CN116483082A - Multi-agent formation recognition method based on interpolation and fuzzy recognition - Google Patents
Multi-agent formation recognition method based on interpolation and fuzzy recognition Download PDFInfo
- Publication number
- CN116483082A CN116483082A CN202310428311.7A CN202310428311A CN116483082A CN 116483082 A CN116483082 A CN 116483082A CN 202310428311 A CN202310428311 A CN 202310428311A CN 116483082 A CN116483082 A CN 116483082A
- Authority
- CN
- China
- Prior art keywords
- formation
- recognition
- membership
- interpolation
- membership function
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000015572 biosynthetic process Effects 0.000 title claims abstract description 88
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000005755 formation reaction Methods 0.000 claims description 83
- 239000003795 chemical substances by application Substances 0.000 claims description 24
- 239000013598 vector Substances 0.000 claims description 16
- 229910003460 diamond Inorganic materials 0.000 claims description 10
- 239000010432 diamond Substances 0.000 claims description 10
- 241000826860 Trapezium Species 0.000 claims description 8
- QNRATNLHPGXHMA-XZHTYLCXSA-N (r)-(6-ethoxyquinolin-4-yl)-[(2s,4s,5r)-5-ethyl-1-azabicyclo[2.2.2]octan-2-yl]methanol;hydrochloride Chemical compound Cl.C([C@H]([C@H](C1)CC)C2)CN1[C@@H]2[C@H](O)C1=CC=NC2=CC=C(OCC)C=C21 QNRATNLHPGXHMA-XZHTYLCXSA-N 0.000 claims description 4
- 238000005259 measurement Methods 0.000 abstract description 5
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000010276 construction Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 5
- 238000001514 detection method Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000004088 simulation Methods 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000003909 pattern recognition Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004064 recycling Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0219—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory ensuring the processing of the whole working surface
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Image Analysis (AREA)
Abstract
The invention belongs to the technical field of artificial intelligence related to formation recognition, and particularly relates to a multi-agent formation recognition method based on interpolation and fuzzy recognition. The method comprises the steps that a plurality of agents are formed and enter a radar identification area, and the radar scans and acquires coordinates of each agent in the identification area to obtain a scattered point set; acquiring four vertexes from the scattered point set; adopting a multipoint interpolation mode, randomly adding interpolation points between vertexes and adding the interpolation points into a scattered point set; constructing a formation according to the connecting lines of the vertexes; identifying the constructed formation by adopting a fuzzy identification method; and iterating for a plurality of times, predicting membership degrees of each next team shape through a fitting function according to the last recognition result, and taking the team shape corresponding to the maximum membership degree as a final recognition result. The invention uses fuzzy recognition, utilizes the membership function as the measurement of the sample and the template, can better reflect the integral characteristics of the mode, and has strong rejection capability for interference and noise in the sample.
Description
Technical Field
The invention belongs to the technical field of artificial intelligence related to formation recognition, and particularly relates to a multi-agent formation recognition method based on interpolation and fuzzy recognition.
Background
A multi-agent system is a collection of agents whose goal is to build a large and complex system into a small, mutually communicating and coordinated, easily managed system. Its research involves the knowledge, goal, skill, planning, and formation identification of the multi-agent system overall team. The current formation recognition technology develops:
a) Rules-based multi-agent formation identification. At present, the formation research of multiple air agents is often influenced by the resolution of detection equipment and measurement errors, single targets in the air formation are difficult to distinguish, and the formation targets have large random change in interval, so that the formation is difficult to capture. The multi-agent formation recognition based on the rules creates formation recognition rules, and the rules are operated by constructing a rule operation platform and leading the aerial operation simulation data to drive the rule operation, so that some typical aerial formation is researched and judged.
b) Formation identification based on Hough transformation and K-means algorithm. In the common formation recognition, the distance between the targets directly affects the capturing precision of the detector, and the smaller the distance between the targets is, the poorer the selectivity of the detector to the specific targets is, but if the target selecting precision of the detector is to be improved, the searching area range of the detector device is required to be reduced, so that the probability of capturing the targets is reduced. The formation can be accurately identified by adopting Hough transformation and a K-means algorithm, the method is easy to implement, the identification result is accurate, the time consumption is less, the formation identification rate can be greatly improved, and the preparation time is shortened.
Through the above analysis, the problems and defects existing in the prior art are as follows: 521
1) The method is greatly influenced by the resolution of the detection equipment and measurement errors, single targets in formation are difficult to distinguish, and the interval between the formed targets is greatly changed randomly, so that the formation is difficult to capture. When the recognition range is enlarged, the recognition accuracy is reduced, the accuracy is improved, and the recognition efficiency is correspondingly reduced.
2) When formation detection is incomplete, the recognition accuracy is low.
3) The current rules are mainly dependent on manual setting, so that the limitation is quite large.
The invention expands on the basis of recognition and provides a formation recognition technology capable of predicting a result as early as possible when the formation is not completely presented.
Disclosure of Invention
The invention aims to provide a multi-agent formation recognition method based on interpolation and fuzzy recognition.
A multi-agent formation recognition method based on interpolation and fuzzy recognition comprises the following steps:
step 1: forming a plurality of agents into a radar identification area, and scanning and acquiring coordinates of each agent in the identification area by using a radar to obtain a scattered point set A;
step 2: four vertexes are obtained from the scattered point set A;
for the scattered point set A, defining the distance between any two points as d; constructing a set B, wherein elements in the set B take the first two groups of scattered points with the largest corresponding distance d; the abscissa of each element in the set B forms a vector X, and the ordinate of each element forms a vector Y, but when the elements in the set B are quite close, the vertex found in this way is meaningless, and a constraint condition is added to the vertex, so that the corresponding vertex is not selected when the situation occurs; the constraint is defined as:
the maximum between the vectors X, Y is made greater than delta, i.e. max|x i -y i |>δ
Wherein x is i ,y i I=1, 2,3,4, being components of the vectors X, Y;
for the scattered point set A, the distances between each scattered point and other scattered points are obtained, all distance results are arranged from large to small, and then the distances are distributed according to the constraint condition max|x i -y i Selecting four scattered points as vertexes, wherein delta is larger than delta;
step 3: adopting a multipoint interpolation mode, randomly adding a interpolation points between vertexes and adding the a interpolation points into a scattered point set A;
the formula of the multipoint interpolation is:
step 4: according to the connection line of each vertex in the set B, constructing a formation, and obtaining four angle characteristics A of the formation 1 、A 2 、A 3 、A 4 The method comprises the steps of carrying out a first treatment on the surface of the Wherein A is 1 、A 2 、A 3 、A 4 Represents the internal angle degree, and A 1 ≥A 2 ≥A 3 ≥A 4 ;
Step 5: identifying the constructed formation by adopting a fuzzy identification method;
suppose that for a given quadrilateral (a 1 ,A 2 ,A 3 ,A 4 ) If A j Is A i Is defined as D (A) i ;A 1 ,A 2 ,A 3 )=A j If A j Is A i Is defined by L (A) i ;A 1 ,A 2 ,A 3 )=A j ;
Step 5.1: calculating a V-shaped Va membership function;
according to the characteristic of the V shape, the inner angle of the V shape is three angles, and four angles of the V shape and the rhombic trapezoid are distinguished, namely when the number of the angles is three, the formation is judged to be the V shape;
when 180 > A 1 >A 2 >A 3 >0,A 4 When=0, va (a 1 ,A 2 ,A 3 ,A 4 )=1
When 180 > A 1 >A 2 >A 3 >A 4 At > 0, va (A 1 ,A 2 ,A 3 ,A 4 )=0
Step 5.2: calculating a diamond Pa membership function;
when A is 2 Is A 1 Is the diagonal of (A), i.e. D (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Pa (A) 1 ,A 2 ,A 3 ,A 4 ) The method comprises the following steps:
when A is 2 Is A 1 When adjacent angles of (a), i.e. L (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Pa (A) 1 ,A 2 ,A 3 ,A 4 ) The method comprises the following steps:
step 5.3: calculating a trapezoidal Tr membership function;
when A is 2 Is A 1 Is the diagonal of (A), i.e. D (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Tr of Tr (A 1 ,A 2 ,A 3 ,A 4 ) The method comprises the following steps:
∧min[|A 2 +L(A 2 ;A 3 ,A 4 )-180|+|A 2 -L(A 2 ;A 3 ,A 4 )|]}
when A is 2 Is A 1 When adjacent angles of (a), i.e. L (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Tr of Tr (A 1 ,A 2 ,A 3 ,A 4 ) The method comprises the following steps:
∧[|A i +L(A i ;A 3 ,A 4 )-180|+|A i -L(A i ;A 3 ,A 4 )|]},i=1,2
step 5.4: calculating other quadrilateral Qu membership functions; the other quadrangles do not have the characteristics of rhombus and trapezium;
Qu(A 1 ,A 2 ,A 3 ,A 4 )=Pa C ∧Tr C
=(1-Pa)∧(1-Tr)
step 5.5: respectively calculating membership degrees of four formations belonging to Va, pa, tr and Qu, and taking the formation corresponding to the largest membership degree as a recognition result, wherein R=max (Va, pa, ta and Qa);
step 6: repeating the step 2 and the step 5 until the set iteration times b; predicting membership Va of each next team through fitting function delta according to the obtained b times of recognition results δ ,Pa δ ,Ta δ ,Qa δ Taking the formation corresponding to the biggest membership degree as a final recognition result R δ =max(Va δ ,Pa, δ Ta δ ,Qa δ )。
The invention has the beneficial effects that:
compared with the prior art, the method and the device can quickly identify the target formation. The results of the invention are fitted in the first few results, and the final formation can be predicted in a short time. The invention uses fuzzy recognition, in the fuzzy pattern recognition, the membership function is used as the measurement of the sample and the template, the integral characteristic of the pattern can be better reflected, and the invention has strong rejection capability for interference and noise in the sample.
Drawings
FIG. 1 is a schematic illustration of a V-shaped multi-agent formation.
Fig. 2 is a schematic diagram of a diamond multi-agent formation.
Fig. 3 is a schematic diagram of a trapezoidal multi-agent formation.
Fig. 4 is a flow chart of the present invention.
FIG. 5 is a construction diagram of a V-shaped formation.
FIG. 6 is a construction diagram of a diamond-shaped formation.
FIG. 7 is a construction diagram of a trapezoidal formation.
Detailed Description
The invention is further described below with reference to the accompanying drawings.
The invention carries out simulation on the formation construction process by deeply analyzing the formation generation elements of multiple agents, including common formation forms generated by the elements, adding random points in formation generation, not changing the formation forms generated by the formation generation angle, gradually increasing the number of formation midpoints from far to near, and the like. Starting from the initial state, repeatedly realizing the work of identifying the vertexes and then constructing the formations, after the construction of a new formation is completed, making a new identification result for each newly constructed formation, ensuring that the detection result can be obtained in the shortest time, repeatedly carrying out until the formation construction is completed, and then fitting the results obtained in the previous times to obtain a final result.
1. Identifying vertices
The invention adopts Manhattan distance, and for the scattered point set A, the distance between any two points is defined as d, namely:
with d (m, n) = |x m -x n |+|y m -y n |
This distance is further optimized, such that b=maxd (m, n),i.e. set B is a subset of set a that maximizes distance d (m, n).
Then, the abscissa of each element in the set B is made into a vector X, and the ordinate of each element is made into a vector Y, but when the elements in the set B are quite close, the vertex thus found is meaningless, and a constraint condition needs to be added to this, so that when the above situation occurs, the corresponding vertex is not selected. The definition is as follows:
let the maximum between the vectors X, Y be greater than delta, i.e. max|x i -y i |>δ
Wherein x is i ,y i I=1, 2, …, n, n being the number of elements in set B, which are components of vector X, Y.
For a stack of scattered points, the points and the distance from the most distant point can be found, and the result is arranged from big to small and then according to max|x i -y i After the delta is eliminated to a part of the vertexes, the first four largest points are selected.
2. Formation construction
In the point set A which is scanned originally, a set A' is obtained by adding some points again, and the fixed point number is added every time of circulation, so that the process of increasing the number of targets from far to near is simulated. After the process is finished, some vertexes can be obtained, then an interpolation method is adopted according to the vertexes, an interpolation formula is calculated, several interpolation formulas are obtained according to different interpolation methods, then random points are taken from the interpolation formulas, and the random points are used as new adding points. And each increment point is incremented on the last set.
The coefficients of the multipoint interpolation formula are:
substituting the y values of each point to obtain an interpolation formula.
3. Formation identification
The invention adopts a fuzzy recognition method to recognize the constructed formation, firstly gives a part of assumptions, and according to the result of formation construction, can obtain some characteristics from the formation, and takes four angles A 1 ,A 2 ,A 3 ,A 4 Wherein A is 1 ,A 2 ,A 3 ,A 4 The internal angle is expressed, and for the sake of calculation and discussion, A is assumed to be 1 ≥A 2 ≥A 3 ≥A 4 。
We assume that for a given quadrilateral (a 1 ,A 2 ,A 3 ,A 4 ) If A j Is A i Is defined by D (A) i ;A 1 ,A 2 ,A 3 )=A j If A j Is A i Is defined by the adjacent angle of L (A) i ;A 1 ,A 2 ,A 3 )=A j ,i=1,2,3,4;j=1,2,3,4,i≠j。
Obviously, if A j Is A i Is a diagonal of (A) i Is A j Is a diagonal of (a); if A j Is A i Adjacent angle of (A) i Is A j I=1, 2,3,4; j=1, 2,3,4, i+.j.
The membership functions of the three formations are discussed separately below
1. Establishment of V-shaped Va membership function
According to the characteristic of the V shape, the inner angle is three angles, and four angles of the trapezoid are distinguished from each other, namely when the number of the angles is three, the formation of the trapezoid is judged to be the V shape.
(I) When 180 > A 1 >A 2 >A 3 >0,A 4 When=0, va (a 1 ,A 2 ,A 3 ,A 4 )=1;
(II) when 180 > A 1 >A 2 >A 3 >A 4 At > 0, va (A 1 ,A 2 ,A 3 ,A 4 )=0
2. Establishment of diamond Pa membership function
Four interior corners of a quadrilateral, of which the second smallest corner A 2 Or is A 1 Is the diagonal of, or A 1 Is discussed in two cases:
(1) When A is 2 Is A 1 Is the diagonal of (A), i.e. D (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Pa (A) 1 ,A 2 ,A 3 ,A 4 ) The following constraints should be satisfied:
(I) When A is 1 =A 2 ,A 3 =A 4 When Pa (A) 1 ,A 2 ,A 3 ,A 4 )=1;
(II) when A 1 =180,A 2 =A 3 =90,A 4 When=0, pa (a 1 ,A 2 ,A 3 ,A 4 )=0;
(III)0≤Pa(A 1 ,A 2 ,A 3 ,A 4 )≤1。
And the closer one angle is to 180 than the sum of adjacent angles, the smaller the difference between the adjacent angles is, the more rhombic is, therefore, the membership function of Pa is defined as
The membership function Pa thus defined obviously satisfies the constraints (I), (II), (III).
(2) When A is 2 Is A 1 When adjacent angles of (a), i.e. L (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Pa (A) 1 ,A 2 ,A 3 ,A 4 ) The following constraints should be satisfied:
(I) When D (A) 1 ;A 3 ,A 4 )=A 1 ,D(A 2 ;A 3 ,A 4 )=A 2 Time Pa (A) 1 ,A 2 ,A 3 ,A 4 )=1;
(II) when A 1 =180,A 2 =180,A 3 =0,A 4 Pa when=0 (a 1 ,A 2 ,A 3 ,A 4 )=0;
(III)0≤Pa(A 1 ,A 2 ,A 3 ,A 4 )≤1。
And the closer 180 is the sum of one angle and any adjacent angle, the smaller is the difference between the two angles, the more like a diamond, therefore, the membership function of Pa can be defined as
The membership function Pa thus defined obviously satisfies the constraints (I), (II), (III).
3. Establishment of membership functions for trapezium Tr
Like the diamond, the following is discussed in two cases:
(1) When A is 2 Is A 1 Is the diagonal of (A), i.e. D (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function of Tr, i.e. Tr (A 1 ,A 2 ,A 3 ,A 4 ) The following constraints should be satisfied:
(I) When A is 1 +A 3 =180,A 1 -A 3 =0,A 1 +A 4 =180,A 1 -A 4 When=0, tr (a 1 ,A 2 ,A 3 ,A 4 )=1;
(II) when A 1 =180,A 2 =180,A 3 =0,A 4 When=0, tr (a 1 ,A 2 ,A 3 ,A 4 )=0;
(III)0≤Tr(A 1 ,A 2 ,A 3 ,A 4 )≤1。
And the trapezium Tr is characterized by an inner angle approximately approaching 180 from any adjacent angle, the difference approaching 0, the quadrilateral is more like a trapezium, therefore, the membership function of Tr is defined as
∧min[|A 2 +L(A 2 ;A 3 ,A 4 )-180|+|A 2 -L(A 2 ;A 3 ,A 4 )|]}
The membership function Tr thus defined obviously satisfies the constraints (I), (II), (III).
(2) When A is 2 Is A 1 When adjacent angles of (a), i.e. L (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function of Tr, i.e. Tr (A 1 ,A 2 ,A 3 ,A 4 ) The following constraints should be satisfied:
(I) When A is 1 +A 2 =180,A 1 -A 2 =0 or a 1 +L(A 1 ;A 3 ,A 4 )=180,A 1 -L(A 1 ;A 3 ,A 4 ) When=180, tr (a 1 ,A 2 ,A 3 ,A 4 )=1;
(II) when A 1 =180,A 2 =180,A 3 =0,A 4 When=0, tr (a 1 ,A 2 ,A 3 ,A 4 )=0;
(III)0≤Tr(A 1 ,A 2 ,A 3 ,A 4 )≤1。
And the trapezium Tr is characterized by an inner angle approximately approaching 180 from any adjacent angle, the difference approaching 0, the quadrilateral is more like a trapezium, therefore, the membership function of Tr is defined as
∧[|A i +L(A i ;A 3 ,A 4 )-180|+|A i -L(A i ;A 3 ,A 4 )|]},i=1,2
The membership function Tr thus defined obviously satisfies the constraints (I), (II), (III).
4. Establishment of other quadrilateral Qu membership functions
Other quadrilaterals do not have the characteristics of a diamond, trapezoid, so the membership function of other quadrilaterals, qu (A) 1 ,A 2 ,A 3 ,A 4 ) Is that
Qu(A 1 ,A 2 ,A 3 ,A 4 )=Pa C ∧Tr C
=(1-Pa)∧(1-Tr)
Based on the above, the multi-agent formation recognition method based on interpolation and fuzzy recognition provided by the invention comprises the following steps:
step 1: forming a plurality of agents into a radar identification area, and scanning and acquiring coordinates of each agent in the identification area by using a radar to obtain a scattered point set A;
step 2: four vertexes are obtained from the scattered point set A;
for the scattered point set A, defining the distance between any two points as d; constructing a set B, wherein elements in the set B take the first two groups of scattered points with the largest corresponding distance d; the abscissa of each element in the set B forms a vector X, and the ordinate of each element forms a vector Y, but when the elements in the set B are quite close, the vertex found in this way is meaningless, and a constraint condition is added to the vertex, so that the corresponding vertex is not selected when the situation occurs; the constraint is defined as:
the maximum between the vectors X, Y is made greater than delta, i.e. max|x i -y i |>δ
Wherein x is i ,y i I=1, 2,3,4, being components of the vectors X, Y;
for the scattered point set A, the distances between each scattered point and other scattered points are obtained, all distance results are arranged from large to small, and then the distances are distributed according to the constraint condition max|x i -y i Selecting four scattered points as vertexes, wherein delta is larger than delta;
step 3: adopting a multipoint interpolation mode, randomly adding a interpolation points between vertexes and adding the a interpolation points into a scattered point set A;
the formula of the multipoint interpolation is:
step 4: according to set BConnecting lines of all vertexes to construct a formation and obtain four angle characteristics A of the formation 1 、A 2 、A 3 、A 4 The method comprises the steps of carrying out a first treatment on the surface of the Wherein A is 1 、A 2 、A 3 、A 4 Represents the internal angle degree, and A 1 ≥A 2 ≥A 3 ≥A 4 ;
Step 5: identifying the constructed formation by adopting a fuzzy identification method;
suppose that for a given quadrilateral (a 1 ,A 2 ,A 3 ,A 4 ) If A j Is A i Is defined as D (A) i ;A 1 ,A 2 ,A 3 )=A j If A j Is A i Is defined by L (A) i ;A 1 ,A 2 ,A 3 )=A j ;
Step 5.1: calculating a V-shaped Va membership function;
according to the characteristic of the V shape, the inner angle of the V shape is three angles, and four angles of the V shape and the rhombic trapezoid are distinguished, namely when the number of the angles is three, the formation is judged to be the V shape;
when 180 > A 1 >A 2 >A 3 >0,A 4 When=0, va (a 1 ,A 2 ,A 3 ,A 4 )=1
When 180 > A 1 >A 2 >A 3 >A 4 At > 0, va (A 1 ,A 2 ,A 3 ,A 4 )=0
Step 5.2: calculating a diamond Pa membership function;
when A is 2 Is A 1 Is the diagonal of (A), i.e. D (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Pa (A) 1 ,A 2 ,A 3 ,A 4 ) The method comprises the following steps:
when A is 2 Is A 1 When adjacent angles of (a), i.e. L (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Pa (A) 1 ,A 2 ,A 3 ,A 4 ) The method comprises the following steps:
step 5.3: calculating a trapezoidal Tr membership function;
when A is 2 Is A 1 Is the diagonal of (A), i.e. D (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Tr of Tr (A 1 ,A 2 ,A 3 ,A 4 ) The method comprises the following steps:
∧min[|A 2 +L(A 2 ;A 3 ,A 4 )-180|+|A 2 -L(A 2 ;A 3 ,A 4 )|]}
when A is 2 Is A 1 When adjacent angles of (a), i.e. L (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Tr of Tr (A 1 ,A 2 ,A 3 ,A 4 ) The method comprises the following steps:
step 5.4: calculating other quadrilateral Qu membership functions; the other quadrangles do not have the characteristics of rhombus and trapezium;
Qu(A 1 ,A 2 ,A 3 ,A 4 )=Pa C ∧Tr C
=(1-Pa)∧(1-Tr)
step 5.5: respectively calculating membership degrees of four formations belonging to Va, pa, tr and Qu, and taking the formation corresponding to the largest membership degree as a recognition result, wherein R=max (Va, pa, ta and Qa);
step 6: repeating the steps 2 and 5 to the set iteration timesA number b; predicting membership Va of each next team through fitting function delta according to the obtained b times of recognition results δ ,Pa δ ,Ta δ ,Qa δ Taking the formation corresponding to the biggest membership degree as a final recognition result R δ =max(Va δ ,Pa, δ Ta δ ,Qa δ )。
Example 1:
the invention starts from three common formations, such as a V shape, a diamond shape and a trapezoid shape in fig. 1,2 and 3, and then the whole implementation process is divided into three steps, such as fig. 4. The invention builds a formation simulation system, simulates and generates multiple intelligent agent formations from less to more in one simulation period, and gradually builds up to 50 targets from the initial 20 targets. And three typical multi-agent formations of V-shape, diamond shape and trapezoid shape are designed.
Membership degrees of four formations belonging to Va, pa, tr and Qu are calculated respectively, and the formation corresponding to the largest membership degree is taken as a recognition result, wherein R=max (Va, pa, ta and Qa). Recycling 5 times to obtain 5 times of identification results.
R 1 =max(Va 1 ,Pa 1 ,Ta 1 ,Qa 1 )
R 2 =max(Va 2 ,Pa 2 ,Ta 2 ,Qa 2 )
R 3 =max(Va 3 ,Pa 3 ,Ta 3 ,Qa 3 )
R 4 =max(Va 4 ,Pa 4 ,Ta 4 ,Qa 4 )
R 5 =max(Va 5 ,Pa 5 ,Ta 5 ,Qa 5 )
Finally, in order to achieve the prediction effect, the invention adopts a fitting method, and a fitting function is assumed to be delta, so that a new membership degree is obtained as a final recognition result. Calculating corresponding membership degree according to the formation obtained after the previous 5 increasing points, and fitting the membership degree calculated in the previous 5 times to predict the membership degree Va of the next time δ ,Pa δ ,Ta δ ,Qa δ 。
Va δ =δ(Va 1 ,Va 2 ,Va 3 ,Va 4 ,Va 5 )
Pa δ =δ(Pa 1 ,Pa 2 ,Pa 3 ,Pa 4 ,Pa 5 )
Ta δ =δ(Ta 1 ,Ta 2 ,Ta 3 ,Ta 4 ,Ta 5 )
Qa δ =δ(Qa 1 ,Qa 2 ,Qa 3 ,Qa 4 ,Qa 5 )
Taking the formation corresponding to the biggest membership degree as a final recognition result, R δ =max(Va δ ,Pa, δ Ta δ ,Qa δ )。
Compared with the prior art, the method and the device can quickly identify the target formation. The results of the invention are fitted in the first few results, and the final formation can be predicted in a short time. The invention uses fuzzy recognition, in the fuzzy pattern recognition, the membership function is used as the measurement of the sample and the template, the integral characteristic of the pattern can be better reflected, and the invention has strong rejection capability for interference and noise in the sample.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (1)
1. The multi-agent formation recognition method based on interpolation and fuzzy recognition is characterized by comprising the following steps of:
step 1: forming a plurality of agents into a radar identification area, and scanning and acquiring coordinates of each agent in the identification area by using a radar to obtain a scattered point set A;
step 2: four vertexes are obtained from the scattered point set A;
for the scattered point set A, defining the distance between any two points as d; constructing a set B, wherein elements in the set B take the first two groups of scattered points with the largest corresponding distance d; the abscissa of each element in the set B forms a vector X, and the ordinate of each element forms a vector Y, but when the elements in the set B are quite close, the vertex found in this way is meaningless, and a constraint condition is added to the vertex, so that the corresponding vertex is not selected when the situation occurs; the constraint is defined as:
the maximum between the vectors X, Y is made greater than delta, i.e. max|x i -y i |>δ
Wherein x is i ,y i I=1, 2,3,4, being components of the vectors X, Y;
for the scattered point set A, the distances between each scattered point and other scattered points are obtained, all distance results are arranged from large to small, and then the distances are distributed according to the constraint condition max|x i -y i Selecting four scattered points as vertexes, wherein delta is larger than delta;
step 3: adopting a multipoint interpolation mode, randomly adding a interpolation points between vertexes and adding the a interpolation points into a scattered point set A;
the formula of the multipoint interpolation is:
step 4: according to the connection line of each vertex in the set B, constructing a formation, and obtaining four angle characteristics A of the formation 1 、A 2 、A 3 、A 4 The method comprises the steps of carrying out a first treatment on the surface of the Wherein A is 1 、A 2 、A 3 、A 4 Represents the internal angle degree, and A 1 ≥A 2 ≥A 3 ≥A 4 ;
Step 5: identifying the constructed formation by adopting a fuzzy identification method;
suppose that for a given quadrilateral (a 1 ,A 2 ,A 3 ,A 4 ) If A j Is A i Is defined as D (A) i ;A 1 ,A 2 ,A 3 )=A j If A j Is A i Is defined by L (A) i ;A 1 ,A 2 ,A 3 )=A j ;
Step 5.1: calculating a V-shaped Va membership function;
according to the characteristic of the V shape, the inner angle of the V shape is three angles, and four angles of the V shape and the rhombic trapezoid are distinguished, namely when the number of the angles is three, the formation is judged to be the V shape;
when 180 > A 1 >A 2 >A 3 >0,A 4 When=0, va (a 1 ,A 2 ,A 3 ,A 4 )=1
When 180 > A 1 >A 2 >A 3 >A 4 At > 0, va (A 1 ,A 2 ,A 3 ,A 4 )=0
Step 5.2: calculating a diamond Pa membership function;
when A is 2 Is A 1 Is the diagonal of (A), i.e. D (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Pa (A) 1 ,A 2 ,A 3 ,A 4 ) The method comprises the following steps:
when A is 2 Is A 1 When adjacent angles of (a), i.e. L (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Pa (A) 1 ,A 2 ,A 3 ,A 4 ) The method comprises the following steps:
step 5.3: calculating a trapezoidal Tr membership function;
when A is 2 Is A 1 Is the diagonal of (A), i.e. D (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Tr of Tr (A 1 ,A 2 ,A 3 ,A 4 ) The method comprises the following steps:
∧min[|A 2 +L(A 2 ;A 3 ,A 4 )-180|+|A 2 -L(A 2 ;A 3 ,A 4 )|]}
when A is 2 Is A 1 When adjacent angles of (a), i.e. L (A) 1 ;A 2 ,A 3 ,A 4 )=A 2 Membership function Tr of Tr (A 1 ,A 2 ,A 3 ,A 4 ) The method comprises the following steps:
∧[|A i +L(A i ;A 3 ,A 4 )-180|+|A i -L(A i ;A 3 ,A 4 )|]},i=1,2
step 5.4: calculating other quadrilateral Qu membership functions; the other quadrangles do not have the characteristics of rhombus and trapezium;
Qu(A 1 ,A 2 ,A 3 ,A 4 )=Pa C ∧Tr C
=(1-Pa)∧(1-Tr)
step 5.5: respectively calculating membership degrees of four formations belonging to Va, pa, tr and Qu, and taking the formation corresponding to the largest membership degree as a recognition result, wherein R=max (Va, pa, ta and Qa);
step 6: repeating the step 2 and the step 5 until the set iteration times b; predicting membership Va of each next team through fitting function delta according to the obtained b times of recognition results δ ,Pa δ ,Ta δ ,Qa δ Taking the largest of themThe formation corresponding to the membership degree is used as a final recognition result R δ =max(Va δ ,Pa, δ Ta δ ,Qa δ )。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310428311.7A CN116483082A (en) | 2023-04-20 | 2023-04-20 | Multi-agent formation recognition method based on interpolation and fuzzy recognition |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310428311.7A CN116483082A (en) | 2023-04-20 | 2023-04-20 | Multi-agent formation recognition method based on interpolation and fuzzy recognition |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116483082A true CN116483082A (en) | 2023-07-25 |
Family
ID=87222635
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310428311.7A Pending CN116483082A (en) | 2023-04-20 | 2023-04-20 | Multi-agent formation recognition method based on interpolation and fuzzy recognition |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116483082A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117372946A (en) * | 2023-09-19 | 2024-01-09 | 日照市规划设计研究院集团有限公司 | Tourist group tourist behavior identification method |
-
2023
- 2023-04-20 CN CN202310428311.7A patent/CN116483082A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117372946A (en) * | 2023-09-19 | 2024-01-09 | 日照市规划设计研究院集团有限公司 | Tourist group tourist behavior identification method |
CN117372946B (en) * | 2023-09-19 | 2024-04-16 | 日照市规划设计研究院集团有限公司 | Tourist group tourist behavior identification method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110012428B (en) | Indoor positioning method based on WiFi | |
CN110490888B (en) | Highway geometric feature vectorization extraction method based on airborne laser point cloud | |
Chen et al. | Probabilistic robust hyperbola mixture model for interpreting ground penetrating radar data | |
CN108805350B (en) | Search and rescue range prediction method based on multi-dimensional Monte Carlo theory | |
CN109034066B (en) | Building identification method based on multi-feature fusion | |
CN104573705A (en) | Clustering method for building laser scan point cloud data | |
CN106651865B (en) | A kind of optimum segmentation scale automatic selecting method of new high-resolution remote sensing image | |
CN110536256B (en) | Indoor positioning method based on double-layer grids | |
CN112135248B (en) | WIFI fingerprint positioning method based on K-means optimal estimation | |
CN116483082A (en) | Multi-agent formation recognition method based on interpolation and fuzzy recognition | |
CN111294921A (en) | RSSI wireless sensor network three-dimensional cooperative positioning method | |
CN110827332A (en) | Registration method of SAR image based on convolutional neural network | |
Chen et al. | Probabilistic conic mixture model and its applications to mining spatial ground penetrating radar data | |
Zheng et al. | Background noise filtering and clustering with 3D LiDAR deployed in roadside of urban environments | |
CN115840205B (en) | Terrain area metering method and system based on laser radar technology | |
CN115131231A (en) | Complex terrain area point cloud hierarchical filtering method assisted by multi-feature clustering | |
CN115267815A (en) | Road side laser radar group optimization layout method based on point cloud modeling | |
CN115512247A (en) | Regional building damage grade assessment method based on image multi-parameter extraction | |
Chen et al. | Optimization and evaluation of multidetector deep neural network for high-accuracy Wi-Fi fingerprint positioning | |
Agapaki et al. | Scene understanding and model generation | |
CN117237902B (en) | Robot character recognition system based on deep learning | |
CN113436223A (en) | Point cloud data segmentation method and device, computer equipment and storage medium | |
CN110310294B (en) | Point cloud segmentation method using self-adaptive segmentation surface fuzzy C-means clustering | |
Javed et al. | Position Vectors Based Efficient Indoor Positioning System. | |
CN117241215A (en) | Wireless sensor network distributed node cooperative positioning method based on graph neural network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |