CN106908066A - The path planning method of the unmanned plane monitoring covering single step optimizing based on genetic algorithm - Google Patents

Abstract

The invention discloses a kind of path planning method of the unmanned plane monitoring covering single step optimizing based on genetic algorithm, its thinking is：The flown region for setting unmanned aerial vehicle group is A, it is S that the appointed task monitor area in the A of region can be flown, the yaw angle independent variable of N frame unmanned planes is defined, and sets the initial time yaw angle of N frame unmanned planes, and position coordinates matrix of the initial time N frame unmanned planes in A respectively；K ∈ { 0,1,2 ..., K }, k represent that kth walks trajectory planning, and K is the trajectory planning total step number of setting, and kth step trajectory planning is designated as into 1 single step to the step trajectory planning of kth+1；Respectively obtaining flight path position of (k+1) t N frame unmanned planes in A can flight domain, i=1,2 ..., N；And then the real-time maximum monitoring covering of the N framves unmanned plane to S of s-th single step completion is obtained, and make s plus 1；The initial value of s is 1；K is made plus 1, until k>K, obtains real-time maximum monitoring covering of the S' N framves unmanned plane of single step completion to S, realizes lasting monitoring covering of the N framves unmanned plane to S maximum magnitudes, and the S' single step walks trajectory planning for the step trajectory plannings of K 1 to K.

Description

The path planning method of the unmanned plane monitoring covering single step optimizing based on genetic algorithm

Technical field

The invention belongs to unmanned air vehicle technique field, more particularly to a kind of unmanned plane monitoring covering single step based on genetic algorithm The path planning method of optimizing, it is adaptable to realize that unmanned aerial vehicle group carries out the lasting monitoring of maximum coverage range to designated area.

Background technology

Unmanned plane (UAV) is the abbreviation of UAV (Unmanned Aerial Vehicle), and it relies on without personnel The characteristics of risk, low cost, good concealment, important application status is occupied in military and civilian receiving；And at nobody In the practical application of machine investigation, sometimes due to some particular tasks need to carry out designated area the monitoring of maximum coverage range； In order to pursue efficient application, it is necessary to cook up the reference track of unmanned plane in advance by ground command center so that unmanned plane Can be required to be flown according to reference track according to investigation.Therefore, unmanned plane covering optimizing trajectory planning technology is unmanned plane during flying The important content of task.

It is less to the general study of unmanned plane region overlay problem both at home and abroad at present, wherein with to multiple no-manned plane region overlay The research of problem is more representative；2004, Ivan Maza and Anibal Ollero proposed a kind of squad's unmanned plane region Covering trajectory planning, some sub-regions are divided into by by whole region, and the elementary path per sub-regions is shaped as " Z " word Shape, when certain frame unmanned plane fails or crashes, is planned again at once；2006, the research of Agarwal was also adopted by region and draws The thought divided, many sub-rectangular areas are divided into by flight range, are divided according to the ability of every frame unmanned plane execution covering task With region, unmanned plane is reduced to only allow the turning of 90 degree and 180 degree, but the shortcoming of this covering scheme does not consider to turn Curved radius；2010, Chen Hai et al. proposed a kind of Path Planning of Convex Polygon Domain, by covering for Convex Polygon Domain Lid trajectory planning problem is converted to the problem for seeking convex polygon width, support parallel lines when unmanned plane need to only occur along width Direction carries out the flight of " Z " font route, but min. turning radius is to zigzag course during it does not account for flight course Influence.

The method of the above region overlay trajectory planning, is directed to required track initiation point with terminal fixation mostly Situation, and general principle is all to form optimal boat by cutting zone, by obstacle avoidance, constraint oil consumption and number of turns Mark so that the covering of specific unmanned plane regional after " ox ploughs formula " flight path realizes cutting；More than study most base Flight path is planned in deterministic algorithm, this kind of algorithm itself has certain defect, when carrying out flight path rule to complex environment on a large scale When drawing, it will so that route searching occurs, and amount of calculation is excessive, inefficient, optimizing ability, it is ensured that not to trajectory planning Computational efficiency and reliability requirement.Additionally, in a practical situation, some tasks may require that unmanned plane is realized continuing to designated area The maximum coverage area of monitoring, the trajectory planning required by this aerial mission is often no fixed starting-point and terminal, For the problem, open report both domestic and external is less, and now method mentioned above can not be completely suitable for solving such boat Mark planning problem.

The content of the invention

For the deficiency that above-mentioned prior art is present, present invention aim at a kind of unmanned plane based on genetic algorithm of proposition The path planning method of monitoring covering single step optimizing, this kind is based on the flight path of the unmanned plane monitoring covering single step optimizing of genetic algorithm Planing method is the single step path planning method that a kind of unmanned aerial vehicle group monitor area based on genetic algorithm covers optimizing, by region Covering trajectory planning problem is organically combined with genetic algorithm, can efficiently solve the aerial mission of unmanned aerial vehicle group, and can be real The trajectory planning problem of the monitoring area coverage maximum of existing designated area and the fixed Origin And Destination of required flight path.

To reach above-mentioned technical purpose, the present invention is adopted the following technical scheme that and is achieved.

A kind of path planning method of the unmanned plane monitoring covering single step optimizing based on genetic algorithm, comprises the following steps：

Step 1, the flown region for setting unmanned aerial vehicle group is A, can fly the appointed task monitor area in the A of region for S, wherein Unmanned aerial vehicle group includes N frame unmanned planes, and an airborne radar is set on every frame unmanned plane, and every frame unmanned plane flies at a constant speed；

Step 2, defines the yaw angle independent variable of N frame unmanned planes, and the initial time driftage of setting N frame unmanned planes respectively Angle, and initial time N framves unmanned plane can fly the position coordinates matrix in the A of region；

Initialization：K ∈ { 0,1,2 ..., K }, k represent that kth walks trajectory planning, and K is the trajectory planning total step number of setting, k's Initial value is 0, and kth step trajectory planning is designated as into 1 single step to the step trajectory planning of kth+1；

Step 3, it is assumed that kt moment the i-th frame unmanned plane is in the flight path position that can fly in the A of regionAnd can fly in the A of region Can fly, so respectively obtain (k+1) t N framves unmanned plane can fly the flight path position in the A of region can flight domain, t represents single Step trajectory planning time interval；I=1,2 ..., N；

Step 4, according to (k+1) t N framves unmanned plane can fly in the A of region respective flight path position can flight domain, obtain Real-time maximum monitoring covering of the N framves unmanned plane that s-th single step is completed to appointed task monitor area S, and make s plus 1；S's is first Initial value is that { 1,2 ..., S'}, S' are total single step number to 1, s ∈；

Step 5, makes k plus 1, and is repeated in performing step 3 and step 4, until k>K, obtains the N that the S' single step is completed Real-time maximum monitoring covering of the frame unmanned plane to appointed task monitor area S, realizes N framves unmanned plane to appointed task monitor area The lasting monitoring covering of S maximum magnitudes, the S' single step is that K-1 walks trajectory planning to K step trajectory plannings.

Beneficial effects of the present invention are：The inventive method is by unmanned aerial vehicle group flight yaw angle variable quantityAs change certainly Amount, is specifying the moment to scout the summation of area coverage as algorithm fitness function unmanned aerial vehicle group, is asked by by trajectory planning Topic is organically combined with genetic algorithm, can solve the problem that a kind of brand-new boats different from traditional area covering optimizing trajectory planning situation Mark planning problem, i.e., do not specify the Origin And Destination of flight path, and is realized to specifying area when requiring unmanned aerial vehicle group with the track flight The maximum trajectory planning problem of the lasting monitoring coverage in domain.

Brief description of the drawings

The present invention is described in further detail with reference to the accompanying drawings and detailed description.

Fig. 1 is a kind of path planning method stream of unmanned plane monitoring covering single step optimizing based on genetic algorithm of the invention Cheng Tu；

To the coverage diagram of specified region S to be monitored, wherein white point is unmanned plane institute to 6 frame unmanned planes when Fig. 2 is 50 step The position at place, black region is 6 frame unmanned plane overlay area summations；

Fig. 3 is the trajectory planning result figure obtained using the inventive method, and wherein solid line enclosing region is 6 frame unmanned planes Can flight range A, dotted line region be specify region S to be monitored, chain-dotted line be 6 every respective flight tracks of frame unmanned plane；

Fig. 4 is the percentage change curve that unmanned aerial vehicle group monitors area coverage, wherein, abscissa is unmanned aerial vehicle group to referring to The monitoring area coverage percentage of fixed region S to be monitored, ordinate is the step number that trajectory planning is carried out using this method, and unit is Step.

Specific embodiment

Reference picture 1, is a kind of trajectory planning of unmanned plane monitoring covering single step optimizing based on genetic algorithm of the invention Method flow diagram；The path planning method of the wherein described unmanned plane monitoring covering single step optimizing based on genetic algorithm, including with Lower step：

Step 1, sets the ambient parameter of trajectory planning problem.Current problem assumes that unmanned plane can flight range what is specified Interior flight, and ensure that the persistently maximum requirement of monitoring area coverage is realized in specified investigation region that pair can be in flight range.Setting The flown region of unmanned plane is A, and it is S that can fly the appointed task monitor area in the A of region；Secondly, according to the System Computer of unmanned plane Dynamic performance, sets a series of unmanned plane kinematic parameters；Finally, the scouting radius of unmanned plane is set according to airborne radar range equation R_s。

The specific sub-step of step 1 is：

When 1.1 unmanned planes perform aerial mission, it is allowed to which the safety zone of unmanned plane during flying can fly region for unmanned plane, if nothing The man-machine region that flies for A, if flying away from the unmanned plane can fly region A, probably by the anti aircraft fire of hostile force, ground-to-air Guided missile force, directed radiation device etc. threaten hit, cause aerial mission to fail.

Unmanned plane can be flown the appointed task monitor area in the A of region and be set to S, the aerial mission requirement of trajectory planning is to this Appointed task monitor area S realizes maximum monitoring covering in real time, radar is sustainably obtained the task and specifies surveillance zone The ground potential threat target of domain S.

1.2 unmanned plane kinematic parameters are the state parameters for representing unmanned plane in ground motion or flight in the air, are passed through The state parameter determines the motion of unmanned plane, wherein the kinematic parameter related to this trajectory planning problem has：Setting unmanned plane Yaw angleFor representing the flying speed direction of unmanned plane and the angle of horizontal coordinates x-axis positive direction；Setting unmanned plane Roll angle γ, for representing unmanned plane symmetrical plane and vertical plane comprising horizontal coordinates x-axis between angle；Setting The turning angle θ of the unmanned plane and radius of turn R of unmanned plane；And an airborne radar, the airborne radar are set on unmanned plane It is both emitter and receiver.

Line between turning starting point and turning end point of the unmanned plane during turning, initially navigates with the unmanned plane To the turning angle θ that formed angle between track is unmanned plane；Unmanned plane original course is formed with the circumscribed line in the new course of unmanned plane Arc radius are the radius of turn R of unmanned plane；Unmanned plane determines the nothing due to the limitation of itself mobility by unmanned aerial vehicle (UAV) control Man-machine roll angle γ, the min. turning radius R that there is the unmanned plane in the case where the roll angle γ of unmanned plane is limited_min, and with this The min. turning radius R of unmanned plane_minCorresponding unmanned plane maximum turning angle, is designated as the maximum turning angle θ of unmanned plane_max；Nobody The turning angle θ of machine cannot be greater than the maximum turning angle θ of the unmanned plane_max, i.e. θ≤θ_max；The radius of turn R of unmanned plane cannot be less than The min. turning radius R of the unmanned plane_min, i.e. R >=R_min；Roll angle γ=30 ° of unmanned plane in the embodiment of the present invention.

It is v to set unmanned plane average flight speed_p, it is used to represent unmanned plane in single step trajectory planning time interval t Flying speed average value；Assuming that flying speed of the unmanned plane in single step trajectory planning time interval t is average in flight course Value v_pRemain constant.

1.3 aerial missions requirement unmanned plane realizes the lasting monitoring of maximum magnitude, the present invention to appointed task monitor area S The monitoring scope of single rack unmanned plane is reduced to this unmanned plane as the center of circle, with airborne radar maximum operating range R_sIt is radius Circle, the airborne radar maximum operating range R_sCan be obtained according to distance by radar equation：

Wherein, P_tThe peak power of airborne radar is represented, G represents the antenna gain of airborne radar, and λ represents that airborne radar is sent out The electromagnetic wavelength penetrated, σ represents the ground potential threat Target scatter section area in airborne radar detection range, and k' represents bohr Hereby graceful constant, T₀Normal room temperature is represented, B represents airborne radar bandwidth, and F represents that airborne radar input signal to noise ratio is believed with output end Make an uproar than ratio, L represents airborne radar own loss, and S represents airborne radar output end signal power, and N is airborne radar output Noise power, (S/N)_ominMinimum output signal-to-noise ratio required for representing airborne radar, subscript omin is represented and is asked minimum output Operation.

Step 2, will realize that the trajectory planning problem of the lasting monitoring of maximum magnitude is abstracted into a mathematics to designated area Optimization problem.First, with N frame unmanned plane yaw anglesVariable quantityAs the independent variable of trajectory planning problem；Secondly, setting The primary condition of trajectory planning problem, with vectorAnd matrix P⁰Represent the yaw angle of zero moment unmanned aerial vehicle group and area can flown The coordinate of present position in the A of domain；Finally, the fitness function and stop criterion of set algorithm, by unmanned aerial vehicle group when specified The summation for scouting area coverage is carved as fitness function, by setting greatest iteration algebraically G come termination algorithm.

Initialization：K ∈ { 0,1,2 ..., K }, k represent that kth walks trajectory planning, and K is the trajectory planning total step number of setting, k's Initial value is 0, and kth step trajectory planning is designated as into 1 single step to the step trajectory planning of kth+1.

The specific sub-step of step 2 is：

The 2.1 yaw angle independents variable for defining N frame unmanned planes Represent Yaw angle variable quantity of the i-th frame unmanned plane in single step trajectory planning time interval t,θ_maxIndicate without Man-machine maximum turning angle, N is unmanned plane main frame number, and N is the positive integer more than 0.

The primary condition of 2.2 setting trajectory planning problems：The initial time yaw angle of N frame unmanned planes is set respectively, and Initial time N framves unmanned plane can fly the position coordinates matrix in the A of region, i.e., respectively with vectorRepresent zero moment N framves nobody The course vector of machine, uses matrix P⁰Represent that zero moment N framves unmanned plane can fly the position coordinates matrix in the A of region, its expression formula Respectively：

Wherein,The yaw angle of zero moment the i-th frame unmanned plane is represented,Represent that zero moment the i-th frame unmanned plane can fly area Flight path position in the A of domain, The i-th frame unmanned plane is can flight path position in flight range A when representing zero moment X-axis coordinate,When representing zero moment the i-th frame unmanned plane can in flight range A flight path position y-axis coordinate, subscript T represents Transposition is operated.

The fitness function stop criterion of 2.3 setting single step Path Plannings：The aerial mission requirement N of this trajectory planning Frame unmanned plane realizes the lasting monitoring of maximum magnitude to appointed task monitor area S, therefore chooses N framves unmanned plane detecing at the kt moment Examine fitness function of the area coverage summation as single step Path Planning, wherein k represents that kth walks trajectory planning, k ∈ 0, 1,2 ..., K }, K is the trajectory planning total step number of setting, and the initial value of k is 0, and kth step is designated as into 1 single step to the step of kth+1； T will be designated as single step trajectory planning time interval, single step refers to that kth walks trajectory planning to the step trajectory planning of kth+1；Single step flight path is advised The fitness function stop criterion of the method for calculating is as follows：The greatest iteration algebraically G of genetic algorithm is set, when the iteration of genetic algorithm is entered Go G times, then terminated this single step trajectory planning task.

Step 3, determines feasible location.

3.1 assume that kt moment the i-th frame unmanned planes are in the flight path position that can fly in the A of regionWhen flight the i-th frame nobody The turning angle θ of machine_iWithout departing from the maximum turning angle θ of unmanned plane_maxWhen, then kt moment the i-th frame unmanned plane can fly in the A of region Flight path position can fly.According to the i-th frame unmanned plane yaw angle in single step trajectory planning time interval tVariable quantity's Difference determines a not only curved stroke, this each point not only in curved stroke can as the i-th frame of (k+1) t nobody Machine can fly the flight path position in the A of regionBecause this not only curved stroke be flight path position can flight position composition collection Close, wherein being only a little real flight path position；Therefore to simplify the process, by this not only curved stroke be approximately one section it is smooth Circular arc, is designated as the arc track of the i-th frame unmanned plane, and this is approximately rational.

3.2 from geometrical relationship, yaw angle variable quantity of the i-th frame unmanned plane in single step trajectory planning time interval tWith the turning angle θ of the i-th frame unmanned plane_iRelation be：As the maximum turning angle θ of unmanned plane_maxAt=22.5 °, Yaw angle variable quantity absolute value of the i-th frame unmanned plane in single step trajectory planning time interval t is less than or equal to 45 °, i.e.,The arc track chord length d of the i-th frame unmanned plane and the arc track of the i-th frame unmanned plane can be then obtained according to geometry derivation Arc length l, and meet d≤l≤1.02d, and it is approximately considered the arc track chord length of the i-th frame unmanned plane and the circle of the i-th frame unmanned plane Arc track arc length value is equal；Therefore, the flight path position in the A of region can flown by (k+1) t the i-th frame unmanned planeInstitute's group Into smooth circular arc be that the flight path position in the A of region can flown with kt moment the i-th frame unmanned planeFor the center of circle, with the i-th frame nobody The arc track chord length d of machine is the smooth circular arc of radius, using the smooth circular arc as (k+1) t the i-th frame unmanned plane can The flight path position flown in the A of region can flight domain.

3.3 make i take 1 to N respectively, are repeated in sub-step 3.1 and 3.2, so respectively obtain the frame of (k+1) t the 1st without It is man-machine can fly the flight path position in the A of region can flight domain to (k+1) t N framves unmanned plane can fly the flight path in the A of region Position can flight domain, be designated as (k+1) t N framves unmanned plane can fly the flight path position in the A of region can flight domain.

Step 4, the single step optimizing trajectory planning based on genetic algorithm.The yaw angle independent variable x of N frame unmanned planes is made first It is the genes of individuals of genetic algorithm, and then determines the genes of individuals linear coding scheme of genetic algorithm.Then genetic algorithm is set In population Z after the g times iteration^g, obtain constituting the population Z after the g times iteration^gL it is individual, and calculate population Z^gL Each individual fitness function value in body, evaluates population Z^gFitness, based on roulette method simulation natural selection, from Middle selection excellent individual is intersected, mutation operation, and the new population constituted to all individualities by intersection, mutation operation is again It is secondary to carry out Fitness analysis, continue to select wherein excellent individual to enter the cross and variation of next round.Circulation above procedure is until full Sufficient greatest iteration algebraically K, chooses solution of the optimum individual in now population as this single step trajectory planning；Wherein g ∈ 1, 2 ..., G }, g represents the g times iteration, and the initial value of g is 1；K ∈ { 0,1,2 ..., K }, k represent that kth walks trajectory planning, K settings Trajectory planning total step number, the initial value of k is 0, and kth step trajectory planning is designated as into single step to the step trajectory planning of kth+1.

The specific sub-step of step 4 is：

4.1 according to (k+1) t N framves unmanned planes can fly the flight path position in the A of region can flight domain, by N frame unmanned planes Yaw angle independent variable be designated asAnd to the yaw angle independent variable of N frame unmanned planesUniform enconding is carried out, is obtained by linear volume The yaw angle independent variable x of the N frame unmanned planes of code,Wherein,Represent the i-th frame Yaw angle variable quantity of the unmanned plane in single step trajectory planning time interval t；Because the i-th frame unmanned plane is in single step trajectory planning Between be spaced t in yaw angle variable quantityWith the turning angle θ of the i-th frame unmanned plane_iRelation be：Therefore, the i-th frame without The man-machine yaw angle variable quantity in single step trajectory planning time interval t∈ is represented and belonged to；In this area Interior selection decimal system linear interpolation coding method carries out uniform enconding to the yaw angle independent variable x of N frame unmanned planes, evenθ_maxRepresent the maximum turning angle of unmanned plane, rand represent one in [0,1] interval it is random Number.

4.2 set the population in genetic algorithm：After the g times iteration being calculated according to decimal system linear interpolation coding method The population Z of N frame unmanned planes when number of individuals is L^g, its expression formula is：

Wherein, x_jThe gene entrained by j-th individuality after the g times iteration is represented, Represent after the g times iteration by the i-th frame unmanned plane, j-th individuality of uniform enconding in single step trajectory planning time interval t Interior yaw angle variable quantity, gene represents yaw angle variable quantity of the N framves unmanned plane in single step trajectory planning time interval t, and L is Even number more than 0.

Fitness function in 4.3 setting genetic algorithms, per each and every one in the population Z of N frame unmanned planes when evaluating number of individuals for L The fitness of body：The population Z of N frame unmanned planes during by number of individuals after the g times iteration for L^gAs single step trajectory planning N framves nobody The variable quantity of machine yaw angle, single step refers to that kth walks trajectory planning to the step trajectory planning of kth+1, and is calculated according to relationship below Feasible location of j-th N framves unmanned plane of individuality in (k+1) t after to the g times iteration Represent after the g times iteration j-th i-th frame of individuality without The man-machine feasible location in (k+1) t, its expression formula is：

Wherein,Represent after the g times iteration j-th it is individual when the i-th frame unmanned plane machine in (k+1) t can line position The x-axis coordinate put,Represent after the g times iteration j-th it is individual when the i-th frame unmanned plane machine in (k+1) t can line position The y-axis coordinate put,The i-th frame unmanned plane is sat in the x-axis that can flying flight path position in the A of region when representing after the g times iteration kt moment Mark, v_pUnmanned plane average flight speed is represented,The yaw angle of kt moment the i-th frame unmanned plane after the g times iteration is represented,Table Show that the i-th frame unmanned plane after the g times iteration by uniform enconding, j-th individuality are inclined in single step trajectory planning time interval t Boat angle variable quantity,The i-th frame unmanned plane is sat in the y-axis that can flying flight path position in the A of region when representing after the g times iteration kt moment Mark, cos represents that complementation string is operated, and sin is represented and sought sinusoidal operation, and subscript T represents that transposition is operated.

4.4 make j take 1 to L respectively, repeat sub-step 4.3, and then respectively obtain the 1st N frame of individuality after the g times iteration Feasible location of the unmanned plane in (k+1) tThe individual N frame unmanned planes of l-th are in (k+ after to the g times iteration 1) feasible location of tL individual corresponding N framves unmanned plane is in (k+1) t after being designated as the g times iteration Feasible locationIts expression formula is：

4.5 heredity by j-th N framves unmanned plane of individuality after the g times iteration in the feasible location of (k+1) t is calculated Method fitness function is designated as Y, is specifically expressed as follows：

Function () is represented and is solved monitor area area function, and gained monitor area area function value is fitness Value；The overall area that the N frame unmanned planes of j-th individuality are monitored is designated as S_j, S_j=S_j1∪…∪S_ji∪…∪S_jN, j=1, 2 ..., L, i=1,2 ..., N, ∪ represents and seeks union operation, S_jiThe region that expression j-th individuality, the i-th frame unmanned plane are monitored, And the region S that j-th individual, the i-th frame unmanned plane is monitored_jiMeet：

Wherein,The region monitored in (k+1) step by j-th individual, the i-th frame unmanned plane after the g times iteration is in x The coordinate of axle,The region monitored in (k+1) step by j-th individual, the i-th frame unmanned plane after the g times iteration is in y-axis Coordinate, x' represents the independent variable of the genetic algorithm fitness function in x-axis of N frame unmanned planes, and y' represents the heredity calculation of N frame unmanned planes Method fitness function y-axis independent variable, R_sIt is airborne radar maximum operating range.

By feasible location of j-th N framves unmanned plane of individuality in (k+1) t after the g times iterationSubstitute into the J-th N frame unmanned plane of individuality be in the genetic algorithm fitness function of the feasible location of (k+1) t after g iteration, It is calculated j-th fitness value f of individuality after the g times iteration_j。

4.6 make j take 1 to L respectively, repeat sub-step 4.5, and then respectively obtain the 1st adaptation of individuality after the g times iteration The individual fitness value of l-th, is designated as Z after angle value to the g times iteration^gMiddle L individual corresponding fitness value, Z^gFor the g times repeatedly The population of N frame unmanned planes during for rear number of individuals for L.

4.7 individual choices are operated.

Using wheel disc bet method in Z^gExcellent individual is chosen in middle L individual corresponding fitness value, by the quantity of population L' wheel selections are carried out, concrete operations are as follows：Often wheel produces the random number in one [0,1] interval, and then obtains L random number, Using L random number as select finger, while wheel disc correspondence is divided intoBlock, n-th piece of correspondenceIt is individual selected general Rate, and n-th piece of size and theIndividual selected probability is into 1:1 ratio,Q ∈ 1,2 ..., L'},L' is total wheel number of setting,Equal with L values, L' is equal with L values.

4.71 select to determine excellent individuals using wheel discs, the in each roundIndividual selected probability is It is any one in L individuality, and repeatable is chosen.

4.72 q is taken turns after theIndividual selected probability is designated as It is any one in L individuality, and can Repeat to be chosen.

4.73 make q take 1 to L' respectively, repeat 4.72, and then respectively obtain after the 1st wheel theIndividual selected probabilityThe after to L' wheelsIndividual selected probabilityIt is designated as Z^gIn L " individual excellent individuals Represent theGene entrained by individual excellent individual；L″ Represent Z^gIn the excellent individual total number that includes；L " is equal with L' values.

The intersection and mutation operation of 4.8 genes of individuals：To Z^gIn L " individual excellent individualsThe friendship of gene is carried out successively Fork and mutation operation, Z^gIn each excellent individual entrained by gene on include N' position respectively, N' is equal with N values And correspond.

4.81 couples of Z^gIn L " individual excellent individualsCarry outWheel crossover operation, often takes turns crossover operation and produces one [0,1] Random number in interval, if jth ' take turns genetic algorithm crossover probability P of the random number more than setting that crossover operation is produced_cross, then A position pos is randomly selected on gene entrained by individual excellent individual_cross,And byIt is individual excellent Gene entrained by elegant individualityWithGene entrained by individual excellent individualTwo parts are divided into respectively, the Gene entrained by individual excellent individualPart IWithGene entrained by individual excellent individualPart IICombination, theGene entrained by individual excellent individualSecond PointWithGene entrained by individual excellent individualPart IGroup Close, after being consequently formed intersectionGene entrained by individual new individualAfter intersectionGene entrained by individual new individualIts specific crossover process is expressed as follows：

4.82 make j' take respectively 1 toRepeat 4.81 and travel through Z simultaneously^gIn L " individual excellent individuals, and then respectively obtain friendship Gene x after fork entrained by the 1st new individual_{c_1}L " the gene x entrained by individual new individual to after intersecting_{c_L}", after being designated as intersecting L " individual new individual,Gene after intersection entrained by each new individual includes N' position,L " is the even number more than 0.

" individual new individual is carried out L after 4.83 pairs of intersectionsTake turns mutation operation, L " withValue is equal, and often wheel produces one [0,1] random number in interval, if theTake turns the genetic algorithm mutation probability P that the random number for producing is not more than setting_mutation, Jth " the gene x entrained by individual new individual after then intersecting_{c_j}" as the gene x entrained by m-th individuality after cross and variation_m, m Initial value is 1, and makes m plus 1；If theTake turns genetic algorithm mutation probability P of the random number more than setting for producing_mutation, then exist Gene x entrained by the individual new individual of jth after intersection "_{c_j}" on randomly select a position pos and carry out genetic mutation, will intersect Jth " the gene x entrained by individual new individual afterwards_{c_j}" element at middle position pos is replaced withIt is [- 2 θ_max,2 θ_max] in the range of a random value, and then obtain the gene x after cross and variation entrained by the m' individuality_m', m' initial values are 1, and make m' plus 1；Its concrete operations is expressed as follows：

Jth " the gene x entrained by individual new individual after intersecting_{c_j}" the N' position for including is designated as And then the gene x after the cross and variation entrained by the m' individuality_m'For：

Wherein, pos ∈ 1,2 ..., N'},Represent jth " the gene x entrained by individual new individual after intersecting_{c_j}" in Element at pos position,Gene x after expression cross and variation entrained by the m' individuality_m'At os position of middle pth Element.

4.84 ordersTake respectively 1 toRepeat 4.83, and then respectively obtain after cross and variation entrained by the 1st individuality GeneGene entrained by L after to cross and variation " individualities" the individual Z that is designated as the L after cross and variation_mutation；Its In, L " withValue is equal and corresponds, and gene after cross and variation entrained by each individuality includes N' position.

Wherein, x_mGene entrained by " representing m after cross and variation " individuality, Represent that by m after the i-th frame unmanned plane, the cross and variation of uniform enconding " individuality is in single step trajectory planning time interval t Interior yaw angle variable quantity, m " ∈ { 1,2 ..., L " }.

4.85 simultaneously, in order to prevent roulette wheel selection from may cause not restraining for solution, is carried out using elite retention strategy Treatment, calculates L " the individual Z after cross and variation_mutationIn m " individual N frame unmanned planes in (k+1) t can Line position is put " the i-th individual frame unmanned plane that represents m In the feasible location of (k+1) t, its expression formula is：

Wherein," i-th frame unmanned plane machine is sat in the x-axis of the feasible location of (k+1) t when individual to represent m Mark,Represent m " when individual i-th frame unmanned plane machine the feasible location of (k+1) t y-axis coordinate,Represent kt The i-th frame unmanned plane can fly the x-axis coordinate of flight path position in the A of region, v during the moment_pUnmanned plane average flight speed is represented,Table Show the yaw angle of kt moment the i-th frame unmanned plane,Represent that by the i-th frame unmanned plane, the m of uniform enconding " individuality is in kth Yaw angle variable quantity in+1 step single step trajectory planning time interval t,The i-th frame unmanned plane can fly region when representing kt moment The y-axis coordinate of flight path position in A, cos represents that complementation string is operated, and sin is represented and sought sinusoidal operation, and subscript T represents that transposition is operated.

4.86 make m " taking 1 to L respectively ", repeat 4.85, and then " the individual individual Z that respectively obtains the L after cross and variation_mutationIn Feasible location of the 1st N framves unmanned plane of individuality in (k+1) tL " individualities after to cross and variation Z_mutationIn L " feasible locations of the individual N framves unmanned plane in (k+1) tAfter being designated as cross and variation Feasible location of the individual corresponding N framves unmanned planes of L " in (k+1) tIts expression formula is：

The individual Z of L after cross and variation "_mutationIn m " individual N frame unmanned planes are in the feasible of (k+1) t PositionGenetic algorithm fitness functionBe specifically expressed as follows：

Function () is represented and is solved monitor area area function, and gained monitor area area function value is fitness Value；∪ represents and seeks union operation, S_jRepresent the overall area that the N frame unmanned planes of j-th individuality are monitored, S_iRepresent j-th individuality The region that i frame unmanned planes are monitored, and the region S that the i-th frame unmanned plane is monitored_iMeet：

The region monitored by j-th individual i-th frame unmanned plane x-axis coordinate,It is j-th individual i-th frame In the coordinate of y-axis, x' represents the genetic algorithm fitness function of N frame unmanned planes in x-axis from becoming in the region that unmanned plane is monitored Amount, y' represents the independent variable of the genetic algorithm fitness function in y-axis of N frame unmanned planes, R_sIt is airborne radar maximum operating range.

4.87 by L " the individual Z after cross and variation_mutationIn m " individual N frame unmanned planes are in (k+1) t Feasible locationSubstitute into L " the individual Z after cross and variation_mutationIn m " individual N frame unmanned planes are in (k+ 1) feasible location of tGenetic algorithm fitness functionIn, it is calculated the L " individualities after cross and variation Z_mutationIn m " individual fitness value f'_m″。

4.88 make m " taking 1 to L respectively ", repeat 4.87, and then " the individual individual Z that respectively obtains the L after cross and variation_mutationIn The 1st fitness value f' of individuality₁The individual Z of L after to cross and variation "_mutationIn L " individual fitness value f'_L", Be designated as the L after cross and variation " individual corresponding fitness value f, f=[f'₁,…,f'_m″,…,f'_L″]。

By Z^g" individual corresponding fitness value is according to suitable with the L after cross and variation for middle L individual corresponding fitness value Answering angle value size carries out descending arrangement, obtains the L+L of fitness value descending arrangement, and " then individuality is arranged from fitness value descending The L+L of row " chooses preceding L individuality, and makes g plus 1, first L for choosing is individual as number of individuals after the g times iteration in individuality The population of N framves unmanned plane during for L, returns to sub-step 4.3.

4.9 repeat sub-step 4.3 to 4.8, until obtaining the kind of N frame unmanned planes when number of individuals after the G times iteration is L Group Z^G, take the population Z of N frame unmanned planes when number of individuals is L after the G times iteration^GGene entrained by the maximum individuality of middle fitness value As (k+1) t N framves unmanned plane in the yaw angle variable quantity optimal value x in single step trajectory planning time interval t_opt, Represent (k+1) t the i-th frame unmanned plane Optimal yaw angle variable quantity in single step trajectory planning time interval t, updates relational expression and is calculated the by following course (k+1) the optimal course vector of t N framves unmanned plane Table Show the yaw angle of kt moment the i-th frame unmanned plane,Represent the yaw angle of (k+1) t the i-th frame unmanned plane.

Being calculated (k+1) t N framves unmanned plane by following location updating relational expression can fly the position in the A of region Put coordinates matrix P^k+1, (k+1) t N framves unmanned plane can fly the position coordinates matrix P in the A of region^k+1It is (k+ 1) t N framves unmanned plane can fly the optimal trajectory position in the A of region.

(k+1) t N framves unmanned plane can fly the position coordinates matrix P in the A of region^k+1Expression formula be：

Wherein,(k+1) t the i-th frame unmanned plane is represented the flight path position in the A of region can be flown,Represent the (k+1) t the i-th frame unmanned plane is in the x-axis coordinate that can fly the flight path position in the A of region,Represent (k+1) t i-th Frame unmanned plane in the y-axis coordinate that can fly the flight path position in the A of region,Represent that kth t the i-th frame unmanned plane can fly region A The x-axis coordinate of interior flight path position,Represent that kth t the i-th frame unmanned plane is sat in the y-axis that can fly the flight path position in the A of region Mark, v_pUnmanned plane average flight speed is represented, cos represents that complementation string is operated, and subscript T represents that transposition is operated.

By the optimal course vector of (k+1) t N frame unmanned planesCan fly with (k+1) t N framves unmanned plane Position coordinates matrix P in the A of region^k+1, the N framves unmanned plane completed as s-th single step is to the reality of appointed task monitor area S When maximum monitoring covering, and make s plus 1；The initial value of s is that { 1,2 ..., S'}, S' are total single step number to 1, s ∈.

Step 5, makes k plus 1, and is repeated in performing step 3 and step 4, until k>K, obtains the N that the S' single step is completed Real-time maximum monitoring covering of the frame unmanned plane to appointed task monitor area S, realizes N framves unmanned plane to appointed task monitor area The lasting monitoring covering of S maximum magnitudes, the S' single step is that K-1 walks trajectory planning to K step trajectory plannings.

Specifically, the optimal course vector of (k+1) t N frame unmanned planes is usedAnd (k+1) t N framves without It is man-machine to fly the position coordinates matrix P in the A of region^k+1The single step optimizing trajectory planning of genetic algorithm is based on as next step Primary condition, the serial process in use time, using the method for step 4, next step is planned within the current step flight time Flight path position, continuously obtains the optimal trajectory position after multiple single step planning, realizes N framves unmanned plane to appointed task surveillance zone Domain S carries out the lasting monitoring of maximum magnitude.

Further checking explanation is made to effect of the present invention by following emulation experiment.

(1) simulated conditions：

Emulation is assumed to scout radius for the unmanned plane of 70km monitors that the appointed task of a piece of 200km × 200km is supervised using 6 framves Viewed area S, the flown region A where unmanned aerial vehicle group is the rectangular area of a piece of 250km × 250km, appointed task monitor area Positioned at the centre that can fly region, the course vector of 6 frame unmanned planes during zero momentAnd zero moment unmanned plane can fly region A Interior position coordinates matrix P⁰Respectively：

The flight path of each step of unmanned aerial vehicle group is all to use a kind of unmanned aerial vehicle group based on genetic algorithm proposed by the present invention to supervise Single step path planning method depending on covering optimizing, experiment gained flight path is the result for having carried out 100 step single steps planning, and detailed is imitative True parameter is referring to table 1 below：

Table 1

(2) emulation content and interpretation of result

The single step trajectory planning of covering optimizing is monitored using a kind of unmanned aerial vehicle group based on genetic algorithm proposed by the present invention Method carries out the result of 100 step trajectory plannings as shown in Fig. 2 to Fig. 4.

To the coverage diagram of specified region S to be monitored, wherein white point is unmanned plane institute to 6 frame unmanned planes when Fig. 2 is 50 step The position at place, black region is 6 frame unmanned plane overlay area summations；From Figure 2 it can be seen that the monitoring coverage of 6 frame unmanned planes can So that specified region S to be monitored is approximately completely covered.

Fig. 3 is the trajectory planning result figure obtained using the inventive method, and wherein solid line enclosing region is 6 frame unmanned planes Can flight range A, dotted line region be specify region S to be monitored, chain-dotted line be 6 every respective flight tracks of frame unmanned plane； From figure 3, it can be seen that planning gained track points are distributed in and can fly in the A of region, the track points that thus explanation this method draws are all It is effective and feasible.

Fig. 4 is the percentage change curve that unmanned aerial vehicle group monitors area coverage, wherein, abscissa is unmanned aerial vehicle group to referring to The monitoring area coverage percentage of fixed region S to be monitored, ordinate is the step number that trajectory planning is carried out using this method, and unit is Step；The flight path drawn based on the inventive method is can be seen that so that unmanned plane from the unmanned aerial vehicle group monitoring area coverage curve of Fig. 4 The area coverage percentage of group is sustainable after convergence to maintain more than 99.5%, it was demonstrated that proposed by the present invention a kind of based on heredity The single step path planning method of the unmanned aerial vehicle group monitoring covering optimizing of algorithm can realize that unmanned aerial vehicle group is carried out most to designated area The lasting monitoring of large coverage.

Obviously, those skilled in the art can carry out various changes and modification without deviating from essence of the invention to the present invention God and scope；So, if these modifications of the invention and modification belong to the scope of the claims in the present invention and its equivalent technologies Within, then the present invention is also intended to comprising these changes and modification.

Claims (7)

1. a kind of unmanned plane monitoring based on genetic algorithm covers the path planning method of single step optimizing, it is characterised in that including Following steps：

Step 1, the flown region for setting unmanned aerial vehicle group is A, can fly appointed task monitor area in the A of region for S, wherein nobody A group of planes includes N frame unmanned planes, and an airborne radar is set on every frame unmanned plane, and every frame unmanned plane flies at a constant speed；

Step 2, defines the yaw angle independent variable of N frame unmanned planes, and sets the initial time yaw angle of N frame unmanned planes respectively, with And initial time N framves unmanned plane can fly the position coordinates matrix in the A of region；

Initialization：K ∈ { 0,1,2 ..., K }, k represent that kth walks trajectory planning, and K is the trajectory planning total step number of setting, and k's is initial It is 0 to be worth, and kth step trajectory planning is designated as into 1 single step to the step trajectory planning of kth+1；

Step 3, it is assumed that kt moment the i-th frame unmanned plane is in the flight path position that can fly in the A of regionAnd can fly fly in the A of region OK, so respectively obtain (k+1) t N framves unmanned plane can fly the flight path position in the A of region can flight domain, t represent single step navigate Mark planning time is spaced；I=1,2 ..., N；

Step 4, according to (k+1) t N framves unmanned plane can fly in the A of region respective flight path position can flight domain, obtain s Real-time maximum monitoring covering of the N framves unmanned plane that individual single step is completed to appointed task monitor area S, and make s plus 1；The initial value of s It is 1, s ∈ { 1,2 ..., S ' }, S ' is total single step number；

Step 5, makes k plus 1, and is repeated in performing step 3 and step 4, until k>K, obtain N framves that the individual single steps of S ' complete without The man-machine real-time maximum monitoring covering to appointed task monitor area S, realizes N framves unmanned plane to appointed task monitor area S most Lasting monitoring covering on a large scale, the individual single steps of S ' walk trajectory planning for K-1 walks trajectory planning to K.

2. a kind of unmanned plane monitoring based on genetic algorithm as claimed in claim 1 covers the trajectory planning side of single step optimizing Method, it is characterised in that in step 1, sets an airborne radar on every frame unmanned plane, also include：

Airborne radar maximum operating range is R_s, its expression formula is：

$R_s={\[\frac{P_t{G}^2{\mathrm{\λ}}^2\mathrm{\σ}}{{\left(4\mathrm{\π}\right)}^3{k}^{\′}{T}_{0}BFL{(S/N)}_{omin}}\]}^{1/4}$

Wherein, P_tThe peak power of airborne radar is represented, G represents the antenna gain of airborne radar, and λ represents airborne radar transmitting Electromagnetic wavelength, σ represents the ground potential threat Target scatter section area in airborne radar detection range, and k ' represents Boltzmann Constant, T₀Normal room temperature is represented, B represents airborne radar bandwidth, and F represents airborne radar input signal to noise ratio and output end signal to noise ratio Ratio, L represents airborne radar own loss, and S represents airborne radar output end signal power, and N is making an uproar for airborne radar output Acoustical power, (S/N)_ominMinimum output signal-to-noise ratio required for representing airborne radar, subscript omin is represented and is asked minimum output operation.

3. a kind of unmanned plane monitoring based on genetic algorithm as claimed in claim 1 covers the trajectory planning side of single step optimizing Method, it is characterised in that the sub-step of step 2 is：

The 2.1 yaw angle independents variable for defining N frame unmanned planes Represent the i-th frame Yaw angle variable quantity of the unmanned plane in single step trajectory planning time interval t,θ_maxRepresent unmanned plane Maximum turning angle, N is unmanned plane main frame number, and N is the positive integer more than 0；

The initial time yaw angle of 2.2 setting N frame unmanned planes, and initial time N framves unmanned plane can fly the position in the A of region Coordinates matrix, i.e., respectively with vectorThe course vector of zero moment N frame unmanned planes is represented, matrix P is used⁰Represent zero moment N framves without Man-machine to fly the position coordinates matrix in the A of region, its expression formula is respectively：

$P^0={\left(\begin{array}{ccccc}{p}_1^0& \mathrm{...}& p_i^0& \mathrm{...}& p_N^0\end{array}\right)}^T,i=1,2,...,N$

Wherein,The yaw angle of zero moment the i-th frame unmanned plane is represented,Represent that zero moment the i-th frame unmanned plane can fly in the A of region Flight path position, When representing zero moment the i-th frame unmanned plane can in flight range A flight path position x-axis Coordinate,When representing zero moment the i-th frame unmanned plane can in flight range A flight path position y-axis coordinate, subscript T represents transposition Operation；

The fitness function stop criterion of 2.3 setting single step Path Plannings：This trajectory planning aerial mission requirement N framves without The man-machine lasting monitoring that maximum magnitude is realized to appointed task monitor area S, therefore choose scouting of the N framves unmanned plane at the kt moment and cover Capping accumulate summation as single step Path Planning fitness function, wherein k represent kth walk trajectory planning, k ∈ 0,1, 2 ..., K }, K is the trajectory planning total step number of setting, and the initial value of k is 0, and kth step is designated as into 1 single step to the step of kth+1；It is single The fitness function stop criterion for walking Path Planning is as follows：The greatest iteration algebraically G of genetic algorithm is set, works as genetic algorithm Iteration carried out G times, then terminate this single step trajectory planning task.

4. a kind of unmanned plane monitoring based on genetic algorithm as claimed in claim 1 covers the trajectory planning side of single step optimizing Method, it is characterised in that in step 3, obtaining (k+1) t N framves unmanned plane can in the flight path position that can fly in the A of region Flight domain, its process is：

3.1 assume that kt moment the i-th frame unmanned planes are in the flight path position that can fly in the A of regionThe i-th frame unmanned plane when flight Turning angle θ_iWithout departing from the maximum turning angle θ of unmanned plane_maxWhen, then kt moment the i-th frame unmanned plane can fly the flight path in the A of region Position can fly；

According to the i-th frame unmanned plane yaw angle in single step trajectory planning time interval tVariable quantityDetermine a smooth circle Arc, each point on the smooth circular arc of this can fly the boat in the A of region as (k+1) t the i-th frame unmanned plane Mark positionAnd this smooth arc is designated as the arc track of the i-th frame unmanned plane；

The arc track arc length l of the 3.2 arc track chord length d for calculating the i-th frame unmanned plane respectively and the i-th frame unmanned plane, and meet d ≤ l≤1.02d, and the flight path position in the A of region can flown by (k+1) t the i-th frame unmanned planeWhat is constituted is smooth Circular arc is can to fly the flight path position in the A of region with kt moment the i-th frame unmanned planeIt is the center of circle, the circular arc with the i-th frame unmanned plane Track chord length d is the smooth circular arc of radius, and then can fly area using the smooth circular arc as (k+1) t the i-th frame unmanned plane Flight path position in the A of domain can flight domain；

3.3 make i take 1 to N respectively, are repeated in sub-step 3.1 and 3.2, and then respectively obtain the frame unmanned plane of (k+1) t the 1st Can fly the flight path position in the A of region can flight domain to (k+1) t N framves unmanned plane can fly the flight path position in the A of region Can flight domain, be designated as (k+1) t N framves unmanned plane can fly the flight path position in the A of region can flight domain.

5. a kind of unmanned plane monitoring based on genetic algorithm as claimed in claim 1 covers the trajectory planning side of single step optimizing Method, it is characterised in that in step 4, the N framves unmanned plane that s-th single step is completed is to the real-time of appointed task monitor area S Maximum monitoring covering, it obtains process and is：

4.1 according to (k+1) t N framves unmanned planes can fly the flight path position in the A of region can flight domain, by the inclined of N frame unmanned planes Boat angle independent variable is designated asAnd to the yaw angle independent variable of N frame unmanned planesUniform enconding is carried out, is obtained by the N of uniform enconding The yaw angle independent variable x of frame unmanned plane, θ_maxThe maximum turning angle of unmanned plane is represented, rand represents a random number in [0,1] interval, and subscript T represents that transposition is operated；

4.2 set the population in genetic algorithm：It is calculated after the g times iteration according to decimal system linear interpolation coding method individual The population Z of N frame unmanned planes when number is for L^g, its expression formula is：

Wherein, x_jThe gene entrained by j-th individuality after the g times iteration is represented, Represent Driftage after the g times iteration by the i-th frame unmanned plane, j-th individuality of uniform enconding in single step trajectory planning time interval t Angle variable quantity, gene represents yaw angle variable quantity of the N framves unmanned plane in single step trajectory planning time interval t, and L is more than 0 Even number；

4.3 by number of individuals after the g times iteration be L when N frame unmanned planes population Z^gN frame unmanned planes as single step trajectory planning are inclined Navigate the variable quantity at angle, and single step refers to kth step to the step of kth+1, and is calculated j-th after the g times iteration according to relationship below Feasible location of the N framves unmanned plane of body in (k+1) t The feasible location of j-th i-th frame unmanned plane of individuality after the g times iteration in (k+1) t is represented, its expression formula is：

$p_{ji}^{k+1}={(x_{ji}^{k+1},y_{ji}^{k+1})}^T$

Wherein,Represent after the g times iteration j-th it is individual when the i-th frame unmanned plane machine the feasible location of (k+1) t x Axial coordinate,Represent after the g times iteration j-th it is individual when the i-th frame unmanned plane machine the feasible location of (k+1) t y Axial coordinate,The i-th frame unmanned plane can fly the x-axis coordinate of flight path position in the A of region, v when representing after the g times iteration kt moment_pTable Show unmanned plane average flight speed,The yaw angle of kt moment the i-th frame unmanned plane after the g times iteration is represented,Represent the g times Yaw angle after iteration by the i-th frame unmanned plane, j-th individuality of uniform enconding in single step trajectory planning time interval t becomes Change amount,The i-th frame unmanned plane can fly the y-axis coordinate of flight path position in the A of region, cos tables when representing after the g times iteration kt moment Show that complementation string is operated, sin is represented and sought sinusoidal operation, subscript T represents that transposition is operated；

4.4 make j take 1 to L respectively, repeat sub-step 4.3, so respectively obtain after the g times iteration the 1st N frame of individuality nobody Feasible location of the machine in (k+1) tThe individual N framves unmanned plane of l-th is in (k+1) t after to the g times iteration The feasible location at quarterL individual corresponding N framves unmanned plane is in the feasible of (k+1) t after being designated as the g times iteration PositionIts expression formula is：

$P_{possible}^{k+1}={\left(\begin{array}{ccccc}{P}_{possible\_1}^{k+1}& \mathrm{...}& P_{possible\_j}^{k+1}& \mathrm{...}& P_{possible\_L}^{k+1}\end{array}\right)}^T;$

4.5 genetic algorithm by j-th N framves unmanned plane of individuality after the g times iteration in the feasible location of (k+1) t is fitted Response function is designated as Y, is specifically expressed as follows：

$Y=function-\left(P_{possible\_j}^{k+1}\right),j=1,2,...,L$

Function () is represented and is solved monitor area area function, and gained monitor area area function value is fitness value；

By feasible location of j-th N framves unmanned plane of individuality in (k+1) t after the g times iterationSubstitute into the g times J-th N frame unmanned plane of individuality is calculated in the genetic algorithm fitness function of the feasible location of (k+1) t after iteration Obtain j-th fitness value f of individuality after the g times iteration_j；

4.6 make j take 1 to L respectively, repeat sub-step 4.5, and then respectively obtain the 1st fitness value of individuality after the g times iteration The individual fitness value of l-th, is designated as Z after to the g times iteration^gMiddle L individual corresponding fitness value, Z^gAfter the g times iteration The population of N frame unmanned planes when number of individuals is L；

4.7 use wheel disc bet method in Z^gExcellent individual is chosen in middle L individual corresponding fitness value, Z is obtained^gIn L it is " individual Excellent individual

4.8 couples of Z^gIn L " individual excellent individualsThe intersection and mutation operation of gene are carried out successively, after obtaining cross and variation The individual corresponding fitness value f of L "；

By Z^g" individual corresponding fitness value is according to fitness with the L after cross and variation for middle L individual corresponding fitness value Value size carries out descending arrangement, obtain fitness value descending arrangement L+L " individuality, then from fitness value descending arrangement L Before being chosen in+L " individualities L it is individual, and make g plus 1, the preceding L individuality that will be chosen as number of individuals after the g times iteration for L when The population of N frame unmanned planes, returns to sub-step 4.3；

4.9 repeat sub-step 4.3 to 4.8, until obtaining the population Z of N frame unmanned planes when number of individuals after the G times iteration is L^G, Take the population Z of N frame unmanned planes when number of individuals is L after the G times iteration^GGene conduct entrained by the maximum individuality of middle fitness value (k+1) t N framves unmanned plane is in the yaw angle variable quantity optimal value x in single step trajectory planning time interval t_opt, Represent that (k+1) t the i-th frame unmanned plane exists Optimal yaw angle variable quantity in single step trajectory planning time interval t；

The optimal course vector that relational expression is calculated (k+1) t N frame unmanned planes is updated by following course The yaw angle of kt moment the i-th frame unmanned plane is represented,Represent (k + 1) yaw angle of the i-th frame of t unmanned plane；

Being calculated (k+1) t N framves unmanned plane by following location updating relational expression can fly the seat of the position in the A of region Mark matrix P^k+1, (k+1) t N framves unmanned plane can fly the position coordinates matrix P in the A of region^k+1During for (k+1) t Carving N framves unmanned plane can fly the optimal trajectory position in the A of region；

(k+1) t N framves unmanned plane can fly the position coordinates matrix P in the A of region^k+1Expression formula be：

$P^{k+1}={\left(\begin{array}{ccccc}{p}_1^{k+1}& ...& p_i^{k+1}& ...& p_N^{k+1}\end{array}\right)}^T$

$p_i^{k+1}={\left(\begin{array}{cc}{x}_i^{k+1}& y_i^{k+1}\end{array}\right)}^T$

Wherein,(k+1) t the i-th frame unmanned plane is represented the flight path position in the A of region can be flown,Represent (k+1) t Moment the i-th frame unmanned plane in the x-axis coordinate that can fly the flight path position in the A of region,Represent (k+1) t the i-th frame nobody Machine in the y-axis coordinate that can fly the flight path position in the A of region,Represent that kth t the i-th frame unmanned plane can fly the boat in the A of region The x-axis coordinate of mark position,Represent kth t the i-th frame unmanned plane in the y-axis coordinate that can fly the flight path position in the A of region, v_p Unmanned plane average flight speed is represented, cos represents that complementation string is operated, and subscript T represents that transposition is operated；

By the optimal course vector of (k+1) t N frame unmanned planesCan fly region A with (k+1) t N framves unmanned plane Interior position coordinates matrix P^k+1, the N framves unmanned plane completed as s-th single step is to the real-time maximum of appointed task monitor area S Monitoring covering.

6. a kind of unmanned plane monitoring based on genetic algorithm as claimed in claim 5 covers the trajectory planning side of single step optimizing Method, it is characterised in that 4.7 process is：

4.71 select to determine excellent individuals using wheel discs, the in each roundIndividual selected probability is

$P_{\hat{l}}=\frac{function-\left(P_{possible\_\hat{l}}^{k+1}\right)}{sum\left(function\right(P_{possible}^{k+1}\left)\right)},\hat{l}\∈\{1,2,...,L\};$

4.72 q is taken turns after theIndividual selected probability is designated as It is any one in L individuality, and repeatable quilt Selection, q ∈ { 1,2 ..., L ' }, L ' are total wheel number of setting, and L ' is equal with L values；And q is taken turns the of selectionIndividuality note It isIndividual excellent individual,Initial value be 1, andThe of selection is taken turns with qIndividuality correspondence, Ran HoulingPlus 1；

4.73 make q take 1 to L ' respectively, and " individual excellent individual is remembered to repeat 4.72, and then respectively obtain the 1st excellent individual to L It is Z^gIn L " individual excellent individuals Represent theIndividual excellent individual institute The gene of carrying；L " represents Z^gIn the excellent individual total number that includes；L " is equal with L ' values.

7. a kind of unmanned plane monitoring based on genetic algorithm as described in claim 5 or 6 covers the trajectory planning of single step optimizing Method, it is characterised in that 4.8 process is：

4.81 couples of Z^gIn L " individual excellent individualsCarry outWheel crossover operation, often takes turns crossover operation and produces one [0,1] interval Interior random number, if genetic algorithm crossover probability P of the random number more than setting that jth ' wheel crossover operation is produced_cross, thenA position pos is randomly selected on gene entrained by individual excellent individual_cross,And byIndividual outstanding Gene entrained by bodyWithGene entrained by individual excellent individualTwo parts are divided into respectively, theIt is individual excellent Gene entrained by elegant individualityPart IWithGene entrained by individual excellent individual Part IICombination, theGene entrained by individual excellent individualPart IIWithGene entrained by individual excellent individualPart IGroup Close, after being consequently formed intersectionGene entrained by individual new individualAfter intersectionGene entrained by individual new individualIts specific crossover process is expressed as follows：

4.82 make j ' take respectively 1 toRepeat 4.81 and travel through Z simultaneously^gIn L " individual excellent individual, so respectively obtain intersection after Gene x entrained by 1st new individual_{c_1}L " the gene x entrained by individual new individual to after intersecting_{c_L″}, it is designated as the L after intersecting " Individual new individual,Gene after intersection entrained by each new individual includes that N ' is individual Position,L " is the even number more than 0；

" individual new individual is carried out L after 4.83 pairs of intersectionsTake turns mutation operation, L " withValue is equal, and often wheel produces one [0,1] Random number in interval, if theTake turns the genetic algorithm mutation probability P that the random number for producing is not more than setting_mutation, then will hand over Gene x entrained by the individual new individual of jth after fork "_{c_j″}As the gene x entrained by m-th individuality after cross and variation_m, m initial values It is 1, and makes m plus 1；If theTake turns genetic algorithm mutation probability P of the random number more than setting for producing_mutation, then after the intersection Gene x entrained by the individual new individual of jth "_{c_j″}On randomly select a position pos and carry out genetic mutation, jth after will intersecting " Gene x entrained by individual new individual_{c_j″}Element at middle position pos is replaced withFor

[-2θ_max,2θ_max] in the range of a random value, and then obtain the gene after cross and variation entrained by m ' individualities x_m′, m ' initial values are 1, and make m ' Jia 1；Its concrete operations is expressed as follows：

Jth " the gene x entrained by individual new individual after intersecting_{c_j″}Comprising the individual positions of N ' be designated as And then the gene x after the cross and variation entrained by m ' individualities_m′For：

Wherein, pos ∈ { 1,2 ..., N ' },Represent jth " the gene x entrained by individual new individual after intersecting_{c_j″}Middle pth os Element at position,Gene x after expression cross and variation entrained by m ' individualities_m′Unit at os position of middle pth Element；

4.84 ordersTake respectively 1 to4.83 are repeated, and then respectively obtains the gene after cross and variation entrained by the 1st individualityGene entrained by L after to cross and variation " individualities" the individual Z that is designated as the L after cross and variation_mutation；Wherein, Gene after cross and variation entrained by each individuality includes the individual positions of N ', L " withValue is equal and corresponds；

Wherein, x_m″Represent m after cross and variation " gene entrained by individuality, Represent that by m after the i-th frame unmanned plane, the cross and variation of uniform enconding " individuality is in single step trajectory planning time interval t Interior yaw angle variable quantity, m " ∈ { 1,2 ..., L " }；

4.85 calculate L " the individual Z after cross and variation_mutationIn m " individual N frame unmanned planes are in (k+1) t Feasible location Represent m " the i-th individual frame nobody In the feasible location of (k+1) t, its expression formula is machine：

$p_{m^{\′\′}i}^{k+1}={(x_{m^{\′\′}i}^{k+1},y_{m^{\′\′}i}^{k+1})}^T$

Wherein,Represent m " when individual i-th frame unmanned plane machine the feasible location of (k+1) t x-axis coordinate, Represent m " when individual i-th frame unmanned plane machine the feasible location of (k+1) t y-axis coordinate,When representing kt moment I-th frame unmanned plane can fly the x-axis coordinate of flight path position in the A of region, v_pUnmanned plane average flight speed is represented,When representing kt The yaw angle of the i-th frame unmanned plane is carved,Represent that by the i-th frame unmanned plane, the m of uniform enconding " individuality is in the step list of kth+1 Yaw angle variable quantity in step trajectory planning time interval t,The i-th frame unmanned plane can fly to be navigated in the A of region when representing kt moment The y-axis coordinate of mark position, cos represents that complementation string is operated, and sin is represented and sought sinusoidal operation, and subscript T represents that transposition is operated；

4.86 make m " taking 1 to L respectively ", repeat 4.85, and then " the individual individual Z that respectively obtains the L after cross and variation_mutationIn the 1st Feasible location of the individual N framves unmanned plane in (k+1) tThe individual Z of L after to cross and variation "_mutation In L " feasible locations of the individual N framves unmanned plane in (k+1) tIt is designated as L " each and every one after cross and variation Feasible location of the corresponding N framves unmanned plane of body in (k+1) tIts expression formula is：

$P_{possible}^{k+1}={\left(\begin{array}{ccccc}{P}_{possible\_1}^{k+1}& \mathrm{...}& P_{possible\_m^{\′\′}}^{k+1}& \mathrm{...}& P_{possible\_L^{\′\′}}^{k+1}\end{array}\right)}^T$

The individual Z of L after cross and variation "_mutationIn m " feasible locations of the individual N framves unmanned plane in (k+1) tGenetic algorithm fitness functionBe specifically expressed as follows：

$\stackrel{\‾}{Y}=function\left(P_{possible\_m^{\′\′}}^{k+1}\right)$

Function () is represented and is solved monitor area area function, and gained monitor area area function value is fitness value；

4.87 by L " the individual Z after cross and variation_mutationIn m " individual N frame unmanned planes in (k+1) t can Line position is putSubstitute into L " the individual Z after cross and variation_mutationIn m " individual N frame unmanned planes are in (k+1) t The feasible location at momentGenetic algorithm fitness functionIn, it is calculated the L " individualities after cross and variation Z_mutationIn m " individual fitness value f '_m″；

4.88 make m " taking 1 to L respectively ", repeat 4.87, and then " the individual individual Z that respectively obtains the L after cross and variation_mutationIn the 1st Individual fitness value f '₁The individual Z of L after to cross and variation "_mutationIn L " individual fitness value f '_L″, it is designated as L after cross and variation " individual corresponding fitness value f, f=[f '₁,…,f′_m″,…,f′_L″]；

4.88 by Z^g" individual corresponding fitness value is according to suitable with the L after cross and variation for middle L individual corresponding fitness value Answering angle value size carries out descending arrangement, obtains the L+L of fitness value descending arrangement, and " then individuality is arranged from fitness value descending The L+L of row " chooses preceding L individuality, and makes g plus 1, first L for choosing is individual as number of individuals after the g times iteration in individuality The population of N framves unmanned plane during for L, returns to sub-step 4.3.