CN101598789A - The time unifying method that is used for single or double base complex high-frequency radar nets - Google Patents

Abstract

The time unifying method that is used for single or double base complex high-frequency radar nets, it relate to high-spectrum remote sensing data emulation with the data fusion technical field, it has solved the detection accuracy and the power of the air missile weapon system of existing employing single target sensor autonomous operation and has reduced greatly, the killing area sharply reduces, cause the air defence weapon system fighting efficiency significantly to descend, even lose the problem of fight capability substantially.If each sensor when starting working, passed through system to the time; In period of time hereafter, each sensor obtains one group of data based on sampling instant.Select the measurement data of sensor, obtain a curve, draw the value of other any times by the curve calculation after the match through data are carried out curve fitting.Utilize this method in the single or double base complex high-frequency radar nets system, also to require to utilize existing equipment to improve the target detection precision as much as possible, to improve the electronic warfare capability of total system.

Description

The time unifying method that is used for single or double base complex high-frequency radar nets

Technical field

The present invention relates to high-spectrum remote sensing data emulation with the data fusion technical field.

Background technology

Since the nineties, the Gulf War, NATO air attack the Federal Republic of Yugoslavia, Afghan War and the war in Iraq have successively taken place in the world from eighties of last century.Make a general survey of these several wars, the offense uses measures such as stealthy scouting, electronic interferences, stealthy bombing and the attack of little radar area cruise missile in a large number, suppress the side's of defence antiaircraft weapon effectively and found, followed the tracks of and measure the ability of target, the feasible detection accuracy and the power of the air missile weapon system of single target sensor autonomous operation of adopting reduced greatly, the killing area sharply reduces, cause the air defence weapon system fighting efficiency significantly to descend, even lose fight capability substantially.In the face of increasingly serious air strike environment, improve the anti-stealthy and antijamming capability of air defence weapon system, keep the overall combat effectiveness of air defence weapon system, be current problem very important and to be solved.

People more and more are concerned about using sensor network monitoring in the past few years.Thereby sensor network can obtain bulk data from the sensor of many independent operatings merges the complementation of formation state, and the sensor network that utilizes multisensor to constitute has advantages such as the reliability of the coverage rate that increases data, enhanced system and robustness.In carrying out data handling procedure, the data-switching that needs localized sensor is recorded for the authentic communication that obtains target is in common coordinate reference system, but owing to have the deviation and the measuring error of sensor, directly change being difficult to guarantee precision and having given play to the superiority of utilizing multisensor, therefore when multi-sensor data is handled, need the registration Algorithm of some sensors.

In the single or double base complex high-frequency radar nets system, also require to utilize existing equipment to improve the target detection precision as much as possible, to improve the electronic warfare capability of total system.

Summary of the invention

The present invention reduces greatly in order to solve the existing detection accuracy and the power of the air missile weapon system of single target sensor autonomous operation of adopting, the killing area sharply reduces, cause the air defence weapon system fighting efficiency significantly to descend, even lose the problem of fight capability substantially, and a kind of time unifying method that is used for single or double base complex high-frequency radar nets has been proposed.

The time unifying method step that is used for single or double base complex high-frequency radar nets of the present invention is as follows:

Step 1: set simulation parameter; Set up rectangular coordinate system with the first base station T as true origin, and the distance of establishing the first base station T and the second base station R is L, set moving target P again, and to establish moving target P initial position be P (L, L), moving target P moves horizontally with the straight line that the speed of vm/s is parallel to the first base station T and R place, second base station; First sensor A among the first base station T is that collection period comes image data with x second, and the second sensor B of the second base station R is that collection period comes image data with y second;

Step 2: carry out emulation according to the simulation parameter in the step 1, and establish first sensor A and the second sensor B when beginning emulation image data, passed through system to the time; In time, the first sensor A and the second sensor B collect one group of sampled data based on the described observed samples time in after this one section observed samples;

Step 3: the measurement data of one or two sensor that is obtained in the selection step 2, through being carried out curve fitting, data obtain a curve, draw the value of other any times by the curve calculation after the match.

Be provided with the first sensor A and the second sensor B, with different frequencies target carried out sampled measurements respectively, each sensor can uniform sampling, also can nonuniform sampling.Each sensor is at sampling instant t ₁A measured value is arranged, be designated as (t _i, y _i), each sensor can obtain one group of measured value like this.Because the sampling period difference of sensor, the time ti value that each sensor obtains data is not quite similar.If directly merge, may make fusion results lose meaning owing to the mistiming, be not so good as the fusion accuracy height of single-sensor on the contrary.Therefore before carrying out data fusion, difference measurement data constantly must be registered on the same time point.

The present invention adopts the time alignment that carries out multisensor based on the method for curve fitting.If each sensor when starting working, passed through system to the time, promptly each sensor carries out measuring the first time to target at one time.This method is used for the emulation of single or double base complex high-frequency radar nets time unifying method data, it has considered that comprehensively thereby sensor network can obtain bulk data and merge the complementation of formation state from the sensor of many independent operatings, and the sensor network that utilizes multisensor to constitute has advantages such as the reliability of the coverage rate that increases data, enhanced system and robustness.Utilize this method in the single or double base complex high-frequency radar nets system, also to require to utilize existing equipment to improve the target detection precision as much as possible, to improve the electronic warfare capability of total system.

Description of drawings

Fig. 1 is the single or double base complex high-frequency radar nets synoptic diagram; Fig. 2 is the observation track plot before the registration; Fig. 3 is the observation track plot behind the registration.

Embodiment

Embodiment one: in conjunction with Fig. 1 present embodiment is described, the present embodiment step is as follows:

Step 1: set simulation parameter; Set up rectangular coordinate system with the first base station T as true origin, and the distance of establishing the first base station T and the second base station R is L, promptly the coordinate at two stations is: T (0,0), R (L, 0); Set moving target P again, and establish moving target P initial position be P (L, L), moving target P moves horizontally with the straight line that the speed of vm/s is parallel to the first base station T and R place, second base station; First sensor A among the first base station T is that collection period comes image data with x second, and the second sensor B of the second base station R is that collection period comes image data with y second;

Step 2: carry out emulation according to the simulation parameter in the step 1, and establish first sensor A and the second sensor B when beginning emulation image data, passed through system to the time, promptly each sensor carries out measurement first time to target at one time; In time, the first sensor A and the second sensor B collect one group of sampled data based on the described observed samples time in after this one section observed samples;

Step 3: the measurement data of one or two sensor that is obtained in the selection step 2, through being carried out curve fitting, data obtain a curve, draw the value of other any times by the curve calculation after the match; Can merge the data that each sensor records with registration this moment by certain criterion.

Embodiment two: present embodiment is described in conjunction with Fig. 1, Fig. 2 and Fig. 3, present embodiment and embodiment one difference are that curve fitting is the spline-fitting of adopting based on least square method in the step 3, splines comes down to a kind of polynomial function piecemeal, if first sensor A is at a certain observed samples time period [a, b] in to target carried out n+1 time the measurement, whole observed samples time interval is divided into by sampling instant: a=t ₀＜t ₁＜...＜t _n=b, given moment point t _iCorresponding observed reading is: f (t _i)=yi (i=0,1 ..., n), construct a cubic spline functions s (x), and make it satisfy following condition: s (t _i)=y _i, i=0,1 ..., n; S (t) is at each minizone [t _i, t _i+ 1] on a cubic polynomial, i=0,1 ..., n-1; S (t) has the Second Order Continuous derivative on [a, b];

It is as follows to make cubic spline functions s (x) satisfy the step of above-mentioned condition:

Splines match based on least square is in the splines space S _KIn (Δ), at first find out the optimal approximation about norm ‖ ù ‖, promptly find S for f (t) ^*(t), make

$\left|\right|f-S^{*}\left|\right|=\underset{S\∈S_k\left(\mathrm{\Δ}\right)}{\min }\left|\right|f-s\left|\right|---\left(1\right)$

Next constructs cubic spline functions:

Note m _i=S ' (t _i) (i=0,1,2 ..., n), at each minizone [t _i, t _i+ 1] (i=0,1,2 ..., n-1) on, utilize the Hermite interpolation formula to write out the computing formula of cubic spline functions s (t):

$s\left(t\right)=(1+2\frac{t-t_i}{t_{i+1}-t_i}){\left(\frac{t-t_{i+1}}{t_i-t_{i+1}}\right)}^2{y}_i+(1+2\frac{t-t_{i+1}}{t_i-t_{i+1}}){\left(\frac{t-t_i}{t_{i+1}-t_i}\right)}^2{y}_{i+1}+---\left(2\right)$

$(t-t_{i+1})\left(\frac{t-t_{i+1}}{t_i-t_{i+1}}\right)m_i(t-t_{i+1}){\left(\frac{t-t_i}{t_{i+1}-t_i}\right)}^2{m}_{i+1}$

Utilize s (t) on [a, b], to have the Second Order Continuous derivative; Then $s^{\′\′}\left(t_i^{-}\right)=s^{\′\′}\left(t_i^{+}\right),$ (i=0,1,2 ..., n-1), and additional boundary condition S " (t ₀)=S " (t _n)=0 can get system of equations:

$\left\{\begin{array}{c}{2\mathrm{<figure-callout\; id="0"\; label="m"\; filenames="A2009100723430008H2.png,A2009100723430009H1.png"\; state="\{\{state\}\}">m</figure-callout>}}_0+a_0{m}_1={\mathrm{\&beta;}}_0\\ (1-a_i)m_{i-1}+{2m}_i+a_i{m}_{i+1}={\mathrm{\&beta;}}_i,(i=\mathrm{0,1,2},...,n-1)\\ (1-a_n)m_{n-1}+{2m}_n={\mathrm{\&beta;}}_n\end{array}\right.---\left(3\right)$

Wherein: a ₀=1, $a_i=\frac{h_{i-1}}{h_{i-1}+h_i},$ i＝(1，2，...n-1)，a _n＝0，

${\mathrm{\&beta;}}_n=\frac{3}{h_{n-1}}(y_n-y_{n-1}),$

${\mathrm{\&beta;}}_0=\frac{3}{h_0}(y_1-y_0),$ ${\mathrm{\&beta;}}_i=3(\frac{1-a_i}{h_{i-1}}(y_i-y_{i-1})+\frac{a_i}{h_i}(y_{i+1}-y_i)),(i=\mathrm{1,2},...,n-1),$

h _i＝t _i+1-t _i，(i＝1，2，...n-1)；

$\left\{\begin{array}{c}{2\mathrm{<figure-callout\; id="0"\; label="m"\; filenames="A2009100723430008H2.png,A2009100723430009H1.png"\; state="\{\{state\}\}">m</figure-callout>}}_0+a_0{m}_1={\mathrm{\&beta;}}_0\\ (1-a_1){\mathrm{<figure-callout\; id="0"\; label="m"\; filenames="A2009100723430008H2.png,A2009100723430009H1.png"\; state="\{\{state\}\}">m</figure-callout>}}_0+{2m}_1+a_1{m}_2={\mathrm{\&beta;}}_1\\ (1-a_2)m_1+{2m}_2+a_2{m}_3={\mathrm{\&beta;}}_2\\ .......\\ (1-a_{n-1})m_{n-2}+{2m}_{n-1}+a_{n-1}{m}_n={\mathrm{\&beta;}}_{n-1}\\ (1-a_n)m_{n-1}+{2m}_n+a_n{m}_{n+1}={\mathrm{\&beta;}}_n\end{array}\right.,(i=\mathrm{0,1,2},...,n-1)---\left(4\right)$

The system of equations matrix of coefficients is a triangular matrix, and its determinant is not 0, so solution of equations exists and be unique; System of equations is found the solution, can be drawn recursion formula:

m _i＝α _im _i+1+b _i(i＝n，n-1，...1，0) (5)

Wherein:

${\mathrm{\&alpha;}}_i=\frac{-a_i}{2+(1-a_i)a_{i-1}},$ $b_i=\frac{{\mathrm{\&beta;}}_i-(1-a_i)b_{i-1}}{2+(1-a_i)a_{i-1}},(i=\mathrm{0,1,2},...,n-1)$

${\mathrm{\&alpha;}}_0=\frac{a_0}{2},$ $b_0=\frac{{\mathrm{\&beta;}}_0}{2}$

Using formula is asked α _i, b _i(i=0,1,2 ..., n-1), make m _I+1=0, obtain m _nAnd m _N-1... m ₀, with the parametric t of being given _i, y _i, m _i(i=0,1,2 ..., n-1) substitution s (x) promptly obtains the cubic spline functions asked; Through the spline interpolation match, can obtain a smooth curve, can be by this curve in the hope of first sensor A value at any time.At this moment again and through system to the time the second sensor B carry out time alignment, can take out the measured value in the corresponding moment from this curve according to the sampling instant of the second sensor B, can merge aligning.Other step is identical with embodiment one.

Claims (2)

1, the time unifying method that is used for single or double base complex high-frequency radar nets is characterized in that its step is as follows:

Step 1: set simulation parameter; (T) sets up rectangular coordinate system as true origin with first base station, and the distance of establishing first base station (T) and second base station (R) is L, set moving target (P) again, and to establish moving target (P) initial position be P (L, L), moving target (P) moves horizontally with the straight line that the speed of vm/s is parallel to first base station (T) and place, second base station (R); First sensor (A) in first base station (T) is that collection period comes image data with x second, and second sensor (B) of second base station (R) is that collection period comes image data with y second;

Step 2: carry out emulation according to the simulation parameter in the step 1, and establish first sensor (A) and second sensor (B) when beginning emulation image data, passed through system to the time; In time, first sensor (A) and second sensor (B) collect one group of sampled data based on the described observed samples time in after this one section observed samples;

Step 3: the measurement data of one or two sensor that is obtained in the selection step 2, through being carried out curve fitting, data obtain a curve, draw the value of other any times by the curve calculation after the match.

2, the time unifying method that is used for single or double base complex high-frequency radar nets according to claim 1 is characterized in that curve fitting is the spline-fitting of adopting based on least square method in the step 3; Splines match based on least square is in the splines space S _KIn (Δ), at first find out for f (t) about norm || ǜ || optimal approximation, promptly find S ^*(t), make

$\left|\right|f-S^{*}\left|\right|=\underset{S\&Element;S_k\left(\mathrm{\&Delta;}\right)}{\min }\left|\right|f-s\left|\right|---\left(1\right)$

Next constructs cubic spline functions:

Note m _i=S ' (t _i) (i=0,1,2 ..., n), at each minizone [t _i, t _i+ 1] (i=0,1,2 ..., n-1) on, utilize the Hermite interpolation formula to write out the computing formula of cubic spline functions s (t):

$s\left(t\right)=(1+2\frac{t-t_i}{t_{i+1}-t_i}){\left(\frac{t-t_{i+1}}{t_i-t_{i+1}}\right)}^2{y}_i+(1+2\frac{t-t_{i+1}}{t_i-t_{i+1}}){\left(\frac{t-t_i}{t_{i+1}-t_i}\right)}^2{y}_{i+1}+$

(2)

$(t-t_{i+1})\left(\frac{t-t_{i+1}}{t_i-t_{i+1}}\right)m_i(t-t_{i+1}){\left(\frac{t-t_i}{t_{i+1}-t_i}\right)}^2{m}_{i+1}$

Utilize s (t) on [a, b], to have the Second Order Continuous derivative; Then $s^{\&prime;\&prime;}\left(t_i^{-}\right)=s^{\&prime;\&prime;}\left(t_i^{+}\right),$ (i=0,1,2 ..., n-1), and additional boundary condition S " (t ₀)=S " (t _n)=0 can get system of equations:

$\left\{\begin{array}{cc}2m_0+a_0{m}_1={\mathrm{\&beta;}}_0& \\ (1-a_i)m_{i-1}+2m_i+& a_i{m}_{i+1}={\mathrm{\&beta;}}_i(i=\mathrm{0,1,2},\&CenterDot;\&CenterDot;\&CenterDot;,n-1)---\left(3\right)\\ (1-a_n)m_{n-1}+2m_n={\mathrm{\&beta;}}_n& \end{array}\right.$

Wherein: a ₀=1, $a_i=\frac{h_{i-1}}{h_{i-1}+h_i},$ i＝(1，2，...n-1)，a _n＝0，

${\mathrm{\&beta;}}_0=\frac{3}{h_0}(y_1-y_0),$ ${\mathrm{\&beta;}}_i=3(\frac{1-a_i}{h_{i-1}}(y_i-y_{i-1})+\frac{a_i}{h_i}(y_{i+1}-y_i))$ (i＝1，2，...，n-1)， ${\mathrm{\&beta;}}_n=\frac{3}{h_{n-1}}(y_n-y_{n-1}),$ h _i＝t _i+1-t _i，(i＝1，2，...n-1)；

$\left\{\begin{array}{c}2m_0+a_0{m}_1={\mathrm{\&beta;}}_0\\ (1-a_1)m_0+2m_1+a_1{m}_2={\mathrm{\&beta;}}_1\\ (1-a_2)m_1+2m_2+a_2{m}_3={\mathrm{\&beta;}}_2\\ \&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\\ (1-a_{n-1})m_{n-2}+2m_{n-1}+a_{n-1}{m}_n={\mathrm{\&beta;}}_{n-1}\\ (1-a_n)m_{n-1}+{2m}_n+a_n{m}_{n+1}={\mathrm{\&beta;}}_n\end{array}\right.,(i=\mathrm{0,1,2},\&CenterDot;\&CenterDot;\&CenterDot;,n-1)---\left(4\right)$

The system of equations matrix of coefficients is a triangular matrix, and its determinant is not 0, so solution of equations exists and be unique;

System of equations is found the solution, can be drawn recursion formula:

m _i＝α _im _i+1+b _i(i＝n，n-1，...1，0) (5)

Wherein:

${\mathrm{\&alpha;}}_i=\frac{-a_i}{2+(1-a_i)a_{i-1}},$ $b_i=\frac{{\mathrm{\&beta;}}_i-(1-a_i)b_{i-1}}{2+(1-a_i)a_{i-1}}$ (i＝0，1，2，...，n-1)

${\mathrm{\&alpha;}}_0=\frac{a_0}{2},$ $b_0=\frac{{\mathrm{\&beta;}}_0}{2}$

Using formula is asked α _i, b _i(i=0,1,2 ..., n-1), make m _I+1=0, obtain m _nAnd m _N-1... m ₀, with the parametric t of being given _i, y _i, m _i(i=0,1,2 ..., n-1) substitution s (x) promptly obtains the cubic spline functions asked; Through the spline interpolation match, can obtain a smooth curve, can be by this curve in the hope of first sensor (A) value at any time.At this moment again and through system to the time second sensor (B) carry out time alignment, can take out the measured value in the corresponding moment from this curve according to the sampling instant of second sensor (B), can merge aligning.

Images