Unmanned rotorcraft power unit fault diagnosis method, system and unmanned rotorcraft
Through the fault diagnosis algorithm based on unknown observers, the problems of insufficient sensor measurement in the fault diagnosis of power unit of the unmanned rotor is solved, and the problem of difficulty in positioning under closed-loop control is achieved, efficient fault identification and positioning is achieved, and the safety of the drone is improved Sex and reliability.
Patent Information
- Application Number
- CN202411235100.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-04
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2044-09-04
AI Technical Summary
There is a lack of direct sensor measurement in the fault diagnosis of unmanned rotor power units, and it is difficult to directly locate the fault under closed-loop control, and the existing methods are difficult to apply to real-fly drones.
Using the power unit fault diagnosis algorithm based on unknown observers, dynamically screen the dynamic kinematic linear system state of the unmanned rotor aircraft, build an extended input observer, estimate the unknown throttle input and ideal speed in the equilibrium state, and combine the estimated error mutation rate to perform Fault identification and location.
The drone rotor failure detection rate is improved, and fault detection can be detected in time to avoid drone accidents and losses caused by failures.
Smart Images

Figure CN118732673B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the safety and reliability technology of an unmanned rotorcraft, in particular to an unmanned rotorcraft power unit fault diagnosis method and system and the unmanned rotorcraft. Background Art
[0002] As an emerging flying vehicle, unmanned rotorcraft has a wide range of applications in military, civil, agricultural and other fields, especially in the promotion and application research of cargo transportation and manned scenes, which will bring major changes to future urban transportation and greatly improve the efficiency of urban resources and population flow. However, with the large-scale application of unmanned rotorcraft, the probability of functional damage of airborne equipment increases, and accidents of unmanned rotorcraft become more frequent. According to relevant statistics, the failure rate of flight control system in unmanned rotorcraft is more than 95%. Among them, more than 45% of the failure events are caused by the power unit composed of actuators. The power unit of the flight control system of unmanned rotorcraft is equivalent to human limbs, which can adjust the flight attitude and motion parameters of the unmanned rotorcraft in real time according to the control instructions. The high efficiency and reliability of the power system directly affect the smooth operation of the unmanned rotorcraft. Therefore, it is crucial to carry out fault diagnosis for the power unit of the unmanned rotorcraft, specifically fault identification and location, and assist the subsequent fault-tolerant control decision-making for the safe operation of the unmanned rotorcraft.
[0003] The fault diagnosis technology of unmanned rotorcraft refers to the use of various measurement methods to perform data processing and quantitative analysis on the characteristic phenomena caused by the fault when the unmanned rotorcraft system fails, so as to determine the time and location of the fault, and provide a decision-making basis for subsequent effective control, maintenance and repair plans. In recent years, scholars and engineers at home and abroad have conducted a lot of research and technical accumulation on the fault diagnosis technology of the power unit of the flight control system of unmanned rotorcraft. However, the application of existing methods mainly meets two constraints: 1) Install suitable sensors at the shaft end of the unmanned rotorcraft, such as current sensors, voltage sensors and speed sensors, and directly detect single-axis faults through sensor data processing; 2) Existing theoretical methods often use the power unit under open-loop control conditions as the verification scenario. Its characteristic is that even if the sensor is not installed at the shaft end of the unmanned rotorcraft, the abnormal situation of the unmanned rotorcraft posture data can be used to identify and locate the fault. However, the above two constraints are difficult to achieve in the actual design of UAVs. On the one hand, due to cost and model structure optimization considerations, it is difficult to install sensors that meet the power unit fault diagnosis needs during the design and development phase of the unmanned rotorcraft; on the other hand, during the actual flight phase of the unmanned rotorcraft, the flight control system including the power unit is mainly based on closed-loop control. Even if the power unit fails, the drone's line position and attitude will not show obvious abnormalities due to the control module, and the sensor data will not show obvious abnormalities, making it difficult to directly locate the power unit from the closed-loop control end. Therefore, the existing power unit fault identification and positioning methods are difficult to apply to actual flight unmanned rotorcraft.
[0004] In view of the technical problems that the current fault diagnosis of UAV power systems lacks direct sensor measurement and the posture sensor data under closed-loop control is difficult to directly locate faults, the present invention provides effective technical support for fault identification and positioning of unmanned rotorcraft power units under closed-loop control. Summary of the invention
[0005] The present invention aims at the technical problems that the current fault diagnosis of unmanned rotorcraft power system lacks direct sensor measurement and the posture sensor data under closed-loop control is difficult to directly locate the fault. The present invention discloses a power unit fault diagnosis algorithm based on an unknown observer under closed-loop control conditions. The present invention aims at the technical problems that the current fault diagnosis of unmanned rotorcraft power system lacks direct sensor measurement and the posture sensor data under closed-loop control is difficult to directly locate the fault. Based on the minimum phase of the linear system and the observer matching principle, a method for dynamically screening the state of the unmanned rotorcraft power kinematic linear system is proposed; on this basis, an observer is constructed based on the extended input to estimate the unknown throttle input in the equilibrium state, and the ideal speed of each shaft is estimated as the basis for fault diagnosis; finally, combined with the ideal speed estimation and the real-time collection of the equivalent speed, a fault identification, location and classification method based on the estimated error mutation rate is proposed.
[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solutions:
[0007] A method for diagnosing a power unit fault of an unmanned rotorcraft under closed-loop conditions based on an unknown input observer, the method comprising the following steps:
[0008] S1: Screening of linear space state description for unmanned rotorcraft dynamic kinematics
[0009] Based on the kinematic and dynamic models of the unmanned rotorcraft, a suitable state set is selected to meet the two conditions for constructing an unknown input observer: the minimum phase system condition and the observer matching condition;
[0010] S2: Construction of an extended unknown input observer in linear space for unmanned rotorcraft dynamic kinematics; constructing an extended state vector containing the original state and unknown input by transforming the state equation ;
[0011] S3: Online estimation of ideal throttle control quantity for unmanned rotorcraft;
[0012] S301: According to the extended state vector Definition, unknown input vector ω(t) Calculate the original state vector x (t) :
[0013]
[0014] in, O is a zero matrix, I is the unit matrix;
[0015] Combined with real-time input sensor data, iteratively estimates the ideal throttle control law under normal conditions;
[0016] S302: based on the estimation of the ideal throttle control amount and the collection of the actual throttle control amount, obtaining the ideal throttle component and the equivalent throttle component according to the following calculation methods;
[0017] M=FU;
[0018] Among them, M represents the throttle component of each axis, U represents the total throttle, and F represents the throttle distribution matrix;
[0019] S303: Based on the throttle component, the motor speed of each axis is calculated using the throttle component according to a third-order polynomial relationship:
[0020] (twenty one)
[0021] Among them, a 0 , a 1 , a 2, a 3 are cubic polynomial coefficients, M i is the throttle component assigned to the i-th rotor axis;
[0022] S4: Unmanned rotorcraft power unit fault status determination
[0023] The equivalent speed of each axis is obtained through the actual throttle sensor data, and the ideal speed of each axis is obtained through the unknown observer speed in step S3. The two speeds are subtracted to obtain the estimated error; the motor position that first detects a large range of estimated errors is the fault position; when a fault point occurs, the one with the largest estimated difference is the fault.
[0024] Preferably, the typical description of the linear state equation considering the unknown input in step S1 is as shown in formula (1):
[0025]
[0026] in, Respectively represent the state variables, known input variables, unknown input variables and observed variables in the linear state space equation; the minimum phase condition and observer matching condition of the system are described in the form of equations (2) and (3):
[0027] 1) Minimum phase condition:
[0028]
[0029] 2) Observer matching conditions:
[0030] rank(CD)=rank(D)=q#(3)
[0031] Where n and q represent the number of states and unknown inputs in the linear system, respectively.
[0032] Preferably, the rotorcraft linear speed is selected in step S1 V E =[u,v,w] T , the Euler angular velocity of the body ω B =[p, q,r] T , rotorcraft line position X E =[x e ,y e ,h e ] T , and the Euler angles of the rotorcraft body η=[ϕ,θ,ψ]T as a candidate set of state spaces; where u,v,w They are respectively the x-axis speed of the rotorcraft ground coordinate system, the y-axis speed of the rotorcraft ground coordinate system, and the z-axis speed of the rotorcraft ground coordinate system , p, q, r They are the body pitch Euler angular velocity, body roll Euler angular velocity, and body yaw Euler angular velocity. ,x e ,y e ,h e They are respectively the x-axis position of the rotorcraft ground coordinate system, the y-axis position of the rotorcraft ground coordinate system, and the y-axis position of the rotorcraft ground coordinate system. , ϕ, θ, ψ They are the body pitch Euler angle, body roll Euler angle, and body yaw Euler angle.
[0033] Preferably, in step S2, according to the state space equation structure described in formula (1), the state vector is set x= [u,v,w,p,q,r] T , known input u=0 , unknown input ω=[dt,de,da,dr] T , observation quantity y=x=[u,v,w,p,q,r] T ,in, dt, de, da, dr They are altitude throttle, pitch throttle, roll throttle, and yaw throttle;
[0034] Construct an extended state vector containing the original state and unknown input by transforming the state equation , the description is shown in formula (4):
[0035]
[0036] in, x(t) and ω(t) are the state variables and unknown input variables in formula (1) respectively;
[0037] In order to adapt the final form of the constructed variables to still satisfy the general structure of the linear equation, the improved description of the output variables is shown in formula (5):
[0038]
[0039] The observation equation of the linear state space equation is rewritten as shown in equation (6):
[0040]
[0041] set up , Substitute into formula (6) and rewrite it into formula (7):
[0042]
[0043] set up , then further rewritten as:
[0044]
[0045] Through the above transformation, the extended state vector is introduced into the observation equation; on this basis, according to the extended state vector The original state vector x(t) is calculated as follows:
[0046]
[0047] Substituting into formula (10) we get:
[0048]
[0049] set up , formula (10) is rewritten as:
[0050]
[0051] The original linear state space (1) is rewritten as shown in formula (12):
[0052]
[0053] Set Existence , then the state transfer equation is rewritten as formula (13):
[0054]
[0055] The expanded linear state equation is expressed as formula (14):
[0056] .
[0057] Preferably, in step S2, according to the extended linear state equation description of formula (14), a suitable observer form is selected to satisfy the state tracking in the extended linear equation. Specifically, the extended state is tracked based on the Luenberger observer, and the observer form is shown in formula (15):
[0058]
[0059] The discretized expression of equation (15) is:
[0060]
[0061] Among them, z(k+1) is used as an intermediate variable and its initial value is set to the zero vector.
[0062] Preferably, in step S4, relative estimation error is used as a basis for subsequent fault location, and the calculation method is as follows:
[0063]
[0064] Among them, ΔΩ(k) is the speed estimation difference, Ω(k) and Represent equivalent speed and ideal speed respectively;
[0065] Based on the calculated ΔΩ(k), due to the motor fault, the speed estimation error will suddenly increase. The first position where it increases is the fault position. The fault occurrence event is calculated as follows:
[0066]
[0067] in, t f is the fault time, m is the number of rotorcraft axes, ΔΩ threshld Indicates the speed estimation error threshold, which is used to distinguish between the estimation error in a normal state and the estimation error in a fault state;
[0068] Then, the location of the fault is determined, and the fault is when the estimated difference at the fault point is the largest; this is obtained by the following calculation method:
[0069]
[0070] Among them, |ΔΩ i | represents the measurement of the estimation error of each shaft, and the maximum value position is the fault location;
[0071] Taking into account the increase of the rotor speed estimation error and the rotor speed estimation error increment when a fault occurs, the relative error weighted summation method is used for calculation; the calculation method is as follows:
[0072]
[0073] in, a The weights of the two errors are set based on subjective experience of experts or automatically optimized based on offline training data;
[0074] Finally, the difference between the estimated error and the ideal speed is calculated based on the fault location to establish the relationship between the estimated error and the fault degree. The calculation method is as follows:
[0075]
[0076] Among them, ΔΩ i represents the error, and P represents the fault degree of each motor.
[0077] Furthermore, the present invention also discloses a fault diagnosis system for an unmanned rotorcraft, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include means for executing the described method.
[0078] Furthermore, the present invention also discloses an unmanned rotorcraft, the unmanned rotorcraft comprising:
[0079] One or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the drone to execute the described method.
[0080] Furthermore, the present invention also discloses a computer-readable storage medium, comprising instructions, which, when executed on a drone, enable the drone to execute the method as claimed in any one of claims 1 to 6.
[0081] Furthermore, the present invention also discloses a computer program product, including a computer program or instructions, which implement the method when executed by a processor.
[0082] The beneficial effects of the present invention are as follows: The method for diagnosing the power unit fault of an unmanned rotorcraft under closed-loop conditions based on an unknown observer in the state estimation of the unmanned aircraft proposed by the present invention comprehensively considers the closed-loop control and sensor perception characteristics of the unmanned aircraft flight control system, and introduces an unknown input observer based on the soft measurement method on the basis of the original unmanned aircraft flight state monitoring, and indirectly measures the rotor speed, and determines the time of occurrence of the fault and the unknown by weighted modeling of the estimation error and the error increment. The method of the present invention can improve the detection rate of unmanned aircraft rotor faults, so that rotor faults can be discovered in time, and unmanned aircraft accidents and losses caused by rotor faults can be avoided. BRIEF DESCRIPTION OF THE DRAWINGS
[0083] Figure 1 This is a flow chart of the steps for locating the power unit fault of an unmanned rotorcraft under closed-loop conditions based on an unknown observer.
[0084] Figure 2 This is a structural diagram of the unmanned rotorcraft power unit fault location method under closed-loop conditions based on an unknown observer.
[0085] Figure 3 This is a framework diagram of the unmanned rotorcraft power unit fault location method under closed-loop conditions based on an unknown observer.
[0086] Figure 4Dynamic screening process for linear state space state vector of unmanned rotorcraft power unit.
[0087] Figure 5 This is an example of the implementation of the rotor speed estimation of a rotary-wing UAV under normal conditions.
[0088] Figure 6 This is a comparative example of rotor speed estimation under a single fault condition.
[0089] Figure 7 This is an example of rotor speed estimation error under a single fault condition.
[0090] Figure 8 This is an embodiment of the error increment of the rotor speed estimation under a single fault state.
[0091] Fig. 9 This is an example of a histogram of rotor speed estimation error indicators under a single fault condition.
[0092] Fig.10 A comparative example of rotor speed estimation under multiple fault conditions is provided.
[0093] Fig.11 This is an example of rotor speed estimation error under multiple fault conditions.
[0094] Fig.12 An embodiment of the error increment of rotor speed estimation under multiple fault conditions.
[0095] Fig.13 This is an example of a histogram of rotor speed estimation error indicators under multiple fault conditions.
[0096] Fig.14 Schematic diagram of the unmanned rotorcraft power unit fault location system under closed-loop conditions based on the unknown input observer. DETAILED DESCRIPTION
[0097] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.
[0098] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.
[0099] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".
[0100] The present invention aims to solve the technical problem of fault location of the power unit of the flight control system without direct sensor measurement. Therefore, the present invention introduces two concepts: ideal speed and equivalent speed. Before introducing in detail the method and device for locating the power unit fault of the unmanned rotorcraft under closed-loop conditions based on an unknown observer disclosed in the present invention, in order to clarify the concepts and definition scope involved in the patent. First, the two concepts of ideal speed and equivalent speed are clearly explained:
[0101] Ideal speed: The speed that the motors of each axis of the power unit of the unmanned rotorcraft should output when the unmanned rotorcraft reaches a balanced state, that is, when the balance point is controlled. In the present invention, the ideal speed can be, but is not limited to being approximately equivalent to the actual speed that the power unit of the unmanned rotorcraft can output when operating normally under the same balance requirements. In the present invention, unless otherwise specified, the ideal speed is defined as the symbol .
[0102] Equivalent speed: The speed of each shaft of the unmanned rotorcraft power unit cannot be directly measured without direct sensor measurement. To solve this problem, the speed can be estimated based on the actual control throttle amount through the physical model of the power system. When the drone is operating normally, the equivalent speed is approximately equal to the actual speed and the ideal speed. When the power unit of the unmanned rotorcraft fails, the actual process can be understood as the power unit being unable to output the speed at a fixed ratio, thereby affecting the lift and torque output; it can also be understood as being able to output the speed in proportion, but unable to output the lift and torque in proportion. The present invention is based on the second assumption, which regards the actual control throttle amount as being able to output the speed according to the established relationship, but the output lift and torque decrease. At this time, the equivalent speed is not necessarily equal to the actual speed. The above assumptions do not affect the basic logic of the method proposed in the present invention. In the present invention, unless otherwise specified, the ideal speed is defined as the symbol Ω.
[0103] Based on the ideal speed and the equivalent speed, the structure of the unmanned rotorcraft power unit fault location method under closed-loop conditions based on the unknown input observer disclosed in the present invention is as follows: Figure 1 The specific algorithm flow is as shown in Figure 2 As shown. Specifically, the power unit fault location method mainly includes the following parts: 1) observer matching based on dynamic screening of spatial states; 2) unknown input observer construction; 3) unknown input (throttle control law) estimation; 4) speed estimation; 5) fault identification and location. In order to clarify the specific operation process of each step, the following takes an 8-axis 16-propeller rotorcraft as an example to introduce in detail the implementation method of the unmanned rotorcraft power unit fault location method under closed-loop conditions based on an unknown input observer disclosed in the present invention.
[0104] 1. Observer matching based on dynamic screening of spatial states
[0105] The construction of the unknown input observer requires that the system satisfies the minimum phase system condition and the observer matching condition. It is not general, considering that the typical description of the linear state equation of the unknown input is shown in equation (1):
[0106]
[0107] in, Represent the state variables, known input variables, unknown input variables and observed variables in the linear state space equation respectively. Therefore, the minimum phase condition and observer matching condition of the system can be described in the form of equations (2) and (3):
[0108] 1) Minimum phase condition:
[0109]
[0110] 2) Observer matching conditions:
[0111] rank(CD)=rank(D)=q#(3)
[0112] in, and Represent the number of states and the number of unknown inputs in the linear equations. However, since the state space equations composed of the kinematic and dynamic models of the unmanned rotorcraft are usually nonlinear, and the combination of different states does not necessarily meet the above two conditions. Therefore, before constructing the unknown input observer, it is necessary to screen out a suitable state set based on the kinematic and dynamic models of the unmanned rotorcraft to meet the two conditions for constructing the unknown input observer. The spatial state dynamic screening process disclosed in the present invention is as follows: Figure 3According to the kinematic and dynamic model characteristics of the unmanned rotorcraft, the present invention selects the rotorcraft linear speed V E =[u,v,w] T , the Euler angular velocity of the body ω B =[p,q,r] T , rotorcraft line position X E = [x e ,y e ,h e ] T , and the Euler angles of the rotorcraft body η=[ϕ,θ,ψ] T as a candidate set of state spaces; where u,v,w They are respectively the x-axis speed of the rotorcraft ground coordinate system, the y-axis speed of the rotorcraft ground coordinate system, and the z-axis speed of the rotorcraft ground coordinate system , p, q, r They are the body pitch Euler angular velocity, body roll Euler angular velocity, and body yaw Euler angular velocity. ,x e ,y e , h e They are respectively the x-axis position of the rotorcraft ground coordinate system, the y-axis position of the rotorcraft ground coordinate system, and the y-axis position of the rotorcraft ground coordinate system. , ϕ, θ, ψ They are the body pitch Euler angle, body roll Euler angle, and body yaw Euler angle, according to Figure 3 The process shown in the figure can obtain the linear state space equation that meets the conditions.
[0113] 2. Construction of unknown input observer
[0114] The present invention uses the state vector selected in step 1 to construct the kinematic and dynamic linear space state equations of the unmanned rotorcraft and serves as a verification example of this step and the following steps. According to the state space equation structure described by formula (1), the state vector is set x=[u,v,w,p,q,r] T , known input u=0 , unknown input ω=[dt,de,da,dr] T , observation quantity y=x=[u, v,w,p,q,r] T , dt, de, da, dr They are altitude throttle, pitch throttle, roll throttle, and yaw throttle.
[0115] The core idea of constructing an unknown input observer by extending the state variable is to construct an extended state vector containing the original state and the unknown input by transforming the state equation. , the description is shown in formula (4):
[0116]
[0117] in, x(t) and ω(t) are the state variables and unknown input variables in equation (1) respectively. Correspondingly, in order to adapt to the final form of the constructed variables and still satisfy the general structure of the linear equation, the output variables can be improved as shown in equation (5):
[0118]
[0119] Therefore, the observation equation of the linear state space equation is rewritten as shown in equation (6):
[0120]
[0121] set up, , Substitute into formula (6) and rewrite it into formula (7):
[0122]
[0123] set up , then further rewritten as:
[0124]
[0125] Through the above transformation, the extended state vector is introduced into the observation equation. On this basis, according to the extended state vector The original state vector x(t) can be calculated as follows:
[0126]
[0127] Substituting into formula (10) we can get:
[0128]
[0129] set up , formula (10) can be rewritten as:
[0130]
[0131] Therefore, the original linear state space (1) can be rewritten as shown in formula (12):
[0132]
[0133] In order to further transform the left side of the state transfer matrix Offset, set existence , then the state transfer equation can be rewritten as formula (13):
[0134]
[0135] Therefore, the expanded linear state equation can be expressed as formula (14):
[0136]
[0137] According to the extended linear state equation description of formula (14), a suitable observer form can be selected to satisfy the state tracking in the extended linear equation. Without loss of generality, the implementation example of the present invention is based on the Luenberger observer to track the extended state, and the observer form is shown in formula (15):
[0138]
[0139] equation( )The discretized expression is:
[0140]
[0141] Among them, z(k+1) is used as an intermediate variable, and its initial value is set to a zero vector. It should be noted that in the implementation example of the present invention, according to the general properties of state tracking, the following relationship can be obtained:
[0142]
[0143] Combining the state transfer equations in equation (14) and equation (15), we can obtain equation (18):
[0144]
[0145] Therefore, in order to achieve stable convergence of the observer, it is necessary is a Hurwitz matrix that satisfies equation (19):
[0146]
[0147] After the above steps, the unknown input observer can be constructed when the linear state equation satisfies the observer matching condition.
[0148] (3) Estimation of unknown input (throttle control law)
[0149] According to formula (9), combined with real-time input sensor data, the ideal throttle control law under normal conditions is iteratively estimated. Since the constructed linear state equation is the model of the unmanned rotorcraft under normal conditions, and under closed-loop control conditions, the posture sensor data of the unmanned rotorcraft remains normal from the outside of the system, the estimated ideal throttle control amount can be approximately regarded as the throttle control under normal conditions, which provides a reference for the subsequent identification and positioning of the faults of each axis of the power unit.
[0150] (4) Speed estimation
[0151] Based on the estimation of the ideal throttle control amount and the acquisition of the actual throttle control amount, the ideal speed and the equivalent speed can be obtained respectively. The conversion basis can be shown in equations (20) and (21):
[0152] M=FU#(20)
[0153] First, according to the total throttle control amount, the throttle component of each rotating shaft can be calculated according to the power distribution principle. Among them, M represents the throttle component of each shaft, U represents the total throttle, and F represents the throttle distribution matrix. Secondly, based on the throttle component, the speed of each shaft is obtained according to the relationship between the throttle component and the motor speed of each shaft. The implementation case of the present invention is described by a third-order polynomial relationship, as shown in formula (21):
[0154] (twenty one)
[0155] Based on the above relationship, the ideal speed and equivalent speed can be obtained for the subsequent identification and location of shaft faults in the unmanned rotorcraft power unit.
[0156] (5) Unmanned rotorcraft power unit shaft fault identification and location
[0157] The identification and location of shaft faults in the power unit of unmanned rotorcraft mainly includes three steps: fault identification, fault location (isolation) and fault classification. Among them, the purpose of fault identification is to determine the time when the fault occurs, the fault location is mainly to determine the location of the shaft where the fault occurs, and the core task of fault classification is to determine the degree of the fault.
[0158] When a rotorcraft fails under closed-loop control conditions, the ratio between the motor throttle, motor speed and lift output at the fault location changes first, causing the overall linear displacement of the rotorcraft and the Euler angle of the fuselage to deviate from the control. Therefore, under the action of the closed loop, the control throttle changes, and the motor at the fault location has insufficient power output, so the speed will increase significantly after control. The speed of each motor also changes accordingly, but generally lags behind the change of the faulty motor, and multiple normal motors compensate for a few faulty motors, so the lifting and lowering adjustment range of normal motors is small. Based on this principle, the rotorcraft power unit fault location under closed-loop conditions proposed by the present invention is combined with the input estimation value obtained by the unknown input observer method, which mainly follows two principles: 1) The motor position that first detects a large range of estimation errors is the fault position; 2) When a fault point occurs, the one with the largest estimation difference is the fault.
[0159] Based on the above principles, the specific implementation plan is as follows:
[0160] First, according to step (4), the equivalent speed of each axis is obtained through the actual throttle sensor data, and the ideal speed of each axis is obtained through the unknown observer speed in step (3), and the two speeds are subtracted to obtain the estimated error. At the same time, in order to reduce the uncertainty of the absolute value of the speed caused by different flight requirements, the present invention uses the relative estimated error as the basis for subsequent fault location, and the calculation method is as follows:
[0161]
[0162] Among them, ΔΩ(k) is the speed estimation difference, Ω(k) and They represent equivalent speed and ideal speed respectively.
[0163] Secondly, based on the calculated ΔΩ(k), due to the occurrence of a motor fault, the speed estimation error will suddenly increase, and according to principle (1), the position where the error increases first is the position where the fault occurs. Therefore, the fault occurrence event can be calculated as follows:
[0164]
[0165] in, t f is the fault time, m is the number of rotorcraft axes, ΔΩ threshld Indicates the speed estimation error threshold, which is used to distinguish between the estimation error in a normal state and the estimation error in a fault state.
[0166] Then, determine the location where the fault occurs. According to principle (2), the fault occurs when the estimated difference at the fault point is the largest. Therefore, it can be obtained by the following calculation method:
[0167]
[0168] Among them, |ΔΩ i | represents the measurement of the estimation error of each axis, and the maximum value position is the fault location.
[0169] Considering that the rotor speed estimation error and the rotor speed estimation error increment will increase simultaneously when a fault occurs, the relative error weighted summation method is used for calculation. The calculation method is as follows:
[0170]
[0171] in, a The weights representing the two errors can be set based on expert experience or automatically optimized based on offline training data.
[0172] Finally, the difference between the estimated error and the ideal speed is calculated based on the fault location to establish the relationship between the estimated error and the fault degree. The calculation method is as follows:
[0173]
[0174] Among them, ΔΩ i represents the error, and P represents the fault degree of each motor.
[0175] Corresponding to the aforementioned embodiment of the method for locating a power unit fault of an unmanned rotorcraft under closed-loop conditions based on an unknown input observer, the present invention also provides an embodiment of a system for locating a power unit fault of an unmanned rotorcraft under closed-loop conditions based on an unknown input observer.
[0176] First, the embodiment verifies the estimation effect of the unknown input observer under closed-loop conditions. The experiment first estimates and tracks each state in the closed loop under the normal state of the drone. Although the position observer directly estimates the throttle control law, the ultimate goal of the observer is to estimate the throttle speed of each rotor, so the experimental effect is directly based on the estimated comparison of the rotor speed.
[0177] Rotor speed estimation example results Figure 5 As shown in Figure 2. After the UAV flight state reaches stability, the unknown input observer starts to estimate the rotation speed of each rotor from 20s. Figure 5 It can be seen that the rotation speed estimation of each rotor can accurately estimate the rotation speed trend within an acceptable range.
[0178] On the basis of verifying the effectiveness of the unknown input observer, the implementation example verifies the effectiveness of fault diagnosis under single rotor fault conditions. First, take the No. 3 rotor fault as an example. It is assumed that after the drone has been flying stably for 200 seconds, the rotor has an offset fault and the power output loss is 40%, that is, the power output ratio drops to 60% of the original power output. Under this condition, the comparison between the ideal speed estimation and the equivalent speed is as follows: Figure 6As shown in the figure, when the 220s single rotor fault is injected, the flight control system will increase the rotor motor speed to ensure the normal output of lift and torque in order to ensure the stable flight state of the drone.
[0179] Depend on Figure 7 It can be seen that when a fault occurs in the UAV, the estimated error of the rotor motor changes suddenly, so the time of the fault can be identified based on this principle. The result of the mutation speed calculation of the estimated error of the single fault rotor motor embodiment is as follows Figure 8 shown.
[0180] On the basis of fault identification, the fault can be located by the speed estimation deviation. The fault diagnosis algorithm disclosed in the present invention comprehensively considers that the rotor speed estimation deviation and the rotor speed estimation deviation increment are increased at the same time when the fault occurs, so the relative error weighted summation method is used for calculation. The calculation method is as follows:
[0181]
[0182] Among them, α represents the weight of the two errors, which can be set according to expert experience or subjective experience, or automatically optimized according to offline training data. In order to facilitate calculation, α=0.5 is selected, that is, the relative estimation error and the relative estimation error increment are equally important. Other settings of are easy for relevant practitioners to think of. As in the previous article, taking the No. 3 rotor fault as an example, the estimated error index bar graph is as follows Fig. 9 As shown in the figure, it can be seen that the error of the rotor speed calculated for rotor No. 3 is the largest, which can be used as the basis for the calculation of rotor No. 3.
[0183] On the basis of verifying the single-rotor fault diagnosis implementation example, the experiment was carried out to verify the effectiveness of fault diagnosis under multi-rotor fault conditions. Taking the No. 3 and No. 5 rotor faults as an example, it is assumed that after the drone has been flying stably for 200 seconds, the rotor has an offset fault and the power output loss is 40%, that is, the power output ratio drops to 60% of the original power output. Under this condition, the comparison between the ideal speed estimate and the equivalent speed is as follows: Fig.10 shown.
[0184] It can be seen from the figure that, similar to the single rotor fault embodiment, when the 220s multi-rotor fault is injected, the flight control system will increase the rotor motor speed to ensure the normal output of lift and torque to ensure the stable flight state of the drone. Fig.11 It can be seen that when a UAV fails, the estimated error of the rotor motor changes suddenly, so the time of the failure can be identified based on this principle. The result of calculating the mutation speed of the estimated error of a single-fault rotor motor is as follows: Fig.12 shown.
[0185] The fault occurrence time can be determined based on the motor speed estimation error increment. Table 2 shows the fault injection experiment under single fault mode and the fault identification time obtained based on algorithm identification. It can be seen from the experimental data that the fault identification time is basically stable at the millisecond level, indicating the effectiveness of fault identification.
[0186] Based on fault identification, the fault is located by the speed estimation deviation. The embodiment selects α=0.5 for the convenience of calculation, that is, the relative estimation error and the relative estimation error increment are equally important. The embodiment takes the No. 3 and No. 5 rotor faults as examples, and the obtained estimation error index bar graph is as follows Fig.13 As shown in the figure, it can be seen that the rotor speed errors calculated for rotors 3 and 5 are the largest, and thus rotors 3 and 5 fail.
[0187] Based on the above experimental results, it can be shown that the fault diagnosis algorithm under closed-loop conditions proposed in this chapter is effective and can meet the needs of rotorcraft UAV power unit fault diagnosis.
[0188] See also Figure 4 An embodiment of the present invention provides an unmanned rotorcraft power unit fault detection system under closed-loop conditions based on an unknown observer, comprising one or more processors for implementing the unmanned rotorcraft power unit fault detection method under closed-loop conditions based on an unknown observer in the above-mentioned embodiment.
[0189] The embodiment of the unmanned rotorcraft power unit fault detection system under closed-loop conditions based on an unknown observer of the present invention can be applied to any device with data processing capabilities, and the device with data processing capabilities can be a device or device such as a computer. The device embodiment can be implemented through software, or through hardware or a combination of software and hardware. Taking software implementation as an example, as a device in a logical sense, it is formed by the processor of any device with data processing capabilities in which it is located reading the corresponding computer program instructions in the non-volatile memory into the internal memory for execution. From a hardware perspective, if Fig.14 As shown, it is a hardware structure diagram of a device with data processing capability in which a power unit fault detection system for an unmanned rotorcraft under closed-loop conditions based on an unknown observer is located. Fig.14 In addition to the processor, memory, network interface, and non-volatile memory shown, any device with data processing capabilities in which the apparatus in the embodiments is located may also include other hardware, generally based on the actual functions of the device with data processing capabilities, which will not be described in detail.
[0190] The implementation process of the functions and effects of each unit in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, and will not be repeated here.
[0191] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of the present invention. Ordinary technicians in this field can understand and implement it without paying creative work.
[0192] An embodiment of the present invention further provides a computer-readable storage medium having a program stored thereon. When the program is executed by a processor, a method for detecting a power unit fault of an unmanned rotorcraft under closed-loop conditions based on an unknown observer in the above embodiment is implemented.
[0193] The computer-readable storage medium may be an internal storage unit of any device with data processing capability described in any of the aforementioned embodiments, such as a hard disk or a memory. The computer-readable storage medium may also be an external storage device, such as a plug-in hard disk, a smart memory card (SmartMedia card, SMC), an SD card, a flash card (Flashcard), etc. equipped on the device. Furthermore, the computer-readable storage medium may also include both an internal storage unit of any device with data processing capability and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capability, and may also be used to temporarily store data that has been output or is to be output.
[0194] Those skilled in the art will readily appreciate other embodiments of the present application after considering the description and practicing the contents disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The description and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the claims.
[0195] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.
[0196] The contents not described in detail in the specification of the present invention belong to the common knowledge of those skilled in the art.
[0197] The above is a description of the embodiments of the present invention. Through the above description of the disclosed embodiments, professionals and technicians in the field can implement or use the present invention. Various modifications to these embodiments will be apparent to professionals and technicians in the field. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown in this article, but will conform to the widest range consistent with the principles and novelties disclosed herein.
Claims
1. A method for fault diagnosis of unmanned rotorcraft power unit under closed-loop conditions based on unknown input observer. It is characterized in that The method comprises the following steps: S1: Screening of linear space state description for unmanned rotorcraft dynamic kinematics Based on the kinematic and dynamic models of the unmanned rotorcraft, a suitable state set is selected to meet the two conditions for constructing an unknown input observer: the minimum phase system condition and the observer matching condition; S2: Construction of an extended unknown input observer in linear space for unmanned rotorcraft dynamic kinematics; constructing an extended state vector containing the original state and unknown input by transforming the state equation ; S3: Online estimation of ideal throttle control quantity for unmanned rotorcraft; S301: According to the extended state vector Definition, unknown input vector ω(t) Calculate the original state vector x(t) : ; in, O is a zero matrix, I is the unit matrix; Combined with real-time input sensor data, iteratively estimates the ideal throttle control law under normal conditions; S302: Based on the estimation of the ideal throttle control amount and the collection of the actual throttle control amount, the ideal throttle component and the equivalent throttle component are obtained according to the following calculation methods: M=FU; Among them, M represents the throttle component of each axis, U represents the total throttle, and F represents the throttle distribution matrix; S303: Based on the throttle component, the motor speed of each axis is calculated using the throttle component according to a third-order polynomial relationship: ; Among them, a 0 , a 1 , a 2 , a 3 are the coefficients of the cubic polynomial, M i is the throttle component assigned to the i-th rotor axis; S4: Unmanned rotorcraft power unit fault status determination The equivalent speed of each axis is obtained through the actual throttle sensor data, and the ideal speed of each axis is obtained through the unknown observer speed in step S3. The two speeds are subtracted to obtain the estimated error; the motor position that first detects a large range of estimated errors is the fault position; when a fault point occurs, the one with the largest estimated difference is the fault.
2. The fault diagnosis method according to claim 1, It is characterized in that The typical description of the linear state equation considering the unknown input in step S1 is shown in equation (1): ; in, Respectively represent the state variables, known input variables, unknown input variables and observed variables in the linear state space equation; the minimum phase condition and observer matching condition of the system are described in the form of equations (2) and (3): 1) Minimum phase condition: ; 2) Observer matching conditions: rank(CD)=rank(D)=q(3) Where n and q represent the number of states and unknown inputs in the linear system, respectively.
3. The fault diagnosis method according to claim 2, It is characterized in that In step S1, select the rotorcraft linear speed V E = [u,v,w] T , the Euler angular velocity of the body ω B =[p,q,r] T , rotorcraft line position X E =[x e ,y e ,h e ] T , and the Euler angles of the rotorcraft body η=[ϕ,θ,ψ] T As a candidate set of state space; among them, u,v,w They are respectively the x-axis speed of the rotorcraft ground coordinate system, the y-axis speed of the rotorcraft ground coordinate system, and the z-axis speed of the rotorcraft ground coordinate system , p, q, r They are the body pitch Euler angular velocity, body roll Euler angular velocity, and body yaw Euler angular velocity. ,x e ,y e ,h e They are respectively the x-axis position of the rotorcraft ground coordinate system, the y-axis position of the rotorcraft ground coordinate system, and the y-axis position of the rotorcraft ground coordinate system. , ϕ, θ, ψ They are the body pitch Euler angle, body roll Euler angle, and body yaw Euler angle.
4. The fault diagnosis method according to claim 3, It is characterized in that In step S2, according to the state space equation structure described by equation (1), the state vector is set x=[u,v,w,p,q,r] T , known input u=0 , unknown input ω=[dt,de,da, dr] T , observation quantity y=x=[u,v,w,p,q,r] T ,in, dt, de, da, dr They are altitude throttle, pitch throttle, roll throttle, and yaw throttle; Construct an extended state vector containing the original state and unknown input by transforming the state equation , the description is shown in formula (4): ; in, x(t) and ω(t) are the state variables and unknown input variables in formula (1) respectively; In order to adapt the final form of the constructed variables to still satisfy the general structure of the linear equation, the improved description of the output variables is shown in formula (5): ; The observation equation of the linear state space equation is rewritten as shown in equation (6): ; set up , Substitute into formula (6) and rewrite it into formula (7): ; set up , then further rewritten as: ; Through the above transformation, the extended state vector is introduced into the observation equation; on this basis, according to the extended state vector The original state vector x(t) is calculated as follows: ; Substituting into formula (10) we get: ; set up , formula (10) is rewritten as: ; The original linear state space (1) is rewritten as shown in formula (12): ; Set Existence , then the state transfer equation is rewritten as formula (13): ; The expanded linear state equation is expressed as formula (14): 。 5. The fault diagnosis method according to claim 4, It is characterized in that In step S2, according to the extended linear state equation description method of formula (14), a suitable observer form is selected to meet the state tracking in the extended linear equation. Specifically, the extended state is tracked based on the Luenberger observer, and the observer form is shown in formula (15): ; The discretized expression of equation (15) is: ; Among them, z(k+1) is used as an intermediate variable and its initial value is set to the zero vector.
6. The fault diagnosis method according to claim 1, It is characterized in that In step S4, the relative estimation error is used as the basis for subsequent fault location, and the calculation method is as follows: ; Among them, ΔΩ(k) is the speed estimation difference, Ω(k) and Represent equivalent speed and ideal speed respectively; Based on the calculated ΔΩ(k), due to the motor fault, the speed estimation error will suddenly increase. The first position where it increases is the fault position. The fault occurrence event is calculated as follows: ; in, t f is the fault time, m is the number of rotorcraft axes, ΔΩ threshld Indicates the speed estimation error threshold, which is used to distinguish between the estimation error in a normal state and the estimation error in a fault state; Then, the location of the fault is determined, and the fault is when the estimated difference at the fault point is the largest; this is obtained by the following calculation method: ; Among them, |ΔΩ i | represents the measurement of the estimation error of each shaft, and the maximum value position is the fault location; Taking into account the increase of the rotor speed estimation error and the rotor speed estimation error increment when a fault occurs, the relative error weighted summation method is used for calculation; the calculation method is as follows: ; in, a The weights of the two errors are set based on subjective experience of experts or automatically optimized based on offline training data; Finally, the difference between the estimated error and the ideal speed is calculated based on the fault location to establish the relationship between the estimated error and the fault degree. The calculation method is as follows: ; Among them, ΔΩ i represents the error, and P represents the fault degree of each motor.
7. A fault diagnosis system for an unmanned rotorcraft, It is characterized in that include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include methods for executing the method described in any one of claims 1-6.
8. An unmanned rotorcraft, It is characterized in that The unmanned rotorcraft comprises: One or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the drone to execute the method as claimed in any one of claims 1 to 6.
9. A computer-readable storage medium comprising instructions, It is characterized in that When the instruction is executed on a drone, the drone executes the method according to any one of claims 1 to 6.
10. A computer program product comprising a computer program or instructions, It is characterized in that When the computer program or instruction is executed by a processor, the method described in any one of claims 1 to 6 is implemented.
