CN115903472A - AUV actuator fault diagnosis method based on fault factors and multiple observers - Google Patents
AUV actuator fault diagnosis method based on fault factors and multiple observers Download PDFInfo
- Publication number
- CN115903472A CN115903472A CN202211240991.1A CN202211240991A CN115903472A CN 115903472 A CN115903472 A CN 115903472A CN 202211240991 A CN202211240991 A CN 202211240991A CN 115903472 A CN115903472 A CN 115903472A
- Authority
- CN
- China
- Prior art keywords
- fault
- propeller
- control surface
- auv
- deformation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Feedback Control In General (AREA)
Abstract
The invention discloses an AUV executing mechanism fault diagnosis method based on fault factors and multiple observers, which mainly comprises the following steps: (1) Describing an execution mechanism of the AUV by using a fault factor, and modeling by using a state equation; (2) Designing a group of fault isolation Extended State Observers (ESOs), wherein each observer is sensitive to only one fault, and judging a generated residual error group through a certain logic rule to realize fault isolation of an actuating mechanism; (3) Fault identification is carried out on a control surface, a propeller and a steering engine in an actuating mechanism; (4) Considering deformation faults of the control surface, and analyzing deformation conditions of the control surface according to the positive and negative of fault factors; (5) And identifying and diagnosing the fault condition of the propeller by analyzing the internal information of the fault factor based on the output force and moment formulas of the propeller and the steering engine.
Description
Technical Field
The invention belongs to the technical field of fault diagnosis of an Underwater robot power system, and relates to an AUV (Autonomous Underwater Vehicle) execution mechanism fault diagnosis method based on fault factors and multiple observers.
Background
In recent years, the discovery and exploration of marine resources has been continuously developed and is becoming mainstream. Autonomous underwater robots (AUVs) play an important role in the marine field as tools for humans to explore and develop marine resources. Since the AUV needs to work in the sea for a long time, the conditions in the sea are extremely severe, the surrounding environment is complicated and variable, and various faults may occur in the actuator in the motion control system. If the fault is not detected in time, the AUV works in an unpredictable mode, so that the service life of the AUV is shortened, the underwater work task of the AUV is influenced, the safety of personnel and equipment is threatened, and finally disastrous results are brought. The actuator is the most common and important failure source of the AUV, and is used as a carrier for working in a complex marine environment, and the condition monitoring and failure diagnosis are the basic and core technologies for the research of AUV safety problems. Therefore, research on fault diagnosis technology of the AUV actuator to improve safety and reliability of the AUV has become an urgent research task, and is one of the hot spots of research in the scientific community at present.
The existing fault diagnosis technologies are generally divided into three categories: data-driven based methods, analytical model based methods, and knowledge based methods. The AUV fault diagnosis method based on the analytical model has a good diagnosis effect in a still water environment, is easy to realize real-time diagnosis, and directly provides useful information for fault-tolerant control or fault recovery of the next step. Unlike other methods, this method relies on an accurate mathematical model of the subject being diagnosed, with relevant parameters in the model being used for the methodological study. The fault factor and the multiple observers belong to a state estimation method in an analytical model, the state estimation method is a method for estimating the internal state of a dynamic system by using available measurement data, the data obtained by measuring the input and the output of the system can only reflect the external characteristics of the system, and the dynamic rule of the system needs to be described by using internal state variables. The state estimation method firstly reconstructs the state of the controlled process, generates a residual sequence by comparing with measurable variables, constructs a proper model and detects faults from the residual sequence by a statistical detection method, and further separates, estimates and makes decisions.
At present, the fault diagnosis technology based on the state estimation method has the following defects: 1) The state estimation method needs to linearize a nonlinear model part, but can not realize decoupling on an unstructured uncertain actual system; 2) The state estimation method needs a relatively accurate model, so that the calculated amount is relatively large, and the stability is relatively poor; 3) The parameter calculation of the state estimation method is more, and more parameters can influence the accuracy of fault diagnosis; 4) The thought of fault isolation and identification is not clear, and the two can not form a complete fault diagnosis mechanism; 5) All faults of a control surface, a steering engine and a propeller in the AUV actuating mechanism are difficult to identify by using a state estimation method, and the identification result has larger errors.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides a fault diagnosis method of an AUV (autonomous Underwater vehicle) actuating mechanism based on fault factors and multiple observers. The method adopts a plurality of extended state observers, so that the calculation amount of parameters is reduced, and the problem of system decoupling is avoided; each observer is sensitive to only one fault and is insensitive to other faults, so that the observer has stronger robustness; and the output formulas of the control surface, the steering engine and the propeller are used for further fault identification, so that the accuracy is high.
In order to solve the technical problems, the invention adopts the following technical scheme:
an AUV actuator fault diagnosis method based on fault factors and multiple observers comprises the following steps:
step 3, designing a plurality of extended state observers by using the scheme in the step 2, and judging residual errors generated by the plurality of extended state observers according to a set cooperation rule to realize fault isolation of a control surface, a steering engine and a propeller in the AUV actuating mechanism;
and 5, when the faults of the propeller and the rudder are isolated in the step 3, designing an output formula of the propeller and the rudder containing multiplicative fault description factors, and analyzing and identifying the fault conditions by analyzing the change of the multiplicative fault description factors.
Further, the specific content and method of step 1 include the following steps:
step 1.1 for a more accurate description of the failure, a description of the failure of the two AUV actuators is given here.
(1) Additive descriptive form:
u=u * +F + (1)
where u is the control input to the AUV actuator and u is * Is a desired input value; f + The fault vector in u is characterized,f + the factor is described for additive faults.
(2) Multiplicative descriptive form:
u=F × u * (2)
in the formula F × To characterize the diagonal matrix of fault amplitudes, defineWhen f is × When not equal to 1, the control input carries fault information, and f is called at the moment × The factors are described for multiplicative faults.
Step 1.2 models the concrete fault in the AUV execution mechanism, the method and the steps are as follows:
(1) Failure of the power propulsion actuator:
the propeller is an actuating mechanism in the AUV, which directly acts with the fluid to generate forward thrust. The relation between the output thrust of the introduced propeller and the rotating speed of the propeller is as follows:
T=K T n|n|+K v nV a (3)
where T is thrust, n is propeller speed, V a The motor propulsion speed; variables related to propeller failure include: coefficient of thrust K T Coefficient of propulsion of motor K v 。
(2) Fault of attitude control actuator:
the attitude control system of the AUV mainly comprises a control surface and a steering engine. The fault of the control surface is small-angle bending deformation of the edge, the deformed control surface acts with the fluid to generate an acting force parallel to the carrier coordinate system, and the additive fault generated by the action of the fluid is as follows:
in the formulaTo/>Additive faults for each degree of freedom; />And F * The fluid forces at failure and no failure, respectively. In an attitude control system, a steering engine is used for providing torque for deflection of a control surface, and a torque formula of the given steering engine is as follows:
M=DK M n|n|+DK v nV a (5)
wherein M is torque; d is the diameter of the propeller; variables related to steering engine failure include: coefficient of moment K M Coefficient of propulsion of motor K v 。
Aiming at a force and moment formula of a propeller and a steering engine, a multiplicative fault description factor f is introduced * And additive fault description factor f + The following can be obtained:
step 1.3 for the modeling and description of the AUV actuator system, consider a nonlinear system of the form:
wherein x, y, u, d are respectively a state vector, an output vector, an input vector and an interference vector; g (x) is a non-linear term; a, B, C and D are constant coefficient matrixes. Separating the change caused by each fault in the actuating mechanism from the state equation, and rewriting the state equation into a form containing fault description factors, so as to obtain:
wherein
Wherein G (-) is a piecewise fault function including additive faults, multiplicative faults and input variables carrying fault vectors; f = [ F = + F × ]=[F 1 F 2 … F n ]The fault distribution matrix is used, wherein the fault matrix can respectively represent the control surface deformation fault, the propeller thrust fault and the steering engine thrust moment fault.
Further, the specific content and method of step 2 include the following steps:
step 2.1 Using the equation of state in step 1.3, define F a =[F a + F a × ]=[F i F i+1 … F j ]And i is more than 1 and less than j and less than n. The residual fault vectors in G form a new matrix which is defined as F b And are combinedRemoving from G, forming a new distribution matrix with the disturbance D, containing only the fault matrix F a Is called G 1 ;
Step 2.2 rewrites the equation of state in step 1.3 to:
The structure is equivalent to the step of classifying a part of the actuator fault vectors as new interference vectorsThe aim of this design is to leave undisturbed the residual error generated by the observer>Is only affected by the new segment fault function G 1 The influence of (c). The new interference vector comprises interference D and partial fault F b So that the residual errors interfere and carry a fault F b Is insensitive to actuator failure G 1 And (4) sensitivity. According to the thought, the fault isolation observers with the same number as the fault types of the actuators can be designed theoretically.
Further, the specific content and method of step 3 include the following steps:
step 3.1 segment failure function G 1 Expanding other system variables into a new state variable x 2 The following can be obtained:
thus, the form of the extended state observer is designed as follows:
wherein i =1,2, \8230, n is an observer corresponding to the ith fault type; z is a radical of 1,i ,z 1,i Are respectively the state quantities x 1,i ,x 1,i Estimation of l 1,i ,l 1,i Is the gain vector of the extended state observer;
step 3.2 calculates the gain of the extended state observer, which can be generally parameterized as:
[l 1,i l 2,i ]=[β 1 ω 0 β 2 ω 0 2 ] (13)
wherein beta is 1 ,β 2 Is a parameter selected such that the characteristic polynomial beta 2 s+β 1 Is Hurwitz. Order to
s (n+1) +β 1 s n +…+β n s+β n+1 =(s+1) (n+1) (14)
It should ensure that the root of the characteristic polynomial is in the left half of the complex plane. Thus, the parameters in equation (6) may be selected as:
[l 1,i l 2,i ]=[2ω 0 ω 0 2 ] (15)
wherein omega 0 In practice, the bandwidth of the observer needs to be constantly changed to ensure that the ESO can correctly estimate the state variables.
Further, the specific content and method of step 4 include the following steps:
and 4.1, carrying out stress analysis on the fault deformation part of the control surface. The control surface in the actuator is regarded as a plane rectangle ABCD, and the angle of the control surface is delta X The area indicated by the dotted line is a part of the rudder where the deformation failure occursThe surface deformation fault is expressed as that a triangle ABC rotates downwards along a straight line AB by an angle which is called as a deformation angle delta f The angle produced on the plane is the fault plane angle r e . The axial fluid acting force under the fault and the axial fluid acting force under the non-fault are subtracted, and the fault acting force in each axial direction can be obtained. Therefore, the fault force and the moment generated by the action of the fluid on the deformed control surface ABC can be approximately calculated:
in the formula x e The vertical distance between the edge of the rudder surface and the z axis; p is the fluid pressure; l is the length of the deformation part of the control surface.
And 4.2, further analyzing the step 4.1, wherein when the control surface has different types of deformation faults, the faults are opposite in positive and negative. The positive and negative of the fault effect caused by all the control surfaces under the possible deformation condition are analyzed, and the following table shows that:
by combining the four kinds of fault force and moment, 8 kinds of basic deformation faults can be obtained. The method considers that only one control surface has faults at the same time, and the positive and negative characteristics of the deformation fault output action of different control surfaces have differences. Therefore, the identification of the fault control surface can be realized by combining the fault estimation result and the upper table.
Further, the specific content and method of step 5 include the following steps:
step 5.1 considers that two unknowns exist in the thrust formula, and the estimation data at one moment can not be used for solving. And the change of the fault degree of the propeller and the steering engine can be ignored in a short time, so two adjacent time periods t with changed parameters are taken 1 ,t 2 The equation (6) is converted into a matrix equation for solving multiplicative fault description factors, and the following data are obtained:
(1) Thrust of the propeller:
(2) Thrust torque of rudder:
step 5.2 utilizes the multiplicative fault matrix F of step 5.1 × (t) by discriminating the multiplicative fault description factor f within × (t) further identifying the propeller and rudder fault information as shown in the following table:
compared with the prior art, the invention has the following advantages and beneficial effects:
1. the invention divides the fault isolation and identification strategy into two stages by using the fault description factor, so that a complete fault diagnosis mechanism is formed. In practical application, a complete and logically correct fault diagnosis system can process faults, so that unnecessary troubles can be avoided in the process, and property loss is reduced.
2. In the multiple extended state observers designed by the invention, each observer is sensitive to a certain fault and is insensitive to other faults, so that the extended state observers have strong robustness. In practical application, when an AUV (autonomous underwater vehicle) breaks down, the fault isolation scheme can locate the fault in time, provide an alarm signal and avoid the accident; if the fault is judged by mistake and corresponding measures are not taken, accidents can happen, and great economic loss is caused.
3. The invention combines the fault description factor with the output formulas of the control surface, the steering engine and the propeller, and the method can identify the fault in the system according to the historical data of the operation in the system and can analyze the fault in the system on line in real time. In practical application, the type of the system fault can be identified in advance, the fault is prevented and processed in advance, and the operation cost is saved, so that the method has important theoretical value and practical significance.
4. The invention can completely identify all faults of the control surface, the steering engine and the propeller in the AUV actuating mechanism. In practical application, the damage degree of the propeller blades can be judged, the specific number of the lost blades is, and whether the degree of the damaged propeller blades needs to be replaced is reached; whether the deformation fault of the control surface is serious or not and the direction angle provided by the steering engine deviates by a certain angle due to the fault.
Drawings
Fig. 1 is a process diagram of a fault diagnosis method based on an analytical model.
Fig. 2 is a schematic view of an autonomous underwater robot in a coordinate system.
FIG. 3 is a flow chart of AUV actuator fault diagnosis based on description factors and multiple observers
FIG. 4 is a flow chart of steps of an embodiment of the present invention.
Fig. 5 is a diagram of an AUV actuator fault isolation scheme.
FIG. 6 is a flow chart of steps two and three of an embodiment of the present invention.
FIG. 7 is a step-by-step flow diagram of an embodiment of the present invention.
Fig. 8 is a schematic view of small angle deformation of the control surface.
FIG. 9 is a flow chart of steps five of an embodiment of the present invention.
FIG. 10a is a fault isolation diagram of the control surface.
Fig. 10b is a fault-recognition diagram of upward deformation of the left rudder.
Fig. 10c is a fault recognition diagram of the deformation of the upper rudder surface to the right.
Fig. 11a is a fault isolation diagram of a propeller.
Fig. 11b is an identification chart of three types of propeller failure.
FIG. 12a is a fault isolation diagram for a steering engine.
Fig. 12b is an identification diagram of three fault types of the steering engine.
Detailed Description
The technical solution of the present invention is further described in detail below with reference to the accompanying drawings.
Fig. 1 is a schematic diagram of an analytic model-based AUV actuator fault diagnosis method according to the present invention. And the controller controls the actuating mechanism to obtain a state value, expands the state observer to obtain an estimated value, finally obtains a residual error through comparison, and processes and decides fault information in a residual error sequence.
Fig. 3 is a flowchart of an embodiment of the method for diagnosing a fault of an AUV actuator based on fault factors and multiple observers, where the embodiment specifically includes the following steps:
in step 1, the detailed contents and design steps of the fault description of the control surface, the steering engine and the propeller of the AUV actuator (as shown in fig. 4) are as follows:
step 1.1 for a more accurate description of the failure, a description of the failure of the two AUV actuators is given here.
(1) Additive descriptive form:
u=u * +F + (19)
where u is the control input to the AUV actuator and u is * Is a desired input value; f + The fault vector in u is characterized and,f + the factor is described for additive faults.
(2) Multiplicative description form:
u=F × u * (20)
in the formula F × To characterize the diagonal matrix of the fault amplitude, defineWhen f is × When not equal to 1, the control input carries fault information, and f is called at the moment × The factors are described for multiplicative faults.
Step 1.2 models for specific faults in the actuator as follows:
(1) Failure of the power propulsion actuator:
the propeller is an actuating mechanism in the AUV, which directly acts with the fluid to generate forward thrust. The relation between the output thrust of the introduced propeller and the rotating speed of the propeller is as follows:
T=K T n|n|+K v nV a (21)
where T is thrust, n is propeller speed, V a The motor propulsion speed; variables related to propeller failure include: coefficient of thrust K T Coefficient of propulsion of motor K v 。
(2) Fault of attitude control actuator:
the attitude control system of the AUV mainly comprises a control surface and a steering engine. The fault of the control surface is small-angle bending deformation of the edge, the deformed control surface acts with the fluid to generate an acting force parallel to the carrier coordinate system, and the additive fault generated by the action of the fluid is as follows:
in the formulaTo>Additive faults for each degree of freedom; />And F * Fluid forces at fault and no fault, respectively. In an attitude control system, a steering engine is used for providing torque for deflection of a control surface, and a torque formula of the given steering engine is as follows:
M=DK M n|n|+DK v nV a (23)
wherein M is torque; d is the diameter of the propeller; variables related to steering engine failure include: coefficient of moment K M Coefficient of propulsion K of motor v 。
Aiming at a force and moment formula of a propeller and a steering engine, a multiplicative fault description factor f is introduced * And additive fault description factor f + The following can be obtained:
step 1.3 for the modeling and description of the AUV actuator system, consider a nonlinear system of the form:
wherein x, y, u, d are respectively a state vector, an output vector, an input vector and an interference vector; g (x) is a non-linear term; a, B, C and D are constant coefficient matrixes. Separating the change caused by each fault in the actuating mechanism from the state equation, and rewriting the state equation into a form containing fault description factors, so as to obtain:
wherein
Wherein G (-) is a piecewise fault function including additive faults, multiplicative faults and input variables carrying fault vectors; f = [ F = + F × ]=[F 1 F 2 … F n ]The fault distribution matrix is used, wherein the fault matrix can respectively represent a control surface deformation fault, a propeller thrust fault and a steering engine thrust moment fault.
in step 2, the specific content and design steps of the fault isolation scheme (as shown in fig. 6) are as follows:
step 2.1 Using the equation of state in step 1.3, define F a =[F a + F a × ]=[F i F i+1 … F j ]And i is more than 1 and less than j and less than n. The residual fault vectors in G form a new matrix which is defined as F b And removing from G and forming a new distribution matrix with the disturbance D, including only the fault matrix F a Is called G 1 ;
Step 2.2 rewrites the state equation in step 1.3 as:
The above structure is equivalent to classifying a part of the actuator fault vectors as new interference vectorsThe aim of this design is to leave undisturbed the residual error generated by the observer>Is only affected by the new segment fault function G 1 The influence of (c). Because the new interference vector contains interference D and partial fault F b So that the residual errors interfere and carry a fault F b Is insensitive to actuator failure G 1 And (4) sensitivity. According to the above thought, the fault isolation observers (as shown in fig. 5) with the same number as the fault types of the actuators can be designed theoretically.
Step 3, designing a plurality of extended state observers by using the scheme in the step 2, and judging residual errors generated by the plurality of extended state observers according to a set cooperation rule to realize fault isolation of a control surface, a steering engine and a propeller in the AUV actuating mechanism;
in step 3, the specific contents and design steps of the plurality of extended state observers designed based on the fault isolation scheme are as follows:
step 3.1 segment Fault function G 1 Expanding system variables other than the one to a new state variable x 2 The following can be obtained:
thus, the form of the extended state observer is designed as follows:
wherein i =1,2, \8230, n is an observer corresponding to the ith fault type; z is a radical of 1,i ,z 1,i Are respectively the state quantities x 1,i ,x 1,i Estimation of l 1,i ,l 1,i Is the gain vector of the extended state observer;
step 3.2 calculates the gain of the extended state observer, which can be parameterized generally as:
[l 1,i l 2,i ]=[β 1 ω 0 β 2 ω 0 2 ] (31)
wherein beta is 1 ,β 2 Is a parameter selected such that the characteristic polynomial beta 2 s+β 1 Is Hurwitz. Order to
s (n+1) +β 1 s n +…+β n s+β n+1 =(s+1) (n+1) (32)
It should ensure that the root of the characteristic polynomial is in the left half of the complex plane. Thus, the parameters in equation (6) may be selected as:
[l 1,i l 2,i ]=…2ω 0 ω 0 2 ] (33)
wherein omega 0 In practice, the bandwidth of the observer needs to be constantly changed to ensure that the ESO can correctly estimate the state variables.
in step 4, the small-angle deformation fault of the control surface is considered, and the deformation condition of the control surface is analyzed according to the positive and negative of the fault force and the moment in the control surface output formula. Specific content and design steps (as shown in fig. 7):
and 4.1, carrying out stress analysis on the fault deformation part of the control surface (as shown in figure 8). The control surface in the actuator is regarded as a plane rectangle ABCD, and the angle of the control surface is delta X The area indicated by the dotted line is the part of the control surface with deformation fault, and the deformation fault is indicated by that a triangle ABC rotates downwards along a straight line AB by an angle, which is called as a deformation angle delta f The angle produced on the plane is the fault plane angle r e . The axial fluid acting force under the fault and the axial fluid acting force under the non-fault are subtracted, and the fault acting force in each axial direction can be obtained. Therefore, the fault force and the moment generated by the action of the fluid on the deformed control surface ABC can be approximately calculated:
in the formula x e The vertical distance between the edge of the rudder surface and the z axis; p is the fluid pressure; l is the length of the deformation part of the control surface.
And 4.2, further analyzing the step 4.1, wherein when the control surface has different types of deformation faults, the faults are opposite in positive and negative. The positive and negative of the fault effect caused by all the control surfaces under the possible deformation condition are analyzed, and the following table shows that:
condition of deformation | f Y | f Z | f M | f N |
Upward deformation of left rudder | - | + | + | + |
Left rudder face |
0 | 0 | 0 | 1 |
Upward deformation of |
1 | 1 | 1 | 0 |
The right lateral rudder face deforms downwards | 1 | 0 | 0 | 0 |
The upper rudder surface deforms leftwards | 1 | 0 | 0 | 1 |
The upper rudder surface deforms to the right | 0 | 0 | 1 | 1 |
Left deformation of |
1 | 1 | 0 | 0 |
Lower rudder face is deformed to right | 0 | 1 | 1 | 0 |
By combining the four kinds of failure force and moment, 8 kinds of basic deformation failures can be obtained. The method considers that only one control surface has faults at the same time, and the positive and negative characteristics of the deformation fault output action of different control surfaces have differences. Therefore, the identification of the fault control surface can be realized by combining the fault estimation result and the upper table.
And 5, when the faults of the propeller and the rudder are isolated in the step 3, designing an output formula of the propeller and the rudder containing multiplicative fault description factors, and analyzing and identifying the fault conditions by analyzing the change of the multiplicative fault description factors.
In step 5, based on the output force and moment formulas of the propeller and the steering engine, the fault condition of the propeller and the steering engine is identified by analyzing the internal information of the fault factor. Specific content and design steps (as shown in fig. 9):
step 5.1 considers that two unknowns exist in the thrust formula, and the estimation data at one moment can not be used for solving. And the change of the fault degree of the propeller and the steering engine can be ignored in a short time, so two adjacent time periods t with the parameters changed are taken 1 ,t 2 The equation (6) is converted into a matrix equation for solving multiplicative fault description factors, and the following data are obtained:
(1) Thrust of the propeller:
(2) Thrust torque of rudder:
step 5.2 utilizes the multiplicative fault matrix F of step 5.1 × (t) by discriminating the multiplicative fault description factor f within × (t) further identifying the propeller and rudder fault information as shown in the following table:
the method of the invention is simulated and verified as follows:
a, writing a six-degree-of-freedom model of an autonomous underwater robot by using an MATLAB program, and designing three extended state observers simultaneously, wherein the three extended state observers respectively correspond to faults of a control surface, a steering engine and a propeller, so that fault isolation of an actuating mechanism is realized;
and B, compiling output formulas of the control plane, the steering engine and the propeller in the MATLAB, and identifying faults by analyzing respective fault description factors.
Fig. 10a is a diagram showing the effect of fault isolation on the control surface. It can be easily found from the figure that an observer designed based on a fault isolation scheme is sensitive to only one fault; therefore, synchronous cooperation of multiple observers can effectively isolate the AUV executing mechanism from faults.
Fig. 10b is a diagram for identifying upward deformation fault of the left rudder surface, and fig. 10c is a diagram for identifying rightward deformation fault of the upper rudder surface. The change condition of each curve in the diagram is observed, and the simulation result is matched with the designed fault type, so that the validity of the identification of the deformation fault of the control surface is verified.
Fig. 11a is a diagram of the effect of the propeller fault isolation, and it can be seen from the diagram that the multi-observation synchronous cooperation can effectively isolate the propeller fault.
Fig. 11b is a graph of the effect of identifying three types of propeller faults. As can be seen from the figure, the multiplicative fault description factor can further analyze the fault type of the propeller, so that the effect of fault identification is achieved.
Fig. 12a is a fault isolation effect diagram of the steering engine. As can be seen from the figure, the synchronous cooperation of multiple observations can effectively isolate the steering engine fault.
Fig. 12b is a diagram of the effect of identifying three types of faults of the steering engine. As can be seen from the figure, the multiplicative fault description factor also has a good effect on analyzing specific fault types inside the steering engine.
As can be seen from the attached drawings, the method can effectively realize fault isolation and identification of the execution mechanism of the autonomous underwater robot, effectively solves the problems of diagnosis of the execution mechanism fault and engineering application thereof, and has important significance for safe operation of the autonomous underwater robot.
The embodiments of the present invention have been described in detail with reference to the drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the gist of the present invention.
Claims (7)
1. An AUV actuator fault diagnosis method based on fault factors and multiple observers is characterized by comprising the following steps:
step 1, introducing multiplicative fault description factors and additive fault description factors, respectively describing faults of a control surface, a steering engine and a propeller in an AUV (autonomous underwater vehicle) actuating mechanism, introducing the fault description factors into a state equation, and modeling faults in the state equation;
step 2, dividing the modeled fault vector into two parts by using the state equation containing the fault of the actuating mechanism in the step 1, synthesizing one part of the fault vector and the interference vector into a new interference vector, and reserving the other part of the fault vector as a precondition of the step 3;
step 3, designing a plurality of extended state observers by using the scheme in the step 2, and judging residual errors generated by the plurality of extended state observers according to a set cooperation rule to realize fault isolation of a control surface, a steering engine and a propeller in the AUV actuating mechanism;
step 4, when the fault of the control surface is isolated in the step 3, the small angle deformation fault of the control surface is considered, a control surface output formula containing an additive fault description factor is designed, and the deformation fault condition of the control surface is analyzed and identified according to the positive and negative of the output force and the moment;
and 5, when the faults of the propeller and the rudder are isolated in the step 3, designing an output formula of the propeller and the rudder containing multiplicative fault description factors, and analyzing and identifying the fault conditions by analyzing the change of the multiplicative fault description factors.
2. The AUV actuator fault diagnosis method based on fault factors and multiple observers according to claim 1, wherein the two description forms of fault description of the control surface, the steering engine and the propeller in the AUV actuator in the step 1 are respectively as follows:
(1) Additive descriptive form:
u=u * +F + (1)
where u is the control input to the AUV actuator, u * Is a desired input value; f + The fault vector in u is characterized,f + is an additive fault description factor;
(2) Multiplicative description form:
u=F × u * (2)
3. The AUV actuator fault diagnosis method based on fault factors and multiple observers according to claim 1, wherein the specific contents and method steps for modeling faults in the state equation in the step 1 are as follows:
(1) Failure of the power propulsion actuator:
the propeller is an actuating mechanism which directly acts with fluid in the AUV to generate forward thrust, and the relation between the output thrust of the propeller and the rotating speed of the propeller is as follows:
T=K T n|n|+K v nV a (3)
where T is thrust, n is propeller speed, V a The motor propulsion speed; variables related to propeller failure include: coefficient of thrust K T Coefficient of propulsion of motor K v ;
(2) Fault of attitude control actuator:
AUV's attitude control system includes control surface and steering wheel, and the control surface trouble is marginal small-angle bending deformation, and deformation control surface and fluid take place the effect and produce the effort parallel with carrier coordinate system, and its additive trouble that receives the fluid effect to produce is:
in the formulaTo>Additive faults for each degree of freedom; />And F * The fluid forces at fault and without fault respectively,
in an attitude control system of an AUV (autonomous Underwater vehicle), a steering engine is used for providing torque for deflection of a control surface, and a torque formula of the given steering engine is as follows:
M=DK M n|n|+DK v nV a (5)
wherein M is torque; d is the diameter of the propeller; variables related to steering engine failure include: coefficient of moment K M Coefficient of propulsion of motor K v ,
Aiming at a force and moment formula of a propeller and a steering engine, a multiplicative fault description factor f is introduced * And additive fault description factor f + The following can be obtained:
for modeling and description of an AUV actuator system, consider a nonlinear system of the form:
wherein x, y, u, d are respectively a state vector, an output vector, an input vector and an interference vector; g (x) is a non-linear term; a, B, C and D are constant coefficient matrixes, changes caused by various faults in the AUV executing mechanism are separated from the state equation, and the state equation is rewritten into a form containing fault description factors, so that the following can be obtained:
wherein
Wherein G (-) is a piecewise fault function including additive faults, multiplicative faults and input variables carrying fault vectors; f = [ F = + F × ]=[F 1 F 2 ... F n ]The fault distribution matrix is used, wherein the fault matrix respectively represents a control surface deformation fault, a propeller thrust fault and a steering engine thrust moment fault.
4. The AUV actuator fault diagnosis method based on fault factors and multiple observers according to claim 1, wherein the specific content and method steps of the step 2 comprise:
step 2.1 Using the equation of state in claim 3, define F a =[F a + F a × ]=[F i F i+1 ... F j ]And 1 < i < j < n, and forming a new matrix by the residual fault vectors in G, wherein the new matrix is defined as F b And removing from G to form a new distribution matrix with interference D, including only the failure matrix F a Is called G 1 ;
Step 2.2 rewrites the state equation in claim 3 into:
The structure is equivalent to the step of classifying a part of fault vectors in the fault vectors of the AUV actuator as new interference vectorsThe aim of this design is to leave undisturbed the residual error generated by the observer>Is only affected by the new segment fault function G 1 Due to the new interference vector comprising the interference D and the partial fault F b Thus the residual error is right for the disturbance and carries the fault F b Is insensitive to actuator failure, and is responsive to actuator failure G 1 And sensitivity is realized, and fault isolation observers with the same number as the fault types of the actuators are designed according to the thought.
5. The AUV actuator fault diagnosis method based on fault factors and multiple observers according to claim 1, wherein the specific content and method steps of the step 3 comprise:
step 3.1 segment failure function G 1 Expanding other system variables into a new state variable x 2 The following can be obtained:
thus, the form of the extended state observer is designed as follows:
wherein i =1, 2.. And n is an observer corresponding to the ith fault type; z is a radical of 1,i ,z 1,i Are respectively the state quantities x 1,i ,x 1,i Estimation of l 1,i ,l 1,i A gain vector for the extended state observer;
step 3.2 calculates the gain of the extended state observer, which can be generally parameterized as:
[l 1,i l 2,i ]=[β 1 ω 0 β 2 ω 0 2 ] (13)
wherein beta is 1 ,β 2 Is a parameter selected such that the characteristic polynomial beta 2 s+β 1 Is Hurwitz, order
s (n+1) +β 1 s n +…+β n s+β n+1 =(s+1) (n+1) (14)
It should be guaranteed that the root of the feature polynomial is in the left half of the complex plane, and therefore the parameters in equation (6) are chosen as:
[l 1,i l 2,i ]=[2ω 0 ω 0 2 ] (15)
wherein omega 0 For the observer's bandwidth, it is necessary in practice to constantly change the bandwidth to ensure that the ESO can correctly estimate the stateAnd (4) variable quantity.
6. The AUV actuator fault diagnosis method based on fault factors and multiple observers according to claim 1, wherein the specific content and method steps of the step 4 comprise:
step 4.1, carrying out stress analysis on the fault deformation part of the control surface, regarding the control surface in the actuating mechanism as a plane rectangle ABCD, and setting the angle of the control surface as delta X The area of the control surface is part of the control surface with deformation fault, and the deformation fault is expressed as that a triangle ABC rotates downwards along a straight line AB by an angle called as a deformation angle delta f The angle produced on the plane is the fault plane angle r e And subtracting the axial fluid acting force under the fault and when the fault does not exist to obtain the fault acting force of each axial direction, and calculating the fault force and the moment generated by the action of the fluid on the deformation control surface ABC according to the fault acting force and the moment:
in the formula x e Is the vertical distance between the edge of the rudder surface and the z axis; p is the fluid pressure; l is the length of the deformation part of the control surface,
step 4.2, the step 4.1 is further analyzed, when the control surfaces have different types of deformation faults, the fault action has the condition of opposite positive and negative, and the positive and negative of the fault action caused by all the control surfaces under the deformation condition are analyzed, as follows:
by combining the four kinds of fault force and moment, 8 kinds of basic deformation faults are obtained, and the situation that only one control surface has faults at the same time and the difference exists between the positive and negative output effects of the deformation faults of different control surfaces is considered, so that the identification of the fault control surface is realized by combining the fault estimation result.
7. The AUV actuator fault diagnosis method based on fault factors and multiple observers according to claim 1, wherein the specific content and method steps of the step 5 comprise:
step 5.1, two unknowns exist in the thrust formula, the two unknowns cannot be solved only through estimation data at one moment, and the change of the fault degree of the propeller and the steering engine in a short time can be ignored, so two adjacent time periods t with parameters changing are taken 1 ,t 2 The equation (6) is converted into a matrix equation for solving multiplicative fault description factors, and the following data are obtained:
(1) Thrust of the propeller:
(2) Thrust torque of rudder:
step 5.2 utilizes the multiplicative fault matrix F of step 5.1 × (t) by discriminating multiplicative fault description factor f within × (t) further identifying the failure information of the propeller and the rudder as follows:
propeller fault identification condition:
f 1 × the value in (t) is less than 1: the thrust is reduced, namely the blade is aged and lost;
f 1 × the values in (t) are 0: the thrust disappears, i.e. it appears as a complete damage of the blade or insufficient control voltage;
f 1 × the values in (t) are not fixed and are all in the range of 0 to 1: the thrust is unstable, namely the blade is wound by foreign matters in water;
and (3) identifying the fault of the rudder:
f 2 × the value in (t) is less than 1: the moment is reduced, namely the locking of the rudder angle is shown;
f 2 × the values in (t) are 0: the moment disappears, namely the rudder angle fault is closed;
f 2 × the values in (t) are not fixed and are all in the range of 0 to 1: the moment is unstable, i.e. it appears as an irregular change in rudder angle.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211240991.1A CN115903472A (en) | 2022-10-11 | 2022-10-11 | AUV actuator fault diagnosis method based on fault factors and multiple observers |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211240991.1A CN115903472A (en) | 2022-10-11 | 2022-10-11 | AUV actuator fault diagnosis method based on fault factors and multiple observers |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115903472A true CN115903472A (en) | 2023-04-04 |
Family
ID=86473467
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211240991.1A Pending CN115903472A (en) | 2022-10-11 | 2022-10-11 | AUV actuator fault diagnosis method based on fault factors and multiple observers |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115903472A (en) |
-
2022
- 2022-10-11 CN CN202211240991.1A patent/CN115903472A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108594788B (en) | Airplane actuator fault detection and diagnosis method based on deep random forest algorithm | |
Zhang et al. | Robust observer-based fault diagnosis for nonlinear systems using MATLAB® | |
CN114813105B (en) | Gear box fault early warning method and system based on working condition similarity evaluation | |
JP5096352B2 (en) | A method for modeling the effects of failures in system behavior. | |
KR101021801B1 (en) | Actuator fault diagnosis of UAVs using adaptive unknown input observers | |
CN108170127A (en) | A kind of fault detection method of UAV Flight Control System | |
CN111460676B (en) | Method for evaluating health performance of multi-rotor aircraft under atmospheric turbulence disturbance | |
CN110989563A (en) | Unmanned naval vessel fault estimation method based on iterative adaptive observer | |
CN114035550B (en) | Autonomous underwater robot actuating mechanism fault diagnosis method based on ESO | |
CN115576184A (en) | Fault online diagnosis and fault-tolerant control method for underwater robot propeller | |
Qin et al. | Sensor fault diagnosis of autonomous underwater vehicle based on LSTM | |
CN116893643A (en) | Intelligent robot driving track safety control system based on data analysis | |
Weinstein et al. | Global aerodynamic modeling using automated local model networks in real time | |
CN114738205A (en) | Method, device, equipment and medium for monitoring state of floating fan foundation | |
CN108388229B (en) | Health degree-based four-rotor random hybrid system health assessment method | |
Hasan et al. | Predictive digital twins for autonomous ships | |
CN115903472A (en) | AUV actuator fault diagnosis method based on fault factors and multiple observers | |
CN111930094A (en) | Unmanned aerial vehicle actuator fault diagnosis method based on extended Kalman filtering | |
CN114037012B (en) | Flight data anomaly detection method based on correlation analysis and deep learning | |
CN106874531B (en) | Method for automatically recovering abnormal measurement value data of atmospheric data system in case of failure | |
WO2023101757A1 (en) | Multicopter online rotor fault diagnosis system | |
CN112990275B (en) | High-speed train running gear system fault diagnosis method based on semi-quantitative information fusion | |
Freeman et al. | Analytical fault detection for a small UAV | |
CN117250970B (en) | Method for realizing AUV fault detection based on model embedding generation countermeasure network | |
CN114217595B (en) | X-type rudder AUV fault detection method based on interval observer |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |