CN115576294A - Fault-tolerant soft-hard hybrid control method for fault diagnosis of aircraft engine sensor - Google Patents
Fault-tolerant soft-hard hybrid control method for fault diagnosis of aircraft engine sensor Download PDFInfo
- Publication number
- CN115576294A CN115576294A CN202211093004.XA CN202211093004A CN115576294A CN 115576294 A CN115576294 A CN 115576294A CN 202211093004 A CN202211093004 A CN 202211093004A CN 115576294 A CN115576294 A CN 115576294A
- Authority
- CN
- China
- Prior art keywords
- sensor
- fault
- soft
- kalman filter
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0243—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D18/00—Testing or calibrating apparatus or arrangements provided for in groups G01D1/00 - G01D15/00
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/20—Pc systems
- G05B2219/24—Pc safety
- G05B2219/24065—Real time diagnostics
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Combined Controls Of Internal Combustion Engines (AREA)
Abstract
The invention belongs to the field of aircraft engine fault diagnosis, and provides a fault-tolerant soft-hard hybrid control method for aircraft engine sensor fault diagnosis. The method realizes the detection of soft/hard faults of the aeroengine sensor and the estimation of the real state of the system through a Kalman filter bank in the soft/hard fault diagnosis system. The designed soft fault diagnosis system further detects whether faults exist by carrying out residual error processing on a sensor measured value and a group of Kalman filter estimated values and carrying out weighted square summation on the residual error processing and the Kalman filter estimated values so as to compare known threshold values; the hard fault diagnostic system is designed to further detect the presence of a fault by comparing the absolute value of the residuals of the sensor measurements and the kalman filter estimates to known thresholds. The verification proves that the method can effectively remove the negative influence of the faults of the aero-engine sensor on the overall performance of the system, so that the system is more stable, and various performance indexes can meet the requirements.
Description
Technical Field
The invention provides a fault-tolerant soft-hard hybrid control method for an aircraft engine sensor, and belongs to the field of aircraft engine fault diagnosis.
Background
Currently, high-performance aircraft engines are developing towards high thrust-weight ratio, high speed, high reliability and the like, the functions implemented by control systems of aircraft engines are more and more complex, and the requirements on the reliability of the control systems are also more and more high. The system fault can directly affect the reliability of the whole system, and the fault diagnosis and fault-tolerant control system is provided aiming at improving the reliability, maintainability and effectiveness of the system. The aeroengine sensor mainly measures and reflects various parameters of the working state of the aeroengine sensor, such as the rotating speed of a rotor, the temperature and the pressure of a part of working section of a gas path, and the like. The accuracy of the sensor measuring parameters directly influences the accuracy of the work and fault diagnosis results of the control system. With the development of aircraft engine technology, higher requirements are also placed on the stability of sensors in the control system of aircraft engines. However, the sensor works in a high-temperature and strong-vibration environment, belongs to an element with lower reliability in a system, and is easy to break down. The classification of sensor faults can be classified into hard faults (generally, faults caused by structural damage, large amplitude and sudden change) and soft faults (generally, variation of characteristics, small amplitude and slow change) according to the fault degree. Hard failures are typically caused by damage to the sensor element, short circuiting, open circuits or strong impulse disturbances of the electrical system. Soft failures are typically caused by component aging, zero drift, and the like. There are three main types of typical aircraft engine sensor failures: intrinsic bias faults, drift faults, impulse disturbance faults.
Since the failure of the sensor affects the measurement output of the control system, the operation of the controller and the implementation of the control algorithm are affected by erroneous feedback, and finally, the stability and other performances of the whole system are negatively affected. Therefore, when the sensor fails, the control system needs to timely isolate the failed sensor and reconstruct the sensor signal when the engine normally works, so that the control system is maintained in a fault-free working state, and precious time is gained for eliminating the engine fault. If the real state of the system can be correctly estimated when a fault occurs, a correct feedback control signal can be obtained, so that the system is prevented from being influenced by the fault. The fault-tolerant control of the sensor opens up a path for improving the reliability, maintainability and effectiveness of the system, and becomes a research focus in the aerospace field. The fundamental characteristic of fault-tolerant control is that when a control system fails, the system can still maintain the self operation in a safe state and meet certain performance index requirements as much as possible. It enables a dynamic system to adapt to significant changes in its environment, avoiding the impact on the stability and other performance of the overall system due to the failure of one or more of the more critical components of the system. Therefore, research into fault-tolerant control methods for aircraft engine sensors is necessary.
The design method of the fault-tolerant control system comprises hardware redundancy and software redundancy. Hardware redundancy is the increase in fault tolerance of a system by providing redundancy to important or failure prone components. Software redundancy is to design a controller to improve the redundancy of the whole system, thereby improving the fault tolerance of the system. Since there are many sensors in an aircraft engine control system, the method of using hardware redundancy to improve the reliability of components and high reliability design is costly, increases the complexity of the system, and affects the system performance. And the cost can be greatly reduced by utilizing the fault diagnosis and fault-tolerant control technology of redundant signals generated by the analytic mathematical model. In the sensor fault, a Kalman filter bank can be used to obtain fault indication signals of a monitored system, and the purpose of fault diagnosis is realized by analyzing the variation trend of each fault indication signal.
The invention provides a fault-tolerant soft-hard hybrid control method for an aircraft engine sensor, which realizes the detection of soft/hard faults of the aircraft engine sensor and the estimation of the real state of the system through a Kalman filter bank in a soft/hard fault diagnosis system. When the sensor has no fault, the measured value of the sensor is directly fed back to the fault-tolerant hybrid controller; when the sensor has a fault, the estimation output of the Kalman filter is fed back to realize the reconstruction of the signal, so that the signal fed back to the fault-tolerant hybrid controller is not fluctuated too much, and the engine is still in a stable working state, thereby realizing the fault-tolerant control of the soft/hard fault of the sensor. The invention is funded by a Chinese postdoctor scientific fund project (2022 TQ 0179), national Chinese natural science fund 61890920, 61890921 and a national key research and development plan project 2018YFB 1700102.
Disclosure of Invention
In order to remove the negative influence of the faults of the aero-engine sensor on the overall performance of the system, so that the system is more stable, and various performance indexes can meet requirements better, the invention provides a fault-tolerant soft-hard hybrid control method for the aero-engine sensor.
The technical scheme of the invention is as follows:
a fault-tolerant soft and hard hybrid control method for an aircraft engine sensor is based on fault diagnosis based on a Kalman filter, the detection of soft/hard faults of the aircraft engine sensor is realized, the real state of a system is estimated when the sensor fails, and a feedback signal is reconstructed, so that the fault-tolerant hybrid control of the soft/hard faults of the sensor is realized, and the method comprises the following steps:
s1, establishing a Kalman filter according to an aircraft engine model.
S1.1, firstly, establishing a small deviation linearization model by using a fitting method based on an aeroengine component-level pneumatic-thermal nonlinear model, and calculating each system matrix of a state space model of the model aiming at a steady-state working point.
The input quantity u of the small deviation linearized model is the fuel flow W fm (ii) a The state variable x is the high-pressure rotor speed XNLPC and the low-pressure rotor speedThe subrotation speed XNHPC; the output y is the rotating speed of the high-pressure rotor and the low-pressure rotor, the static pressure P of the outlet of the high-pressure compressor and the total temperature T of the inlet of the low-pressure turbine; and using the corresponding parameter value at the steady-state operating point as a reference value, i.e. (x) S ，u S ，y S ) Wherein x is S Representing the state variable u at the steady-state operating point S Represents the input quantity at the steady-state operating point, y S Representing the output at the steady state operating point.
Let Δ x and Δ u be the deviations of the state variable and the input quantity, respectively, and Δ y be the deviation of the output quantity, at the steady-state operating point (x) S ，u S ，y S ) The state space equation for the resulting small-deviation linearized model is as follows:
where, denotes the first derivative, x = x S +Δx，u＝u S +Δu，y＝y S + Δ y, and then the system matrices a, B, C, D are calculated by program fitting.
S1.2, establishing a Kalman filter: at steady state operating point (x) S ，u S ，y S ) On the basis of the established linear model, for the linear model of the engine with the input not containing the deviation amount of the health parameter, the following formula is formed:
in the formula, w and v are respectively system noise and measurement noise, and w and v are assumed to be white gaussian noise with a mean value of zero, and covariance matrices thereof are respectively Q and R. Next, a Kalman filter may be built as follows:
in the formula (I), the compound is shown in the specification,
representing an estimate of a variable, K represents a kalman gain matrix; p is a solution of the ricalifting Equation (Riccati Equation) shown in Equation (4):
AP+PA T -PC T R -1 CP+Q＝0. (4)
and S2, inputting the output signal of the aircraft engine into a hard fault diagnosis system or a soft fault diagnosis system of the sensor according to the amplitude of the fault signal and the soft/hard switching rule.
And S3, designing a hard fault diagnosis system of the aeroengine sensor.
And (2) building a sensor hard fault diagnosis system by adopting the Kalman filter built in the step (S1), inputting the output value of the measurement parameter of the aeroengine sensor into the Kalman filter, and then calculating the difference between the estimation value of the measurement parameter after passing through the Kalman filter and the output value of the measurement parameter of the sensor, wherein the difference between the estimation value of the measurement parameter and the output value of the measurement parameter of the sensor is the residual error of the corresponding Kalman filter.
When the absolute value of the residual error of the Kalman filter corresponding to the measured value of one or more sensors exceeds the threshold value (determined by the characteristics of the components corresponding to the measured parameters of the sensors), the fault of the sensor can be judged, the sensor is isolated, the sensor with the fault is cut off, the estimated value of the Kalman filter replaces the measured value of the sensor with the measured value of the fault and is fed back to the controller, and the reconstruction of the signal is realized.
When the absolute value of the Kalman filter residual value corresponding to one or more sensor measurement values is smaller than a threshold value, the sensor can be judged to be faultless, and the output of the sensor is fed back to the controller.
And S4, designing a soft fault diagnosis system of the aero-engine sensor.
When a sensor soft fault occurs, the amplitude of the output value of the sensor measurement parameter deviating from the normal value is small, and the sensor is easily submerged by interference noise. Therefore, it cannot be determined by simply comparing whether the residual exceeds the threshold, and a set of kalman filters is required to perform a weighted sum-of-squares process on the residual sequence. The method comprises the following specific steps:
s4.1 assume that the fault diagnosis model shares n outputsParameters which respectively correspond to n sensors and adopt n +1 Kalman filters; wherein: input y of Kalman filter 0 0 Using all n sensor measurements, representing a normal state; input y of Kalman filter i i All the remaining n-1 sensor measurements except the corresponding own sensor are used. Input value y of Kalman filter i And estimated output value after filtering
The difference is the residual error.
S4.2 obtaining weighted average sum WSSR of residual errors of Kalman filter i i (i =1,2, 3.. Once.) when one of the sensors fails, the characteristics of the residual between the measured engine parameters and the estimated filter output change accordingly, and the WSSR changes i A large variation occurs. Then, the statistic WSSR of each filter is calculated i Value WSSR from Normal mode 0 Subtracting to obtain statistic SR i ：
SR i ＝WSSR 0 -WSSR i .
S4.3 when detecting the soft fault of the sensor, obtaining each statistic SR i Maximum value MAX (SR) of (i =1,2,3.. Times.) i ) Comparing the measured value with a set threshold value, if the measured value does not exceed the threshold value, determining that the system has no sensor fault, and directly feeding back the measured value of the sensor to the controller; if the threshold is exceeded, the SR is considered i The sensor corresponding to the maximum value has a fault; then sequentially judging other SR i Whether the maximum of the values exceeds a threshold until all sensors are detected. And if the fault occurs, the estimated value of the Kalman filter is used for replacing the measured value of the fault sensor to be fed back to the controller, and a feedback signal is reconstructed.
And S5, constructing a fault-tolerant soft-hard hybrid control system of the aero-engine sensor.
And inputting the output signal of the aircraft engine into a hard fault diagnosis system or a soft fault diagnosis system of the sensor according to the amplitude of the fault signal and the soft/hard switching rule. When the sensor has no fault, the signal is directly fed back to the input end of the control system; when the sensor has a fault, the fault detection and diagnosis system diagnoses the fault, gives an alarm in time, and reconstructs a feedback signal at the same time so that the engine is still in a normal stable working state; and the serious consequences of the whole system can not occur due to the distortion of the feedback signal caused by the fault of the sensor when the engine works normally. At the moment, the fault-tolerant hybrid control system established for soft/hard faults of the aeroengine sensor has fault-tolerant capability, and the negative influence of the sensor faults on the overall performance of the system can be relieved to a great extent.
The invention has the beneficial effects that:
aiming at the problem of fault diagnosis of an aircraft engine sensor, the invention designs a fault-tolerant hybrid control system based on a Kalman filter. The designed soft fault diagnosis system further detects whether faults exist by carrying out residual error processing on a sensor measured value and a group of Kalman filter estimated values and carrying out weighted square summation on the residual error processing and the Kalman filter estimated values so as to compare known threshold values; the hard fault diagnostic system is designed to further detect the presence of a fault by comparing the absolute value of the residuals of the sensor measurements and an estimate of the kalman filter with known thresholds. The fault-tolerant soft-hard hybrid control system of the aero-engine sensor realizes effective detection of soft/hard faults of the aero-engine sensor, estimates the real state of the system when the sensor faults occur, reconstructs sensor signals when the engine works normally, thereby realizing fault-tolerant hybrid control of the soft/hard faults of the sensor, effectively removing negative effects of the faults of the aero-engine sensor on the overall performance of the system, and ensuring stable operation of the aero-engine control system under the condition that the sensor faults exist. Compared with a design method for improving the reliability of the sensor by adopting hardware redundancy, the method has the advantages that the cost is greatly reduced; meanwhile, the method not only can realize the fault-tolerant control of the sensors, but also can position which sensor or sensors have faults, and can estimate the size and severity of the faults.
Drawings
FIG. 1 is a schematic structural diagram of a fault-tolerant soft-hard hybrid control system of an aircraft engine sensor;
FIG. 2 is W of fault-tolerant control of an aircraft engine sensor hard/soft failure system fm A signal simulation diagram; FIG. 2 (a) is W of the fault-tolerant-free control of the hard fault system of the aeroengine sensor fm Signal simulation diagram, FIG. 2 (b) is W of soft fault system without fault-tolerant control of aeroengine sensor fm And (4) signal simulation diagram.
FIG. 3 is a W with fault tolerant control for an aircraft engine sensor hard/soft failure system fm A signal simulation diagram; FIG. 3 (a) is a W with fault tolerant control for an aircraft engine sensor hard failure system fm Signal simulation diagram, FIG. 3 (b) is W with fault tolerant control for soft fault system of aircraft engine sensor fm And (4) signal simulation diagrams.
Detailed Description
The invention will be further explained with reference to the drawings.
FIG. 1 is a schematic structural diagram of a fault-tolerant soft-hard hybrid control system of an aircraft engine sensor. Wherein a left-side dotted line frame in fig. 1 is an aircraft engine sensor hard fault diagnosis system, and a right-side dotted line frame in fig. 1 is an aircraft engine sensor soft fault diagnosis system.
As shown in fig. 1, the fault-tolerant soft-hard hybrid control system of the aircraft engine sensor inputs an output signal of the aircraft engine into a hard fault diagnosis system or a soft fault diagnosis system of the sensor according to the amplitude of a fault signal and a soft/hard switching rule; when the sensor has no fault, the signal is directly fed back to the input end of the control system; when the sensor has a fault, the fault detection and diagnosis system diagnoses the fault, gives an alarm in time, and reconstructs a feedback signal at the same time so that the engine is still in a normal stable working state; and the serious consequences of the whole system can not occur due to the distortion of the feedback signal caused by the fault of the sensor when the engine works normally.
As shown by a dashed line box on the left side of FIG. 1, in the designed aircraft engine sensor hard fault diagnosis system, the difference between the output value y of the sensor measurement parameter and the measurement parameter estimation value after passing through the Kalman filter is the differenceIs residual r, if used
Representing the estimated values of the kalman filter, there are:
when the absolute value of the Kalman filter residual error r corresponding to one or more sensor measurement values exceeds the threshold value of the Kalman filter residual error r, the sensor can be judged to have a fault, the sensor with the fault is isolated, the sensor with the fault is cut off, the estimation value of the Kalman filter replaces the measurement value of the fault sensor and is fed back to the controller, and the signal reconstruction is realized. When the absolute value of the residual value r of the Kalman filter is smaller than the threshold value, the sensor can be judged to be faultless, and the output of the sensor is fed back to the controller.
As shown in the right dashed box of fig. 1, the aircraft engine sensor soft fault diagnosis system is designed such that the filter 0 uses all sensor measurements to represent a normal state. Input y of other filters i Are the outputs of all other sensors except the corresponding own sensor output. The reason for this is that, assuming that a certain sensor has a fault, only the kalman filter corresponding to the sensor does not use the measurement value of the faulty sensor, and therefore, only the estimation result obtained by the kalman filter is correct, while the other kalman filters all use the measurement value of the faulty sensor, and therefore, the estimation results of the other kalman filters deviate from the accurate value to different degrees, so that the existence of the fault can be further determined. Input value y of Kalman filter i And estimated output value after filtering
The difference is the residual r, and for the kalman filter i, there are:
thereby can obtainTo residual sequence r i When the sensor is not in fault and the filtering process tends to be stable, the residual vector r i Obeying a multidimensional normal distribution
Sensor fault indication variables can thus be constructed:
WSSR i ＝(r i ) T .(∑ i ) -1 .r i ，
∑ i ＝diag[σ i ] 2 ，
WSSR i (Weight Sum of Squared Residuals) is called the residual weighted Sum of squares, since
So WSSR i Compliance chi 2 And (4) distribution. When one sensor fails, the characteristics of residual errors between the engine measurement parameters and the estimated output of the filter are correspondingly changed, and the WSSR i A large variation occurs. Then, the statistic WSSR of each filter is calculated i (i =1,2,3..) and the normal modal value WSSR of filter 0 0 Subtracting to obtain statistic SR i ：
SR i ＝WSSR 0 -WSSR i .
When detecting the soft fault of the sensor, obtaining each statistic SR i Maximum value MAX (SR) of (i =1,2,3. -) i ) Comparing the measured value with a set threshold value, if the measured value does not exceed the threshold value, determining that the system has no sensor fault, and directly feeding back the measured value of the sensor to the controller; if the threshold is exceeded, the SR is considered i The sensor corresponding to the maximum value has a fault; then sequentially judging other SR i Whether the maximum of the values exceeds a threshold until all sensors are detected. And if the fault occurs, the estimated value of the Kalman filter is used for replacing the measured value of the fault sensor to be fed back to the controller, and a feedback signal is reconstructed.
The method comprises the following steps that the hard fault and the soft fault of the low-voltage rotor rotating speed XNLPC sensor are added into an implementation case 1 and an implementation case 2 respectivelyAnd the main fuel flow W before and after the sensor hard fault/soft fault system and the fault-tolerant hybrid control measure are fed back to the controller fm And comparing the signals, and verifying the fault-tolerant effect of the fault-tolerant soft-hard hybrid control system of the aero-engine sensor through a simulation example.
FIG. 2 is a W of fault-tolerant control of a hard/soft failure system of an aircraft engine sensor fm A signal simulation diagram; FIG. 2 (a) is W of the fault-tolerant-free control of the hard fault system of the aeroengine sensor fm Signal simulation diagram, FIG. 2 (b) is W of soft fault system without fault-tolerant control of aeroengine sensor fm And (4) signal simulation diagram. FIG. 3 is a W with fault tolerant control for an aircraft engine sensor hard/soft failure system fm A signal simulation diagram; FIG. 3 (a) is W with fault tolerant control for an aircraft engine sensor hard failure system fm Signal simulation diagram, FIG. 3 (b) is W with fault tolerant control for soft fault system of aircraft engine sensor fm And (4) signal simulation diagrams.
In embodiment 1, a hard fault of a low-voltage rotor rotating speed XNLPC sensor is added, and the simulation is as follows: at t =20s, the low pressure rotor speed XNLPC sensor output suddenly increases by 50 revolutions for a duration of 2 seconds. As can be seen from fig. 2 (a) and fig. 3 (a), when a hard failure of the sensor occurs and no fault-tolerant adjustment is applied, the system is adjusted by the controller, and although the system can finally achieve stability, the dynamic process fluctuates greatly. After fault-tolerant control measures are added, the hard fault diagnosis system of the sensor diagnoses faults in time, simultaneously feeds back the estimation output of the Kalman filter and reconstructs feedback signals, so that the engine is still in a normal stable working state, the fluctuation of the dynamic process of the system and the dynamic response time are obviously and greatly reduced, and a good fault-tolerant control effect is achieved.
In the embodiment 2, the soft fault of the low-voltage rotor rotating speed XNLPC sensor is added, the fault drifting along with the time is simulated and selected, so that the soft fault of the sensor is simulated, and the fault signal occurs between 20 seconds and 30 seconds. As can be seen from fig. 2 (b) and fig. 3 (b), when no fault-tolerant adjustment measure is applied after a soft fault of the sensor occurs, the output signal of the fault system gradually deviates from the steady-state value with the passage of time until t =30s after fault eliminationGradually returning to a steady state value; after fault-tolerant control measures are added, drift faults are accumulated to a certain degree, and fault judgment marks MAX (SR) in a sensor soft fault diagnosis system i ) The signal reconstruction switch is triggered to switch to the estimation output of the Kalman filter, so that the signal fed back to the controller is not fluctuated greatly, and the aim of maintaining the normal and stable whole system is fulfilled.
In conclusion, the fault-tolerant soft-hard hybrid control method for the aircraft engine sensor is feasible, the method is based on fault diagnosis based on a Kalman filter, detection of soft/hard faults of the aircraft engine sensor is achieved, and the real state of the system is estimated when the sensor fails, so that fault-tolerant hybrid control of the soft/hard faults of the sensor is achieved, negative effects of the faults of the aircraft engine sensor on the overall performance of the system can be effectively removed, the system is more stable, and various performance indexes can meet requirements.
The above-mentioned embodiments only express the embodiments of the present invention, but not should be understood as the limitation of the scope of the invention patent, it should be noted that, for those skilled in the art, many variations and modifications can be made without departing from the concept of the present invention, and these all fall into the protection scope of the present invention.
Claims (2)
1. A fault-tolerant soft-hard hybrid control method for aircraft engine sensor fault diagnosis is characterized in that the control method is based on fault diagnosis based on a Kalman filter, detection of soft/hard faults of an aircraft engine sensor is achieved, the real state of a system is estimated when the sensor fault occurs, a feedback signal is reconstructed, and therefore fault-tolerant hybrid control over the soft/hard faults of the sensor is achieved, and the method comprises the following steps:
s1, establishing a Kalman filter according to an aircraft engine model;
s2, inputting an output signal of the aero-engine into a hard fault diagnosis system or a soft fault diagnosis system of the sensor according to the amplitude of the fault signal and a soft/hard switching rule;
s3, designing a hard fault diagnosis system of the aeroengine sensor;
adopting the Kalman filter established in the step S1 to build a sensor hard fault diagnosis system, inputting the output value of the measurement parameter of the aeroengine sensor into the Kalman filter, and then calculating the difference between the estimation value of the measurement parameter after passing through the Kalman filter and the output value of the measurement parameter of the sensor, wherein the difference is the residual error of the corresponding Kalman filter;
when the absolute value of the residual error of the Kalman filter corresponding to the measured value of one or more sensors exceeds the threshold value of the Kalman filter, judging that the sensors have faults, isolating the sensors, cutting off the sensors with the faults, and feeding back the measured values of the sensors with the estimated values of the Kalman filter to the controller to realize signal reconstruction;
when the absolute value of the Kalman filter residual value corresponding to one or more sensor measurement values is smaller than a threshold value, judging that the sensor has no fault, and feeding back the output of the sensor to the controller;
s4, designing a soft fault diagnosis system of the aero-engine sensor;
when a sensor soft fault occurs, a group of Kalman filters are adopted, and the weighted sum of squares processing is carried out on a residual sequence:
s4.1, assuming that n output parameters are shared in the fault diagnosis model and respectively correspond to n sensors, adopting n +1 Kalman filters; wherein: input y of Kalman filter 0 0 Using all n sensor measurements, representing a normal state; input y of Kalman filter i i Using all the measured values of the other n-1 sensors except the corresponding sensors; input value y of Kalman filter i And estimated output value after filtering
The difference is the residual error;
s4.2 obtaining weighted average sum WSSR of residual errors of Kalman filter i i (i =1,2,3, \ 8230;) when one of the sensors fails, it is sent outThe characteristics of the residual between the measured parameters of the motor and the estimated output of the filter are changed correspondingly, WSSR i A change occurs; then, the statistic WSSR of each filter is calculated i Value WSSR from Normal mode 0 Subtracting to obtain statistic SR i :
SR i ＝WSSR 0 -WSSR i .
S4.3 when detecting the soft fault of the sensor, obtaining each statistic SR i Maximum value MAX (SR) of (i =1,2,3, \ 8230;) i ) Comparing the measured value with a set threshold value, if the measured value does not exceed the threshold value, determining that the system has no sensor fault, and directly feeding back the measured value of the sensor to the controller; if the threshold is exceeded, the SR is considered i The sensor corresponding to the maximum value has a fault; then sequentially judging other SR i Whether a maximum of the values exceeds a threshold until all sensors are detected; if the fault occurs, replacing the measured value of the fault sensor with the estimated value of the Kalman filter and feeding back the value to the controller to reconstruct a feedback signal;
s5, constructing a fault-tolerant soft-hard hybrid control system of the aero-engine sensor;
inputting the output signal of the aeroengine into a hard fault diagnosis system or a soft fault diagnosis system of the sensor according to the amplitude of the fault signal and a soft/hard switching rule; when the sensor has no fault, the signal is directly fed back to the input end of the control system; when the sensor has faults, the fault detection and diagnosis system diagnoses the faults, gives an alarm in time, and reconstructs a feedback signal at the same time, so that the engine is in a normal stable working state.
2. The fault-tolerant soft-hard hybrid control method for the fault diagnosis of the sensor of the aircraft engine according to claim 1, wherein the step S1 of establishing the Kalman filter comprises the following specific steps:
s1.1, firstly, establishing a small deviation linearization model by using a fitting method based on an aeroengine component-level pneumatic-thermal nonlinear model, and calculating each system matrix of a state space model of the model aiming at a steady-state working point;
the input quantity u of the small deviation linearization model is the fuel flow W fm (ii) a The state variable x is the rotating speed XNLPC of the high-pressure rotor and the rotating speed XNHPC of the low-pressure rotor; the output y is the rotating speed of the high-pressure rotor and the low-pressure rotor, the static pressure P of the outlet of the high-pressure compressor and the total temperature T of the inlet of the low-pressure turbine; at the corresponding parameter value (x) at the steady state operating point S ，u S ，y S ) Is a reference value, wherein x S Representing the state variable, u, at the steady-state operating point S Indicating the input at steady state operating point, y S Representing an output quantity at a steady state operating point;
let Δ x and Δ u be the deviations of the state variable and the input quantity, respectively, and Δ y be the deviation of the output quantity, at the steady-state operating point (x) S ，u S ，y S ) The state space equation for the resulting small-deviation linearized model is as follows:
wherein x = x S +Δx,u＝u S +Δu,y＝y S + delta y, and then calculating a system matrix A, B, C and D by program fitting;
s1.2, establishing a Kalman filter: at steady state operating point (x) S ，u S ，y S ) On the basis of the established linear model, for the linear model of the engine with the input not containing the deviation amount of the health parameter, the following formula is formed:
in the formula, w and v are respectively system noise and measurement noise, and assume that w and v are Gaussian white noise with the mean value of zero, and covariance matrixes of the W and v are respectively Q and R; establishing a Kalman filter as follows:
in the formula (I), the compound is shown in the specification,
representing an estimate of a variable, K represents a kalman gain matrix; p is the solution of the ricati equation shown in equation (4):
AP+PA T -PC T R -1 CP+Q＝0 (4)。
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211093004.XA CN115576294A (en) | 2022-09-08 | 2022-09-08 | Fault-tolerant soft-hard hybrid control method for fault diagnosis of aircraft engine sensor |
| PCT/CN2022/121216 WO2024050883A1 (en) | 2022-09-08 | 2022-09-26 | Fault-tolerant soft-hard hybrid control method for fault diagnosis of sensor of aero-engine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211093004.XA CN115576294A (en) | 2022-09-08 | 2022-09-08 | Fault-tolerant soft-hard hybrid control method for fault diagnosis of aircraft engine sensor |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115576294A true CN115576294A (en) | 2023-01-06 |
Family
ID=84582095
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211093004.XA Pending CN115576294A (en) | 2022-09-08 | 2022-09-08 | Fault-tolerant soft-hard hybrid control method for fault diagnosis of aircraft engine sensor |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN115576294A (en) |
| WO (1) | WO2024050883A1 (en) |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9002678B1 (en) * | 2014-01-10 | 2015-04-07 | King Fahd University Of Petroleum And Minerals | Unified approach to detection and isolation of parametric faults using a kalman filter residual-based approach |
| CN109857094B (en) * | 2019-03-14 | 2020-06-02 | 杭州电子科技大学 | Two-stage Kalman filtering algorithm-based aeroengine fault diagnosis method |
-
2022
- 2022-09-08 CN CN202211093004.XA patent/CN115576294A/en active Pending
- 2022-09-26 WO PCT/CN2022/121216 patent/WO2024050883A1/en unknown
Also Published As
| Publication number | Publication date |
|---|---|
| WO2024050883A1 (en) | 2024-03-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109857094B (en) | Two-stage Kalman filtering algorithm-based aeroengine fault diagnosis method | |
| CA2354944C (en) | Sensor fault detection, isolation and accomodation | |
| US9151786B2 (en) | Systems, methods, and apparatus for detecting shifts in redundant sensor signals | |
| EP3458697A1 (en) | Controlling a gas turbine considering a sensor failure | |
| US20130191681A1 (en) | Systems, methods, and apparatus for signal processing-based fault detection, isolation and remediation | |
| US20130226492A1 (en) | Systems, methods, and apparatus for detecting and removing sensor signal impulse disturbances | |
| Kobayashi et al. | Evaluation of an enhanced bank of Kalman filters for in-flight aircraft engine sensor fault diagnostics | |
| JP5965406B2 (en) | System, method and apparatus for detecting irregular sensor signal noise | |
| CN110991024A (en) | Method for monitoring sudden change of gas circuit component under concurrent fault of aircraft engine control system | |
| CN110609479B (en) | Gas turbine sensor fault-tolerant control method based on linear variable parameter model | |
| US20230315084A1 (en) | Methods of health degradation estimation and fault isolation for system health monitoring | |
| CN114282570A (en) | Fault tolerance method and fault tolerance system for sensor of aircraft engine | |
| CN115576294A (en) | Fault-tolerant soft-hard hybrid control method for fault diagnosis of aircraft engine sensor | |
| Samy et al. | Fault detection and isolation (FDI) | |
| WO2012050473A1 (en) | Systems, methods, and apparatus for detecting agreement for individual channels among redundant sensor signals | |
| CN112749789A (en) | Aero-engine multiple fault diagnosis device based on self-association neural network | |
| CN113678107A (en) | Fault location in redundant acquisition systems | |
| Xue et al. | Application of Kalman filters for the fault diagnoses of aircraft engine | |
| Guo et al. | A distributed fault-detection and diagnosis system using on-line parameter estimation | |
| Panov et al. | Model-based compensation of sensor failure in industrial gas turbine | |
| Yang et al. | A Fault-Tolerant Soft/Hard Hybrid Control for Fault Diagnosis of Aero-engine Sensors | |
| Simani | Fault diagnosis of a simulated industrial gas turbine via identification approach | |
| Zhao et al. | Sensor Fault Diagnosis and Estimation Based on Multiple-Model Approach for Aeroengine | |
| WO2012050470A1 (en) | Systems, methods, and apparatus for remediating confidence values from redundant sensors | |
| CN117687379A (en) | Aeroengine control system actuating mechanism fault detection method based on unknown input observer |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |