A control system for adaptive regulation of launch vehicle gain and network parameters
Through the control method of adaptive adjustment, the arrow posture control parameters are adjusted online using the adaptive law and adjustment coefficient, which solves the problem of reduced stability margin caused by the time-variability and uncertainty of model parameters of the rocket attitude control system, and achieves a higher level adaptability and robustness.
Patent Information
- Application Number
- CN202211215779.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-09-30
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2042-09-30
AI Technical Summary
Due to the role of complex aerodynamic loads, the launch vehicle has time-degeneration and uncertainty in the arcuate model parameters, resulting in a decrease in the stability margin of the rocket attitude control system, and may even lead to instability.
Adaptive adjustment control method is adopted to obtain the arrow body posture deviation signal, perform adaptive law operations to obtain adjustment coefficients, adjust the arrow body posture deviation signal to generate the arrow body command signal, and then control the arrow body posture angle and/or angular velocity. . The method includes adaptive adjustment of gain and network parameters, using high-pass filters and low-pass filters to extract real-time elastic vibration energy signals, and combining adaptive law and limiting processing to realize online adaptive of attitude angle gain parameters and network parameters. adjust.
By adjusting the control parameters in real time, the conservatism of the ground fixed parameter design is overcome, the adaptability and robustness of the control system are significantly improved, the impact of high-order elasticity on attitude angle control is effectively suppressed, and the reliability of the control system is enhanced.
Smart Images

Figure CN115657473B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of launch vehicles, and in particular to a control system for adaptively adjusting the gain and network parameters of a launch vehicle. Background Art
[0002] With the improvement of the carrying capacity of launch vehicles, their mass, thrust and slenderness ratio increase, and the booster and core stage engine are often used to swing together to participate in attitude control. At the same time, due to the effect of complex aerodynamic loads, the overall rocket presents characteristics such as elastic low-frequency and dense-frequency modes, strong coupling vibrations, and complex local deformations, resulting in a large degree of time-varying and uncertainty in the rocket body model parameters, which will lead to a reduction in the stability margin of the rocket attitude control system and even greatly increase the probability of instability. Traditional attitude control design selects characteristic second points, performs extreme deviations on the model parameters, and designs a set of fixed parameters that can adapt to the rated and deviation states at multiple flight times. This deviation-based design method is widely used in engineering, but it is relatively conservative. In addition, for situations where the model parameter deviations are large, the parameters designed based on this method are difficult to adapt to different flight times. Summary of the invention
[0003] The present application provides an adaptive control method and control system, the purpose of which is to control elastic vibration using amplitude stabilization to reduce the amplitude of elastic vibration until the adaptive law reaches the latest equilibrium state.
[0004] In a first aspect, a control method for adaptive regulation is provided, characterized by comprising:
[0005] Acquire a first arrow body attitude deviation signal, where the first arrow body attitude deviation signal is used to indicate a deviation between an actual arrow body attitude and a theoretical arrow body attitude;
[0006] Performing an adaptive law operation according to the first arrow body attitude deviation signal to obtain an adjustment coefficient;
[0007] Adjust the second arrow body attitude deviation signal according to the adjustment coefficient to obtain an arrow body command signal;
[0008] According to the arrow body command signal, the attitude angle and / or angular velocity of the arrow body is controlled.
[0009] The adjustment coefficient includes a gain adjustment coefficient and / or a network adjustment coefficient.
[0010] On the basis of traditional PID control, a gain and network parameter adaptive adjustment module is added. The real-time high-order elastic vibration energy signal is extracted from the command signal through high-pass filter and low-pass filter, and then the gain and network adjustment coefficients are obtained by combining adaptive law and limiting processing. The adjustment coefficients are fed back to the original gain and network to realize the online adaptive adjustment of attitude angle gain parameters and network parameters.
[0011] Compared with the prior art, the solution provided by this application includes at least the following beneficial technical effects:
[0012] (1) The control parameters are adaptively adjusted according to the strength of the elastic signal during real-time flight, overcoming the conservatism of the ground fixed parameter design and greatly improving the adaptability of the control system.
[0013] (2) Controlling with adjusted gains and / or network parameters can effectively suppress the influence of high-order elasticity on attitude angle control and enhance the robustness and reliability of the control system.
[0014] (3) The solution proposed by the present invention is simple, reliable and easy to implement in engineering.
[0015] In combination with the first aspect, in certain implementations of the first aspect, the second rocket body attitude deviation signal and the first rocket body attitude deviation signal are the same information.
[0016] It can be regulated in real time according to the posture deviation of the arrow body in the current cycle.
[0017] In combination with the first aspect, in certain implementations of the first aspect, a period corresponding to the first rocket body attitude deviation signal is before a period corresponding to the second rocket body attitude deviation signal.
[0018] The rocket body attitude deviation signal can be processed in parallel, which is convenient for application in high-frequency scenarios.
[0019] In combination with the first aspect, in some implementations of the first aspect, performing an adaptive law operation according to the first rocket body attitude deviation signal to obtain an adjustment coefficient includes:
[0020] Performing high-pass filtering on the first arrow body attitude deviation signal to obtain a first elastic energy signal;
[0021] Performing low-pass filtering on the first elastic energy signal to obtain a second elastic energy signal;
[0022] An adaptive law operation is performed according to the second elastic energy signal to obtain the adjustment coefficient.
[0023] The design of the adaptive regulation law of gain and network parameters needs to take the elastic energy signal extracted by the low-pass filter as input. When the control parameters are reasonably designed, that is, the correction network has good elastic vibration suppression in each flight segment, the elastic signal in the instruction is close to zero, so the rate of change of the regulation coefficient should be 0, and the regulation coefficient is always kept at around the initial value of 1. If the elastic signal suddenly increases during the flight, causing high-frequency elastic vibration signals in the attitude angle control instruction, the elastic energy of the attitude angle instruction after high-pass and low-pass filtering increases, and the gain regulation coefficient and network regulation coefficient should be gradually reduced. The reduction of the regulation coefficient brings two effects: one is the reduction of the control gain, and the other is that the network parameters are adjusted, and the filtering effect on the elastic amplitude is enhanced. The above two measures are equivalent to using amplitude stabilization to control the elastic vibration, which promotes the reduction of the elastic vibration amplitude until the adaptive law reaches the latest equilibrium state.
[0024] In combination with the first aspect, in some implementations of the first aspect, performing an adaptive law operation according to the first rocket body attitude deviation signal to obtain an adjustment coefficient includes:
[0025] Performing an adaptive law operation according to the first rocket body attitude deviation signal to obtain an initial value of an adjustment coefficient;
[0026] When the initial value of the adjustment coefficient is greater than the maximum preset adjustment coefficient, the adjustment coefficient is the maximum preset adjustment coefficient;
[0027] When the initial value of the adjustment coefficient is less than the minimum preset adjustment coefficient, the adjustment coefficient is the minimum preset adjustment coefficient;
[0028] When the initial value of the adjustment coefficient is between the minimum preset adjustment coefficient and the maximum preset adjustment coefficient, the adjustment coefficient is the initial value of the adjustment coefficient.
[0029] The adjustment coefficient is limited to keep it within a reasonable range. To ensure that the adaptive gain and network adjustment range are within the design envelope, the upper and lower limits of the gain are pulled 20% for margin analysis during normal frequency domain design. At the same time, there is still enough stability margin at the first-order elastic frequency after the time-varying network link is connected to the main channel network of the first-stage flight segment.
[0030] In conjunction with the first aspect, in some implementations of the first aspect, the first elastic energy signal satisfies:
[0031] y hp =G hp (s)δ c , G hp (s) represents the high-pass filtering process, δ c Represents the arrow body command signal.
[0032] In combination with the first aspect, in some implementations of the first aspect, the second elastic energy signal satisfies:
[0033] G lp (s) represents the low-pass filtering process, y hp represents the first elastic energy signal.
[0034] In combination with the first aspect, in some implementations of the first aspect, performing the adaptive law operation according to the first rocket body attitude deviation signal includes:
[0035] according to Determine the adjustment coefficient k, y s =D aL (z)·(D aH (z)·L zl ) 2 , α and β are adaptive law coefficients greater than zero, D aH (z), D aL (z) represent the high-pass filter network and the low-pass filter network respectively, L zl For control instructions,
[0036] Adaptive law operation can take into account elasticity, which is specifically reflected in -p hi (y s )·α·k, to achieve the control of the amplitude stability caused by elastic vibration.
[0037] In conjunction with the first aspect, in some implementations of the first aspect, the activation function satisfies: a, b, and m are constant coefficients greater than zero.
[0038] The activation function is the elastic energy signal y s The function can change according to the strength of the high-frequency elasticity contained in the attitude angle channel control instruction, and can be flexibly designed according to actual conditions.
[0039] In combination with the first aspect, in some implementations of the first aspect,
[0040] Combining the lower limit design of attitude control design is conducive to improving the rationality of adaptive law design.
[0041] In a second aspect, a control system with adaptive regulation is provided, wherein the system is used to execute the method described in any one of the implementations of the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1A schematic flow chart of an adaptive regulation control method provided in an embodiment of the present application.
[0043] Figure 2 A schematic structural diagram of an adaptive control system provided in an embodiment of the present application.
[0044] Figure 3 A schematic structural diagram of another adaptive adjustment control system provided in an embodiment of the present application.
[0045] Figure 4 is the activation function p under different b / a hi (y s ) schematic diagram. DETAILED DESCRIPTION
[0046] The present application is further described in detail below with reference to the accompanying drawings and specific embodiments.
[0047] Figure 1 It is a schematic flow chart of an adaptive adjustment control method provided in an embodiment of the present application.
[0048] 110. Obtain a first arrow body attitude deviation signal.
[0049] 120. Perform an adaptive law operation according to the first rocket body attitude deviation signal to obtain an adjustment coefficient.
[0050] 130. Adjust the second rocket body attitude deviation signal according to the adjustment coefficient to obtain a rocket body command signal.
[0051] 140, controlling the attitude angle and / or angular velocity of the rocket body according to the rocket body command signal.
[0052] In the embodiment of the present application, the adjustment coefficient may include a gain adjustment coefficient and / or a network adjustment coefficient.
[0053] Figure 2 Schematic diagram of a control system for adaptive regulation provided by an embodiment of the present application. Figure 2 , describing an adaptive adjustment control method provided in an embodiment of the present application.
[0054] The control system may include a high-pass filter, a low-pass filter, an adaptive law 1, an adaptive law 2, a limiting process 1, and a limiting process 2.
[0055] In cycle 1, assuming that the theoretical attitude angle 1 of the rocket body is The theoretical angular velocity 1 is ω c1 , the actual attitude angle 1 of the rocket body is The actual angular velocity 1 is ω1 (the actual attitude angle and actual angular velocity can be detected by the inertial component and the rate gyro). and The deviation of c1 The deviation from ω1 can give the rocket attitude deviation signal 1.
[0056] According to the network adjustment coefficient k of cycle 1 w1 (For example, the initial network adjustment coefficient, or the δ of the previous cycle of cycle 1 c0 Determine) and the gain adjustment coefficient k of cycle 1 a1 (For example, the initial gain adjustment coefficient, or the delta of the previous cycle of cycle 1 c0 The arrow body attitude deviation signal 1 is adjusted to obtain the adjusted arrow body attitude deviation signal 2. The arrow body command signal δ can be determined according to the arrow body attitude deviation signal 2. c1 The rocket body command signal can be used to instruct the servo mechanism to control the state of the engine to reduce the deviation between the theoretical attitude angle and the actual attitude angle of the rocket body, and the deviation between the theoretical angular velocity and the actual angular velocity of the rocket body.
[0057] Specifically, for gain adjustment, we can consider directly multiplying the main channel static gain a0 by the adjustment coefficient k1, that is, a0 = k1·a0; network adjustment can consider adjusting the second-order network link, for example The molecular damping term ξ1 is adjusted, and the elastic high-frequency filtering is adjusted by changing k2, that is, ξ1 = k2·ξ2.
[0058] In some embodiments, δ c1 It can also be used to indicate the network adjustment coefficient k of period 2 w2 and gain adjustment factor k a2 . Network adjustment coefficient k w2 and gain adjustment factor k a2 It can be used to adjust the rocket body attitude deviation signal 3 of period 2 in period 2, and obtain the rocket body attitude deviation signal 4 adjusted in period 2.
[0059] The specific process is as follows:
[0060] In cycle 2, assuming that the theoretical attitude angle 2 of the rocket body is The theoretical angular velocity 2 is ω c2 , the actual attitude angle 2 of the rocket body is The actual angular velocity 2 is ω2. and The deviation of c2 The deviation from ω2 can give the rocket attitude deviation signal 3.
[0061] According to the network adjustment coefficient k of period 2 w2 and the gain adjustment factor k for cycle 2 a2, adjust the arrow body attitude deviation signal 3 to obtain the adjusted arrow body attitude deviation signal 4. According to the arrow body attitude deviation signal 4, the arrow body command signal δ can be determined c2 δ c2 It can be used to indicate the state of the servo mechanism controlling the engine to reduce the deviation between the theoretical attitude angle and the actual attitude angle of the rocket body, as well as the deviation between the theoretical angular velocity and the actual angular velocity of the rocket body.
[0062] In some other embodiments, δ c1 It can also be used to correct the network adjustment coefficient k of cycle 1 w1 and gain adjustment factor k a1 , and obtain the corrected network adjustment coefficient k w1’ and gain adjustment factor k a1’ . Network adjustment coefficient k w1’ and gain adjustment factor k a1’ It can be used to adjust the rocket body attitude deviation signal 1 of cycle 1 in cycle 1, and obtain the rocket body attitude deviation signal 5 adjusted in cycle 1. According to the rocket body attitude deviation signal 5, the rocket body command signal δ can be determined c1’ δ c1’ It can be used to indicate the state of the servo mechanism controlling the engine to reduce the deviation between the theoretical attitude angle and the actual attitude angle of the rocket body, as well as the deviation between the theoretical angular velocity and the actual angular velocity of the rocket body.
[0063] Figure 3 This is another adaptive control system provided by the embodiment of the present application. Figure 3 , describing an adaptive adjustment control method provided in an embodiment of the present application.
[0064] The control system may include a high-pass filter, a low-pass filter, an adaptive law 1, an adaptive law 2, a limiting process 1, and a limiting process 2.
[0065] In cycle 1, assuming that the theoretical attitude angle 1 of the rocket body is The theoretical angular velocity 1 is ω c1 , the actual attitude angle 1 of the rocket body is The actual angular velocity 1 is ω1. and The deviation of c1 The deviation from ω1 can be used to obtain the rocket body attitude deviation signal 1. The rocket body attitude deviation signal 1 can be used to indicate the network adjustment coefficient k of period 1 w1 and gain adjustment factor k a1 According to the network adjustment coefficient k of cycle 1 w1 and the gain adjustment factor k for cycle 1 a1, adjust the rocket body attitude deviation signal 1 to obtain the adjusted rocket body attitude deviation signal 2. According to the rocket body attitude deviation signal 2, the rocket body command signal δ c1 .
[0066] The following describes the process of determining the adjustment coefficient.
[0067] δ c It can pass through a high-pass filter. Combined with the high-order elastic frequency deviation range of the rocket body, a high-pass filter G is designed. hp (s), the high-order elastic information y in the control command obtained through the network and gain calculation hp Extracted, there are: y hp =G hp (s)δ c The extracted elastic information is squared to obtain the elastic energy signal
[0068] Elastic energy signal It can pass through a low-pass filter. According to the response capability of the arrow body, a low-pass filter G is designed. lp The frequency band of (s) is within the rigid body cutoff frequency, ensuring that the elastic signal entering the adaptive law changes smoothly. The elastic energy signal is low-pass filtered to obtain:
[0069] The elastic vibration signal is processed by high-pass and low-pass filters and then input into the adaptive law. Generally, the adaptive law is designed as a differential function of the adjustment coefficient k, which can change according to the strength of the high-frequency elasticity contained in the attitude angle channel control command and can be flexibly designed according to actual conditions.
[0070] In some embodiments provided in the present application, the adaptive law of the gain and the network are designed to be in the form of a differential function of the adjustment coefficient k:
[0071]
[0072] Where p hi (y s ) is an activation function, which varies in the range of 0 to 1 according to the strength of the high-frequency elasticity contained in the attitude angle channel control instruction, and α and β are adaptive law coefficients greater than zero.
[0073] The activation function can satisfy: y s =D aL (z)·(D aH (z)·L zl ) 2 , D aH (z), D aL (z) represent the high-pass filter network and the low-pass filter network respectively, Lzl is the control instruction, a, b, m are constant coefficients greater than zero.
[0074] When the elastic energy y in the instruction s When the activation function is very small, it is approximately 0, and the elastic part of the adaptive law -p hi (y s )·α·k is 0, and elasticity has no effect on gain adjustment. s When it increases to a certain value, the activation function gradually increases to 1, and the effect of elasticity on gain adjustment gradually increases. s Magnification factor.
[0075] a and b determine the activation function p hi (y s ), b / a determines the activation threshold. The smaller it is, the lower the activation threshold is. Figure 4 As shown. b determines the speed at which the function rises. The faster the rise, the faster the gain and network adjustment, and the more sensitive it is to changes in elasticity. From an engineering practice perspective, it is necessary to avoid adjusting too quickly, and combine simulation and actual flight results analysis.
[0076] In some embodiments, the adjustment coefficient output by the adaptive law may be an initial value of the adjustment coefficient. The initial value of the adjustment coefficient may be subjected to a limiting operation. Specifically, in combination with the frequency domain stability analysis, the adjustment coefficient calculated by the adaptive law is subjected to a limiting operation to output a final value of the adjustment coefficient.
[0077] Gain adjustment coefficient k a , network adjustment coefficient k w The limit value needs to be designed based on the actual frequency domain stability envelope range and the actual deviation range. a , k w The initial value of is 1.0, and the limiting process is performed:
[0078] 1) Limiter for gain adjustment
[0079] Limit at Within the range, that is
[0080]
[0081] 2) Limiting for time-varying network regulation
[0082] Limit at Within the range, that is
[0083]
[0084] Then the adjusted control gain is: a0 = k aa0, adjusted time-varying network damping: ξ1 = k w ·ξ1.
[0085] For the design of adaptive law parameters α and β, consider the equilibrium point of the adaptive law, that is, have:
[0086]
[0087] Design upper and lower limits according to attitude control, combined with upper and lower limits of gain deviation When the gain reaches the lower limit When:
[0088]
[0089] For the adjustment of network parameters, consider adjusting the damping parameters of the second-order filter network at a fixed elastic frequency. Suppose the adjustment range of the molecular damping parameter ξ1 is When ξ1 reaches the lower limit When:
[0090]
[0091] The parameter selection principles are as follows:
[0092] 1) If you want to activate the function p hi (y s )=1, the gain adjustment coefficient reaches the limit value, then In this case, the gain will not reach the limit value easily, and the adjustment speed is not fast;
[0093] 2) If you want to activate the function in the rising section p hi (y s )<1, the gain adjustment coefficient reaches the limit value, then In this way, the overall adaptive law is more sensitive to elasticity, and the effect of reducing elastic vibration by reducing gain is better;
[0094] 3) If you want to activate the function p hi (y s )=1, the network adjustment coefficient reaches the limit value, then In this case, the parameter will not easily reach the limit value, and the adjustment speed is not fast;
[0095] 4) If you want to activate the function in the rising section p hi (y s )<1, the network adjustment coefficient reaches the limit value, then In this way, the overall adaptive law is more sensitive to elasticity, and the effect of reducing elastic vibration by adjusting network parameters is better.
[0096] In order to suppress the large elastic deviation amplitude crossing, the upper limit of the adjustment coefficient is specified. Both are 1.
[0097] Figure 2 and Figure 3 The adaptive control systems shown can all use attitude angle control instructions δ c The input is the static gain adjustment coefficient k1 and the network adjustment coefficient k2. Based on the PID-based controller + correction network, the following two functions are achieved through adaptive adjustment of the gain and network adjustment coefficient:
[0098] (1) When uncertain elastic vibration causes coupling between the elastic signal and the control signal, the influence of the elastic vibration is weakened by adjusting the attitude angle channel gain online and adjusting the network parameters to enhance the filtering;
[0099] (2) When the original controller can achieve stable attitude angle control well, try not to change the gain and network parameters.
[0100] Although the present invention is disclosed as above in the form of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art may make possible changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be based on the scope defined by the claims of the present invention.
Claims
1. A control method for adaptive regulation, characterized in that: include: Acquire a first arrow body attitude deviation signal, where the first arrow body attitude deviation signal is used to indicate a deviation between an actual arrow body attitude and a theoretical arrow body attitude; Performing an adaptive law operation according to the first arrow body attitude deviation signal to obtain an adjustment coefficient; Adjust the second arrow body attitude deviation signal according to the adjustment coefficient to obtain an arrow body command signal; Controlling the attitude angle and / or angular velocity of the arrow body according to the arrow body command signal; The second arrow body attitude deviation signal is the same as the first arrow body attitude deviation signal; The period corresponding to the first rocket body attitude deviation signal is before the period corresponding to the second rocket body attitude deviation signal; The performing of the adaptive law operation according to the first rocket body attitude deviation signal to obtain the adjustment coefficient includes: performing high-pass filtering on the first rocket body attitude deviation signal to obtain a first elastic energy signal; Performing low-pass filtering on the first elastic energy signal to obtain a second elastic energy signal; Performing an adaptive law operation according to the second elastic energy signal to obtain the adjustment coefficient; The step of performing an adaptive law operation according to the first rocket body attitude deviation signal to obtain an adjustment coefficient comprises: performing an adaptive law operation according to the first rocket body attitude deviation signal to obtain an initial value of the adjustment coefficient; When the initial value of the adjustment coefficient is greater than the maximum preset adjustment coefficient, the adjustment coefficient is the maximum preset adjustment coefficient; When the initial value of the adjustment coefficient is less than the minimum preset adjustment coefficient, the adjustment coefficient is the minimum preset adjustment coefficient; When the initial value of the adjustment coefficient is between the minimum preset adjustment coefficient and the maximum preset adjustment coefficient, the adjustment coefficient is the initial value of the adjustment coefficient.
2. The method according to claim 1, characterized in that The first elastic energy signal satisfies: hp =G hp (s)δ c , G hp (s) represents the high-pass filtering process, δ c Represents the arrow body command signal.
3. The method according to claim 1, characterized in that The second elastic energy signal satisfies: G lp (s) represents the low-pass filtering process, y hp represents the first elastic energy signal.
4. The method according to claim 1, characterized in that: The performing the adaptive law operation according to the first rocket body attitude deviation signal comprises: according to Determine the adjustment coefficient k, y s =D aL (z)·(D aH (z)·L zl ) 2 , α and β are adaptive law coefficients greater than zero, D aH (z), D aL (z) represent the high-pass filter network and the low-pass filter network respectively, L zl For control instructions, 5. The method according to claim 4, characterized in that The activation function satisfies: a, b, and m are constant coefficients greater than zero.
6. The method according to claim 4, characterized in that 7. The method according to any one of claims 1 to 5, characterized in that The adjustment coefficient includes a gain adjustment coefficient and / or a network adjustment coefficient.
8. An adaptive control system, characterized in that: The system is used to perform the method according to any one of claims 1 to 7.
Citation Information
Patent Citations
Spacecraft fault tolerance attitude cooperation tracking control method based on normalized neural network
CN105404304A
Adaptive augmented interference-resisting fault tolerance method for aiming at active segment flight of vertical take-off and landing carrier
CN109541941A
