Filtering method for relative line-of-sight angle of on-orbit target
By using universal variable method and the Kalman filtering method of "current" statistical model when the spacecraft is approaching non-cooperative targets, the problem of obtaining accurate relative line of sight angles is solved, and the accuracy and error control of spacecraft guidance are improved. Suitable for spacecraft on-orbit non-cooperative target tracking missions.
Patent Information
- Application Number
- CN202410933722.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-12
- Publication Date
- 2024-11-08
AI Technical Summary
When spacecraft approaches non-cooperative targets in space, it is difficult to obtain accurate relative line of sight information. The existing filtering methods have errors in real applications and lack effective accurate solutions to the position and velocity of three-dimensional space orbits.
Using the target orbit design based on universal variable method and the "current" statistical model, the relative motion model of the spacecraft and the target is constructed, and the target position and velocity are calculated through Newton's iterative calculation, and the three-dimensional relative line-of-sight angle filtering system is designed to reduce the Measurement error and system error.
It improves the accuracy of relative line of sight angle and the accuracy of application of filtering methods, reduces the seeker measurement error and system error, and is suitable for tasks where spacecraft approaches non-cooperation goals in orbit.
Smart Images

Figure CN118916982A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to aerospace technology, and in particular relates to a relative sight angle filtering method for an on-orbit target. Background Art
[0002] With the deepening of human exploration, development and utilization of outer space and the demand for on-orbit services of spacecraft, higher requirements are placed on rendezvous and docking technology. Topics such as on-orbit capture and repair of faulty spacecraft, removal of space debris, and changing the orbit of asteroids have become issues that need to be faced and solved in the development of aerospace technology, and the technology of obtaining accurate relative line of sight angles in the process of approaching non-cooperative targets is a key technology necessary to solve these problems. Space non-cooperative targets generally refer to a class of space objects that cannot provide effective information, including faulty or failed satellites, space debris, asteroids, etc. In order to meet the need for accurate relative line of sight angle input during the approach of spacecraft to non-cooperative targets, it is of great value to study a relative line of sight angle filtering method for on-orbit targets.
[0003] There are two main tracking methods for small spacecraft approaching non-cooperative targets in orbit: active and passive. Active tracking mainly uses radar tracking, which can obtain comprehensive status information such as target position and speed, but the consequences are high cost, limited observation distance, and large size. Passive tracking does not emit signals, and has the characteristics of being difficult to be discovered, fast response, and long observation distance. Passive tracking mainly tracks the target in real time by obtaining relative line of sight angle information, thereby guiding the spacecraft to hand over the target. Therefore, obtaining accurate relative line of sight angles is of great significance for spacecraft passive tracking targets.
[0004] As early as the 1950s, Wax et al. proposed the idea of target tracking, which has attracted great attention from many researchers. Among them, target tracking mainly processes data by filtering, estimates the target's motion state, and constructs a state space to describe the discrete or continuous dynamic behavior of the target. The classic filtering methods mainly include least squares filtering, Wiener filtering, and Kalman filtering, which have good estimation effects on linear systems. When the acceleration of the target motion state changes, Singer proposed using Kalman filters to estimate the target motion state in this case in 1970. Some researchers have studied the acceleration of the target motion state from a statistical perspective, estimating the acceleration as a disturbance input with random statistical characteristics, and then estimating the frequency and time constant of the target state acceleration change according to the designed filtering method, and finally obtaining the target state acceleration change law. Zhang Biao et al. combined the "current" statistical model with the volumetric Kalman filter, which has a good effect on the estimation of the target state acceleration change problem with high accuracy. Later, some researchers used the particle filter method to solve this problem, and estimated the current state information of the target by the weighted sum of a batch of weighted particles.
[0005] In fact, in practical engineering applications, in order to verify the effect of filtering methods on reducing the error of the relative sight angle of the target obtained by the spacecraft, it is not only necessary to design the filtering method, but also to consider the construction of the target spacecraft's orbit, obtain the target's real-time position and speed, and obtain the true value of the relative sight angle. At the same time, it is also necessary to build a relative motion model between the spacecraft and the target in the orbital plane, obtain the relationship between the relative sight angle of the spacecraft and time in the process of tracking the target, and construct the state equation. Therefore, based on this situation, the present invention combines the theory of spacecraft orbital mechanics and target tracking theory, and designs a CS-based relative sight angle KF method for on-orbit targets in order to filter the relative sight angle in the process of approaching the target, and finally calculates the root mean square error of the relative sight angle after filtering. Summary of the invention
[0006] The present invention proposes a relative sight angle filtering method for an on-orbit target, which effectively reduces the error of the relative sight angle of the target obtained by the spacecraft during the process of tracking the target.
[0007] The technical solution to realize the present invention is: a method for filtering the relative sight angle of an on-orbit target, comprising the following steps:
[0008] Step 1: Design target trajectory based on universal variable method:
[0009] According to the principle that the target is only affected by the earth's gravity in a vacuum environment, the target motion equation in the earth coordinate system is constructed to obtain the functional relationship between the target position and velocity over time. At the same time, the target angular momentum and energy equations are established. A universal variable τ is introduced to solve the target motion trajectory using the universal variable method. The target position r at any time is obtained by Newton iteration calculation. T1 and speed v T1 .
[0010] Step 2: Construct the relative motion model between the spacecraft and the target:
[0011] According to the changing law of the position and velocity of the spacecraft in three-dimensional space, the relative motion relationship between the spacecraft M and the target T1, which are split into the horizontal plane xoy and the vertical plane yoz in the inertial coordinate system, is obtained, and the relative motion model between the two is obtained. The azimuth and altitude angles of the spacecraft velocity, and the azimuth and altitude angles of the approach speed of the spacecraft and the target are calculated; the target relative line of sight angle motion model is simplified, and the state equation of relative line of sight angle change is obtained.
[0012] Step 3: Filter the relative sight angle:
[0013] According to the relative line of sight angle motion model of the spacecraft to the target, the state equation and observation equation of the relative line of sight angle KF system based on CS are constructed to describe the state change relationship of the relative line of sight azimuth, azimuth velocity, azimuth acceleration, pitch angle, pitch angle velocity, and pitch angle acceleration in three-dimensional conditions; the "current" statistical model is designed to calculate the state prediction value and the prediction error covariance P(k+1 / k), calculate the Kalman filter gain K(k), update the covariance matrix P(k) and the system state quantity X(k+1), that is, the filtered relative line of sight angle state quantity, and the relative line of sight angle accuracy is improved after filtering.
[0014] A method for filtering relative sight angles of on-orbit targets is based on the on-orbit target relative sight angle filtering process.
[0015] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, a filtering process based on a relative sight angle facing an on-orbit target is implemented.
[0016] A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the relative sight angle Kalman filtering process based on CS for an on-orbit target is implemented based on a relative sight angle filtering method for an on-orbit target.
[0017] Compared with the prior art, the present invention has the following significant advantages: the present invention considers the relative sight angle filtering method in three-dimensional space in the process of the spacecraft approaching the on-orbit target, and designs a method for solving the target position and speed using the universal variable method, taking into account the actual changing environment in the entire spacecraft guidance process, improving the accuracy of the relative sight angle filtering method, and reducing the measurement error and system error of the relative sight angle measured by the seeker. It is more in line with the actual operation law of the spacecraft than the existing target tracking filtering method that lacks the accurate solution of the target orbital position and speed, and is closer to engineering application. It can be widely used in satellite recovery missions when the spacecraft approaches non-cooperative targets in orbit. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] Figure 1 The present invention is a flow chart of a relative sight angle Kalman filtering (KF) method based on a "current" statistical model (CS) for an on-orbit target.
[0019] Figure 2 It is a geometric relationship diagram of the spacecraft and the target.
[0020] Figure 3 It is the three-axis error diagram of the estimated target orbit position.
[0021] Figure 4 It is the three-axis error diagram of the estimated velocity of the target at different times.
[0022] Figure 5 It is the error map of the estimated absolute position and velocity of the target track in three-dimensional space.
[0023] Figure 6 It is a graph showing the relationship between the relative line of sight azimuth and the true azimuth before and after filtering and how they change over time.
[0024] Figure 7 It is a graph showing the relationship between the root mean square error of the relative line of sight azimuth before and after filtering and time.
[0025] Figure 8 It is a graph showing the relationship between the relative sight line height angles before and after filtering and the true height and low values changing with time.
[0026] Fig. 9 It is a graph showing the relationship between the root mean square error of the relative sight height angle before and after filtering and time. DETAILED DESCRIPTION
[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0028] In the present invention, unless otherwise clearly specified and limited, the terms "connection", "fixation" and the like should be understood in a broad sense. For example, "fixation" can be a fixed connection, a detachable connection, or an integral connection; "connection" can be a mechanical connection or an electrical connection. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0029] In addition, the technical solutions between the various embodiments of the present invention can be combined with each other, but it must be based on the fact that ordinary technicians in the field can implement it. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by the present invention.
[0030] The following will further introduce the specific implementation method, as well as the technical difficulties and inventive points of this invention in combination with this design example.
[0031] Combination Figure 1 The present invention proposes a relative sight angle filtering method for on-orbit targets, which reduces the measurement error and system error of the relative sight angle between the spacecraft and the target in the input data brought by the seeker during the entire terminal guidance process. The specific steps are as follows:
[0032] Step 1: Design target trajectory based on universal variable method
[0033] In the task of spacecraft handing over on-orbit targets, it is necessary to build an on-orbit motion model of the target for simulation verification. In the orbital task of calculating on-orbit targets in space research, the commonly used method is to use conic sections to approximate the initial orbit of the design target and obtain the initial value of the launch; then the target spacecraft orbit is calculated by accurate numerical calculation methods. The use of classical conic section analysis methods will involve many inevitable selection problems, which is not conducive to the use of computer programming to solve, resulting in a huge workload, and when calculating parabolic orbits, the numerical accuracy is often not high. The universal variable was first proposed by Sundman, and later Battin proposed a formula for using universal variables to calculate the running time of the target in orbit in his research. Compared with the classical conic section method, the universal variable method has a simpler calculation process and is more friendly to computer programming, especially more suitable for calculating parabolic orbits. Therefore, the present invention introduces the universal variable method in orbital mechanics into the orbital design of on-orbit targets, which improves the calculation accuracy on the one hand, and provides convenient conditions for computer programming on the other hand, and improves the efficiency of guidance law design.
[0034] Since the target is only affected by the earth's gravity in a vacuum environment, the target motion equation in the earth coordinate system is:
[0035]
[0036] Among them, R T1 Represents the position vector of the target in the earth coordinate system; V T1 Represents the velocity vector of the target in the earth coordinate system; G T1 Represents the gravitational acceleration vector of the target in the Earth coordinate system.
[0037] Solve the equations of motion:
[0038] The specific forms of angular momentum h and energy ξ are as follows:
[0039]
[0040] Where r and v are the orbital radius and velocity of the target at a certain moment, u is the gravitational constant, and a is the semi-major axis of the orbit.
[0041] If v is decomposed into radial components and the lateral component Then the energy equation can be written as:
[0042]
[0043] Solve have to:
[0044]
[0045] Where p is the semi-sine focal chord, and the universal variable method is applied to find the target motion trajectory, introducing a universal variable τ:
[0046]
[0047] Where r is the position vector of the spacecraft.
[0048] After a series of transformation integrations, the relationship between the given time t and τ and the initial state of the spacecraft is obtained as follows:
[0049]
[0050]
[0051] Among them, the intermediate variable Intermediate variables Intermediate variables t n Given the target initial state and initial value τ n , r0 represents the initial position of the target, and v0 represents the initial velocity of the target.
[0052] The approximate value of τ is obtained according to Newton's iteration method, which is as follows:
[0053]
[0054] t is the given flight time, It can be obtained from the above formula.
[0055] because E is the eccentric anomaly angle. For a complete orbit, ΔE = 2π. Therefore, the first approximate value of τ can be:
[0056]
[0057] Where T is the orbital period of the spacecraft, and t0 represents the time when the target is in its initial state.
[0058] The τ corresponding to a given flight time can be solved through Newton iteration.
[0059] Since Kepler motion is limited to one plane, namely the orbital plane of the target, the current state vector R, V of the target and the initial state vector R0, V0 are coplanar, so:
[0060] R=fR0+gV0
[0061] Differentiating the above formula, we can get:
[0062]
[0063] f,g, is a scalar quantity related to time. Combining the trigonometric identities of the sine and cosine of an angle and the definitions of z, C(z), and S(z), we can get f, g, The relationship with τ is as follows:
[0064]
[0065]
[0066] The motion trajectory of the target can be calculated using the above formula.
[0067] Step 2: Construct the relative motion model between the spacecraft and the target
[0068] In order to study the relative motion law of spacecraft in three-dimensional space, the relative motion relationship between spacecraft M and target T1 is divided into horizontal plane and vertical plane in the inertial coordinate system for study. The horizontal plane xoy refers to the projection of the relative position relationship between the spacecraft and the target on the XOY plane of the inertial coordinate system, and the vertical plane yoz is always the same vertical plane where the spacecraft and the target are located. The geometric relationship between the spacecraft and the target is as follows: Figure 2 shown.
[0069] Where oy is the projection of the line of sight MT on the horizontal plane XOY, α is the line of sight azimuth, and β is the line of sight pitch angle. The position vector of the spacecraft M and the target T1 relative to the center of the earth in the inertial coordinate system is R M , R T1 The target moves under the influence of gravity and does not maneuver. The passive external forces on the spacecraft include gravity and aerodynamic thrust. According to Newtonian kinematics, the relative motion model between the two can be expressed as:
[0070]
[0071] Among them A T1 is the target acceleration caused by the earth's gravity, A M is the acceleration of the spacecraft caused by the Earth’s gravity and the thrust of the cold gas jet. The Earth’s gravitational constant G = 6.67259 × 10 -11 N·m 2 / kg 2 , Earth mass M d =5.965×10 24 kg, A control is the control command acceleration acting on the spacecraft, Respectively represent R M , R T1 The second derivative of .
[0072] Depend on Figure 2 The relative position vector R and the absolute position vector R between the spacecraft and the target can be obtained T1 , R M The relative position relationship is R = R T1 -R M .
[0073] From the motion relationship between the spacecraft and the target in three-dimensional space, the azimuth and altitude angles of the spacecraft velocity, and the azimuth and altitude angles of the closing velocity of the spacecraft and the target can be obtained as follows:
[0074]
[0075] where α vm , β vm are the azimuth and pitch angles of the spacecraft velocity vector, α v , β v are the azimuth and pitch angles of the approaching velocity vector, R mx , R my , R mz The absolute position vector R m Projection on the x, y, and z axes, R x , R y , R z are the projections of the relative position vector R on the x, y, and z axes respectively.
[0076] According to the relative motion characteristics of the spacecraft and the target, the relative line of sight angle between the spacecraft and the target changes continuously. Considering the aerodynamic thrust and the gravity of the earth on the spacecraft, the acceleration change law under different circumstances is not uniform. Therefore, in the present invention, it is assumed that the spacecraft is used as a fixed-point platform and the target is used as a maneuvering target relative to the fixed-point spacecraft to construct a simplified target relative line of sight angle motion model, and its equation can be expressed as:
[0077] x k+1 =Ax k +Bu k +Cd k
[0078] in:
[0079]
[0080]
[0081] In the above formula, time k = 0, 1, 2, ..., α, β are the azimuth and elevation angles relative to the line of sight, T t is the period, w ak ,w bk is the angular velocity of rotation relative to the line of sight azimuth and pitch angle, is the angular acceleration relative to the azimuth and pitch angle of the line of sight, which is in a time-varying state. A is the state coefficient matrix, B is the control variable coefficient matrix, C is the second control variable matrix coefficient, and x k is the current state matrix, x k+1 is the state matrix at the next moment, u k is the control variable matrix at the current moment, d k is the second control variable matrix.
[0082] Step 3: Filter the relative sight angle:
[0083] The uniform velocity model in traditional filtering models the acceleration of the target as Gaussian white noise with a mean of zero. When the acceleration changes, it does not conform to the statistical characteristics of the acceleration of the target at this time. The Singer model models the acceleration as exponential autocorrelated zero-mean random noise. In the actual target motion process, the acceleration mean is not zero mean. The "current" statistical (CS) model assumes that the target maneuvering acceleration is a non-zero mean and takes it as the current acceleration prediction value. This model can better reflect the statistical characteristics of the change in the target acceleration.
[0084] The CS model models the target acceleration as a first-order time-dependent model with non-zero mean:
[0085]
[0086]
[0087] In the above formula, x(t) is the target state quantity, is the second-order derivative of x(t), is the target non-zero acceleration mean; a(t) is the zero-mean colored noise of acceleration at time t, and its autocorrelation function refers to the Singer model; It is the inverse of the time constant. The more drastic the acceleration change, the larger the corresponding time constant. w(t) is zero-mean white noise.
[0088] Assume that the acceleration is measured at time t Then the derivative of a1(t) is for:
[0089]
[0090] Then the state equation of the continuous-time system of the CS model is:
[0091]
[0092] After discretization, the state equation of the system is:
[0093]
[0094] Among them, the state vector X(k / k)=[x(k / k), U(k) represents the process input matrix, W(k) represents the discrete time white noise sequence, is the first-order derivative of x(t), is the third-order derivative of x(t).
[0095] The state transfer matrix Φ(k+1,k) is:
[0096]
[0097] The process input matrix U(k) is:
[0098]
[0099] W(k) is a discrete-time white noise sequence, and j represents a time; and:
[0100]
[0101]
[0102] In the above formula: is the acceleration variance, q represents the coefficient matrix, Q(k) represents the covariance of the white noise sequence; and:
[0103]
[0104] Considering that the spacecraft mainly filters the received relative line of sight angle information during the process of tracking the target, the observation equation is presented as a linear equation. According to the simplified target relative line of sight motion model, the state equation and observation equation of the relative line of sight angle KF system based on CS are constructed.
[0105] The system state equation can be expressed as:
[0106]
[0107] In the above formula: F(k+1,k) is the state transfer matrix; Ω(k) is the process input matrix; are the state vectors, which are the azimuth, azimuth velocity, azimuth acceleration, pitch angle, pitch velocity, and pitch acceleration of the target relative to the line of sight in three dimensions; is the average acceleration of the azimuth and pitch angle relative to the line of sight, where are the variances of azimuth acceleration and pitch acceleration relative to the line of sight, respectively.
[0108] The state transfer matrix F(k+1,k) of the system is:
[0109]
[0110] The process input matrix Ω(k) is:
[0111]
[0112]
[0113] The covariance Q(k) of the white noise sequence is:
[0114]
[0115] The system observation equation Y(k / k) is:
[0116] Y(k / k)=H(k)X(k / k)+V(k)
[0117] The measurement data of the spacecraft mainly includes the azimuth angle α relative to the target m (k) and pitch angle β m (k), 0 3×3 represents a 3×3 zero matrix, V(k) is the measurement noise sequence, and its covariance is R(k); the measurement matrix is H(k), which is as follows:
[0118] Y(k / k)=[α m (k),β m (k)] T
[0119] H(k)=[1,0,0,0,0,0;0,0,0,1,0,0]
[0120] The specific method is as follows:
[0121] 1) Initialization. The initial measurement value of the relative line of sight is used as the initial state quantity X(0). The covariance R(0) and the initial estimated covariance P(0) are constructed according to the measurement error of the spacecraft measurement equipment. The preset minimum negative angular acceleration and the preset maximum positive angular acceleration (a) of the relative line of sight azimuth and pitch angle angular acceleration changes are determined according to the space orbit dynamics and the aerodynamic thrust of the spacecraft. αmin ,a αmax ,a βmin ,a βmax ).
[0122] 2) Calculate the state prediction value and the prediction error covariance. According to the relationship between the angular acceleration estimate and the variance of the current statistical model, the variance of the azimuth and pitch angle angular acceleration relative to the current line of sight is calculated. For positive time, Current relative line of sight azimuth angular acceleration When it is negative, Similarly, The calculation of One-step prediction of the state for:
[0123]
[0124] Among them, the one-step prediction of the angular acceleration state can be regarded as the mean of the current angular acceleration The specific formula is as follows:
[0125]
[0126] Then the one-step prediction for this state will become as follows:
[0127]
[0128] in:
[0129] Intermediate variables Intermediate variables
[0130] The one-step state prediction error covariance P(k+1 / k) is:
[0131] P(k+1 / k)=Φ(k+1,k)P(k)Φ(k+1,k) T +Q(k)
[0132] 3) Calculate the Kalman filter gain K(k):
[0133] K(k)=P(k+1 / k)H(k) T (H(k)P(k+1 / k)H(k) T +R(k)) -1
[0134] Where R(k) represents the covariance;
[0135] 4) Update the covariance matrix P(k):
[0136] P(k)=P(k+1 / k)-K(k)H(k)P(k+1 / k)
[0137] 5) Update the system state X(k+1):
[0138]
[0139] The present invention also proposes a relative line of sight angle Kalman filter (KF) system based on a "current" statistical model (CS) for on-orbit targets. Based on the relative line of sight angle filtering method, the requirements for reducing the relative line of sight angle measurement error and system error of spacecraft measurements are achieved.
[0140] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, based on the relative line of sight angle filtering method, the requirement of reducing the relative line of sight angle measurement error and system error of spacecraft measurement is achieved.
[0141] A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, based on the relative line-of-sight angle filtering method, the requirement of reducing the relative line-of-sight angle measurement error and system error of a spacecraft is achieved.
[0142] In summary, the filtering method of traditional target tracking is to filter the position and other state information of the target in the short-range atmosphere. The tracking filtering of the three-dimensional perspective of the spacecraft on the target in the orbital space lacks a real environment verification. Therefore, the target position velocity is first solved based on the universal variable method to determine the orbit. Based on this situation, the relative motion model of the spacecraft and the target is constructed to determine the state equation of the relative line of sight angle change. Finally, a relative line of sight angle filtering method based on CS is designed to filter the measured relative line of sight angle to reduce its measurement error and system error.
[0143] Example
[0144] In order to verify the effectiveness of the scheme of the present invention, the following experiment was carried out.
[0145] Simulation environment construction
[0146] When the distance between the spacecraft and the target reaches 5 kilometers, it begins to enter the guidance and tracking target phase. The target orbit altitude is 36000.22m, and the spacecraft orbit altitude is 36003.25m. The initial position and velocity of the spacecraft terminal guidance are the state information when the distance between the spacecraft and the target is 5 kilometers, and the initial attitude angle of the spacecraft body is consistent with the initial velocity angle.
[0147] Assume that when the end guidance starts, the initial position (xt0, yt0, zt0) and velocity (vxt0, vyt0, vzt0) of the target are:
[0148]
[0149] The initial position (xm0, ym0, zm0) and velocity (vxm0, vym0, vzm0) of the spacecraft are:
[0150]
[0151] The initial yaw and pitch angles of the spacecraft correspond to θ0 = 224.2°, The roll angle is always 0, the initial relative distance between the spacecraft and the target is 4990m, the azimuth and elevation angles of the spacecraft and the target relative line of sight are α0=189.2°, β0=-7.2°, the measurement angle error is 0.1°, and the initial relative speed and direction between the spacecraft and the target can be expressed as:
[0152]
[0153] Simulation content and result analysis
[0154] The initialization information for relative sight azimuth filtering is set as shown in Table 1 below:
[0155] Table 1 Initialization information of the relative line of sight azimuth filter between the spacecraft and the target
[0156]
[0157]
[0158] Similarly, the initialization information of the relative line of sight height and low angle filtering between the spacecraft and the target is set as shown in Table 2 below:
[0159] Table 2 Initialization information of the relative line of sight height and low angle filtering between the spacecraft and the target
[0160]
[0161] The three-dimensional orbital position of the on-orbit target after simulation includes the estimated position and the real position. The real position is the orbital position simulated by STK, and the estimated position is obtained by iterative calculation using the universal variable method.
[0162] Comparing the actual position of the target orbit with the estimated position, the error diagram of the estimated target orbit position in the three-axis direction in the inertial coordinate system is as follows: Figure 3 shown.
[0163] correspond Figure 3 , the specific values of the error of the estimated target orbit position at different times are shown in Table 3 below:
[0164] Table 3. Estimated three-axis error values of target orbit position at different times
[0165]
[0166] It is observed that in the three-axis directions under the inertial coordinate system, the error between the estimated target orbit position at the same time and the true target orbit position is in an increasing relationship with time t. As the time for estimating the target orbit increases, the estimated error increases. Based on the research environment of the present invention, when the target orbit position is estimated for a maximum of 60s, the errors in the three-axis directions are all less than 0.1m, which is negligible compared to the target size. It can be seen that the target orbit estimated by the universal variable method can be approximated to the true orbit, which plays an irreplaceable role in the study of the terminal approach guidance law in aerospace missions.
[0167] The estimated speed of the target calculated by the iterative universal variable method is compared with the actual speed derived from the STK simulation, and the actual speed of the target at the same time is compared with the estimated speed. The error diagram of the estimated speed of the target at different times in the orbit in the three-axis direction in the inertial coordinate system is shown in the figure below. Figure 4 shown.
[0168] In addition, the specific values of the errors of the target speed in the three-axis directions in the inertial coordinate system at different times can be shown in the following Table 4:
[0169] Table 4. Estimated three-axis error values of target speed at different times
[0170]
[0171] according to Figure 4 As shown in Table 4, the target speed estimated by the universal variable method has a small error compared with the actual target speed within 60 seconds. It can be seen that the estimated speed error does not increase with time, and the speed error is stable near the average value. The average error in the X-axis direction is about 4×10 -4 m / s, and the average error in the Y-axis direction is about 5×10 - 4 m / s, and the average error in the Z-axis direction is about 2.5×10 -4 m / s. From the above simulation environment, we can know that the speed of the target in the high orbit is about 3.04km / s. It can be seen that the estimated speed error is negligible compared with the target running speed, and the estimated target speed can be approximated to the true target speed.
[0172] The absolute position error and absolute velocity error obtained by comparing the estimated target orbit position and velocity in three-dimensional space with the actual orbit position and velocity are as follows: Figure 5 shown.
[0173] In addition, the estimated track absolute position and velocity error values of the target in three-dimensional space at different times are shown in Table 5 below:
[0174] Table 5. Estimated track absolute position and velocity error values of the target in three-dimensional space at different times
[0175]
[0176]
[0177] It is observed that the error of the estimated target’s three-dimensional position in orbit within 60 seconds increases with the increase of estimation time, and the maximum error reaches 0.0337m, which is much closer to 0 than the target size. The error of the estimated target velocity fluctuates within 60 seconds with an average value of about 0.6×10 -3 m / s, and the actual speed of the target orbit is also close to 0. Therefore, the target orbit and speed estimated by the universal variable method can completely approximate the actual orbit position and speed in the space short-range guidance research, and the error can be ignored.
[0178] Based on this orbit and the relative line of sight angle filtering method, the relationship between the relative line of sight azimuth angle between the spacecraft and the target before and after filtering and the true value over time during the entire spacecraft guidance process is as follows Figure 6 As shown:
[0179] In addition, in order to more clearly compare the effect of the filtering method on the measurement error, the root mean square error of the measured value of the relative line of sight azimuth between the spacecraft and the target and the root mean square error of the azimuth after filtering were calculated 100 times during the entire spacecraft terminal guidance process. The relationship between the root mean square error of the relative line of sight azimuth before and after filtering and time is as follows Figure 7 shown.
[0180] Similarly, the relationship between the relative sight height angle between the spacecraft and the target before and after filtering and the true value over time during the entire spacecraft terminal guidance process is as follows: Figure 8 shown.
[0181] After 100 Monte Carlo simulations, the relationship between the relative line of sight height angle between the spacecraft and the target before and after filtering and the time change of the root mean square error is as follows: Fig. 9 shown.
[0182] from Figure 8 and Figure 6 It can be seen that the relative sight angle KF method based on CS can respond quickly to reduce the measurement error, making the filtered relative sight angle closer to the true value without causing large fluctuations. Fig. 9 and Figure 7 It can be found that the response time of the filtering method is about 5s, and the root mean square error of the relative line of sight azimuth and elevation angle before filtering is reduced from 0.0017rad to 0.00025rad. It can be seen that the filtering method is effective in reducing measurement errors.
[0183] In summary, the relative line of sight angle filtering method for on-orbit targets proposed in the present invention is more suitable for practical engineering applications in which spacecraft approach non-cooperative targets in space.
[0184] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0185] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.
Claims
1. A relative sight angle filtering method for an on-orbit target, characterized in that: The following steps are involved: Step 1: Design target trajectory based on universal variable method: According to the principle that the target is only affected by the earth's gravity in a vacuum environment, the target motion equation in the earth coordinate system is constructed to obtain the functional relationship between the target position and velocity over time. At the same time, the target angular momentum and energy equations are established. A universal variable τ is introduced to solve the target motion trajectory using the universal variable method. The target position r at any time is obtained by Newton iteration calculation. T1 and speed v T1 ; Step 2: Construct the relative motion model between the spacecraft and the target: According to the change law of the position and speed of the spacecraft in three-dimensional space, the relative motion relationship between the spacecraft M and the target T1, which are split into the horizontal plane xoy and the vertical plane yoz in the inertial coordinate system, is obtained, and the relative motion model between the two is obtained. The azimuth and elevation angle of the spacecraft speed, and the azimuth and elevation angle of the approach speed of the spacecraft and the target are calculated; the target relative line of sight angle motion model is simplified, and the relative line of sight angle change state equation is obtained; Step 3: Filter the relative sight angle: According to the relative line of sight angle motion model of the spacecraft to the target, the state equation and observation equation of the relative line of sight angle KF system based on CS are constructed to describe the state change relationship of the relative line of sight azimuth, azimuth velocity, azimuth acceleration, pitch angle, pitch angle velocity, and pitch angle acceleration in three-dimensional conditions; the "current" statistical model is designed to calculate the state prediction value and the prediction error covariance P(k+1 / k), calculate the Kalman filter gain K(k), update the covariance matrix P(k) and the system state quantity X(k+1), that is, the relative line of sight angle state quantity after filtering, and the relative line of sight angle accuracy is improved after filtering.
2. The method for filtering relative sight angles of on-orbit targets according to claim 1, characterized in that: Step 1: Design the target spacecraft orbit based on the universal variable method, specifically: The position and velocity variables are calculated and solved in the target's orbital plane, and the current state vector position R and velocity V are solved using the universal variable method, as shown in the following formula: R=fR0+gV0 Where R0 and V0 are the initial position vector and velocity vector of the target. Differentiating the above equation, we get: f,g, Both are scalar quantities related to time; Define intermediate variables z, C(z), S(z): Where a is the semi-major axis of the orbit; τ is a universal variable, Combining the trigonometric identities for the sine and cosine of an angle, and z, C(z), S(z), we get f, g, The relationship with τ is as follows: Where t represents time, r0 represents the initial position of the target, u is the gravitational constant and r is the position vector of the spacecraft; The motion trajectory of the target can be calculated by solving the above formula.
3. The method for filtering relative sight angle of an on-orbit target according to claim 2, characterized in that: Step 2: Construct the relative motion model between the spacecraft and the target. The specific method is as follows: The target moves under the action of gravity and does not maneuver. The passive external forces on the spacecraft include gravity and aerodynamic thrust. According to Newtonian kinematics, the relative motion model between the two is expressed as: Among them A T1 is the target acceleration caused by the earth's gravity, A M is the acceleration of the spacecraft caused by the Earth’s gravity and the thrust of the cold gas jet. The Earth’s gravitational constant G = 6.67259 × 10 -11 N·m 2 / kg 2 , Earth mass M d =5.965×10 24 kg, A control is the control command acceleration acting on the spacecraft, Respectively represent R M , R T1 The second derivative of The relative position vector R and the absolute position vector R between the spacecraft and the target T1 , R M The relative position relationship is R = R T1 -R M ; According to the relative motion characteristics of the spacecraft and the target, the relative line of sight angle between the spacecraft and the target changes continuously. Assuming that the spacecraft is used as a fixed-point platform and the target is used as a maneuvering target relative to the fixed-point spacecraft, a simplified target relative line of sight angle motion model is constructed, and its equation can be expressed as: x k+1 =Ax k +Bu k +Cd k in: In the above formula, time k = 0, 1, 2, ..., α, β are the azimuth and elevation angles relative to the line of sight, T t is the period, w ak ,w bk is the angular velocity of rotation relative to the line of sight azimuth and pitch angle, is the angular acceleration relative to the azimuth and pitch angle of the line of sight, which is in a time-varying state. A is the state coefficient matrix, B is the control variable coefficient matrix, C is the second control variable matrix coefficient, and x k is the current state matrix, x k+1 is the state matrix at the next moment, u k is the control variable matrix at the current moment, d k is the second control variable matrix.
4. The method for filtering relative sight angles of on-orbit targets according to claim 3, characterized in that: Step 3: filter the relative sight angle. The specific method is as follows: According to the relative line of sight angle motion model of the spacecraft to the target described in step 2, the state equation and observation equation of the relative line of sight angle KF system based on CS are constructed, where the system state equation is expressed as: In the above formula: F(k+1,k) is the state transfer matrix; Ω(k) is the process input matrix; are the state vectors, which are the azimuth, azimuth velocity, azimuth acceleration, pitch angle, pitch velocity, and pitch acceleration of the target relative to the line of sight in three dimensions; is the average acceleration of the azimuth and pitch angle relative to the line of sight, where are the variances of the azimuth acceleration and the pitch acceleration relative to the line of sight respectively; W(k) is a discrete time white noise sequence; The state transfer matrix F(k+1,k) of the system is: The process input matrix Ω(k) is: The covariance Q(k) of the white noise sequence is: The system observation equation Y(k / k) is: Y(k / k)=H(k)X(k / k)+V(k) The measurement data of the spacecraft include the azimuth angle α relative to the line of sight between the spacecraft and the target. m (k) and pitch angle β m (k), 0 3×3 represents a 3×3 zero matrix, V(k) is the measurement noise sequence, whose covariance is R(k); q represents the coefficient matrix; Φ(k+1,k) is the state transfer matrix, U(k) is the process input matrix, It is the reciprocal of the time constant. The more drastic the acceleration change, the larger the corresponding time constant. are the variances of azimuth acceleration and pitch acceleration relative to the line of sight respectively; The measurement matrix is H(k), which is as follows: Y(k / k)=[a m (k),b m (k)] T H(k)=[1,0,0,0,0,0;0,0,0,1,0,0].
5. The method for filtering relative sight angles of on-orbit targets according to claim 1, characterized in that: In step 3, the relative sight angle is filtered as follows: 1) Initialization: The initial measurement value of the relative line of sight is used as the initial state quantity X(0), and the covariance R(0) and the initial estimated covariance P(0) are constructed according to the measurement error of the spacecraft measurement equipment. The preset minimum negative angular acceleration and the preset maximum positive angular acceleration (a) of the relative line of sight azimuth and pitch angle angular acceleration changes are determined according to the space orbit dynamics and the aerodynamic thrust of the spacecraft. αmin ,a αmax ,a βmin ,a βmax ); 2) Calculate the state prediction value and prediction error covariance: According to the relationship between the angular acceleration estimate and the variance of the current statistical model, the variance of the azimuth and pitch angle angular acceleration relative to the current line of sight is calculated; the angular acceleration of the current azimuth relative to the line of sight is calculated. For positive time, Current relative line of sight azimuth angular acceleration When it is negative, Similarly, The calculation of One-step prediction of the state for: Among them, the one-step prediction of the angular acceleration state can be regarded as the mean of the current angular acceleration The specific formula is as follows: Then the one-step prediction for this state will become as follows: in: Intermediate variables Intermediate variables The one-step state prediction error covariance P(k+1 / k) is: P(k+1 / k)=Φ(k+1,k)P(k)Φ(k+1,k) T +Q(k) 3) Calculate the Kalman filter gain K(k): K(k)=P(k+1 / k)H(k) T (H(k)P(k+1 / k)H(k) T +R(k)) -1 Where R(k) represents the covariance; 4) Update the covariance matrix P(k): P(k)=P(k+1 / k)-K(k)H(k)P(k+1 / k) 5) Update the system state X(k+1):
6. A method for filtering relative sight angles of on-orbit targets, characterized in that: A relative sight angle filtering process for on-orbit targets based on any one of claims 1-5.
7. A computer device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, a relative sight angle filtering process for an on-orbit target according to any one of claims 1 to 5 is implemented.
8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, based on the relative sight angle filtering method for on-orbit targets described in any one of claims 1 to 5, a CS-based relative sight angle Kalman filtering process for on-orbit targets is implemented.
Citation Information
Cited By
Driver response time detection method for real vehicle scene and vehicle control method
CN121040913A
