Unmanned helicopter attitude control method based on model-assisted disturbance observer
Through the attitude control method based on the model assisted perturbation observer, the tracking differential and state observer estimating errors and compensating total perturbation is solved, and the unmanned helicopter's insufficient immunity performance under model uncertainty and external interference is achieved, and more efficient attitude control is achieved.
Patent Information
- Application Number
- CN202510643184.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-19
- Publication Date
- 2025-08-29
AI Technical Summary
The existing unmanned helicopter attitude control method has insufficient immunity when facing model uncertainty and external interference. Traditional self-immunity control depends on the mathematical model of unmanned helicopters and is difficult to apply in practice, and there is a lack of specific technical solutions for model-assisted perturbation observers.
The attitude control method based on the model-assisted perturbation observer is adopted. The attitude control model includes a tracking differentializer, a state observer and a nonlinear control law. The smooth reference input is obtained through the tracking differentializer. The state observer estimates the error and compensates for the total perturbation, improving the disturbance estimation accuracy and control performance.
It improves the control performance and immunity of the unmanned helicopter, reduces the difficulty of parameter setting, improves the attitude tracking accuracy and estimation accuracy of disturbances, and enhances the robustness of the system.
Smart Images
Figure CN120560313A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of attitude control of unmanned helicopters, and more particularly to an attitude control method of an unmanned helicopter based on a model-assisted disturbance observer. Background Art
[0002] In the aviation field, unmanned helicopters have demonstrated unique advantages in military reconnaissance, logistics distribution, agricultural and forestry plant protection, emergency rescue and other scenarios with their vertical take-off and landing, hovering in the air, and high maneuverability. Moreover, they have low costs and low losses, which can effectively reduce manpower input and improve operational efficiency. Their application prospects are very broad.
[0003] Attitude control for unmanned helicopters is fundamental to their autonomous flight missions. Most designs for attitude control in unmanned helicopters are based on PID control. PID control is a classic method in automatic control. It achieves fast response through a proportional phase, eliminates steady-state errors through an integral phase, and predicts change trends through a differential phase. It boasts a simple structure and intuitive principles. While it generally meets requirements under normal flight conditions, its interference rejection performance is limited. When the unmanned helicopter experiences model uncertainty or strong external interference, control performance degrades. Active disturbance rejection control (ADRC) has been improved upon PID control through a series of improvements. By designing an extended state observer to estimate the total disturbance and compensate it for the control variable, the control algorithm's interference rejection performance is enhanced. In the field of unmanned helicopter attitude control, traditional ADRC is a control algorithm that does not rely on a mathematical model of the unmanned helicopter. In fact, the state observer in traditional ADRC can be improved by leveraging the mathematical model of the unmanned helicopter, resulting in a model-assisted disturbance observer. However, the mathematical model of an unmanned helicopter is difficult to establish completely and accurately, which is why model-dependent control methods have not been widely used in attitude control of unmanned helicopters. However, it is possible to leverage information from a simplified mathematical model of the unmanned helicopter to reduce the bandwidth of the state observer and improve its accuracy in estimating the total disturbance, thereby further enhancing the unmanned helicopter's disturbance rejection performance. The model uncertainty generated during the simplification process is also included as part of the total disturbance, resulting in a model-assisted disturbance observer with high robustness. However, this approach has yet to be implemented in the field of unmanned helicopter control, providing a practical and feasible technical solution. Summary of the Invention
[0004] In order to overcome the shortcomings and deficiencies in the prior art, the purpose of the present invention is to provide an unmanned helicopter attitude control method based on a model-assisted disturbance observer; this method can improve the control performance and anti-disturbance capability of the unmanned helicopter attitude control.
[0005] To achieve the above object, the present invention is implemented through the following technical solutions: a method for controlling the attitude of an unmanned helicopter based on a model-assisted disturbance observer, wherein the attitude of the unmanned helicopter is controlled using an attitude control model; the attitude control model includes a tracking differentiator, a state observer, and a nonlinear control law; the method for controlling the attitude of the unmanned helicopter includes the following steps:
[0006] X1, the tracking differentiator provides a transition process for the reference input v to obtain a smooth reference input x r1 and its differential x r2 ;
[0007] X2, according to the reference input x r1 and its differential x r2 Calculate the errors e1 and e2 with the angular velocity z1 and angular acceleration z2 estimated by the state observer;
[0008] X3, preliminarily calculate the control variable u0 using the nonlinear control law based on the errors e1 and e2;
[0009] X4, then compensate the control quantity u0 according to the total disturbance z3 estimated by the state observer to obtain the control quantity u c , as the attitude control input of the unmanned helicopter at the current moment.
[0010] Preferably, in step X1, the reference input x r1 and its differential x r2 Calculated by the following formulas:
[0011]
[0012] in, and is x r1 with x r2 The first-order derivative of ; h0 is the step size of the fhan function; r0 is the convergence rate coefficient.
[0013] Preferably, the fhan function is in the following form:
[0014]
[0015] Among them, d, α0, y T ,α1,α2,s y ,α,s a It is the intermediate variable in the calculation process of the fhan function; sign is the symbol function:
[0016]
[0017] Preferably, the estimated state z of the state observer is set to [z1 z2 z3] Tare the angular velocity, angular acceleration, and total disturbance of the unmanned helicopter, respectively. The angular velocity z1 and angular acceleration z2 in step X2 and the total disturbance z3 in step X4 are calculated using the following formula:
[0018]
[0019] Among them, z(k) is the estimated state at time k; z(k+1) is the estimated state at time k+1; y d (k) is the output of the state observer; u d (k)=[u(k)y(k)] T , is the combination of the input u(k) and output y(k) of the unmanned helicopter; L c Observer gain matrix; F E is the state transfer matrix, describing the relationship between the states before and after, G E is the input matrix, reflecting the impact of the input on the state, H E is the output matrix, reflecting the relationship from state to output, J E It is a direct transfer matrix that represents the influence of output on output.
[0020] Preferably, the F E , G E 、H E 、J E They are:
[0021] F E =F-FL c H; G E =[G FL c ]; H E =IL c H; J E =[0L c ];
[0022] Where, F = I + hA; G = hB; H = C; h is the system sampling period;
[0023] C =
[100] ;
[0024] The observer gain matrix L c for:
[0025]
[0026] Among them, the relationship between the poles of the discrete state observer and the poles of the continuous state observer is: ω0 is the pole of the continuous state observer; a 00 with a 11 is an element in the state transfer matrix A, which describes the relationship between states.
[0027] Preferably, in step X2, the errors e1 and e2 are respectively:
[0028] e1=x r1 -z1;
[0029] e2=x r2 -z2.
[0030] Preferably, in step X3, the control amount u0 is preliminarily calculated as:
[0031] u0=k1fal(e1,a e1 ,δ)+k2fal(e2,a e2 ,δ);
[0032] Among them, k1 and k2 are the gains of angular velocity error and angular acceleration error in the nonlinear control law respectively; a e1 with a e2 are the attenuation rates of the angular velocity tracking error and its differential respectively; the fal function is a nonlinear function; the form of the fal function is as follows:
[0033]
[0034] Among them, a is the decay rate of the error; δ is the filter factor of the fal function.
[0035] Preferably, in step X4, the control amount u c for:
[0036]
[0037] Among them, b0 is the compensation coefficient, which describes the relationship between the control quantity and the total disturbance.
[0038] A readable storage medium stores a computer program, which, when executed by a processor, enables the processor to execute the unmanned helicopter attitude control method based on a model-assisted disturbance observer.
[0039] A computer device comprises a processor and a memory for storing a program executable by the processor. When the processor executes the program stored in the memory, the unmanned helicopter attitude control method based on a model-assisted disturbance observer is implemented.
[0040] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0041] The present invention fully utilizes the model information of the unmanned helicopter in the design of the extended state observer, reduces the bandwidth of the extended state observer, effectively improves the estimation accuracy of the disturbance, and improves the control performance and anti-disturbance capability of the unmanned helicopter. At the same time, the use of model information reduces the difficulty of parameter setting; the reference input is smoothed by a tracking differentiator to reduce the tracking difficulty; the control quantity is constructed by a nonlinear control law of state error feedback and compensation for the total disturbance, effectively improving the accuracy of attitude tracking. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1 This is a control principle diagram of the unmanned helicopter attitude control method based on the model-assisted disturbance observer of the present invention;
[0043] Figure 2 Schematic diagram of the construction process of the posture control model of the present invention;
[0044] Figure 3 It is the lateral channel posture of the unmanned helicopter in embodiment 1;
[0045] Figure 4 is the lateral channel angular velocity of the unmanned helicopter in Example 1;
[0046] Figure 5 It is the longitudinal channel posture of the unmanned helicopter in embodiment 1;
[0047] Figure 6 is the longitudinal channel angular velocity of the unmanned helicopter in Example 1. DETAILED DESCRIPTION
[0048] The present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments.
[0049] Example 1
[0050] This embodiment provides an unmanned helicopter attitude control method based on a model-assisted disturbance observer, which controls the attitude of the unmanned helicopter using an attitude control model; the attitude control model includes a tracking differentiator, a state observer, and a nonlinear control law.
[0051] Unmanned helicopter attitude control method, such as Figure 1 As shown, the following steps are included:
[0052] X1, the tracking differentiator provides a transition process for the reference input v to obtain a smooth reference input x r1 and its differential x r2 .
[0053] Reference input x r1 and its differential x r2 Calculated by the following formulas:
[0054]
[0055] in, and is x r1 with x r2 The first-order derivative of ; h0 is the step size of the fhan function; r0 is the convergence rate coefficient.
[0056] The fhan function is in the following form:
[0057]
[0058] Among them, d, α0, y T ,α1,α2,s y ,α,s a It is the intermediate variable in the calculation process of the fhan function.
[0059] sign is the sign function:
[0060]
[0061] X2, according to the reference input x r1 and its differential x r2 The errors e1 and e2 are calculated with the angular velocity z1 and angular acceleration z2 estimated by the state observer LESO.
[0062] Set the estimated state of the state observer z = [z1 z2 z3] T are the angular velocity, angular acceleration, and total disturbance of the unmanned helicopter respectively; the angular velocity z1, angular acceleration z2, and total disturbance z3 are calculated using the following formula:
[0063]
[0064] Among them, z(k) is the estimated state at time k; z(k+1) is the estimated state at time k+1; y d (k) is the output of the state observer; u d (k)=[u(k)y(k)] T , is the combination of the input u(k) and output y(k) of the unmanned helicopter; L c Observer gain matrix.
[0065] The F E , G E 、H E 、J E They are:
[0066] F E =F-FL c H; G E =[G FL c ]; HE =IL c H; J E =[0L c ];
[0067] Where, F = I + hA; G = hB; H = C; h is the system sampling period;
[0068] C =
[100] ;
[0069] The observer gain matrix L c for:
[0070]
[0071] Among them, the relationship between the poles of the discrete state observer and the poles of the continuous state observer is: ω0 is the pole of the continuous state observer; a 00 with a 11 is an element in the state transfer matrix A, which describes the relationship between states.
[0072] The errors e1 and e2 are:
[0073] e1=x r1 -z1;
[0074] e2=x r2 -z2.
[0075] X3. Preliminary calculation of the control variable u0 using the nonlinear control law NLSEF based on the errors e1 and e2:
[0076] u0=k1fal(e1,a e1 ,δ)+k2fal(e2,a e2 ,δ);
[0077] Among them, k1 and k2 are the gains of angular velocity error and angular acceleration error in the nonlinear control law respectively; a e1 with a e2 are the attenuation rates of the angular velocity tracking error and its differential respectively; the fal function is a nonlinear function; the form of the fal function is as follows:
[0078]
[0079] Among them, a is the decay rate of the error; δ is the filter factor of the fal function.
[0080] X4, then compensate the control quantity u0 according to the total disturbance z3 estimated by the state observer LESO to obtain the control quantity u c , as the attitude control input of the unmanned helicopter at the current moment.
[0081]
[0082] Among them, b0 is the compensation coefficient, which describes the relationship between the control quantity and the total disturbance.
[0083] The method for constructing the posture control model of the present invention is as follows: Figure 2 As shown, the following steps are included:
[0084] S1. Based on first principles, a nonlinear dynamic model of the unmanned helicopter is established, and the angular velocity mathematical model is simplified to obtain a single-input single-output mathematical model of the attitude control horizontal channel;
[0085] S2. Based on the active disturbance rejection theory, a tracking differentiator is designed for the angular velocity control of the unmanned helicopter. The transition process is arranged for the reference input to obtain a smooth reference input.
[0086] S3. Design a model-assisted extended state observer based on the single-input single-output mathematical model of the horizontal channel to estimate the disturbance in real time and compensate the control variable in real time;
[0087] S4. Based on the current Euler method, the designed model-assisted state observer is discretized and applied to actual flight to obtain real-time compensation;
[0088] S5. Design a nonlinear control law based on angular velocity state error feedback, calculate the control quantity based on the smooth reference input, state feedback, and real-time compensation, and perform parameter tuning.
[0089] In step S1, the nonlinear dynamic model of the unmanned helicopter attitude is:
[0090]
[0091] in, are the roll angle, pitch angle, and yaw angle of the unmanned helicopter, is the first-order derivative of the roll angle, pitch angle, and yaw angle, [pqr] is the roll angular velocity, pitch angular velocity, and yaw angular velocity of the unmanned helicopter in the body coordinate system, is the first-order derivative of the roll angular velocity, pitch angular velocity, and yaw angular velocity, [I xx I yy I zz ] is the moment of inertia of the unmanned helicopter around the three axes of the body coordinate system, L M The rolling moment generated by the main rotor, L T is the rolling moment generated by the tail rotor, M M The pitching moment generated by the main rotor, Q M The reaction torque generated by the main rotor, N Tis the yaw moment generated by the tail rotor, a1 is the longitudinal flapping angle of the main rotor, b1 is the lateral flapping angle of the main rotor, is the first derivative of the longitudinal flapping angle, is the first derivative of the lateral flapping angle, A lon is the longitudinal channel control quantity δ lon The steady-state gain to the longitudinal flapping angle a1, B lat is the transverse channel control quantity δ lat The steady-state gain to the lateral flapping angle b1, A b With B a is the main rotor cross-coupling term, τ roll represents the time constant of the main rotor lateral channel, τ pitch Represents the time constant of the main rotor longitudinal channel. Ignoring the coupling terms in the flapping motion, simplifying the flapping dynamic equation, ignoring the differential terms in the kinematic equation, and simplifying it, and based on the relationship between the torque and the flapping angle, the single-input and single-output transfer function mathematical model of the horizontal lateral channel and the longitudinal channel can be simplified to:
[0092]
[0093] Where s is the Laplace variable, p(s) is the Laplace transform of the rolling angular velocity, and δ lat (s) is the Laplace transform of the lateral channel control quantity, q(s) is the Laplace transform of the pitch angular velocity, δ lon (s) is the Laplace change of the longitudinal channel control quantity. It can be converted into a differential equation expression as follows:
[0094]
[0095] in, is the second-order derivative of the roll angular velocity, is the second-order derivative of the pitch angular velocity, f roll and f pitch The parameter τ is the total disturbance on the lateral channel and longitudinal channel of the unmanned helicopter, including external interference and model uncertainty. The external interference is the interference of wind on the attitude, and the model uncertainty is the part that is not considered or simplified in the modeling process. roll , τ pitch 、ω np 、ω nq 、A lat 、A lon It can be obtained through system identification methods.
[0096] In step S2, design a tracking differentiator for the reference input:
[0097]
[0098] Among them, v is the expected input, x r1 To track the smooth output of the differentiator, x r2 is x r1 The differential signal of and is x r1 with x r2 The first derivative of . The fhan function used here is as follows:
[0099]
[0100] Among them, h0 is the step size of fhan function, which can be an integer multiple of the system sampling period, r0 is the convergence speed coefficient, the larger the value of r0, the faster x r1 The faster the response speed of tracking v, sign is the sign function:
[0101]
[0102] A tracking differentiator is used to arrange a transition process for the longitudinal channel and transverse channel angular velocity reference input of the horizontal channel to obtain smooth reference input and differential signals, thereby reducing the tracking difficulty.
[0103] In step S3, the state observer is designed as follows:
[0104] The general mathematical model of the second-order system differential equation is as follows:
[0105]
[0106] Among them, y is the system output, is the first-order derivative of y, is the second-order derivative of y, w is unknown, part b is known (the known part is b0), f u =w+(b-b0)u is the total disturbance including external disturbance and unknown information of the controlled object. Select the state variable as x1=y, x3=f, then the system is expressed as a state space expression:
[0107]
[0108] in, C =
[100] , f u The first-order derivative of . The linear extended state observer for this system is designed as follows:
[0109]
[0110] Among them, the observed state z=[z1 z2 z3] T are the angular velocity, angular acceleration, and total disturbance of the unmanned helicopter. L = [l1 l2 l3] T is the observer gain matrix to be designed. A convenient method for parameter tuning is to configure the poles of the observer's characteristic equation at the same position -ω0:
[0111] λ(s)=|sI-(A-LC)|=(s+ω0) 3
[0112] According to the coefficient relationship, the observer gain matrix is solved as follows:
[0113]
[0114] The lateral channel and longitudinal channel of the unmanned helicopter are regarded as two independent single-input and single-output channels. The coupling of the lateral channel and the longitudinal channel's swinging dynamics, which is ignored in the simplification process, is also considered as part of the disturbance, and compensation is based on the observation of the disturbance. State observers are designed for the lateral channel and the longitudinal channel respectively. According to the mathematical model of the single-input and single-output differential equations of the lateral channel and the longitudinal channel, the gain matrices of the state observers are:
[0115]
[0116] Among them, ω 0r is the pole of the transverse channel expansion state observer, which is also its bandwidth, ω 0p is the pole of the longitudinal channel expanded state observer and also its bandwidth.
[0117] In step S4, the designed extended state observer is discretized. The discretization is based on the current Euler method and uses the following formula:
[0118]
[0119] Among them, x(k) is the state at time k, x(k+1) is the state at time k+1, h is the system sampling period, and the The discrete state space expression is:
[0120] x(k+1)=(I+hA)x(k)+hBu(k)=Fx(k)+Gu(k)
[0121] Where F = I + hA, G = hB, H = C, and the discrete extended state observer is constructed as follows:
[0122]
[0123] Among them, z(k) is the estimated state at time k, z(k+1) is the estimated state at time k+1, and y d (k) is the output of the state observer, F E =F-FL c H, G E =[G FL c ], H E =IL c H, J E =[0L c ],u d (k)=[u(k)y(k)] T , is the combination of system input and output. L c The observer gain matrix places the observer poles at β:
[0124] |zI-F E |=(z-β) 3
[0125] Solve L according to the corresponding relationship of coefficients c for;
[0126]
[0127] The corresponding parameters are calculated based on the single-input single-output differential equation mathematical model of the unmanned helicopter's transverse channel and longitudinal channel. For the transverse channel, a 11 =1 / τ roll , For the longitudinal channel, a 11 =1 / τ pitch , The relationship between the poles of the discrete state observer and the poles of the continuous state observer is: ω0 is the pole of the continuous state observer and also the bandwidth of the observation signal.
[0128] Furthermore, in step S5, a nonlinear feedback law is designed based on the state calculation error observed by the state observer. The control variable is constructed using the nonlinear function fal function. The form of the fal function is as follows:
[0129]
[0130] Among them, a is the decay rate of the error, which is a constant located in [0,1]. The smaller the value of a, the faster the error decays, but at the same time it will amplify the influence of high-frequency noise. δ is the filter factor of the fal function. The larger the value of δ, the smaller the bandwidth of the fal function, the better the filtering effect on noise, and the higher the smoothness of the output, but at the same time it will increase the lag. In order to balance the filtering effect and tracking speed, 5h≤δ≤10h is taken. The essence of the fal function is to increase the gain when the error is large and reduce the gain when the error is small. Compared with the traditional low-pass filter, the fal function not only has a good filtering effect, but also has better tracking performance. Let e1=x r1 -z1,e2=x r2 -z2, where x r1 with x r2 In order to track the smooth reference input and differential signal of the differentiator, z1 and z2 are the angular velocity and angular acceleration of the single channel of the unmanned helicopter observed by the state observer. The nonlinear control law is designed as follows:
[0131] u0=k1fal(e1,a e1 ,δ)+k2fal(e2,a e2 ,δ)
[0132] Among them, k1 and k2 are the gains of angular velocity error and angular acceleration error in the control law, a e1 with a e2 is the decay rate of the angular velocity tracking error and its derivative. Based on the state observer, the total disturbance estimate z3 including external interference and model uncertainty is used to compensate the control quantity:
[0133]
[0134] The unmanned helicopter is controlled based on the model-assisted disturbance observer, and the control task is executed cyclically at a certain control frequency. In each cycle, the tracking differentiator arranges the transition process as the reference input, calculates the state error based on the smoothed reference input and the state observer's estimate of the state, calculates the preliminary control quantity based on the state error, and then compensates the preliminary control quantity based on the state observer's estimate of the disturbance to construct the final control quantity. Construct the lateral channel control quantity δ lat and longitudinal channel control quantity δ lon , acts as the input to the automatic tilter of the unmanned helicopter to control the attitude of the unmanned helicopter.
[0135] Here is a set of experimental simulations:
[0136] The parameters of the unmanned helicopter lateral channel model used in the simulation are: 1 / τ roll =9.05,ω np =36.59, A lat=9.39, the bandwidth of the transverse channel expansion state observer is ω 0r =80; longitudinal channel model parameters are: 1 / τ pitch =10.80,ω nq =20.66, A lon =6.64, the bandwidth of the longitudinal channel expansion state observer is ω 0p =40; the tracking differentiator parameters are r0=40, h0=0.02; the system sampling period h=0.005s; the lateral channel state error feedback nonlinear control law gains are 5200 and 40; the lateral channel state error feedback nonlinear control law gains are 8000 and 40.
[0137] like Figure 3 and Figure 4 As shown in the figure, the unmanned helicopter can stably, quickly and accurately track the reference input in the lateral channel, and can quickly respond to external interference to return the unmanned helicopter to a balanced position. Figure 5 and Figure 6 As shown, the unmanned helicopter can stably, quickly and accurately track the reference input in the longitudinal channel, and can respond quickly in the presence of external interference to return the unmanned helicopter to a balanced position.
[0138] Example 2
[0139] This embodiment provides a readable storage medium, wherein the readable storage medium stores a computer program. When the computer program is executed by a processor, the processor executes the unmanned helicopter attitude control method based on the model-assisted disturbance observer described in the first embodiment.
[0140] Example 3
[0141] This embodiment provides a computer device, including a processor and a memory for storing a program executable by the processor. When the processor executes the program stored in the memory, the unmanned helicopter attitude control method based on the model-assisted disturbance observer described in the first embodiment is implemented.
[0142] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be considered as equivalent replacement methods and are included in the scope of protection of the present invention.
Claims
1. A method for attitude control of an unmanned helicopter based on a model-assisted disturbance observer, characterized in that: The attitude of an unmanned helicopter is controlled using an attitude control model; the attitude control model includes a tracking differentiator, a state observer, and a nonlinear control law; the attitude control method of an unmanned helicopter includes the following steps: X1, the tracking differentiator provides a transition process for the reference input v to obtain a smooth reference input x r1 and its differential x r2 ; X2, according to the reference input x r1 and its differential x r2 Calculate the errors e1 and e2 with the angular velocity z1 and angular acceleration z2 estimated by the state observer; X3, preliminarily calculate the control variable u0 using the nonlinear control law based on the errors e1 and e2; X4, then compensate the control quantity u0 according to the total disturbance z3 estimated by the state observer to obtain the control quantity u c , as the attitude control input of the unmanned helicopter at the current moment.
2. The unmanned helicopter attitude control method based on model-assisted disturbance observer according to claim 1, characterized in that: In step X1, the reference input x r1 and its differential x r2 Calculated by the following formulas: in, and is x r1 with x r2 The first-order derivative of ; h0 is the step size of the fhan function; r0 is the convergence rate coefficient.
3. The unmanned helicopter attitude control method based on model-assisted disturbance observer according to claim 2, characterized in that: The fhan function is in the following form: Among them, d, α0, y T ,α1,α2,s y ,α,s a It is the intermediate variable in the calculation process of the fhan function; sign is the symbol function:
4. The unmanned helicopter attitude control method based on model-assisted disturbance observer according to claim 1, characterized in that: Set the estimated state z of the state observer to [z1z2z3] T are the angular velocity, angular acceleration, and total disturbance of the unmanned helicopter, respectively. The angular velocity z1 and angular acceleration z2 in step X2 and the total disturbance z3 in step X4 are calculated using the following formula: Among them, z(k) is the estimated state at time k; z(k+1) is the estimated state at time k+1; y d (k) is the output of the state observer; u d (k)=[u(k)y(k)] T , is the combination of the input u(k) and output y(k) of the unmanned helicopter; L c Observer gain matrix; F E is the state transfer matrix; G E is the input matrix; H E is the output matrix; J E is a direct transfer matrix.
5. The unmanned helicopter attitude control method based on model-assisted disturbance observer according to claim 4 is characterized in that: The F E , G E 、H E 、J E They are: F E =F-FL c H;G E =[G FL c ];H E =I-L c H;J E =[0L c ]; Where, F = I + hA; G = hB; H = C; h is the system sampling period; C＝[100]； The observer gain matrix L c for: Among them, the relationship between the poles of the discrete state observer and the poles of the continuous state observer is: ω0 is the pole of the continuous state observer; a 00 with a 11 is an element in the state transition matrix A.
6. The unmanned helicopter attitude control method based on model-assisted disturbance observer according to claim 1, characterized in that: In step X2, the errors e1 and e2 are: e1=x r1 -z1; e2=x r2 -z2。 7. The unmanned helicopter attitude control method based on model-assisted disturbance observer according to claim 1, characterized in that: In step X3, the control quantity u0 is preliminarily calculated as: u0=k1fal(e1,a e1 ,δ)+k2fal(e2,a e2 ,d); Among them, k1 and k2 are the gains of angular velocity error and angular acceleration error in the nonlinear control law respectively; a e1 with a e2 are the attenuation rates of the angular velocity tracking error and its differential respectively; the fal function is a nonlinear function; the form of the fal function is as follows: Among them, a is the decay rate of the error; δ is the filter factor of the fal function.
8. The unmanned helicopter attitude control method based on model-assisted disturbance observer according to claim 1, characterized in that: In step X4, the control quantity u c for: Among them, b0 is the compensation coefficient.
9. A readable storage medium, characterized in that The storage medium stores a computer program, which, when executed by a processor, enables the processor to execute the unmanned helicopter attitude control method based on a model-assisted disturbance observer according to any one of claims 1 to 8.
10. A computer device comprising a processor and a memory for storing a program executable by the processor, characterized in that: When the processor executes the program stored in the memory, the unmanned helicopter attitude control method based on the model-assisted disturbance observer according to any one of claims 1 to 8 is implemented.