CN108459610B - Method for inhibiting liquid sloshing during lander power descent - Google Patents

Abstract

The invention relates to a lander power descending attitude control method, in particular to a liquid shaking suppression attitude control method during lander power descending, and belongs to the technical field of spacecraft attitude control. Aiming at the problem of suppressing liquid shaking in a power descending section, the method adopts two observers with different filter coefficients to respectively estimate slowly-changing engine interference torque and the sum of the engine interference torque and the shaking interference torque. According to the historical values of the two estimated disturbance moments for a period of time, the state of the disturbance moment and the estimated state of the observer are judged, one estimated value of the disturbance moment is selected, attitude control feedforward compensation is carried out, and attitude control precision is improved.

Description

Method for inhibiting liquid sloshing during lander power descent

Technical Field

The invention relates to a lander power descending attitude control method, in particular to a liquid shaking suppression attitude control method during lander power descending, and belongs to the technical field of spacecraft attitude control.

Background

During the power descending of the detector, the guidance system gives a target attitude according to the position and speed information of the detector, the attitude control system determines the given current attitude according to the attitude, the detector is controlled to tend to the target attitude, and finally the detector is in soft landing on the surface of the moon.

When the lander propellant storage tank is a surface tension storage tank, in the landing process, besides the interference torque generated by the engine, the liquid propellant is easy to shake to generate shaking torque, so that the landing safety is endangered.

Disclosure of Invention

The technical problem to be solved by the invention is as follows: the method overcomes the defects of the prior art, provides a method for inhibiting the liquid shaking in the power descent aiming at the attitude control of the lander with the liquid propellant in the power descent period, and improves the attitude control precision.

The technical solution of the invention is as follows:

a method for restraining the dynamic descent liquid shaking of a lander can estimate the angular velocity and the disturbance moment of the lander based on a Longberger observer method, and the angular velocity coefficient and the disturbance moment coefficient of the Longberger observer are both 0-1;

when the disturbance moment coefficient of the Robert observer is larger than zero and smaller than 0.5, the estimation result is slowly-changed engine disturbance moment and corresponding angular velocity, and is defined as slow-to-interference torque and slow-to-interference angular velocity, and the coefficient of the Robert observer at the moment is defined as slow-to-interference torque coefficient and slow-to-interference torque coefficient;

when the disturbance moment coefficient of the humper observer is more than or equal to 0.5 and less than 1, the estimation result is the sum of the engine disturbance moment and the shaking disturbance moment and the corresponding angular velocity, and the sum is defined as a fast-changing disturbance moment and a fast-changing angular velocity, and the coefficient of the humper observer at the moment is defined as a fast-changing angular velocity coefficient and a fast-changing disturbance moment coefficient;

according to the statistical characteristic of the difference value of the fast-changing disturbance moment and the slow-changing disturbance moment in a period of time, determining to adopt the fast-changing disturbance moment or the slow-changing disturbance moment as a feedforward compensation moment for attitude control;

the method is based on the estimation of the angular speed and the disturbance moment of the lander by a Romberg observer, and comprises the following specific steps:

(1) setting the initial state of the lander, the angular velocity ω is changed rapidly_{_e1}(0)＝[0 0 0]^TQuickly changing disturbance moment M_{d_e1}(0)＝[0 0 0]^T(ii) a Slowly changing angular velocity omega_{_e2}(0)＝[0 0 0]^TSlow-to-break disturbance torque M_{d_e2}(0)＝[0 0 0]^T(ii) a Initial attitude control moment M_c(0)＝[0 0 0]^T；

(2) Estimating the fast changing angular speed omega in the ith (i is more than or equal to 1) control period_{_e1}(i)，ω_{_e1}(i)＝ω_{_e1}(i-1)+J^-1·(M_c(i-1)+M_{d_e1}(i-1)-cross(ω_{_e1}(i-1),J·ω_{_e1}(i-1)))·Δt+L₁₁·(ω(i)-ω_{_e1}(i-1)-J^-1·(M_c(i-1)+M_{d_e1}(i-1)-cross(ω_{_e1}(i-1),J·ω_{_e1}(i-1)))·Δt)；

Wherein: omega_{_e1}(i-1) fast upward-periodic angular velocity, M_{d_e1}(i-1) is the upper-cycle fast-varying disturbance moment, L₁₁For fast changing angular velocity coefficient, M_c(i-1) is an upper period attitude control moment, omega (i) is an ith control period, an attitude angular velocity is obtained through measurement, J is a lander inertia matrix, and delta t is a control period;

(3) estimating the fast-changing disturbance moment M in the ith control period_{d_e1}(i)，M_{d_e1}(i)＝M_{d_e1}(i-1)+L₂₁·J·(ω(i)-ω_{_e1}(i-1)-J^-1·(M_c(i-1)+M_{d_e1}(i-1)-cross(ω_{_e1}(i-1),J·ω_{_e1}(i-1)))·Δt)/Δt；

Wherein: l is₂₁The interference moment coefficient is changed rapidly;

(4) estimating the slow change angular velocity omega in the ith control period_{_e2}(i)，ω_{_e2}(i)＝ω_{_e2}(i-1)+J^-1·(M_c(i-1)+M_{d_e2}(i-1)-cross(ω_{_e2}(i-1),J·ω_{_e2}(i-1)))·Δt+L₁₂·(ω(i)-ω_{_e2}(i-1)-J^-1·(M_c(i-1)+M_{d_e2}(i-1)-cross(ω_{_e2}(i-1),J·ω_{_e2}(i-1)))·Δt)；

Wherein: omega_{_e2}(i-1) upper-period slow variation of angular velocity, M_{d_e2}(i-1) Upper-cycle Slow Dry disturbance Torque, L₁₂Is a slow varying angular velocity coefficient;

(5) estimating the slow-to-interference torque M in the ith control period_{d_e2}(i)，M_{d_e2}(i)＝M_{d_e2}(i-1)+L₂₂·J·(ω(i)-ω_{_e2}(i-1)-J^-1·(M_c(i-1)+M_{d_e2}(i-1)-cross(ω_{_e2}(i-1),J·ω_{_e2}(i-1)))·Δt)/Δt；

Wherein: l is₂₂Is the slow-to-break disturbance torque coefficient;

(6) ith controlIn a period, obtaining a torque difference historical value matrix dM_deD matrix dM_deIs 3 lines N_wMatrix of columns, N_wCounting the number of cycles; the method comprises the step of obtaining M in the step (3)_{d_e1}(i) Subtracting M of step (5)_{d_e2}(i) Difference value M of_{d_e1}(i)-M_{d_e2}(i) Assigning to a torque difference historical value matrix dM_deMod (i, N) of_w) Columns, wherein: the function mod () is a remainder function;

(7) in the ith control cycle, if i<N_wAnd then: feedforward compensation moment DeltaM_d(i)＝[0 0 0]^T；

(8) In the ith control period, if i is more than or equal to N_wAnd then: calculating the mean value of the rolling moment difference Meanx (i),

the rolling moment difference variance stdx (i) is calculated,

wherein: j is 1 to N_w(ii) a If | dM_de(1,i)-Meanx(i)|>stdx (i), then: will M_{d_e2}(i) Is assigned to Δ M_d(i) The 1 st element of (a); if | dM_de(1, i) -Meanx (i) < stdx (i), then: will M_{d_e1}(i) Is assigned to Δ M_d(i) The 1 st element of (a);

(9) in the ith control period, if i is more than or equal to N_wAnd then: calculating the mean value of the difference between the pitching moments mean (i),

the difference variance stdy (i) of the pitching moment differences is calculated,

wherein: j is 1 to N_w(ii) a If | dM_de(2,i)-Meany(i)|>stdy (i), then: will M_{d_e2}(i) Is assigned to Δ M_d(i) The 2 nd element of (1); if | dM_de(2, i) -means (i) < stdy (i), then: will M_{d_e1}(i) Is assigned to Δ M_d(i) The 2 nd element of (1);

(10) in the ith control period, if i is more than or equal to N_wAnd then: calculating the mean value of the yaw moment difference Meanz (i),

calculating the yaw moment difference variance stdz (i),

wherein: j is 1 to N_w(ii) a If | dM_de(3,i)-Meanz(i)|>stdz (i), then: will M_{d_e2}(i) Is assigned to Δ M_d(i) The 3 rd element of (a); if | dM_de(3, i) -Meanz (i) ≦ stdz (i), then: will M_{d_e1}(i) Is assigned to Δ M_d(i) The 3 rd element of (a);

(11) in the ith control period, calculating attitude control moment M_c(i)，M_c(i)＝T_PID(i)+ΔM_d(i) Wherein: t is_PID(i) The torque is controlled by PID.

Compared with the prior art, the invention has the beneficial effects that:

aiming at the problem of liquid shaking inhibition in a power descending section, the attitude control method adopts a humper observer with different filter coefficients to respectively estimate slowly-changing engine interference torque and the sum of the engine interference torque and shaking interference torque. According to the historical values of the two estimated disturbance moments for a period of time, the state of the disturbance moment and the estimated state of the observer are judged, one estimated value of the disturbance moment is selected, attitude control feedforward compensation is carried out, and attitude control precision is improved. The invention discloses a lander dynamic descent attitude control method, which aims at the problem of liquid sloshing inhibition in a dynamic descent section and adopts two observers with different filter coefficients to respectively estimate slowly-changing engine interference torque and the sum of the engine interference torque and the sloshing interference torque. According to the historical values of the two estimated disturbance moments for a period of time, the state of the disturbance moment and the estimated state of the observer are judged, one estimated value of the disturbance moment is selected, attitude control feedforward compensation is carried out, and attitude control precision is improved.

Detailed Description

The invention has the following implementation steps:

(1) setting the initial state of the lander, the angular velocity ω is changed rapidly_{_e1}(0)＝[0 0 0]^TQuickly changing disturbance moment M_{d_e1}(0)＝[0 0 0]^T(ii) a Slowly changing angular velocity omega_{_e2}(0)＝[0 0 0]^TSlow-to-break disturbance torque M_{d_e2}(0)＝[0 0 0]^T(ii) a Initial attitude control moment M_c(0)＝[0 0 0]^T；

(2) Estimating the fast changing angular speed omega in the ith (i is more than or equal to 1) control period_{_e1}(i)，ω_{_e1}(i)＝ω_{_e1}(i-1)+J^-1·(M_c(i-1)+M_{d_e1}(i-1)-cross(ω_{_e1}(i-1),J·ω_{_e1}(i-1)))·Δt+L₁₁·(ω(i)-ω_{_e1}(i-1)-J^-1·(M_c(i-1)+M_{d_e1}(i-1)-cross(ω_{_e1}(i-1),J·ω_{_e1}(i-1)))·Δt)；

Wherein: omega_{_e1}(i-1) fast upward-periodic angular velocity, M_{d_e1}(i-1) is the upper-cycle fast-varying disturbance moment, L₁₁For fast changing angular velocity coefficient, M_c(i-1) is an upper period attitude control moment, omega (i) is an ith control period, an attitude angular velocity is obtained through measurement, J is a lander inertia matrix, and delta t is a control period;

(3) estimating the fast-changing disturbance moment M in the ith control period_{d_e1}(i)，M_{d_e1}(i)＝M_{d_e1}(i-1)+L₂₁·J·(ω(i)-ω_{_e1}(i-1)-J^-1·(M_c(i-1)+M_{d_e1}(i-1)-cross(ω_{_e1}(i-1),J·ω_{_e1}(i-1)))·Δt)/Δt；

Wherein: l is₂₁The interference moment coefficient is changed rapidly;

(4) estimating the slow change angular velocity omega in the ith control period_{_e2}(i)，ω_{_e2}(i)＝ω_{_e2}(i-1)+J^-1·(M_c(i-1)+M_{d_e2}(i-1)-cross(ω_{_e2}(i-1),J·ω_{_e2}(i-1)))·Δt+L₁₂·(ω(i)-ω_{_e2}(i-1)-J^-1·(M_c(i-1)+M_{d_e2}(i-1)-cross(ω_{_e2}(i-1),J·ω_{_e2}(i-1)))·Δt)；

Wherein: omega_{_e2}(i-1) upper-period slow variation of angular velocity, M_{d_e2}(i-1) Upper-cycle Slow Dry disturbance Torque, L₁₂Is a slow varying angular velocity coefficient;

(5) estimating the slow-to-interference torque M in the ith control period_{d_e2}(i)，M_{d_e2}(i)＝M_{d_e2}(i-1)+L₂₂·J·(ω(i)-ω_{_e2}(i-1)-J^-1·(M_c(i-1)+M_{d_e2}(i-1)-cross(ω_{_e2}(i-1),J·ω_{_e2}(i-1)))·Δt)/Δt；

Wherein: l is₂₂Is the slow-to-break disturbance torque coefficient;

(6) in the ith control period, obtaining a torque difference historical value matrix dM_deD matrix dM_deIs 3 lines N_wMatrix of columns, N_wCounting the number of cycles; the method comprises the step of obtaining M in the step (3)_{d_e1}(i) Subtracting M of step (5)_{d_e2}(i) Difference value M of_{d_e1}(i)-M_{d_e2}(i) Assigning to a torque difference historical value matrix dM_deMod (i, N) of_w) Columns, wherein: the function mod () is a remainder function;

(7) in the ith control cycle, if i<N_wAnd then: feedforward compensation moment DeltaM_d(i)＝[0 0 0]^T；

(8) In the ith control period, if i is more than or equal to N_wAnd then: calculating the mean value of the rolling moment difference Meanx (i),

the rolling moment difference variance stdx (i) is calculated,

wherein: j is 1 to N_w(ii) a If | dM_de(1,i)-Meanx(i)|>stdx (i), then: will M_{d_e2}(i) Is assigned to Δ M_d(i) The 1 st element of (a); if | dM_de(1, i) -Meanx (i) < stdx (i), then: will M_{d_e1}(i) Is assigned to Δ M_d(i) The 1 st element of (a);

(9) the ithWhen controlling the period, if i is more than or equal to N_wAnd then: calculating the mean value of the difference between the pitching moments mean (i),

the difference variance stdy (i) of the pitching moment differences is calculated,

wherein: j is 1 to N_w(ii) a If | dM_de(2,i)-Meany(i)|>stdy (i), then: will M_{d_e2}(i) Is assigned to Δ M_d(i) The 2 nd element of (1); if | dM_de(2, i) -means (i) < stdy (i), then: will M_{d_e1}(i) Is assigned to Δ M_d(i) The 2 nd element of (1);

(10) in the ith control period, if i is more than or equal to N_wAnd then: calculating the mean value of the yaw moment difference Meanz (i),

calculating the yaw moment difference variance stdz (i),

wherein: j is 1 to N_w(ii) a If | dM_de(3,i)-Meanz(i)|>stdz (i), then: will M_{d_e2}(i) Is assigned to Δ M_d(i) The 3 rd element of (a); if | dM_de(3, i) -Meanz (i) ≦ stdz (i), then: will M_{d_e1}(i) Is assigned to Δ M_d(i) The 3 rd element of (a);

(11) in the ith control period, calculating attitude control moment M_c(i)，M_c(i)＝T_PID(i)+ΔM_d(i) Wherein: t is_PID(i) The torque is controlled by PID.

Examples

(1) Lander inertia matrix

The control period Δ t is 0.128. Fast changing angular velocity coefficient L₁₁0.9, fast-varying disturbance moment coefficient L₂₁0.5, slow change angular velocity coefficient L₁₂＝0.9，Slow-dry disturbance torque coefficient L₂₂0.02. Counting the number of cycles N_w＝50。

(2) Controlling the number of cycles to be less than N_wTake the control cycle number equal to 1 as an example. Estimating fast varying angular velocity ω_{_e1}(1)＝[0.009 0.18 0.27]^TQuick change interference moment M_{d_e1}(1)＝[0.17 2.62 3.93]^TSlowly changing angular velocity omega_{_e2}(1)＝[0.009 0.18 0.27]^TSlow-to-disturbance torque M_{d_e2}(1)＝[0.007 0.1 0.16]^T(ii) a Moment difference historical value matrix dM_deColumn 1 of (2) is [ 0.1632.523.77]^T(ii) a Calculating attitude control moment M_c(1)＝[0.01 0.07 0.08]^T。

(3) Controlling the number of cycles to be greater than or equal to N_wTake the control cycle number equal to 50 as an example. Fast changing angular velocity omega_{_e1}(50)＝[0.05 0.22 0.32]^TQuickly changing disturbance moment M_{d_e1}(50)＝[0.05 15 26]^TSlowly changing angular velocity omega_{_e2}(50)＝[0.06 0.2 0.3]^TSlow-to-break disturbance torque M_{d_e2}(50)＝[0.02 10 20]^T(ii) a Moment difference historical value matrix dM_deColumn 50 of (2) is [ 0.0356]^T(ii) a Rolling moment difference mean value Meanx (50) is 0.02, rolling moment difference variance stdx (50) is 0.06, pitching moment difference mean value Meany (50) is 0.2, pitching moment difference variance stdy (50) is 10, yawing moment difference mean value Meanz (50) is 0.3, yawing moment difference variance stdz (50) is 11, Δ M_d(50)＝[0.05 15 25]^TPID control moment T_PID(50)＝[0.01 0.4 0.5]^T. Attitude control moment M_c(50)＝[0.06 15.4 25.5]^T. Mathematical simulation shows that the attitude control moment acts on the lander to inhibit the liquid of the propellant of the lander from shaking, and high-precision attitude control in the process of power descent is realized. The angular velocity is given in degrees/s and the moment in Nm.

Those skilled in the art will appreciate that those matters not described in detail in the present specification are well known in the art.

Claims (10)

1. A method for inhibiting the liquid sloshing of lander power descending is characterized by comprising the following specific steps:

(1) setting the initial state of the lander, the angular velocity ω is changed rapidly_{_e1}(0)＝[0 0 0]^TQuickly changing disturbance moment M_{d_e1}(0)＝[0 0 0]^T(ii) a Slowly changing angular velocity omega_{_e2}(0)＝[0 0 0]^TSlow-to-break disturbance torque M_{d_e2}(0)＝[0 0 0]^T(ii) a Initial attitude control moment M_c(0)＝[0 0 0]^T；

(2) In the ith control period, a fast-changing angular velocity omega is estimated by using a Romberg observer_{_e1}(i) A value of (d); i is more than or equal to 1;

(3) in the ith control period, a Romberg observer is used for estimating the fast-changing disturbance moment M_{d_e1}(i) A value of (d);

(4) in the ith control period, a Romberg observer is used for estimating the slow-changing angular velocity omega_{_e2}(i) A value of (d);

(5) in the ith control period, the slow-drying disturbance moment M is estimated by using a Romberg observer_{d_e2}(i) A value of (d);

(6) in the ith control period, obtaining a torque difference historical value matrix dM_de；

(7) In the ith control cycle, if i<N_wAnd then: feedforward compensation moment DeltaM_d(i)＝[0 0 0]^T；N_wCounting the number of cycles;

(8) in the ith control period, if i is more than or equal to N_wAnd then: calculating a rolling moment difference mean value Meanx (i); calculating a rolling moment difference variance stdx (i); wherein: j is 1 to N_w(ii) a If | dM_de(1,i)-Meanx(i)|>stdx (i), then: will M_{d_e2}(i) Is assigned to Δ M_d(i) The 1 st element of (a); if | dM_de(1, i) -Meanx (i) < stdx (i), then: will M_{d_e1}(i) Is assigned to Δ M_d(i) The 1 st element of (a);

(9) in the ith control period, if i is more than or equal to N_wAnd then: calculating a mean value mean (i) of the differences of the pitching moments; calculating a pitching moment difference variance stdy (i); wherein: j is 1 to N_w(ii) a If | dM_de(2,i)-Meany(i)|>stdy (i), then: will M_{d_e2}(i) Is assigned to Δ M_d(i) The 2 nd element of (1); if | dM_de(2, i) -means (i) < stdy (i), then: will M_{d_e1}(i) Is assigned to Δ M_d(i) The 2 nd element of (1);

(10) in the ith control period, if i is more than or equal to N_wAnd then: calculating a yaw moment difference mean value Meanz (i); calculating a yaw moment difference variance stdz (i); wherein: j is 1 to N_w(ii) a If | dM_de(3,i)-Meanz(i)|>stdz (i), then: will M_{d_e2}(i) Is assigned to Δ M_d(i) The 3 rd element of (a); if | dM_de(3, i) -Meanz (i) ≦ stdz (i), then: will M_{d_e1}(i) Is assigned to Δ M_d(i) The 3 rd element of (a);

(11) in the ith control period, calculating attitude control moment M_c(i)。

2. The landing gear power-down liquid sloshing suppression method according to claim 1, wherein: in the step (2), the angular velocity ω is rapidly changed_{_e1}(i) The values of (A) are:

ω_{_e1}(i)＝ω_{_e1}(i-1)+J^-1·(M_c(i-1)+M_{d_e1}(i-1)-cross(ω_{_e1}(i-1),J·ω_{_e1}(i-1)))·Δt+L₁₁·(ω(i)-ω_{_e1}(i-1)-J^-1·(M_c(i-1)+M_{d_e1}(i-1)-cross(ω_{_e1}(i-1),J·ω_{_e1}(i-1)))·Δt)；

wherein: omega_{_e1}(i-1) fast upward-periodic angular velocity, M_{d_e1}(i-1) is the upper-cycle fast-varying disturbance moment, L₁₁For fast changing angular velocity coefficient, M_cAnd (i-1) is an upper period attitude control moment, omega (i) is an ith control period, the attitude angular velocity is obtained through measurement, J is a lander inertia matrix, and delta t is a control period.

3. The landing gear power-down liquid sloshing suppression method according to claim 2, wherein: in the step (3), the disturbance moment M is changed rapidly_{d_e1}(i) The values of (A) are:

M_{d_e1}(i)＝M_{d_e1}(i-1)+L₂₁·J·(ω(i)-ω_{_e1}(i-1)-J^-1·(M_c(i-1)+M_{d_e1}(i-1)-cross(ω_{_e1}(i-1),J·ω_{_e1}(i-1)))·Δt)/Δt；

wherein: l is₂₁The fast-changing disturbance moment coefficient.

4. The landing gear power-down liquid sloshing suppression method according to claim 1, wherein: in the step (4), the angular velocity omega is slowly changed_{_e2}(i) The values of (A) are: omega_{_e2}(i)＝ω_{_e2}(i-1)+J^-1·(M_c(i-1)+M_{d_e2}(i-1)-cross(ω_{_e2}(i-1),J·ω_{_e2}(i-1)))·Δt+L₁₂·(ω(i)-ω_{_e2}(i-1)-J^-1·(M_c(i-1)+M_{d_e2}(i-1)-cross(ω_{_e2}(i-1),J·ω_{_e2}(i-1)))·Δt)；

Wherein: omega_{_e2}(i-1) upper-period slow variation of angular velocity, M_{d_e2}(i-1) Upper-cycle Slow Dry disturbance Torque, L₁₂Is a slowly varying angular velocity coefficient.

5. The landing gear power-down liquid sloshing suppression method according to claim 4, wherein: in the step (5), the slow-drying disturbance torque M_{d_e2}(i) The values of (A) are:

M_{d_e2}(i)＝M_{d_e2}(i-1)+L₂₂·J·(ω(i)-ω_{_e2}(i-1)-J^-1·(M_c(i-1)+M_{d_e2}(i-1)-cross(ω_{_e2}(i-1),J·ω_{_e2}(i-1)))·Δt)/Δt；

wherein: l is₂₂Is the slow-to-disturb torque coefficient.

6. The landing gear power-down liquid sloshing suppression method according to claim 1, wherein: in the step (6), the historical value matrix dM_deIs 3 lines N_wMatrix of columns, N_wCounting the number of cycles; the method comprises the step of obtaining M in the step (3)_{d_e1}(i) Subtracting M of step (5)_{d_e2}(i) Difference value M of_{d_e1}(i)-M_{d_e2}(i) Assigning to a torque difference historical value matrix dM_deMod (i, N) of_w) Columns, wherein: function mod () is a remainder function.

7. The landing gear power-down liquid sloshing suppression method according to claim 1, wherein: in the step (8), the rolling moment difference mean value meanx (i) is:

the rolling moment difference variance stdx (i) is:

8. the landing gear power-down liquid sloshing suppression method according to claim 1, wherein: in the step (9), the mean value of the difference between the pitching moments mean (i) is:

the pitching moment difference variance stdy (i) is:

9. the landing gear power-down liquid sloshing suppression method according to claim 1, wherein: in the step (10), the mean yaw moment difference mean value meanz (i) is:

the yaw moment difference variance stdz (i) is:

10. the landing gear power-down liquid sloshing suppression method according to claim 1, wherein: the steps are(11) In, the attitude control moment is: m_c(i)＝T_PID(i)+ΔM_d(i) Wherein: t is_PID(i) The torque is controlled by PID.