A passive fault-tolerant control method for multi-rotor actuator failure
By designing a passive fault-tolerant control method suitable for multi-rotors, the safety problem of micro-rotor vehicles when rotor failure is solved, and the safety and stability of multi-rotor vehicles are improved without adding hardware.
Patent Information
- Application Number
- CN202310631548.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-31
- Publication Date
- 2025-08-15
- Estimated Expiration
- 2043-05-31
AI Technical Summary
The prior art is difficult to improve the safety of micro multi-rotor vehicles without adding hardware, especially fault-tolerant control problems when multiple rotors fail.
A passive fault-tolerant control method suitable for the failure of multiple actuators of multi-rotors is designed. By designing a control allocation matrix model, calculating virtual control amounts, estimating rotor failure disturbances and performing low-pass filtering, the control allocation optimization goal is finally achieved, which is suitable for multi-rotor vehicles with different rotor numbers.
The fault-tolerant control that can deal with one or more rotor failures without adding hardware is realized, simplifying the fault-tolerant controller structure of multi-rotor aircraft and improving the safety and stability of the aircraft.
Smart Images
Figure CN116699986B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of multi-rotor fault-tolerant control, including quadrotors, hexacoptors, octorotors and other multi-rotor aircraft, and specifically relates to a passive fault-tolerant control method suitable for failure of multiple actuators of a multi-rotor. Background Art
[0002] Multirotor drones, due to their simple structure and ability to easily perform complex motion control tasks, have seen widespread application in various fields in recent years. This has come with safety concerns. For micro-multirotors in particular, available safety measures are constrained by their size and payload capacity. Specifically, adding spare rotors reduces the multirotor's payload capacity, increases overall system complexity, and increases the cost of the multirotor. Therefore, a common approach to safeguarding micro-multirotors against rotor failure is to employ fault-tolerant control algorithms. This approach improves multirotor safety without adding any additional hardware. Researching a passive fault-tolerant control method for multirotor actuators is crucial for promoting the widespread adoption of multirotor drones and ensuring their safer use in society. Summary of the Invention
[0003] The present invention provides a passive fault-tolerant control method for failure of multiple actuators of a multi-rotor. The method has the following characteristics: the method can be applicable to different multi-rotor aircraft, such as quadrotors, hexacoptors, octorotors, etc.; the method has a simple structure and does not require switching control but can be applicable to different rotor failures. For example, for a quadrotor, the method can be used to achieve fault-tolerant control under the failure of one rotor, fault-tolerant control under the failure of two rotors, and fault-tolerant control under the failure of three rotors. Users do not need to carry out special fault-tolerant control design for different rotor failure types.
[0004] To achieve the above purpose, the implementation steps of the present invention are as follows: Figure 6 As shown;
[0005] Step 1: Design the control allocation matrix model
[0006] Define the body coordinate system of the multi-rotor aircraft. The center of mass of the aircraft is the origin, and the distance from the origin to the front of the aircraft is the x-axis, that is, x b . From the center of mass to the right and with x b The vertical direction is the y-axis, that is, y b . Vertically downward from the center of mass and with x b ,y b The direction perpendicular to the z axis is z b The inertial coordinate system uses the "North-East" orientation.
[0007] Design the control allocation matrix of a multi-rotor aircraft. The core difference between multi-rotor aircraft with different numbers of rotors is the selection of the control allocation matrix. Figure 1 As shown, for a multi-rotor with n rotors, the propellers are marked in clockwise direction from i = 1 to i = n; the body o b x b The angle between the shaft and the support arm where each rotor motor is located is The unit is radian; the distance between the center of the body and the i-th motor is The unit is meter; the relationship between the anti-torque moment and the pulling force of a single rotor is expressed as M i =cT i , T i The unit of force for a single rotor is Newton, M i The unit of the anti-torque generated by a single rotor is Nm, c is the conversion coefficient from force to anti-torque; δ i Indicates the direction of rotation of the rotor. If the rotor rotates clockwise when viewed from above, then δ i = -1, if it rotates counterclockwise, then δ i =1; then the designed control allocation matrix M can be calculated according to formula (1):
[0008]
[0009] The control allocation matrix M represents the linear transformation from rotor thrust to thrust acting on the aircraft system and torque about the aircraft axis. Different multirotors have different control allocation matrices M.
[0010] Step 2: Calculate the expected virtual control amount:
[0011] Step 2.1: Define the rotation matrix. Get the rotation matrix of the current multi-rotor aircraft from the multi-rotor navigation system This matrix consists of 9 elements to represent the attitude of the multirotor aircraft. These 9 elements are needed in the subsequent controller design, so the symbols of these 9 elements are defined here as:
[0012]
[0013] Step 2.2: Get the angular velocity and expected information of the aircraft. Get the current angular velocity of the multirotor aircraft from the multirotor navigation system: ω = [pqr] T The unit is radians per second, and the speed v perpendicular to the ground is obtained at the same time z The unit is meter per second, where p, q, and r represent the multi-rotor aircraft's rotation around the body x b axis, y b Axis, z b The angular velocity of the axis in radians per second.3,x Expectation value n 3,x,d , get n 3,y The expected value n 3,y,d , get v z The expected value v z,d , get the expected value r of r d These expected values can be set manually, such as by a remote controller, or by a higher-level controller designed separately.
[0014] Step 2.3: Design a virtual control quantity generation algorithm. Calculate the expected total body tension f d , around x b Expected moment τ of the shaft p,d , around y b Expected moment τ of the shaft q,d , around z b Expected moment τ of the shaft r,d , as shown in formula (3):
[0015]
[0016]
[0017]
[0018] τ r,d =-J z k r (rr d )
[0019] Among them: J x , J y , J z The multi-rotor aircraft revolves around the body axis x b ,y b , z b The moment of inertia of the vehicle can be obtained through parameter identification. m represents the mass of the vehicle in kilograms. g represents the local acceleration of gravity, for example, 9.81 m / s 2 . are the derivatives of A, B, C, and D respectively, and are calculated as follows:
[0020]
[0021] The adjustable parameters in formula (3) are: k v,z Used to control the vertical velocity gain, k ω Used to control the angular rate gain, Used to adjust the correction force of the multi-rotor aircraft's tilt, k cp is the compensation coefficient of the inertia tensor coupling term, k r is the yaw rate gain.
[0022] The desired virtual control quantity is expressed as a vector u d =[f d τ p,d τ q,d τ r,d ] T .
[0023] Step 3: Estimate the disturbance caused by the rotor failure:
[0024] Step 3.1: Get the current desired motor force T∈[T1 T2 … T n ] T , where 0≤T i ≤T max , i=1,2,…,n; T max Indicates the maximum thrust that the rotor of a multi-rotor aircraft can produce, in Newtons.
[0025] Step 3.2: Calculate the numerical derivative of vertical velocity and body angular rate. Calculate vertical velocity v z Numerical differentiation of Calculate the numerical derivatives of angular rates p, q, and r As shown in formula (5):
[0026]
[0027] where v z,old , p old ,q old , r old Represents v of the previous control cycle respectively z , p, q, r; ΔT represents the control period in seconds.
[0028] Step 3.3: Design the perturbation estimation equation. Calculate
[0029]
[0030] in represents the disturbance vector caused by the rotor failure of the multi-rotor, d f represents the disturbance acting on the body tension, Indicates the effect on the body x b The disturbance torque on the axis, Indicates the effect on the body y b The disturbance torque on the axis, Indicates the effect on the body b Disturbance torque on the axis.
[0031] Step 3.4: Low-pass filter the disturbance estimate. Low-pass filter d to obtain This operation process can be expressed in Laplace form as
[0032]
[0033] Where: s is the Laplace symbol, ∈ is the time constant of the low-pass filter, which needs to be adjusted according to the vibration conditions of the aircraft.
[0034] Step 4: Design control allocation instruction algorithm:
[0035]
[0036] Where: u c K is the control allocation instruction obtained. rotor is an adjustable parameter matrix, specifically in the form of k rotor and k rotor,z It is an adjustable parameter.
[0037] Step 5: Design the control allocation optimization target. The design control allocation optimization target is as follows:
[0038]
[0039] Get the new T. Where: Represents the optimized weight matrix, and its diagonal elements satisfy the relationship w f Indicates the control distribution weight of the tension channel, Represents x around the body b The control distribution weight of the torque channel of the axis, Represents y around the body b The control distribution weight of the torque channel of the axis, Represents z around the body b The control distribution weight of the torque channel of the axis is converted into the actual rotor control command.
[0040] At this point, the passive fault-tolerant control method suitable for multiple actuator failures of a multi-rotor has been designed.
[0041] The advantages and beneficial effects of the present invention are that the present invention is applicable to multi-rotors with different numbers of rotors to achieve fault-tolerant control of actuator failures; the present invention can achieve fault-tolerant control from complete failure of one rotor to complete failure of multiple rotors, without the need to design special fault-tolerant controllers for different rotor failure types, thereby making the fault-tolerant controller structure of the multi-rotor simpler. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1 It is a symbolic diagram of the control distribution matrix of a multirotor aircraft.
[0043] Figure 2 This is a schematic diagram of the quadrotor serial number.
[0044] Figures 3a-3d This is a schematic diagram of the degree of body tilt when the number of rotors of a quadrotor is stopped.
[0045] Figure 4 This is a six-rotor serial number diagram.
[0046] Figures 5a-5d This is a schematic diagram of the body tilt degree when the number of rotors of a six-rotor is stopped.
[0047] Figure 6 It is a flow chart of the control method of the present invention. DETAILED DESCRIPTION
[0048] This invention provides a passive fault-tolerant control method for multiple actuator failures in a multirotor system. The specific implementation is further described using the fault-tolerant control of common quadrotors and hexacopters as examples. Simulations and calculations were performed on a 2.90 GHz computer with 24.0 GB of memory, running MATLAB R2022b on the Windows 11 operating system.
[0049] Taking a quadrotor aircraft as an example, the controller designed by the present invention implements the following steps:
[0050] According to step 1: Figure 1 and Figure 2 , the control allocation matrix of the quadrotor is selected as
[0051]
[0052] in: d1=d2=d3=d4=0.125,δ1=δ3=1,δ2=δ4=-1,c=0.0166。 If it is other types of multi-rotor aircraft, such as hexa-rotor and octa-rotor, then refer to Figure 1 The control allocation matrix can be calculated.
[0053] According to step 2: the control expectation is set to n 3,x,d =0,n 3,y,d =0,v z,d = 0. Calculate the desired virtual control quantity based on formula (3). The parameters involved in formula (3) are selected as k v,z =1,k ω =14,k r =5, J x =J y =0.0056, J z =0.0104, g=9.81, m=0.7, Tmax =8.5, ΔT=0.0025, ∈=0.06, k rotor =1.3, k rotor,z =0.1, w f =1,
[0054] According to step three: calculate the disturbance caused by the rotor failure based on formula (6).
[0055] According to step 4: calculate the control allocation instruction based on formula (8).
[0056] According to step 5: select w f =1, According to the control assignment instruction u obtained from step 4 c , the desired motor pull T is obtained using the optimization objective described by formula (9).
[0057] Simulation process: Build a quadcopter model in MATLAB / Simulink and implement the above controller according to steps 1 to 5. During the simulation, the quadcopter model provides the controller with the required rotation matrix R eb , the body's angular rate p, q, r, and the body's vertical velocity v z The operating frequency of the controller is 400Hz.
[0058] Result analysis: The rotor number of the quadrotor aircraft is as follows: Figure 2 As shown. 3,x , n 3,y Describes the multirotor aircraft z b The projection of the axes in the inertial coordinate system. If they are all 0, it means that the quadcopter is in a horizontal hovering state. The larger their values are, the greater the degree of tilt of the quadcopter. So, n 3,x , n 3,y It can be used to describe whether the attitude of a multi-rotor aircraft is stable. The simulation results are shown in Figure 3, which shows the n of the quadrotor aircraft after the rotor stops. 3,x , n 3,y Constructing a two-dimensional plane scatter plot, we can find that they are around the expected value without divergence. The origin at coordinate [0,0] represents the expected value, which is n 3,x,d =0,n 3,y,d =0. Figure 3a Corresponding to the No. 1 rotor stopping, Figure 3b Corresponding to the stop of rotors 1 and 2, Figure 3c Corresponding to the stop of rotors 1 and 3, Figure 3d This corresponds to the stalling of rotors 1, 2, and 3. The above results demonstrate the effectiveness of the proposed passive fault-tolerant control method for multi-rotor systems with multiple actuator failures.
[0059] Taking a six-rotor aircraft as an example, the controller designed by the present invention implements the following steps:
[0060] According to step 1: Figure 1 and Figure 4 , the control allocation matrix of the six-rotor is selected as
[0061]
[0062] in: d1=d2=d3=d4=d5=d6=0.336, δ2=δ4=δ5=1, δ1=δ3=δ6=-1, c=0.0166.
[0063] According to step 2: the control expectation is set to n 3,x,d =0,n 3,y,d =0,v z,d = 0. Calculate the desired virtual control quantity based on formula (3). The parameters involved in formula (3) are selected as k v,z =1,k ω =14,k r =5, J x =J y =0.1741, J z =0.3033, g=9.81, m=3.75, T max =23.5, ΔT=0.0025, ∈=0.06, k rotor =1.3, k rotor,z =0.1, w f =1,
[0064] According to step three: calculate the disturbance caused by the rotor failure based on formula (6).
[0065] According to step 4: calculate the control allocation instruction based on formula (8).
[0066] According to step 5: select w f =1, According to the control assignment instruction u obtained from step 4 c , the desired motor pull T is obtained using the optimization objective described by formula (9).
[0067] Simulation process: Build a six-rotor aircraft model in MATLAB / Simulink and implement the above controller according to steps 1 to 5. During the simulation process, the multi-rotor aircraft model provides the required rotation matrix R to the controller. eb , the body's angular rate p, q, r, and the body's vertical velocity v zThe operating frequency of the controller is 400Hz.
[0068] Result analysis: The rotor number of the six-rotor aircraft is as follows: Figure 4 The simulation results are shown in Figure 5, which shows the n of the six-rotor aircraft after the rotor stops. 3,x , n 3,y Constructing a two-dimensional plane scatter plot, we can find that they are around the expected value without divergence. The origin at coordinate [0,0] represents the expected value, which is n 3,x,d =0,n 3,y,d =0. Figure 5a Corresponding to the No. 1 rotor stopping, Figure 5b Corresponding to the stop of rotors 1 and 2, Figure 5c Corresponding to the stop of rotors 1 and 3, Figure 5d This corresponds to the stalling of rotors 1, 2, and 3. The above results demonstrate the effectiveness of the proposed passive fault-tolerant control method for multi-rotor systems with multiple actuator failures.
Claims
1. A passive fault-tolerant control method for failure of multiple actuators of a multi-rotor, characterized in that: The steps include: Step 1: Design the control allocation matrix model; Step 2: Calculate the desired virtual control quantity: including: defining the rotation matrix; obtaining the angular velocity and desired information of the aircraft; designing the virtual control quantity generation algorithm; Step 3: Estimating the disturbance caused by the rotor failure: This includes: obtaining the current expected motor thrust; calculating the numerical differentials of the vertical velocity and the body angular rate; designing the disturbance estimation equation; and low-pass filtering the disturbance estimate. Step 4: Design control allocation instruction algorithm: Where: u c is the obtained control allocation instruction; K rotor is the adjustable parameter matrix, s is the Laplace symbol, ∈ is the time constant of the low-pass filter, u d is the desired virtual control quantity, d represents the disturbance vector caused by the rotor failure of the multi-rotor, and d is obtained by low-pass filtering. M represents the linear conversion relationship from rotor thrust to thrust acting on the airframe and torque around the airframe axis, and T is the desired motor thrust; The specific form is k rotor and k rotor,z It is an adjustable parameter; Step 5: Design control allocation optimization objectives; Get a new T; where: Represents the optimized weight matrix, and its diagonal elements satisfy the relationship w f Indicates the control distribution weight of the tension channel, Represents x around the body b The control distribution weight of the torque channel of the axis, Represents y around the body b The control distribution weight of the torque channel of the axis, Represents z around the body b The control distribution weight of the torque channel of the axis; the obtained T is converted into the actual rotor control command.
2. The passive fault-tolerant control method for multiple actuator failures of a multi-rotor according to claim 1, characterized in that: In step 1, define the body coordinate system of the multi-rotor aircraft; the center of mass of the aircraft is the origin, and the distance from the origin to the front of the aircraft is the x-axis, that is, x b ; from the center of mass to the right and with x b The vertical direction is the y-axis, that is, y b ; Vertically downward from the center of mass and with x b ,y b The direction perpendicular to the z axis is z b ; The inertial coordinate system uses the "northeast" orientation; Design the control allocation matrix of a multirotor aircraft; the core difference between multirotor aircraft with different numbers of rotors is the selection of the control allocation matrix; for a multirotor aircraft with n rotors, mark the propellers in clockwise direction from i=1 to i=n; the body o b x b The angle between the shaft and the support arm where each rotor motor is located is The unit is radian; the distance between the center of the body and the i-th motor is The unit is meter; the relationship between the anti-torque moment and the pulling force of a single rotor is expressed as M i =cT i , T i The unit of force for a single rotor is Newton, M i The unit of the anti-torque generated by a single rotor is Nm, c is the conversion coefficient from force to anti-torque; δ i Indicates the direction of rotation of the rotor. If the rotor rotates clockwise when viewed from above, then δ i = -1, if it rotates counterclockwise, then δ i =1; then the designed control allocation matrix M is calculated according to formula (1): The control allocation matrix M represents the linear conversion relationship from rotor thrust to thrust acting on the aircraft system and torque around the body axis; different multi-rotor aircraft have their own different control allocation matrix M.
3. The passive fault-tolerant control method for multiple actuator failures of a multi-rotor according to claim 1, characterized in that: In step 2, the rotation matrix of the current multirotor aircraft is obtained from the multirotor navigation system This matrix consists of 9 elements to represent the attitude of the multirotor aircraft. These 9 elements are needed in the subsequent controller design. The symbols of these 9 elements are defined as:
4. The passive fault-tolerant control method for multiple actuator failures of a multi-rotor according to claim 3, characterized in that: In step 2, the current multi-rotor aircraft body angular velocity ω = [pqr] is obtained from the multi-rotor navigation system. T The unit is radians per second, and the speed v perpendicular to the ground is obtained at the same time z The unit is meter per second, where p, q, and r represent the multi-rotor aircraft's rotation around the body x b axis, y b Axis, z b The angular velocity of the axis, in radians per second; get n 3,x Expectation value n 3,x,d , get n 3,y The expected value n 3,y,d , get v z The expected value v z,d , get the expected value r of r d .
5. The passive fault-tolerant control method for multiple actuator failures of a multi-rotor according to claim 3 or 4, characterized in that: In step 2, calculate the expected total body pull f d , around x b Expected moment τ of the shaft p,d , around y b Expected moment τ of the shaft q,d , around z b Expected moment τ of the shaft r,d , as shown in formula (3): Among them: J x , J y , J z The multi-rotor aircraft revolves around the body axis x b ,y b , z b The moment of inertia of the vehicle is obtained through parameter identification; m represents the mass of the vehicle in kilograms; g represents the local acceleration of gravity; are the derivatives of A, B, C, and D respectively, and are calculated as follows: The adjustable parameters in formula (3) are: k v,z Used to control the vertical velocity gain, k ω Used to control the angular rate gain, Used to adjust the correction force of the multi-rotor aircraft's tilt, k cp is the compensation coefficient of the inertia tensor coupling term, k r is the yaw rate gain; The desired virtual control quantity is expressed as a vector u d =[f d τ p,d τ q,d τ r,d ] T .
6. The passive fault-tolerant control method for multiple actuator failures of a multi-rotor according to claim 5, characterized in that: In step 3, the desired motor force T∈[T1 T2 … T n ] T , where 0≤T i ≤T max , i=1,2,…,n; T max Indicates the maximum thrust that the rotor of the multi-rotor aircraft can generate, in Newtons; calculate the vertical speed v z Numerical differentiation of Calculate the numerical derivatives of angular rates p, q, and r As shown in formula (5): where v z,old , p old ,q old , r old Represents v of the previous control cycle respectively z , p, q, r; ΔT represents the control period in seconds.
7. The passive fault-tolerant control method for multiple actuator failures of a multi-rotor according to claim 6, characterized in that: In step 3, the perturbation estimation equation is: in represents the disturbance vector caused by the rotor failure of the multi-rotor, d f represents the disturbance acting on the body tension, Indicates the effect on the body x b The disturbance torque on the axis, Indicates the effect on the body y b The disturbance torque on the axis, Indicates the effect on the body b Disturbance torque on the axis.
8. The passive fault-tolerant control method for multiple actuator failures of a multi-rotor according to claim 7, characterized in that: In step three, d is low-pass filtered to obtain This operation process is expressed in Laplace form as: Where: s is the Laplace symbol, ∈ is the time constant of the low-pass filter, which needs to be adjusted according to the vibration conditions of the aircraft.