Electric aircraft rotor pitch calibration method and equipment thereof

Abstract

The invention provides an electric aircraft rotor pitch calibration method and device, and belongs to the field of rotor pitch calibration, and the method comprises the steps: controlling a steering engine to drive a rotor to rotate; in the process of controlling the steering engine to drive the rotor wing to rotate, recording a steering engine stroke value in the steering engine rotating process; measuring the total pitch value, the longitudinal periodic variable pitch value and the transverse periodic variable pitch value of the rotor wing when the steering engine moves to each steering engine travel value; performing calculation on each steering engine travel value and the corresponding three propeller pitch coefficients through a cubic fitting method to obtain a positive transformation matrix between the steering engine travel values and the three propeller pitch coefficients; generating a training sample through the forward transformation matrix; and inputting the training sample into a neural network for training to obtain an inverse transformation relation between the steering engine travel value and the three pitch coefficients, and calibrating the rotor pitch of the electric aircraft through the forward transformation matrix and the inverse transformation relation.

Description

Technical Field

[0001] The present application relates to the technical field of rotor pitch calibration, and in particular to a method and device for calibrating the rotor pitch of an electric aircraft. Background Art

[0002] As aircraft begin to enter people's production and life, the aircraft manufacturing industry has flourished, and various small and medium-sized electric aircraft have provided convenience for people's production and life. Using electrical equipment to replace mechanical equipment and realizing the control of electric aircraft through electrical equipment control strategies has become a current research hotspot, including the main propulsion motor of the aircraft rotor. Compared with the traditional rotor drive and control methods, controlling the operation of the main propulsion motor can achieve precise control of the rotor speed and rotation angle, and controlling the electric servo can achieve precise control of the servo stroke value. This has brought about a fundamental change in the control method of the rotor, but at the same time it has also put forward higher requirements on the accuracy of pitch calibration. The traditional pitch calibration method still has a lot of room for improvement. Therefore, in the context of aircraft electrification, it is necessary to propose a new rotor pitch calibration method that is suitable for electric drive rotors.

[0003] Since the research on aircraft electrification has only gradually emerged in recent years, the method suitable for electric aircraft rotor pitch calibration needs to be optimized. Therefore, it is urgent to propose a new electric aircraft rotor pitch calibration method to calibrate the electric aircraft rotor pitch more accurately. Summary of the invention

[0004] In view of the technical problems existing in the background technology, the present application provides an electric aircraft rotor pitch calibration method and equipment thereof. The electric aircraft rotor pitch calibration method obtains the forward transformation matrix relationship and inverse transformation relationship between the servo stroke value and the total pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value through cubic fitting and neural network training, thereby realizing the calibration of the electric aircraft rotor pitch.

[0005] In a first aspect, an embodiment of the present application provides a method for calibrating rotor pitch of an electric aircraft, comprising:

[0006] Control the servo to drive the rotor to rotate;

[0007] In the process of controlling the steering gear to drive the rotor to rotate, the steering gear stroke value during the steering gear rotation process is recorded; and the total pitch value, longitudinal cyclic pitch value and lateral cyclic pitch value of the rotor are measured when the steering gear moves to each of the steering gear stroke values;

[0008] By using a cubic fitting method, each of the steering gear stroke values ​​and the corresponding collective value, the longitudinal cyclic pitch value, and the lateral cyclic pitch value are calculated to obtain a positive transformation matrix between the collective value, the longitudinal cyclic pitch value, the lateral cyclic pitch value and the steering gear stroke value;

[0009] Randomly generate a plurality of groups of the total pitch value, the longitudinal periodic pitch value and the lateral periodic pitch value, and input them into the forward transformation matrix to generate the steering gear stroke value corresponding to the steering gear stroke value, so as to form training samples;

[0010] The training samples are input into a neural network for training to obtain an inverse transformation relationship between the total pitch value, the longitudinal cyclic pitch value, the lateral cyclic pitch value and the steering gear stroke value.

[0011] In the technical solution of the embodiment of the present application, the forward transformation matrix relationship and inverse transformation relationship between the servo stroke value and the total pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value are obtained by cubic fitting and neural network training, thereby realizing the calibration of the rotor pitch of the electric aircraft.

[0012] In some embodiments, the measuring of the rotor's total pitch value, longitudinal cyclic pitch value, and lateral cyclic pitch value when the servo moves to each servo stroke value includes:

[0013] Setting the advance control angle of the rotor;

[0014] Controlling the rotation of the rotor according to the servo stroke value;

[0015] Controlling the rotation of the blades of the rotor and measuring the pitch angles of the blades at 0°, 90°, 180°, and 270°;

[0016] The total pitch value, the longitudinal cyclic pitch value, and the lateral cyclic pitch value corresponding to the servo stroke value are calculated based on the advance control angle and the propeller angles measured at 0°, 90°, 180°, and 270° azimuths.

[0017] In this embodiment, the pitch angles measured at four orientations of the blade (0°, 90°, 180°, 270°) ensure that the pitch angles of the blade at different orientations are uniform, thereby providing more accurate parameter support for the rotor pitch calibration process.

[0018] In some embodiments, the measuring of the rotor's total pitch value, longitudinal cyclic pitch value, and lateral cyclic pitch value when the servo moves to each servo stroke value further includes:

[0019] Setting the initial installation angle, lateral deflection angle and longitudinal deflection angle of the rotor;

[0020] After obtaining the propeller angle pitches at 0°, 90°, 180°, and 270°, the propeller angle pitches at 0°, 90°, 180°, and 270° are corrected by using the initial installation angle, the lateral deflection angle, and the longitudinal deflection angle to obtain the corrected propeller angle pitch;

[0021] The total pitch value, the longitudinal cyclic pitch value, and the lateral cyclic pitch value corresponding to the servo stroke value are calculated based on the advance control angle and the corrected propeller angle measured at 0°, 90°, 180°, and 270° azimuths.

[0022] In this embodiment, the propeller angle pitch at 0°, 90°, 180°, and 270° is corrected by the initial installation angle, the lateral deflection angle, and the longitudinal deflection angle, and the corrected propeller angle pitch can be used to calculate a more accurate total pitch value θ. 7 、Longitudinal periodic pitch θ s 、Horizontal periodic pitch θ c .

[0023] In some embodiments, in the process of controlling the servo to drive the rotor to rotate, recording the servo stroke value during the servo rotation process includes:

[0024] Set the total distance, the highest point of the servo travel, the lowest point of the servo travel, the minimum displacement difference and the maximum displacement difference;

[0025] Generate a servo stroke value sequence according to the total distance number, the highest point of the servo stroke, the lowest point of the servo stroke, the minimum displacement difference and the maximum displacement difference;

[0026] The steering gear stroke value is obtained by calculating the steering gear stroke value sequence.

[0027] In this embodiment, the servo stroke value obtained by calculating the servo stroke value sequence can greatly improve the control accuracy of the servo, optimize the power output and response time, and prevent over-travel damage.

[0028] In some embodiments, in calculating the positive transformation matrix between the servo stroke value and the total pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value, the positive transformation matrix is:

[0029]

[0030] Where: L 1 , L 2 , L 3 is a set of servo travel values, θ 7 is the total distance value, θ s is the longitudinal periodic pitch value, θ cis the lateral cyclic pitch value, and A, B, C, and C are the coefficients obtained by three-time fitting of each group of servo stroke values ​​and the corresponding total pitch value, longitudinal cyclic pitch value, and lateral cyclic pitch value.

[0031] In this embodiment, the positive transformation matrix between the servo stroke value and the total pitch value, the longitudinal periodic pitch value and the lateral periodic pitch value is obtained by cubic fitting. It is found in actual experiments that the relationship between the servo stroke value and the three pitch coefficients (total pitch value, longitudinal periodic pitch value and lateral periodic pitch value) is approximately a straight line in the image, but it is not completely fitted, and there is still a certain deviation. Therefore, compared with the traditional linear fitting, the cubic fitting method used in this embodiment is closer to the actual situation, has a smaller error, and has a better fitting effect, which can better realize the calibration of the rotor pitch of the electric aircraft.

[0032] In some embodiments, after calculating the positive transformation matrix between the total pitch value, the longitudinal cyclic pitch value, the lateral cyclic pitch value and the servo stroke value, the target total pitch value, the target longitudinal cyclic pitch value and the target lateral cyclic pitch value are input into the positive transformation matrix to calculate the target servo stroke value, the rotation of the drive rotor is controlled according to the target servo stroke value, and the actual total pitch value, actual longitudinal cyclic pitch value and actual lateral cyclic pitch value of the rotor are measured, and the accuracy of the positive transformation matrix is ​​verified by comparing the target total pitch value, the target longitudinal cyclic pitch value and the target lateral cyclic pitch value with the actual total pitch value, the actual longitudinal cyclic pitch value and the actual lateral cyclic pitch value.

[0033] In this embodiment, the reliability of the servo stroke value obtained by the forward transformation matrix is ​​determined by verifying the accuracy of the forward transformation matrix, which can effectively improve the stability and reliability of the rotor pitch calibration.

[0034] In some embodiments, inputting the training sample into a neural network for training includes:

[0035] Establish a BP neural network with 3, 10, and 3 neurons in the input layer, hidden layer, and output layer respectively;

[0036] The servo stroke value in the training sample is used as input, and the total pitch value, the longitudinal periodic pitch value, and the lateral periodic pitch value corresponding to the servo stroke value are used as output. The BP neural network automatically adjusts the weight coefficients of the connections between each layer by comparing the errors between the network output and the target output, thereby completing the training of the BP neural network.

[0037] In the present embodiment, in the forward transformation matrix obtained by cubic fitting, the servo stroke value and the three pitch coefficients are not in a one-to-one correspondence, and there may be inverse transformation errors. Therefore, if the reverse cubic fitting is still used to implement the inverse transformation process, the error generated will not be able to explain which formula produces it, which leads to the possibility of errors in the servo stroke value when it is inversely transformed into three pitch coefficients. Therefore, in order to solve the problem that there may be errors in the inverse transformation process implemented by the above-mentioned reverse cubic fitting, in the present embodiment, a large number of data samples are obtained by using the cubic fitting equation, and the neural model is trained with the data samples to realize the inverse transformation between the servo stroke value and the three pitch coefficients, thereby effectively reducing the inverse transformation error between the servo stroke value and the three pitch coefficients.

[0038] In some embodiments, after obtaining the inverse transformation relationship between the collective pitch value, the longitudinal cyclic pitch value, the lateral cyclic pitch value and the servo stroke value, a set of given servo stroke values ​​are input into the BP neural network, and the theoretical collective pitch value, the theoretical longitudinal cyclic pitch value and the theoretical lateral cyclic pitch value corresponding to the given servo stroke value are calculated by the BP neural network, the rotation of the drive rotor is controlled according to the given servo stroke value, and the actual collective pitch value, the actual longitudinal cyclic pitch value and the actual lateral cyclic pitch value of the rotor are measured, and the accuracy of the inverse transformation relationship between the collective pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value and the servo stroke value is verified by comparing the theoretical collective pitch value, the theoretical longitudinal cyclic pitch value and the theoretical lateral cyclic pitch value with the actual collective pitch value, the actual longitudinal cyclic pitch value and the actual lateral cyclic pitch value.

[0039] In this embodiment, the accuracy of the BP neural network inverse transformation relationship is verified to determine the credibility of the pitch coefficient obtained by the BP neural network inverse transformation relationship, thereby improving the stability and reliability of the rotor pitch calibration.

[0040] The present application also discloses a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the electric aircraft rotor pitch calibration method as described in any one of the above items is implemented.

[0041] The present application also discloses an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor is used to implement the electric aircraft rotor pitch calibration method as described in any one of the above items when executing the computer program stored in the memory.

[0042] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solution of the present application, the following is a brief introduction to the drawings used in the present application. Obviously, the drawings described below are only some embodiments of the present application, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0044] Figure 1 This is a flow chart of the rotor pitch calibration method in an embodiment of the present application;

[0045] Figure 2 This is a three-layer BP neural network model diagram in the embodiment of this application;

[0046] Figure 3 This is a flow chart of the BP neural network algorithm in the embodiment of this application. DETAILED DESCRIPTION

[0047] The following embodiments of the technical solution of the present application are described in detail in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present application, and are therefore only used as examples, and cannot be used to limit the scope of protection of the present application.

[0048] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0049] The term "comprising" herein indicates the presence of the described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their collections. The terms "comprising", "including", "having" and their variations all mean "including but not limited to", excluding that they are specially emphasized in other ways. Hereinafter, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating one or more of the features. In the description of the embodiments of the present application, unless otherwise stated, the meaning of "multiple" is two or more.

[0050] In this article, the term "and / or" is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0051] As aircraft begin to enter people's production and life, the aircraft manufacturing industry has flourished, and various small and medium-sized electric aircraft have provided convenience for people's production and life. Using electrical equipment to replace mechanical equipment and realizing the control of electric aircraft through electrical equipment control strategies has become a current research hotspot, including the main propulsion motor of the aircraft rotor. Compared with the traditional rotor drive and control methods, controlling the operation of the main propulsion motor can achieve precise control of the rotor speed and rotation angle, and controlling the electric servo can achieve precise control of the servo stroke value. This has brought about a fundamental change in the control method of the rotor, but at the same time it has also put forward higher requirements on the accuracy of pitch calibration. The traditional pitch calibration method still has a lot of room for improvement. Therefore, in the context of aircraft electrification, it is necessary to propose a new rotor pitch calibration method that is suitable for electric drive rotors.

[0052] In order to solve the technical problem of how to calibrate the rotor pitch of an electric aircraft more accurately through a new electric aircraft rotor pitch calibration method, the present application provides an electric aircraft rotor pitch calibration method and equipment thereof, wherein the electric aircraft rotor pitch calibration method obtains the forward transformation matrix relationship and inverse transformation relationship between the servo stroke value and the total pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value through cubic fitting and neural network training, thereby achieving the technical effect of calibrating the electric aircraft rotor pitch.

[0053] As aircraft begin to enter people's production and life, the aircraft manufacturing industry has flourished, and various small and medium-sized electric aircraft provide convenience for people's production and life. Using electrical equipment to replace mechanical equipment and realizing the control of electric aircraft through electrical equipment control strategies has become a current research hotspot, including the main propulsion motor of the aircraft rotor. Compared with the traditional rotor drive and control method, controlling the operation of the main propulsion motor can achieve precise control of the rotor speed and rotation angle, and controlling the electric servo can achieve precise control of the servo stroke value. The flight attitude and direction of the electric aircraft are mainly controlled by three independent servo systems, namely the longitudinal cyclic pitch servo, the lateral cyclic pitch servo and the collective pitch servo. Among them, the longitudinal cyclic pitch servo changes the angle of the front and rear blades of the main rotor to affect the front and rear distribution of lift, thereby controlling the pitch action of the helicopter; the lateral cyclic pitch servo changes the angle of the left and right blades of the main rotor to affect the left and right distribution of lift, thereby controlling the roll action of the helicopter; the collective pitch servo controls the overall pitch angle of the main rotor to affect the lift of the helicopter, that is, to control the helicopter to rise or fall. The three groups of servos mentioned above respectively control the total pitch value, longitudinal cyclic pitch value, lateral cyclic pitch value and other parameters of the electric aircraft's main rotor to maintain the stability of the helicopter and achieve different flight maneuvers.

[0054] This embodiment discloses a method for calibrating the rotor pitch of an electric aircraft. Figure 1 , including: controlling the steering gear to drive the rotor to rotate; in the process of controlling the steering gear to drive the rotor to rotate, recording the steering gear stroke value during the steering gear rotation process; and measuring the rotor's total pitch value, longitudinal cyclic pitch value and lateral cyclic pitch value when the steering gear moves to each steering gear stroke value; each steering gear stroke value and the corresponding total pitch value, longitudinal cyclic pitch value, lateral cyclic pitch value are calculated by a cubic fitting method to obtain a positive transformation matrix between the total pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value and the steering gear stroke value;

[0055] A plurality of groups of total pitch values, longitudinal cyclic pitch values ​​and lateral cyclic pitch values ​​are randomly generated and input into a forward transformation matrix to generate steering gear stroke values ​​corresponding to the steering gear stroke values, so as to form training samples; the training samples are input into a neural network for training, and an inverse transformation relationship between the total pitch value, longitudinal cyclic pitch value and lateral cyclic pitch value and the steering gear stroke value is obtained.

[0056] In this embodiment, the forward transformation matrix and inverse transformation relationship between the servo stroke value and the total pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value are obtained by cubic fitting and neural network training, wherein the forward transformation matrix describes the relationship between the servo stroke value and the total pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value, and the inverse transformation relationship is used to derive the total pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value from the servo stroke value. The calibration of the rotor pitch of the electric aircraft can be achieved through the above-mentioned positive transformation matrix and inverse transformation relationship.

[0057] Secondly, in this embodiment, the positive transformation matrix between the servo stroke value and the total pitch value, the longitudinal periodic pitch value and the lateral periodic pitch value is obtained by cubic fitting. It is found in actual experiments that the relationship between the servo stroke value and the three pitch coefficients (total pitch value, longitudinal periodic pitch value and lateral periodic pitch value) is approximately a straight line in the image, but it is not completely fitted, and there is still a certain deviation. Therefore, compared with the traditional linear fitting, the cubic fitting method used in this embodiment is closer to the actual situation, has a smaller error, and has a better fitting effect, which can better realize the calibration of the rotor pitch of the electric aircraft.

[0058] However, since the servo stroke value and the three pitch coefficients are not in a one-to-one correspondence in the forward transformation matrix obtained by cubic fitting, there may be inverse transformation errors. Therefore, if the reverse cubic fitting is still used to implement the inverse transformation process, the error generated will not be able to explain which formula produces it, which leads to the possibility of errors in the servo stroke value when it is inversely transformed into three pitch coefficients. Therefore, in order to solve the problem that there may be errors in the inverse transformation process implemented by the above-mentioned reverse cubic fitting, in this embodiment, a large number of data samples are obtained by using the cubic fitting equation, and the neural model is trained with the data samples to realize the inverse transformation between the servo stroke value and the three pitch coefficients, thereby effectively reducing the inverse transformation error between the servo stroke value and the three pitch coefficients.

[0059] In this embodiment, measuring the rotor's total pitch value, longitudinal cyclic pitch value, and lateral cyclic pitch value when the steering gear moves to each steering gear stroke value includes:

[0060] Set the advance control angle of the rotor; and control the rotation of the rotor according to the servo travel value;

[0061] After the steering gear controls the rotor to rotate to a predetermined angle, the rotor blades are controlled to rotate, and the blade angles are measured at 0°, 90°, 180°, and 270°.

[0062] The total pitch value, longitudinal cyclic pitch value, and lateral cyclic pitch value corresponding to the servo stroke value are calculated based on the advance control angle and the propeller angles measured at 0°, 90°, 180°, and 270° azimuths.

[0063] Specifically, firstly, the advance control angle α of the electric aircraft rotor is obtained through experiments; and the rotation of the rotor is controlled according to the servo stroke value;

[0064] After the servo controls the rotor to rotate to a predetermined angle, the rotor blades are controlled to rotate. After the electric aircraft rotor is running stably, the pitch angles measured by the sensor at four blade positions (0°, 90°, 180°, 270°) are recorded and marked as the pitch angle θ. 0 ,θ 90,θ 180 ,θ 270 ;

[0065] ) Given the advance control angle α, the pitch of the blade at the azimuth angle ψ satisfies the following relationship:

[0066] θ=θ 7 +θ s cos(ψ+α)+θ c sin(ψ+α) (1)

[0067] Among them, θ, θ 7 ,θ s ,θ c They are, respectively, the pitch angle of the blade at the azimuth angle ψ, the total pitch of the rotor under this working condition, the longitudinal cyclic pitch, and the lateral cyclic pitch.

[0068] Substituting the pitch angle data at the above four positions into the above formula (1), we can get:

[0069]

[0070] Expressed in matrix form, we can get:

[0071]

[0072] In the above formula (3), θ 0 ,θ 90 ,θ 180 ,θ 270 are all known, so θ in the above formula s ,θ c ,θ 7 It can be expressed as:

[0073]

[0074] Among them, H * is the MP generalized inverse matrix of H.

[0075] Then, the θ calculated by this method is 7 ,θ s ,θ c They are the total pitch, longitudinal cyclic pitch, and lateral cyclic pitch of the rotor under the corresponding stroke of the servo.

[0076] In a possible embodiment, in order to obtain a more accurate pitch angle, after obtaining the pitch angles at 0°, 90°, 180°, and 270°, it is also necessary to correct the pitch angles at 0°, 90°, 180°, and 270°.

[0077] Specifically, the initial installation angle θ of the rotor needs to be set in advance before calibration0 , transverse deflection angle β, longitudinal deflection angle γ;

[0078] After obtaining the propeller angle pitches at 0°, 90°, 180°, and 270°, the propeller angle pitches at 0°, 90°, 180°, and 270° are corrected using the initial installation angle, lateral deflection angle, and longitudinal deflection angle according to the following formula:

[0079]

[0080] The pitch angles at the four directions after correction can be obtained

[0081] Given the advance control angle α, the pitch of the blade at the azimuth angle ψ satisfies the following relationship:

[0082] θ=θ 7 +θ s cos(ψ+α)+θ c sin(ψ+α) (1)

[0083] Among them, θ, θ 7 ,θ s ,θ c They are, respectively, the pitch angle of the blade at the azimuth angle ψ, the total pitch of the rotor under this working condition, the longitudinal cyclic pitch, and the lateral cyclic pitch.

[0084] Substituting the corrected pitch angle data at the four azimuths into the above formula, we can obtain:

[0085]

[0086] Expressed in matrix form, we can get:

[0087]

[0088] In the above formula, are all known, so θ in the above formula s ,θ c ,θ 7 It can be expressed as:

[0089]

[0090] Among them, H * is the MP generalized inverse matrix of H.

[0091] According to this method, the corrected pitch angle A more accurate total distance value θ can be calculated 7 、Longitudinal periodic pitch θ s 、Horizontal periodic pitch θ c .

[0092] In this embodiment, in the process of controlling the steering gear to drive the rotor to rotate, recording the steering gear stroke value during the steering gear rotation process includes:

[0093] Set the total calibration distance n and the lowest point of the servo travel x min 、Servo travel highest point x max , minimum displacement difference y min , maximum displacement difference y max . Take the lowest point of the servo travel x min is the starting point and the highest point of the servo travel x max As the end point, the total distance number n is calibrated as the sequence length, and the intervals are evenly distributed to generate the total distance sequence x; with the minimum displacement difference y min is the starting point and the maximum displacement difference y max is the midpoint, minimum displacement difference y min As the end point, the total distance n is calibrated as the sequence length, and the interval is evenly distributed to generate the displacement difference sequence y. The following formula is used to generate the servo travel value sequence L:

[0094] x i +y i *I

[0095] Where I is an empirical matrix with 4 rows and 3 columns. After processing, 4n groups of servo travel values ​​can be obtained. Each group of servo travel values ​​is a matrix containing L 1 , L 2 , L 3 The matrix of three data, L 1 , L 2 , L 3 Respectively represent the travel values ​​of the three servos.

[0096] In this embodiment, the servo stroke value is calculated by the servo stroke value sequence, which can greatly improve the control accuracy of the servo, optimize the power output and response time, and prevent over-travel damage.

[0097] In this embodiment, in the forward transformation matrix between the steering gear stroke value and the total pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value, the forward transformation matrix is ​​calculated as follows:

[0098]

[0099] Where: L 1 , L 2 , L 3 is a set of servo travel values, θ 7 is the total distance value, θ s is the longitudinal periodic pitch value, θ cis the lateral cyclic pitch value, and A, B, C, and C are the coefficients obtained by three-time fitting of each group of servo stroke values ​​and the corresponding total pitch value, longitudinal cyclic pitch value, and lateral cyclic pitch value.

[0100] Specifically, the cubic fitting method is used to obtain the servo travel value L and three pitch coefficients θ s ,θ c ,θ 7 The relationship can be fitted into the following matrix form:

[0101]

[0102] in,

[0103] Expanding the above formula, we can get the following three equations:

[0104]

[0105] Each set of servo travel values ​​and three pitch coefficients can get three equations. In order to solve the 30 unknowns in the 4 coefficient matrices A, B, C, and D in the formula, at least 10 sets of measurement data are required. Assuming that n (n>10) sets of data are obtained through measurement, a total of 3n equations can be listed, which are divided into 3 sets of equations as follows:

[0106]

[0107] This overdetermined set of equations can be transformed into the standard matrix equation PX = Q by rearranging and transforming:

[0108]

[0109]

[0110] Then use the formula X=(P T P) -1 P T Q can be used to find the least square solution of X, and the remaining equations can be solved in this way to find the values ​​of the 30 unknowns in the 4 coefficient matrices, and then get the relationship between the servo travel value and the pitch coefficient.

[0111] In this embodiment, after calculating the forward transformation matrix between the total pitch value, the longitudinal cyclic pitch value, the lateral cyclic pitch value and the servo stroke value, the target total pitch value, the target longitudinal cyclic pitch value and the target lateral cyclic pitch value are input into the forward transformation matrix to calculate the target servo stroke value, the rotor is driven to rotate according to the target servo stroke value, and the actual total pitch value, the actual longitudinal cyclic pitch value and the actual lateral cyclic pitch value of the rotor are measured, and the accuracy of the forward transformation matrix is ​​verified by comparing the target total pitch value, the target longitudinal cyclic pitch value and the target lateral cyclic pitch value with the actual total pitch value, the actual longitudinal cyclic pitch value and the actual lateral cyclic pitch value, and the accuracy of the forward transformation matrix is ​​verified, and the credibility of the servo stroke value obtained by the forward transformation matrix is ​​judged by verifying the accuracy of the forward transformation matrix, thereby improving the stability and reliability of the rotor pitch calibration.

[0112] Secondly, in this embodiment, three target pitch coefficients can also be given, and the servo stroke value is calculated by linear fitting and cubic fitting methods respectively, and the pitch calibration test is performed with the stroke value as input to obtain the measured values ​​of the three pitch coefficients. The target values ​​and measured values ​​of the three pitch coefficients are compared, and the mean square error MSE is used as the evaluation index to calculate and compare the errors and fitting effects of the two methods. The results are shown in Table 1 below. The pitch calibration method using cubic fitting in this application can greatly improve the fitting effect compared with linear fitting, which is of great help to engineering practice.

[0113] Table 1 Comparison of linear fitting and cubic fitting errors

[0114]

[0115]

[0116] See also Figure 2-3 In this embodiment, when inputting the training samples into the neural network for training:

[0117] Use the newff function of the MATLAB neural network toolbox to build a BP neural network with 3, 10, and 3 neurons in the input layer, hidden layer, and output layer respectively;

[0118] The hidden layer and output layer neuron transfer functions are tansig function and purelin function respectively, and the network training algorithm adopts Levenberg-Marquardt algorithm trainlm.

[0119] The servo stroke value in the training sample is taken as input, and the total distance value, longitudinal periodic pitch value, and lateral periodic pitch value corresponding to the servo stroke value are taken as output. The BP neural network automatically adjusts the weight coefficients of the connections between each layer by comparing the error between the network output and the target output. The training of the BP neural network is completed, and the inverse transformation relationship between the total distance value, longitudinal periodic pitch value, lateral periodic pitch value and the servo stroke value is obtained.

[0120] In this embodiment, since the servo stroke value and the three pitch coefficients are not in a one-to-one correspondence in the forward transformation matrix obtained by cubic fitting, there may be inverse transformation errors. Therefore, if the reverse cubic fitting is still used to implement the inverse transformation process, the error generated will not be able to explain which formula produces it, which leads to the possibility of errors in the servo stroke value when it is inversely transformed into three pitch coefficients. Therefore, in order to solve the problem that there may be errors in the inverse transformation process implemented by the above-mentioned reverse cubic fitting, in this embodiment, a large number of data samples are obtained by using the cubic fitting equation, and the neural model is trained with the data samples to realize the inverse transformation between the servo stroke value and the three pitch coefficients, thereby effectively reducing the inverse transformation error between the servo stroke value and the three pitch coefficients. Therefore, the BP neural network can quickly and accurately realize the inverse transformation process to meet the needs of engineering practice.

[0121] In this embodiment, after obtaining the inverse transformation relationship between the total pitch value, the longitudinal cyclic pitch value, the lateral cyclic pitch value and the steering gear stroke value, a set of given steering gear stroke values ​​is input into the BP neural network, and the theoretical total pitch value, the theoretical longitudinal cyclic pitch value and the theoretical lateral cyclic pitch value corresponding to the given steering gear stroke value are calculated by the BP neural network, the rotor is driven to rotate according to the given steering gear stroke value, and the actual total pitch value, the actual longitudinal cyclic pitch value and the actual lateral cyclic pitch value of the rotor are measured, and the accuracy of the inverse transformation relationship between the total pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value and the steering gear stroke value is verified by comparing the theoretical total pitch value, the theoretical longitudinal cyclic pitch value and the theoretical lateral cyclic pitch value with the actual total pitch value, the actual longitudinal cyclic pitch value and the actual lateral cyclic pitch value, and the accuracy of the inverse transformation relationship between the total pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value and the steering gear stroke value is verified, and the credibility of the pitch coefficient obtained by the inverse transformation relationship of the BP neural network is judged by verifying the accuracy of the inverse transformation relationship of the BP neural network, thereby improving the stability and reliability of the rotor pitch calibration.

[0122] Secondly, in this embodiment, given a set of servo stroke values, the theoretical values ​​of the three pitch coefficients are solved by using the methods of reverse linear fitting, reverse cubic fitting and BP neural network respectively. At the same time, in the pitch calibration test, when the servo is placed at the given servo stroke value, the measurement is performed to obtain the measured values ​​of the three pitch coefficients. The target values ​​and measured values ​​of the three pitch coefficients are compared, and the mean square error MSE is used as the evaluation index to calculate and compare the errors and fitting effects of the three methods. The results are shown in Table 2 below. It is found that the error of the result calculated by the BP neural network is basically negligible. In addition, it is found that the calculation rates of the three methods are similar during the solution. Therefore, it is believed that the BP neural network can quickly and accurately realize the inverse transformation process to meet the needs of engineering practice.

[0123] Table 2 Comparison of errors among reverse linear fitting, reverse cubic fitting and BP neural network

[0124]

[0125] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and constraints involved in the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0126] Among them, when the embodiment of the present application is implemented by software, it can be implemented in whole or in part in the form of a computer program product. That is, the present application implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electric carrier signals and telecommunication signals.

[0127] The present application also discloses an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor is configured to implement any of the above-mentioned electric aircraft rotor pitch calibration methods when executing the computer program stored in the memory.

[0128] It should be noted that the present application is not limited to the above-mentioned embodiments. The above-mentioned embodiments are only examples, and the embodiments having the same structure as the technical idea and the same effect as the technical solution of the present application are all included in the technical scope of the present application. In addition, without departing from the scope of the main purpose of the present application, various modifications that can be thought of by those skilled in the art to the embodiments and other methods of combining some of the constituent elements in the embodiments are also included in the scope of the present application.

Claims

1. A method for calibrating the rotor pitch of an electric aircraft, characterized in that: include: Control the servo to drive the rotor to rotate; During the process of controlling the steering gear to drive the rotor to rotate, recording the steering gear stroke value during the steering gear rotation process; and measuring the total pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value of the rotor when the steering gear moves to each of the steering gear stroke values; By using a cubic fitting method, each of the steering gear stroke values ​​and the corresponding collective value, the longitudinal cyclic pitch value, and the lateral cyclic pitch value are calculated to obtain a positive transformation matrix between the collective value, the longitudinal cyclic pitch value, the lateral cyclic pitch value and the steering gear stroke value; Randomly generate a plurality of groups of the total pitch value, the longitudinal periodic pitch value and the lateral periodic pitch value, and input them into the forward transformation matrix to generate the steering gear stroke value corresponding to the steering gear stroke value, so as to form training samples; The training samples are input into a neural network for training to obtain an inverse transformation relationship between the total pitch value, the longitudinal cyclic pitch value, the lateral cyclic pitch value and the steering gear stroke value.

2. The electric aircraft rotor pitch calibration method according to claim 1, characterized in that: The measuring of the rotor's total pitch value, longitudinal cyclic pitch value, and lateral cyclic pitch value when the steering gear moves to each steering gear stroke value includes: Setting the advance control angle of the rotor; Controlling the rotation of the rotor according to the servo stroke value; Controlling the rotation of the blades of the rotor and measuring the pitch angles of the blades at 0°, 90°, 180°, and 270°; The total pitch value, the longitudinal cyclic pitch value, and the lateral cyclic pitch value corresponding to the servo stroke value are calculated based on the advance control angle and the propeller angles measured at 0°, 90°, 180°, and 270° azimuths.

3. The electric aircraft rotor pitch calibration method according to claim 2, characterized in that: The measuring of the rotor's total pitch value, longitudinal cyclic pitch value and lateral cyclic pitch value when the steering gear moves to each steering gear stroke value also includes: Setting the initial installation angle, lateral deflection angle and longitudinal deflection angle of the rotor; After obtaining the propeller angle pitches at 0°, 90°, 180°, and 270°, the propeller angle pitches at 0°, 90°, 180°, and 270° are corrected by using the initial installation angle, the lateral deflection angle, and the longitudinal deflection angle to obtain the corrected propeller angle pitch; The total pitch value, the longitudinal cyclic pitch value, and the lateral cyclic pitch value corresponding to the servo stroke value are calculated based on the advance control angle and the corrected propeller angle measured at 0°, 90°, 180°, and 270° azimuths.

4. The electric aircraft rotor pitch calibration method according to claim 1, characterized in that: In the process of controlling the steering gear to drive the rotor to rotate, recording the steering gear stroke value during the steering gear rotation process includes: Set the total distance, the highest point of the servo travel, the lowest point of the servo travel, the minimum displacement difference and the maximum displacement difference; Generate a servo stroke value sequence according to the total distance number, the highest point of the servo stroke, the lowest point of the servo stroke, the minimum displacement difference and the maximum displacement difference; The steering gear stroke value is obtained by calculating the steering gear stroke value sequence.

5. The electric aircraft rotor pitch calibration method according to claim 1, characterized in that: In calculating the positive transformation matrix between the steering gear stroke value and the total pitch value, the longitudinal periodic pitch value and the lateral periodic pitch value, the positive transformation matrix is: Where: L1, L2, L3 are a set of servo travel values, θ7 is the total distance value, θ s is the longitudinal periodic pitch value, θ c is the lateral cyclic pitch value, and A, B, C, and C are coefficients obtained by three-fold fitting of each servo stroke value and the corresponding total pitch value, longitudinal cyclic pitch value, and lateral cyclic pitch value.

6. The electric aircraft rotor pitch calibration method according to claim 1, characterized in that: After calculating the forward transformation matrix between the collective pitch value, the longitudinal cyclic pitch value, the lateral cyclic pitch value and the servo stroke value, the target collective pitch value, the target longitudinal cyclic pitch value and the target lateral cyclic pitch value are input into the forward transformation matrix to calculate the target servo stroke value, the rotation of the rotor is controlled according to the target servo stroke value, and the actual collective pitch value, actual longitudinal cyclic pitch value and actual lateral cyclic pitch value of the rotor are measured to obtain the actual collective pitch value, the actual longitudinal cyclic pitch value and the actual lateral cyclic pitch value, and the accuracy of the forward transformation matrix is ​​verified by comparing the target collective pitch value, the target longitudinal cyclic pitch value and the target lateral cyclic pitch value with the actual collective pitch value, the actual longitudinal cyclic pitch value and the actual lateral cyclic pitch value.

7. The electric aircraft rotor pitch calibration method according to claim 1, characterized in that: Inputting the training samples into a neural network for training includes: Establish a BP neural network with 3, 10, and 3 neurons in the input layer, hidden layer, and output layer respectively; The servo stroke value in the training sample is used as input, and the total pitch value, the longitudinal periodic pitch value, and the lateral periodic pitch value corresponding to the servo stroke value are used as output. The BP neural network automatically adjusts the weight coefficients of the connections between each layer by comparing the errors between the network output and the target output, thereby completing the training of the BP neural network.

8. The electric aircraft rotor pitch calibration method according to claim 7, characterized in that: After obtaining the inverse transformation relationship between the collective pitch value, the longitudinal cyclic pitch value, the lateral cyclic pitch value and the steering gear stroke value, a set of given steering gear stroke values ​​are input into the BP neural network, and the theoretical collective pitch value, the theoretical longitudinal cyclic pitch value and the theoretical lateral cyclic pitch value corresponding to the given steering gear stroke value are calculated by the BP neural network, the rotation of the rotor is controlled according to the given steering gear stroke value, and the actual collective pitch value, the actual longitudinal cyclic pitch value and the actual lateral cyclic pitch value of the rotor are measured to obtain the actual collective pitch value, the theoretical longitudinal cyclic pitch value and the theoretical lateral cyclic pitch value, and the accuracy of the inverse transformation relationship between the collective pitch value, the longitudinal cyclic pitch value and the lateral cyclic pitch value and the steering gear stroke value is verified by comparing the theoretical collective pitch value, the theoretical longitudinal cyclic pitch value and the theoretical lateral cyclic pitch value with the actual collective pitch value, the actual longitudinal cyclic pitch value and the actual lateral cyclic pitch value.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the electric aircraft rotor pitch calibration method according to any one of claims 1 to 8 is implemented.

10. An electronic device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor is used to implement the electric aircraft rotor pitch calibration method as described in any one of claims 1 to 8 when executing the computer program stored in the memory.

Images