Fault identification method and system for dissimilar redundancy actuation system

Abstract

The invention provides a fault identification method and system for a dissimilar redundancy actuation system. Comprising the steps of determining a state space model of the NT-DRAS dissimilar redundancy actuation system in a main working mode, determining a state space model after a fault is injected, establishing a multi-dimensional fault representation model, performing fault identification on the NT-DRAS dissimilar redundancy actuation system with an actual fault, and the like. According to the method, fault identification based on the multi-dimensional fault characterization model is adopted, comprehensive judgment is carried out in combination with physical failure analysis and a physical characterization model set, on the basis that the model set is established, the generated fault can be identified at low cost during subsequent fault diagnosis, and the fault diagnosis accuracy is improved. Therefore, the health management efficiency is improved, the fault-tolerant control strategy is accurately executed on the basis, and the engineering applicability is higher.

Description

Technical Field

[0001] The invention relates to a fault identification method and system for a non-similar redundant actuation system, belonging to the technical field of fault identification for non-similar redundant actuation systems. Background Art

[0002] The ultra-long flight time and harsh working conditions of near-space vehicles require them to have extremely high reliability, especially the reliability of the actuation system, which directly determines the success rate of the execution of related tasks. In order to further improve the reliability of the aircraft, the aircraft in the aerospace field have increasingly higher requirements for the fault tolerance of the control system, especially in the field of large civil aircraft and unmanned aerial vehicles with high reliability and ultra-long flight time. In order to ensure the success rate of flight mission execution under harsh working conditions, dissimilar redundant actuation technology is adopted. Electro-hydrostatic actuator (EHA) and electromechanical actuator (EMA) have the advantages of distributed and flexible layout, and can be further integrated into a new type of dissimilar redundant actuation system (NT-DRAS), so as to deal with common cause / common mode failures of the aircraft actuation system based on the fault-tolerant architecture design, while meeting the design requirements of high power-to-weight ratio, high reliability and high intelligence of the aircraft.

[0003] The new non-similar redundant actuation system has abundant redundancy resources and fault-tolerant potential, which makes it possible for aircraft to further improve the intelligence level and mission reliability based on fault-tolerant design. However, the premise for the smooth implementation of reconstruction fault tolerance is the accurate identification of the faults of this type of actuation system. Only on the premise of knowing the current fault status information of the new non-similar redundant actuation system can we formulate the corresponding offline health management strategy and the fault-tolerant processing strategy for the current fault. Therefore, the design of fault identification technology for the new non-similar redundant actuation system is of great significance.

[0004] Existing fault identification methods mainly include data-based, model-based, and expert system-based methods. Data-based diagnostic methods require real-time data and a large amount of historical data, and the data processing and calculation process are complex. In model-based methods, there are various modeling methods, and the established models include dynamic models and bond graph models, etc., which require complex calculations and processing when used for fault diagnosis. Similarly, expert systems need to establish a detailed expert database and require more sophisticated processing. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a method and system for fault identification of a non-similar redundant actuation system that can complete the identification of specific faults at a relatively low cost and has strong engineering applicability.

[0006] The technical solution of the present invention is a method for fault identification of a non-similar redundant actuation system, comprising the following steps:

[0007] The first step is to determine the state space model of the NT-DRAS non-similar redundant actuation system in the main working mode.

[0008]

[0009] Where u(t) is the model input, y(t) is the model output, and w(t) = F L is the external gust interference, x(t) is the state variable vector, A is the state matrix, B is the input matrix, C is the output matrix, and G is the interference matrix;

[0010] The second step is to determine the state space model after the fault is injected based on the state space model of the NT-DRAS non-similar redundant actuation system in the first step.

[0011] A2.1. Determine the fault characterization-related parameters, which include the total leakage coefficient C of the EHA actuator. ehal , EHA equivalent damping coefficient B eha 、The internal resistance of the motor R e and the bulk elastic modulus E of the EHA actuator eha , Fault-1 is the electrostatic hydraulic redundancy actuator leakage fault, and the total leakage coefficient C of the associated EHA actuator ehal , Fault-2 is the increase of actuation damping, and the equivalent damping coefficient B of the associated EHA eha , Fault-3 is the contamination of electrostatic fluid redundancy oil, and the volume elastic modulus E of the associated EHA actuator eha , Fault-3 is the increase of internal resistance, which is related to the internal resistance Re of the motor;

[0012] A2.2. Determine the state space model after the fault is injected based on the fault characterization correlation parameters.

[0013]

[0014] Among them, matrices ΔA and ΔB are characteristic matrices of model parameter uncertainty caused by system slow-varying faults;

[0015] The third step is to establish a multi-dimensional fault characterization model based on the fault correlation parameters determined in the second step and the state space model after the fault is injected.

[0016] A3.1. Determine a multi-dimensional fault characterization characteristic signal, which includes a displacement fault characterization characteristic signal x eha , speed fault characterization signal Pressure fault characterization signal P eha And the motor speed fault characterization characteristic signal ω eha ;

[0017] A3.2. Construct a mapping relationship between the fault mode Fault-i and the multi-dimensional fault characterization characteristic signal changes determined in step A3.1. Fault-1 is associated Fault-2 Correlation Fault-3 Correlation Fault-4 Correlation in Indicates that the pressure in the hydraulic cylinder drops. Indicates that the motor speed increases. Indicates that the system speed is decreasing, x eha ↓ indicates that the displacement of the system decreases, It indicates that the pressure in the hydraulic cylinder increases. Indicates that the pressure in the hydraulic actuator is unstable and vibrating. Indicates that the system speed has increased. Indicates that the motor speed decreases;

[0018] A3.3. Based on the fault mode and the state space model after the fault injection in step A2.2, a multi-dimensional fault characterization model of formula (6) is established.

[0019]

[0020] Where ΔA i , ΔB i The characteristic matrix of the multidimensional fault characterization model, with subscript i = 1, 2, 3, 4, corresponds to the fault mode;

[0021] The fourth step is to identify the fault of the NT-DRAS non-similar redundant actuation system with actual faults based on the multi-dimensional fault characterization model determined in the third step.

[0022] A4.1. Input the same detection signal to the NT-DRAS non-similar redundant actuation system with actual fault and the multi-dimensional fault characterization model determined in the third step, and obtain the actual output x of the NT-DRAS non-similar redundant actuation system with actual fault eha , P eha ,ω eha and the output of the multidimensional fault characterization model

[0023] A4.2. Calculate the actual output x eha , P eha ,ω eha And the model output The difference |Δx ehai |、 |ΔP ehai |,|Δω ehai |, and the output value error threshold Δx Model , ΔP Model , Δω Model By comparison, if the difference values ​​of a certain model all meet the output value error threshold, the fault mode corresponding to the model is determined;

[0024] A4.3. According to the fault mode determined in step A4.2, determine whether the change of the multi-dimensional fault characterization characteristic signal under the fault mode conforms to the mapping relationship between the fault mode Fault-i constructed in step A3.2 and the multi-dimensional fault characterization characteristic signal change determined in step A3.1. If so, determine that the actual fault is this fault mode.

[0025] A fault identification system for a non-similar redundant actuation system, characterized in that it comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements any step of the fault identification method when executing the computer program.

[0026] The beneficial effects of the present invention compared with the prior art are as follows:

[0027] (1) The present invention adopts fault identification based on a multi-dimensional fault characterization model, and combines physical failure analysis and a physical characterization model set for comprehensive judgment. On the basis of establishing a model set, the fault that occurs can be identified at a relatively low cost during subsequent fault diagnosis, thereby improving the efficiency of health management and accurately executing the fault-tolerant control strategy on this basis, with stronger engineering applicability.

[0028] (2) The characterization model of the present invention is based on the dynamic model of the actuator system. According to the fault type of the failure physical analysis, a corresponding characterization model containing fault factor parameters is established. After a specific detection signal is injected, the model output can be used to determine whether the corresponding fault occurs, which has the advantages of short-time and high efficiency.

[0029] (3) The multi-dimensional fault characterization model of the present invention is established based on the failure mechanism of the non-similar redundant actuation system, and the residual quantization threshold of the multi-dimensional signal for fault judgment is designed. Multiple conditions ensure the accuracy of fault identification;

[0030] (4) The present invention can obtain the fault status information of the current non-similar redundant actuation system by injecting the detection signal and running the model matching algorithm. It is fast and efficient, saves fault detection resources and time, and can efficiently provide support information for subsequent real-time fault tolerance and offline health management;

[0031] (5) The fault identification method of the present invention can realize the accurate identification of specific faults under fault conditions, thereby improving the efficiency of fault-tolerant control execution and the health management efficiency of the faulty system. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a flow chart of the present invention;

[0033] Figure 2 It is a schematic diagram of the composition and working principle of the new non-similar redundant actuation system;

[0034] Figure 3 A schematic diagram of a fault identification mechanism based on a multi-dimensional fault characterization signal of the present invention;

[0035] Figure 4 This is a schematic diagram of the principle of the fault identification model matching algorithm of the present invention;

[0036] Figure 5 This is a simulation result diagram of multi-dimensional state characteristic signal characterization under different fault conditions of the example of the present invention;

[0037] Figure 6 This is a diagram of a fault identification simulation result based on a multi-dimensional fault characterization model of an example of the present invention;

[0038] Figure 7 This is a simulation diagram of the characteristic model output error under the fault condition of the example of the present invention. DETAILED DESCRIPTION

[0039] In the field of high-value aircraft such as near-space aircraft that require high reliability and ultra-long flight time, the new non-similar redundant actuation system in the present invention has the advantages of high power-to-weight ratio and high reliability. The present invention integrates electrostatic hydraulic actuators (EHA) and electromechanical actuators (EMA) as basic redundancy into a new non-similar redundant actuation system (NT-DRAS). In order to smoothly implement fault-tolerant control strategies, a fault identification method and system based on a multi-dimensional fault characterization model are proposed.

[0040] Based on the explanation of the novel non-similar redundant actuation system architecture and its working mode, the present invention analyzes the occurrence mechanism of the preset fault for the main working mode, and then diagnoses the mode of the fault based on the multi-dimensional fault characterization model and identification method, and conducts example verification based on NT-DRAS to illustrate its usability.

[0041] The present invention is described in detail below with reference to specific examples and accompanying drawings.

[0042] The present invention Figure 1 As shown, a method for fault identification of a non-similar redundant actuation system is provided, comprising the following steps:

[0043] The first step is to determine the working mode of the NT-DRAS non-similar redundant actuation system and the basic mathematical model under normal working conditions. The NT-DRAS non-similar redundant actuation system is an electrostatic hydraulic actuator (EHA) and an electromechanical actuator (EMA) as the basic redundancy integration.

[0044] This step includes:

[0045] A1.1. Integrate the NT-DRAS non-similar redundant actuation system and ensure that it has a working mode.

[0046] like Figure 2 As shown in Figure 1, the composition and working principle diagram of the NT-DRAS non-similar redundant actuation system in this step are shown in Table 1. The NT-DRAS actuation system in this step can have three working modes. Working mode 1 (EHA active / EMA passive) is used as the main working mode, and working mode 2 (EHA passive / EMA active) and working mode 3 (EHA active / EMA active) are used as alternative working modes under specific working conditions to perform aircraft fault tolerance and expected performance response.

[0047] Table 1

[0048]

[0049] A1.2. Construct the mathematical model of the NT-DRAS non-similar redundant actuation system in the main working mode.

[0050] In this step, the NT-DRAS non-similar redundant actuation system is modeled using the state space method, and the state variable vector is selected as where x eha , P eha ,ω eha They are the output displacement, velocity, actuator pressure and motor speed of EHA in the main working mode respectively.

[0051] Based on the working principle of the NT-DRAS non-similar redundant actuation system, the differential equations satisfied by the above variables are in the form of:

[0052]

[0053] In the above formula, B eha , B ema , B dare the equivalent damping parameters of the EHA actuator piston, EMA transmission mechanism and control surface respectively; m eha 、m ema 、m d are the equivalent masses of the EHA actuator piston, EMA transmission mechanism and control surface respectively; A eha is the equivalent cross-sectional area of ​​the EHA actuator piston; F L is the gust disturbance acting on the rudder surface, V eha is the volume of the EHA actuator; E eha is the bulk elastic modulus of the EHA actuator; C ehal is the total leakage coefficient of the EHA actuator; V P is the output of the EHA pump; J m is the total inertia of the motor and pump of EHA; B me is the equivalent damping parameter of the motor; K m is the torque coefficient of the motor; R e is the internal resistance of the motor.

[0054] The state space model is in the form of

[0055]

[0056] Where u(t) is the model input, y(t) is the model output, and w(t) = F L For external gust interference, the state, input, output, and interference matrix of the model are as follows:

[0057]

[0058] The mathematical model in this step is a well-known technology in the art.

[0059] The second step is to determine the state space model after the fault is injected based on the mathematical model of the NT-DRAS non-similar redundant actuation system.

[0060] In this step, the relevant parameters in the mathematical model of the NT-DRAS non-similar redundant actuation system are subjected to fault identification analysis based on physical meaning and working mechanism, the fault characterization associated parameters are determined, and the state space model after the fault is injected is determined based on the fault characterization associated parameters.

[0061] The other working modes of the NT-DRAS non-similar redundant actuation system are usually switched to other working modes after a system failure or when the dynamic characteristics under high load are achieved. In the present invention, only the failure of the main working mode of the NT-DRAS non-similar redundant actuation system is analyzed, because the failure in this working mode is the first to occur and the most priority to be handled for the system.

[0062] Furthermore, this step specifically includes:

[0063] A2.1. Analyze the fault mode under working mode 1 and determine the fault characterization associated parameter as the total leakage coefficient C of the EHA actuator. ehal , EHA equivalent damping coefficient B eha 、The internal resistance of the motor R e and the bulk elastic modulus E of the EHA actuator eha .

[0064] In this working mode, the EHA is in active working mode and the EMA follows. Therefore, the EMA and the control surfaces can be regarded as the load of the EHA. Therefore, B in the model ema and B d can be regarded as a constant. Similarly, when there is no defect caused by structural damage to the aircraft, the weight of EHA, EMA and the control surface remains unchanged, that is, m eha 、m ema 、m d The volume and cross-sectional area of ​​the EHA hydraulic actuator remain unchanged in the absence of structural defects, that is, parameter A eha and V eha The output parameter V of the EHA pump is also kept constant. P Also designed as a constant. Total inertia of the motor and pump J m And the torque coefficient K of the motor m Also remains constant.

[0065] In addition to the parameters related to the above-mentioned structural solidification design, other parameters of the new non-similar redundant actuation system, as shown in Table 2, will cause parameter drift due to their respective different failure modes.

[0066] Table 2

[0067] Fault identification Physical meaning of failure mode related parameters Fault-1 <![CDATA[Electro-hydrostatic redundancy actuator leakage fault: leakage coefficient C ehal <!-- 5 -->]]> Fault-2 <![CDATA[Actuation damping increase: damping coefficient B eha > Fault-3 <![CDATA[Electro-hydrostatic redundancy oil contamination causes changes in elastic modulus: Elastic modulus E eha > Fault-4 <![CDATA[Increased internal resistance: The armature resistance R of the motor e >

[0068] The specific failure mechanism and change trend are as follows: With the increase of working time, friction and wear cause the leakage of EHA to gradually increase, so the total leakage coefficient C ehal will change; as the friction of EHA increases with working time, the damping increases, so the equivalent damping coefficient of EHA B eha will also change; because EHA oil will be contaminated by bubbles, excess wear, etc., which will eventually cause its elastic modulus correlation parameter E eha Changes; the EHA drive motor has heating problems, resulting in the internal resistance related parameter R e changes have occurred.

[0069] A2.2. Determine the state space model after the fault is injected based on the fault characterization correlation parameters.

[0070] The above fault mode is a slow-changing fault caused by the long-term operation of the new non-similar redundant actuation system, so the parameter changes caused by the fault are all within a certain range. Since the change of the fault parameter is unknown, the above fault is modeled as an uncertainty module during fault injection, as follows:

[0071]

[0072] The matrices ΔA and ΔB are the uncertainty modules of model parameters caused by slow-varying faults in the system. The above correlation matrix is ​​marked as follows:

[0073]

[0074] In the above formula, matrices A and B have been defined in the initial model of formula (3), a 22 、a 23 、a 32 、a 33 、a 34 、a 43 、a 44 、b 41 is the element in the matrix, Δa 22 =-ΔB eha / (m eha +m ema +m d ) is the equivalent damping fault factor; Δa 32 =-4ΔE eha A eha / V eha and Δa 34 =4ΔE eha V P / V eha are elastic modulus failure factors; Δa 33 =-[4(ΔE eha )·(ΔC ehal )] / V eha is the comprehensive failure factor of elastic modulus / total leakage coefficient; Δb 41 =K m / (J m ΔR e ) is the motor internal resistance fault factor, ΔB eha is the drift of the equivalent damping parameter of the EHA actuator piston due to the fault, ΔE eha is the drift of the EHA actuator bulk elastic modulus due to the fault, ΔC ehal is the drift of the total leakage coefficient of the EHA actuator due to the fault, ΔR e It is the drift of the motor internal resistance due to fault.

[0075] The third step is to establish a multidimensional fault characterization model based on the fault association parameters determined in the second step.

[0076] In this step, fault analysis is performed based on the working failure mechanism of the NT-DRAS non-similar redundant actuation system, and a set of fault characterization models (multidimensional fault characterization model) is established based on fault correlation parameters.

[0077] Furthermore, this step specifically includes:

[0078] A3.1. Determine a multi-dimensional fault characterization characteristic signal, which includes a displacement fault characterization characteristic signal, a speed fault characterization characteristic signal, a pressure fault characterization characteristic signal, and a motor speed fault characterization characteristic signal. The displacement fault characterization characteristic signal x eha ↓ Characterizes the decline of effective output displacement performance at the system level; characteristic signal of speed fault characterization Characterizes the degradation of effective output speed performance at the system level; Characterizes characteristic signals of pressure faults and Respectively represent the rise and fall of pressure in the EHA actuator, Characterizes the unstable sudden jitter of the pressure in the actuator; Characterizes the characteristic signal of motor speed failure and Respectively represent the increase and decrease of the EHA motor speed.

[0079] In this step, the fault reasoning and analysis of the NT-DRAS non-similar redundant actuation system are carried out, and its output x eha , P eha ,ω eha Under fault conditions, they all have a certain degree of change trend and representation. For specific fault conditions, the comprehensive information of multiple representation signals can be used to judge the current fault mode of the NT-DRAS non-similar redundant actuation system. The specific representation rules are as follows: Figure 3 As shown, the different symptoms caused by the failure are listed as follows:

[0080] (a) Multi-dimensional fault characterization characteristic signal 1 (displacement characterization): x eha ↓Represents the degradation of effective output displacement performance at the system level;

[0081] (b) Multi-dimensional fault characterization feature signal 2 (speed characterization): Characterizes the degradation of effective output speed performance at the system level;

[0082] (c) Multi-dimensional fault characterization characteristic signal 3 (pressure characterization): and Respectively represent the rise and fall of pressure in the EHA actuator; Characterizes the unstable sudden jitter of the pressure in the actuator;

[0083] (d) Multi-dimensional fault characterization characteristic signal 4 (motor speed characterization): and Respectively represent the increase and decrease of the EHA motor speed.

[0084] A3.2. Construct a mapping relationship between the fault mode Fault-i and the multi-dimensional fault characterization characteristic signal changes determined in step A3.1. Fault-1 is associated Fault-2 Correlation Fault-3 Correlation Fault-4 Correlation in Indicates that the pressure in the hydraulic cylinder drops. Indicates that the motor speed increases. Indicates that the system speed is decreasing, x eha ↓ indicates that the displacement of the system decreases, It indicates that the pressure in the hydraulic cylinder increases. Indicates that the pressure in the hydraulic actuator is unstable and vibrating. Indicates that the system speed has increased. Indicates that the motor speed is decreasing.

[0085] The details are as follows:

[0086] When Fault-1 occurs, leakage of the hydraulic cylinder will cause a significant drop in the pressure inside the cylinder. To maintain the pressure at the desired level, the motor speed will increase Used to compensate for the drop in pressure in the actuator. Due to the occurrence of the fault, the increase in motor speed cannot completely compensate for the drop in pressure in the actuator, so the speed and displacement at the system level are still lower than before the fault.

[0087] When Fault-2 occurs, due to the increase in the system equivalent damping, additional reaction force is generated on the EHA hydraulic actuator piston, causing the pressure on the piston surface to increase significantly. Also in order to maintain the desired pressure, the motor speed will increase Generate positive pressure to overcome the additional reaction force caused by the increase in damping. Similarly, due to the occurrence of the damping increase fault, the increase in motor speed cannot completely overcome the additional damping torque, so the system-level speed and displacement are still lower than before the fault.

[0088] When Fault-3 occurs, the fluid characteristics show an unstable mutation due to the entry of air bubbles into the oil, which causes the pressure in the actuator to show an obvious unstable shaking state. Even if the motor speed increases The system output speed is also improved However, due to the unstable state of the contaminated fluid, the effective target displacement performance at the system level is still reduced (x eha ↓).

[0089] When Fault-4 occurs, the motor speed will drop significantly due to the increase of internal resistance of the motor. This directly causes the pressure in the actuator to drop. Correspondingly, the system-level output speed performance is reduced The displacement performance also decreases (x eha ↓).

[0090] A3.3. Establish a multi-dimensional fault characterization model of formula (6).

[0091] In this step, based on the fault analysis of the NT-DRAS non-similar redundant actuation system, a model set consisting of a multi-dimensional fault characterization model is determined, and its specific form is as follows:

[0092]

[0093] Where i = 1, 2, 3, 4, corresponding to the fault mode, the characteristic matrix and parameter form of the multidimensional fault characterization model are:

[0094] Model-1 for Fault-1:

[0095]

[0096] Model-2 for Fault-2

[0097]

[0098] Model-3 for Fault-3

[0099]

[0100] Model-4 for Fault-4

[0101]

[0102] The fourth step is to perform fault identification on the NT-DRAS non-similar redundant actuation system with actual faults based on the multi-dimensional fault characterization model determined in the third step.

[0103] In this step, based on the model matching between the multi-dimensional fault characterization model and the actual fault system, the fault threshold in the judgment condition is determined to complete the fault identification.

[0104] Specifically include:

[0105] A4.1. Input the same detection signal to the NT-DRAS non-similar redundant actuation system with actual fault and the multi-dimensional fault characterization model determined in the third step, and obtain the actual output x of the NT-DRAS non-similar redundant actuation system with actual fault eha , P eha ,ω eha and the output of the multidimensional fault characterization model

[0106] In this step, the detection signal is a step instruction with a given fixed amplitude.

[0107] A4.2. Calculate the actual output x eha , P eha ,ω eha And the model output The difference |Δx ehai |、 |ΔP ehai |,|Δω ehai |, and the output value error threshold Δx Model , ΔP Model , Δω Model By comparison, if the difference values ​​of a certain model meet the output value error threshold, the fault mode corresponding to the model is determined.

[0108] Furthermore, in this step, the output value error threshold standard is met when the output value is less than or equal to the output value error threshold.

[0109] Furthermore, in this step, the output value error threshold is determined based on the system performance index. The smaller the output value error threshold is, the higher the system accuracy requirement is.

[0110] Furthermore, in this step, the calculation of the difference between the actual output and the model output is determined according to the actual situation, and can be an average value or an integral error.

[0111] A specific matching example of the multi-dimensional fault characterization model is as follows: Figure 4 As shown, x eha , P eha ,ω eha is the actual output of the NT-DRAS non-similar redundant actuation system, is the output of the multi-dimensional fault characterization model determined in the third step, Δx Model, ΔP Model , Δω Model are the set thresholds for the four output quantity errors between the actual system and the characterization model.

[0112] When a multi-dimensional fault characterization model is used for fault detection, specific detection instructions are sent to the NT-DRAS non-similar redundant actuation system and the multi-dimensional fault characterization model, so as to compare and match the model based on the output of the system and the model. Figure 4 The integral error algorithm structure of the selected time interval, when the four output error quantities simultaneously meet the set threshold of a certain model, the state of the current NT-DRAS non-similar redundant actuation system completes the match with a certain specific characterization model, thereby determining the failure mode of the current NT-DRAS non-similar redundant actuation system.

[0113] A4.3. According to the fault mode determined in step A4.2, determine whether the change of the multi-dimensional fault characterization characteristic signal under the fault mode conforms to the mapping relationship between the fault mode Fault-i constructed in step A3.2 and the multi-dimensional fault characterization characteristic signal change determined in step A3.1. If so, determine that the actual fault is this fault mode.

[0114] Furthermore, the present invention also provides a fault identification system for a non-similar redundant actuation system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the steps of the above-mentioned fault identification method are implemented when the processor executes the computer program.

[0115] In the present invention, fault identification is simulated and verified. Under normal working conditions, the parameter matrix of the NT-DRAS non-similar redundant actuation system is as follows:

[0116]

[0117] The parameter matrix under normal working conditions is injected with a bounded form of uncertainty, and the selected fault parameter characterization range is shown in Table 3. The above faults are injected respectively to verify whether the fault analysis of the present invention is correct. The specific verification results are shown in Figure 5 As shown in the figure: Under the four fault conditions, the fault response of the NT-DRAS non-similar redundant actuation system is consistent with the fault analysis results, so it can be determined that the fault analysis is correct.

[0118] Table 3

[0119]

[0120] For the unknown fault state of the NT-DRAS non-similar redundant actuation system, based on the fault detection matching in the fourth step: the fault response curve of the NT-DRAS non-similar redundant actuation system after injecting the detection command (a given fixed amplitude step command) is as follows: Figure 6 As shown, the error threshold of each characterization quantity is used for judgment, and the error response curve is shown in Figure 7 The specific error quantization results are shown in Table 4. The error quantization values ​​of the multidimensional fault characterization model 1 all meet the set threshold at the same time. It is further judged that the change trend of the multidimensional fault characterization signal is consistent with Fault-1. Therefore, it can be determined that the fault mode of the current new non-similar redundant actuation system is Fault-1.

[0121] Table 4

[0122]

[0123] Parts of the present invention that are not described in detail are well known to those skilled in the art.

Claims

1. A fault identification method for a non-similar redundant actuation system, characterized in that: The following steps are involved: The first step is to determine the state space model of the NT-DRAS non-similar redundant actuation system in the main working mode. Where u(t) is the model input, y(t) is the model output, and w(t) = F L is the external gust interference, x(t) is the state variable vector, A is the state matrix, B is the input matrix, C is the output matrix, and G is the interference matrix; The second step is to determine the state space model after the fault is injected based on the state space model of the NT-DRAS non-similar redundant actuation system in the first step. A2.

1. Determine the fault characterization-related parameters, which include the total leakage coefficient C of the EHA actuator. ehal , EHA equivalent damping coefficient B eha 、The internal resistance of the motor R e and the bulk elastic modulus E of the EHA actuator eha , Fault mode Fault-1 is the leakage fault of the electrostatic hydraulic redundancy actuator, and the total leakage coefficient C of the associated EHA actuator ehal , Fault-2 is the increase of actuation damping, and the equivalent damping coefficient B of the associated EHA eha , Fault-3 is the contamination of electrostatic fluid redundancy oil, and the volume elastic modulus E of the associated EHA actuator eha , Fault-3 is the increase of internal resistance, the internal resistance R of the associated motor e ; A2.

2. Determine the state space model after the fault is injected based on the fault characterization correlation parameters. Among them, matrices ΔA and ΔB are characteristic matrices of model parameter uncertainty caused by system slow-varying faults; The third step is to establish a multi-dimensional fault characterization model based on the fault correlation parameters determined in the second step and the state space model after the fault is injected. A3.

1. Determine a multi-dimensional fault characterization characteristic signal, which includes a displacement fault characterization characteristic signal x eha , speed fault characterization signal Pressure fault characteristic signal P eha And the motor speed fault characterization characteristic signal ω eha ; A3.

2. Construct a mapping relationship between the fault mode Fault-i and the multi-dimensional fault characterization characteristic signal changes determined in step A3.

1. Fault-1 is associated Fault-2 Correlation Fault-3 Correlation Fault-4 Correlation in Indicates that the pressure in the hydraulic cylinder drops. Indicates that the motor speed increases. Indicates that the system speed is decreasing, x eha ↓ indicates that the displacement of the system decreases, It indicates that the pressure in the hydraulic cylinder increases. Indicates that the pressure in the hydraulic actuator is unstable and vibrating. Indicates that the system speed has increased. Indicates that the motor speed decreases; A3.

3. Based on the fault mode and the state space model after the fault injection in step A2.2, a multi-dimensional fault characterization model of formula (6) is established. Among them, ΔA i , ΔB i The characteristic matrix of the multidimensional fault characterization model, with subscript i = 1, 2, 3, 4, corresponds to the fault mode; The fourth step is to identify the fault of the NT-DRAS non-similar redundant actuation system with actual faults based on the multi-dimensional fault characterization model determined in the third step. A4.

1. Input the same detection signal to the NT-DRAS non-similar redundant actuation system with actual fault and the multi-dimensional fault characterization model determined in the third step, and obtain the actual output x of the NT-DRAS non-similar redundant actuation system with actual fault eha , P eha ,ω eha and the output of the multidimensional fault characterization model A4.

2. Calculate the actual output x eha , P eha ,ω eha And the model output The difference |Δx ehai |、 |ΔP ehai |,|Δω ehai |, and the output value error threshold Δx Model , ΔP Model , Δω Model By comparison, if the difference values ​​of a certain model all meet the output value error threshold, the fault mode corresponding to the model is determined; A4.

3. According to the fault mode determined in step A4.2, determine whether the change of the multi-dimensional fault characterization characteristic signal under the fault mode conforms to the mapping relationship between the fault mode Fault-i constructed in step A3.2 and the multi-dimensional fault characterization characteristic signal change determined in step A3.

1. If so, determine that the actual fault is this fault mode.

2. The fault identification method of a non-similar redundant actuation system according to claim 1, characterized in that: In step A2.2, ΔB=[0 0 0 Δb 41 ] T , Δa 22 =-ΔB eha / (m eha +m ema +m d ) is the equivalent damping fault factor, Δa 32 =-4ΔE eha A eha / V eha , Δa 34 =4ΔE eha V P / V eha is the elastic modulus failure factor, Δa 33 =-[4(ΔE eha )·(ΔC ehal )] / V eha is the comprehensive failure factor of elastic modulus / total leakage coefficient, Δb 41 =K m / (J m ΔR e ) is the motor internal resistance fault factor, ΔB eha is the drift of the equivalent damping parameter of the EHA actuator piston due to the fault, ΔE eha is the drift of the EHA actuator bulk elastic modulus due to the fault, ΔC ehal is the drift of the total leakage coefficient of the EHA actuator due to the fault, ΔR e It is the drift of the motor internal resistance due to fault.

3. The fault identification method of a non-similar redundant actuation system according to claim 2, characterized in that: The characteristic matrix and parameter form of the multidimensional fault characterization model in step A3.3 are as follows: Model-1 for Fault-1, Model-2 for Fault-2, Model-3 for Fault-3, Model-4 for Fault-4, 4. The fault identification method of a non-similar redundant actuation system according to claim 3 is characterized in that: The first step comprises, A1.

1. Determine the working modes of the NT-DRAS non-similar redundant actuation system, including working mode 1, working mode 2 and working mode 3, where working mode 1 is EHA active / EMA passive, working mode 2 is EHA passive / EMA active, and working mode 3 is EHA active / EMA active. Working mode 1 is the main working mode, and working mode 2 and working mode 3 are alternative working modes under specific working conditions for executing aircraft fault tolerance and expected performance response; A1.

2. Construct the state space model of the NT-DRAS non-similar redundant actuation system in the main working mode. Where u(t) is the model input, y(t) is the model output, and w(t) = F L is the external gust interference, x(t) is the state variable vector, A is the state matrix, B is the input matrix, C is the output matrix, and G is the interference matrix.

5. The fault identification method of a non-similar redundant actuation system according to claim 4, characterized in that: The state variable vector in step A1.2 where x eha , P eha ,ω eha They are the output displacement, velocity, actuator pressure and motor speed of EHA in the main working mode respectively.

6. The fault identification method of a non-similar redundant actuation system according to claim 5, characterized in that: In step A1.2 Among them, B eha , B ema , B d are the equivalent damping parameters of the EHA actuator piston, EMA transmission mechanism and control surface respectively; m eha 、m ema 、m d are the equivalent masses of the EHA actuator piston, EMA transmission mechanism and control surface respectively; A eha is the equivalent cross-sectional area of ​​the EHA actuator piston; F L is the gust disturbance acting on the rudder surface, V eha is the volume of the EHA actuator; E eha is the bulk elastic modulus of the EHA actuator; C ehal is the total leakage coefficient of the EHA actuator; V P is the output of the EHA pump; J m is the total inertia of the motor and pump of EHA; B me is the equivalent damping parameter of the motor; K m is the torque coefficient of the motor; R e is the internal resistance of the motor.

7. A fault identification method for a non-similar redundant actuation system according to claim 6, characterized in that: In step A4.2, the output value error threshold standard is met when the output value error threshold is less than or equal to the output value error threshold.

8. The fault identification method of a non-similar redundant actuation system according to claim 6, characterized in that: The output value error threshold in step A4.2 is determined based on system performance indicators.

9. The fault identification method of a non-similar redundant actuation system according to claim 6, characterized in that: In the present step A4.1, the detection signal is a step instruction with a given fixed amplitude.

10. A fault identification system for a non-similar redundant actuation system, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the fault identification method according to any one of claims 1 to 9 are implemented.

Images