-1) is the root mean square error at time t-1, and N... L,M (tL R ) for tL R Predicted low-pressure rotor speed at time N L,E (tL R ) for tL R The low-pressure rotor speed sensor reading at any given time.

[0056] In a preferred embodiment of the sensor fault dynamic diagnosis method for fault-tolerant engine control systems, the root mean square error of the low-pressure rotor speed is used to calculate the dynamic threshold for fault diagnosis of the i-th sensor as a function of the root mean square error R(t), expressed as:

[0057] ,

[0058] Where, δ i (t) is the dynamic threshold for fault diagnosis of the i-th sensor, ε i,k It is the proportionality constant, ε i,b It is the bias coefficient.

[0059] In a preferred embodiment of the sensor fault dynamic diagnosis method for fault tolerance in engine control systems, in step 400, the fault detection counter for the i-th sensor is represented as follows:

[0060]

[0061] The counter value is a non-negative number, i.e., m. i (t)≥0.

[0062] In a preferred embodiment of the sensor fault dynamic diagnosis method for fault tolerance in engine control systems, in step 400, the method for calculating the value of the fault detection counter is as follows: when the residual The absolute value is less than the dynamic threshold δ i At time (t), the counter m i (t) decreases if m i (t) < 0, then m i (t) is set to 0, when the residual The absolute value exceeds the dynamic threshold δ i At time (t), the counter m i (t) increases.

[0063] In one embodiment, in step 500, the fault detection counter reports faults in the following manner: if the smoothed residual... Exceeding the dynamic threshold δ for a continuous period of time i (t), that is, satisfying m i (t)>m f,i When the i-th sensor fails, a fault is reported. The counting boundary m... f,i This is a constant, and its value can be set according to specific needs.

[0064] In one embodiment, such as Figure 1 As shown, the method includes,

[0065] Step 100: Record the sensor measurements of the engine's air path parameters and the predicted air path parameters of the engine component-level model;

[0066] Step 200: Calculate the predicted residuals of the high-pressure rotor speed, the low-pressure compressor outlet pressure, the high-pressure compressor outlet pressure, and the low-pressure turbine outlet temperature, and filter the residual signals.

[0067] Step 300: Calculate the root mean square error corresponding to the predicted low-pressure rotor speed, which is used as the dynamic threshold for fault diagnosis of the high-pressure rotor speed sensor, the low-pressure compressor outlet pressure sensor, the high-pressure compressor outlet pressure sensor, and the low-pressure turbine outlet temperature sensor.

[0068] Step 400: Set the sensor fault detection counter and calculate the value of the fault detection counter by comparing the filtered residual signal with the dynamic threshold.

[0069] Step 500: Compare the counter value with the counter setting value. If the counter value exceeds the setting value, report a fault.

[0070] In step 100, the sensor measurements of the engine include engine input parameters and airflow parameters, and the input parameter I at time t is... E (t)={T t0 (t), P t0 (t), W f (t), A8(t)}, where T t0 (t) represents the total intake temperature, P t0 (t) represents the total intake pressure, W f (t) represents the fuel flow rate, and A8(t) represents the nozzle area. The gas path parameters O at time t are... E (t)={N L,E (t), N H,E (t),P t25,E (t), P t31,E (t), T t6,E (t)}, where N L,E (t) represents the measured value of the low-pressure rotor speed, N H,E (t) represents the measured value of the high-pressure rotor speed, P t25,E (t) represents the measured pressure at the outlet of the low-pressure compressor, P t31,E (t) represents the measured pressure at the outlet of the high-pressure compressor, T. t6,E (t) represents the measured temperature at the outlet of the low-pressure turbine.

[0071] In step 100, the predicted values ​​of the air path parameters of the engine component-level model are based on the engine input parameter I. E (t) is used to calculate the predicted value of the gas path parameters O at time t. M (t)={N L,M (t), N H,M (t), P t25,M (t), P t31,M (t),T t6,M (t)}, where N L,M (t) represents the predicted low-pressure rotor speed calculated by the model, N H,M (t) represents the predicted high-pressure rotor speed calculated by the model, P t25,M (t) represents the predicted pressure at the low-pressure compressor outlet calculated by the model, P t31,M (t) represents the predicted pressure at the high-pressure compressor outlet calculated by the model, Tt6,M (t) represents the predicted temperature at the low-pressure turbine outlet calculated by the model.

[0072] In step 200, the prediction residual is calculated as follows: Sensor N... H P t25 P t31 and T t6 The numbers are sequentially recorded as 1 to 4, and the measurement value of the i-th sensor at time t is recorded as O. E,i (t), let O be the predicted value of the i-th sensor at time t. M,i (t), calculate the prediction residual of the i-th sensor at time t. .

[0073] In step 200, the filtering calculation formula for the residual signal is as follows:

[0074]

[0075] in, is the smoothed residual at time t, and θ is the smoothing factor between 0 and 1.

[0076] In step 300, the predicted low-pressure rotor speed N is calculated. L,M The root mean square error R(t) corresponding to (t) is calculated iteratively using the following formula:

[0077]

[0078] Among them, L R This calculates the window length, R(t-1) is the root mean square error at time t-1, and N... L,M (tL R ) for tL R Predicted low-pressure rotor speed at time N L,E (tL R ) for tL R The low-pressure rotor speed sensor reading at any given time.

[0079] In step 300, the root mean square error of the low-pressure rotor speed, used as the dynamic threshold for fault diagnosis of the i-th sensor, is designed as a function of the root mean square error R(t), which can be expressed as:

[0080]

[0081] Where, δ i (t) is the dynamic threshold for fault diagnosis of the i-th sensor, ε i,k It is the proportionality constant, ε i,b It is the bias coefficient.

[0082] In step 400, the fault detection counter's fault detection counting method for the i-th sensor can be expressed as follows:

[0083]

[0084] The counter value is a non-negative number, i.e., m. i (t)≥0.

[0085] In step 400, the method for calculating the value of the fault detection counter is as follows: when the residual The absolute value is less than the dynamic threshold δ i At time (t), the counter m i (t) decreases if m i (t) < 0, then m i (t) is set to 0. When the residual The absolute value exceeds the dynamic threshold δ i At time (t), the counter m i (t) increases,

[0086] In step 500, the fault detection counter reports faults in the following manner: if the smoothed residual... Exceeding the dynamic threshold δ for a continuous period of time i (t), that is, satisfying m i (t)>m f,i When the i-th sensor fails, a fault is reported. The counting boundary m... f,i This is a constant, and its value can be set according to specific needs.

[0087] In one embodiment, this disclosure provides a dynamic diagnostic method for sensor faults in an engine control system to improve fault tolerance, using sensor P t25 For example, a fault:

[0088] Step 100, record engine input parameter I E (t)={T t0 (t), P t0 (t), W f (t), A8(t)}, where T t0 (t) represents the total intake temperature, P t0 (t) represents the total intake pressure, W f (t) represents the fuel flow rate, and A8(t) represents the nozzle area. Record the engine sensor measurements O. E (t)={N L,E (t), N H,E (t), P t25,E (t), P t31,E (t), T t6,E (t)}, where N L,E(t) represents the measured value of the low-pressure rotor speed, N H,E (t) represents the measured value of the high-pressure rotor speed, P t25,E (t) represents the measured pressure at the outlet of the low-pressure compressor, P t31,E (t) represents the measured pressure at the outlet of the high-pressure compressor, T. t6,E (t) represents the measured temperature at the outlet of the low-pressure turbine.

[0089] Step 200, based on engine input parameter I E (t), calculate the output O of the component-level model. M (t)={N L,M (t), N H,M (t), P t25,M (t), P t31,M (t), T t6,M (t)}, where N L,M (t) represents the predicted low-pressure rotor speed calculated by the model, N H,M (t) represents the predicted high-pressure rotor speed calculated by the model, P t25,M (t) represents the predicted pressure at the low-pressure compressor outlet calculated by the model, P t31,M (t) represents the predicted pressure at the high-pressure compressor outlet calculated by the model, T t6,M (t) represents the predicted temperature at the low-pressure turbine outlet calculated by the model.

[0090] Step 300, sensor N H P t25 P t31 and T t6 If the numbers are sequentially recorded as 1 to 4, then sensor P t25 The measured value is recorded as O. E,2 (t), sensor P t25 The predicted value is denoted as O. M,2 (t), calculate sensor P t25 Predicted residuals .

[0091] Step 400, for sensor P t25 The residual signal r2(t) is filtered, and the filtering calculation formula is as follows:

[0092]

[0093] in, At time t, sensor P t25 The smoothed residuals are obtained, and θ is taken as 0.8.

[0094] Step 500: Calculate the predicted low-pressure rotor speed N. L,M The root mean square error R(t) corresponding to (t) is calculated iteratively using the following formula:

[0095]

[0096] Wherein, the calculation window length L R Set it to 20.

[0097] Step 600, for sensor P t25 The dynamic threshold δ2(t) for fault diagnosis is a function of the root mean square error R(t) of low-pressure speed prediction, and its calculation formula is as follows:

[0098]

[0099] Wherein, the proportionality coefficient ε i,k Set to 1.1, bias coefficient ε i,b Set to 0.005.

[0100] Step 700, set sensor P t25 Configure a fault detection counter, with the counting method as follows:

[0101]

[0102] Wherein, the counter value m2(t) is a non-negative number, that is, m2(t)≥0.

[0103] Step 800, calculate sensor P t25 The value of the fault detection counter. When the residual When the absolute value of the counter is less than the dynamic threshold δ2(t), the counter m2(t) decreases; if m2(t) < 0, then m2(t) is set to 0. When the residual... When the absolute value of the counter exceeds the dynamic threshold δ2(t), the counter m2(t) increases.

[0104] Step 900, if smoothing residuals If the value exceeds the dynamic threshold δ2(t) for a sustained period of time, i.e., if m2(t) > m f,2 When this happens, sensor P will report. t25 Fault. Wherein, the counting boundary m f,2 Set it to 20.

[0105] Bias fault manifests as a large, fixed deviation in sensor measurements starting from a certain moment; its simulation method is as follows:

[0106]

[0107] Drift faults manifest as a deviation in sensor measurements at a certain rate starting from a certain moment; the simulation method is as follows:

[0108]

[0109] Among them, y i (t) represents the signal value when the i-th sensor is fault-free, y i,bf (t) represents the signal value when the i-th sensor biases fault, y i,df (t) represents the signal value when the i-th sensor experiences a drift fault, Δy i t represents the magnitude of the sensor signal bias amplitude. f k represents the time when the sensor failure occurred. df This indicates the drift rate of the sensor signal.

[0110] The method of this disclosure is verified by injecting sensor faults into an engine model.

[0111] Figure 2(a) shows sensor P t25 The fault is a bias fault with a bias magnitude of Δy2 = 5000 Pa, and the fault occurs at 16.06 s.

[0112] Figure 2(b) shows sensor P t25 The fault is a drift fault, and the drift rate is k. df = 2500Pa / s, the fault occurred at 16.06s.

[0113] Figure 3(a) shows sensor P t25 The bias fault diagnosis result showed that the sensor fault was diagnosed in 16.54s, and the diagnosis time was 0.48s.

[0114] Figure 3(b) shows sensor P t25 The drift fault diagnosis result showed that the sensor fault was diagnosed at 18.24s, and the diagnosis time was 2.18s.

[0115] As can be seen from Figures 3(a) and 3(b), the dynamic diagnostic method for sensor faults proposed in this invention, which is geared towards fault tolerance in engine control systems, demonstrates excellent diagnostic performance for both sensor bias and drift faults. This method can quickly and accurately diagnose sensor faults without triggering false alarms.

[0116] In one embodiment, the method includes,

[0117] The engine component-level model calculates the predicted values ​​of the sensors based on the input parameters; calculates the predicted residuals of each sensor; filters the sensor residual signals; calculates the root mean square error between the predicted low-pressure rotor speed and the measured low-pressure speed; uses the low-pressure rotor speed error as a benchmark to calculate the dynamic thresholds corresponding to other sensors besides the low-pressure speed; designs a fault indicator: if the residual between the sensor measured value and the model predicted value is greater than the dynamic threshold, the indicator value increases; otherwise, the indicator value decreases; when the fault indicator value exceeds a set counting boundary, a sensor fault is reported. This invention can solve the problems of high false alarm rate and high false alarm rate in sensor fault diagnosis methods based on fixed thresholds, and can timely diagnose sensor bias and drift faults. It has the characteristics of simple algorithm and easy implementation.

[0118] A dynamic fault diagnosis system for sensors aimed at improving the fault tolerance of engine control systems includes:

[0119] The measurement unit detects and records the sensor measurements of the engine's air path parameters, and calculates the predicted values ​​of the air path parameters of the engine component-level model based on the engine component-level model corresponding to the engine.

[0120] The residual unit calculates the residual signal between the measured values ​​of each gas path parameter sensor and the predicted values ​​of the corresponding gas path parameter, and filters the residual signal.

[0121] The threshold calculation unit calculates the root mean square error of the predicted low-pressure rotor speed, which is denoted as the prediction error of the low-pressure rotor speed. Based on the prediction error, it calculates the dynamic threshold corresponding to the fault diagnosis of other gas path sensors.

[0122] The fault detection counter calculates its value by comparing the filtered residual signal with the dynamic threshold.

[0123] The comparison unit compares the counter value with the counter set value. If the counter value exceeds the set value, a fault is reported.

[0124] A computer storage medium including computer instructions that, when run on a computer, cause the computer to perform the method.

[0125] An electronic device, the electronic device comprising:

[0126] Memory, processor, and computer programs stored in memory and executable on the processor, wherein,

[0127] The processor implements the method when executing the program.

[0128] Although embodiments of the present invention have been described above in conjunction with the accompanying drawings, the present invention is not limited to the specific embodiments and application fields described above. The specific embodiments described above are merely illustrative and instructive, and not restrictive. Those skilled in the art can make many other forms based on the guidance of this specification and without departing from the scope of protection of the claims of the present invention, and all of these are within the scope of protection of the present invention.

Claims

1. A dynamic diagnostic method for sensor faults in an engine control system, characterized in that, Includes the following steps: Step 100: Detect and record the sensor measurements of the engine, and calculate the predicted values ​​of the air path parameters of the engine component-level model based on the engine component-level model corresponding to the engine. Step 200: Calculate the residual signal between the measured value of the gas path parameter sensor and the predicted value of the corresponding gas path parameter, and filter the residual signal; Step 300: Calculate the root mean square error of the predicted low-pressure rotor speed, denoted as the prediction error of the low-pressure rotor speed, and use the prediction error as a benchmark to calculate the dynamic threshold corresponding to the fault diagnosis of other gas path sensors. Step 400: Set the sensor fault detection counter and calculate the value of the fault detection counter by comparing the filtered residual signal with the dynamic threshold. Step 500: Compare the counter value with the counter set value. If the counter value exceeds the set value, report a fault. in, The sensor measurements include engine input parameters and airflow parameters, where I is the engine input parameter at time t. E (t)={T t0 (t), P t0 (t), W f (t), A8(t)}, where T t0 (t) represents the total intake temperature, P t0 (t) represents the total intake pressure, W f (t) represents the fuel flow rate, A8(t) represents the nozzle area, and the gas path parameters O at time t are... E (t)={N L,E (t), N H,E (t), P t25,E (t), P t31,E (t),T t6,E (t)}, where N L,E (t) represents the measured value of the low-pressure rotor speed, N H,E (t) represents the measured value of the high-pressure rotor speed, P t25,E (t) represents the measured pressure at the outlet of the low-pressure compressor, P t31,E (t) represents the measured pressure at the outlet of the high-pressure compressor, T. t6,E (t) represents the measured temperature at the outlet of the low-pressure turbine; Based on engine input parameter I E (t) Calculate the predicted gas path parameters O at time t. M (t)={N L,M (t), N H,M (t),P t25,M (t), P t31,M (t), T t6,M (t)}, where N L,M (t) represents the predicted low-pressure rotor speed calculated from the engine component-level model, N. H,M (t) represents the predicted high-pressure rotor speed calculated from the engine component-level model, P t25,M (t) represents the predicted pressure at the low-pressure compressor outlet calculated from the engine component-level model, P. t31,M (t) represents the predicted pressure at the high-pressure compressor outlet calculated from the engine component-level model, T. t6,M (t) represents the predicted temperature at the low-pressure turbine outlet calculated by the engine component-level model; In step S200, the prediction residual is calculated as follows: Sensor N... H P t25 P t31 and T t6 The numbers are sequentially recorded as 1 to 4, and the measurement value of the i-th sensor at time t is recorded as O. E,i (t), let O be the predicted value of the i-th sensor at time t. M,i (t), calculate the prediction residual of the i-th sensor at time t. .

2. The dynamic diagnostic method for sensor faults in engine control systems as described in claim 1, characterized in that, The engine includes a low-pressure rotor, a high-pressure rotor, a low-pressure compressor, a high-pressure compressor, a high-pressure turbine, and a low-pressure turbine.

3. The dynamic diagnostic method for sensor faults in engine control systems as described in claim 1, characterized in that, In step 200, the filtering calculation formula for the residual signal is as follows: ， in, is the smoothed residual at time t, and θ is the smoothing factor between 0 and 1.

4. The dynamic diagnostic method for sensor faults in engine control systems as described in claim 3, characterized in that, In step 300, the predicted low-pressure rotor speed N at time t is calculated. L,M The root mean square error R(t) of (t) is used to calculate the predicted low-pressure rotor speed N. L,M The dynamic threshold for fault diagnosis of the high-pressure rotor speed sensor, the low-pressure compressor outlet pressure sensor, the high-pressure compressor outlet pressure sensor, and the low-pressure turbine outlet temperature sensor other than (t), and the root mean square error R(t) are calculated iteratively using the following formula: , where L R This calculates the window length, R(t-1) is the root mean square error at time t-1, and N... L,M (tL R ) for tL R Predicted low-pressure rotor speed at time N L,E (tL R ) for tL R The low-pressure rotor speed sensor reading at any given time.

5. A dynamic fault diagnosis system for sensors aimed at improving the fault tolerance of engine control systems, characterized in that, include: The measurement unit is used to: detect and record the sensor measurement values ​​of the engine's air path parameters, and calculate the predicted values ​​of the air path parameters of the engine component-level model based on the engine component-level model corresponding to the engine. The residual unit is used to: calculate the residual signal between the measured value of each gas path parameter sensor and the predicted value of the corresponding gas path parameter, and filter the residual signal; The threshold calculation unit is used to: calculate the root mean square error of the predicted low-pressure rotor speed, denoted as the prediction error of the low-pressure rotor speed, and use the prediction error as a benchmark to calculate the dynamic threshold corresponding to the fault diagnosis of other gas path sensors. The fault detection counter is used to calculate the value of the fault detection counter by comparing the filtered residual signal with the dynamic threshold. The comparison unit is used to compare the counter value with the counter set value; if the counter value exceeds the set value, a fault is reported. in, The sensor measurements include engine input parameters and airflow parameters, where I is the engine input parameter at time t. E (t)={T t0 (t), P t0 (t), W f (t), A8(t)}, where T t0 (t) represents the total intake temperature, P t0 (t) represents the total intake pressure, W f (t) represents the fuel flow rate, A8(t) represents the nozzle area, and the gas path parameters O at time t are... E (t)={N L,E (t), N H,E (t), P t25,E (t), P t31,E (t),T t6,E (t)}, where N L,E (t) represents the measured value of the low-pressure rotor speed, N H,E (t) represents the measured value of the high-pressure rotor speed, P t25,E (t) represents the measured pressure at the outlet of the low-pressure compressor, P t31,E (t) represents the measured pressure at the outlet of the high-pressure compressor, T. t6,E (t) represents the measured temperature at the outlet of the low-pressure turbine; Based on engine input parameter I E (t) Calculate the predicted gas path parameters O at time t. M (t)={N L,M (t), N H,M (t),P t25,M (t), P t31,M (t), T t6,M (t)}, where N L,M (t) represents the predicted low-pressure rotor speed calculated from the engine component-level model, N. H,M (t) represents the predicted high-pressure rotor speed calculated from the engine component-level model, P t25,M (t) represents the predicted pressure at the low-pressure compressor outlet calculated from the engine component-level model, P. t31,M (t) represents the predicted pressure at the high-pressure compressor outlet calculated from the engine component-level model, T. t6,M (t) represents the predicted temperature at the low-pressure turbine outlet calculated by the engine component-level model; In the residual unit, the prediction residual is calculated as follows: The sensor N... H P t25 P t31 and T t6 The numbers are sequentially recorded as 1 to 4, and the measurement value of the i-th sensor at time t is recorded as O. E,i (t), let O be the predicted value of the i-th sensor at time t. M,i (t), calculate the prediction residual of the i-th sensor at time t. .

6. A computer storage medium, characterized in that, The storage medium includes computer instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1-4.

7. An electronic device, characterized in that, The electronic device includes: Memory, processor, and computer programs stored in memory and executable on the processor, wherein, When the processor executes the program, it implements the method as described in any one of claims 1-4.

Images