CN108549402B - Unmanned aerial vehicle group task allocation method based on quantum crow group search mechanism - Google Patents
Unmanned aerial vehicle group task allocation method based on quantum crow group search mechanism Download PDFInfo
- Publication number
- CN108549402B CN108549402B CN201810224721.9A CN201810224721A CN108549402B CN 108549402 B CN108549402 B CN 108549402B CN 201810224721 A CN201810224721 A CN 201810224721A CN 108549402 B CN108549402 B CN 108549402B
- Authority
- CN
- China
- Prior art keywords
- quantum
- crow
- unmanned aerial
- task
- equal
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 43
- 239000011159 matrix material Substances 0.000 claims abstract description 29
- 238000004364 calculation method Methods 0.000 claims abstract description 15
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 14
- 238000009826 distribution Methods 0.000 claims abstract description 9
- 238000013507 mapping Methods 0.000 claims abstract description 6
- 230000006870 function Effects 0.000 claims description 56
- 238000004519 manufacturing process Methods 0.000 claims description 6
- 230000008447 perception Effects 0.000 claims description 6
- 238000009827 uniform distribution Methods 0.000 claims description 5
- 230000015572 biosynthetic process Effects 0.000 claims description 4
- 238000012854 evaluation process Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 3
- 238000013461 design Methods 0.000 abstract description 3
- 238000005457 optimization Methods 0.000 description 6
- 239000002245 particle Substances 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 206010040954 Skin wrinkling Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000002922 simulated annealing Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Images
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
- G05D1/10—Simultaneous control of position or course in three dimensions
- G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
- G05D1/104—Simultaneous control of position or course in three dimensions specially adapted for aircraft involving a plurality of aircrafts, e.g. formation flying
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)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to an unmanned aerial vehicle cluster task allocation method based on a quantum crow group search mechanism, which comprises the following steps: establishing an unmanned aerial vehicle group task distribution model from a plurality of starting points to a plurality of tasks, wherein the unmanned aerial vehicle group task distribution model comprises an unmanned aerial vehicle type number, a starting point terminal point and a distribution model; initializing a quantum crow group; carrying out fitness calculation on each quantum crow according to a fitness function, wherein the position of the quantum crow corresponding to the minimum value of the calculated fitness function is stored as a global optimal food position; updating the quantum position and position of each quantum crow; and carrying out fitness calculation on each quantum crow according to a fitness function, determining the hidden food position of each quantum crow, simultaneously finding the optimal food position so far, outputting the overall optimal food position if the maximum iteration algebra is reached, and mapping the overall optimal food position into a task allocation matrix. The invention solves the problem of solving a discrete multi-constraint objective function, designs a discrete quantum crow algorithm as an evolution strategy, and has the advantages of high convergence speed and high convergence precision.
Description
Technical Field
The invention relates to a task allocation method for an unmanned aerial vehicle cluster, in particular to a task allocation method for the unmanned aerial vehicle cluster based on a quantum crow group search mechanism, and belongs to the field of autonomous control of unmanned aerial vehicles.
Background
Unmanned Aerial vehicles (also known as Unmanned Aerial Vehicles (UAVs)) do not need to carry operators during use, provide lift through aerodynamic force, can fly remotely or autonomously under the control of a predetermined program, and carry out specific tasks through carrying task equipment. Unmanned aerial vehicle has small in size, uses advantage such as nimble, the disguise is good, strong adaptability, can accomplish certain specific work and task that some mankind can't reach and undertake under various abominable, dangerous and extreme environment. The development, production and use costs of the unmanned aerial vehicle are far lower than those of a piloted aircraft, so that the unmanned aerial vehicle has wide application space in military and civil fields.
Unmanned aerial vehicle task allocation is one of key technologies for unmanned aerial vehicle autonomous control, and is an important factor for unmanned aerial vehicle realization of intellectualization, autonomous flight and task execution. The task allocation of the unmanned aerial vehicle means that whether the unmanned aerial vehicle executes the task and what kind of task is executed are determined for the unmanned aerial vehicle through a certain task allocation method in the whole task execution process, and reasonable task allocation can ensure that the cost of the unmanned aerial vehicle is minimum on one hand and can complete various tasks best on the other hand.
Through the search of the prior art documents, it is found that the Tang Dynasty et al in electro-optic and control
The 'multi-UCAV ground attack target allocation based on game theory' published in (2011, Vol.18, No.10, pp.28-31) provides a task allocation model, and optimal task allocation is sought by using a game theory algorithm, but the algorithm model is complex, the accuracy is not high enough, and the calculation amount is large. Mehmet"applying the optimal mapping for applying target assignment by fuzzy reasoning" published in Information Sciences (2014, Vol.255, No.10, pp.28-31) solves the weapon-target assignment problem by fuzzy reasoning, but the reasoning process is complex, the calculation amount is large, and the practicability is not high. With the development of the intelligent heuristic computing technology, the intelligent optimization technology is applied to the task allocation problem of multiple unmanned aerial vehicles. The 'multi-unmanned-plane task allocation method based on the particle swarm algorithm' published by liew et al in 'control and decision' (2010, vol.25, No.9, pp.1359-1364) applies the particle swarm algorithm to the task allocation problem of the multi-unmanned plane, but the particle swarm algorithm is easy to fall into local optimization, and the convergence precision needs to be improved. The simulated annealing algorithm is calculated by 'cooperative air combat fire power distribution based on SA-DPSO hybrid optimization algorithm' proposed by Libinggan et al in aerospace science and newspaper (2014, Vol.25, No.9, pp.1626-631)The unmanned aerial vehicle task allocation is carried out by combining the method and the particle swarm algorithm, the method has better convergence speed, but is easy to fall into dimension disaster, and the optimizing performance is not enough.
Because the unmanned aerial vehicle task allocation methods are non-linear solving methods, local extrema are easy to fall into in the solving process, and the global optimal solution is difficult to obtain. The existing unmanned aerial vehicle task allocation method rarely comprehensively considers various evaluation indexes and constraints in task allocation of the unmanned aerial vehicle cluster, so that the application range of the existing unmanned aerial vehicle task allocation method is limited. Therefore, it is valuable to find a new task allocation method for improving the performance of unmanned aerial vehicle battles.
Disclosure of Invention
Aiming at the prior art, the invention provides an unmanned aerial vehicle group task allocation method based on a quantum crow group search mechanism, which considers multiple starting points and multiple end points and is suitable for the discrete problem.
In order to solve the technical problem, the invention discloses an unmanned aerial vehicle cluster task allocation method based on a quantum crow group search mechanism, which comprises the following steps of:
the method comprises the following steps: initializing the maximum iteration algebra to TmaxEstablishing a task distribution model of the unmanned aerial vehicle cluster from a plurality of starting points to a plurality of tasks: suppose that unmanned aerial vehicles of U types execute Q tasks from M starting points;
let the m-th starting point coordinate of the unmanned aerial vehicle beWherein M is more than or equal to 1 and less than or equal to M, and the coordinate of the qth task of the unmanned aerial vehicle isQ is more than or equal to 1 and less than or equal to Q, and all unmanned aerial vehicles are divided into L types according to starting points and models, wherein L is U multiplied by M, namely the unmanned aerial vehicles of the same type have the same starting points and belong to the same model;
obtaining the starting point coordinate of the first type unmanned aerial vehicle asWhere L is 1, 2.. and L, the distance between the starting point of the L-th drone and the q-th task is Dl,qAnd satisfies the following conditions:
task allocation matrix for unmanned aerial vehicle cluster L row Q column allocation matrix a ═ al,q|al,q∈{0,1}}L×QThat is, if the ith model of unmanned aerial vehicle executes the qth task, al,q1, otherwise al,q=0;
Assuming that each drone has D weapons, the probability that the No. l drone uses the No. D weapon isThe manufacturing cost of the No. l unmanned plane for using the No. d weapon isl,dWherein D is more than or equal to 1 and less than or equal to D, L is more than or equal to 1 and less than or equal to L, and the killing rate of the D-th weapon to the q-th task isWherein D is more than or equal to 1 and less than or equal to D, Q is more than or equal to 1 and less than or equal to Q, and a damage probability matrix P of the unmanned aerial vehicle is set as { P ═ Pl,q,d|Pl,q,d∈[0,1]}L×Q×D,Pl,q,dIs the damage degree of the No. l unmanned plane to the No. q task by using the No. d weapon and meets the following conditions:the damage threshold of the q-th task is WqWherein Q is more than or equal to 1 and less than or equal to Q, and the value of the qth task is set as VqAnd the number of the first type unmanned planes is BlAnd the maximum formation number of unmanned aerial vehicles attacking the qth task is CqThe maximum range of the No. l unmanned plane is RlThe maximum range of all unmanned aerial vehicles is OmaxThe flight speed of the No. l unmanned plane is ZlThe maximum flight time of all unmanned aerial vehicles is Zmax;
The unmanned aerial vehicle task allocation model is respectively represented by a target value gain function, a flight distance function, a consumption volume cost function and a target coverage rate function:
(1) the normalized target value revenue function is:where A is the task allocation matrix, and A ═ al,q|al,q∈{0,1}}L×Q,Pl,q,dThe damage probability of the q task by using d weapons for the type I unmanned aerial vehicle, d is the type of the weapons, the type of the weapons used by the type I unmanned aerial vehicle needs to be set in advance, VqThe value of the qth task, N the number of drones actually participating in the task allocation,max is a function for solving the maximum value for the maximum task value;
(2) the normalized flight distance function is:wherein λ1,λ2Is a weight of two factors, λ1+λ2=1,λ1,λ2≥0,Is the length of the longest path and,Rlthe maximum range of the type I unmanned aerial vehicle;
(3) the normalized cost of draw function is:where the type of weapon used by the type l drone needs to be set in advance,maxin order to be the most cost-intensive,
the task allocation model of the unmanned aerial vehicle group meets the following constraint conditions:
(1) and (4) mission force constraint: the number of missions of each type of drone cannot exceed the number of drones with that type,
(2) unmanned aerial vehicle combat radius constraint: ensure that the flight distance of the unmanned aerial vehicle is within the operation radius, al,q×(Dl,q-Rl)≤0(l=1,2,...,L;q=1,2,...,Q);
(3) Constraint on the damage degree of the target: the damage degree of the unmanned aerial vehicle executing the task q to the task q is not less than the damage degree threshold value of the task,Pl,q,dthe damage probability of the ith unmanned plane to the qth task by using d weapons is given, d is the type of the weapon used, and the type of the weapon used by the ith unmanned plane needs to be set in advance;
(4) constraint on the number of drones attacking the target: the number of drones attacking the qth task does not exceed its maximum number of convoy, i.e.
(5) Constraint of the range of the attack target: i.e. the range of the attack mission does not exceed a given maximum range,Omaxthe maximum range of all unmanned aerial vehicles;
(6) time constraints of the attack targets: the time of the attack task does not exceed a given maximum time,Zmaxthe maximum flight time of all unmanned aerial vehicles;
determining population specifications of quantum crow groupModulo K, dimension J of optimization problem is L multiplied by Q, task allocation matrix A of unmanned aerial vehicle group is arranged according to increasing L and Ql,q|al,q∈{0,1}}L×QThe elements in (1) areCorrespondingly recording elements in the task allocation matrix A of the unmanned aerial vehicle group;
step two: initializing a quantum crow group:
quantum position of ith quantum of crowEach dimension of (a) is set asWherein i is more than or equal to 1 and less than or equal to K, J is more than or equal to 1 and less than or equal to J, and measuring the quantum position of the ith quantum crow to obtain the position of the ith quantum crowInitializing the hidden food position of the ith quantum crowWherein i is more than or equal to 1 and less than or equal to K, t is iteration times, and t is set to be 0 at the beginning;
dimension j of quantum position of ith quantum crowMeasuring to obtain the jth dimension of the position of the ith quantum crowWherein i is more than or equal to 1 and less than or equal to K, J is more than or equal to 1 and less than or equal to J,is a random number satisfying uniform distribution;
step three: carrying out fitness calculation on each quantum crow according to the fitness function, and calculating the corresponding quantity of the minimum value of the fitness functionThe location of the crow's feet is stored as the global optimum food location
Step four: updating the quantum position and position of each quantum crow:
the ith quantum crow randomly selects another quantum crow in the quantum crow group, and then follows the quantum crow to discover the food position hidden by the quantum crow, the perception probability that the quantum crow discovers to be followed is mu, if the perception probability is mu, the quantum crow is not hidden by the quantum crowQuantum crow i updates the quantum position through a strategy 1, otherwise, quantum crow i updates the quantum position through a strategy 2;
policy 2 satisfies: the ith quantum crow updates the position according to the food position hidden by the ith quantum crow and the optimal food position, and the update equation of the j-th dimension quantum rotation angle of the ith quantum crow isWherein e2,e3Determining the influence degree of the position of the quantum crow on the evolution of the quantum crow as a constant;
the evolution process of the quantum position is as follows:
where ζ is 0.15/J, the variation probability, and abs () is the function of the absolute value;
measuring the quantum position of the quantum crow to obtain the position of the quantum crow, wherein the measurement rule is as follows:
wherein i is more than or equal to 1 and less than or equal to K, J is more than or equal to 1 and less than or equal to J,is a random number satisfying uniform distribution;
step five: carrying out fitness calculation on each quantum crow according to a fitness function, determining a hidden food position of each quantum crow, and finding out an optimal food position of the current iteration algebra;
the position of the ith quantum of crowAssign a value to the task assignment matrix A in accordance withCalculating the fitness;
Step six: if the maximum iteration algebra T is reachedmaxThe algorithm is terminated, and step seven is executed; otherwise, making t equal to t +1, and returning to the fourth step to continue the operation;
step seven: and outputting the global optimal food position, and mapping the global optimal food position into a task allocation matrix.
The invention relates to an unmanned aerial vehicle cluster task allocation method based on a quantum crow group search mechanism, which further comprises the following steps:
the fitness evaluation process in the third step is as follows: firstly, the position of ith quantum crow of the t generationAssigning to a task allocation matrix A, wherein the jth generation is the jth dimension of the ith quantum crowIs assigned to al,qAccording toPerforming fitness calculation, wherein
c1,c2,c3,c4,c5,c6As a penalty factor, ω1,ω2,ω3,ω4As a weighting factor, ω1+ω2+ω3+ω4=1,0≤ω1,ω2,ω3,ω4And max is less than or equal to 1, the maximum function is obtained, and min is the minimum function.
The invention has the beneficial effects that: the invention provides an unmanned aerial vehicle cluster task allocation model considering multiple starting points and multiple end points, and simultaneously provides a quantum crow group search mechanism suitable for a discrete problem for solving the task allocation problem of the unmanned aerial vehicle cluster, aiming at the defects of the existing unmanned aerial vehicle cluster task allocation method. Compared with the prior art, the invention fully considers the situation that a plurality of tasks are executed from a plurality of starting points in the task allocation process of the unmanned aerial vehicle group, and simultaneously considers a plurality of targets of a target value gain function, a flight distance function, a bomb consumption cost function and a target coverage rate function, and has the following advantages:
(1) the invention solves the problem of solving discrete multi-constraint objective functions, designs a novel discrete quantum crow algorithm as an evolution strategy, and processes different objective functions by utilizing linear weight.
(2) Compared with the existing unmanned aerial vehicle group task allocation method, the method can effectively solve the multi-target constraint requirement on the unmanned aerial vehicle group, and has wider applicability.
(3) Simulation results show that the unmanned aerial vehicle group task allocation method provided by the invention can obtain a more reasonable unmanned aerial vehicle task allocation scheme than a Particle Swarm Optimization (PSO) method, thereby demonstrating the effectiveness of the method.
Drawings
FIG. 1 is a flow chart of task allocation for an unmanned aerial vehicle fleet;
FIG. 2 is a flow chart of a quantum crow location update;
fig. 3 is a convergence curve for implementing multi-drone task allocation by two methods.
Detailed Description
The following describes embodiments of the present invention with reference to the drawings.
As shown in fig. 1, the technical scheme of the invention comprises the following steps:
the method comprises the following steps: and establishing a task allocation model of the unmanned aerial vehicle group from a plurality of starting points to a plurality of tasks, and assuming that the unmanned aerial vehicles with U types execute Q tasks from M starting points.
Let the m-th starting point coordinate of the unmanned aerial vehicle beWherein M is more than or equal to 1 and less than or equal to M, and the coordinate of the qth task of the unmanned aerial vehicle isWherein Q is more than or equal to 1 and less than or equal to Q. All drones can be divided into L types according to their starting point and model, where L is U × M, i.e. the same type of drone has the same starting point and belongs to the same model.
According to the starting point of the model of the type I unmanned aerial vehicle, the starting point coordinate of the type I unmanned aerial vehicle can be obtained asWhere L1, 2.. and L, the distance D between the starting point of the L-th drone and the q-th missionl,qIs composed of
The task allocation matrix of the unmanned aerial vehicle cluster can be an L-row Q-column allocation matrix A ═ al,q|al,q∈{0,1}}L×QThat is, if the ith model of unmanned aerial vehicle executes the qth task, al,q1, otherwise al,q=0。
Assuming that each drone has D weapons, the probability that the No. l drone uses the No. D weapon isThe manufacturing cost of the No. l unmanned plane for using the No. d weapon isl,dWherein D is more than or equal to 1 and less than or equal to D, and L is more than or equal to 1 and less than or equal to L. The kill rate of the qth weapon on the qth task isWherein D is more than or equal to 1 and less than or equal to D, Q is more than or equal to 1 and less than or equal to Q, and a damage probability matrix P of the unmanned aerial vehicle is set as { P ═ Pl,q,d|Pl,q,d∈[0,1]}L×Q×D,Pl,q,dThe damage degree of the No. l unmanned plane to the No. q task by using the No. d weapon isThe damage threshold of the q-th task is WqWherein Q is more than or equal to 1 and less than or equal to Q. Let the value of the qth task be VqAnd the number of the first type unmanned planes is BlAnd the maximum formation number of unmanned aerial vehicles attacking the qth task is CqThe maximum range of the No. l unmanned plane is RlThe maximum range of all unmanned planes is OmaxThe flight speed of the No. l unmanned plane is ZlThe maximum flight time of all unmanned planes is Zmax。
The unmanned aerial vehicle task allocation model may be represented by a target value gain function, a flight distance function, a volume cost function, and a target coverage function.
(1) The normalized target value revenue function is:where A is the task allocation matrix, and A ═ al,q|al,q∈{0,1}}L×Q,Pl,q,dThe damage probability of the qth task by using d weapons for the type I unmanned aerial vehicle is determined, d is the type of the weapons used, and the type of the weapons used by the type I unmanned aerial vehicle needs to be set in advance. VqThe value of the qth task, N the number of drones actually participating in the task allocation, max is the function of finding the maximum value for the maximum mission value.
(2) The normalized flight distance function is:wherein λ1,λ2Is a weight of two factors, λ1+λ2=1,λ1,λ2≥0,Is the length of the longest path and,Rlthe maximum range of the No. l unmanned plane.
(3) The normalized cost of draw function is:where the type of weapon used by the type l drone needs to be set in advance,maxin order to be the most cost-intensive,
in addition, the following constraints should be satisfied:
(1) and (5) performing military force restriction on the task. I.e. the number of moves of each type of drone cannot exceed the number of drones having that type.
(2) Unmanned aerial vehicle radius of war restraint. I.e. it must be guaranteed that the flight distance of the drone is within its operational radius. a isl,q×(Dl,q-Rl)≤0(l=1,2,...,L;q=1,2,...,Q)。
(3) And (4) constraint on the damage degree of the target. Namely, the damage degree of the unmanned aerial vehicle executing the task q to the task q is not less than the damage degree threshold value of the task.Pl,q,dThe damage probability of the qth task by using d weapons for the type I unmanned aerial vehicle is determined, d is the type of the weapons used, and the type of the weapons used by the type I unmanned aerial vehicle needs to be set in advance.
(4) A constraint on the number of drones attacking the target. The number of drones attacking the qth task does not exceed its maximum number of convoy, i.e.
(5) The constraints of the range of the attack target. I.e. the flight of the attack mission does not exceed a given maximum flight.OmaxThe maximum range of all drones.
(6) Time of attack target. I.e. the time of the attack task does not exceed a given maximum time.ZmaxThe maximum flight time for all drones.
Then, determining the population scale K of the quantum crow group, optimizing the dimension J of the problem to be L multiplied by Q, and arranging the task allocation matrix A of the unmanned aerial vehicle group to be { a ] according to the mode that L is gradually increased and Q is gradually increasedl,q|al,q∈{0,1}}L×QThe elements in (1) areAnd correspondingly recording elements in the unmanned aerial vehicle group task allocation matrix A.
Step two: initializing the quantum crow group.
Quantum position of ith quantum of crowEach dimension of (a) is set asWherein i is more than or equal to 1 and less than or equal to K, J is more than or equal to 1 and less than or equal to J, and the quantum position of the ith quantum crow is measured to obtain the position of the ith quantum crowInitializing the hidden food position of the ith quantum crowWherein i is more than or equal to 1 and less than or equal to K. t is the iteration number, and t is 0 at the beginning.
Dimension j of quantum position of ith quantum crowMeasuring to obtain the jth dimension of the position of the ith quantum crowWherein i is more than or equal to 1 and less than or equal to K, J is more than or equal to 1 and less than or equal to J,is to satisfy the uniform distributionRandom number of cloth.
Step three: carrying out fitness calculation on each quantum black-crow according to a fitness function, wherein the position of the quantum black-crow corresponding to the minimum value of the calculated fitness function is stored as the global optimal food position
The fitness evaluation process is as follows:
firstly, the position of ith quantum crow of the t generationAssigning to a task allocation matrix A, wherein the jth generation is the jth dimension of the ith quantum crowIs assigned to al,q. According toPerforming fitness calculation, wherein
c1,c2,c3,c4,c5,c6As a penalty factor, ω1,ω2,ω3,ω4As a weighting factor, ω1+ω2+ω3+ω4=1,0≤ω1,ω2,ω3,ω4And max is less than or equal to 1, the maximum function is obtained, and min is the minimum function.
Step four: and updating the quantum position and position of each quantum crow.
As shown in fig. 2, the ith quantum crow randomly selects another quantum crow in the quantum crow group, and then follows the quantum crow to find the food position hidden by the quantum crow. The perception probability that quantum crow s finds to be followed is mu. If it isAnd (3) updating the quantum position of the quantum crow i through a strategy 1, or updating the quantum position of the quantum crow i through a strategy 2.
Strategy 1: and the ith quantum crow updates the position according to the food position hidden by the quantum crow. The j dimension quantum rotation angle of the ith quantum crow has the updated equation ofWherein e1And determining the influence degree of the position of the quantum crow on the evolution of the quantum crow as a constant, wherein H is the flight length.
Strategy 2: and the ith quantum crow updates the position according to the food position hidden by the ith quantum crow and the optimal food position. The j dimension quantum rotation angle of the ith quantum crow has the updated equation ofWherein e2,e3The number of the quantum crow is constant, and the influence degree of guiding the position of the quantum crow on the evolution of the quantum crow is determined.
The evolution process of the quantum position is as follows:
where ζ is 0.15/J, the mutation probability, and abs () is the absolute value function.
And measuring the quantum position of the quantum crow to obtain the position of the quantum crow. The measurement rules are as follows:
wherein i is more than or equal to 1 and less than or equal to K, J is more than or equal to 1 and less than or equal to J,is a random number satisfying uniform distribution.
Step five: and carrying out fitness calculation on each quantum crow according to a fitness function, determining the hidden food position of each quantum crow, and simultaneously finding the optimal food position so far.
The position of the ith quantum of crowAnd assigning to the task allocation matrix A. According toAnd calculating the fitness.
Step six: if the maximum iterative algebra is reached, the algorithm is terminated, and a seventh step is executed; otherwise, let t be t +1, return to step four and continue to go on.
Step seven: and outputting the global optimal food position, and mapping the global optimal food position into a task allocation matrix.
The specific embodiment is as follows:
the model parameters are set as follows:
the number U of the unmanned aerial vehicle is 4, the number M of the starting points of the unmanned aerial vehicle is 3, the coordinates of the starting points are (368,319,150), (264,44,264) and (296,242,347.5), the number Q of the tasks of the unmanned aerial vehicle is 10, the coordinate of the 1 st task is (264,715,800), the task value is 5, and the damage threshold is 0.5; the coordinate of the 2 nd task is (225,605,670), the task value is 5, and the damage threshold values are all 0.5; the coordinate of the 3 rd task is (168,538,340), the task value is 2, and the damage threshold values are all 0.5; the 4 th task has the coordinate of (180,455,670), the task value of 1 and the damage threshold of 0.5; sit on task 5Labeled (120,400,600), task value is 2, damage threshold is 0.5; the coordinate of the 6 th task is (96,304,233), the task value is 5, and the damage threshold values are all 0.5; the coordinate of the 7 th task is (10,451,233), the task value is 5, and the damage threshold values are all 0.5; the coordinate of the 8 th task is (162,660,233), the task value is 5, the damage threshold values are all 0.5, the coordinate of the 9 th task is (110,561,45), the task value is 5, and the damage threshold values are all 0.5; the 10 th task has coordinates (105,473,1830), a task value of 5, and a damage threshold of 0.5. The unmanned aerial vehicle weapon type D is 2, wherein the unmanned aerial vehicles of the 1 st type and the 2 nd type use the 2 nd weapon, the unmanned aerial vehicles of the 3 rd type and the 4 th type use the 1 st weapon, the manufacturing cost of the 1 st weapon is 5 units, the manufacturing cost of the 2 nd weapon is 3 units, the probability that the unmanned aerial vehicle of the 1 st type selects the 1 st weapon is 0.67, and the probability that the 2 nd weapon is 0.78; the probability of selecting the 1 st weapon by the unmanned plane of the 2 nd model is 0.67, and the probability of selecting the 2 nd weapon is 0.78; the probability of selecting the 1 st weapon by the unmanned plane of the 3 rd model is 0.92, and the probability of selecting the 2 nd weapon is 0.92; the probability of selecting a weapon of type 1 for a drone of type 4 is 0.92, and the probability of selecting a weapon of type 2 is 0.92. The kill rate of the 1 st weapon to the 1 st and 2 nd tasks is 0.92; the kill rate of the 1 st weapon to the 3 rd task, the 4 th task and the 5 th task is 0.8; the kill rate of the 1 st weapon on the 6 th task, the 7 th task and the 8 th task is 0.94; the kill rate for type 1 weapons on the 9 th and 10 th tasks was 0.6. The killing rate of the 2 nd weapon on the 1 st task, the 2 nd task, the 3 rd task, the 4 th task and the 5 th task is 0.8; the kill rate of the 2 nd weapon on the 6 th task, the 7 th task and the 8 th task is 0.92; the kill rate of the 2 nd weapon on the 9 th task was 0.97; the kill rate of the 2 nd weapon on the 10 th task was 0.6. The number of type 1 drones is 5, the maximum range is 300, the number of type 2 drones is 6, and the maximum range is 900. The number of type 3 drones is 6, and the maximum range is 900. The number of type 4 drones is 15, and the maximum range is 1700. The number of the 5 th type unmanned aerial vehicles is 3, and the maximum range is 300. The number of the type 6 unmanned aerial vehicles is 5, the maximumThe voyage is 900. The number of type 7 drones is 6, and the maximum range is 900. The number of type 8 drones is 4 and the maximum range is 1700. The number of type 9 drones is 5, and the maximum range is 300. The number of 10 th type unmanned planes is 10, and the maximum range is 900. The number of type 11 drones is 5, and the maximum range is 900. The number of type 12 drones is 10 and the maximum range is 1700. The maximum formation number of unmanned aerial vehicles attacking the task is 8. Weight λ1=1,λ 20, objective function weight ω1=0.322,ω2=0.214,ω3=0.1856,ω40.2784. Penalty factor c1=c2=c3=c6=50,c4=c 50. The unit of the coordinates and the voyage is km.
The parameter setting of the unmanned aerial vehicle group task allocation method based on the quantum crow group search mechanism is as follows: the population scale K is 20, the maximum iteration number is 200, the perception probability mu is 0.1, and the influence degree e on the quantum crow evolution1=0.06,e2=0.03,e3The flight length H is 0.01 and 2.
The parameter setting of the unmanned aerial vehicle group task allocation method based on the particle swarm optimization is disclosed in 'control and decision' (2010, Vol.25No.9, pp.1359-1364) of Li Wei et al, and other parameters are the same as those of the unmanned aerial vehicle group task allocation method based on the quantum crow group search mechanism.
As shown in fig. 3, under the above parameter setting condition, the convergence curves of the multi-drone task allocation are realized for the two methods, and the present invention has a faster convergence effect.
The results of the unmanned aerial vehicle group task allocation method based on the quantum crow group search mechanism are shown in the table:
table 1 model assignment of corresponding tasks of unmanned aerial vehicle at each starting point
Where M1 denotes the first start point, M2 denotes the first start point, and M3 denotes the third start point. Q1 through Q10 represent the 1 st through 10 th tasks, respectively. U1 represents a model 1 drone, U2 represents a model 2 drone, U3 represents a model 3 drone, U4 represents a model 4 drone, and 0 represents no drone performing the task from this starting point.
The invention solves the problems that the traditional algorithm has low searching speed and large calculation amount, is difficult to find the optimal task allocation of the unmanned aerial vehicle group, and the existing unmanned aerial vehicle group task allocation design based on intelligent calculation rarely comprehensively considers various evaluation indexes and constraints, so that the application range is limited. A task allocation model considering the unmanned aerial vehicle cluster is provided, and a discrete quantum crow group searching mechanism is provided for solving the task allocation problem of the unmanned aerial vehicle cluster. The steps of the method are required to be: firstly, establishing an unmanned aerial vehicle group task distribution model from a plurality of starting points to a plurality of tasks, wherein the unmanned aerial vehicle group task distribution model comprises an unmanned aerial vehicle model number, a starting point terminal point and a distribution model. And secondly, initializing the quantum crow group. And thirdly, calculating the fitness of each quantum crow according to the fitness function, wherein the position of the quantum crow corresponding to the minimum value of the calculated fitness function is stored as the global optimal food position. And fourthly, updating the quantum position and the position of each quantum of the crow. And fifthly, calculating the fitness of each quantum crow according to a fitness function, determining the hidden food position of each quantum crow, finding the optimal food position so far, outputting the global optimal food position if the maximum iterative algebra is reached, and mapping the global optimal food position into a task allocation matrix.
Claims (2)
1. A quantum crow group search mechanism-based unmanned aerial vehicle cluster task allocation method is characterized by comprising the following steps:
the method comprises the following steps: initializing the maximum iteration algebra to TmaxEstablishing a task distribution model of the unmanned aerial vehicle cluster from a plurality of starting points to a plurality of tasks: suppose that unmanned aerial vehicles of U types execute Q tasks from M starting points;
let the m-th starting point coordinate of the unmanned aerial vehicle beWherein M is more than or equal to 1 and less than or equal to M, and the coordinate of the qth task of the unmanned aerial vehicle isQ is more than or equal to 1 and less than or equal to Q, and all unmanned aerial vehicles are divided into L types according to starting points and models, wherein L is U multiplied by M, namely the unmanned aerial vehicles of the same type have the same starting points and belong to the same model;
obtaining the starting point coordinate of the first type unmanned aerial vehicle asWhere L is 1, 2.. and L, the distance between the starting point of the L-th drone and the q-th task is Dl,qAnd satisfies the following conditions:
task allocation matrix for unmanned aerial vehicle cluster L row Q column allocation matrix a ═ al,q|al,q∈{0,1}}L×QThat is, if the ith model of unmanned aerial vehicle executes the qth task, al,q1, otherwise al,q=0;
Assuming that each drone has D weapons, the probability that the No. l drone uses the No. D weapon isThe manufacturing cost of the No. l unmanned plane for using the No. d weapon isl,dWherein D is more than or equal to 1 and less than or equal to D, L is more than or equal to 1 and less than or equal to L, and the killing rate of the D-th weapon to the q-th task isWherein D is more than or equal to 1 and less than or equal to D, Q is more than or equal to 1 and less than or equal to Q, and a damage probability matrix P of the unmanned aerial vehicle is set as { P ═ Pl,q,d|Pl,q,d∈[0,1]}L×Q×D,Pl,q,dIs the damage degree of the No. l unmanned plane to the No. q task by using the No. d weapon and meets the following conditions:the damage threshold of the q-th task is WqWherein Q is more than or equal to 1 and less than or equal to Q, and the value of the qth task is set as VqAnd the number of the first type unmanned planes is BlAnd the maximum formation number of unmanned aerial vehicles attacking the qth task is CqThe maximum range of the No. l unmanned plane is RlThe maximum range of all unmanned aerial vehicles is OmaxThe flight speed of the No. l unmanned plane is ZlThe maximum flight time of all unmanned aerial vehicles is Zmax;
The unmanned aerial vehicle task allocation model is respectively represented by a target value gain function, a flight distance function, a consumption volume cost function and a target coverage rate function:
(1) the normalized target value revenue function is:where A is the task allocation matrix, and A ═ al,q|al,q∈{0,1}}L×Q,Pl,q,dThe damage probability of the q task by using d weapons for the type I unmanned aerial vehicle, d is the type of the weapons, the type of the weapons used by the type I unmanned aerial vehicle needs to be set in advance, VqThe value of the qth task, N the number of drones actually participating in the task allocation,max is a function for solving the maximum value for the maximum task value;
(2) the normalized flight distance function is:wherein λ1,λ2Is a weight of two factors, λ1+λ2=1,λ1,λ2≥0,Is the length of the longest path and,Rlthe maximum range of the type I unmanned aerial vehicle;
(3) the normalized cost of draw function is:where the type of weapon used by the type l drone needs to be set in advance,maxin order to be the most cost-intensive,
the task allocation model of the unmanned aerial vehicle group meets the following constraint conditions:
(1) and (4) mission force constraint: the number of missions of each type of drone cannot exceed the number of drones with that type,
(2) unmanned aerial vehicle combat radius constraint: ensure that the flight distance of the unmanned aerial vehicle is within the operation radius, al,q×(Dl,q-Rl)≤0(l=1,2,...,L;q=1,2,...,Q);
(3) Constraint on the damage degree of the target: the damage degree of the unmanned aerial vehicle executing the task q to the task q is not less than the damage degree threshold value of the task,Pl,q,dthe damage probability of the ith unmanned plane to the qth task by using d weapons is given, d is the type of the weapon used, and the type of the weapon used by the ith unmanned plane needs to be set in advance;
(4) constraint on the number of drones attacking the target: the number of drones attacking the qth task does not exceed its maximum number of convoy, i.e.
(5) Constraint of the range of the attack target: i.e. the range of the attack mission does not exceed a given maximum range,Omaxthe maximum range of all unmanned aerial vehicles;
(6) time constraints of the attack targets: the time of the attack task does not exceed a given maximum time,Zmaxthe maximum flight time of all unmanned aerial vehicles;
determining the population scale K of the quantum crow group, optimizing the dimension J of the problem to be L multiplied by Q, arranging the task allocation matrix A of the unmanned aerial vehicle group to be a in a mode of increasing L and Q progressivelyl,q|al,q∈{0,1}}L×QThe elements in (1) areCorrespondingly recording elements in the task allocation matrix A of the unmanned aerial vehicle group;
step two: initializing a quantum crow group:
quantum position of ith quantum of crowEach dimension of (a) is set asWherein i is more than or equal to 1 and less than or equal to K, J is more than or equal to 1 and less than or equal to J, and measuring the quantum position of the ith quantum crow to obtain the position of the ith quantum crowInitializing the hidden food position of the ith quantum crowWherein i is more than or equal to 1 and less than or equal to K, t is iteration times, and t is set to be 0 at the beginning;
dimension j of quantum position of ith quantum crowMeasuring to obtain the jth dimension of the position of the ith quantum crowWherein i is more than or equal to 1 and less than or equal to K, J is more than or equal to 1 and less than or equal to J,is a random number satisfying uniform distribution;
step three: carrying out fitness calculation on each quantum black-crow according to a fitness function, wherein the position of the quantum black-crow corresponding to the minimum value of the calculated fitness function is stored as the global optimal food position
Step four: updating the quantum position and position of each quantum crow:
the ith quantum crow randomly selects another quantum crow in the quantum crow group, and then follows the quantum crow to discover the food position hidden by the quantum crow, the perception probability that the quantum crow discovers to be followed is mu, if the perception probability is mu, the quantum crow is not hidden by the quantum crowQuantum crow i updates the quantum position through a strategy 1, otherwise, quantum crow i updates the quantum position through a strategy 2;
strategy 1 satisfies: the ith quantum crow updates the position according to the food position hidden by the quantum crow, and the update equation of the j-th dimension of the ith quantum rotary angle of the ith quantum crow isWherein e1Is a constant, blockDetermining the influence degree of the position for guiding the quantum crow on the evolution of the quantum crow, wherein H is the flight length;
policy 2 satisfies: the ith quantum crow updates the position according to the food position hidden by the ith quantum crow and the optimal food position, and the update equation of the j-th dimension quantum rotation angle of the ith quantum crow isWherein e2,e3Determining the influence degree of the position of the quantum crow on the evolution of the quantum crow as a constant;
the evolution process of the quantum position is as follows:
where ζ is 0.15/J, the variation probability, and abs () is the function of the absolute value;
measuring the quantum position of the quantum crow to obtain the position of the quantum crow, wherein the measurement rule is as follows:
wherein i is more than or equal to 1 and less than or equal to K, J is more than or equal to 1 and less than or equal to J,is a random number satisfying uniform distribution;
step five: carrying out fitness calculation on each quantum crow according to a fitness function, determining a hidden food position of each quantum crow, and finding out an optimal food position of the current iteration algebra;
the position of the ith quantum of crowAssign a value to the task assignment matrix A in accordance withCalculating the fitness;
wherein:
c1,c2,c3,c4,c5,c6as a penalty factor, ω1,ω2,ω3,ω4As a weighting factor, ω1+ω2+ω3+ω4=1,0≤ω1,ω2,ω3,ω4Max is less than or equal to 1, max is a function for solving the maximum value, and min is a function for solving the minimum value;
Step six: if the maximum iteration algebra T is reachedmaxThe algorithm is terminated, and step seven is executed; otherwise, making t equal to t +1, and returning to the fourth step to continue the operation;
step seven: and outputting the global optimal food position, and mapping the global optimal food position into a task allocation matrix.
2. The quantum crow group search mechanism-based unmanned aerial vehicle cluster task allocation method according to claim 1, characterized in that: the fitness evaluation process in the third step is as follows:
firstly, the position of ith quantum crow of the t generationAssigning to a task allocation matrix A, wherein the jth generation is the jth dimension of the ith quantum crowIs assigned to al,qAccording toPerforming fitness calculation, whereinc1,c2,c3,c4,c5,c6As a penalty factor, ω1,ω2,ω3,ω4As a weighting factor, ω1+ω2+ω3+ω4=1,0≤ω1,ω2,ω3,ω4And max is less than or equal to 1, the maximum function is obtained, and min is the minimum function.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810224721.9A CN108549402B (en) | 2018-03-19 | 2018-03-19 | Unmanned aerial vehicle group task allocation method based on quantum crow group search mechanism |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810224721.9A CN108549402B (en) | 2018-03-19 | 2018-03-19 | Unmanned aerial vehicle group task allocation method based on quantum crow group search mechanism |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108549402A CN108549402A (en) | 2018-09-18 |
CN108549402B true CN108549402B (en) | 2020-11-10 |
Family
ID=63516652
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810224721.9A Active CN108549402B (en) | 2018-03-19 | 2018-03-19 | Unmanned aerial vehicle group task allocation method based on quantum crow group search mechanism |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108549402B (en) |
Families Citing this family (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109460056B (en) * | 2018-11-06 | 2021-12-24 | 哈尔滨工程大学 | Unmanned aerial vehicle cluster combat game decision method based on quantum krill cluster evolution mechanism |
CN109656136B (en) * | 2018-12-14 | 2022-03-18 | 哈尔滨工程大学 | Underwater multi-AUV (autonomous underwater vehicle) co-location formation topological structure optimization method based on acoustic measurement network |
CN109740954B (en) * | 2019-01-10 | 2021-05-25 | 北京理工大学 | Large-scale unmanned aerial vehicle rapid marshalling method for disaster rescue task |
CN109507891B (en) * | 2019-01-21 | 2021-07-27 | 闽江学院 | Semi-active fuzzy control method |
CN110083173B (en) * | 2019-04-08 | 2022-01-11 | 合肥工业大学 | Optimization method for unmanned aerial vehicle formation inspection task allocation |
CN111476965B (en) * | 2020-03-13 | 2021-08-03 | 深圳信息职业技术学院 | Method for constructing fire detection model, fire detection method and related equipment |
CN111766901B (en) * | 2020-07-22 | 2022-10-04 | 哈尔滨工程大学 | Multi-unmanned aerial vehicle cooperative target distribution attack method |
CN112046467B (en) * | 2020-09-03 | 2021-06-04 | 北京量子信息科学研究院 | Automatic driving control method and system based on quantum computing |
CN112596373B (en) * | 2020-10-27 | 2023-05-23 | 西北工业大学 | Unmanned aerial vehicle attitude control parameter intelligent setting method based on quantum firefly algorithm |
CN113009934A (en) * | 2021-03-24 | 2021-06-22 | 西北工业大学 | Multi-unmanned aerial vehicle task dynamic allocation method based on improved particle swarm optimization |
CN113077082B (en) * | 2021-03-26 | 2022-09-23 | 安徽理工大学 | Mining area mining subsidence prediction method based on improved crow search algorithm |
CN113608546B (en) * | 2021-07-12 | 2022-11-18 | 哈尔滨工程大学 | Unmanned aerial vehicle group task distribution method based on quantum sea lion mechanism |
CN114995492A (en) * | 2022-05-27 | 2022-09-02 | 哈尔滨工程大学 | Multi-unmanned aerial vehicle disaster rescue planning method |
CN114815896B (en) * | 2022-05-27 | 2024-09-13 | 哈尔滨工程大学 | Heterogeneous multi-unmanned aerial vehicle collaborative task allocation method |
CN117556979B (en) * | 2024-01-11 | 2024-03-08 | 中国科学院工程热物理研究所 | Unmanned plane platform and load integrated design method based on group intelligent search |
CN118504928A (en) * | 2024-07-10 | 2024-08-16 | 中国人民解放军国防科技大学 | Task planning method based on multi-objective combined optimization |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101136081B (en) * | 2007-09-13 | 2010-06-02 | 北京航空航天大学 | Unmanned aircraft multiple planes synergic tasks distributing method based on ant colony intelligence |
CN104102791B (en) * | 2014-08-01 | 2017-05-24 | 哈尔滨工程大学 | Antenna array spare construction method based on quantum glowworm search mechanism |
US9896202B2 (en) * | 2014-12-03 | 2018-02-20 | X Development Llc | Systems and methods for reliable relative navigation and autonomous following between unmanned aerial vehicle and a target object |
CN105225003B (en) * | 2015-09-23 | 2018-11-30 | 西北工业大学 | A kind of method that cuckoo searching algorithm solves the problems, such as UAV multitask investigation decision |
CN107045458B (en) * | 2017-03-09 | 2020-05-12 | 西北工业大学 | Unmanned aerial vehicle cooperative task allocation method based on multi-target quantum particle swarm algorithm |
CN107622327B (en) * | 2017-09-15 | 2020-11-03 | 哈尔滨工程大学 | Multi-unmanned aerial vehicle flight path planning method based on culture ant colony search mechanism |
-
2018
- 2018-03-19 CN CN201810224721.9A patent/CN108549402B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN108549402A (en) | 2018-09-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108549402B (en) | Unmanned aerial vehicle group task allocation method based on quantum crow group search mechanism | |
CN112733421B (en) | Task planning method for cooperation of unmanned aerial vehicle with ground fight | |
CN109631900B (en) | Unmanned aerial vehicle three-dimensional flight path multi-target particle swarm global planning method | |
CN111240353B (en) | Unmanned aerial vehicle collaborative air combat decision method based on genetic fuzzy tree | |
CN111722643B (en) | Unmanned aerial vehicle cluster dynamic task allocation method imitating wolf colony cooperative hunting mechanism | |
CN108985549A (en) | Unmanned plane method for allocating tasks based on quantum dove group's mechanism | |
CN111091273B (en) | Multi-bullet collaborative task planning method based on capability prediction | |
CN107589663B (en) | Unmanned aerial vehicle cooperative reconnaissance coverage method based on multi-step particle swarm optimization | |
CN105302153A (en) | Heterogeneous multi-UAV (Unmanned Aerial Vehicle) cooperative scouting and striking task planning method | |
CN112733251B (en) | Collaborative flight path planning method for multiple unmanned aerial vehicles | |
CN113608546B (en) | Unmanned aerial vehicle group task distribution method based on quantum sea lion mechanism | |
Lei et al. | Path planning for unmanned air vehicles using an improved artificial bee colony algorithm | |
CN111797966B (en) | Multi-machine collaborative global target distribution method based on improved flock algorithm | |
CN110986960A (en) | Unmanned aerial vehicle track planning method based on improved clustering algorithm | |
CN112947541A (en) | Unmanned aerial vehicle intention track prediction method based on deep reinforcement learning | |
CN113324545A (en) | Multi-unmanned aerial vehicle collaborative task planning method based on hybrid enhanced intelligence | |
CN113887919A (en) | Hybrid-discrete particle swarm algorithm-based multi-unmanned aerial vehicle cooperative task allocation method and system | |
CN114740883B (en) | Coordinated point reconnaissance task planning cross-layer joint optimization method | |
Jing et al. | Cooperative task assignment for heterogeneous multi-UAVs based on differential evolution algorithm | |
Qingtian | Research on cooperate search path planning of multiple UAVs using Dubins curve | |
CN117008641A (en) | Distribution method and device for cooperative low-altitude burst prevention of multiple heterogeneous unmanned aerial vehicles | |
Qian et al. | Route planning of UAV based on improved ant colony algorithm | |
CN115186378A (en) | Real-time solution method for tactical control distance in air combat simulation environment | |
CN115457809A (en) | Multi-agent reinforcement learning-based flight path planning method under opposite support scene | |
CN115422776A (en) | Multi-base heterogeneous unmanned cluster multi-wave-time unified cooperative task allocation method and system |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |