CN103292782B - Infrared target passive ranging method based on genetic algorithm and particle filtering - Google Patents

Abstract

The invention discloses a ranging method for medium and long-distance targets, mainly solving the shortage that the conventional infrared passive ranging method is difficultly applied to the medium and long-distance targets. The ranging method is realized by the following steps of: measuring a target pitch angle beta by utilizing an infrared detector; creating a ternary nonlinear equation set comprising a target distance R by utilizing a geometrical relationship among the target pitch angle beta, the infrared detector and the target motion; solving the nonlinear equation set by using a genetic algorithm to obtain a set of possible solutions of the nonlinear equation set; selecting all the possible solutions, selecting the first 100 possible solutions with high fitness as initial particles for particle filtering, and carrying out particle filtering on all the initial particles to obtain the target distance R. The ranging method has the characteristics of high invisibility, small measurement variable, good feasibility and high accuracy, and is suitable for distance estimation of a target with a distance of 40-80km from the infrared detector.

Description

Based on the infrared target passive ranging method of genetic algorithm and particle filter

Technical field

The invention belongs to infrared early warning technical field, relate to passive tracking system and to adjust the distance the estimation of the target range within the scope of infrared eye 40 ~ 80km, can be used for target following.

Background technology

In photoelectronic warfare, UAV system or missile-borne infrared detection system are operated in passive working method because of it, not outwardly emittance, significantly enhance stealth capabilities and penetration ability, become one of focus of research at present.But, because it does not have range observation function, the Guidance Law of various advanced person cannot be applied, reduces its guidance precision, limit its development.Therefore one of passive ranging technology gordian technique becoming electro-optical countermeasure svstem.Document " Barbaric Z P; Bondzulic B P; Mitrovic S T.Passive ranging using image intensity and contrast measurements [J] .Electronics letters; 2012; 48 (18): 1122-1123. " is under the prerequisite of known extinction coefficient, by calculating intensity and the contrast of target image, the good communication to target range can be realized, but do not possess distance measurement function.Document " Dowski E R Jr; and Cathey W T.Single-lens single-image incoherent passive-ranging systems [J] .Applied Optics; 1994; 33 (29): 6762-6773. " proposes based on wavefront coded monocular passive ranging technology, its principle is: in optical system, introduce the optical cover meeting certain condition, make the optical transfer function of imaging system form cyclical variation zero point that is a series of and target object distance dependent.Finally, by carrying out the distance of estimating target object to the spectrum analysis of sampled images, this method cannot realize adjusting the distance the distance estimations of detector 40km and above target.Document " Macdonald D J.Passive Ranging Using Infra-Red Atmospheric Attenuation [D] .Air Force Institute of Technology (AU); Master's thesis; the Mar 2010. " IRFT of high s/n ratio processing gain, to CO near 2 μm ₂spectrum absorption spectra is analyzed and researched, and distance estimations scope can be made in theory to expand to 50km, but this expansion also needs to do current optical system further to improve and could realize, and not easily realizes.

Summary of the invention

Object of the present invention is the deficiency for above-mentioned prior art, proposes a kind of infrared target passive ranging method based on genetic algorithm and particle filter, to realize having Target Distance Estimation far away under stealth capabilities.

Realizing technical scheme of the present invention is: by the measurement target angle of pitch, set up the ternary Nonlinear System of Equations comprising target range information, solved this system of equations by genetic algorithm and particle filter, obtain infrared detection system to the estimated value of target range, concrete steps are as follows:

(1) infrared eye is utilized to record target pitch angle β;

(2) according to the geometric relationship of infrared eye and target travel, the ternary Nonlinear System of Equations comprising target range is set up:

$f1=R+R_0\sin \mathrm{\β}-\sqrt{R_0^2{\sin }^2\mathrm{\β}+2h{R}_0+h^2}=0$

$f2=R^2-h^2-{\left({2R}_0\sin \frac{\mathrm{\θ}}{2}\right)}^2-4h{R}_0{\sin }^2\frac{\mathrm{\θ}}{2}=0$

f3＝-h+R*sin(β)/cos(θ)+R ₀/cos(θ)-R ₀＝0

The unknown quantity of this ternary Nonlinear System of Equations is R, θ, h, and wherein, R is the distance that target arrives infrared eye, and θ is target and the infrared eye angle relative to the earth's core, and h is the distance of target to the earth's core and the difference of earth radius, R ₀for earth radius, f1, f2, f3 represent the value of three nonlinear equations respectively;

(3) random generation 1000 target range values in interval 30 ~ 40km, are expressed as R _i, i=1 ~ 1000, form the initial population q of target range value; All the other two unknown quantity θ=0 in the above-mentioned system of equations of initialization, h=28km;

(4) by the target range value R of 1000 in initial population q _iand initialized θ, h are combined into 1000 groups of iterative value, are expressed as a _i={ R _iθ h}, respectively as the iterative initial value of quasi-Newton method, utilizes the Nonlinear System of Equations set up in quasi-Newton method solution procedure (2), obtains 1000 and target range value R _ithe solution of corresponding Nonlinear System of Equations, and gained solution is substituted in Nonlinear System of Equations, obtain each target range value R _ithe value f1 of three corresponding nonlinear equations _i, f2 _i, f3 _i, i=1 ~ 1000;

(5) the value f1 of the Nonlinear System of Equations obtained in step (4) is utilized _i, f2 _i, f3 _i, calculate 1000 target range value R in initial population q _icorresponding fitness fitness _i:

fitness _i＝1/max{|f1 _i| |f2 _i| |f3 _i|}，i=1～1000；

(6) population q is selected, intersect, variation, obtain and there are 1000 target range value Ra _iprogeny population q ₁, i=1 ~ 1000;

(7) step (4) ~ (5) are utilized to calculate progeny population q respectively ₁in each target range value Ra _ithe solution of corresponding Nonlinear System of Equations and fitness fitness _i, the solution of Nonlinear System of Equations is expressed as { Rx _iθ x _ihx _i, i=1 ~ 1000, wherein: Rx _ii-th distance value that target arrives infrared eye, Rx _iexistence range be 40km ~ 80km; θ x _itarget and infrared eye i-th angle value relative to the earth's core, θ x _iexistence range be 0rad ~ 0.0157rad; Hx _ii-th distance value of target to the earth's core and the difference of earth radius, hx _iexistence range be 0km ~ 80km;

(8) according to above Rx _i, θ x _i, hx _iscope acquired all solutions are selected, reject the solution outside this scope, retain the solution within this scope;

(9) solution retained in step (8) is carried out fully intermeshing according to fitness size, select front 100 solutions that fitness is large as the primary xa of particle filter _j={ Ry _jθ y _jhy _j, j=1 ~ 100, wherein: Ry _jthe jth distance value that target arrives infrared eye, θ y _jtarget and the infrared eye jth angle value relative to the earth's core, hy _jjth the distance value of target to the earth's core and the difference of earth radius;

(10) according to the primary xa in surveyed target pitch angle β and step (9) _j, calculate the weight w of each primary _j:

w _j＝|ln(|β-arcsin(((hy _j+R ₀)*cos(θy _j)-R ₀)/Ry _j)|)|；

(11) total weight value of all primaries is calculated utilize the weight w of each primary of total weight value g normalization _j, obtain the weights gw after by normalization _j, j=1 ~ 100;

(12) re-sampling operations is carried out to all primaries, obtain the particle x after 100 resamplings _j={ Rz _jθ z _jhz _j, the weights of each particle after resampling are designated as gz _j, j=1 ~ 100, wherein: Rz _jthe jth distance value that target arrives infrared eye, θ z _jtarget and the infrared eye jth angle value relative to the earth's core, hz _jjth the distance value of target to the earth's core and the difference of earth radius;

(13) according to the particle x after resampling _jin components R z _jand each particle x after resampling _jweights gz _j, calculate the distance of target to infrared eye

Wherein, Rz _jfor target is to a jth distance value of infrared eye, j=1 ~ 100.

Tool of the present invention has the following advantages:

1) the present invention measures owing to utilizing the angle of pitch of infrared eye to target, thus not outwardly emittance, namely under passive mode, realizes range finding, significantly enhances the stealth capabilities of range measurement system;

2) the present invention utilizes genetic algorithm and particle filter, the Nonlinear System of Equations comprising target range is solved, overcome the shortcoming that existing solution methods of nonlinear equations is large to initial value dependence and result is not unique, improve the feasibility of ranging process and the accuracy of result;

3) the present invention only need utilize infrared detector measurement target pitch angle β, so it is few to measure item, makes ranging process be easier to realize;

4) experiment shows, the present invention adjusts the distance the target of infrared eye 40km ~ 80km, can reach good distance estimations effect.

Accompanying drawing explanation

Fig. 1 is the location diagram of infrared eye and target travel in the present invention;

Fig. 2 is workflow diagram of the present invention;

Range estimation when Fig. 3 is simulation objectives linear uniform motion of the present invention and distance actual value comparison diagram;

Range error curve map when Fig. 4 is simulation objectives linear uniform motion of the present invention;

Range estimation when Fig. 5 is simulation objectives uniformly accelrated rectilinear motion of the present invention and distance actual value comparison diagram;

Range error curve map when Fig. 6 is simulation objectives uniformly accelrated rectilinear motion of the present invention;

Range estimation when Fig. 7 is simulation objectives Horizontal sinusoidal of the present invention motion and distance actual value comparison diagram;

Range error curve map when Fig. 8 is simulation objectives Horizontal sinusoidal of the present invention motion.

Embodiment

With reference to Fig. 2, performing step of the present invention is as follows:

Step 1, utilizes infrared detector measurement target pitch angle β:

Infrared charge coupling element CCD is arranged on the transverse axis of the three-dimensional turntable being fixed on ground, by infrared charge coupling element CCD optical axis alignment target, reads the output of the optical electric axial angle encoder be connected on three-dimensional turntable transverse axis, obtain target pitch angle β.

Step 2, with reference to Fig. 1, set up the ternary Nonlinear System of Equations comprising target range:

(2a) at the position M by target, the position O of infrared eye, in the triangle OKM of the position K composition in the earth's core, if OM=R, KM=R ₀+ h, OK=R ₀, wherein, R is the distance that target arrives infrared eye, R ₀for earth radius, h is the distance of target to the earth's core and the difference of earth radius; The cosine law is utilized to obtain equation: and this equation is solved, obtain the distance of target to infrared eye:

$R=-R_0\sin \mathrm{\β}+\sqrt{R_0^2\mathrm{si}{n}^2\mathrm{\β}+2h{R}_0+h^2},---\left[1\right],$

Wherein β is the target pitch angle recorded in step 1;

(2b) be connected with geocentric position K by target location M, the intersection point of this line and earth sphere is N, in Δ OMN, if OM=R, MN=h, is obtained by geometric relationship wherein, θ is target and the infrared eye angle relative to the earth's core; The cosine law is utilized to obtain equation:

$R^2=h^2+{\left({2R}_0\sin \frac{\mathrm{\θ}}{2}\right)}^2-2h\left(2R_0\sin \frac{\mathrm{\θ}}{2}\right)\cos (\frac{\mathrm{\π}}{2}+\frac{\mathrm{\θ}}{2}),$ Abbreviation is carried out to this formula, obtains the equation after abbreviation:

$R^2-h^2-{\left({2R}_0\sin \frac{\mathrm{\θ}}{2}\right)}^2-4h{R}_0\mathrm{si}{n}^2\frac{\mathrm{\θ}}{2}=0,---\left[2\right];$

(2c) at infrared eye position O, target location M, on the plane OMK that the earth's core K forms, cross O point and be earth sphere tangent line OL, cross M point and do the vertical line of tangent line OL and OL meets at D point, the line of M point and K point is MK, and line segment OD and line segment MK intersects at B point, forms two triangle Δ OKB and Δ BMD; According to BN=R in Δ OKB ₀/ cos (θ)-R ₀with BM=R*sin (β) in Δ BMD/cos (θ), obtain the distance of target to the earth's core and the difference of earth radius:

h＝R*sin(β)/cos(θ)+R ₀/cos(θ)-R ₀， [3]；

(2d) the above-mentioned equation of simultaneous [1], [2], [3], obtain the ternary Nonlinear System of Equations containing unknown quantity R, θ, h:

$f1=R+R_0\sin \mathrm{\β}-\sqrt{R_0^2{\sin }^2\mathrm{\β}+2h{R}_0+h^2}=0$

$f2=R^2-h^2-{\left({2R}_0\sin \frac{\mathrm{\θ}}{2}\right)}^2-4h{R}_0{\sin }^2\frac{\mathrm{\θ}}{2}=0---\left[4\right];$

f3＝-h+R*sin(β)/cos(θ)+R ₀/cos(θ)-R ₀＝0，

Wherein, f1, f2, f3 represent the value of three nonlinear equations respectively.

Step 3, in interval 30 ~ 40km, random generation 1000 target range values, are expressed as R _i, i=1 ~ 1000, form the initial population q of target range value; All the other two unknown quantity θ=0 in the above-mentioned system of equations of initialization [4], h=28km.

Step 4, by the target range value R of 1000 in initial population q _iand initialized θ, h are combined into 1000 groups of iterative value, are expressed as a _i={ R _iθ h}, respectively as the iterative initial value of quasi-Newton method, utilizes the Nonlinear System of Equations set up in quasi-Newton method solution procedure 2, obtains 1000 and target range value R _ithe solution of corresponding Nonlinear System of Equations, and gained solution is substituted in Nonlinear System of Equations, obtain each target range value R _ithe value f1 of three corresponding nonlinear equations _i, f2 _i, f3 _i, i=1 ~ 1000.

The common method of the Nonlinear System of Equations in the solution procedure 2 described in this step has quasi-Newton method, variable step Newton iteration method and gradient method etc., this example used but be not limited to for quasi-Newton method, quasi-Newton method routine call is from what Chongqing of ＂, computing machine typical number value-based algorithm and program (C++ version), People's Telecon Publishing House, 2003.7.1 ＂, corresponding chapters and sections are 8.2.8.

Step 5, utilizes the value f1 of the Nonlinear System of Equations obtained in step 4 _i, f2 _i, f3 _i, calculate 1000 target range value R in initial population q _icorresponding fitness fitness _i:

fitness _i＝1/max{|f1 _i| |f2 _i| |f3 _i|}，i=1～1000。

Step 6, selects initial population q, intersects, mutation operation.

(6a) the fitness fitness of each target range value in initial population q is utilized _i, the selection of gambling dish is carried out to initial population q, produces the population qa after comprising the selection of 1000 target range values, i=1 ~ 1000;

(6b) all target range values of the population qa after selection are carried out binary coding, and the target range value after all codings is divided into 500 groups at random, often group has two target range values;

(6c) two the target range values often organized are carried out two-point crossover, produce 1000 intersect after target range value;

(6d) with Probability p c=0.05, in steps (6c) obtain 1000 intersect after target range value carry out basic bit mutation, produces the target range values of 1000 renewals;

(6e) all target range values that step (6d) obtains are converted to decimal value, obtain the progeny population q with 1000 target range values ₁={ Ra ₁..., Ra _i..., Ra ₁₀₀₀, wherein, Ra _ibe i-th target range value, i=1 ~ 1000.

Step 7, utilizes step 4 ~ step 5 to calculate progeny population q respectively ₁in each target range value R _ithe solution of corresponding Nonlinear System of Equations and fitness fitness _i, the solution of Nonlinear System of Equations is expressed as { Rx _iθ x _ihx _i, i=1 ~ 1000, wherein: Rx _ii-th distance value that target arrives infrared eye, Rx _iexistence range be 40km ~ 80km; θ x _itarget and infrared eye i-th angle value relative to the earth's core, θ x _iexistence range be 0rad ~ 0.0157rad; Hx _ii-th distance value of target to the earth's core and the difference of earth radius, hx _iexistence range be 0km ~ 80km.

Step 8, according to above Rx _i, θ x _i, hx _iscope acquired all solutions are selected, reject the solution outside this scope, retain the solution within this scope.

Step 9, carries out fully intermeshing by the solution retained in step 8 according to fitness size, selects large front 100 solutions of fitness as the primary xa of particle filter _j={ Ry _jθ y _jhy _j, j=1 ~ 100, wherein: Ry _jthe jth distance value that target arrives infrared eye, θ y _jtarget and the infrared eye jth angle value relative to the earth's core, hy _jjth the distance value of target to the earth's core and the difference of earth radius.

Step 10, according to the primary xa in surveyed target pitch angle β and step 9 _j, calculate the weight w of each primary _j:

w _j＝|ln(|β-arcsin(((hy _j+R ₀)*cos(θy _j)-R ₀)/Ry _j)|)|。

Step 11, calculates the total weight value of all primaries the weight w of each primary of normalization _j, obtain the weights gw after normalization _j=g/w _j, j=1 ~ 100.

Step 12, carries out re-sampling operations to all primaries.

(12a) the accumulative weights s of each primary is calculated _m, m=1 ~ 100, computing method are: s ₁=gw ₁, s ₂=gw ₁+ gw ₂, s ₃=gw ₁+ gw ₂+ gw ₃..., s ₁₀₀=gw ₁+ gw ₂+ gw ₃..., gw ₁₀₀;

(12b) [0,1] upper equally distributed random number u is produced;

(12c) the accumulative weights s of all primaries obtained in step (12a) _min, search makes s _mthe accumulation weights s of the primary that the subscript m that>=u sets up is minimum _m, and record this primary: l=xa _m, m=1 ~ 100;

(12d) repeat step (12b) ~ step (12c) totally 100 times, note j is multiplicity, obtains the particle x after 100 resamplings _j={ Rz _jθ z _jhz _j,

Wherein: Rz _jthe jth distance value that target arrives infrared eye, θ z _jtarget and the infrared eye jth angle value relative to the earth's core, hz _jjth the distance value of target to the earth's core and the difference of earth radius, j=1 ~ 100,

By particle x after resampling _jweights be expressed as gz _j, j=1 ~ 100.

Step 13, according to the particle x after resampling _jin components R z _jand each particle x after resampling _jweights gz _j, calculate the distance of target to infrared eye

Wherein, Rz _jfor target is to a jth distance value of infrared eye, j=1 ~ 100.

Step 14, judges that ranging process is the need of end: if do not need ranging process to proceed down, then stop the measurement to target pitch angle β and distance estimations; Otherwise, proceed to step 1, again start ranging process.

Correctness of the present invention can be proven by relevant l-G simulation test.

1. simulated conditions:

Infrared eye is on the ground static all the time, if target is 65km to the initial distance of infrared eye, the elemental height on target range ground is 20km, earth radius R ₀be taken as 6372.797km; Infrared eye is measured target pitch angle β every 0.05s, and infrared eye is 0s ~ 50s to the observation time of target.

2. emulate content:

Utilize the present invention to do linear uniform motion to target, uniformly accelrated rectilinear motion, Horizontal sinusoidal three kinds of motion conditions of moving emulate:

Emulation 1, target range when utilizing the present invention to do horizontal linear uniform motion to infrared eye direction to target is estimated, wherein, target velocity is 200m/s, the contrast of gained Target Distance Estimation value and target range actual value as shown in Figure 3, range error as shown in Figure 4, in the diagram, the maximal value of range error is 5.2145%, and the average of range error is-0.1931%.

Emulation 2, target range when utilizing the present invention to do horizontal uniformly accelrated rectilinear motion to infrared eye direction to target is estimated, wherein, target velocity is 200m/s, and acceleration is 1m/s ², as shown in Figure 5, as shown in Figure 6, in figure 6, the maximal value of range error is 7.5990% to range error, and the average of range error is 0.0111% in the contrast of gained Target Distance Estimation value and target range actual value,

Emulation 3, target range when utilizing the present invention to do Horizontal sinusoidal motion to infrared eye direction to target is estimated, wherein, target in the horizontal direction speed is 200m/s, done the in the vertical direction sinusoidal motion of target is expressed as: y=510*sin (0.196*t), t is the target travel time, y is target travel distance, the contrast of gained Target Distance Estimation value and target range actual value as shown in Figure 7, range error as shown in Figure 8, in fig. 8, the maximal value of range error is 5.7912%, and the average of range error is-0.1797%.

From Fig. 3 ~ Fig. 8, the present invention can obtain good Target Distance Estimation effect.

Claims (4)

1., based on an infrared target passive ranging method for genetic algorithm and particle filter, comprise the steps:

(1) infrared eye is utilized to record target pitch angle β;

(2) according to the geometric relationship of infrared eye and target travel, the ternary Nonlinear System of Equations comprising target range is set up:

$f1=R+R_0\sin \mathrm{\β}-\sqrt{R_0^2{\sin }^2\mathrm{\β}+2h{R}_0+h^2}=0$

$f2=R^2-h^2-{\left(2R_0\sin \frac{\mathrm{\θ}}{2}\right)}^2-4h{R}_0{\sin }^2\frac{\mathrm{\θ}}{2}=0$

f3＝-h+R*sin(β)/cos(θ)+R ₀/cos(θ)-R ₀＝0

The unknown quantity of this ternary Nonlinear System of Equations is R, θ, h, and wherein, R is the distance that target arrives infrared eye, and θ is target and the infrared eye angle relative to the earth's core, and h is the distance of target to the earth's core and the difference of earth radius, R ₀for earth radius, f1, f2, f3 represent the value of three nonlinear equations respectively;

(3) random generation 1000 target range values in interval 30 ~ 40km, are expressed as R _i, i=1 ~ 1000, form the initial population q of target range value; All the other two unknown quantity θ=0 in the above-mentioned system of equations of initialization, h=28km;

(4) by the target range value R of 1000 in initial population q _iand initialized θ, h are combined into 1000 groups of iterative value, are expressed as a _i={ R _iθ h}, respectively as the iterative initial value of quasi-Newton method, utilizes the Nonlinear System of Equations set up in quasi-Newton method solution procedure (2), obtains 1000 and target range value R _ithe solution of corresponding Nonlinear System of Equations, and gained solution is substituted in Nonlinear System of Equations, obtain each target range value R _ithe value f1 of three corresponding nonlinear equations _i, f2 _i, f3 _i, i=1 ~ 1000;

(5) the value f1 of the Nonlinear System of Equations obtained in step (4) is utilized _i, f2 _i, f3 _i, calculate 1000 target range value R in initial population q _icorresponding fitness fitness _i:

fitness _i＝1/max{|f1 _i| |f2 _i| |f3 _i|}，i＝1～1000；

(6) population q is selected, intersect, variation, obtain and there are 1000 target range value Ra _iprogeny population q ₁, i=1 ~ 1000;

(7) step (4) ~ (5) are utilized to calculate progeny population q respectively ₁in each target range value Ra _ithe solution of corresponding Nonlinear System of Equations and fitness fitness _i, the solution of Nonlinear System of Equations is expressed as { Rx _iθ x _ihx _i, i=1 ~ 1000, wherein: Rx _ii-th distance value that target arrives infrared eye, Rx _iexistence range be 40km ~ 80km; θ x _itarget and infrared eye i-th angle value relative to the earth's core, θ x _iexistence range be 0rad ~ 0.0157rad; Hx _ii-th distance value of target to the earth's core and the difference of earth radius, hx _iexistence range be 0km ~ 80km;

(8) according to above Rx _i, θ x _i, hx _iscope acquired all solutions are selected, reject the solution outside this scope, retain the solution within this scope;

(9) solution retained in step (8) is carried out fully intermeshing according to fitness size, select front 100 solutions that fitness is large as the primary xa of particle filter _j={ Ry _jθ y _jhy _j, j=1 ~ 100, wherein: Ry _jthe jth distance value that target arrives infrared eye, θ y _jtarget and the infrared eye jth angle value relative to the earth's core, hy _jjth the distance value of target to the earth's core and the difference of earth radius;

(10) according to the primary xa in surveyed target pitch angle β and step (9) _j, calculate the weight w of each primary _j:

w _j＝|ln(|β-arcsin(((hy _j+R ₀)*cos(θy _j)-R ₀)/Ry _j)|)|；

(11) total weight value of all primaries is calculated utilize the weight w of each primary of total weight value g normalization _j, obtain the weights gw after normalization _j, j=1 ~ 100;

(12) re-sampling operations is carried out to all primaries, obtain the particle x after 100 resamplings _j={ Rz _jθ z _jhz _j, the weights of each particle after resampling are designated as gz _j, j=1 ~ 100, wherein: Rz _jthe jth distance value that target arrives infrared eye, θ z _jtarget and the infrared eye jth angle value relative to the earth's core, hz _jjth the distance value of target to the earth's core and the difference of earth radius;

(13) according to the particle x after resampling _jin components R z _jand each particle x after resampling _jweights gz _j, calculate the distance of target to infrared eye

Wherein, Rz _jfor target is to a jth distance value of infrared eye, j=1 ~ 100.

2. infrared target passive ranging method according to claim 1, the foundation wherein described in step (2) comprises the ternary Nonlinear System of Equations of target range, carries out as follows:

(2a) remember that the position of target is M, the position of infrared eye is O, and the position in the earth's core is K, in Δ OKM, and OM=R, KM=R ₀+ h, OK=R ₀, wherein, R is the distance that target arrives infrared eye, R ₀for earth radius, h is the distance of target to the earth's core and the difference of earth radius; The cosine law is utilized to obtain equation: and this equation is solved, obtain the distance of target to detector:

$R=-R_0\sin \mathrm{\β}+\sqrt{R_0^2{\sin }^2\mathrm{\β}+2h{R}_0+h^2}---\left[1\right],$

Wherein β is the target pitch angle recorded;

(2b) remember that the line of target location M and geocentric position K and the intersection point of earth sphere are N, in Δ OMN, OM=R, MN=h, obtained by geometric relationship wherein, θ is target and the infrared eye angle relative to the earth's core; The cosine law is utilized to obtain equation:

$R^2=h^2+{\left(2R_0\sin \frac{\mathrm{\θ}}{2}\right)}^2-2h\left(2R_0\sin \frac{\mathrm{\θ}}{2}\right)\cos (\frac{\mathrm{\π}}{2}+\frac{\mathrm{\θ}}{2}),$ Abbreviation is carried out to this formula, obtains the equation after abbreviation:

$R^2-h^2-{\left(2R_0\sin \frac{\mathrm{\θ}}{2}\right)}^2-4h{R}_0{\sin }^2\frac{\mathrm{\θ}}{2}=0---\left[2\right];$

(2c) at infrared eye position O, target location M, on the plane OMK that the earth's core K forms, cross O point and be earth sphere tangent line OL, cross M point and do the vertical line of tangent line OL and OL meets at D point, the line of M point and K point is MK, and line segment OD and line segment MK intersects at B point, forms two triangle Δ OKB and Δ BMD; By in Δ OKB, BN=R ₀/ cos (θ)-R ₀with in Δ BMD, BM=R*sin (β)/cos (θ), obtains the distance of target to the earth's core and the difference of earth radius:

h＝R*sin(β)/cos(θ)+R ₀/cos(θ)-R ₀[3]；

(2d) the above-mentioned equation of simultaneous [1], [2], [3], obtain the ternary Nonlinear System of Equations containing unknown quantity R, θ, h:

$f1=R+R_0\sin \mathrm{\β}-\sqrt{R_0^2{\sin }^2\mathrm{\β}+2h{R}_0+h^2}=0$

$f2=R^2-h^2-{\left(2R_0\sin \frac{\mathrm{\θ}}{2}\right)}^2-4h{R}_0{\sin }^2\frac{\mathrm{\θ}}{2}=0$

f3＝-h+R*sin(β)/cos(θ)+R ₀/cos(θ)-R ₀＝0

Wherein, f1, f2, f3 represent the value of three nonlinear equations respectively.

3. infrared target passive ranging method according to claim 1, the Nonlinear System of Equations set up in the solution procedure (2) wherein described in step (4), changes to variable step Newton iteration method or gradient method by employing quasi-Newton method.

4. infrared target passive ranging method according to claim 1, selecting population q wherein described in step (6), intersects, and variation obtains and has 1000 target range value Ra _iprogeny population q ₁, i=1 ~ 1000, carry out as follows:

(6a) the fitness fitness of each target range value in population q is utilized _i, the selection of gambling dish is carried out to population q, produces the population qa after comprising the selection of 1000 target range values, i=1 ~ 1000;

(6b) all target range values of the population qa after selection are carried out binary coding, and the target range value after all codings is divided into 500 groups at random, often group has two target range values;

(6c) two the target range values often organized are carried out two-point crossover, produce 1000 new target range values;

(6d) with Probability p c=0.05, to the target range value that (6c) 1000 of obtaining are new in steps carry out basic bit mutation, produce 1000 target range values upgraded;

(6e) all target range values that step (6d) obtains are converted to decimal value, obtain the progeny population q with 1000 target range values ₁={ Ra ₁..., Ra _i..., Ra ₁₀₀₀, wherein, Ra _ibe i-th target range value, i=1 ~ 1000.