CN111797966A - Multi-machine cooperative global target distribution method based on improved sheep swarm algorithm - Google Patents
Multi-machine cooperative global target distribution method based on improved sheep swarm algorithm Download PDFInfo
- Publication number
- CN111797966A CN111797966A CN202010354151.2A CN202010354151A CN111797966A CN 111797966 A CN111797966 A CN 111797966A CN 202010354151 A CN202010354151 A CN 202010354151A CN 111797966 A CN111797966 A CN 111797966A
- Authority
- CN
- China
- Prior art keywords
- unmanned aerial
- target
- cooperative
- sheep
- algorithm
- 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.)
- Granted
Links
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 82
- 241001494479 Pecora Species 0.000 title claims abstract description 64
- 238000000034 method Methods 0.000 title claims abstract description 31
- 230000009191 jumping Effects 0.000 claims abstract description 9
- 244000144980 herd Species 0.000 claims description 15
- 239000000446 fuel Substances 0.000 claims description 12
- 238000005457 optimization Methods 0.000 claims description 11
- 244000144992 flock Species 0.000 description 17
- 230000002068 genetic effect Effects 0.000 description 11
- 230000008901 benefit Effects 0.000 description 7
- 238000011161 development Methods 0.000 description 5
- 230000003993 interaction Effects 0.000 description 5
- 230000007246 mechanism Effects 0.000 description 5
- 238000004088 simulation Methods 0.000 description 5
- 230000006399 behavior Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 241001074086 Peprilus triacanthus Species 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 235000019687 Lamb Nutrition 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 210000000349 chromosome Anatomy 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000002147 killing effect Effects 0.000 description 2
- 239000002245 particle Substances 0.000 description 2
- 241000282461 Canis lupus Species 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000019637 foraging behavior Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000000670 limiting effect Effects 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000036961 partial effect Effects 0.000 description 1
- 238000009304 pastoral farming Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000002922 simulated annealing Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/04—Constraint-based CAD
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a multi-machine cooperative global target distribution method based on an improved sheep swarm algorithm, which comprises the following steps: step 1, determining cost and income to ensure that the overall efficiency of the cooperative combat of the multiple unmanned aerial vehicles is maximized; step 2, determining constraint conditions to ensure that the overall efficiency of the cooperative combat of the multiple unmanned aerial vehicles is maximized; step 3, determining a multi-machine cooperative global target distribution model; and 4, solving a multi-machine cooperative global object distribution model. The multi-machine cooperative global target distribution method based on the improved sheep swarm algorithm is high in stability, high in optimal solution quality and high in convergence speed, has the capability of jumping out of a local optimal solution, and can effectively guarantee the maximization of the overall efficiency of multi-unmanned aerial vehicle cooperative combat.
Description
Technical Field
The invention relates to the technical field of information of multi-machine cooperative global target distribution and a novel flocks algorithm, in particular to a multi-machine cooperative global target distribution method based on an improved flocks algorithm.
Background
With the development maturity of unmanned aerial vehicle technology and the continuous improvement of intelligent level, unmanned aerial vehicle will become the leader of future sky and the main equipment of each country armed strength in the world, has huge combat potential in the future battlefield. In modern wars of informatization, networking and systematization for resisting high-speed development, tasks such as information reconnaissance, battlefield striking and the like executed by a single unmanned aerial vehicle can not meet the requirements of the current tasks, and the cooperative execution of combat tasks by utilizing a plurality of unmanned aerial vehicles aiming at a plurality of targets becomes a necessary trend. Because the complexity of the operation environment and the number of unmanned aerial vehicles are increased, the cooperative target allocation of the multiple unmanned aerial vehicles is particularly important for improving the operation efficiency, and the cooperative target allocation of the multiple unmanned aerial vehicles is a key technology for independently and cooperatively completing tasks of the multiple unmanned aerial vehicles, and the practicability of cooperative target allocation of the unmanned aerial vehicles and reasonable target allocation of the unmanned aerial vehicles is determined.
In recent years, research on the problem of target allocation of multiple unmanned aerial vehicles is widely concerned, and common methods include mathematical programming, a negotiation-based method and a heuristic algorithm. The mathematical programming is a deterministic method for solving target distribution in a centralized manner, the method has specific requirements on research objects, a mathematical model needs to be changed and adjusted, and when the scale of the model is overlarge, the calculation amount for solving is increased exponentially. The method based on negotiation belongs to a distributed task planning method, and a contract network and an auction algorithm are commonly used. The method is suitable for task allocation and decision under the scenes of strong uncertainty, high dynamic and high real-time requirement. In contrast, heuristic algorithms represented by a particle swarm algorithm, a genetic algorithm, a simulated annealing algorithm, an ant colony algorithm, a grayish wolf algorithm and the like have low computational complexity, are flexible to apply and easy to implement, are widely used for solving a target distribution problem at present, but a reliable initial distribution scheme is difficult to find by a traditional heuristic algorithm, and the convergence speed is not ideal.
The Sheep swarm algorithm (SO) is a novel intelligent cluster algorithm for simulating Sheep swarm foraging behavior proposed in 2018 by qu dac and schrenxiang et al. The algorithm is issued from the core of a cluster intelligent algorithm, and three corresponding strategies of global search, local development and local optimum jumping out in the algorithm are designed by simulating three behaviors of sheep-head sheep leading, sheep-flock interaction and shepherd dog supervision. Compared with a particle swarm algorithm, the algorithm can obtain a solution with higher quality, and has higher convergence rate and better stability. However, in the common herd algorithm, each sheep is searched based on a continuous space (interval), the initial position and the position updating mode are continuous functions, variables in the multi-unmanned aerial vehicle cooperative target allocation problem are discrete, in addition, the stability of the herd operation on the algorithm for solving the target allocation problem is greatly influenced, and the basic herd algorithm is improved correspondingly according to the problems.
Aiming at the problems that the optimal solution quality of the traditional heuristic algorithm is unreliable and the convergence rate is not ideal, a target distribution model with multiple constraint conditions is established according to the characteristic of multi-machine cooperative global target distribution, and an improved herd algorithm is adopted for solving.
Disclosure of Invention
Aiming at the problems, the invention provides a multi-unmanned-aerial-vehicle cooperative global target allocation method based on an improved sheep swarm algorithm, so that the method has the advantages of high stability, high optimal solution quality, high convergence speed, capability of jumping out of local optimal solutions, and capability of more effectively ensuring the maximization of the overall efficiency of multi-unmanned-aerial-vehicle cooperative combat.
Step 1: cost and benefit determination guarantee multi-unmanned aerial vehicle cooperative combat overall efficiency maximization
In order to ensure the maximization of the overall efficiency of the cooperative combat of multiple unmanned aerial vehicles, the fuel consumption cost of the unmanned aerial vehicles and the damage cost when attacking targets are required to be as small as possible, and the income cost when attacking the targets is as large as possible. Therefore, cost benefits are determined from three aspects of fuel consumption cost, attack damage cost and attack benefit cost, and the whole efficiency of the unmanned aerial vehicle cooperative battle is maximized.
(1) Cost of fuel consumption
The problem of fuel consumption needs to be considered when the unmanned aerial vehicle completes a task, fuel consumption of the flight of the unmanned aerial vehicle is related to flight range and flight time, and if the flight speed of the unmanned aerial vehicle is fixed, the shorter the flight distance is, the less the fuel consumption is, and the fuel consumption cost can be represented by the size of the flight range. The total fuel consumption cost can be expressed as:
in the formula (1), di(Tj,Tj+1) Express unmanned plane UiThe distance from the jth target to the j +1 th target in the distributed targets represents the unmanned plane U when j is 1iDistance from the departure point to the first target point. When the target is pre-distributed, the flight range of the unmanned aerial vehicle is represented by the linear distance between the unmanned aerial vehicle and a target point in the application because the specific flight range of the unmanned aerial vehicle cannot be predicted.
(2) Attack damage cost
When the unmanned aerial vehicle attacks the target, the unmanned aerial vehicle is affected by the threats such as enemy firepower, terrain in the flying environment, obstacles and the like, and the minimum damage cost ensures that the threat degree of the unmanned aerial vehicle in the task execution process is minimum.
Suppose unmanned plane UiAttack target TiThe damage probability of the unmanned aerial vehicle is hijThen, the attack damage cost of all the drones is:
(3) cost of attack profit
The attack profit cost refers to the target value profit which can be obtained by the unmanned aerial vehicle when attacking the target. The target attack income cost is maximized as a target, and in order to conveniently calculate a target function value, the target residual value amount is adopted to evaluate the cost.
Suppose unmanned plane UiAttack target TiAt the same time, the pair,
The killing probability of the target is pijTarget TiHas a value of vjThen, the attack profit cost of all the drones is:
in actual combat, each index can not be guaranteed to be optimal, different indexes can be processed by a normalization method for a multi-target problem according to the relative weight of each target, and the problem is converted into a single-target optimization problem. The optimization objective function of the multi-drone cooperative target allocation can be expressed as:
min f=c1α1fF+c2α2fA+c3α3fV(4)
in the formula (4), c1,c2,c3The weight coefficient represents the importance degree of each optimization index, and the value range is [0,1 ]]And satisfy c1+c2+c3=1,α1,α2,α3And the method is a scaling factor, and ensures that all cost values are in the same order.
Step 2: determining constraint conditions to ensure overall efficiency maximization of cooperative combat of multiple unmanned aerial vehicles
Many unmanned aerial vehicle cooperation target distribution is a complicated many restraint optimization problem, for guaranteeing unmanned aerial vehicle cooperative operation overall efficiency maximize, the constraint condition that this application was considered includes:
(1) maximum range constraint
Unmanned aerial vehicle is limited by airborne fuel, and single flight distance is limited, assuming that unmanned aerial vehicle UiTotal distance of flight Li,Express unmanned plane UiThe maximum flying distance is less than the total range of the unmanned aerial vehicle for executing the taskIts maximum distance of flight, i.e.
(2) Maximum execution capacity constraint
The number of ammunition that each unmanned aerial vehicle can carry can receive the restriction of load-carrying capacity, assumes that each ammunition can only attack once target, NumiFor unmanned plane UiThe maximum loading capacity of the unmanned aerial vehicle, the number of targets that can be attacked by the unmanned aerial vehicle is less than or equal to the maximum execution capacity of the unmanned aerial vehicle, that is, the maximum loading capacity of the unmanned aerial vehicle is
(3) Target execution order constraints
When the unmanned aerial vehicle cooperatively attacks the targets, important targets need to be attacked firstly, and meanwhile, the priority of certain specific targets is required to be higher than that of other targets. Suppose a target TiIs higher than the target TjThe execution order needs to satisfy:
tj>ti+Δt(7)
in the formula (7), ti,tjAre respectively a target TiAnd a target TjThe time of attack, Δ t, is the minimum time interval between two targets being attacked, Δ t > 0.
(4) Decision variable constraints
Because the size relation between the number of the unmanned aerial vehicles and the number of the targets is different, the constraints on the unmanned aerial vehicles and the targets during target distribution are also different. When u is larger than t, each unmanned aerial vehicle attacks at least one target; when u < t, each target is attacked at least once. The constraint may be expressed as:
and step 3: determining multi-machine cooperative global object allocation model
3.1 target Allocation battle Scenario
In the actual combat process, the unmanned aerial vehicle and the target can not be determined, and for the purpose that the efficiency of the application in the target distribution process is higher, the application establishes a target distribution model for a research background by using a plurality of unmanned aerial vehicles to execute a cooperative attack task on a plurality of targets.
Suppose that one has U (U ≧ 1) unmanned aerial vehicle U ═ 11U2... UuT (T is more than or equal to 1) determined targets T ═ T for enemies1T2... TtThe flight performance and the load size of each airplane are different, and different targets also have different attack values and resistance capacities.
Because unmanned aerial vehicle and target quantity are uncertain during actual combat, therefore this application is to three kinds of typical condition establishment target distribution models, specifically are:
when u is t, the unmanned aerial vehicles are required to be in one-to-one correspondence with the targets, the model is simple, and the cooperative constraint relationship is few;
when u is larger than t, each unmanned aerial vehicle is only allocated with one target, and the situation that a plurality of unmanned aerial vehicles attack one target in a cooperative mode exists, wherein the requirement on the time cooperation of the unmanned aerial vehicles allocated with the same target in formation is high;
when u is less than t, each target is only allocated once, and a situation that one unmanned aerial vehicle attacks a plurality of targets exists, and the allocated targets need to follow the timing constraint of task execution.
When the target is distributed, the corresponding relation between the unmanned aerial vehicle and the target is determined by a decision variable xijDetermining, which is defined as shown in formula (9):
for different quantitative relationships, the decision variables can be expressed as:
in equation (10), i 'represents the drone number of the attack target j, and j' represents the attack target number of the drone i.
For the multi-objective optimization problem, an objective function needs to be established as an optimization index to judge the quality of a target distribution result, the optimization index considered by the method comprises the fuel consumption cost of the unmanned aerial vehicle, the damage cost when the unmanned aerial vehicle attacks the target and the income cost of the attack target, and meanwhile, the constraint conditions such as the flight distance, the flight time, the load size and the target execution sequence of the unmanned aerial vehicle need to be met.
3.2 construct fitness function
The constraint conditions of the multi-machine cooperative target allocation problem are numerous, and a proper mode needs to be selected for processing to obtain a fitness function. The method adopts a penalty function method to process the constraint condition, and the corresponding penalty function is as follows:
(1) maximum range
If the flying distance of a certain unmanned aerial vehicle exceeds the maximum flying distance of the unmanned aerial vehicle in the target distribution result, punishment is applied to the unmanned aerial vehicle:
in formula (11), L represents the flight distance of the unmanned aerial vehicle, and LmaxAnd l is a penalty value applied when the maximum range constraint is not met.
(2) Maximum execution capacity
If some unmanned aerial vehicle exceeds the maximum execution capacity constraint in the target distribution result, applying punishment to the unmanned aerial vehicle:
in the equation (12), Num represents the target number allocated to the drone, Num represents the payload of the drone, and n is a penalty value applied when the maximum execution capacity constraint is not satisfied.
Target execution order
If some unmanned aerial vehicle in the target distribution result does not meet the target execution order constraint, applying punishment to the unmanned aerial vehicle:
after a plurality of constraint conditions are processed by adopting a penalty function method, the fitness function when the optimal target distribution scheme is solved by the herd algorithm can be expressed as follows:
in the formula (14), f is an objective function value corresponding to a certain sheep, C is a penalty term, and when C is 0, the individual is feasible.
And 4, step 4: multi-machine cooperative global object allocation model solution
4.1 sheep flock initialization
Because the selection of the initialization mode directly affects the search efficiency of the algorithm and the result of the distribution problem, when the discrete lamb swarm algorithm is used for solving the problem of the multi-unmanned aerial vehicle cooperative target distribution, a proper population initialization mode needs to be selected.
In the discrete sheep swarm algorithm, each sheep represents an alternative solution, and the position of the whole sheep swarm is updated through sheep head leading, sheep swarm interaction and shepherd dog supervision behaviors so as to find the optimal solution. This application adopts nimble initialization mode to set up the first generation flocks of sheep according to big or small relation and the constraint between unmanned aerial vehicle and the target. The dimension of the solution represented by each sheep depends on the situation of current target assignment, assuming that the dimension of the solution is NcThen N iscThe value taking mode is as follows:
when the number of the unmanned aerial vehicles is larger than the number of the targets, the dimension is the total number of the unmanned aerial vehicles, and when the number of the targets to be distributed is larger than or equal to the number of the unmanned aerial vehicles, the dimension is the total number of the targets to be distributed. The flocks are represented by multidimensional arrays, as shown in FIGS. 2-4.
Taking u < t (u-4, t-8) as an example, assuming that the final allocation result is the solution as shown in the 1 st sheep in fig. 2, the corresponding decision variable matrix is as follows:
the initialization mode can meet the constraint condition of decision variables during target distribution, the advantages and disadvantages of the initial population directly influence the result of the offspring after evolution, and in order to ensure the validity of the schemes, the initial population is obtained, whether each scheme meets the constraint condition is judged, and if not, the scheme is initialized again.
4.2 sheep flock Algorithm improvement strategy and solving step
The present application improves flock movement into a way of generating random integer update locations. Assuming that a certain sheep executes the operation of leading the first sheep, the position updating mode of the common sheep swarm algorithm is as follows:
the improved sheep swarm algorithm adopts the following leading updating mode:
the Step function in equation (18) is shown in table 1.
TABLE 1 Step function flow
The steps of multi-machine cooperative global object allocation using the improved herd algorithm are shown in table 2.
TABLE 2 improved herd algorithm procedure
The method has the advantages that small changes of each dimension of the solution in the target distribution problem can greatly affect the fitness function value, reinitialization of the herded sheep greatly affects the stability of the algorithm, the shepherd dog supervision mechanism is combined with the genetic algorithm, each sheep is regarded as a chromosome in the genetic algorithm, reinitialization operation is improved into a mode of gene crossing of the same chromosome, the stability of the algorithm is guaranteed, and the capacity of jumping out of local optimum is achieved. Assuming that u-t-6, a particular sheep is grazed, the operation is shown in fig. 5.
The flock algorithm realizes rapid global exploration by leading flocks by simulating head flocks, so that the flocks are rapidly close to a known global optimization solution; local development is realized through mutual movement among flocks of sheep, and the convergence speed is further accelerated; and (4) judging whether to enter local optimum or not and rapidly jumping out of a local optimum solution by applying a shepherd dog supervision mechanism.
(1) Guide collar for sheep head
The first sheep refers to the sheep with the optimal fitness function value in the sheep group, the leading of the first sheep refers to the behavior of each sheep moving towards the first sheep, the global exploration mechanism of the algorithm is corresponded, and in order to guarantee the searching performance, the position of the new sheep is updated only when the fitness function value of the new sheep is better than that of the old sheep, as shown in fig. 6.
(2) Herd interaction
Local development mechanism of corresponding algorithm of sheep flock interactive behaviors, wherein each sheep in a sheep flock is xiWill randomly choose another random sheep xjAnd carrying out a flock interaction strategy. If sheep x is selectediThe fitness value of the sheep is superior to that of a random sheep xjThen xiTo be far away from xjLocation update of (1), xjTowards xiAnd otherwise, the reverse operation is performed. Also to ensure the performance of the search, the sheep position is updated only when the value of the fitness function of the new sheep is better than that of the old sheep, as shown in fig. 7.
(3) Shepherd dog supervision
When the difference value of the fitness function of the first generation sheep and the previous generation sheep is smaller than a threshold value, a shepherd dog supervision mechanism is introduced to jump out of local optimization. Each sheep will be grazed by the shepherd dog with a certain probability p, i.e. the sheep is reinitialized with probability p, as shown in fig. 8.
Compared with the traditional heuristic algorithm and the basic flocks algorithm, the method has the advantages that the quality of the optimal solution obtained in the multi-unmanned-aerial-vehicle cooperative global target distribution is more reliable, the capability of jumping out of the local optimal solution is realized, the convergence speed is higher, the stability is higher, and the maximization of the overall efficiency of the multi-unmanned-aerial-vehicle cooperative combat can be effectively guaranteed.
Drawings
FIG. 1 is a block diagram of steps of a multi-machine cooperative global target distribution method based on an improved sheep swarm algorithm.
Fig. 2 to 4 show the case of population initialization, i.e., u ═ t (u ═ t ═ 6), u < t (u ═ 4, t ═ 8), and u > t (u ═ 8, t ═ 4).
Fig. 5 shows a situation where a certain sheep is grazed when u-t-6 is in grazing operation.
Fig. 6 is a flow chart of the lead algorithm for the sheep.
Fig. 7 is a flow chart of the flock interaction algorithm.
Fig. 8 is a flowchart of a shepherd dog supervision algorithm.
Detailed Description
The systems and methods of use according to the present invention are further illustrated by the following examples:
example 1
The invention relates to a multi-machine cooperative global target distribution method based on an improved sheep swarm algorithm, which specifically comprises the following steps:
step 1: and determining cost and income to ensure the maximization of the overall efficiency of the cooperative operation of the unmanned aerial vehicles. In order to ensure the maximization of the overall efficiency of the cooperative combat of multiple unmanned aerial vehicles, the fuel consumption cost of the unmanned aerial vehicles and the damage cost when attacking targets are required to be as small as possible, and the income cost when attacking the targets is as large as possible.
Step 2: and determining constraint conditions to ensure the maximization of the overall efficiency of the cooperative combat of the multiple unmanned aerial vehicles. The multi-unmanned aerial vehicle cooperative target allocation is a complex multi-constraint optimization problem, and the constraint conditions considered by the application comprise: the maximum flight constraint, the maximum execution capacity constraint, the target execution order constraint and the decision variable constraint are respectively expressed as:
tj>ti+Δt
wherein, unmanned plane UiTotal distance of flight Li,Express unmanned plane UiMaximum flight distance, NumiFor unmanned plane UiMaximum loading of ti,tjAre respectively a target TiAnd a target TjThe time of attack, Δ t, is the minimum time interval between two targets being attacked, Δ t > 0.
And step 3: determining a multi-machine cooperative global target distribution model, determining a target distribution battle scene and constructing a fitness function, wherein the fitness function when the optimal target distribution scheme is solved by the herd algorithm can be expressed as follows:
wherein, f is an objective function value corresponding to a certain sheep, C is a penalty term, and when C is 0, the individual is feasible.
And 4, step 4: and (3) solving the multi-machine cooperative global object distribution model, selecting a proper population initialization mode, and determining an improved strategy and an improved step of the sheep swarm algorithm.
Preferably, a target distribution model with multiple constraint conditions is established according to the characteristics of multi-machine cooperative global target distribution, an improved lamb swarm algorithm is adopted for solving, the quality of the optimal solution obtained in the multi-machine cooperative global target distribution is reliable, the capability of jumping out of the local optimal solution is achieved, the convergence speed is high, the stability is high, and the maximization of the overall efficiency of the multi-unmanned aerial vehicle cooperative combat can be effectively guaranteed.
Example 2
In order to verify the effectiveness of the multi-machine cooperative global target distribution problem of the sheep swarm algorithm, MATLAB simulation experiments are carried out and performance comparison is carried out with a Genetic Algorithm (GA).
Step one, algorithm initialization
Assuming that 8 combat drones are available for calling before combat, 8 target points to be attacked are available, and the initial parameter information classification of the drones and the target points is shown in table 3 and table 4.
TABLE 3 initial information table of unmanned aerial vehicle
TABLE 4 destination Point initial information Table
Assuming that the probability of killing each target by the unmanned aerial vehicle is the same, the damage probability after attacking the target is shown in table 5. The algorithm parameters are set as follows: the population size NP is 50, the maximum iteration number is 100, and the threshold value is 10-8The reset probability p is 0.2.
Table 5 damage probability table after unmanned aerial vehicle attacks target
Because unmanned aerial vehicle quantity is uncertain with target quantity during actual combat, this application carries out simulation experiment to three kinds of different application scenes, and each scene setting is as follows:
scene one: t, unmanned plane U1U6Attack target point T1T6;
Scene two: u > t, unmanned plane U1U8Attack target point T1T4;
Scene three: u < t, noneMan-machine U1U4Attack target point T1T8。
Step two, algorithm simulation
The present application simulates the distribution problem in different scenarios, and the specific distribution results are shown in tables 6, 7, and 8:
TABLE 6 Scenario-one assignment results
Table 7 scene allocation results
TABLE 8 Scenario three Allocation results
Tables 6-8 show that the multi-machine cooperative global target distribution problem under different quantity relationships and multiple constraint conditions can be solved by using the herd algorithm, and reasonable results can be obtained. Due to the fact that the global target distribution problem has extremely high requirements on the quality, the convergence rate and the stability of the solution, the improved herd algorithm and the original genetic algorithm are compared in performance.
Step three, performance analysis
In order to verify the search efficiency of the multi-machine cooperative global target distribution problem solved by the sheep swarm algorithm, the improved sheep swarm algorithm is compared with the genetic algorithm in performance, the same parameter of the two algorithms is set to be the same value, and the cross probability P in the genetic algorithm is set to be the samec0.9, probability of mutation Pm0.1. The initial population mean fitness function values after population initialization are shown in table 9.
Table 9 mean fitness function value of initial population under each scene
In order to avoid the influence of accidental factors in a single experiment, 30 times of simulation tests are respectively carried out on each scene, the results are recorded, and the average value of the results is calculated. The cost function value of the primary population under different scenes is stable: scene one is about 33.7, scene two is about 29.8, and scene three is about 31.6. Table 9 shows that the cost function value of the initial population of 30 simulation experiments has no influence on the performance of the algorithm. The improved sheep swarm algorithm is superior to the genetic algorithm in the average cost function value after population updating for the first time: under the scene one, the ISO is 16.57, the GA is 33.61, under the scene two, the ISO is 16.14, the GA is 29.88, under the scene three, the ISO is 20.29, and the GA is 31.52, and the improved sheep swarm algorithm has the capability of fast global search. The results show that the improved herd algorithm has more ideal convergence speed, better final solution quality and better stability compared with the genetic algorithm, and the improved herd algorithm has fewer parameters compared with the genetic algorithm.
Analysis results show that the convergence speed, the solution quality and the stability of the algorithm are obviously superior to those of a genetic algorithm, the algorithm has fewer parameters, can quickly converge in a few iterations and is stabilized to be near the optimal solution, the capability of jumping out of local optimal is realized, and the problem of multi-machine cooperative global target distribution can be better solved.
The above examples of the present invention are intended to be illustrative only and not limiting of the embodiments of the present invention. It will be apparent to those skilled in the art that any modification or partial replacement without departing from the spirit and scope of the invention is intended to be covered by the appended claims.
Claims (2)
1. A multi-machine cooperative global target distribution method based on an improved sheep swarm algorithm is characterized by comprising the following specific steps:
step 1: determining cost and income to ensure that the overall efficiency of the cooperative combat of the unmanned aerial vehicles is maximized; in order to ensure the maximization of the overall efficiency of the cooperative combat of the multiple unmanned aerial vehicles, the fuel consumption cost of the unmanned aerial vehicles and the damage cost when attacking targets are required to be as low as possible, and the income cost when attacking the targets is as high as possible;
step 2: determining constraint conditions to ensure that the overall efficiency of the cooperative combat of the multiple unmanned aerial vehicles is maximized; the multi-unmanned aerial vehicle cooperative target allocation is a complex multi-constraint optimization problem, and the constraint conditions considered by the application comprise: the maximum flight constraint, the maximum execution capacity constraint, the target execution order constraint and the decision variable constraint are respectively expressed as:
tj>ti+Δt
wherein, unmanned plane UiTotal distance of flight Li,Express unmanned plane UiMaximum flight distance, NumiFor unmanned plane UiMaximum loading of ti,tjAre respectively a target TiAnd a target TjThe time of attack, delta t is the minimum time interval of two targets being attacked, and delta t is more than 0;
and step 3: determining a multi-machine cooperative global target distribution model, determining a target distribution battle scene and constructing a fitness function, wherein the fitness function when the optimal target distribution scheme is solved by the herd algorithm can be expressed as follows:
fitness=f+C
wherein f is an objective function value corresponding to a certain sheep, C is a penalty term, and when C is 0, the individual is feasible;
and 4, step 4: and (3) solving the multi-machine cooperative global object distribution model, selecting a proper population initialization mode, and determining an improved strategy and an improved step of the sheep swarm algorithm.
2. The multi-machine cooperative global object allocation method according to claim 1, characterized in that a multi-constraint-condition object allocation model is established according to the characteristics of multi-machine cooperative global object allocation, an improved herd algorithm is adopted for solving, the quality of an optimal solution obtained in multi-machine cooperative global object allocation is reliable, the capability of jumping out of a local optimal solution is provided, the convergence speed is high, the stability is high, and the overall efficiency maximization of multi-unmanned aerial vehicle cooperative combat can be effectively ensured.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010354151.2A CN111797966B (en) | 2020-04-28 | 2020-04-28 | Multi-machine collaborative global target distribution method based on improved flock algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010354151.2A CN111797966B (en) | 2020-04-28 | 2020-04-28 | Multi-machine collaborative global target distribution method based on improved flock algorithm |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111797966A true CN111797966A (en) | 2020-10-20 |
CN111797966B CN111797966B (en) | 2024-04-05 |
Family
ID=72806701
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010354151.2A Active CN111797966B (en) | 2020-04-28 | 2020-04-28 | Multi-machine collaborative global target distribution method based on improved flock algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111797966B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113190041A (en) * | 2021-04-30 | 2021-07-30 | 哈尔滨工业大学 | Unmanned aerial vehicle cluster online target distribution method based on constraint relaxation technology |
CN115129082A (en) * | 2022-07-13 | 2022-09-30 | 安徽大学 | Sensing detection unmanned aerial vehicle cluster cooperative control method for pollution source search |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2007080584A2 (en) * | 2006-01-11 | 2007-07-19 | Carmel-Haifa University Economic Corp. Ltd. | Uav decision and control system |
CN110766254A (en) * | 2018-07-27 | 2020-02-07 | 深圳市白麓嵩天科技有限责任公司 | Multi-unmanned aerial vehicle cooperative task allocation method based on improved genetic algorithm |
-
2020
- 2020-04-28 CN CN202010354151.2A patent/CN111797966B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2007080584A2 (en) * | 2006-01-11 | 2007-07-19 | Carmel-Haifa University Economic Corp. Ltd. | Uav decision and control system |
CN110766254A (en) * | 2018-07-27 | 2020-02-07 | 深圳市白麓嵩天科技有限责任公司 | Multi-unmanned aerial vehicle cooperative task allocation method based on improved genetic algorithm |
Non-Patent Citations (1)
Title |
---|
王勋;姚佩阳;孙昱;李锴;荣庆;: "文化算法在有人/无人机协同作战目标分配中的应用", 空军工程大学学报(自然科学版), no. 02, 25 April 2016 (2016-04-25) * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113190041A (en) * | 2021-04-30 | 2021-07-30 | 哈尔滨工业大学 | Unmanned aerial vehicle cluster online target distribution method based on constraint relaxation technology |
CN115129082A (en) * | 2022-07-13 | 2022-09-30 | 安徽大学 | Sensing detection unmanned aerial vehicle cluster cooperative control method for pollution source search |
CN115129082B (en) * | 2022-07-13 | 2023-09-29 | 安徽大学 | Perception detection unmanned aerial vehicle cluster cooperative control method for pollution source search |
Also Published As
Publication number | Publication date |
---|---|
CN111797966B (en) | 2024-04-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113741508B (en) | Unmanned aerial vehicle task allocation method based on improved wolf pack algorithm | |
CN108549402B (en) | Unmanned aerial vehicle group task allocation method based on quantum crow group search mechanism | |
CN110889625A (en) | Task planning method for swarm unmanned aerial vehicle system | |
CN110083971B (en) | Self-explosion unmanned aerial vehicle cluster combat force distribution method based on combat deduction | |
CN113485456B (en) | Unmanned aerial vehicle group distributed online self-adaptive task planning method | |
CN111859541B (en) | PMADDPG multi-unmanned aerial vehicle task decision method based on transfer learning improvement | |
CN111240353A (en) | Unmanned aerial vehicle collaborative air combat decision method based on genetic fuzzy tree | |
CN108594645B (en) | Planning method and system for single-station multi-unmanned aerial vehicle distribution and flight route | |
CN106779210A (en) | Algorithm of Firepower Allocation based on ant group algorithm | |
CN113031650B (en) | Unmanned aerial vehicle cluster cooperative target distribution design method under uncertain environment | |
CN111311049A (en) | Multi-agent task allocation method based on income maximization | |
CN111813144B (en) | Multi-unmanned aerial vehicle collaborative route planning method based on improved flocks of sheep algorithm | |
CN111797966B (en) | Multi-machine collaborative global target distribution method based on improved flock algorithm | |
CN112364972A (en) | Unmanned fighting vehicle team fire power distribution method based on deep reinforcement learning | |
CN116702633B (en) | Heterogeneous warhead task reliability planning method based on multi-objective dynamic optimization | |
CN112801540A (en) | Intelligent cooperative architecture design based on unmanned cluster | |
CN115222271A (en) | Weapon target distribution method based on neural network | |
CN113887919A (en) | Hybrid-discrete particle swarm algorithm-based multi-unmanned aerial vehicle cooperative task allocation method and system | |
CN116225049A (en) | Multi-unmanned plane wolf-crowd collaborative combat attack and defense decision algorithm | |
Bergsma et al. | Adaptive spatial reasoning for turn-based strategy games | |
CN110766125A (en) | Multi-target weapon-target allocation method based on artificial fish swarm algorithm | |
CN115829230A (en) | Heterogeneous unmanned aerial vehicle group cooperative attack task allocation method | |
CN113608546B (en) | Unmanned aerial vehicle group task distribution method based on quantum sea lion mechanism | |
CN115047907B (en) | Air isomorphic formation command method based on multi-agent PPO algorithm | |
Cheng et al. | Weapon-target assignment of ballistic missiles based on Q-learning and genetic algorithm |
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 |