KR20080103228A - Flexible launch vehicle ground test-bed and adaptive controller - Google Patents

Abstract

A flexible launch vehicle ground test-bed and adaptive controller is provided to realize the performance test about various controlling method which is impossible to perform actual repetitive experiment, so it is actually applied to a projectile. A flexible launch vehicle ground test-bed and adaptive controller includes a projectile model copying the shape of projectile(11), a fuel acceptance model copying the fuel of making the projectile model promoted(12), a thrust vector module for controlling the rotation of the projectile model, an angular velocity detection sensor installed between the projectile model and the fuel acceptance model, an air pad for removing the frictional force due to the structure support(15a,15b), and an apparatus having for copying the disturbance of the projectile model.

Description

Flexible Launch Vehicle Ground Test-bed and Adaptive Controller

Figure 1a shows a schematic form of a projectile ground test apparatus according to the present invention.

Figure 1b shows a modeling form for mathematical analysis of the projectile ground test apparatus according to the present invention.

2A and 2B illustrate an embodiment of a notch filter that can be applied to a projectile ground test apparatus according to the present invention.

Figure 3 illustrates an embodiment of a non-linear adaptive controller for the control of the projectile ground test apparatus according to the present invention.

Figure 4 shows an embodiment of a test apparatus for the control of the projectile ground test apparatus according to the present invention.

Figures 5a, 5b and 5c graphically shows the control test results for the projectile ground test apparatus according to the present invention.

The present invention relates to a flexible space projectile ground test apparatus having a thruster vector launcher, and specifically, a flexible space projectile capable of analyzing the performance change of the flexible space vehicle control system according to the relative position of the sensor and the driving apparatus. It relates to a ground test apparatus.

A prior invention on projectiles is US Pat. No. 5,529,264. The projectile system of the prior invention comprises at least one main solid fuel motor, a shroud for receiving the payload and an attitude control system. And the attitude control system has a circular outer wall and a liquid fuel motor for propulsion along the longitudinal axis of the rocket system.

The attitude control command of the liquid propulsion rocket can be calculated from the attitude and angular velocity measured in the inertial device located in the rocket mounted portion including the attitude data according to the rigid motion of the rocket and the rotational motion element by elastic deformation. However, a problem has been raised that the sensor and the driving device for the attitude data are located at different places, thereby increasing the attitude control instability. Therefore, in the design process of the rocket attitude control system, the attitude control instability due to such structural feedback should be taken into consideration. For this purpose, the filter research has been conducted to minimize the effect of structural vibration near the natural frequency with very large deformation. However, despite the progress of the study of the controller design technique, it is difficult to repeat the experiment through actual flight due to the characteristics of the projectile, and the study of the control technique was analyzed based only on the results of computer simulation.

The present invention is to propose a ground test apparatus for modeling a flexible structure projectile and reviewing the controller design technique and analyzing the dynamic characteristics and control system performance of the projectile through experiments. Have

An object of the present invention is to provide a projectile ground test apparatus capable of testing the dynamic characteristics of the attitude control system of the flexible space projectile and the performance of the control technique.

Another object of the present invention is to provide an adaptive controller capable of testing the performance of a flexible projectile ground test apparatus.

Another object of the present invention is to provide a flexible projectile ground test performance system.

According to a preferred embodiment of the present invention, a flexible projectile ground testing apparatus includes a projectile model that simulates the shape of an actual projectile; A fuel accommodating model that simulates fuel for propelling a projectile model; A thrust vector module for controlling the rotation of the projectile motel; An angular velocity sensor mounted between the projectile model and the fuel receiving model; An air pad to remove frictional forces due to the structure support; And an excitation device for simulating the disturbance of the projectile model, wherein a hose capable of supplying and discharging liquid at a constant speed is connected to the fuel receiving model and the air pad is operated by compressed air.

According to another suitable embodiment of the invention, the thrust vector module comprises a compressed air hose whose direction and displacement are adjusted by the adjustment of the motor.

According to another suitable embodiment of the present invention, a method for testing the performance of a flexible projectile ground test apparatus that simulates a projectile includes measuring the angular velocity to obtain a natural frequency with a multi-mode notch filter to filter the angular velocity signal; And calculating the output signal at the filtered angular and angular velocity to obtain a moment and rotating the flexible projectile ground test apparatus, wherein the natural frequency is measured in real time so that the gain value of the control command is optimized through a neural network algorithm. do.

According to another suitable embodiment of the present invention, an apparatus for testing the performance of a flexible projectile ground test apparatus includes: a flexible projectile ground test model that simulates a projectile; An angular velocity measuring device for measuring an angular velocity of the flexible projectile ground test model; A signal conversion board for receiving a signal of the angular velocity measuring device; An attitude control computer having an adaptive notch filter and an adaptive controller for receiving a signal from the signal conversion board and generating a control command; A thrust vector driver for driving thrust of the flexible projectile ground test motel according to a command signal from the signal change board; And a disturbance simulation apparatus for simulating the disturbance of the flexible projectile ground test model.

According to another suitable embodiment of the present invention, the angular velocity is input to the signal conversion board at 200 Hz.

According to another suitable embodiment of the present invention, a ground test apparatus for a liquid-propelled rocket having an elongate structure includes a projectile model that simulates the shape of an actual projectile; A transparent vinyl jacket that receives and discharges water at a constant speed by a hose; A plurality of air exhaust holes located on top of the transparent vinyl jacket; A thrust vector module for controlling the rotation of the projectile motel; An angular velocity sensor mounted between the projectile model and the fuel receiving model; An air pad to remove frictional forces due to the structure support; And an excitation device for simulating the disturbance of the projectile model, wherein the ground test device can be modeled with a center of rotation, flexible beam, and thrust vector control module.

According to another suitable embodiment of the present invention, the rate of water supply and discharge through the hose is from 0.02 kg to 0.06 kg per second.

According to another suitable embodiment of the present invention, a ground test apparatus for performance analysis of a control system of a liquid propulsion rocket is provided.

The invention is described in detail below using the accompanying drawings and examples. The examples presented are exemplary and are not intended to limit the scope of the invention.

The projectile testing device must meet flight boundary conditions. The following two conditions were set as flight boundary conditions.

(i) there shall be no frictional forces due to the support of the structure, except for aerodynamic forces, in the event of vibration and posture change of the structure; And

(ii) Easily attach and support additional structures, such as thrust vector control modules for research, or reaction wheels to simulate the forces causing disturbances on projectiles.

In order to meet these conditions, a device such as that shown in FIG. 1A was designed.

Figure 1a shows a schematic form of a projectile ground test apparatus according to the present invention.

1A, a test apparatus according to the present invention includes a projectile model 11 simulating the shape of an actual projectile; A fuel accommodating model 12 that simulates fuel for propelling the projectile model 11 in the firing process; A motor 13 for controlling the thrust direction of the projectile model 11; And a thrust model 131 which generates thrust in the projectile model 11 while discharging the fuel of the fuel accommodating model 12.

The projectile model 11 is supported by a pivot and installed so that the direction can be changed freely according to vibration or thrust control transmitted from the outside. The projectile model 11 can be varied in dynamics by various factors and this change in dynamics is sensed by the angular velocity sensor 18 installed between the projectile model 11 and the fuel acceptance model 12. As already described above, air pads 15a and 15b may be installed so that the frictional force due to the support of the structure other than the air force does not act on the test apparatus. In addition, an appropriate transparent vinyl packet may be installed in the fuel accommodating model 12 to simulate the consumption of fuel during the test process.

Water that mimics the liquid fuel of the fuel containment model 12 may be stored in a transparent vinyl packet and the transparent vinyl packet is connected to a hose 122 that regulates emissions by an electric pump. The hose 122 is similarly used when water is supplied to the transparent plastic packet of the fuel accommodating model 12. When water is supplied to the fuel accommodating model 12, since air exists inside the transparent vinyl packet, the internal pressure increases as the water is filled. Therefore, the air discharge hole 121 is formed in the upper end of the transparent vinyl packet. When an experiment is conducted to simulate the test of the projectile, the water supply and discharge rate through the hose 122 may be 0.02 to 0.06 kg per second. The air pads 15a and 15b are operated by the compressed air supplied through the compressed air hose 151, and four of the two air pads are installed at the front and rear in FIG. Can be installed. Air pads 15a and 15b support the projectile ground test apparatus to prevent friction with the ground. Since the actual projectile is disturbed during the propulsion process, the excitation device 14 is installed in the projectile model 11 to simulate the disturbance of the actual projectile. The thrust model 131 may be made of a compressed air hose in which the direction and the discharge amount are adjusted by the adjustment of the motor 13 to be operated by compressed air.

In order to confirm whether the projectile ground test apparatus manufactured as described above is similar to the actual small satellite projectile and the performance of the test apparatus, a vibration test was performed. For the vibration test was modeled as shown in Figure 1b for the analysis according to the finite analysis method of the projectile ground test apparatus.

The projectile model 11 has a moment of inertia; In addition, since the fuel accommodating model 12 may be represented by linear density, respectively, the projectile model 11 may include a center of rotation 21 and the fuel accommodating model 12 may include a flexible beam 22 for mathematical analysis. The motor 13 and the propeller model 131 may be represented by a single thrust vector control module 23. The manufactured projectile ground test apparatus was discretized to four material points (m ₁ , m ₂ , m ₃ and m _t ) and modeled with the following constant values.

Linear density of the flexible beam ρ _m = 0.2 to 0.5 kg / m; Strain per unit length E = 30 Gpa; Cross section first moment value I = 7 × 10 ^-11 m ⁴ ; Moment of inertia at center of rotation I _c = 0 kg㎡, mass of the end of the flexible beam m _t = 0.3 kg; Radius of rotation center l ₀ = 0 m; Length of flexible beam l = 1.3 m

For the projectile ground test apparatus manufactured by applying such constant conditions, natural frequency was obtained through an analysis study on vibration mode and an actual vibration test, and the results are shown in Table 1.

Table 1: Natural Frequency

Analytical model Vibration experiment model Initial state (ρ _m = 0.5 kg / m) Final state (ρ _m = 0.2 kg / m) Initial state Final status First mode 1.5 2.0 2.8 3.8 Second mode 8.3 12.8 7.8 9.3 Third mode 13.9 26.9 15.8 18.0

In Table 1, the analytical model refers to a numerical model constructed by assuming that the ground test apparatus is a simple beam. The first, second and third modes mean natural frequency modes of the configured experimental apparatus. ρ _m can have a maximum density of 0.5 kg / m when filled with water at the linear density of the flexible beam and at least 0.2 kg / m without water.

In Table 1, the initial state is that the weight of the test device is heavy due to the liquid that simulates the projectile fuel, and the final state is light weight of the test device due to the ejection of liquid. Indeed, in the case of projectiles, the mass changes drastically, but the manufactured projectile ground test apparatus showed a difference of about 1 Hz in the first bending mode and 1.5 Hz in the secondary mode due to the limitation of the test equipment.

A nonlinear adaptive controller was designed and applied for the attitude control of the projectile ground test apparatus. The nonlinear adaptive controller can be configured with a notch filter. First, the multi-mode adaptive notch filter applied to the projectile ground test apparatus according to the present invention will be described.

adaptation Notch  Filter design

An adaptive notch filter is used to compensate for the shortcomings of the notch filter sensitive to modeling error. The adaptive notch filter uses the sensor signal to accurately estimate the frequency of the real system and the design parameters of the notch filter are updated in real time to accurately estimate the constant of the real system.

An autoregressive moving average (ARMA) model is known as a filter structure for estimating the frequency of the input of the multiple sinusoidal signals. The coefficient estimation method of the ARMA model can be obtained by the nonlinear least-squares method, but it has the disadvantage that it can be wrongly converged to the local minimum and it takes a lot of computation time. To compensate for these drawbacks, multi-block notch frequencies are estimated by arranging blocks of second-order notch filters. 2A illustrates an embodiment of a multi-mode notch filter.

2A, μ (n) is the input function; y _k (n) is an output function; H _k (z ⁻¹ ) represents a network function for a plurality of frequencies z, and y _k (n) is fed back to the k-th block. You can calculate the natural frequencies of the projectile ground tester model by selecting the network function, but the convergence of the kth block is updated separately from the block after k so that the notch frequency converged in the block after k is again the kth block. The problem arises that we can converge at. In addition, a problem arises that the notch frequencies that are well converging with each other may converge again. In order to solve this problem, a multi-mode notch filter of FIG. 2B is used. As shown in Fig. 2b, the output value of the notch block of the final stage is fed back to each block, reducing the occurrence of the problem raised above.

Nonlinear Adaptive Controller Design

The nonlinear adaptive controller, which is the attitude controller of the projectile ground test apparatus, may be designed by a method of gain scheduling when the coefficient of the system changes with time. However, the presence of modeling errors in the application process can lead to system instability. To prevent this, the nonlinear adaptive controller is designed to include the multiple adaptive notch filter described above.

3 schematically illustrates a configuration of a nonlinear adaptive controller including multiple adaptive notch filters.

In general, the projectile control algorithm obtains the model as accurately as possible and designs it by the gain planning method in advance. However, in the case of the controller for the experiment of the model according to the present invention, the controller is designed to be robust to the model error by estimating the natural frequency of the model in real time. The controller includes a controller and a filter, and the controller optimizes the gain value of the controller using a neural network algorithm when the model changes using the neural network to sufficiently follow the instruction. Also, in flexible structures, if the driver and sensor are not placed in the same place, phase difference may occur in the signal of the sensor and the controller may become unstable. In this case, the gain and phase margin of the controller can be reduced in the portion corresponding to the natural frequency of the structure. In the case of the controller according to the present invention, the gain and phase margin are recovered by removing the signal using a notch filter. In the controller according to the present invention, the adaptive notch filter is applied to automatically follow the model error when the natural frequency changes in real time. Looking at the control process for the terrestrial experiment apparatus, first, the signal obtained from the angular velocity sensor 31 of the experimental apparatus is received to obtain the first and second natural frequencies using the multi-mode notch filter 32. In the case of the controller according to the present invention, the block of the multi-mode notch filter 32 obtains one natural frequency, and the more the block is used, the higher the natural frequency of the higher mode of the experimental apparatus can be estimated. The angular velocity signal is filtered by applying the second-order notch filter using the obtained natural frequency. The filtered angular velocity information is used as an input of the controller 33, and the angular velocity information is integrated using the angular velocity information. The controller 33 calculates the control output using the filtered angle and angular velocity information. According to the control command corresponding to the calculated output, the motor used as the driver by the thrust vector angle positions the thrust vector module 34, and the product of the thrust vector force and the distance from the center of rotation acts as a moment to rotate the experimental apparatus. Let's go.

Application of nonlinear adaptive controller

In order to apply the non-linear adaptive controller described above to the projectile ground test apparatus according to the present invention, the test apparatus is configured as shown in FIG.

Referring to FIG. 4, the rotational speed of the body of the flexible projectile ground testing device 45 is measured by the precision angular velocity sensor 47 and read at 200 Hz on the signal conversion board 43. The signal conversion board 43 means A / D (analog / digital) and D / A (digital / analog) board 43. The angle information is integrated by using the angular velocity sensor 47 and the control computer including the adaptive notch filter 41 and the adaptive controller 42 performs calculation of the adaptive controller 42 using the adaptive notch filter 41 and the neural network. do. The thrust vector driver 44 uses the thrust vector control (TVC) method and sends a control command of the TVC driver 44 at 25 Hz from the control computer. A disturbance simulation device 46 that simulates external disturbances using an amplifier and an exciter simulates the actual projectile disturbance and delivers it to the flexible projectile ground test device 45. The thrust vector angle calculated by the adaptive controller 42 is input to the input of the other flexible projectile ground test apparatus 45. The thrust vector driver 44 positions the thrust vector module at the calculated command angle. Accordingly, a corresponding moment is generated to cause the flexible projectile ground test device 45 to rotate, and the rotation speed is output again, and is measured by the angular velocity measuring device 47.

result

The test results are shown in FIGS. 5A, 5B and 5C.

Figure 5a shows the experimental results composed of a notch filter and the PD-type controller. Notch filter coefficients initially had an error. Actually applied notch filter frequency is 2.4Hz and 6.7Hz, which has initial error of about 14%. As can be seen from the experimental result of FIG. 5A, the angular error and the angular velocity do not stabilize as time passes. This is the result of incorrect notch filter design. The error of the initial notch filter is large due to the increase of the bending mode frequency of the system over time, so the system becomes unstable over time. In order to supplement the above experimental results, a nonlinear adaptive controller using an adaptive notch filter and a neural network was tested and the results are shown in FIG. 5B. As in the previous case, a sine wave with a period of 12 seconds from the initial stop state was used as a command. Initial 1st and 2nd notch frequency is 1.5Hz and designed to have error more than 46%. As shown in FIG. 5B, when the adaptive controller is used, the posture error is small and the vibration phenomenon is eliminated in the angular velocity signal. Unlike the PD controller using the notch filter, the system is stabilized. Figure 5c shows the bending mode frequency of the estimated system. It can be seen that the frequency of the adaptive notch filter after 10 seconds actually estimates the bending mode frequency of the real terrestrial model.

According to the present invention, a flexible projectile ground test apparatus was manufactured in consideration of mass changes similar to an actual projectile. And the ground test equipment manufactured was performance tested with nonlinear controller. The present invention allows the performance test for various control techniques that are difficult to repeat the experiment through real flight. And the result of the performance test has the advantage that it can be applied to the actual projectile.

Claims (8)

In the flexible projectile ground test apparatus, A projectile model that simulates the shape of an actual projectile; A fuel accommodating model that simulates fuel for propelling a projectile model; A thrust vector module for controlling the rotation of the projectile motel; An angular velocity sensor mounted between the projectile model and the fuel receiving model; An air pad to remove frictional forces due to the structure support; And Includes an excitation device for simulating the disturbance of the projectile model, And a hose capable of supplying and discharging liquid at a constant rate to the fuel receiving model, wherein the air pad is operated by compressed air. The apparatus of claim 1, wherein the thrust vector module includes a compressed air hose whose direction and displacement are controlled by the adjustment of the motor. In the method for testing the performance of the flexible projectile ground test apparatus that simulates the projectile, Filtering the angular velocity signal by measuring the angular velocity to obtain a natural frequency with a multi-mode notch filter; And calculating the output signal at the filtered angular and angular velocity to obtain a moment to rotate the flexible projectile ground test apparatus. The natural frequency is measured in real time, and the gain value of the control command is a performance test method of the flexible projectile ground test apparatus, characterized in that the optimization through the neural network algorithm. An apparatus for testing the performance of a flexible projectile ground test apparatus, Flexible projectile ground test model that simulates the projectile; An angular velocity measuring device for measuring an angular velocity of the flexible projectile ground test model; A signal conversion board for receiving a signal of the angular velocity measuring device; An attitude control computer having an adaptive notch filter and an adaptive controller for receiving a signal from the signal conversion board and generating a control command; A thrust vector driver for driving thrust of the flexible projectile ground test motel according to a command signal from the signal change board; And A performance testing device of the flexible projectile ground testing apparatus, comprising a disturbance simulation apparatus for simulating the disturbance of the flexible projectile ground testing model. The performance test apparatus according to claim 4, wherein the angular velocity is input to the signal conversion board at 200 Hz. In the ground test apparatus of the liquid propulsion rocket having an elongate structure, A projectile model that simulates the shape of an actual projectile; A transparent vinyl jacket that receives and discharges water at a constant speed by a hose; A plurality of air exhaust holes located on top of the transparent vinyl jacket; A thrust vector module for controlling the rotation of the projectile motel; An angular velocity sensor mounted between the projectile model and the fuel receiving model; An air pad to remove frictional forces due to the structure support; And Includes an excitation device for simulating the disturbance of the projectile model, The ground test apparatus is a ground test apparatus of the liquid propulsion rocket, characterized in that can be modeled as the center of rotation, flexible beam and thrust vector control module. The method according to claim 6, 4. The ground test apparatus of claim 3 wherein the rate of water supply and discharge through the hose is between 0.02 kg and 0.06 kg per second. Ground test apparatus according to claim 1 for performance analysis of a control system of a liquid propulsion rocket.

Images