Active fault diagnosis method for autonomous underwater robot propeller

Abstract

The invention discloses an active fault diagnosis method for a propeller of an autonomous underwater robot. The method comprises the following steps: step 1, constructing a kinetic model of the autonomous underwater robot, a corresponding nonlinear model and a propeller fault model; 2, establishing a plurality of local linear models of the nonlinear model under an IF-THEN rule according to the nonlinear model and the propeller fault model; 3, constructing a T-S fuzzy model according to the local linear model, a fuzzy rule and a redundant actuator; 4, taking the unit step signal as an auxiliary signal, designing a corresponding auxiliary signal distributor, and constructing a ToMFIR residual error according to the T-S fuzzy model and the auxiliary signal distributor; and 5, setting a fault threshold, and diagnosing the fault through the constructed residual error. According to the method, the redundant actuator and the auxiliary signal distributor are introduced in the ToMFIR residual construction process, auxiliary signal excitation of the ToMFIR residual is added when the fault occurs, and detection of the early fault is facilitated.

Description

Technical Field

[0001] The invention relates to the technical field of active fault diagnosis of thrusters, and in particular to an active fault diagnosis method for thrusters of an autonomous underwater robot. Background Art

[0002] As an important engineering equipment for marine resource exploration and operation, AUV (Autonomous Underwater Vehicle) has the ability to autonomously perform tasks in deep-sea environments. However, due to the complex working environment and harsh mission intensity in the deep sea, the propulsion system that powers the AUV and maintains its attitude is often prone to failure, thus affecting the performance and safety of the AUV. Therefore, in the early stages of AUV propulsion system failures, fault detection and diagnosis has become a hot topic in the field of AUV system safety and reliability research in recent years.

[0003] ToMFIR (Total Measurable Fault Information Residual) residual is a fault detection and diagnosis technology for closed-loop control systems that has emerged in recent years. It is mainly used to solve the problem of false alarms or missed alarms caused by observer residuals in model-based fault diagnosis schemes. At present, fault detection and diagnosis methods are mainly divided into passive and active types, among which ToMFIR residual belongs to a typical passive fault detection and diagnosis scheme. Due to the complex structure of AUV, the strong dynamic nonlinearity of the system, and the uncertainty brought by the deep-sea working environment, new challenges and problems have been brought to fault detection and diagnosis, especially in the early stage of AUV thruster system failure, the detection and diagnosis performance of early faults is not ideal. The existing technology has the following defects: the existing ToMFIR residual is carried out under a "passive" framework, that is, only the input and output data of the system are used to generate residual signals for indicating faults, and no additional input signals are applied to the system, so as to have no effect on the evolution law of the system. When the fault characteristics contained in the data are not obvious, the diagnosis effect is poor, especially for control systems under complex working conditions, the performance of early fault detection and diagnosis is difficult to improve. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides an active fault diagnosis method for an autonomous underwater robot thruster to solve the technical problem that the prior art has poor diagnostic effect, especially for control systems under complex working conditions, and the performance of early fault detection and diagnosis is difficult to improve.

[0005] The present invention provides an active fault diagnosis method for an autonomous underwater robot thruster, comprising the following steps:

[0006] Step 1: Construct the dynamic model of the autonomous underwater robot and the corresponding nonlinear model and thruster fault model;

[0007] Step 2: According to the nonlinear model and the thruster fault model, several local linear models of the nonlinear model under the IF-THEN rule are established;

[0008] Step 3: Construct the TS fuzzy model based on the local linear model, fuzzy rules and redundant actuators;

[0009] Step 4: Take the unit step signal as the auxiliary signal, design the corresponding auxiliary signal distributor, and construct the ToMFIR residual according to the TS fuzzy model and the auxiliary signal distributor;

[0010] Step 5: Set the fault threshold, obtain the real-time AUV thruster status, and diagnose the fault through the ToMFIR residual constructed in step 4.

[0011] Furthermore, in step 3, the redundant actuator is: B2=B1, and the state space equation of the true local linear model in the TS fuzzy model is:

[0012]

[0013] The state space equation of the regular local linear model in the TS fuzzy model is:

[0014]

[0015] In the formula, x(t), u * (t), y(t), d * (t) are the state vector, control input vector, system output vector and external disturbance vector of the real system after adding the auxiliary signal; x0(t), u0(t), y0(t) are the state vector, control input vector and system output vector of the regular system respectively; A i , C i are the input matrix and output matrix of the real system under TS fuzzy rules; [B f B2] is the input matrix of the real system augmented with redundant actuators.

[0016] Furthermore, in step 4, the auxiliary signal is specifically: a unit step signal with the same weight in all frequency ranges.

[0017] Further, in step 4, the auxiliary signal distributor is:

[0018] α=[α1 α2] T

[0019] In the formula, α1=-α2, both are m×1 dimensional filters, and satisfy [B1 B2]α=0, where B1 and B2 are both n×m dimensional.

[0020] Furthermore, the objective function of the auxiliary signal distributor is:

[0021] J=max||L -1 (P ry (s)λ|| rms

[0022]

[0023] Where P ry (s) is the transfer function from auxiliary input to residual; λ(P ry (s)) is the transfer function P ry The extreme point of (s); min is the minimum pole; Im(λ min ) and Re(λ min ) are the real and imaginary parts respectively.

[0024] Furthermore, in step 4, the ToMFIR residual is constructed, where, when the thruster is not faulty:

[0025] ToMFIR i =G 0i d

[0026] When there is a thruster failure:

[0027] ToMFIR i =G i u d -G 0i u+(G i -G 0i )(-α1γ)

[0028] In the formula, G 0i is the transfer function of the regular system under fuzzy rules; d is the input disturbance; G i is the transfer function of the real system under fuzzy rules; u is the input quantity in the real system; u d is the perturbed input quantity in the real system.

[0029] Furthermore, when the thruster fails, the auxiliary signal excitation in the ToMFIR residual needs to satisfy:

[0030] (G i -G 0i )(-α1γ)>0

[0031] Furthermore, in step 5, when actively locating a fault, the fault threshold is set as:

[0032]

[0033] Where, d j is the external disturbance component on a certain degree of freedom, j=1...6.

[0034] Beneficial effects of the present invention:

[0035] The present invention introduces the effects of redundant actuators and auxiliary signal distributors in the process of constructing the ToMFIR residual. The auxiliary input signal has no excitation effect on the ToMFIR residual of the fault-free system. When a system fault occurs, the ToMFIR residual with the auxiliary signal input increases the auxiliary signal excitation compared to the ToMFIR residual without the auxiliary signal input. Therefore, the auxiliary signal has an excitation effect on the ToMFIR residual of the faulty system. At the same time, adjusting the auxiliary signal distributor can increase the excitation degree, which is beneficial to the detection of early faults. In the present invention, the redundant actuators and auxiliary signal distributors reduce the requirements of the auxiliary input signal for active fault detection. The present invention uses a simple unit step signal as the auxiliary input signal, and the design process is simple. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings, which are schematic and should not be construed as limiting the present invention in any way. In the accompanying drawings:

[0037] Figure 1 It is a schematic diagram of a flow chart of a specific embodiment of the present invention;

[0038] Figure 2 is a schematic diagram of an autonomous underwater robot in a coordinate system in a specific embodiment of the present invention;

[0039] Figure 3 This is the construction flow chart of the passive ToMFIR residual under fuzzy rules.

[0040] Figure 4 It is the construction flow chart of the passive global ToMFIR residual;

[0041] Figure 5 is a flow chart of the construction of the ToMFIR residual of the present invention;

[0042] Figure 6 This is a step flow of setting a threshold value in a specific embodiment of the present invention;

[0043] Figure 7 is a fault detection threshold determined according to the upper bound of the system disturbance in a specific embodiment of the present invention;

[0044] Figure 8It is a comparison of fault detection performance under fixed evolution rate and different fault amplitudes;

[0045] Fig. 9 It is a comparison of fault detection performance under fixed fault amplitude and different evolution rates. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0047] The present invention is further illustrated below in conjunction with specific embodiments. Those skilled in the art should understand that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention, and modifications to various equivalent forms of the present invention fall within the scope defined by the appended claims of this application.

[0048] like Figure 1 As shown, the present invention provides an active fault diagnosis method for an autonomous underwater robot thruster, comprising the following steps:

[0049] Step 1: Construct the dynamic model of the autonomous underwater robot and the corresponding nonlinear model and thruster fault model. The specific steps are as follows:

[0050] Step 11: Construct AUV dynamics model:

[0051] The coordinate diagram of the underwater robot is as follows: Figure 2 As shown in the figure, the underwater robot model used is the nonlinear dynamic model proposed by Fossen, which can be expressed in the moving coordinate system as follows:

[0052]

[0053] Where M = M RB +M A ; C(v)=C RB (v)+C A (v); M A is the additional mass generated by the underwater robot accelerating in water; C A (v) is the additional mass due to the Coriolis force-centrifugal force; D(v) is the damping matrix; g(η) is the restoring force and torque; η is the posture vector in the static coordinate system; τ is the control input of the actuator.

[0054] Assuming that the origin of the carrier coordinate system coincides with the center of gravity, the six-degree-of-freedom AUV motion equation with disturbances under buoyancy and hydrodynamics can be expressed as:

[0055]

[0056] In the formula, I x ,I y ,I z They represent the moment of inertia of the AUV about the x, y, and z axes in the carrier coordinate system; g ...N g Represents the force and moment generated by buoyancy and gravity; X w ...N w represents the force and torque generated by hydrodynamics; X d ...N d Represents the forces and moments generated by external disturbance forces.

[0057] The relevant parameters in AUV are defined as shown in Table 1:

[0058] Table 1

[0059] Step 12: Build the AUV Thruster Breakdown:

[0060] AUV generally uses thrusters as motion control components, and the thrust and torque generated by them can be expressed as:

[0061] T=[XYZ]

[0062] T'=[KMN]

[0063] The AUV system structure is complex and has certain nonlinearity. In addition, the complexity of its fault mechanism and the uncertainty of the working environment make fault diagnosis difficult.

[0064] The thrust generated by a thruster failure is less than the thrust generated when the failure does not occur. Therefore, the actual output of the thruster under the failure condition can be expressed as:

[0065] u f =u-Δu

[0066] Where u = [XYZKMN] T is the actual output of the thruster under normal conditions; Δu is the impact caused by thruster failure.

[0067] When considering the propeller failure of AUV, a parameter failure weight matrix can be introduced: θ = diag{θ1...θ m}, indicating that the AUV under study has m degrees of freedom, and θ1...θm ∈[0,1], that is, the failure of the thruster can be expressed as: Δu=θu.

[0068] When θ m = 0, it means that the thruster controlling the mth degree of freedom has no fault; when θ m = 1, indicating that the thruster controlling the mth degree of freedom is completely ineffective; when θ m ∈(0,1), it means that at least one of the thrusters controlling the mth degree of freedom has a certain degree of failure.

[0069] Since the actual thrust output of the thruster acts on the AUV after a failure, the control input τ under the thruster failure condition can be recalculated by combining the input matrix B in the state space equation. f :

[0070] τ f =Bu f =Bu-BΔu=B(I-θ)u

[0071] According to the above formula, in order to facilitate the derivation of active fault detection, in the specific embodiment of the present invention, B is used f It represents B(I-θ). Bθ is caused by the failure of the propeller. When the propeller is not faulty, B f =B, that is, it is considered that the failure of the thruster causes the input matrix B in the state space equation to become B f , the magnitude of the change is Bθ.

[0072] A scalar δ related to faults and disturbances is used as a fault scalar to measure the magnitude of the fault. The specific expression of δ is as follows:

[0073]

[0074] Assume that δ and

[0075] are the minimum and maximum values ​​of the early fault scalar, respectively. When a fault occurs, when there is , the fault in the system is an early fault.

[0076] When considering the changing trend of faults, the following expression is defined to distinguish between sudden early faults and slow early faults:

[0077]

[0078] In the formula, t f is the fault injection time; β is a scalar, indicating the fault change rate. When β is large, it is used to indicate a sudden fault; when β is small, it is used to indicate a slow fault. When the set θ satisfies When β is larger or smaller, it can be used to characterize sudden premature failure or slow premature failure, respectively.

[0079] Step 2: According to the nonlinear model and the thruster fault model, several local linear models of the nonlinear model under the IF-THEN rule are established. The specific steps are as follows:

[0080] The nonlinear model of the six-degree-of-freedom AUV can be written as follows:

[0081]

[0082] Where x(t), u(t), and y(t) are the system state vector, control input vector, and system output vector, respectively; g(x) is a nonlinear term; d(t) is the external disturbance of the system; the constant coefficient matrices A, B, C, and D can all be derived from the six-degree-of-freedom AUV motion equation, where the structures of x(t), u(t), and y(t) are as follows:

[0083] x(t)=y(t)=[uvwpqr] T

[0084] u(t)=[XYZKMN] T

[0085] From the moving coordinate system, we can know that the disturbance force of the near-water wave will act on the AUV body together with the force generated by the actuator. Compared with the force generated by the actuator, the disturbance force of the near-water wave is relatively small. In general, B and D are reversible. Let’s assume a matrix H that satisfies:

[0086] BH=D

[0087] Right now

[0088] Dd(t)=BHd(t)

[0089] Let d'(t) = Hd(t), we can get:

[0090] Dd(t)=Bd'(t)

[0091] After the conversion, the near-surface wave disturbance force can be approximately regarded as a system input disturbance, which is beneficial to the derivation of the active fault detection process after the above transformation.

[0092] The transformed six-DOF AUV nonlinear model can be written as follows:

[0093]

[0094] The local linear model of the real system of the six-degree-of-freedom AUV is:

[0095]

[0096] In the formula, A i , C i is a constant coefficient matrix; θ1(t), θ2(t), …, θ r (t) is the prior variable; is a fuzzy set; r is the number of fuzzy rules.

[0097] The corresponding global fuzzy system can be expressed as:

[0098]

[0099] In the formula, is the membership function; The following conditions must be met:

[0100]

[0101] Local linear model of the six-degree-of-freedom AUV canonical system:

[0102]

[0103] The corresponding global fuzzy system can be expressed as:

[0104]

[0105] Step 3: Construct the TS fuzzy model based on the local linear model, fuzzy rules and redundant actuators, as follows:

[0106] First, the TS fuzzy model is constructed by using the local linear model and fuzzy rules. In the process of constructing the TS fuzzy model, the redundant actuator is introduced: B2 = B1. After the introduction, the input matrix of the real local linear model is expanded to: [B f B2], where B f and B2 are both n×m dimensions. When there is no fault, θ=0, and the input matrix is: [B1 B2], so the input matrix of the regular local linear model is expanded to: [B1 B2], and the control variables are expanded to: The disturbance variable is expanded to: The input variables of the true local linear model are expanded to: The regular local linear model input variables are expanded to: Where 0 is an m×1 matrix vector.

[0107] The augmented true local linear model G i * and the regularized local linear model G 0i * The state space equations are:

[0108]

[0109] Step 4: Take the unit step signal as the auxiliary signal, design the corresponding auxiliary signal distributor, and construct the ToMFIR residual according to the TS fuzzy model and the auxiliary signal distributor, as follows:

[0110] A unit step signal with the same weight in all frequency ranges is selected as the auxiliary signal γ. The auxiliary signal γ must be introduced into the system through an auxiliary signal distributor. The corresponding auxiliary signal distributor is: α = [α1 α2] T , are all m×1 dimensional filters.

[0111] After the auxiliary signal distributor is introduced, the control variable G i * Input variables The output of the regularized local linear model is y0 * (t) = G 0i * u0 * (t), the output of the true local linear model is y * (t) = G i * u γ * (t), output residual r y * (t) = y * (t)-y0 * (t) = G i * u d * (t)-G 0i * u0 * (t), control quantity residual

[0112] In order to make sure that when there is no fault, u γ * (t) and r u * In (t), the auxiliary signal introduced by the auxiliary signal distributor can be offset, so it is necessary to satisfy [B1 B2]α=0, so α1=-α2.

[0113] The design of the auxiliary signal distributor is directly related to the impact of its output signal on the residual generation. In order to increase the impact of the auxiliary signal on the residual, the following objective function is designed:

[0114] J=max||L -1 (P ry (s)γ||rms

[0115]

[0116] Where P ry (s) is the transfer function between the auxiliary input and the residual, λ(P ry (s)) represents the transfer function P ry The extreme point of (s), λ min is the minimum pole, Im(λ min ) and Re(λ min ) are the real and imaginary parts respectively.

[0117] The specific P ry Substituting (s) and γ, we can obtain:

[0118] J=max||L -1 (G i * (s)-G 0i * (s))(-α1γ)|| rms

[0119]

[0120] like Figure 3-5 As shown, the global ToMFIR residual is constructed as follows:

[0121] Let the output of the regular local linear model be: y0(t) = G 0i u0(t); the output of the true local linear model is: y(t) = G i u d (t), where u d (t) = u(t) + d(t), then the output of the regular local linear model driven by the real local linear model input u(t) is: The output residual is recorded as: r y (t) = y(t) - y0(t) = G i u d (t)-G 0i u0(t), the control residual is: r u (t) = u(t) - u0(t).

[0122] Define fuzzy rule i:

[0123]

[0124] When the system has no faults:

[0125] ToMFIR i (t) = G 0i d(t)

[0126] Get ToMFIR under each fuzzy rule i (t), the global ToMFIR residual can be further constructed as follows:

[0127]

[0128] After introducing redundant actuators, auxiliary signals and auxiliary signal distributors, the logical mechanism of fault detection is described as follows:

[0129] It is known that the redundant actuator B2 = B1, the auxiliary signal distributor α1 = -α2,

[0130] When there is no thruster failure:

[0131]

[0132] When the thruster has no fault, the ToMFIR residual with auxiliary signal input is exactly the same as the ToMFIR residual without auxiliary signal input, so the auxiliary signal has no excitation effect on the ToMFIR residual of the normal system.

[0133] In the event of a thruster failure:

[0134]

[0135] When the thruster fails, the ToMFIR residual with auxiliary signal input is greater than the ToMFIR residual without auxiliary signal input. i -G 0i )(-α1γ), therefore, the auxiliary signal has an excitation effect on the ToMFIR residual of the faulty system. At the same time, adjusting the auxiliary signal distributor can increase the excitation degree, which is beneficial to the detection of early faults.

[0136] In the actual fault detection process, the auxiliary signal excitation must be positive excitation to increase the ToMFIR residual. Otherwise, it will cause the ToMFIR residual to decrease. Therefore, the auxiliary signal excitation must satisfy the following conditions:

[0137] (G i -G 0i )(-α1γ)>0

[0138] Right now

[0139] -θC i (sI-A i ) -1 B(-α1γ)>0

[0140] When considering thruster failure, the fault weight matrix θ is positive; C i (sI-Ai ) -1 B is for G 0i The transfer function should be positive; γ is the set unit step signal, which is positive; so when α1 is positive, (G i -G 0i )(-α1γ)>0.

[0141] Step 5: Set the fault threshold, obtain the real-time AUV thruster status, and diagnose the fault through the ToMFIR residual constructed in step 4, as shown in Figure 6 , 7 As shown, specifically:

[0142] In active fault detection, the six state variables of the six-freedom AUV can correspond to six ToMFIR residuals. If one of the ToMFIR residuals exceeds the threshold, it means that a fault has occurred. To simplify the processing process, the norm ||ToMFIR(t)|| can be taken as an indicator. It can be seen from the ToMFIR residuals when there is no fault that when the system does not fail, external disturbances are the main factor affecting the ToMFIR residuals. Therefore, the following thresholds are designed:

[0143]

[0144] In active fault location, in order to determine the location of the fault, it is necessary to design the following thresholds for the corresponding ToMFIR residuals on each degree of freedom:

[0145]

[0146] Where, d j is the external disturbance component on a certain degree of freedom, j=1...6.

[0147] The present invention is verified and its effect is demonstrated by simulation as follows:

[0148] Use MATLAB to write the TS fuzzy model program of the AUV nonlinear real system and the regular system, and add fault description; use MATLAB to write the AUV thruster active fault diagnosis program based on the combination of ToMFIR residual and auxiliary input signal design.

[0149] According to the calculation, the maximum value of θ is 0.1 and the minimum value is 0.02. The smaller the absolute value of θ is, the smaller the fault amplitude is.

[0150] When θ changes, Figure 8As shown in Table 2 below, the residual and warning time when θ changes from 0.02 to 0.1 and the fault evolution rate β = 0.004 are given. It can be seen that when the absolute value of θ used to characterize the fault amplitude is smaller, the alarm time will be correspondingly delayed. However, after adding the auxiliary signal, the warning time will be correspondingly advanced.

[0151]

[0152] Table 2

[0153] When the fault evolution rate β changes, such as Fig. 9 As shown in Table 3 below, the residuals and warning times when β ranges from 0.004 to 0.08 and θ = 0.05 are given. It can be seen that when the value of β is smaller, that is, the fault evolution trend is slower, the warning time will be correspondingly delayed. However, after adding the auxiliary signal, the warning time will be correspondingly advanced.

[0154]

[0155] Table 3

[0156] Therefore, for faults with smaller amplitudes and slower evolution trends, the method proposed in the present invention is more conducive to rapid and accurate detection and diagnosis of early faults.

[0157] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. An active fault diagnosis method for an autonomous underwater robot thruster, characterized in that: The steps include: Step 1: Construct the dynamic model of the autonomous underwater robot and the corresponding nonlinear model and thruster fault model; Step 2: According to the nonlinear model and the thruster fault model, several local linear models of the nonlinear model under the IF-THEN rule are established; Step 3: Construct the TS fuzzy model based on the local linear model, fuzzy rules and redundant actuators; Step 4: Take the unit step signal as the auxiliary signal, design the corresponding auxiliary signal distributor, and construct the ToMFIR residual according to the TS fuzzy model and the auxiliary signal distributor; Step 5: Set the fault threshold, obtain the real-time AUV thruster status, and diagnose the fault through the ToMFIR residual constructed in step 4.

2. The active fault diagnosis method for an autonomous underwater robot thruster according to claim 1, characterized in that: In step 3, the redundant actuator is: B2=B1, and the state space equation of the true local linear model in the TS fuzzy model is: The state space equation of the regular local linear model in the TS fuzzy model is: In the formula, x(t), u * (t), y(t), d * (t) are the state vector, control input vector, system output vector and external disturbance vector of the real system after adding the auxiliary signal; x0(t), u0(t), y0(t) are the state vector, control input vector and system output vector of the regular system respectively; A i , C i are the input matrix and output matrix of the real system under TS fuzzy rules; [B f B2] is the input matrix of the real system augmented with redundant actuators.

3. The active fault diagnosis method for an autonomous underwater robot thruster according to claim 2, characterized in that: In step 4, the auxiliary signal is specifically a unit step signal with the same weight in all frequency ranges.

4. The active fault diagnosis method for an autonomous underwater robot thruster according to claim 3, characterized in that: In step 4, the auxiliary signal distributor is: α＝[α1 α2] T In the formula, α1=-α2, both are m×1 dimensional filters, and satisfy [B1 B2]α=0, where B1 and B2 are both n×m dimensional.

5. The active fault diagnosis method for an autonomous underwater robot thruster according to claim 4, characterized in that: The objective function of the auxiliary signal distributor is: Where P ry (s) is the transfer function from auxiliary input to residual; λ(P ry (s)) is the transfer function P ry The extreme point of (s); min is the minimum pole; Im(λ min ) and Re(λ min ) are the real and imaginary parts respectively.

6. The active fault diagnosis method for an autonomous underwater robot thruster according to claim 4 or 5, characterized in that: In step 4, the ToMFIR residual is constructed, where the thruster is not faulty: <h2 style=";text-align:left;direction:ltr">ToMFIR<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> =G<h2 style=";text-align:left;direction:ltr"> 0i <h2 style=";text-align:left;direction:ltr"> d When there is a thruster failure: ToMFIR i =G i u d -G 0i u+(G i -G 0i )(-α1γ) In the formula, G 0i is the transfer function of the regular system under fuzzy rules; d is the input disturbance; G i is the transfer function of the real system under fuzzy rules; u is the input quantity in the real system; u d is the perturbed input quantity in the real system.

7. The active fault diagnosis method for an autonomous underwater robot thruster according to claim 6, characterized in that: When the thruster fails, the auxiliary signal excitation in the ToMFIR residual needs to meet the following requirements: (G i -G 0i )(-α1γ)>0.

8. The active fault diagnosis method for an autonomous underwater robot thruster according to claim 1, characterized in that: In step 5, when actively locating a fault, the fault threshold is set as: Where, d j is the external disturbance component on a certain degree of freedom, j=1...6.

Images