CN106017480B - Depth-integrated navigation method for deep space exploration capture segment - Google Patents

Abstract

The present invention provides a kind of depth Combinated navigation method towards deep space exploration capture section, including preproduction phase and filtering stage；Establish that the dynamics of orbits model of deep space probe, direction finding model, ranging model, test the speed model the preproduction phase；The filtering stage is filtered using extended Kalman filter, state transition model in Navigation Filter is dynamics of orbits model, measurement model in Navigation Filter is optionally comprised in pulse observation cycle, direction finding model or the model that tests the speed are selected when not obtaining ranging information, and the received pulse signal of X-ray sensitive device is compensated；When pulse signal accumulation, which finishes, obtains ranging information, ranging model is selected；Navigation Filter utilizes ranging model, according to acquisition navigation desired position and velocity vector.Present invention inhibits pulse arrival time Doppler's deviations, filter convergence, and positioning accuracy is high, and very low to sensor requirements.Therefore, the present invention has important practical significance to Spacecraft Autonomous Navigation.

Description

Depth-integrated navigation method for deep space exploration capture segment

technical field

The invention belongs to the field of autonomous navigation of spacecraft, and in particular relates to a combined navigation method for direction finding/ranging/velocity measurement and depth oriented to the capture section of deep space exploration.

Background technique

Navigation information is the premise of guidance and is crucial to the success or failure of deep space exploration. Affected by the ultra-long distance and the long time delay, the ground station cannot provide real-time high-precision navigation information, especially in the acquisition section. Since 1990, deep space exploration missions have failed 7 times, 4 of which were related to the capture segment. The spacecraft autonomous navigation system can provide real-time high-precision autonomous navigation for deep space probes by measuring and solving celestial body information. Therefore, for the capture segment, astronomical autonomous navigation is extremely important. The capture segment is a highly dynamic environment whose orbital dynamics model is strongly nonlinear and time-varying. It is also extremely difficult to achieve high-precision autonomous navigation in this environment.

At present, in the field of deep space exploration, there are mainly the following autonomous navigation measurement methods: (1) X-ray pulsar ranging navigation. X-ray pulsar ranging and navigation By observing the pulsar radiation signals and accumulating them according to the pulse period, a high signal-to-noise ratio cumulative profile can be obtained, and the pulse TOA (time-of-arrival, time-of-arrival) can be obtained by comparing it with the standard profile. ), and high-precision ranging information can be obtained by solving TOA. However, in the highly dynamic environment of the capture segment, due to the variable acceleration flight of the spacecraft, the pulsar radiation signal is greatly affected by the Doppler effect, and it is difficult to compensate. The pulse profile is greatly distorted, and the pulse arrival time has a large deviation. This will seriously affect navigation performance. (2) Direction finding navigation. Direction-finding navigation is a traditional astronomical navigation method, which obtains the orientation information of the spacecraft relative to the celestial body by measuring the near celestial body. However, this method cannot provide high-precision distance information between spacecraft and near celestial bodies. Especially for Venus, a planet without natural satellites, the spacecraft can only obtain position information relative to a celestial body. The distance information between the spacecraft and the near celestial body is completely unavailable. This navigation filter is very prone to divergence. (3) Speed navigation. Velocimetry navigation obtains information on the speed of the spacecraft relative to the star by measuring the spectral frequency shift of the sun. Its speed measurement accuracy is high. However, the speed measurement navigation method cannot directly provide the position information, and the position information is obtained by integrating the speed information. Therefore, the velocity error accumulates for a long time, and there must be a large integral error in the position information. In general, speed navigation cannot work alone, and is often used as an auxiliary means for other navigation methods.

The above three methods have their own advantages and disadvantages. Some scholars have combined them, such as: angle measurement/ranging integrated navigation (Ma Jie, Liu Jin, Tian Jinwen. A pulsar/CNS integrated navigation method, national invention patent, ZL 2009100632674), ranging/speed integrated navigation (Liu J, Kang Z W, White P, Ma J, Tian J W. Doppler/XNAV-integrated navigation system using small-area X-ray sensor, IET Radar, Sonar and Navigation. 5(9):1010-1017). However, this is not a depth combination, and the influence of the Doppler effect of pulsar ranging and navigation is not considered, and it is even more incapable of adapting to the high dynamic environment of the capture segment.

To sum up, in the high dynamic environment of the deep space exploration capture segment, the X-ray pulsar ranging navigation system is seriously affected by the Doppler effect, and its measurement information has a large deviation; the speed measurement navigation cannot provide high-precision positioning. Information; DF navigation has acceptable tangential accuracy, but extremely low radial accuracy. That is to say, a single navigation method cannot be used for the task of high-precision autonomous navigation in the deep space exploration and capture segment.

SUMMARY OF THE INVENTION

The invention proposes a combined navigation method of direction finding/ranging/velocity measurement for the deep space exploration capture section, aiming at providing the spacecraft with high-precision positioning and constant speed autonomous navigation information in the deep space exploration capture section. The word "depth" now uses the direction finding and speed measurement information to assist the acquisition of ranging information, specifically to use the direction finding and speed measurement information to provide high-precision three-dimensional velocity estimation information, and use it to compensate for the arrival time of pulsar photons caused by the high-speed flight of the spacecraft error, so as to achieve the effect of suppressing Doppler deviation. Integrated navigation uses direction finding, ranging, and speed measurement information to update the state of the navigation filter to achieve multi-source information fusion.

The present invention provides a depth integrated navigation method oriented to the deep space exploration capture segment, including a preparation stage and a filtering stage,

The preparatory stage, including establishing various models required for navigation filtering, includes the following steps:

Step A1, establishing an orbital dynamics model of the deep space probe, which is implemented as follows,

Let the state vector X of the deep space probe be,

Among them, r=[x, y, z] ^T and v=[v _x , v _y , v _z ] ^T are the position and velocity vectors of the deep space detector, respectively, and x, y, z are the The components of the position on the three axes, v _x , v _y , and v _z are the components of the speed of the deep space probe on the three axes;

Then the orbital dynamics model of the deep space probe is,

in, are the derivatives of x, y, z, v _x , v _y , v _z , respectively,

Formula (2) is expressed as,

in, is the derivative of the state vector X, for time t f(X,t) is the state transition model of the deep space probe, [x ₁ , y ₁ , z ₁ ] and [x ₂ , y ₂ , z ₂ ] are the relative position vectors of Venus and Earth with respect to the center of mass of the solar system, respectively , μ _s , μ _v , μ _e are the gravitational constants of the sun, Venus and the Earth, respectively, _{rp s , r pv ,} r _pe are the distances from the deep space probe to the center of mass of the Sun, the center of Venus and the center of mass of the Earth, respectively; r _sv , r _se are the distances from the centroid of Venus, the centroid of the Earth to the centroid of the sun, respectively; ω(t) is the noise of the navigation system of the deep space probe at time t;

Step A2, establishing a direction finding model;

Step A3, establishing a ranging model;

Step A4, establishing a speed measurement model;

The filtering stage utilizes extended Kalman filter filtering, the state transition model included in the navigation filter is an orbital dynamics model, and the measurement model selection in the navigation filter comprises the following steps,

The filtering stage utilizes extended Kalman filter filtering, the state transition model included in the navigation filter is an orbital dynamics model, and the measurement model selection in the navigation filter comprises the following steps,

Step B1, in the current pulse observation period, when the ranging information of the current pulse observation period is not obtained, select a direction finding model or a velocity measurement model, and use the epoch stacking method based on Doppler compensation to detect the pulse received by the X-ray sensor. The signal is compensated, and the compensation is realized as follows,

Step B11, the X-ray sensor records the arrival time of a single X-ray photon;

Step B12, performing Doppler compensation on the arrival time of the X-ray photons, the process is as follows,

(a) Estimate the current speed of the spacecraft When the filter has feedback, the value adopts the feedback value; otherwise, it is obtained by integrating equation 3;

(b) Utilize Compensate the arrival time of X-ray photons according to Equation 4;

The i-th sub-pulse Doppler compensation amount is expressed as follows,

Among them, the i-th pulse period is P _i , the number of pulses in the pulsar observation period is N, t _i is the arrival time of the i-th sub-pulse, n is the pulsar azimuth vector, T is the transposition, and c is the k-th speed of light The pulse period is P _k , is the spacecraft velocity vector in the kth pulse period;

Step B13, superimpose the photons according to the predicted pulse period to obtain the pulse TOA, and obtain the ranging information;

Step B2, when the accumulation of pulse signals is completed and ranging information is obtained, a ranging model is selected;

Step B3, the navigation filter uses the ranging model to process the received pulse TOA, Venus azimuth and Doppler velocity to obtain a state vector, and obtain the position and velocity vector required for navigation; after the current pulse observation period ends, return to Step B1, continue the navigation of the next pulse observation period.

Moreover, in step A2, the direction finding model is established as follows,

where Z is the direction finding value, r _v is the Venus position vector, and υ is the direction finding noise;

Moreover, in step A3, the ranging model is established as follows,

where, t and t _b are the times when the pulse reaches the spacecraft and the solar system mass center, respectively; n is the pulsar azimuth vector; D ₀ is the distance from the pulsar to the solar system mass center, b is the position vector of the solar system mass center relative to the solar system mass center, and c is The speed of light, σ is the TOA measurement noise, and |·| denotes the magnitude of the vector.

Moreover, in step A4, the establishment of the speed measurement model is as follows,

where V is the speed measurement value and υ is the speed measurement noise.

The invention suppresses the pulse arrival time Doppler deviation, the filter converges, the positioning accuracy is high, and the requirements for the sensor are very low. Therefore, the present invention has important practical significance for autonomous navigation of spacecraft.

The advantages of the present invention compared with the prior art are:

(1) The autonomous navigation system of spacecraft has lower requirements on sensors. In general, high-precision navigation requires high-precision sensors. However, the sensors (astronomical optical navigation camera, X-ray sensor, spectrometer) required by the present invention are all existing, and need not be re-developed or collected, which saves cost and time. In addition, this method can obtain high-precision navigation and positioning information even if a low-resolution astronomical optical navigation camera, a small-area X-ray sensor, and a low-precision spectrometer are used, which can meet the needs of deep space exploration. Only one of the above detectors is needed, which reduces the load.

(2) The present invention can realize high-precision navigation in the capture section. The capture segment is a highly dynamic environment. The orbital dynamics model at this time is strongly nonlinear and time-varying. The pulsar signal is greatly affected by Doppler, and the effect of Doppler compensation method is limited. In addition, the navigation filter is very prone to divergence. The invention suppresses the Doppler effect by using the direction finding and speed measurement information, and fully utilizes the direction finding, ranging and speed measurement information, and realizes high-precision deep autonomous navigation in the capture section.

Description of drawings

FIG. 1 is a flow chart of combined direction finding/ranging/velocity measurement depth navigation according to an embodiment of the present invention.

Detailed ways

The technical solution of the present invention can support the automatic running process by means of computer software. The technical solutions of the present invention will be described in detail below with reference to the accompanying drawings and embodiments. EKF, short for Extended Kalman Filter, that is, Extended Kalman Filter.

The acquisition segment of deep space exploration is a highly dynamic environment, and the pulsar signal is greatly affected by the Doppler effect. During the pulsar observation period, the present invention uses direction finding and speed measurement information to compensate for the Doppler deviation in the pulsar signal, which reflects the word "depth"; the integrated navigation method includes establishing an orbital dynamics model, and direction finding, ranging , the velocity measurement navigation model, filtered using the extended Kalman filter. The present invention takes the Venus detector as an embodiment.

First, the Venus Express orbit is given, as shown in Table 1.

Table 1 Initial orbit parameters of Venus Express

Embodiments can be divided into a preparation phase and a filtering phase.

The preliminary stage of the embodiment establishes various models required for navigation filtering, specifically:

Step A1: Establish the orbital dynamics model of the deep space probe. The orbital dynamics model is the prediction model of the extended Kalman filter. The specific implementation process is as follows:

Because the state vector X of the deep space probe is:

Among them, r=[x, y, z] ^T and v=[v _x , v _y , v _z ] ^T are the position and velocity vectors of the deep space detector, respectively, and x, y, z are the The components of the position on the three axes, v _x , v _y , and v _z are the components of the speed of the deep space probe on the three axes respectively; then the orbital dynamics model of the deep space probe is:

in, are the derivatives of x, y, z, v _x , v _y , v _z , respectively,

Formula (2) can be expressed as:

in, is the derivative of the state vector X, for time t f(X,t) is the state transition model of the deep space probe, [x ₁ , y ₁ , z ₁ ] and [x ₂ , y ₂ , z ₂ ] are the relative position vectors of Venus and Earth with respect to the center of mass of the solar system, respectively , μ _s , μ _v , μ _e are the gravitational constants of the sun, Venus and the earth, respectively;

r _ps , r _pv , and r _pe are the distances from the deep space probe to the center of mass of the Sun, the center of Venus and the center of the Earth, respectively. The calculation formula is:

are the distances from the centroid of Venus, the centroid of the Earth to the centroid of the sun, respectively; the noise of the navigation system of the deep space probe ω=[0,0,0,ΔF _x ,ΔF _y ,ΔF _z ] ^T , where ΔF _x ,ΔF _y and ΔF _z are the perturbation force, and ω(t) is the navigation system noise of the deep space probe at time t.

Step A2: Establish a direction finding model.

where Z is the DF value, r _v is the Venus position vector, and υ is the DF noise.

Step A3: Establish a ranging model.

where t and t _b are the times when the pulse reaches the spacecraft and the solar system's center of mass, respectively. n is the pulsar azimuth vector. _D0 is the distance from the pulsar to the solar system barycenter, and b is the position vector of the solar system barycenter relative to the sun's barycenter. c is the speed of light. σ is the TOA measurement noise. where |·| represents the magnitude of the vector.

Step A4: Build a speed measurement model.

where V is the speed measurement value and υ is the speed measurement noise.

During specific implementation, the execution order of steps A2, A3, and A4 can be adjusted in sequence, or executed in parallel.

The filtering stage of the embodiment utilizes extended Kalman filter filtering, which is specifically implemented as follows:

The state transition model in the navigation filter is an orbital dynamics model. The measurement model selection method in the navigation filter is as follows:

Step B1: In the current pulse observation period, when the ranging information of the current pulse observation period is not obtained, on the one hand, a direction finding model or a speed measurement model is selected, and in the specific implementation, the corresponding model can be used according to the specific situation, that is, the direction finding information is obtained. When the speed measurement information is obtained based on the astronomical optical navigation camera, the direction finding model is used, and the speed measurement model is used based on the spectrometer. On the other hand, the pulse signal received by the X-ray sensor is compensated by the epoch stacking method based on Doppler compensation, as follows:

Step B11: The X-ray sensor records the arrival time of a single X-ray photon.

Step B12: Doppler compensation for the X-ray photon arrival time, the process is as follows:

(a) Estimate the current speed of the spacecraft When the filter has feedback in step B4, the value is the feedback value, that is, the ranging information of the last pulse observation period obtained by the last execution of step B4 is used; otherwise, it is obtained by integrating formula (3). Wherein, for formula (3), refer to step A1.

(b) Utilize According to formula (7), the X-ray photon arrival time is compensated.

The i-th sub-pulse Doppler compensation amount can be expressed as:

Among them, the i-th pulse period is P _i , the number of pulses in the pulsar observation period is N, and the current speed of the spacecraft is That is, the spacecraft speed in the ith pulse period, t _i is the arrival time of the ith sub-pulse, n is the pulsar azimuth vector, T represents the transposition, and c is the speed of light. Correspondingly, the kth pulse period is P _k , is the spacecraft velocity vector in the kth pulse period.

Step B13: Superimpose the photons according to the predicted pulse period. That is, accumulating according to the pulse period, a high signal-to-noise ratio accumulation profile can be obtained, and the pulse TOA (time-of-arrival) can be obtained by comparing it with the standard profile. The pulse period duration unit is usually milliseconds.

Step B2: When the accumulation of pulse signals is completed and ranging information (ie, pulse TOA) is obtained, a ranging model is selected, and high-precision ranging information can be obtained by solving TOA. That is, after obtaining the pulse TOA in step B13, the ranging model can be selected instead.

Step B3: The navigation filter uses the currently selected ranging model to process the received pulse TOA, Venus azimuth and Doppler velocity to obtain a state vector. This value is the position and velocity vector required for navigation. In addition, this value can also be used to compensate the X-ray photon arrival time in the subsequent execution B12. After the current pulse observation period ends, return to step B1 to continue the navigation of the next pulse observation period. The pulse observation period is the ranging observation period.

The design of the extended Kalman filter involves prediction models, measurement models and related parameters. After the model and parameters are set, the filter can filter the measured values, and the obtained state estimate includes the position and velocity information components, which is the navigation result.

The filter parameters are shown in Table 2:

Table 2 Navigation filter parameters

where P(0) is the initial state error matrix, Q is the state noise covariance, i.e. the square of q ₁ , That is, the square of _q2 .

The specific embodiments described herein are merely illustrative of the spirit of the invention. Those skilled in the art to which the present invention pertains can make various modifications or additions to the described specific embodiments or substitute in similar manners, but will not deviate from the spirit of the present invention or go beyond the definitions of the appended claims range.

Claims (1)

1. A deep integrated navigation method facing a deep space exploration capturing section is characterized in that deep integrated navigation is carried out based on a Venus probe, and comprises a preparation stage and a filtering stage,

the preparation phase, including the building of the various models required for navigation filtering, includes the following steps,

step A1, building an orbit dynamics model of the deep space probe, which is realized as follows,

the state vector X of the deep space probe is set as,

wherein r ═ x, y, z]^TAnd v ═ v_x,v_y,v_z]^TRespectively the position and velocity vector of the deep space probe, x, y, z are the components of the position of the deep space probe in three axes, respectively, v_x,v_y,v_zThe components of the speed of the deep space probe on three axes are respectively;

the orbit dynamics of the deep space probe is modeled as,

wherein,are respectively x, y, z, v_x,v_y,v_zThe derivative of (a) of (b),

the formula (2) is represented by the formula,

wherein,is the derivative of the state vector X and,at a time tf (X, t) is the state transition model of the deep space probe, [ X ]₁,y₁,z₁]And [ x ]₂,y₂,z₂]Respectively, the relative position vectors, mu, of the Venus and Earth with respect to the center of mass of the solar system_s,μ_v,μ_eThe gravitational forces of the sun, the Venus and the EarthNumber r_ps,r_pv,r_peThe distances from the deep space probe to the sun centroid, the golden star centroid and the earth centroid are respectively; r is_sv,r_seThe distances from the star centroid and the earth centroid to the sun centroid respectively; omega (t) is the navigation system noise of the deep space probe at the moment t;

step a2, a direction-finding model is built as follows,

wherein Z is a direction finding value, r_vIs a Venus position vector, upsilon is direction finding noise;

step a3, a ranging model is established as follows,

wherein t and t_bRespectively the time of arrival of the pulse at the spacecraft and solar system centroids; n is the pulsar azimuth vector; d₀Is the distance from the pulsar to the solar system centroid, b is the position vector of the solar system centroid relative to the solar system centroid, c is the speed of light, σ is the TOA measurement noise, | · | represents the mode of the vector;

step a4, a velocity model is established as follows,

v is a velocity measurement value, and upsilon is velocity measurement noise;

the filtering stage utilizes an extended Kalman filter for filtering, the state transition model included in the navigation filter is an orbit dynamics model, the measurement model selection in the navigation filter comprises the following steps,

step B1, when the distance measurement information of the current pulse observation period is not obtained in the current pulse observation period, selecting a direction-finding model or a speed-measuring model, and compensating the pulse signal received by the X-ray sensor by using an epoch stacking method based on Doppler compensation, wherein the compensation is realized as follows,

step B11, the X-ray sensor records the arrival time of the single X-ray photon;

step B12, doppler compensation is performed on the X-ray photon arrival time, as follows,

(a) estimating a current velocity of a spacecraftWhen the filter has feedback, the value adopts a feedback value; otherwise, obtaining the result through an integral formula III;

(b) by usingCompensating the X-ray photon arrival time according to the formula four;

i sub-pulse Doppler compensation quantityAs shown below, the following description is given,

wherein the ith pulse period is P_iThe number of pulses in the pulsar observation period is N, t_iIs the arrival time of the ith sub-pulse, n is the orientation vector of the pulsar, T represents transposition, c is the speed of light, k is the pulse period P_k，Is the spacecraft velocity vector in the kth pulse period;

step B13, overlapping the photons according to the predicted pulse period to obtain a pulse TOA and obtain ranging information;

step B2, when the pulse signal is accumulated and the ranging information is obtained, selecting a ranging model;

step B3, the navigation filter processes according to the received pulse TOA, the Venus azimuth and the Doppler velocity by using a ranging model to obtain a state vector and obtain a position and a velocity vector required by navigation; after the current pulse observation period is finished, the procedure returns to step B1 to continue the navigation of the next pulse observation period.