Typical Fault Diagnosis Method for Electro-Hydraulic Servo Mechanism of Air Inlet Turbine Disk Type Control Valve in Altitude Test Stand

Abstract

The present invention provides a typical fault diagnosis method for the electro-hydraulic servo mechanism of an intake turbine disk type regulating valve in a high altitude test stand, comprising the following steps: Step 1: According to the pressure requirement, the intake system controller outputs a command signal for the valve opening, takes the command signal as the input of the electro-hydraulic servo actuator, the hydraulic cylinder outputs a displacement corresponding to the valve opening requirement, and feeds it back to the controller through a displacement sensor; Step 2: Inject faults into the displacement sensor and assign corresponding fault labels, extract the characteristics of the displacement sensor fault signal, obtain a prediction model in which the change trend of the modulus maximum value forms a corresponding relationship with the corresponding fault label, and perform fault identification and classification; Step 3: Establish a state space model of the actuator, design a sliding mode observer and corresponding thresholds, and form a fault detection strategy; Step 4: Simulate typical faults by changing the dominant characteristic parameters, and judge the faulty components and fault types based on the co-simulation results, so as to complete the verification.

Description

Technical Field

[0001] This specification relates to the technical field of aero-engines, and particularly to a typical fault diagnosis method for an electro-hydraulic servo mechanism of an air intake disk type regulating valve in a high-altitude test stand. Background Art

[0002] In the existing high-altitude simulation test machines, the types, numbers of test stands, test subjects, and test hours are all showing an increasing trend. Along with this, the test risks of the high-altitude test stand have also increased significantly. The safety and reliability of the key closed-loop regulation system of the air intake of the high-altitude test stand are important links to ensure the safety of the test of the engine under test. It mainly achieves the purpose of regulating the controlled parameters based on the control of displacement sensors and electro-hydraulic servo actuators. During the process of the high-altitude test stand performing the dynamic regulation tasks of the flight altitude and Mach number of the engine under test, the displacement sensors, drive, and regulating components in the system will all work for a long time under harsh working conditions of large loads, strong vibrations, and high-frequency use, with relatively large potential safety hazards. Moreover, technicians cannot quickly judge the location of the fault and cannot effectively replace the equipment. Therefore, it is necessary to carry out research on the online fault diagnosis method for the sensing and actuator of the intake pressure regulation system, so as to achieve the purpose of realizing real-time online diagnosis of typical fault states and effectively avoiding test risks.

[0003] As a hydraulic servo system, most of the hydraulic components and displacement sensors of the intake pressure regulation actuator are placed outdoors, facing the influences of large-span environmental temperature, humidity, dust, vibration shock, etc. for a long time. At the same time, the measurement signals in the intake pressure control system are easily contaminated by random noise, which is extremely likely to cause component failures. Conducting fault sorting for its electro-hydraulic servo actuator, and carrying out theoretical research on the diagnostic methods of displacement sensor signal analysis and processing, actuator fault analysis models, and adaptive sliding mode observers has guiding significance for improving the control accuracy and reliability of the intake pressure regulation system, and also provides a reference for the subsequent engineering application of fault diagnosis and fault-tolerant control technology methods for the intake pressure regulation displacement sensors and actuators of the high-altitude test stand environmental simulation system. Summary of the Invention

[0004] In view of this, the embodiments of this specification provide a typical fault diagnosis method for an electro-hydraulic servo mechanism of an air intake disk type regulating valve in a high-altitude test stand to improve the control accuracy and reliability of the intake pressure regulation system of the high-altitude test stand.

[0005] The technical solution of the present invention is as follows: A typical fault diagnosis method for the electro-hydraulic servo mechanism of the intake turbine disk type regulating valve in a high-altitude test stand, comprising the following steps: Step 1: According to the required pressure requirement, the intake system controller outputs an instruction signal for the valve opening, takes the instruction signal as the input of the electro-hydraulic servo actuator, the hydraulic cylinder outputs a displacement corresponding to the valve opening requirement, and feeds it back to the controller through a displacement sensor; Step 2: Inject faults into the displacement sensor and assign corresponding fault labels, extract the characteristics of the displacement sensor fault signal through the wavelet analysis method, use the supervised learning method to obtain a prediction model in which the changing trend of the modulus maximum value extracted under fault conditions forms a corresponding relationship with the corresponding fault label, and perform fault identification and classification; Step 3: Establish a state space model of the actuator, design a sliding mode observer and corresponding thresholds based on the state space model of the actuator, and form a fault detection strategy; Step 4: Utilize the established AMESim model of the actuator, simulate typical faults by changing the dominant characteristic parameters, and judge the faulty components and fault types based on the AMESim / MATLAB co-simulation results, so as to complete the verification of the typical fault diagnosis method for the electro-hydraulic servo actuator.

[0006] Further, Step 2 includes: Step 2.1, expanding the fault samples, injecting faults into the displacement sensor to obtain interference and power-off fault samples, and assigning corresponding fault labels.

[0007] Further, Step 2 also includes: Step 2.2, obtaining the multi-layer wavelet analysis results through the wavelet analysis method formula to obtain the modulus maximum value of each layer and perform linear connection, so as to extract the characteristics of the displacement sensor fault signal, where x(t) is the sensor output signal, φ a,τ (t) is the wavelet basis function, WT X (a,τ) is the multi-layer wavelet analysis signal decomposed from the sensor output signal, a is the scale parameter, and τ is the displacement parameter.

[0008] Further, Step 2 also includes: Step 2.3, using the existing supervised learning method in the MATLAB toolbox, taking the changing trends of the modulus maximum values and the corresponding fault labels of the interference fault samples and power-off fault samples as the training inputs, obtaining a prediction model in which the changing trend of the modulus maximum value forms a corresponding relationship with the corresponding fault label, and realizing the identification and classification of unknown interference and power-off faults through the prediction model.

[0009] Further, Step 3 includes: Step 3.1, establishing a state space model of the electro-hydraulic servo actuator, and establishing a sliding mode observer through the state space model, thereby forming an adaptive sliding mode observer.

[0010] Further, Step 3 further includes: Step 3.2, performing adaptive threshold design using a recursive algorithm.

[0011] Further, Step 3 further includes Step 3.3: When a fault occurs, the fault detection strategy is e y >Thmaxrore y <Th_minr; when no fault occurs, the fault detection strategy is Th_maxr < e y <Th_minr, where Th_maxr represents the upper bound of the threshold, Th_minr represents the lower bound of the threshold, and e y is the output error.

[0012] Specifically, Step 4 is as follows: Taking the spool displacement, the speed of the hydraulic cylinder piston rod, and the pressure in the high-pressure chamber of the hydraulic cylinder as observable state variables, using the AMESim / MATLAB co-simulation results as the input data for the sliding mode observer to provide fault diagnosis. The observer observes the observable state variables and outputs the residuals. According to the influence of different faults on the corresponding state observers and the comparison between the residuals and the thresholds, the faulty component and the type of fault are judged, thereby completing the verification of the typical fault diagnosis method for the electro-hydraulic servo actuator.

[0013] Compared with the prior art, the at least one technical solution adopted in the embodiments of this specification can achieve at least the following beneficial effects: The present invention realizes that when typical faults occur in the special type of disk regulating valve actuator, it can effectively analyze the fault parameter characteristics of important components, quickly and accurately locate the fault location, enabling technicians to replace the equipment in a timely and effective manner, effectively saving energy consumption and personnel requirements during fault troubleshooting. Description of the Drawings

[0014] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0015] Figure 1 is the method schematic diagram of the embodiments of the present invention;

[0016] Figure 2 is the schematic diagram of the wavelet analysis result of the displacement sensor open circuit fault;

[0017] Figure 3 is the schematic diagram of the change trend of the modulus maximum value of the detail coefficient of different decomposition layers of the open circuit fault;

[0018] Figure 4 is the wavelet analysis result of the displacement sensor interference fault;

[0019] Figure 5 It is a schematic diagram of the change trend of the modulus maximum value of the detail coefficients at different decomposition levels of the interference fault;

[0020] Figure 6 It is a schematic diagram of the fault classification accuracy of the supervised learning algorithm;

[0021] Figure 7 It is a schematic diagram of the valve displacement observation effect in the case of internal leakage fault;

[0022] Figure 8 It is the residual value and the corresponding threshold obtained from the valve displacement and its observed value in the case of internal leakage fault;

[0023] Figure 9 It is a schematic diagram of the cylinder piston speed observation effect in the case of internal leakage fault;

[0024] Figure 10 It is the residual value and the corresponding threshold obtained from the cylinder piston speed and its observed value in the case of internal leakage fault;

[0025] Figure 11 It is a schematic diagram of the cylinder high-pressure chamber pressure observation effect in the case of internal leakage fault;

[0026] Figure 12 It is the residual value and the corresponding threshold obtained from the cylinder high-pressure chamber pressure and its observed value in the case of internal leakage fault; Specific implementation mode

[0027] The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0028] The following uses specific specific examples to illustrate the implementation manners of the present application. Those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The present application can also be implemented or applied through other different specific implementation manners. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that, without conflict, the following embodiments and the features in the embodiments can be combined with each other. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without making creative efforts belong to the scope of protection of the present application.

[0029] As Figures 1 to 11 shown, the embodiments of the present invention provide a typical fault diagnosis method for the electro-hydraulic servo mechanism of the air intake turbine disk type regulating valve at high altitude platforms, including:

[0030] Step 1: According to the required pressure in the altitude test chamber, the intake system controller outputs a command signal for the valve opening. This signal serves as the input to the electro-hydraulic servo actuator. The hydraulic cylinder outputs a displacement corresponding to the valve opening requirement, and is feedback to the controller through a displacement sensor;

[0031] Step 2: First, fault injection is performed on the displacement sensor to obtain 2000 groups of interference and power-off fault samples each, and corresponding fault labels are assigned. Secondly, the wavelet analysis method is used to extract the characteristics of the displacement sensor fault signal. The extracted signal is divided into 6 decomposition layers, and the modulus maximum value of each layer is taken to obtain the change trend of the modulus maximum value of different decomposition layers. Finally, using the existing supervised learning method in the MATLAB toolbox, the change trend of the modulus maximum value extracted from the above fault samples and its fault label are used as the training input, so as to obtain a prediction model in which the change trend of the modulus maximum value extracted under fault conditions corresponds to the corresponding fault label, and then realize the identification and classification of unknown interference and power-off faults;

[0032] Step 3: Establish a state space model of the actuator, design a sliding mode observer based on this model, and design corresponding thresholds to form a fault detection strategy;

[0033] Step 4: Use the AMESim model of the actuator to simulate typical faults by changing the dominant characteristic parameters. The spool displacement, the speed of the hydraulic cylinder piston rod, and the pressure in the high-pressure chamber of the hydraulic cylinder are observable state variables. The combined AMESim / MATLAB simulation results provide input data for fault diagnosis to the sliding mode observer. The observer observes the above state variables and outputs residuals. According to the influence of different faults on the corresponding state observer and the comparison between the residuals and the thresholds, the faulty components and fault types are judged, thus completing the verification of the typical fault diagnosis method for the electro-hydraulic servo actuator.

[0034] Further, the specific steps of Step 2 include:

[0035] (1) Fault sample expansion:

[0036] Fault injection is performed on the displacement sensor (the input signal of the sensor is processed by equal difference increment, such as the amplitude of the interference fault), so as to obtain 2000 groups of interference and power-off fault samples each, and corresponding fault labels are assigned.

[0037] (2) Sensor fault signal extraction method:

[0038] The wavelet analysis method is used for the fault feature extraction of the displacement sensor, and its expression is as follows:

[0039]

[0040] where x(t) is the sensor output signal, φa,τ (t) is the wavelet basis function, WT X (a, τ) is the 6-layer wavelet analysis signal decomposed from the sensor output signal, a is the scale parameter (controlling the decomposition layer number, here it is 6), and τ is the displacement parameter.

[0041] Take the modulus of the maximum value of each layer to obtain 6 modulus maxima and linearly connect these 6 points, so as to obtain the changing trend of the modulus maxima with different decomposition layer numbers. Thus, the feature extraction of the sensor fault signal is completed.

[0042] (3) Supervised learning method:

[0043] Adopt the existing supervised learning method in the MATLAB toolbox, and use the changing trend of the modulus maxima and the corresponding fault labels of 2000 groups of interference fault samples and 2000 groups of power-off fault samples as the training input. Its training criterion is: the changing trend of the modulus maxima extracted from the fault samples forms a corresponding relationship with its fault label, so as to obtain a prediction model in which the changing trend of the modulus maxima extracted under fault conditions forms a corresponding relationship with the corresponding fault label. Finally, the identification and classification of unknown interference and power-off faults are realized.

[0044] Furthermore, the specific steps of step 3 include:

[0045] (1) Sliding mode observer design method:

[0046] The established state space model of the electro-hydraulic servo actuator is:

[0047]

[0048] Among them, x ∈ R n (n = 4), u ∈ R m (m = 1), y ∈ R p (p = 4) are the state, input, and output vectors of the electro-hydraulic servo actuator respectively. g(x) is a known nonlinear term that satisfies the Lipschitz condition, The specific expression is shown in Equation (3).

[0049] B = [0 0 0 0.7758], x 1 、x 2 、x 3 、x 4 are the speed of the piston rod, the pressure on the high-pressure side of the hydraulic cylinder, the pressure on the low-pressure side of the hydraulic cylinder, and the displacement of the servo valve spool respectively; u is the input current of the electro-hydraulic servo actuator; y 1 、y 2 、y 3 、y4 They are respectively the speed of the hydraulic cylinder piston rod, the pressure on the high-pressure side of the hydraulic cylinder, the pressure on the low-pressure side of the hydraulic cylinder, and the displacement of the servo valve spool.

[0050]

[0051] Design sliding mode observers as shown in Equation (4) for the speed of the hydraulic cylinder piston rod, the displacement of the servo valve spool, and the pressure on the high-pressure side of the hydraulic cylinder in the electro-hydraulic servo actuator system established in Equation (2):

[0052]

[0053] Where: G n is the gain of the nonlinear term, and its function is to expand the sliding mode surface. The Gs of the hydraulic cylinder piston rod speed observer, the servo valve spool displacement observer, and the hydraulic cylinder high-pressure side pressure observer n are respectively taken as: G l is the observer gain matrix, and its function is to make the observation error asymptotically converge to 0. The Gs of the hydraulic cylinder piston rod speed observer, the servo valve spool displacement observer, and the hydraulic cylinder high-pressure side pressure observer l are respectively taken as: e y is the output error; v(t) is the sliding mode variable structure input signal, and its expression is:

[0054]

[0055] Among them, η(t) is the time-varying parameter designed for the adaptive law, which changes with the error and time, and its form is as shown in Equation (6). The design of the adaptive law and C2, D2 is to enable the nonlinear term to switch around the sliding mode surface and drive e y to move to the sliding mode surface. δ is a very small integer, and take δ = 0.001; the sliding mode surface is defined as s = {e y : e y = 0}, and there is:

[0056]

[0057] Among them, ρ, η 0 are constants greater than 0 and are 0.8 and 1.25 respectively.

[0058] (2) Threshold design method:

[0059] The application of fixed thresholds is more common in fault detection. However, when the uncertainty of the system changes significantly, a large number of false alarms and missed alarms will occur. To reduce the false alarm rate and missed alarm rate, an adaptive threshold can be designed for judgment. The design of the adaptive threshold is as follows:

[0060] The residual output by the observer is r, its mathematical expectation E(r), and the standard deviation is S(r).

[0061] The upper bound of the threshold is selected as:

[0062] Th_maxr = E(r) + n 1 S(r) (7)

[0063] where Th_maxr represents the upper bound of the threshold, and n 1 is a constant greater than 0.

[0064] The lower bound of the threshold is selected as:

[0065] Th_minr = r - n 2 S(r) (8)

[0066] where Th_minr represents the lower bound of the threshold, and n 2 is a constant greater than 0.

[0067] The mathematical expectation and standard deviation can be replaced by the average value of the residuals statistically calculated online the mean value and the sample mean standard deviation instead.

[0068]

[0069]

[0070] where N is the number of sampling times, and r i is the residual output by the observer at each sampling moment.

[0071] To reduce the computational load and storage space requirements, a recursive algorithm can be used:

[0072]

[0073]

[0074] (3) The fault detection strategy is as follows:

[0075]

[0076] Furthermore, the specific steps of step 4 include:

[0077] Using the established AMESim model of the actuator, typical faults are simulated by changing the dominant characteristic parameters. The spool displacement, the speed of the hydraulic cylinder piston rod, and the pressure in the high-pressure chamber of the hydraulic cylinder are observable state variables, and the results of the AMESim / MATLAB co-simulation provide the input data for the fault diagnosis of the sliding mode observer (as Figure 4 shown). The observer observes the above state variables and outputs the residuals. According to the influence of different faults on the corresponding state observers and the comparison between the residuals and the thresholds, the faulty components and the types of faults are judged, thus completing the verification of the fault diagnosis method for the electro-hydraulic servo actuator. Among them:

[0078] The fault simulation method based on the AMESim model is as follows:

[0079] (1) Leakage fault: The internal leakage of the cylinder is essentially caused by the increase in the sealing gap between the piston and the cylinder body or the pressure difference between the high and low pressure chambers (such as wear or overload). Considering that the pressure difference between the two chambers is mainly determined by the load and working conditions, the gap between the piston and the cylinder wall is taken as the dominant characteristic parameter. Based on the AMESim model of the actuator, the gap size corresponding to the leakage amount is set to simulate the leakage fault of the hydraulic cylinder.

[0080] (2) Spool jamming fault: Spool jamming is due to the accumulation of particles in the valve, that is, the spool cannot move at a certain moment. Therefore, the spool displacement is taken as the dominant characteristic parameter corresponding to this fault, and the limit condition corresponding to the jamming degree is set based on the actuator model to simulate the spool jamming.

[0081] (3) Electrical open circuit fault: Open circuit is a common electrical fault form of the servo valve, and generally, there will be an obvious change in the coil current. The input current signal of the servo valve is taken as the dominant characteristic parameter, and the current signal corresponding to the open circuit of the servo valve is set based on the actuator model to simulate this electrical fault.

[0082] Based on the displacement sensor model of a certain type of high-altitude test stand special wheel disc regulating valve, the typical fault diagnosis simulation of the displacement sensor is carried out.

[0083] Figure 2 And Figure 3 shown is the simulation result of extracting the open circuit fault of the displacement sensor, Figure 2 is the wavelet analysis result of the open circuit fault of the displacement sensor. The time points of the fault occurrence can be seen from the 6-layer wavelet decomposition results of the output signal of the displacement sensor with an open circuit fault, and as the decomposition layer increases, the detail coefficients obtained by the wavelet decomposition gradually increase. Figure 3 is the changing trend of the modulus maximum value of the detail coefficients of different decomposition layers for the open circuit fault. When the output signal of the displacement sensor has an open circuit fault, as the decomposition layer increases, the modulus maximum value of the detail coefficients obtained by the wavelet decomposition shows an increasing trend. Comparing Figure 2 and Figure 3It is concluded that when the same open - circuit fault occurs in different signals, their wavelet analysis results are basically the same, and the variation trends of the modulus maxima of the detail coefficients obtained by wavelet decomposition are also basically the same.

[0084] As Figure 4 and Figure 5 shown in the simulation results of extracting the interference fault of the displacement sensor, Figure 4 is the wavelet analysis result of the interference fault of the displacement sensor. The time points of the fault occurrence can be seen from the 6 - layer wavelet decomposition results of the output signal of the displacement sensor with interference faults. As the decomposition layer increases, the detail coefficients obtained by wavelet decomposition become smaller and smaller. Figure 5 is the variation trend of the modulus maxima of the detail coefficients with different decomposition layers for the interference fault. When the output signal of the displacement sensor has an interference fault, as the decomposition layer increases, the modulus maxima of the detail coefficients obtained by wavelet decomposition show a decreasing trend. By comparing Figure 4 and Figure 5 It is concluded that when the same interference fault occurs in different signals, their wavelet analysis results are basically the same, and the variation trends of the modulus maxima of the detail coefficients obtained by wavelet decomposition are also basically the same.

[0085] As Figure 6 shown in the fault classification accuracy of the supervised learning algorithm. Here, three faults are selected. From the results in the figure, it can be seen that 171 faults of type 1 and 287 faults of type 2 are misclassified as fault 3. The classification accuracy of fault 1 is 79%, the classification accuracy of fault 2 is 82%, and the classification accuracy of fault 3 is 100%. The overall classification accuracy of the algorithm is 88.5%.

[0086] Based on the nonlinear model of the electro - hydraulic servo actuator of a certain type of high - altitude test stand special wheel - disc regulating valve, the typical fault diagnosis simulation of the actuator is carried out by means of co - simulation.

[0087] As Figures 7 to 12 shown in the simulation results of the leakage fault of the actuator hydraulic cylinder, where Figure 7 is the schematic diagram of the valve displacement observation effect in the case of internal leakage fault; Figure 8 is the residual value and the corresponding threshold obtained by comparing the valve displacement with its observed value in the case of internal leakage fault; Figure 9 is the schematic diagram of the cylinder piston speed observation effect in the case of internal leakage fault; Figure 10 is the residual value and the corresponding threshold obtained by comparing the cylinder piston speed with its observed value in the case of internal leakage fault; Figure 11 is the schematic diagram of the high - pressure chamber pressure observation effect of the cylinder in the case of internal leakage fault; Figure 12It is the residual value and the corresponding threshold obtained from the pressure in the high-pressure chamber of the cylinder and its observed value in the case of internal leakage failure. By setting the distance between the piston rod and the piston cylinder in the model to simulate the failure situation with large internal leakage, when internal leakage occurs, obvious residuals appear in both the observed displacement of the main spool of the servo valve and the observed speed of the piston rod compared with the actual values. Therefore, it is possible to judge whether a leakage failure has occurred based on the residual and the set adaptive threshold.

[0088] As Figures 7 to 12 Shown in the figure is the state parameters, the output of the corresponding observer, the residual and the threshold when the hydraulic cylinder of the actuator has a leakage failure. It can be concluded that when the hydraulic cylinder has a leakage failure, except for the spool displacement, the other observed values fail to track other state parameters and the residuals are all lower than the lower limit of the threshold. By comparing with the simulation results of other failures, it forms a difference from the observation effects of other failures, and this is used as the basis for judging the leakage failure.

[0089] In summary, for the typical fault diagnosis method of the electro-hydraulic servo mechanism of the air intake turbine disc regulating valve in the high-altitude test stand based on the adaptive sliding mode observer and the displacement sensor signal analysis technology, through the simulation study on the extraction of the fault characteristics of the sensor, the change trends of the characteristic parameters are different for different faults. Therefore, the linear identification of the sensor fault adopted has a high accuracy rate, providing a favorable reference basis for judging the displacement sensor fault. By designing an adaptive sliding mode fault observer with good observation ability through the state space model, it is possible to judge the location and type of the fault based on the dynamic parameter residual values of each typical fault and the set corresponding adaptive thresholds, providing a reference for the fault diagnosis technology of the high-altitude test stand actuator, and also providing a reference for the engineering application of the high-altitude test stand fault diagnosis and the fault tolerance control technology method.

[0090] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any change or replacement that can be easily thought of by those skilled in the art within the technical scope disclosed in the present application should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A typical fault diagnosis method for the electro-hydraulic servo mechanism of the intake turbine disk type regulating valve in a high-altitude test stand, characterized in that, it includes the following steps: Step 1: According to the required pressure requirement, the intake system controller outputs a command signal for the valve opening, takes the command signal as the input of the electro-hydraulic servo actuator, the hydraulic cylinder outputs a displacement corresponding to the valve opening requirement, and feeds it back to the controller through a displacement sensor; Step 2: Inject faults into the displacement sensor and assign corresponding fault labels, extract the characteristics of the displacement sensor fault signal through the wavelet analysis method, use the supervised learning method to obtain a prediction model in which the change trend of the modulus maximum value extracted under fault conditions forms a corresponding relationship with the corresponding fault label, and perform fault identification and classification; Step 3: Establish a state space model of the actuator, design a sliding mode observer and corresponding thresholds based on the state space model of the actuator, and form a fault detection strategy; Step 4: Use the established AMESim model of the actuator, simulate typical faults by changing the dominant characteristic parameters, and judge the faulty components and fault types based on the AMESim / MATLAB co-simulation results, so as to complete the verification of the typical fault diagnosis method for the electro-hydraulic servo actuator; The said Step 2 includes: Step 2.1, fault sample expansion, inject faults into the displacement sensor to obtain interference and power-off fault samples, and assign corresponding fault labels; Step 2.2, through the wavelet analysis method formula Obtain the multi-layer wavelet analysis results, and then obtain the modulus maxima of each layer and perform linear connection to extract the features of the displacement sensor fault signal, where is the sensor output signal, is the wavelet basis function, is the multi-layer wavelet analysis signal decomposed from the sensor output signal, is the scale parameter, is the displacement parameter; Step 2.3, use the existing supervised learning method in the MATLAB toolbox, take the change trend of the modulus maximum value and the corresponding fault label of the interference fault sample and the power-off fault sample as the training input, obtain a prediction model in which the change trend of the modulus maximum value forms a corresponding relationship with the corresponding fault label, and realize the identification and classification of unknown interference and power-off faults through the prediction model; The said Step 4 is specifically: Take the spool displacement, the speed of the hydraulic cylinder piston rod and the pressure of the high-pressure chamber of the hydraulic cylinder as observable state variables, use the AMESim / MATLAB co-simulation results as the input data for fault diagnosis of the sliding mode observer, the observer observes the said observable state variables and outputs residuals, and judge the faulty components and fault types according to the influence of different faults on the corresponding state observer and the comparison between the residuals and the thresholds, so as to complete the verification of the typical fault diagnosis method for the electro-hydraulic servo actuator.

2. The typical fault diagnosis method for the electro-hydraulic servo mechanism of the intake turbine disk type regulating valve according to claim 1, characterized in that, the said Step 3 includes: Step 3.1, establish a state space model of the electro-hydraulic servo actuator, and establish a sliding mode observer through the state space model, and then form an adaptive sliding mode observer.

3. The typical fault diagnosis method for the electro-hydraulic servo mechanism of the intake turbine disk type regulating valve according to claim 2, characterized in that, the said Step 3 further includes: Step 3.2, adopt a recursive algorithm for adaptive threshold design.

4. The typical fault diagnosis method for the electro-hydraulic servo mechanism of the intake turbine disk type regulating valve according to claim 3, characterized in that, Step 3 further includes Step 3.3: When a fault occurs, the fault detection strategy is ； When no fault occurs, the fault detection strategy is , where represents the upper threshold, represents the lower threshold, and e y is the output error.

Images