Active-disturbance-rejection control algorithm for aviation electro-mechanical actuator aiming at large-load and large-disturbance scenes

Abstract

The invention discloses an active-disturbance-rejection control algorithm of an aviation electro-mechanical actuator for a heavy-load and large-disturbance scene. The active-disturbance-rejection control algorithm comprises the following steps: step 1, establishing a first-order nonlinear active-disturbance-rejection vector control method framework for motor control; 2, establishing a mathematical physical model of an open-loop system of the permanent magnet synchronous motor; step 3, establishing a speed loop active disturbance rejection controller; step 4, establishing a position loop first-order active disturbance rejection controller; step 5, establishing a current loop active disturbance rejection controller; step 6, based on a Landau discrete recursion method, adding a rotational inertia identification algorithm into the speed loop active disturbance rejection controller in the step 3; 7, establishing a load torque observer based on an integral sliding mode, and introducing the torque observer into the speed loop active disturbance rejection controller in the step 3; and step 8, optimizing the speed loop active disturbance rejection controller in the step 3. According to the method, the influence of the external environment is fully considered, the adaptability of the system is greatly improved, and the situation of large load and large disturbance can be effectively dealt with.

Description

Technical Field

[0001] The present invention relates to the field of aviation electromechanical technology, and in particular to an aviation electromechanical actuator self-disturbance rejection control algorithm for large load and large disturbance scenarios. Background Art

[0002] AC servo control technology is widely used in the field of engineering control due to its excellent performance such as high precision, fast response and high efficiency. With the development of semiconductor devices, microprocessors and large-scale integrated circuits, AC servo control technology plays an important role in the field of engineering control. At present, servo control technology for precision control and tracking is widely used in military science and technology fields such as aerospace, large-load weapon systems and military radars, and is making continuous progress in these fields. In the aviation field, with the continuous breakthroughs in technology, the requirements for equipment carrying technology are getting higher and higher. Researching a servo control technology that can meet the requirements is an urgent problem that needs to be solved.

[0003] Electromechanical Actuator (EMA) is a general term for a type of position servo control system. It controls the movement of the load directly or indirectly by controlling the operation of the motor or electrical appliance to achieve position control of the control target. Electromechanical actuators are mainly used in mechanisms such as retraction, steering, and thrust vectoring. The electromechanical actuation system consists of three parts: motor control, drive control, and mechanical transmission. The core is the motor control of the motor. Permanent magnet synchronous motor (PMSM) is the most frequently used motor as the control object due to its excellent properties such as high system efficiency, simple system structure, and high safety.

[0004] In the actual environment of aviation, these actuators are affected by factors such as air turbulence, temperature changes, mechanical vibration, strong magnetic interference, etc., which may cause the permanent magnet synchronous motor to experience sudden acceleration and deceleration and load mutation during operation. In order to meet the performance of the system, the motor needs to respond and recover quickly and stably reach the target speed or position. In addition, aviation electromechanical actuators are also required to have excellent properties such as fast response speed, high precision, and strong reliability to better meet the requirements of aircraft for flight control and improve the maneuverability, stability and safety of the aircraft.

[0005] At present, the working conditions of the electromechanical servo system for aviation are characterized by high power, large load disturbance, and short-term high load, which puts forward higher requirements on the controller performance of the electromechanical servo system under the conditions of high specific power, fast response and large power change, and stability under large load disturbance. The existing control technology cannot show good performance in dealing with these situations, so there is an urgent need for an aviation electromechanical actuator control algorithm for large load and large disturbance scenarios. Summary of the invention

[0006] To meet the above needs, the present invention aims to provide an active disturbance rejection control algorithm for aviation electromechanical actuators for large load and large disturbance scenarios. The algorithm is based on a first-order nonlinear active disturbance rejection controller (ADRC), and introduces moment of inertia identification and load torque observation functions in the speed loop active disturbance rejection controller to identify the moment of inertia and load torque of the servo system, and compensate them to the system in the form of forward feedback, so that the algorithm fully considers the influence of the external environment, greatly improves the adaptability of the system, and can effectively cope with large load and large disturbance scenarios.

[0007] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0008] The anti-disturbance control algorithm for aviation electromechanical actuators in large load and large disturbance scenarios includes the following steps:

[0009] Step 1: According to the parameters of the motor of the electromechanical actuator, a first-order nonlinear active disturbance rejection control FOC framework for motor control is established;

[0010] Step 2: According to the first-order nonlinear active disturbance rejection control FOC framework, a mathematical and physical model of a permanent magnet synchronous motor open-loop system with the q-axis current of the motor as input and the speed as output is established;

[0011] Step 3: Establish a speed loop active disturbance rejection controller according to the mathematical and physical model and the first-order nonlinear active disturbance rejection control method;

[0012] Step 4: Establish a first-order anti-disturbance controller for the position loop with the motor angle as input and the motor angular velocity as output;

[0013] Step 5: derive the system equation of the motor current loop according to the motor stator voltage equation, and establish the current loop anti-disturbance controller accordingly;

[0014] Step 6: Based on the Landau discrete recursive method, the moment of inertia identification algorithm is added to the speed loop anti-disturbance control controller in step 3;

[0015] Step 7: Establish a load torque observer based on integral sliding film, and introduce the torque observer into the speed loop anti-disturbance controller in step 3;

[0016] Step 8: Based on the identification compensation and load observation methods in steps 6 and 7, optimize the speed loop anti-disturbance control controller in step 3.

[0017] Preferably, the first-order nonlinear active disturbance rejection control FOC framework includes forward and inverse park transformation, forward and inverse clark transformation, and space vector pulse width modulation SVPWM;

[0018] The Clark transformation is as follows:

[0019]

[0020] The Clark inverse transform is as follows:

[0021]

[0022] Where: i α ,i β is the equivalent value of the three-phase current in the α-β dimension, i a ,i b ,i c It is the current value in the original three-phase coordinate system;

[0023] The Park transform is as follows:

[0024]

[0025] The inverse Park transform is as follows:

[0026]

[0027] Where: i d ,i q is the vector value of the current on the dq axis after rotating by an angle of θ.

[0028] Preferably, step 2 comprises the following steps:

[0029] Step 2.1: Convert the mathematical model of the motor in the three-phase coordinate system to the dq coordinate system, then:

[0030] Voltage equation:

[0031]

[0032] Among them, u d ,u q ,i d ,i q ,L d ,L q ,ω e ,R,ψ f They represent the voltage, current, inductance of the d and q axes in the motor dq coordinate system, as well as the angular velocity, resistance, and flux of the motor. The definitions of these parameters in the following equations are exactly the same; at the same time, an important idea of ​​the foc algorithm is that the torque can only be obtained through the q-axis current i q To control, and always keep the d-axis current i d Control is 0;

[0033] Magnetic flux equation:

[0034]

[0035] Among them, ψ d ,ψ q is the component of the motor flux on the d-axis and q-axis;

[0036] Torque equation:

[0037]

[0038] Among them, T e is the electromagnetic torque of the motor itself, p is the number of magnetic poles of the motor, and since i is always controlled d =0, so the torque equation here can be simplified to:

[0039]

[0040] Mechanical equation:

[0041]

[0042] Among them, T e is the electromagnetic torque of the motor, T f is the static friction torque, T m is the mechanical torque, ω m is the motor speed, J is the moment of inertia, and B is the damping coefficient;

[0043] Step 2.2: Take the motor's q-axis current i q is the input, speed ω m As the output, the open-loop formula is as follows:

[0044]

[0045] Preferably, step 3 comprises the following steps:

[0046] Step 3.1: Preprocess the open-loop formula of step 2 and set the output ω m is the state variable X1, input i q is u, and the rest is combined with the external disturbance v and recorded as the total disturbance f as follows:

[0047]

[0048] The total disturbance f is then expanded into a new state variable X2, and the disturbance is observed to further compensate for the impact of the disturbance on the system:

[0049] X2=f(X1)

[0050]

[0051] count The expanded state equation of the first-order system is as follows:

[0052]

[0053] Step 3.2: According to the extended state equation, a speed loop anti-disturbance control is established, including:

[0054] First-order tracking differentiator TD: as follows:

[0055]

[0056] Among them, the optimal comprehensive control function is defined as follows:

[0057]

[0058] Among them, a i ,δ is the parameter that needs to be debugged, sign is the sign function;

[0059] The extended state observer ESO is as follows:

[0060]

[0061] Among them, ω is the actual speed output of the motor measured by the speed sensor, z1, z2 are the observed value of the motor speed and the observed value of the disturbance, β1, β2 are the parameters to be debugged for the algorithm, As a function of the deviation e, take

[0062] The nonlinear state error feedback control rate NSEF is as follows:

[0063] u0=k fal(v1-z1,α1',δ')

[0064] Among them, k, α1', δ' are the parameters to be debugged, and v1 is the reference target value of the speed output by the first-order tracking differentiator;

[0065] The disturbance compensation of the control rate is as follows:

[0066]

[0067] Preferably, step 4 comprises:

[0068] The motor position loop state equation:

[0069]

[0070] Among them, u=ω, y=θ, x=θ are the speed and angle of the motor respectively.

[0071] Tracking Differentiator:

[0072]

[0073] Among them, τ is the parameter for arranging the transition process.

[0074] Extended State Observer:

[0075]

[0076] Among them, theta sensor measures the motor position, and z1 is the observed value of the motor position.

[0077] Linear state error control law:

[0078] u0=K p (θ1-z1)

[0079] Among them, θ is the motor position measured by the sensor, and z1 is the observed value of the motor position;

[0080] The disturbance compensation for the control rate is as follows:

[0081]

[0082] Preferably, step 5 comprises:

[0083] According to the motor stator voltage equation:

[0084]

[0085] Among them, L q =L d =L,U=[u d ,u q ] T ,Y=[i d ,i q ] T ,X1=[i d ,i q ] T , The state equation is as follows:

[0086]

[0087] Among them, f is the total disturbance term, including internal disturbance and external disturbance W, and the specific form is as follows:

[0088]

[0089] Expand the total disturbance f into a new state variable and rewrite the above formula as follows:

[0090]

[0091] Based on the existing state equation, a first-order anti-disturbance controller of the current loop is constructed. The current loop controller needs to have the characteristics of fast response, and no tracking differentiator is set. The rest of the parts are as follows:

[0092] Extended State Observer:

[0093]

[0094] Among them, i is the actual value of the current detected by the current sensor of the system, and z1 is the observed value of the current actually flowing in the control system;

[0095] State error control rate:

[0096] u0=K p (z1-i1)

[0097] Among them, i1 is the target value of current, z1 is the observed value of current;

[0098] The disturbance compensation of the control rate is as follows:

[0099]

[0100] Preferably, step 6 comprises the following steps:

[0101] Step 6.1: Since the change of the large load motor is not obvious during actual operation, the load change rate can be considered to be 0. Based on this, the motor mechanical equation given in step 2 is discretized as follows:

[0102]

[0103] Among them, T L is the load moment, T s is a discrete interval time;

[0104] Substituting T in the above formula L The constant coefficient reference model can be obtained by eliminating the term:

[0105]

[0106] remember ΔT e (k-1) = T e (k-1)-T e (k-2), the constant coefficient reference model is abbreviated as:

[0107] ω m (k)=2ω m (k-1)-ω m (k-2)+bΔT e (k-1)

[0108] Step 6.2: Based on the form of the constant coefficient reference model in step 4.1, the adjustable model is established as follows:

[0109]

[0110] Step 6.3: According to the formula in step 4.2, the parameters to be identified are transformed to obtain:

[0111]

[0112] Among them, β α is the adaptive gain of the system, which mainly determines the identification deviation weight and ensures the convergence of the identification results;

[0113] because Then we have:

[0114]

[0115] Among them, b * is the value after identification and compensation. After adding the moment of inertia identification, the ADRC is obtained.

[0116] Preferably, step 7 comprises:

[0117] Assuming that the load torque of the system does not change within the cycle time, combined with the motor mathematical model in the speed loop, the following form is obtained:

[0118]

[0119] The expansion torque observer is:

[0120]

[0121] in, are the estimated values ​​of the motor angular velocity and load torque, k1 and k2 are parameters to be adjusted, and both are positive values;

[0122] The velocity tracking error of the system is defined as:

[0123] e ω =ω * -ω

[0124] Combined with the motor mathematical model in the speed loop, we can get:

[0125]

[0126] The sliding surface of the system is selected as:

[0127]

[0128] in:

[0129]

[0130] Among them, β is the design parameter of error shaping. The performance of the function can be adjusted by adjusting the parameter. The principle of adjustment is that small error leads to large gain, and large error leads to small gain.

[0131] The differential treatment of the sliding surface of the system is as follows:

[0132]

[0133] Use exponential approach rate to improve the dynamic performance of the approach curve:

[0134]

[0135] Among them, ρ>0, ε>0;

[0136] Combining the sliding surface differential formula and the exponential approach rate, the control rate is as follows:

[0137]

[0138] In order to weaken the chattering of the system, a small gain function is used to replace the sign function sgn(s). The setting of the small gain function is as follows:

[0139]

[0140] The control rate is replaced by the observation result of the above observer, which is:

[0141]

[0142] Preferably, step 8 comprises the following steps:

[0143] Step 8.1: Reconstruct the velocity loop state equation in step 3 to reduce the observed load under large load and large disturbance conditions and improve the performance of the overall actuation system. Specifically:

[0144]

[0145] in, is the observable disturbance in the total disturbance, and its value is given according to the identification and observation results in steps 6 and 7, specifically:

[0146]

[0147] The friction coefficient B of the system is replaced by the calibrated value of the motor;

[0148] Step 8.2: Optimize part of the structure of the speed loop ADRC in step 3 as follows:

[0149] Extended State Observer:

[0150]

[0151] Among them, ω is the motor speed actually measured by the speed sensor, and z1 is the observed value of the motor speed at the current moment in the control system.

[0152] State error feedback control rate:

[0153] u0=k fal(ω1-z1,α1',δ')

[0154] Among them, k, α1', δ' are the parameters to be debugged, ω1, z1 are the target value and observed value of the speed;

[0155] The disturbance compensation of the control rate is as follows:

[0156]

[0157] The present invention fully considers the influence of disturbance factors of the external environment on the electromechanical servo system, and by adding identification compensation technology, the adaptability of the system is greatly improved, and it can effectively cope with large load and large disturbance scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0158] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0159] Figure 1 It is a principle block diagram of the present invention.

[0160] Figure 2 This is the schematic diagram of the motor vector control strategy (foc).

[0161] Figure 3 This is the basic structure diagram of the active disturbance rejection controller.

[0162] Figure 4 The actual structure of the speed loop anti-disturbance controller with moment of inertia identification added.

[0163] Figure 5 This is the principle diagram of the vector control strategy (foc) with the integral synovial load observer added. DETAILED DESCRIPTION

[0164] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings.

[0165] like Figure 1 As shown, the present invention discloses an anti-disturbance control algorithm for an aircraft electromechanical actuator for a large load and large disturbance scenario, comprising the following steps:

[0166] Step 1: Determine the relevant parameters of the controlled motor and derive the first-order nonlinear active disturbance rejection control FOC framework and its component modules of motor control, including forward and inverse park transform, forward and inverse clark transform, space vector pulse width modulation SVPWM, etc. Specifically:

[0167] Step 1.1: First-order nonlinear active disturbance rejection control vector control strategy principle (foc) framework as follows Figure 2 As shown, the illustrated structure is built based on simulink simulation.

[0168] Step 1.2: Convert the motor variables originally in the three-dimensional coordinate system to the α-β two-dimensional coordinate system according to the principle of corresponding decomposition. This is the Clark transformation, and the opposite is called the Clark inverse transformation. The transformation method from the stationary α-β coordinate system to the rotating dq coordinate system is called the Park transformation, and the opposite is called the Park inverse transformation. The mathematical formulas of the positive and negative Park changes in the framework and the positive and negative Clark transformations are as follows:

[0169] Clark Transformation:

[0170]

[0171] Among them, i α ,i β is the equivalent value of the three-phase current in the α-β dimension, i a ,i b ,i c is the current value in the original three-phase coordinate system, and the following inverse transformation parameter definitions are the same here;

[0172] Clark inverse transform:

[0173]

[0174] Park Transform:

[0175]

[0176] Among them, i d ,i q is the vector value of the current on the dq axis after rotating by angle θ;

[0177] Inverse Park transform:

[0178]

[0179] The definitions of the parameters are the same as above.

[0180] Step 2: Using the motor's q-axis current as input and the speed as output, derive the mathematical and physical model of the permanent magnet synchronous motor open-loop system, as follows:

[0181] Step 2.1: First, according to the method described in step 1, the mathematical model of the motor in the three-phase coordinate system is converted to the dq coordinate system, then:

[0182] Voltage equation:

[0183]

[0184] Among them, u d ,u q ,i d ,i q ,L d ,L q ,ω e ,R,ψ f They represent the voltage, current, inductance of the d and q axes in the motor dq coordinate system, as well as the angular velocity, resistance, and flux of the motor. The definitions of these parameters in the following equations are exactly the same; at the same time, an important idea of ​​the foc algorithm is that the torque can only be obtained through the q-axis current i q To control, and always keep the d-axis current i d Control is 0;

[0185] Magnetic flux equation:

[0186]

[0187] Among them, ψ d ,ψ q is the component of the motor flux on the d-axis and q-axis;

[0188] Torque equation:

[0189]

[0190] Among them, T e is the electromagnetic torque of the motor itself, p is the number of magnetic poles of the motor, and since i is always controlled d =0, so the torque equation here can be simplified to:

[0191]

[0192] Mechanical equation:

[0193]

[0194] Among them, T e is the electromagnetic torque of the motor, Tf is the static friction torque, T m is the mechanical torque, ω m is the motor speed, J is the moment of inertia, and B is the damping coefficient.

[0195] Step 2.2: Combine the motor mathematical model in step 2.1 and take the motor's q-axis current i q is the input, speed ω m For the output, the formula is as follows:

[0196]

[0197] The definitions of all parameters are the same as those in the above step 2.1.

[0198] Step 3: Combine the first-order nonlinear anti-disturbance control to establish the foc speed loop anti-disturbance control, specifically:

[0199] Step 3.1: Preprocess the open-loop formula obtained in step 2 and set the output ω m is the state variable X1, input i q is u, and the rest is combined with the external disturbance v and recorded as the total disturbance f, specifically:

[0200]

[0201] The parameters of each part are the same as above. The total disturbance f is expanded into a new state variable X2, and the disturbance is observed to better compensate for the impact of the disturbance on the system, specifically:

[0202] X2=f(X1)

[0203]

[0204] count The expanded state equation of the first-order system is written as follows:

[0205]

[0206] Step 3.2: The basic structure of the ADRC is as follows: Figure 3 According to the state equation given in step 3.1, the speed loop anti-disturbance controller is designed, and its various parts are specifically structured as follows:

[0207] First-order tracking differentiator TD:

[0208]

[0209] Among them, the optimal comprehensive control function is defined as follows:

[0210]

[0211] Among them, a i ,δ is the parameter that needs to be debugged, and sign is the sign function.

[0212] Extended State Observer ESO:

[0213]

[0214] Among them, ω is the actual speed output of the motor measured by the speed sensor, z1, z2 are the observed value of the motor speed and the observed value of the disturbance, β1, β2 are the parameters to be debugged for the algorithm, is a function of the deviation e, usually taken as You can achieve better results.

[0215] Nonlinear state error feedback control rate NSEF:

[0216] u0=k fal(v1-z1,α1',δ')

[0217] Among them, k, α1', δ' are the parameters to be debugged, v1 is the reference target value of the speed output by the first-order tracking differentiator, and the other parameters are defined as above. The control rate is subjected to disturbance compensation as follows:

[0218]

[0219] The parameter definitions are the same as in the previous steps. Connect and encapsulate these structures to get Figure 3 structure.

[0220] Step 4: A first-order motor position loop anti-disturbance controller with motor angle as input and motor angular velocity as output is as follows:

[0221] The motor position loop state equation:

[0222]

[0223] Among them, u=ω, y=θ, x=θ are the speed and angle of the motor respectively.

[0224] Tracking Differentiator:

[0225]

[0226] Among them, τ is the parameter for arranging the transition process.

[0227] Extended State Observer:

[0228]

[0229] Among them, theta sensor measures the motor position, and z1 is the observed value of the motor position.

[0230] Linear state error control law:

[0231] u0=K p (θ1-z1)

[0232] Among them, the θ sensor measures the motor position, and z1 is the observed value of the motor position. Disturbance compensation is performed on the control rate as follows:

[0233]

[0234] The parameter definitions are the same as in the previous steps.

[0235] Step 5: According to the motor stator voltage equation:

[0236]

[0237] Among them, L q =L d =L,U=[u d ,u q ] T ,Y=[i d ,i q ] T ,X1=[i d ,i q ] T , The state equation is as follows:

[0238]

[0239] Among them, f is the total disturbance term, including internal disturbance and external disturbance W, and the specific form is as follows:

[0240]

[0241] Expand the total disturbance f into a new state variable and rewrite the above formula as follows:

[0242]

[0243] Based on the existing state equation, a first-order anti-disturbance controller of the current loop is constructed. The current loop controller needs to have the characteristics of fast response, and no tracking differentiator is set. The rest of the parts are as follows:

[0244] Extended State Observer:

[0245]

[0246] Among them, i is the actual value of the current detected by the current sensor of the system, and z1 is the observed value of the current actually flowing in the control system.

[0247] State error control rate:

[0248] u0=K p (z1-i1)

[0249] Among them, i1 is the target value of the current, and z1 is the observed value of the current. The control rate is subjected to disturbance compensation as follows:

[0250]

[0251] The definitions of various parameters are the same as those in the previous steps.

[0252] Step 6: The Landau discrete recursive algorithm is a type of model reference adaptive system (MRAS). It is necessary to set a constant coefficient reference model and an adjustable model, and adopt an appropriate adaptive algorithm to make the error value of the parameters in the two models reach an acceptable range. Based on the Landau discrete recursive idea, the moment of inertia identification algorithm is added to the speed loop anti-disturbance controller optimized in step 3. The details are as follows:

[0253] Step 6.1: Since the change of the large load motor is not obvious during actual operation, the load change rate can be considered to be 0. Based on this, the motor mechanical equation given in step 2 is discretized, specifically:

[0254]

[0255] Among them, T L is the load moment, T s is the discrete interval time, and the definitions of other parameters are the same as above. L The constant coefficient reference model can be obtained by eliminating the term:

[0256]

[0257] remember ΔT e (k-1) = T e (k-1)-T e (k-2), the constant coefficient reference model is abbreviated as:

[0258] ω m (k)=2ω m (k-1)-ω m (k-2)+bΔT e (k-1)

[0259] Step 6.2: Based on the form of the constant coefficient reference model in step 4.1, design the adjustable model as follows:

[0260]

[0261] Step 6.3: According to the formula in step 4.2, the parameters to be identified are transformed to obtain:

[0262]

[0263] Among them, β α is the adaptive gain of the system, which mainly determines the identification deviation weight and ensures the convergence of the identification result. Then we have:

[0264]

[0265] Among them, b * is the value after identification and compensation. The complete structure of the ADRC after adding the moment of inertia identification is as follows: Figure 4 shown.

[0266] Step 7: The actual working conditions of electromechanical actuation systems with large loads and large disturbances are complex, and they usually encounter sudden increases or decreases in load torque. The traditional speed loop controller has poor control performance in this case. By designing a load torque observer based on integral sliding film, the load torque is acquired in real time and fed back to the system in the form of feedforward compensation. The system flow chart is as follows: Figure 5 , the specific method is as follows:

[0267] Step 7.1: Similar to the assumption in step 6, it can be assumed here that the load torque of the system does not change during the cycle time. Combined with the motor mathematical model in the speed loop, the following form can be obtained:

[0268]

[0269] The expansion torque observer is:

[0270]

[0271] in, are the estimated values ​​of the motor angular velocity and load torque, k1 and k2 are the parameters to be adjusted, and both are positive values.

[0272] The velocity tracking error of the system is defined as:

[0273] e ω =ω * -ω

[0274] Combined with the motor mathematical model in the speed loop, we can get:

[0275]

[0276] The sliding surface of the system is selected as:

[0277]

[0278] in:

[0279]

[0280] Among them, β is the design parameter of error shaping. By adjusting the parameters reasonably to adjust the performance of the function, it is necessary to follow the idea of ​​small error and large gain, and large error and small gain. Differentiation of the sliding surface of the system is performed, specifically:

[0281]

[0282] Use exponential approach rate to improve the dynamic performance of the approach curve:

[0283]

[0284] Among them, ρ>0, ε>0. Combining the sliding surface differential formula and the exponential approach rate, the control rate can be obtained as follows:

[0285]

[0286] In order to weaken the chattering of the system, a small gain function is used to replace the sign function sgn(s). The setting of the small gain function is as follows:

[0287]

[0288] The control rate is replaced by the observation result of the above observer, which is:

[0289]

[0290] Step 8: Use the identification result of the moment of inertia in step 6 and the observation result of the load torque in step 7 to optimize the speed loop anti-disturbance controller proposed in step 3, specifically:

[0291] Step 8.1: Reconstruct the velocity loop state equation in step 3 to reduce the observed load under large load and large disturbance conditions and improve the performance of the overall actuation system. Specifically:

[0292]

[0293] in, is the observable disturbance in the total disturbance, and its value can be given according to the identification and observation results in steps 6 and 7, specifically:

[0294]

[0295] The definitions of the various parameters in the formula are the same as in the previous steps. Since the change of the friction coefficient B of the system is not obvious, the calibrated value of the motor is used instead. It is the unknown disturbance part. The extended state observer of the current velocity loop's anti-disturbance controller only needs to observe this part, which greatly reduces the burden of the observer and improves the performance of the controller.

[0296] Step 8.2: Based on the design in step 8.1, the partial structure of the speed loop ADRC in step 3 is now optimized as follows:

[0297] Extended State Observer:

[0298]

[0299] Among them, ω is the motor speed actually measured by the speed sensor, and z1 is the observed value of the motor speed at the current moment in the control system.

[0300] State error feedback control rate:

[0301] u0=k fal(ω1-z1,α1',δ')

[0302] Among them, k, α1', δ' are the parameters to be debugged, ω1, z1 are the target value and observed value of the speed, and the other parameters are defined as above. Disturbance compensation is performed on the control rate, as follows:

[0303]

[0304] The parameter definition is the same as step 3.

[0305] Through comparative simulation experiments, the auto-disturbance rejection controller based on identification compensation designed by this method can show excellent control performance under large load and large disturbance conditions.

[0306] Of course, the present invention may have many other embodiments. Without departing from the spirit and essence of the present invention, technicians familiar with the field may make various corresponding changes and deformations based on the present invention, but these corresponding changes and deformations should all fall within the scope of protection of the claims attached to the present invention.

Claims

1. The anti-disturbance control algorithm for aviation electromechanical actuators in large load and large disturbance scenarios is characterized by: The following steps are involved: Step 1: According to the parameters of the motor of the electromechanical actuator, a first-order nonlinear active disturbance rejection control vector control method framework for motor control is established; Step 2: According to the first-order nonlinear active disturbance rejection control FOC framework, a mathematical and physical model of a permanent magnet synchronous motor open-loop system with the q-axis current of the motor as input and the speed as output is established; Step 3: Establish a speed loop active disturbance rejection controller according to the mathematical and physical model and the first-order nonlinear active disturbance rejection control method; Step 4: Establish a first-order anti-disturbance controller for the position loop with the motor angle as input and the motor angular velocity as output; Step 5: derive the system equation of the motor current loop according to the motor stator voltage equation, and establish the current loop anti-disturbance controller accordingly; Step 6: Based on the Landau discrete recursive method, the moment of inertia identification algorithm is added to the speed loop anti-disturbance control controller in step 3; Step 7: Establish a load torque observer based on integral sliding film, and introduce the torque observer into the speed loop anti-disturbance controller in step 3; Step 8: Based on the identification compensation and load observation methods in steps 6 and 7, optimize the speed loop anti-disturbance control controller in step 3.

2. The active disturbance rejection control algorithm for aircraft electromechanical actuators in large load and large disturbance scenarios according to claim 1 is characterized in that: The first-order nonlinear active disturbance rejection control FOC framework includes forward and inverse park transformation, forward and inverse clark transformation, and space vector pulse width modulation SVPWM; The Clark transformation is as follows: The Clark inverse transform is as follows: Where: i α ,i β is the equivalent value of the three-phase current in the α-β dimension, i a ,i b ,i c It is the current value in the original three-phase coordinate system; The Park transform is as follows: The inverse Park transform is as follows: Where: i d ,i q is the vector value of the current on the dq axis after rotating by an angle of θ.

3. The active disturbance rejection control algorithm for aircraft electromechanical actuators in large load and large disturbance scenarios according to claim 1 is characterized in that: Step 2 includes the following steps: Step 2.1: Convert the mathematical model of the motor in the three-phase coordinate system to the dq coordinate system, then: Voltage equation: Among them, u d ,u q ,i d ,i q ,L d ,L q ,ω e ,R,ψ f They represent the voltage, current, inductance of the d and q axes in the motor dq coordinate system, as well as the angular velocity, resistance, and flux of the motor. These parameters are in the following equation The definition of is exactly the same; at the same time, an important idea of ​​the foc algorithm is that the torque can only be passed through the q-axis current i q To control, and always keep the d-axis current i d Control is 0; Magnetic flux equation: Among them, ψ d ,ψ q is the component of the motor flux on the d-axis and q-axis; Torque equation: Among them, T e is the electromagnetic torque of the motor itself, p is the number of magnetic poles of the motor, and since i is always controlled d =0, so the torque equation here can be simplified to: Mechanical equation: Among them, T e is the electromagnetic torque of the motor, T f is the static friction torque, T m is the mechanical torque, ω m is the motor speed, J is the moment of inertia, and B is the damping coefficient; Step 2.2: Take the motor's q-axis current i q is the input, speed ω m As the output, the open-loop formula is as follows:

4. The active disturbance rejection control algorithm for aircraft electromechanical actuators in large load and large disturbance scenarios according to claim 3 is characterized in that: Step 3 includes the following steps: Step 3.1: Preprocess the open-loop formula of step 2 and set the output ω m is the state variable X1, input i q is u, and the rest is combined with the external disturbance v and recorded as the total disturbance f as follows: The total disturbance f is then expanded into a new state variable X2, and the disturbance is observed to further compensate for the impact of the disturbance on the system: X2=f(X1) count The expanded state equation of the first-order system is as follows: Step 3.2: According to the extended state equation, a speed loop anti-disturbance control is established, including: First-order tracking differentiator TD: as follows: Among them, the optimal comprehensive control function is defined as follows: Among them, a i ,δ is the parameter that needs to be debugged, sign is the sign function; The extended state observer ESO is as follows: Among them, ω is the actual speed output of the motor measured by the speed sensor, z1, z2 are the observed value of the motor speed and the observed value of the disturbance, β1, β2 are the parameters to be debugged for the algorithm, As a function of the deviation e, take The nonlinear state error feedback control rate NSEF is as follows: u0=k fal(v1-z1,α1',δ') Among them, k, α1', δ' are the parameters to be debugged, and v1 is the reference target value of the speed output by the first-order tracking differentiator; The disturbance compensation of the control rate is as follows:

5. The active disturbance rejection control algorithm for aircraft electromechanical actuators in large load and large disturbance scenarios according to claim 4 is characterized in that: Step 4 includes: The motor position loop state equation: Among them, u=ω, y=θ, x=θ are the speed and angle of the motor respectively. Tracking Differentiator: Among them, τ is the parameter for arranging the transition process. Extended State Observer: Among them, theta sensor measures the motor position, and z1 is the observed value of the motor position. Linear state error control law: u0=K p (θ1-z1) Among them, θ is the motor position measured by the sensor, and z1 is the observed value of the motor position; The disturbance compensation for the control rate is as follows:

6. The active disturbance rejection control algorithm for aircraft electromechanical actuators in large load and large disturbance scenarios according to claim 4 is characterized in that: Step 5 includes: According to the motor stator voltage equation: Among them, L q =L d =L,U=[u d ,u q ] T ,Y=[i d ,i q ] T ,X1=[i d ,i q ] T , The state equation is as follows: Among them, f is the total disturbance term, including internal disturbance and external disturbance W, and the specific form is as follows: Expand the total disturbance f into a new state variable and rewrite the above formula as follows: Based on the existing state equation, a first-order anti-disturbance controller of the current loop is constructed. The current loop controller needs to have the characteristics of fast response, and no tracking differentiator is set. The rest of the parts are as follows: Extended State Observer: Among them, i is the actual value of the current detected by the current sensor of the system, and z1 is the observed value of the current actually flowing in the control system; State error control rate: u0=K p (z1-i1) Among them, i1 is the target value of current, z1 is the observed value of current; The disturbance compensation of the control rate is as follows:

7. The active disturbance rejection control algorithm for aircraft electromechanical actuators in large load and large disturbance scenarios according to claim 6 is characterized in that: Step 6 includes the following steps: Step 6.1: Since the change of the large load motor is not obvious during actual operation, the load change rate can be considered to be 0. Based on this, the motor mechanical equation given in step 2 is discretized as follows: Among them, T L is the load moment, T s is a discrete interval time; Substituting T in the above formula L Eliminating the term, we can get the constant coefficient reference model: remember ΔT e (k-1) = T e (k-1)-T e (k-2), the constant coefficient reference model is abbreviated as: oh m (k)=2ω m (k-1)-ω m (k-2)+bΔT e (k-1) Step 6.2: Based on the form of the constant coefficient reference model in step 4.1, the adjustable model is established as follows: Step 6.3: According to the formula in step 4.2, the parameters to be identified are transformed to obtain: Among them, β α is the adaptive gain of the system, which mainly determines the identification deviation weight and ensures the convergence of the identification results; because Then we have: Among them, b * is the value after identification and compensation. After adding the moment of inertia identification, the ADRC is obtained.

8. The active disturbance rejection control algorithm for aircraft electromechanical actuators in large load and large disturbance scenarios according to claim 7 is characterized in that: Step 7 includes: Assuming that the load torque of the system does not change within the cycle time, combined with the motor mathematical model in the speed loop, the following form is obtained: The expansion torque observer is: in, are the estimated values ​​of the motor angular velocity and load torque, k1 and k2 are parameters to be adjusted, and both are positive values; The velocity tracking error of the system is defined as: e ω =ω * -oh Combined with the motor mathematical model in the speed loop, we can get: The sliding surface of the system is selected as: in: Among them, β is the design parameter of error shaping. The performance of the function can be adjusted by adjusting the parameter. The principle of adjustment is that small error leads to large gain, and large error leads to small gain. The differential treatment of the sliding surface of the system is as follows: Use exponential approach rate to improve the dynamic performance of the approach curve: Among them, ρ>0,ε>0; Combining the sliding surface differential formula and the exponential approach rate, the control rate is as follows: In order to weaken the chattering of the system, a small gain function is used to replace the sign function sgn(s). The setting of the small gain function is as follows: The control rate is replaced by the observation result of the above observer, which is:

9. The active disturbance rejection control algorithm for aircraft electromechanical actuators in large load and large disturbance scenarios according to claim 8 is characterized in that: Step 8 includes the following steps: Step 8.1: Reconstruct the velocity loop state equation in step 3 to reduce the observed load under large load and large disturbance conditions and improve the performance of the overall actuation system. Specifically: in, is the observable disturbance in the total disturbance, and its value is given according to the identification and observation results in steps 6 and 7, specifically: The friction coefficient B of the system is replaced by the calibrated value of the motor; Step 8.2: Optimize part of the structure of the speed loop ADRC in step 3 as follows: Extended State Observer: Among them, ω is the motor speed actually measured by the speed sensor, and z1 is the observed value of the motor speed at the current moment in the control system. State error feedback control rate: u0=k fal(ω1-z1,α′1,δ′) Among them, k, α′1, δ′ are the parameters to be debugged, ω1, z1 are the target value and observed value of the speed; The disturbance compensation of the control rate is as follows:

Images