Online Detection Method of Actuator Fault

Abstract

The present invention relates to an online detection method for actuator faults in the field of online health monitoring technology, comprising: designing a nominal model of actuator dynamic characteristics, and rewriting the system form using the nominal model of actuator dynamic characteristics; designing an extended state observer based on the nominal model of actuator dynamic characteristics, and obtaining the extended state of the system using the extended state observer; and designing a fault identification strategy based on the extended state. Through the above method, online diagnosis and identification of faults of actuator dynamic characteristics of complex uncertain nonlinear systems can be achieved without actuator sensors, and is not affected by external interference.

Description

Technical Field

[0001] The present invention relates to the technical field of online health monitoring, and in particular to an online detection method for actuator faults without sensors. Background Art

[0002] With the rapid development of industrial technology, the safety and reliability requirements of the system are constantly increasing. To this end, people have conducted a lot of research on fault diagnosis and isolation technology. The theory and application of nonlinear observers have gradually become the mainstream of fault diagnosis research.

[0003] For closed-loop control systems, the dynamic characteristics of actuators (such as the speed of aircraft servo deflection, etc., the dynamic characteristics of actuators are usually described by bandwidth) play a vital role in system stability. If the dynamic characteristics of actuators are too slow, the performance of closed-loop control systems will be reduced or even diverge. In engineering, when designing complex control laws for various controlled objects, it is usually assumed in advance that the bandwidth of the actuators is large enough, that is, the dynamic characteristics are fast enough (generally, the bandwidth of the actuators is required to be more than 3 to 5 times the control bandwidth of the closed-loop system), that is, it is assumed that there is no amplitude attenuation, phase lag, and time delay between the control signal given by the control law and the actual output of the actuators, so the dynamic characteristics of the actuators are ignored when designing the control law. In actual engineering, due to the lack of reliability of the actuators themselves, as well as the effects of external forces and thermal loads, especially for actuators that work for a long time, the dynamic performance of the actuators will be reduced or even fail, resulting in inevitable errors between the actual output force and torque of the actuators and the control signals required by the control system, affecting the overall performance of the closed-loop system, and even causing the control system to diverge under extreme conditions. Therefore, in actual engineering, there is an urgent need for corresponding fault diagnosis technology that can identify the performance degradation or even failure of the actuator in real time, so that corresponding fault isolation measures can be taken.

[0004] However, under the influence of uncertain factors such as the uncertainty of the model parameters of the controlled object itself and the existence of strong external interference, it is difficult to identify the attenuation or failure of the dynamic characteristics of the actuator alone without sensors for measuring the input and output of the actuator. Summary of the invention

[0005] In order to overcome the defects of the above-mentioned prior art, the purpose of the present invention is to provide an online detection method for actuator faults, which can realize online diagnosis and identification of faults in the dynamic characteristics of actuators of complex uncertain nonlinear systems without the need for actuator sensors and is not affected by external interference.

[0006] In order to achieve the above-mentioned object of the invention, the technical solution of the present invention is:

[0007] The present invention provides an actuator fault online detection method, comprising:

[0008] Designing a nominal model of dynamic characteristics of an actuator, and rewriting the system form using the nominal model of dynamic characteristics of the actuator;

[0009] Designing an extended state observer according to the nominal model of the dynamic characteristics of the actuator, and using the extended state observer to obtain the extended state of the system;

[0010] A fault identification strategy is designed according to the expansion state.

[0011] Beneficial effects:

[0012] According to the actuator fault online diagnosis method of the present invention, based on the extended state observer to observe the system state response and actuator input, the corresponding fault diagnosis strategy can accurately identify the actuator fault from the uncertainty of the complex model of the controlled object and the strong external interference, and realize the online diagnosis of the dynamic characteristics of the actuator of the complex uncertain nonlinear system. The method has strong robustness to the composite interference other than the dynamic characteristics of the actuator, has high sensitivity and recognition accuracy to the actuator fault, and has strong engineering practicality. In addition, the present invention does not need to design and configure the measurement sensor specifically for the actuator, but can achieve the above-mentioned identification effect through the state of the controlled object (already available when the control system is designed) and the ideal control signal (given by the control law), which can effectively reduce the system cost and have a high economic effect after popularization and application. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0014] Figure 1 A flowchart of an actuator fault online detection method provided by an embodiment of the present invention is schematically shown;

[0015] Figure 2 A schematic diagram showing a simulation diagram of the longitudinal pitch angle tracking effect of an actuator under normal conditions of a certain aircraft in the atmosphere provided by an embodiment of the present invention, when the aircraft is subject to the uncertainty of its own model parameters and the influence of external interference factors;

[0016] Figure 3 A schematic diagram showing a simulation diagram of the fault diagnosis and identification effect of an actuator under normal conditions of a certain atmospheric aircraft under the condition that the aircraft is subject to the uncertainty of its own model parameters and external interference factors provided by an embodiment of the present invention;

[0017] Figure 4 A schematic diagram showing a simulation diagram of input-output curves of an actuator structure of an aircraft in an atmosphere under normal conditions, provided by an embodiment of the present invention, when the aircraft is subject to the uncertainty of its own model parameters and the influence of external interference factors;

[0018] Figure 5 A schematic diagram showing a simulation of the longitudinal pitch angle tracking effect of a certain aircraft in the atmosphere under the condition of actuator failure provided by an embodiment of the present invention, when the aircraft is subject to the uncertainty of its own model parameters and external interference factors;

[0019] Figure 6 A schematic diagram showing a fault diagnosis and identification effect simulation diagram of a certain atmospheric aircraft under an actuator fault condition provided by an embodiment of the present invention, when the aircraft is subject to the uncertainty of its own model parameters and the influence of external interference factors;

[0020] Figure 7 A diagram schematically shows a simulation diagram of an actuator structure input-output curve under actuator failure conditions of an aircraft in an atmosphere provided by an embodiment of the present invention, when the aircraft is subject to uncertainty in its own model parameters and external interference factors. DETAILED DESCRIPTION

[0021] The description of the embodiments of this specification should be combined with the corresponding drawings, which should be considered as part of the complete specification. In the drawings, the shape or thickness of the embodiments may be enlarged and indicated for simplification or convenience. Furthermore, the parts of each structure in the drawings will be described separately. It is worth noting that the elements not shown in the drawings or not described in words are in a form known to ordinary technicians in the relevant technical field.

[0022] The description of the embodiments herein and any reference to directions and orientations are only for the convenience of description and are not to be construed as any limitation on the scope of protection of the present invention. The following description of the preferred embodiments may involve combinations of features, which may exist independently or in combination, and the present invention is not particularly limited to the preferred embodiments. The scope of the present invention is defined by the claims.

[0023] Figure 1 The steps of the actuator fault online detection method according to the present embodiment are schematically shown.

[0024] like Figure 1 As shown, the actuator fault online detection method comprises the following steps:

[0025] Step 110, designing a nominal model of the dynamic characteristics of the actuator, and rewriting the system form using the nominal model of the dynamic characteristics of the actuator;

[0026] Step 120: designing an extended state observer according to the nominal model of the dynamic characteristics of the actuator, and using the extended state observer to obtain the extended state of the system;

[0027] Step 130: Design a fault identification strategy according to the expansion state.

[0028] In this embodiment, the above method is designed for the following n-dimensional single-input single-output system, and the system model is:

[0029]

[0030] Where x = [x1, x2, ..., x n ] T is the system status, is the nonlinear term in the system, unknown; is the external disturbance, y(t) is the system output, u is the control signal given by the control law, and δ(t) is the actual output of the actuator. Usually the dynamic characteristics of the actuator are simulated using a first-order model, τ is the inertia time constant of the first-order model, and the smaller τ is, the faster the dynamic characteristics of the actuator are.

[0031] In step 110, the dynamic characteristic nominal model of the actuator is designed as:

[0032]

[0033] Among them, u is the control signal given by the control law, τ0 is the nominal value of the inertia time constant τ of the first-order model simulating the dynamic characteristics of the actuator, and δ0 is the actuator output predicted by the nominal model of the dynamic characteristics of the actuator.

[0034] In addition, the nominal model of the actuator dynamic characteristics is used to rewrite the system into the following form that is convenient for the design of the extended state observer:

[0035]

[0036] in, is the total system disturbance used in the design of the extended state observer, including the total difference between the original system and the target system; τ0 is the nominal value of the inertia time constant τ of the first-order model simulating the dynamic characteristics of the actuator, and δ0 is the actuator output predicted by the nominal model of the dynamic characteristics of the actuator.

[0037] In step 120, the designed extended state observer is:

[0038]

[0039] in, For the system state x = [x1, x2, ..., x n ]T The estimated value of is the total disturbance of the system The estimated value of , design the gains α1, α2, ..., α of the extended state observer n+1 , so that the polynomial s n+1 +α1s n +α2s n-1 +…α n s+α n+1 is Hurwitz's, and the extended state observer scaling factor ε is a small positive constant. This formula can also be called a fault identification filter based on the extended state observer.

[0040] In step 130, the process of designing a fault identification strategy according to the expansion state includes: designing a residual evaluation function according to the expansion state, and using the residual evaluation function to diagnose and identify the fault of the actuator according to an appropriate judgment threshold. The designed residual evaluation function (i.e., the fault identification strategy) is:

[0041]

[0042] in, is the estimation error of the extended state observer; design J is the judgment threshold, indicating the health of the system, that is, when the inertia time constant τ=τ0 of the first-order model simulating the dynamic characteristics of the actuator, the initial value x0 of the system state has the maximum influence on the residual evaluation function.

[0043] The fault diagnosis logic for judging the threshold is as follows:

[0044]

[0045] In order to fully and in detail illustrate the specific process of the method design of the present invention and the beneficial effects obtained, this embodiment selects the longitudinal dynamics model of the following aircraft in the atmosphere as a verification example, and obtains simulation results and analyzes the results for the longitudinal pitch angle tracking process of the aircraft through the above-mentioned online fault detection method. The longitudinal dynamics model of the following aircraft in the atmosphere is:

[0046]

[0047]

[0048] Where α is the angle of attack (°), q is the pitch velocity (° / s), and δ is the elevator deflection angle (°). d1(t) and d2(t) represent all disturbances caused by the unmodeled dynamics of the system, external wind disturbances, and changes in the aerodynamic parameters of the system. f1(α), f2(α), b1(α), and b2 depend on the aerodynamic coefficients. The aerodynamic coefficients of the aircraft at Mach 3, 6095 meters (20,000 ft.) altitude, and angle of attack |α|≤20 degrees are:

[0049]

[0050]

[0051]

[0052]

[0053] The rudder surface deflection dynamics are represented by the following first-order model:

[0054]

[0055] Where u is the control command given by the control law, and τ>0 is the first-order inertia time constant. The actual control surface deflection angle |δ|≤30°. The system control law is taken as nonlinear dynamic inverse control based on disturbance observer. The output of the aircraft is taken as:

[0056] x1=α+k q q (14)

[0057] where k q =0.06. From equations (7) to (13), the aircraft dynamics system can be rewritten as:

[0058]

[0059] Where f = f1(α) + k q f2(α)+q+d1(t)+k q d2(t)(16), b=b1+k q b2(17),

[0060] For equation (15), the following fault diagnosis strategy is designed according to equations (4) to (6):

[0061]

[0062] The parameters in the above formula are ε=0.1,α1=1,α2=0.5,τ0=0.1,Δt=4,J th = 6000. So far, the actuator fault online diagnosis strategy based on the improved extended state observer can be obtained.

[0063] The effectiveness of the actuator fault online detection method of the present invention is given by the following simulation example analysis. The flight control command signal is selected as a sinusoidal signal with an amplitude of 10° and a frequency of 1rad / s. The simulation adopts a discrete form and the sampling time is 1 / 200s. In the following example, the aerodynamic parameter f1(α) in the above formula (9) becomes 125% of the nominal value at the 12th second, the aerodynamic parameter f2(α) in the above formula (10) becomes 125% of the nominal value at the 14th second, the control gain b1 becomes 90% of the nominal value at the 16th second, and the control gain b2 becomes 90% of the nominal value at the 18th second. A step disturbance d1(t) with an amplitude of 0.1 is introduced at the 20th second, and a sinusoidal disturbance d2(t) with an amplitude of 5 and a frequency of 10rad / s is introduced at the 22nd second.

[0064] Figure 2 This is the system state command value and its response value. It can be seen from the figure that due to the influence of various interferences, the tracking effect of the control system has been weakened. Figure 3 It can be seen that the evaluation function is not affected by these disturbances. Figure 4 Indicates that the actuator is working well.

[0065] The above indicates that the fault diagnosis strategy designed in the present invention is not affected by the uncertainty of the controlled object itself and external interference.

[0066] On this basis, an actuator fault is introduced at the 30th second in the simulation, that is, the inertia time constant τ in the above formula (13) changes from 0.1 to 1, and the dynamic characteristics of the actuator are greatly reduced. Figure 5 It can be seen that due to the failure of the actuator, the system output response value deviates from the command value. Figure 6 It can be seen that the actuator fault can be accurately identified by the fault diagnosis strategy of the present invention. Figure 7 The fault state curve of the actuator can be seen.

[0067] As can be seen from the above, the above actuator fault online detection method is targeted at uncertain nonlinear systems and is not affected by the uncertainty of the controlled object itself and external interference. When the dynamic characteristics of the actuator decay, this method can identify it in a timely and accurate manner. At the same time, there is no need to set up actuator sensors separately, which provides a new idea for the actuator fault diagnosis of complex nonlinear systems represented by aircraft.

[0068] The serial numbers of the above-mentioned steps involved in the method of the present invention do not mean the order of execution of the method. The execution order of each step should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the implementation mode of the present invention.

[0069] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An actuator fault online detection method, comprising: Designing a nominal model of dynamic characteristics of an actuator, and rewriting the system form using the nominal model of dynamic characteristics of the actuator; Designing an extended state observer according to the nominal model of the dynamic characteristics of the actuator, and using the extended state observer to obtain the extended state of the system; designing a fault identification strategy based on the expansion state; The nominal model of the dynamic characteristics of the actuator is: Wherein, u is the control signal given by the control law, τ0 is the nominal value of the inertia time constant τ of the first-order model simulating the dynamic characteristics of the actuator, and δ0 is the actuator output predicted by the nominal model of the dynamic characteristics of the actuator; The system is an n-dimensional single-input single-output system, and the system model is: Where x=[x1,x2,…,x n ] T is the system status, is the nonlinear term in the system, unknown; is the external disturbance, y(t) is the system output, u is the control signal given by the control law, and δ(t) is the actual output of the actuator. The dynamic characteristics of the actuator are simulated using a first-order model, and τ is the inertia time constant of the first-order model. The smaller τ is, the faster the dynamic characteristics of the actuator are. The system is rewritten into the following form that is convenient for the design of the extended state observer by using the nominal model of the dynamic characteristics of the actuator: in, is the total system disturbance used in the design of the extended state observer, including the total difference between the original system and the target system; τ0 is the nominal value of the inertia time constant τ of the first-order model simulating the dynamic characteristics of the actuator, and δ0 is the actuator output predicted by the nominal model of the dynamic characteristics of the actuator.

2. The method according to claim 1, characterized in that The extended state observer is: in, For the system state x=[x1,x2,…,x n ] T The estimated value of is the total disturbance of the system The estimated value of the extended state observer is designed to gain α1, α2, …, α n+1 , so that the polynomial s n+1 +α1s n +α2s n-1 +…α n s+α n+1 is Hurwitz's, extended state observer scaling factor ε is a small positive constant.

3. The method according to claim 1, characterized in that The process of designing a fault identification strategy according to the expansion state includes: designing a residual evaluation function according to the expansion state, and using the residual evaluation function to diagnose and identify the fault of the actuator according to a judgment threshold.

4. The method according to claim 3, characterized in that The residual evaluation function is: in, is the estimated error of the extended state observer; design is the judgment threshold, indicating system health, that is, when the inertia time constant τ=τ0 of the first-order model simulating the dynamic characteristics of the actuator, the maximum influence of the initial value x0 of the system state on the residual evaluation function.

5. The method according to claim 4, characterized in that The fault diagnosis logic of the judgment threshold is:

Images