CN101520328A - Method for autonomous navigation using geomagnetic field line map - Google Patents

Abstract

The invention discloses a method for autonomous navigation using a geomagnetic field line map. Firstly, a plurality of characteristic quantities of the geomagnetic filed on a path of an aerial vehicle are measured continuously according to a preset frequency, and measurement data are used to build a matched line map of the corresponding characteristic quantities in a sliding window mode with fixed-point number; and a matched line map of the plurality of characteristic quantities is matched and compared with a reference map by using an algorithm for fining global optimum according to a matching similarity rule and a matching result fusion rule to acquire the position information of the aerial vehicle. The technology makes full use of the characteristics of the plurality of characteristic quantities of the geomagnetic field to calculate the accurate position of the aerial vehicle, avoids navigation accumulated error under a condition of long flight period, is particularly suitable for navigation in environments without typical geomorphic features such as ocean and plain, can meet requirements of future cruise missiles, unmanned aerial vehicles, submarines and the like for passive, all-sky time, all-weather and all-terrain navigation, and also can be used in civil area.

Description

Method for autonomous navigation using geomagnetic field line map

Technical field

The present invention relates to a kind of air navigation aid, particularly method for autonomous navigation using geomagnetic field line map.

Background technology

The navigator fix technology is in fundamental position in development of modern scientific technology, permeate in various dual-use fields, demonstrates more and more important effect.Yet ripe at present navigation locating method remains in very big deficiency under conditions such as long-term navigation task, no typical morphologic characteristics environment, needs replenishing of other navigate mode.Submarine for example is if only adopt INS technology, i.e. inertial navigation technology, though can independent navigation and precision is higher in short-term, there is accumulated error, and is difficult to by other prior art at deep water, submarine has to regularly emerge revise, and has reduced its disguise and security.Cruise missile for another example, though can receive radio navigation signal, the wartime signal is subjected to enemy's shielding and interference easily, navigation base station itself is also under attack easily; In addition, when flights such as over strait, desert and forest, because the gray scale and the texture of landform are basic identical, the terrain match method also can't realize.

The ultimate principle of earth magnetism coupling navigation is exactly to measure the line chart that the terrestrial magnetic field characteristic quantity of carrier locality forms with flight path, and by doing matching ratio with the earth magnetism pattern of storage in advance, thereby realize accurate navigator fix.Magnetic map is exactly parameters such as base area magnetic field intensity or direction and a series of isoline of drawing.These patterns are followed the rule that earth ball model forms generally, but also obviously are subjected to the influence of the surface distributed and the local geographic entity of the earth's crust, the final feature pattern that enriches that forms.Simultaneously, the terrestrial magnetic field also has more intensity and angle character amount, and the operability of coupling is stronger than landform and image.

Document " based on the earth magnetism matching process [J] of isoline constraint. space science journal; 2007; 27 (6): 505～511. " a kind of earth magnetism coupling air navigation aid disclosed, this method has been used for reference the terrain contour matching auxiliary navigation method, matching process is divided into thick matching stage and smart matching stage, utilizes the terrestrial magnetic field isoline to retrain to assist INS to position.But in actual applications, these methods have also continued the deficiency of terrain match auxiliary navigation method: can not break away from INS and work alone, exist under the big initial alignment sum of errors course error situation at INS, the bearing accuracy that is obtained by the earth magnetism matching algorithm reduces greatly; Geomagnetic field information is utilized insufficient, only used individually that the magnetic field amplitude characteristic quantity mates, thereby be absorbed in the mistake coupling easily; Just can carry out relevant matches after a series of magnetic survey data in accumulation and locate, thereby cause the real-time of navigating relatively poor etc.

Summary of the invention

Geomagnetic field information is utilized insufficient in order to overcome traditional earth magnetism coupling air navigation aid, the navigation real-time is strong and need to rely on weakness such as INS co-operation, the invention provides a kind of method for autonomous navigation using geomagnetic field line map, according to a plurality of terrestrial magnetic field characteristic quantity on the setpoint frequency continuous coverage aircraft flight path, and according to the coupling line chart of measurement data structure character pair amount, according to coupling similarity criterion and matching result fusion criterion, adopt global optimum's searching algorithm that the coupling line chart of a plurality of characteristic quantities is done matching ratio with corresponding reference map, obtain the positional information of aircraft.

The technical solution adopted for the present invention to solve the technical problems: a kind of method for autonomous navigation using geomagnetic field line map is characterized in comprising the steps:

(1) select a plurality of earth magnetism characteristic quantities as the matching characteristic amount;

(2) at sampling instant t, measurements and calculations terrestrial magnetic field matching characteristic amount;

(3) adopt the moving window mode to construct the coupling line chart of these earth magnetism characteristic quantities;

(4) adopt global optimum's match search algorithm that each coupling line chart is done matching ratio with corresponding reference map respectively, this step need be calculated the similarity measurement value and the normalization similarity measurement value of coupling line chart, and the normalization metric of different characteristic amount need be merged;

(5) the output aircraft is in the locating information of sampling instant t;

(6) at next sampling instant t=t+1, repeated execution of steps 2～5, thus realize location continuously.

First-selected geomagnetic fieldvector intensity is as the matching characteristic amount, has only local magnetic vector intensity fluctuation hour, just adds terrestrial magnetic field angle character amount jointly as the matching characteristic amount; Getting observation station is initial point, gets geographic north to, geographical east orientation be respectively x, y vertically downward and the rectangular coordinate system of the forward of z axle each earth magnetism characteristic quantity of giving a definition, and is calculated the value of all the other characteristic quantities of terrestrial magnetic field by formula (1)

$H=\sqrt{X^2+Y^2}$

tan?D＝Y/X

sin?D＝Y/H

(1)

tan?I＝Z/H

sin?I＝Z/F

$F=\sqrt{H^2+Z^2}=\sqrt{X^2+Y^2+Z^2}$

In the formula, F is that terrestrial magnetic field total intensity, Y are that east component, X are that north component, Z are that vertical component, H are that horizontal component, D are that magnetic declination, I are magnetic dips;

Adopt the moving window mode of fixedly counting to construct the coupling line chart,, suppose to mate to count to be the N point, at every turn at new measured value u promptly to each characteristic quantity _iAfter the arrival, with u _iBe added in the slotted line graphic sequence of this characteristic quantity, and with u _I-NFrom this sequence, reject, stay u _I-N+1To u _iBetween altogether N measured value as newly mating line chart; At each sampling instant t _iFinish after the magnetic survey task, at once new coupling line chart and reference map are mated, the pairing navigation position of this sampling instant of output matched position sequence.

The coupling similarity criterion of characteristic quantity is MAD criterion, MSD criterion, NPROD criterion or Hausdroff distance criterion, in actual applications, selects a kind of criterion to carry out the similarity coupling according to navigation request; If u and v are two sequences doing coupling, counting of sequence is N, D (u, v) or H (u v) is the similarity measurement value, and then each criterion is defined as

The a.MAD criterion $D(u,v)=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}|u_i-v_i|---\left(2\right)$

The b.MSD criterion $D(u,v)=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(u_i-v_i)}^2---\left(3\right)$

The c.NPROD criterion $D(u,v)=\underset{i=1}{\overset{i=N}{\mathrm{\Σ}}}{u}_i{v}_i/\left(\sqrt{\underset{i=1}{\overset{i=N}{\mathrm{\Σ}}}{u_i}^2}\sqrt{\underset{i=1}{\overset{i=N}{\mathrm{\Σ}}}{v_i}^2}\right)---\left(4\right)$

D.Hausdroff distance criterion H (u, v)=max (h (u, v), h (v, u)) (5)

5, method for autonomous navigation using geomagnetic field line map according to claim 1 is characterized in that: method for normalizing is

$f_i^{\′}=\frac{f_i-f_{\min }}{f_{\max }-f_{\min }}---\left(6\right)$

In the formula, f and

Be respectively the similarity measurement value of normalization front and back, f _MaxAnd f _MinBe respectively the preceding minimum and maximum similarity measurement value of normalization.

Blending algorithm is weighting maximum/minimum value principle and normalization and principle, in actual applications, comes a kind of fusion criterion of choose reasonable according to navigation request; Each blending algorithm is defined as

A. weighting maximum/minimum value principle

B. normalization and principle

$F_I=f_{I,1}^{\′}+f_{I,2}^{\′}+...+f_{I,N}^{\′}---\left(8\right)$

In the formula, N is the number of matching characteristic amount, F _IFinal fusion similarity measurement,

Be the normalization similarity measurement value of matched position at feature j.

Global optimum's match search algorithm adopts swarm intelligence algorithms such as genetic algorithm, simulated annealing and hybrid algorithm thereof.

The invention has the beneficial effects as follows: owing to only rely on the magnetic field of the earth feature to carry out independent navigation, when not being subjected to day and the restriction of weather, can under the condition of no typical morphologic characteristicss such as Plain and sea, not lose efficacy yet, can be dual-use aircraft provides the accurate positioning service of long voyage schedule, passive, round-the-clock, round-the-clock and full region, is particularly suitable for being applied to the special occasions of contour or approximate contour motion such as cruise missile, unmanned plane, submarine and civil aviaton.Utilizing a plurality of characteristic quantity common match of terrestrial magnetic field, also is the multiple chart coupling, and realization makes full use of geomagnetic field information, thereby reduces the mistake coupling effectively; Take after the magnetic survey task is finished in each sampling instant the mode of coupling at once, the real-time of therefore navigating is better; Can break away from INS independently uses.

Below in conjunction with drawings and Examples the present invention is elaborated.

Description of drawings

Fig. 1 is the definition synoptic diagram of each characteristic quantity of terrestrial magnetic field of method for autonomous navigation using geomagnetic field line map employing of the present invention.

Fig. 2 is the process flow diagram of method for autonomous navigation using geomagnetic field line map of the present invention.

Embodiment

With reference to Fig. 1～2.The present invention adopts a plurality of earth magnetism characteristic quantities as the matching characteristic amount.The system of selection of matching characteristic amount is: first-selected ground magnetic vector intensity is as the matching characteristic amount, has only local magnetic vector intensity fluctuation hour, just adds earth magnetism angle character amount jointly as the matching characteristic amount.Definition such as Fig. 1 of terrestrial magnetic field total intensity F, east component Y, north component X, vertical component Z, horizontal component H, magnetic declination D, magnetic dip I, to get observation station be initial point to used rectangular coordinate system among the figure, gets geographic north to, geographical east orientation be respectively x, y vertically downward and the forward of z axle.Mathematical relation between each characteristic quantity of terrestrial magnetic field is

$H=\sqrt{X^2+Y^2}$

tan?D＝Y/X

sin?D＝Y/H

(9)

tan?I＝Z/H

sin?I＝Z/F

$F=\sqrt{H^2+Z^2}=\sqrt{X^2+Y^2+Z^2}$

As long as recorded terrestrial magnetic field east orientation, north orientation and vertical component intensity, just can calculate the value of all the other characteristic quantities of terrestrial magnetic field by formula (9).

At sampling instant t, measurements and calculations terrestrial magnetic field matching characteristic amount.

Adopt the moving window mode of fixedly counting to construct the coupling line chart,, suppose to mate to count to be the N point, at every turn at new measured value u promptly to each characteristic quantity _iAfter the arrival, with u _iBe added in the slotted line graphic sequence of this characteristic quantity, and with u _I-NFrom this sequence, reject, stay u _I-N+1To u _iBetween altogether N measured value as newly mating line chart.After the magnetic survey task is finished in each sampling instant, at once new coupling line chart and reference map are mated the navigation position of this sampling instant correspondence in the output matched position sequence.This mode does not need need wait until that as the navigation of traditional earth magnetism coupling having accumulated one section just can carry out relevant matches after measuring sequence, therefore can realize continuous location, thereby improve the real-time of coupling.

Global optimum's match search algorithm can adopt swarm intelligence algorithms such as genetic algorithm, simulated annealing and hybrid algorithm thereof to come the search matched position, and these searching algorithms can converge on globally optimal solution preferably.Swarm intelligence algorithms such as genetic algorithm, simulated annealing and hybrid algorithm thereof are personage's method in common knowledge in the industry, do not repeat them here.

The coupling similarity criterion of characteristic quantity is chosen as: MAD criterion, MSD criterion, NPROD criterion and Hausdroff distance criterion.If u and v are two sequences doing coupling, counting of sequence is N, D (u, v) or H (u v) is the similarity measurement value, and then each criterion is defined as follows:

The a.MAD criterion $D(u,v)=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}|u_i-v_i|---\left(10\right)$

The b.MSD criterion $D(u,v)=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(u_i-v_i)}^2---\left(11\right)$

The c.NPROD criterion $D(u,v)=\underset{i=1}{\overset{i=N}{\mathrm{\Σ}}}{u}_i{v}_i/\left(\sqrt{\underset{i=1}{\overset{i=N}{\mathrm{\Σ}}}{u_i}^2}\sqrt{\underset{i=1}{\overset{i=N}{\mathrm{\Σ}}}{v_i}^2}\right)---\left(12\right)$

D.Hausdroff distance criterion H (u, v)=max (h (u, v), h (v, u)) (13)

In actual applications, come a kind of criterion of choose reasonable to carry out the similarity coupling according to navigation request.

Normalized purpose is the order of magnitude of the similarity measurement value of unified different characteristic amount, and the normalization formula is as follows:

$f_i^{\′}=\frac{f_i-f_{\min }}{f_{\max }-f_{\min }}---\left(14\right)$

Wherein, f and

Be respectively the similarity measurement value of normalization front and back, f _MaxAnd f _MinBe respectively the preceding minimum and maximum similarity measurement value of normalization.

Adoptable blending algorithm is weighting maximum/minimum value principle or normalization and principle.Each blending algorithm is defined as follows:

A. weighting maximum/minimum value principle

B. normalization and principle

$F_I=f_{I,1}^{\′}+f_{I,2}^{\′}+...+f_{I,N}^{\′}---\left(16\right)$

In the formula, N is the number of matching characteristic amount, F _IBe final fusion similarity measurement,

Be the normalization similarity measurement value of matched position at feature j; Because the measuring accuracy of every kind of feature is variant, therefore introduce weight w ∈ [0,1], w is big more, and then the signal to noise ratio (S/N ratio) of character pair is high more, and the confidence level of this feature is big more.In actual applications, come a kind of fusion criterion of choose reasonable according to navigation request.

Be example to adopt genetic algorithm as global optimum's match search algorithm below, the present invention is described.

1. the design of genetic manipulation algorithm

1. coded system: the coding form that adopts real-valued " initial position+initial velocity+measuring point acceleration ".For example individual

Wherein 1000 and 999 represent the initial position of aircraft on the x-y coordinate system, 101.4 and 98.1 represent the initial velocity on the x-y axle, the acceleration on the x-y of 8.5 and 9.4 first measurement points of representative arrival, and remaining gene is for arriving the acceleration of other measurement points.

2. initial population produces: the matching result by kinematic constraint and last time produces the coupling flight path.When mating first, produce initial population at random by acceleration constraint, the constraint of initial velocity error and position error constraint just, first position in Pi Pei the initial population and initial velocity add that by the positioning result of last time coupling the random quantity that satisfies the error constraint produces backward, acceleration is produced by the acceleration constraint, and the acceleration constraint is determined according to the motion priori of aircraft.

3. fitness calculates: the position (p that is extrapolated each measuring point earlier by initial position, initial velocity and the measuring point acceleration of individuality by formula (9) _x, p _y), select the MSD criterion to calculate the similarity measurement value of each individual characteristic quantity by formula (10) then, by formula (14) normalization similarity measurement, select weighting minimum value principle to merge the normalization similarity measurement of each characteristic quantity at last, as final fitness value by formula (15).

$\left\{\begin{array}{c}{v}_i=v_{i-1}+a_{i}t\\ p_i=p_{i-1}+v_{i-1}t+\frac{a_i{t}^2}{2}\end{array}\right.---\left(17\right)$

In the formula, v, a and t are respectively velocity series, acceleration sequence and the sampling time interval of aircraft.

$D(u,v)=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(u_i-v_i)}^2---\left(18\right)$

In the formula, v is the measurement sequence-coupling line chart of certain characteristic quantity, u is for hereditary individuality being converted to the benchmark geomagnetic data in this characteristic quantity reference map behind the corresponding position sequence P, if P not on grid intersection, is then calculated the earth magnetism reference data at this some place by bilinear interpolation.

4. select operator to adopt roulette back-and-forth method and elite to protect strategy, the reorganization operator adopts linear recombination method, and mutation operator adopts evenly variation.

5. shut down criterion: adopt the restriction of threshold value and iterations.If individual MSD value is less than threshold value, perhaps iteration has reached maximum iteration time, and then termination of iterations is got the individuality of optimal adaptation degree correspondence and separated for coupling.

2. based on the performing step of the matching process of genetic algorithm

Based on above-mentioned genetic manipulation design, the specific implementation step of matching process is as follows:

(1) gets north orientation, east orientation and vertical direction absolute force as the matching characteristic amount; Reference map adopts the geomagnetic data of matrix form record grid intersection position with the geographical position gridization of reality, and the grid specification is 100 * 100m, and is stored in the computing machine of aircraft; Aircraft is pressed close to the average velocity flight of sea level with 150m/s.

(2) setting the legacy parameter is: 150 of population numbers, recombination fraction 0.95, aberration rate 0.10, maximum iteration time 70 times.The kinematic constraint of aircraft is: acceleration range-35～35m/s ², first position error range-100～100m,, initial velocity error range-20～20m/s.

(3) when each coupling, the coupling of each coupling line chart is counted and is taken as 10 points.After aircraft is started working, three axis magnetometer is according to north orientation, east orientation and the vertical direction absolute force of 10 measuring points on the sample frequency continuous acquisition flight path of 1Hz, the data of same characteristic quantity are lined up a sequence according to the sequencing of gathering, 3 sequences have just constituted 3 coupling line charts, and coupling is just utilized these 3 coupling line charts first; After this, from obtaining the data of the 11st measuring point, just by fixing the coupling line chart that 10 moving window modes of counting are constructed each matching characteristic amount.

(4) after sampling instant t finishes magnetic survey, use genetic algorithm at once the coupling line chart is carried out match search, finishing once, the substep of the genetic manipulation of coupling is:

1. produce 100 initial individualities according to above-mentioned coded system and initial population generation way, and calculate each individual fitness value according to MSD criterion and weighting minimum value principle;

2. population is cooked selection, intersection and mutation operation;

3. calculate new individual fitness value according to MSD criterion and weighting minimum value principle again;

4. judge to satisfy shut down criterion not,, then export the best individuality of fitness in the population, optimum individual is converted to the position sequence of correspondence by formula (17) as optimum solution if satisfy; If do not satisfy, then the iteration execution in step 2.～4..

(5) the pairing position of this sampling instant in the outgoing position sequence.

(6) at next sampling instant t=t+1, measure geomagnetic fieldvector, after the magnetic survey task is finished, mate according to 3～5 pairs of magnetic survey data of above-mentioned steps at once, thereby realize location continuously.

Claims (7)

1, a kind of method for autonomous navigation using geomagnetic field line map is characterized in that comprising the steps:

(1) select a plurality of earth magnetism characteristic quantities as the matching characteristic amount;

(2) at sampling instant t, measurements and calculations terrestrial magnetic field matching characteristic amount;

(3) adopt the moving window mode to construct the coupling line chart of earth magnetism matching characteristic amount;

(4) adopt global optimum's match search algorithm that each coupling line chart is done matching ratio with corresponding reference map respectively, this step need be calculated the similarity measurement value and the normalization similarity measurement value of coupling line chart, and the normalization metric of different characteristic amount need be merged;

(5) the output aircraft is in the locating information of sampling instant t;

(6) at next sampling instant t=t+1, repeated execution of steps 2～5, thus realize location continuously.

2, method for autonomous navigation using geomagnetic field line map according to claim 1, it is characterized in that: first-selected geomagnetic fieldvector intensity is as the matching characteristic amount, have only local magnetic vector intensity fluctuation hour, just add terrestrial magnetic field angle character amount jointly as the matching characteristic amount; Getting observation station is initial point, gets geographic north to, geographical east orientation be respectively x, y vertically downward and the rectangular coordinate system of the forward of z axle each earth magnetism characteristic quantity of giving a definition, and is calculated the value of all the other characteristic quantities of terrestrial magnetic field by formula (1)

$H=\sqrt{X^2+Y^2}$

tan?D＝Y/X

sin?D＝Y/H (1)

tan?I＝Z/H

sin?I＝Z/F

$F=\sqrt{H^2+Z^2}=\sqrt{X^2+Y^2+Z^2}$

In the formula, F is that terrestrial magnetic field total intensity, Y are that east component, X are that north component, Z are that vertical component, H are that horizontal component, D are that magnetic declination, I are magnetic dips;

3, method for autonomous navigation using geomagnetic field line map according to claim 1 is characterized in that: adopt the moving window mode of fixedly counting to construct the coupling line chart, promptly to each characteristic quantity, suppose to mate to count to be the N point, at new measured value u at every turn _iAfter the arrival, with u _iBe added in the slotted line graphic sequence of this characteristic quantity, and with u _I-NFrom this sequence, reject, stay u _I-N+1To u _iBetween altogether N measured value as newly mating line chart; At each sampling instant t _iFinish after the magnetic survey task, at once new coupling line chart and reference map are mated, the pairing navigation position of this sampling instant of output matched position sequence.

4, method for autonomous navigation using geomagnetic field line map according to claim 1, it is characterized in that: the coupling similarity criterion of characteristic quantity is MAD criterion, MSD criterion, NPROD criterion or Hausdroff distance criterion, in actual applications, select a kind of criterion to carry out the similarity coupling according to navigation request; If u and v are two sequences doing coupling, counting of sequence is N, D (u, v) or H (u v) is the similarity measurement value, and then each criterion is defined as

The a.MAD criterion $D(u,v)=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}|u_i-v_i|---\left(2\right)$

The b.MSD criterion $D(u,v)=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(u_i-v_i)}^2---\left(3\right)$

The c.NPROD criterion $D(u,v)=\underset{i=1}{\overset{i=N}{\mathrm{\Σ}}}{u}_i{v}_i/\left(\sqrt{\underset{i=1}{\overset{i=N}{\mathrm{\Σ}}}{u_i}^2}\sqrt{\underset{i=1}{\overset{i=N}{\mathrm{\Σ}}}{v_i}^2}\right)---\left(4\right)$

D.Hausdroff distance criterion H (u, v)=max (h (u, v), h (v, u)) (5)

5, method for autonomous navigation using geomagnetic field line map according to claim 1 is characterized in that: method for normalizing is

$f_i^{\′}=\frac{f_i-f_{\min }}{f_{\max }-f_{\min }}---\left(6\right)$

In the formula, f and $f_i^{\′}\∈\left[\mathrm{0,1}\right]$ Be respectively the similarity measurement value of normalization front and back, f _MaxAnd f _MinBe respectively the preceding minimum and maximum similarity measurement value of normalization.

6, method for autonomous navigation using geomagnetic field line map according to claim 1 is characterized in that: blending algorithm is weighting maximum/minimum value principle and normalization and principle, in actual applications, comes a kind of fusion criterion of choose reasonable according to navigation request; Each blending algorithm is defined as

A. weighting maximum/minimum value principle

B. normalization and principle

$F_I=f_{I,1}^{\′}+f_{I,2}^{\′}+...+f_{I,N}^{\′}---\left(8\right)$

In the formula, N is the number of matching characteristic amount, F _IBe final fusion similarity measurement,

Be the normalization similarity measurement value of matched position at feature j.

7, method for autonomous navigation using geomagnetic field line map according to claim 1 is characterized in that: global optimum's match search algorithm adopts swarm intelligence algorithms such as genetic algorithm, simulated annealing and hybrid algorithm thereof.

Images