CN103674034B - Multi-beam test the speed range finding revise robust navigation method - Google Patents

Abstract

Multi-beam test the speed range finding revise robust navigation method, first carry out inertial navigation calculating, then carry out range finding revise, next carry out speed correction, range finding revise adopt information source for find range sensor; The signal source of speed correction is divided into two kinds, one is distance measuring signal, and another kind is tachometer signal: when distance measuring signal is effective, and flying height meets and tests the speed when revising introducing scope, the height solved by double distance measuring signal carries out difference, obtains the speed of gravity direction; When tachometer signal is effective, and flying height meets and tests the speed when revising introducing scope, directly obtains the speed along the sensor beam direction that tests the speed.Then utilize the speed in inertial navigation the first two cycle to calculate the lander velocity inertial calculating range finding or tachometer signal correspondence moment, be converted to the speed of relative lunar surface afterwards and project to short transverse or the beam direction that tests the speed.The present invention is simple and practical, and reliability is high, ensure that the high precision that navigation is omnidistance.

Description

Multi-beam test the speed range finding revise robust navigation method

Technical field

The present invention relates to a kind of multi-beam test the speed range finding revise robust navigation method, belong to the independent navigation field of the moon and survey of deep space soft landing task.

Background technology

Objects outside Earth soft landing is a brand-new task to China.For GNC system, obtaining navigation data is accurately the prerequisite ensureing the safe soft landing of lander.Because soft landing process time is short, independence is high, therefore external existing similar tasks, the Lunar Landing Mission that the Luna series etc. comprising Apollo, Surveyor of the U.S. and USSR (Union of Soviet Socialist Republics) has been implemented and European Space Agency and Japan plan the Lunar Landing Mission that carries out all using inertial navigation as prime navaid mode.

Inertial navigation technology relies on the direct measured angular speed of gyro and integration obtains detector attitude angle or sets up stable platform with gyro and obtain attitude by frame corners; Rely on accelerometer measures specific force, and integration obtains speed and position.This air navigation aid has the advantages that operational process does not rely on external information.Because inertial navigation is a kind of extrapolation algorithm, therefore the precision of inertial navigation depends on the error of navigation initial value and inertia device, and it presents ever-increasing feature in time.

For soft landing task, the initial value of inertial navigation is provided by ground observing and controlling and star sensor.And for objects outside Earth, the limited precision of ground observing and controlling, detector position error reaches km magnitude.In this, as navigating, initial value is dangerous concerning soft landing.The topographic relief of adding celestial body surface also has uncertainty.Therefore it is infeasible for relying on merely inertial navigation to complete soft landing navigation task.

For this reason, external objects outside Earth soft landing task is all equipped with microwave or laser ranging or the sensor that tests the speed, to revise inertial navigation.But due to the difference of the difference of various countries' sensor technical merit, the difference of detector scale and soft landing guidance algorithm, the concrete grammar utilizing range finding or the information that tests the speed to revise inertial navigation is not identical yet.Such as, the Apollo lunar module of the U.S. have employed Microwave Velocity, range finding sensor, and this sensor has three test the speed wave beam and distance measurement beams.Three wave beams that test the speed work respectively, and then geometric solution calculates the speed of lunar module to lunar surface.This sensor has two different gears in addition, can ensure that distance measurement beam point to lunar surface as far as possible, thus obtain the elevation information of lunar module according to the attitude of lunar module.Refer to " BernardA.Kriegsman; Radar-updatedinertialnavigationofacontinuously-poweredsp acevehicle; IEEEtransactiononAerospaceandElectronicSystems, Vol.AES-2, No.4; July; 1966,549-565 " and " FloydV.Bennett, Apollolunardescentandascenttrajectories; NASATMX58040, March1970 ".The method of filter correction have employed the kalman filter method of simplification.The shortcoming that said method exists is: (1) needs three wave beams simultaneously effectively, and fault-tolerance is bad; (2) only having a distance measurement beam, in order to adapt to the adjustment of large attitude, needing movable part, reduce the reliability of system.

The navigation sensor that China lunar exploration second phase soft landing adopts is different from external inter-related task, except for except the IMU of inertial navigation, is also equipped with tellurometer survey, test the speed sensor and laser ranging sensor be for revising inertial navigation.In order to avoid sensor rotating mechanism fault works the mischief to landing mission, tellurometer survey, the sensor that tests the speed of China are different from Apolline way, are fixed installation.But in order to ensure can both to test the speed and ranging information in 90 ° of attitude slewing areas in landing mission, the Microwave Velocity of outfit, range finding sensor have three different test the speed wave beam and two different distance measurement beams pointed to pointed to.Add laser ranging sensor and also there are two different distance measurement beams pointed to, therefore can be used in China's soft landing task testing the speed, the number of beams revised of finding range is much larger than Apollo.These wave beams are measured also asynchronous, and precision is different, how utilizing these information and taking into account the needs of System Fault Tolerance effectively and reasonably, and this becomes the difficult problem that navigation algorithm needs to solve.

Summary of the invention

Technology of the present invention is dealt with problems: overcome the deficiencies in the prior art, provide the robust navigation method that a kind of simple and practical multi-beam tests the speed range finding correction, and reliability is high, ensure that the precision that navigation is omnidistance.

The technical solution adopted in the present invention is: multi-beam test the speed range finding revise robust navigation method, performing step is as follows

(1) inertial reference calculation

The angle step of gyro to measure in the previous sampling period is utilized to calculate the inertial attitude of current time lander, the speed of the speed increment accumulation calculating current time lander then utilizing accelerometer measures to obtain, and integration obtains the position of lander under inertial coordinates system;

(2) range finding is revised

If receive the valid data of range finding sensor, and flying height meets the introducing scope that range finding is revised, then these data are solved altitude information by installation matrix and the conversion of lander attitude matrix; Utilize the inertial reference calculation in the first two cycle simultaneously, calculate the lander height on time point corresponding to range finding sensor data, the height that inertial navigation and stadimeter are resolved respectively is done difference, and according to the site error in altitude correction factor corrected altitude direction; This altitude correction factor is with the change of lander highly linear, and lander height is higher, and described altitude correction factor is less, and highly more low clearance correction factor is larger;

(3) speed correction

The signal source of speed correction is divided into two kinds, and one is distance measuring signal, and another kind is tachometer signal, when distance measuring signal is effective, and flying height meet test the speed revise introducing scope time, the height solved by double distance measuring signal carries out difference, obtain the speed of gravity direction, utilize the lander velocity inertial in the speed calculation distance measuring signal in the first two cycle in inertial reference calculation corresponding moment simultaneously, and be converted to the speed of relative lunar surface, the speed that the speed obtain distance measuring signal and inertial navigation obtain does difference, and according to the speed in speed correction factor corrected altitude direction, this speed correction factor changes with lander speed linearity, lander speed is faster, speed correction factor is less, speed is less, speed correction factor is larger, effective for tachometer signal, utilize the lander velocity inertial in the speed calculation tachometer signal in the first two cycle in inertial reference calculation corresponding moment, and be converted to the speed of relative lunar surface, the speed obtain tachometer signal and the speed that obtains of inertial navigation do difference, and according to the speed of speed correction factor correction beam direction, this speed correction factor changes with lander speed linearity, lander speed is faster, speed correction factor is less, and speed is less, and speed correction factor is larger,

(4) each control cycle is repeated to the step of (1) ~ (3), complete navigation calculation.

Described altitude correction factor and speed correction factor generally change between 0 ~ 1.

The present invention's advantage is compared with prior art: the present invention does not require range finding, the wave beam that tests the speed has movable part, adds the reliability of system; In navigation procedure, any one wave beam effectively can both be revised the result of inertial reference calculation in a certain direction, is conducive to improving system survivability, reduces the complicacy of fault diagnosis algorithm; Filtering have employed the correction factor of linear change, and comparison with standard Kalman filtering calculates simple, automatically can adjust again for the different accuracy of different phase sensor to filter correction parameter, to ensure omnidistance high precision of navigating.

Accompanying drawing explanation

Fig. 1 is that multi-beam of the present invention tests the speed the theory diagram of robust navigation method that range finding revises;

Fig. 2 is the definition of reference frame;

Fig. 3 is the curve of actual lander change in location in soft lunar landing process;

Fig. 4 is the change in location curve that in soft lunar landing process, navigational system provides;

Fig. 5 is the curve of actual lander velocity variations in soft lunar landing process;

The speed change curves that in Fig. 6 soft lunar landing process, navigational system provides.

Embodiment

As shown in Figure 1, multi-beam of the present invention test the speed range finding revise the concrete computation process of robust navigation method as follows:

The first step, carries out inertial reference calculation

(1) foundation of reference frame and navigational system initial value (initialization)

Body series (O _b-X _by _bz _b) connect firmly at lander, initial point overlaps with barycenter, and three axles are parallel to the lander principal axis of inertia.

Navigational system selects landing inertial coordinates system (O _m-X _iy _iz _i) as reference coordinate system: initial point is moon heart O _m, land the moment in nominal, X _iaxle points to nominal landing point, Z _iaxle points to direction of motion, Y in the landing orbit plane of planning _iaxle forms right hand orthogonal coordinate system, as shown in Figure 2.If lander lands in nominal landing point with nominal attitude at nominal time, so landing inertial coordinates system is parallel with three axles of body coordinate system.

The initial value of navigational system comprises position, speed and attitude.Wherein position, speed are provided by ground observing and controlling, and attitude is obtained by star sensor and gyro to measure.It should be noted that, ground observing and controlling and the quick coordinate system measuring employing of star may not be landing inertial coordinates system, such as moon heart J2000 coordinate system.But due to nominal landing point and nominal landing times known, therefore just can easily these states be transferred in landing inertial coordinates system according to moon ephemeris.If position vector is r ⁱ, velocity is v ⁱ, relatively the land attitude matrix of inertial system of body is C _bi(corresponding attitude quaternion is q).

(2) attitude determination algorithm

The object of attitude determination algorithm upgrades attitude matrix C _bi, it can obtain according to gyro to measure recursion.If the angular velocity of gyro to measure is ω ^b, then Attitude kinematic function can be with quaternion representation:

$\stackrel{\·}{q}=\frac{1}{2}\mathrm{Eq}\left(q\right){\mathrm{\ω}}^b---\left(1\right)$

Wherein

$\mathrm{Eq}\left(q\right)=\left[\begin{array}{ccc}{q}_4& -q_3& q_2\\ q_3& q_4& -q_1\\ -q_2& q_1& q_4\\ -q_1& -q_2& -q_3\end{array}\right]---\left(2\right)$

The renewal that integration can complete hypercomplex number is carried out to (1) formula, can attitude matrix be calculated further:

$C_{\mathrm{bi}}=\left[\begin{array}{ccc}{q}_1^2-q_2^2-q_3^2+q_4^2& 2(q_1{q}_2+q_3+q_4)& 2(q_1{q}_3-q_2{q}_4)\\ 2(q_1{q}_2-q_3{q}_4)& -q_1^2+q_2^2-q_3^2+q_4^2& 2(q_2{q}_3+q_1{q}_4)\\ 2(q_1{q}_3+q_2{q}_4)& 2(q_2{q}_3-q_1{q}_4)& -q_1^2-q_2^2+q_3^2+q_4^2\end{array}\right]---\left(3\right)$

(3) position and speed recursion, the speed of the speed increment accumulation calculating current time lander namely utilizing accelerometer measures to obtain, and integration obtains the position of lander under inertial coordinates system

The inertial navigation fundamental equation described under landing inertial coordinates system is

${\stackrel{\·\·}{r}}^I=g^I+f^I---\left(4\right)$

Wherein, g ⁱfor moon's gravity acceleration, when ignoring moon aspheric Gravitational perturbation, have

$g^I={\mathrm{\μ}}_m\frac{r^I}{r^3}---\left(5\right)$

Wherein, μ _mfor lunar gravitation constant.F ⁱbe the non-gravitational acceleration (specific force) represented under inertial system, it can obtain according to accelerometer measures.But because accelerometer measures represents under body series, i.e. f ^b, therefore need to change according to attitude matrix.

$f^I=C_{\mathrm{ib}}{f}^b=C_{\mathrm{bi}}^T{f}^b---\left(6\right)$

Due to therefore the velocity v that two repeated integrals just can obtain lander is respectively carried out to (4) formula ⁱwith position vector r ⁱ.

Behind acquisition position, the height of relative lunar surface can also be calculated

h=||r ^I||-r _M(7)

R _mfor r ⁱcorresponding reference lunar surface height.

Second step, carries out correction (independent navigation filtering) of finding range and test the speed

Utilization is tested the speed, range radar can revise inertial navigation recursion position and speed.Range finding sensor provides line-of-sight distance, therefore utilizes attitude information and surveys sensor beam direction (direction of visual lines), line-of-sight distance can be converted into the height of the relative lunar surface of lander.What the sensor that tests the speed obtained is the component of speed at beam direction of the relative lunar surface of lander, therefore utilizes attitude information and the sensor that tests the speed is installed, and compensates according to detector position geostrophic ground velocity of landing on the moon, and just can obtain the velocity inertial of lander.Obtain compared with the position of information and inertial reference calculation that range finding and the sensor that tests the speed are obtained, velocity information and newly cease, just can be revised the position of inertial navigation, velocity error by Kalman filtering.

When reality uses, owing to testing the speed, finding range sensor Measuring Time inconsistent, to be subject on star software again and to calculate quantitative limitation, the EKF in navigation algorithm is by matching and optimize filter correction coefficient and realize, and tests the speed, finds range to revise and separately carry out.

(1) range finding is revised

According to the height that the height of the relative lunar surface reference surface of lander measuring the moment is converted to range finding sensor data, computation and measurement residual error; Carry out position correction according to optimization correction factor, and utilize the current gravitational acceleration of position calculation revised.Specific algorithm is:

A. navigation data extrapolation

Because the time of ranging information is often asynchronous with the time of inertial reference calculation, therefore, need to utilize navigation data extrapolation to carry out synchronizing time point.If time corresponding to ranging information is t _u, the time that first two steps inertial navigation the most close with it calculates is t _n-2and t _n-1, corresponding position calculation value is with then utilize the t that inertial navigation is predicted _umoment lander position is

$r_u^I=r_{n-1}^I+\frac{r_{n-1}^I-r_{n-2}^I}{t_{n-1}-t_{n-2}}(t_u-t_{n-1})---\left(\text{8}\right)$

B. error calculation

The positional information calculated by inertial navigation can obtain the height of relative lunar surface

$h_{u,\mathrm{ins}}=\left|\right|r_u^I\left|\right|-r_M---\left(9\right)$

R _mfor corresponding reference lunar surface height.

If range finding sensor wave beam is oriented in the installation of body series then utilize attitude information it can be forwarded in landing inertial coordinates system

$u_R^I=C_{\mathrm{ib}}{u}_R^b---\left(10\right)$

If find range, sensor is measured as ρ _u, then can calculate and be highly

$h_{u,\mathrm{radar}}={\mathrm{\&rho;}}_u\frac{<u_R^I\&CenterDot;-r_u^I>}{\left|\right|r_u^I\left|\right|}---\left(11\right)$

Thus, the error that can obtain high computational is

δh _u=h _u,radar-h _u,ins(12)

C. filter correction

Get altitude correction factor w _hincrease gradually with highly declining, namely

$w_h=w_h^{*}(1-h/h_h^{*})---\left(13\right)$

with for given parameters, adjustment can be carried out according to different task features general value between 0 ~ 1, therefore $0\&le;w_h\&le;w_h^{*}.$

Then height correction is carried out to inertial navigation

r ^I=r ^I+w _h·δh _u·r ^I/||r ^I||(14)

(2) to test the speed correction

I. utilize elevation information to the algorithm of gravity direction speed correction

Utilize twice not in the same time distance measurement beam measure the height of the oblique distance information clearing lander relative lunar surface obtained respectively, and difference obtains the speed component of the gravity direction of lander, and computing velocity measures residual error, and carries out speed correction according to optimization correction factor.Specific algorithm is:

A. navigation data extrapolation

The same with during ranging data process, need according to the Data Extrapolation of first two steps inertial navigation test the speed sensor measure moment t _uspeed.If t _n-2and t _n-1moment speed calculated value is with then

$v_u^I=v_{n-1}^I+\frac{v_{n-1}^I-v_{n-2}^I}{t_{n-1}-t_{n-2}}(t_u-t_{n-1})---\left(15\right)$

B. error calculation

The speed that inertial navigation calculates is relative inertness system speed, is then converted into the speed of relative lunar surface.

$v_{g,\mathrm{ins}}^I=v_u^I-{\mathrm{\&omega;}}_m^I\&times;r_u^I---\left(16\right)$

Its component at gravity direction is

$v_{h,\mathrm{ins}}\left(t_k\right)=\frac{<r^I\&CenterDot;v_{g,\mathrm{ins}}^I>}{\left|\right|r^I\left|\right|}---\left(17\right)$

If t _k-1moment and t _kthe oblique distance information that moment distance measurement beam records respectively is ρ (t _k-1) and ρ (t _k), then in conjunction with the sensing of distance measurement beam and the attitude in corresponding moment and positional information, calculate height.With t _kmoment is example, if the installation of distance measurement beam under body series is oriented to p ^b, then it is at the sensing p of inertial system ⁱ(t _k) be

p ^I(t _k)=C _ib(t _k)p ^b(18)

So can calculate height according to range finding sensor distance measurement information, be designated as

$h_{\mathrm{radar}}\left(t_k\right)=\frac{<p^I\left(t_k\right),r^I\left(t_k\right)>}{\left|\right|r^I\left(t_k\right)}\mathrm{\&rho;}\left(t_k\right)---\left(19\right)$

Equally can according to t _k-1the ranging information computed altitude h in moment _radar(t _k-1).So difference acquisition lander in the speed of gravity direction is

$v_{h,\mathrm{radar}}\left(t_k\right)=\frac{h_{\mathrm{\&rho;}}\left(t_k\right)-h_{\mathrm{\&rho;}}\left(t_{k-1}\right)}{t_k-t_{k-1}}---\left(20\right)$

Thus, the difference that can obtain the gravity direction speed calculated by two kinds of means of different is

δv _h=v _h,radar-v _h,ins(21)

C. filter correction

Get speed correction factor w _vreduce with speed and increase gradually, namely

$w_v=w_v^{\&CenterDot;}(1-|\left|v^I\right||/v_v^{*})---\left(19\right)$

with for given parameters, can adjust according to different task features, generally value between 0 ~ 1, therefore

Then the speed of inertial navigation short transverse is revised

v ^I=v ^I+w _v·δv _h·r ^I/||r ^I||(20)

II. the algorithm of sensor to the correction of beam direction speed that test the speed is utilized

The speed component of the speed of the relative lunar surface of the lander utilizing knotmeter to measure and the beam direction that tests the speed of inertial reference calculation, computing velocity measures residual error, and carries out speed correction according to optimization correction factor.Specific algorithm is:

A. navigation data extrapolation

Utilize t _n-2and t _n-1the speed calculated value of moment inertial navigation is with test the speed according to the extrapolation of (15) formula the lander velocity inertial in sensor measurement data corresponding moment.

B. error calculation

First needing is the speed of relative lunar surface by the rate conversion of inertial reference calculation according to (16) formula, and then projected to beam direction, specific algorithm is as follows.

If velocity radar wave beam being oriented at body series then utilize attitude information it can be forwarded in landing inertial coordinates system

$u_v^I=C_{\mathrm{ib}}{u}_v^b---\left(21\right)$

Then the component of the speed of inertial reference calculation in radar beam direction is

$v_{u,\mathrm{ins}}=<v_{g,\mathrm{ins}}^I\&CenterDot;u_v^I>---\left(22\right)$

The velocity survey of the actual acquisition of velocity radar is v _{u, radar}, therefore, both differences are

δv _u=v _u,radar-v _u,ins(23)

C. filter correction

Get speed correction factor according to (19) formula to increase gradually with speed reduction, speed correction is carried out to inertial navigation

$v^I=v^I+w_v\&CenterDot;\mathrm{\&delta;}{v}_u\&CenterDot;u_v^I---\left(20\right)$

(3) correction strategy

Above-mentioned range finding and the correction algorithm that tests the speed are all carry out for single beam, as long as namely any one wave beam of current time effectively just may be used for carrying out navigation correction.In makeover process, range finding revise mainly for be altitude channel site error, test the speed revise mainly for be the velocity error of beam direction.For soft landing task, elevation information is the most important, the safety that direct relation lands; And the accuracy requirement of horizontal position information is not high especially, the position correction therefore adopting the present invention to carry out is suitable.For correction of testing the speed, horizontal and vertical speed is all quite important, and it is directly connected to the inclined extent of landing speed, affects task success or failure.Although speed correction of the present invention is mainly along beam direction, as long as there is the wave beam of three different directions to measure, the correction of three-dimensional velocity just can be completed.Therefore, this algorithm is also simple and effective.

Simulation analysis.Simulation analysis is carried out for the goddess in the moon's No. three Lunar satellite orbit.Simulated conditions is as follows:

● preliminary orbit: the perilune position of 100km × 15km track and speed is highly 15km, and horizontal velocity is that 1692m/s(perilune and apocynthion height all exist 1km stochastic error).

● initial attitude: the attitude angle of landing inertial system is the angle of pitch 110 °, driftage and roll angle 0 ° relatively.

● initial navigation error: orbit error comprises site error 2km(R direction 0.45km, T direction 1.8km, N direction 0.6km), velocity error 2m/s(R direction 1.65m/s, T direction 1m/s, N direction 0.6m/s), attitude error 0.03 °.

● tellurometer survey is tested the speed sensor five groups of antennas totally, wherein finds range two groups, tests the speed three groups, be arranged on lander-Z face.Specifically be oriented to: the expression of unit vector in lander body series of three groups of each beam positions of rate antenna (containing receiving antenna and emitting antenna) is respectively R1=[cos133 °, cos90 °, cos137 °], R2=[cos154.3265 °, cos75.3330 °, cos110.5823 °], R3=[cos154.3265 °, cos104.6670 °, cos110.5823 °]; Two groups of telemetry antenna (containing receiving antenna and emitting antenna) beam position R4 and R5 respectively along lander body series-Z axis and-X-axis.

● correction factor is taken as $w_h^{*}=0.5,h_h^{*}=8000m,w_v^{*}=0.4,v_v^{*}=300m/s.$

Navigation strategy is: highly introduce ranging information apart from lunar surface 8km, highly introduces the information that tests the speed apart from lunar surface 3.5km; Apart from lunar surface more than height 8km and below 30m, only utilize inertial navigation.

Fig. 3 give landing mission reality with the change in location curve under the landing inertial system that provides of navigating, the error of navigation is as shown in Figure 4.As seen from Figure 4 before outside range finding and the information that tests the speed are introduced, inertial navigation location compute is dispersed; Contribute in the introducing of below 8km height (about after 360s) ranging information the site error revising x direction (short transverse), but capability for correcting is not had for the site error of y, z direction (horizontal direction); Introduce in below 3.5km (about after the 450s) information that tests the speed, it can reduce the divergence speed of horizontal level to a certain extent by the speed revising y, z direction, improve the precision of navigation.

Fig. 5 give power dropping process reality with the change curve of lander velocity inertial provided that navigates, the error of navigation is as shown in Figure 6.Data from Fig. 6, before Microwave Velocity data introduce navigational system, the velocity error of inertial reference calculation increases gradually; After Microwave Velocity data are introduced, velocity error is revised, and ensure that the precision of navigation.

Non-elaborated part of the present invention belongs to techniques well known.

Claims (2)

1. multi-beam test the speed range finding revise robust navigation method, it is characterized in that performing step is as follows:

(1) inertial reference calculation

Utilize the inertial attitude of the angular speed calculation current time lander of gyro to measure in the previous sampling period, the specific force then utilizing accelerometer measures to obtain calculates the speed of current time lander, and integration obtains the position of lander under inertial coordinates system;

(2) range finding is revised

If receive the valid data of range finding sensor, and flying height meets the introducing scope that range finding is revised, then these data are solved altitude information by installation matrix and the conversion of lander attitude matrix; Utilize the inertial reference calculation in the first two cycle simultaneously, calculate the lander height on time point corresponding to range finding sensor data, the height that inertial navigation and range finding sensor are resolved respectively is done difference, and according to the site error in altitude correction factor corrected altitude direction; This altitude correction factor is with the change of lander highly linear, and lander height is higher, and described altitude correction factor is less, and highly more low clearance correction factor is larger;

(3) speed correction

The signal source of speed correction is divided into two kinds, and one is distance measuring signal, and another kind is tachometer signal, when distance measuring signal is effective, and flying height meet test the speed revise introducing scope time, the height solved by double distance measuring signal carries out difference, obtain the speed of gravity direction, utilize the lander velocity inertial in the speed calculation distance measuring signal in the first two cycle in inertial reference calculation corresponding moment simultaneously, and be converted to the speed of relative lunar surface, the speed that the speed obtain distance measuring signal and inertial navigation obtain does difference, and according to the speed of speed correction factor correction gravity direction, this speed correction factor changes with lander speed linearity, lander speed is faster, speed correction factor is less, speed is less, speed correction factor is larger, when tachometer signal is effective, utilize the lander velocity inertial in the speed calculation tachometer signal in the first two cycle in inertial reference calculation corresponding moment, and be converted to the speed of relative lunar surface, the speed obtain tachometer signal and the speed that obtains of inertial navigation do difference, and according to the speed of speed correction factor correction beam direction, this speed correction factor changes with lander speed linearity, lander speed is faster, speed correction factor is less, and speed is less, and speed correction factor is larger,

(4) each control cycle is repeated to the step of (1) ~ (3), complete navigation calculation.

2. multi-beam according to claim 1 test the speed range finding revise robust navigation method, it is characterized in that: described altitude correction factor and speed correction factor are 0-1.