A method and system for decomposing reconnaissance tasks and allocating sensor resources

Abstract

The present invention discloses a method for decomposing reconnaissance tasks and allocating sensor resources, which includes the following steps: S1. Obtain the task type issued by the superior and various parameters of the target, and construct a task information database; S2. Obtain sensor resource parameters and construct a sensor resource information database; S3. Evaluate the probability of the sensor discovering the target to prepare for the subsequent sensor resource allocation task; S4. According to the known information in the current scenario, decompose the task issued by the superior using a hierarchical task network; S5. Execute each atomic task in sequence according to the above task decomposition result, and generate a task resource allocation plan in combination with the existing sensor resources. The present invention improves the quality and stability of the solution, realizes the reasonable and optimized allocation of reconnaissance resources, and helps to improve the quality and efficiency of reconnaissance decision-making.

Description

Technical Field

[0001] The present invention relates to the technical field of network task decomposition, and in particular to a method and system for reconnaissance task decomposition and sensor resource allocation based on a hierarchical task network method. Background Art

[0002] Reconnaissance mission decomposition and sensor resource allocation are the core links of command and control and coordinated operations, and are key areas of modern military operations. Optimal reconnaissance perception and decision support can be achieved through reasonable mission planning and sensor scheduling. Reconnaissance mission decomposition refers to the scientific and reasonable decomposition of the complex superior combat mission into multiple executable subtasks based on the grasp of the environmental conditions and real-time situation, and the generation of a primary task list in order to clarify the combat objectives, better allocate resources, coordinate actions and improve combat efficiency. Task decomposition determines the specific tasks of each tactical unit, sensor and other resources, and is an indispensable part. Sensor resource allocation aims to optimize the use of sensors according to mission requirements and environmental changes, and maximize intelligence acquisition and reconnaissance perception effects. Reasonable sensor scheduling can significantly improve the timeliness and accuracy of intelligence, thereby supporting mission execution.

[0003] At present, traditional task decomposition mainly relies on the commander's experience and manual planning, which is inefficient and easily affected by human factors. In addition, as the reconnaissance environment changes, the task requirements may change significantly, and the task decomposition plan needs to be adjusted in real time. As the complexity of tasks and the number of sensors increase, especially when the real-time requirements of operations increase, the traditional static allocation method is inefficient and difficult to cope with complex tasks. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for reconnaissance task decomposition and sensor resource allocation, which solves the technical problems in the prior art such as low efficiency of task allocation and susceptibility to human factors.

[0005] To achieve the above object, the present invention provides the following technical solution: a method for reconnaissance task decomposition and sensor resource allocation, comprising the following steps:

[0006] S1. Obtain the task type issued by the superior, as well as all parameters such as the location, quantity, threat level, type and frequency band of the radiation source signal carried by the target, and build a task information database;

[0007] S2. Obtain all parameters of sensor resources, including type, quantity, location, working frequency band, status, detection range, etc., and build a sensor resource information database;

[0008] S3, evaluate the probability of the sensor finding the target and prepare for the subsequent sensor resource allocation task;

[0009] S4. Decompose the tasks issued by the superior using a hierarchical task network according to the known information in the current scenario;

[0010] S5: Execute each atomic task in sequence according to the task decomposition result of S4, and generate a task resource allocation plan in combination with the existing sensor resources.

[0011] In a specific embodiment, in S1, the task types include multi-target positioning and single-target high-precision positioning. Each parameter of the target includes the position, quantity, threat level, type and frequency band of the radiation source signal carried by the target; in S2, the sensor resource parameters include the type, quantity, position, operating frequency band, and detection range of the sensor resources.

[0012] In a specific embodiment, in S3, the probability that the sensor discovers the target includes:

[0013] Probability that a single sensor discovers the target:

[0014]

[0015] Probability that multiple sensors discover the target:

[0016]

[0017] Among them,

[0018]

[0019] In the formula, D is the detection range of the sensor; d indicates whether the target can be detected by the sensor; s c is the type of sensor resource, s c ∈{tz, lz}, tz i is the i-th communication reconnaissance sensor, lz j is the j-th radar reconnaissance sensor; i, j are the numbers of the sensors, i = 1, 2,..., m, j = 1, 2,..., n; s m is the type of radiation source carried by the target, s m ∈{tz, lz}; r is the distance between the target and the sensor; f m is the frequency band of the target; f c is the operating frequency band of the sensor.

[0020] In a specific embodiment, S4 includes the following steps:

[0021] S41. Select a multi-target positioning task or a single-target high-precision positioning task according to the task type;

[0022] S42. Select the next task according to the task constraint condition method:

[0023] In the multi-target positioning task: the condition for constraint condition method 0 is that the target signal type is only communication signal, the condition for constraint condition method 1 is that the target signal type is only radar signal, and the condition for constraint condition method 2 is that the target signal type includes communication signal and radar signal;

[0024] In the single-target high-precision positioning task: the condition for constraint condition method 0 is that the target signal type is only communication signal, and the condition for constraint condition method 1 is that the target signal type is only radar signal;

[0025] S43. Split the task into executable atomic tasks according to the constraint conditions and task information; the atomic tasks include the communication and reconnaissance sensors available for each target, determine the number of communication and reconnaissance sensors to be used, allocate communication and reconnaissance sensors for multi-targets, and allocate communication and reconnaissance sensors for single-target.

[0026] In a specific embodiment, S5 includes the following steps:

[0027] S51. Determine the set of sensor resources available for each target;

[0028] S52. Determine the number of sensor resources used by each target;

[0029] S53. Allocate sensor resources for each target.

[0030] In a specific embodiment, in S52, the number of targets is two types: single-target and multi-target;

[0031] The number of sensor resources used by the single-target is the number of all available resources in the set of sensor resources available for the target; or,

[0032] Calculate the highest accuracy macc for simultaneous positioning of n resources n , n ∈ [2, size(S)-1], when |macc n+1 - macc n | is less than the set value k, then num = m, where macc m = min(macc n , macc n+1 ); if the condition cannot be met, then num = m, where macc m = min(macc n );

[0033] In the formula, S is the set of sensor resources available for the target; size(S) is the number of elements in the set of sensor resources available for the target; num is the number of sensor resources used by the single-target; m is the number of sensor resources that makes the accuracy the highest; macc mis the highest precision at the set value k;

[0034] As described above, the number of sensor resources used for multiple targets is: 2 available sensor resources are allocated to each target; if the number of available sensor resources for a certain target is less than 2, the positioning task of this target is abandoned and no sensor resources are allocated to it.

[0035] In a specific embodiment, in S53, if the number of sensor resources used is the number of elements size(S) in the set of available sensor resources for the target, the single-target optimal allocation scheme is to allocate all the resources in the set S of available sensor resources for the target; otherwise, the scheme of taking num resources for simultaneous positioning with the highest precision is taken as the optimal allocation scheme;

[0036] In the formula, num is the number of sensor resources used for a single target;

[0037] The allocation of sensor resources for multiple targets includes the following steps:

[0038] S531. Based on the sets of available sensor resources for each target, use the genetic algorithm to randomly generate a certain number of chromosomes as the initial population, and the chromosomes are the initial candidate allocation schemes;

[0039] S532. Calculate the fitness value of each chromosome in the current population;

[0040] S433. Select, cross, and mutate the initial population to obtain the next generation population;

[0041] S534. Perturb the next generation population obtained after the above steps using the simulated annealing method to generate a new solution, calculate the fitness values of the solution before and after perturbation respectively, and obtain the increment Δf between the initial solution and the new solution;

[0042] S535. Apply the Meteopolis rule to determine whether to accept the new solution. If Δf < 0, accept the new solution and go to S536; otherwise, generate a random number ξ = U(0,1). If it satisfies then accept the new solution;

[0043] S536. Judge whether the thermal equilibrium is reached. If it is reached, go to the next step; otherwise, go to S534;

[0044] S537. Perform the cooling operation: when the temperature T i is less than the set temperature threshold, the annealing process ends; otherwise, T i+1 = αT i , α ∈ (0,1), i = i + 1, and continue to go to S534 for annealing;

[0045] S538. Determine whether the termination condition is satisfied. If not, go to S532; otherwise, output the individual with the maximum fitness value in the current population as the optimal solution to the problem sought, representing the best allocation plan for sensor allocation.

[0046] In a specific embodiment, the specific method of S531 is as follows: Allocate sensor resources to each target randomly in descending order of threat level, and add them to the set of allocated sensor resources:

[0047]

[0048] Y = Y ∪ C i ;

[0049] where C is the resource set of the final solution for a single target; Y is the set of allocated sensor resources; S i is the set of sensor resources available for the i-th target; size(S) is the number of elements in the set of sensor resources available for the target; x is the first sensor selected; y is the second sensor selected; is an empty set;

[0050] The specific method of S532 is as follows:

[0051] Calculate the probability that the target can be successfully located:

[0052]

[0053] l > 0 means it may be successfully located, l = 0 means it cannot be successfully located;

[0054] Determine the number of localizations weighted according to the threat level of the target:

[0055]

[0056] Determine the average accuracy of multi-target localization weighted according to the threat level of the target:

[0057]

[0058] The fitness value is:

[0059] f = kf1 + (1 - k)f2;

[0060] where w i is the threat level of each target; l i is the probability that the target can be successfully located; g(C i ) is the localization accuracy of the i-th target; p + q is the gene position of the chromosome, that is, the total number of targets; k is a set value.

[0061] In a specific embodiment, the specific method of S533 is as follows:

[0062] The process of the selection operation is as follows:

[0063] The selection operation is carried out in the way of paired crossover inheritance, and the selection mechanism is the fitness proportion selection mechanism. The selection probability of each chromosome is:

[0064]

[0065] The process of the crossover operation is as follows:

[0066] Randomly select two points in the two selected parental chromosomes to determine the matching crossover gene strings. After swapping the gene strings between the two points to obtain two new chromosomes, compare them with the original chromosomes respectively, and save the excellent individuals;

[0067] The process of the mutation operation is as follows:

[0068] Adopt the method based on substitution mutation, that is, randomly select a substitution position, and replace the gene information at the current position with the gene information that meets the constraints in the current individual, so as to obtain a new chromosome, then compare it with the original chromosome, and save the excellent individuals.

[0069] A reconnaissance task decomposition and sensor resource allocation system, characterized in that the system is used to implement a reconnaissance task decomposition and sensor resource allocation method, including:

[0070] Information acquisition module: used to acquire sensor resource parameters; acquire the task type issued by the superior, as well as the location, quantity, threat level of the target, and the type and frequency band of the radiation source signal carried by the target;

[0071] Database module: including sensor resource information database and task information database;

[0072] Evaluation module: used to evaluate the probability of the sensor discovering the target;

[0073] Decomposition module: decompose the task issued by the superior by using the hierarchical task network according to the known information in the current scenario;

[0074] Allocation module: sequentially execute each atomic task according to the task decomposition result, and generate a task resource allocation plan in combination with the existing sensor resources.

[0075] The present invention also provides an electronic device, including a memory and a processor. When the processor executes the computer management program stored in the memory, it realizes the steps of the reconnaissance task decomposition and sensor resource allocation method.

[0076] The present invention also provides a computer-readable storage medium, on which a computer management program is stored. When the computer management program is executed by a processor, the steps of the reconnaissance task decomposition and sensor resource allocation method are implemented.

[0077] Compared with the prior art, the present invention has the following beneficial effects:

[0078] By constructing a database of sensor resource information and task information, the present invention uses a hierarchical task network to decompose the tasks issued by the superior. The complex reconnaissance tasks are decomposed into atomic tasks at multiple levels. Through a clear hierarchical structure, high-level goals are transformed into low-level executable tasks, making the task planning more systematic, structured, and easy to understand and implement.

[0079] In the process of implementing atomic tasks, the present invention uses a method combining genetic algorithm and simulated annealing algorithm to seek the global optimal solution in the resource allocation problem, improving the quality and stability of the solution, and realizing the reasonable and optimized allocation of resources quickly. The present invention combines the task decomposition based on the hierarchical task network with the genetic-simulated annealing algorithm optimization, and can provide a reconnaissance task decomposition and resource allocation method with clear structure, flexible and efficient, rapid convergence and strong adaptability, which helps to improve the quality and efficiency of decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 is a schematic flow chart of the present invention;

[0081] Figure 2 is a schematic framework diagram of the task decomposition part of the present invention;

[0082] Figure 3 is a schematic flow chart of the single-target sensor resource allocation of the present invention;

[0083] Figure 4 is a schematic flow chart of the multi-target sensor resource allocation of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0084] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments; based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0085] Embodiment 1:

[0086] Please refer to Figure 1 , a reconnaissance task decomposition and sensor resource allocation method, including the following steps:

[0087] S1. Obtain the task type issued by the superior, as well as all parameters such as the location, quantity, threat level of the target, and the type and frequency band of the radiation source signal carried by the target, and construct a task information database;

[0088] S2. Obtain all parameters such as the type, quantity, location, working frequency band, status, detection range of the sensor resources, and construct a sensor resource information database;

[0089] S3. Evaluate the probability of the sensor detecting the target to prepare for the subsequent sensor resource allocation task.

[0090] The task type includes multi-target positioning and single-target high-precision positioning; the type s of the radiation source carried by the target m includes two types, communication detection sensors and mine detection sensors, s m ∈{tz, lz}.

[0091] Since it is necessary to consider whether the detection range, type, frequency band of the sensor match the distance between the target and the sensor, the type and frequency band of the carried radiation source. Therefore, the probability of the sensor detecting the target includes:

[0092] Since the attenuation of electromagnetic waves in space is proportional to the square of the distance, then, the probability of a single sensor detecting the target:

[0093]

[0094] When there are n sensors searching for the same target, assuming they are independent of each other, the probability of their joint detection of the target is:

[0095]

[0096] Among them,

[0097]

[0098] In the formula, D is the detection range of the sensor; d is whether the target can be detected by the sensor; s c is the type of sensor resource, s c ∈{tz, lz}, tz i is the i-th communication detection sensor, lz j is the j-th mine detection sensor; i, j are the numbers of the sensors, i = 1, 2,..., m, j = 1, 2,..., n; s m is the type of the radiation source carried by the target, s m ∈{tz, lz}; r is the distance between the target and the sensor; f m is the frequency band of the target; f c is the working frequency band of the sensor; x i is the probability of the i-th sensor detecting the target alone.

[0099] S4. According to the known information in the current scenario, use a hierarchical task network to decompose the tasks issued by the superior. As Figure 2 shown, the specific steps are as follows:

[0100] S41: According to the task type, select the corresponding composite task, and the composite task includes multi-target positioning and single-target high-precision positioning.

[0101] S42: Select the next task according to the constraint condition method in the composite task.

[0102] In the multi-target positioning task:

[0103] The condition of the constraint condition method0 is that the target signal type is only communication signal; the condition of the constraint condition method1 is that the target signal type is only radar signal; the condition of the constraint condition method2 is that the target signal type includes both communication signal and radar signal.

[0104] In the single-target high-precision positioning task:

[0105] The condition of the constraint condition method0 is that the target signal type is only communication signal, and the condition of the constraint condition method1 is that the target signal type is only radar signal.

[0106] S43: Split the task into executable atomic tasks according to the constraint conditions and task information. The atomic tasks include the communication and reconnaissance sensors available for each target, determining the number of communication and reconnaissance sensors to be used, allocating communication and reconnaissance sensors for multi-targets, and allocating communication and reconnaissance sensors for single-targets, etc.

[0107] S5: Execute each atomic task in turn according to the task decomposition result of S5, and generate the best resource allocation plan for the task in combination with the existing sensor resources.

[0108] It includes the following steps:

[0109] S51. Determine the set of sensor resources available for each target; the number of targets includes two types: single-target and multi-target.

[0110] The number of single-targets (taking communication and reconnaissance sensors as an example) is:

[0111]

[0112] The number of multi-targets (taking communication and reconnaissance sensors and radar and reconnaissance sensors as an example) is:

[0113] Define that there are p communication and reconnaissance sensor targets to be located in the air and q radar and reconnaissance sensor targets to be located; then there are,

[0114]

[0115] Wherein, S is the set of sensor resources available to the target; is the set of sensor resources available to the i-th target; indicates whether the target can be detected by the i-th communication reconnaissance sensor; indicates whether the target can be detected by the j-th communication reconnaissance sensor; indicates whether the target can be detected by the k-th mine reconnaissance sensor; tz i is the i-th communication reconnaissance sensor; tz j is the j-th communication reconnaissance sensor; lz k is the k-th mine reconnaissance sensor; i = 1, 2, …, m + n; j = 1, 2, …, p; k = 1, 2, …, q.

[0116] S52. Determine the number of sensor resources used by each target;

[0117] The number of sensor resources used by the single target is the number of all available resources in the set of sensor resources available to the target; or,

[0118] Calculate the highest precision macc for simultaneous positioning of n resources n , n ∈ [2, size(S) - 1], when |macc n+1 - macc n | is less than the set value k, then num = m, where macc m = min(macc n , macc n+1 ); if the condition cannot be met, then num = m, where macc m = min(macc n );

[0119] Wherein, S is the set of sensor resources available to the target; size(S) is the number of elements in the set of sensor resources available to the target; num is the number of sensor resources used by the single target; m is the number of sensor resources with the highest precision; macc m is the highest precision under the set value k.

[0120] The number of sensor resources used by the multi-target is:

[0121] To ensure the mission success rate and maximize resource utilization, 2 available sensor resources are allocated to each target; if the number of sensor resources available to a certain target is less than 2, the positioning task of this target is abandoned and no sensor resources are allocated to it.

[0122] S53. Allocate sensor resources to each target to obtain the optimal allocation plan, as Figure 3 shown.

[0123] For a single target, if the number of sensor resources used is size(S), the optimal allocation scheme is to allocate all resources in S; otherwise, the optimal allocation scheme is to select num resources to locate the target with the highest accuracy at the same time.

[0124] The allocation of sensor resources for multiple targets is as Figure 4 shown, and includes the following steps:

[0125] S531. Set the genes of the chromosome to p + q bits (the total number of targets). Each gene bit is the number of the sensor resources selected for the location of the target. Based on the set of sensor resources available for each target, use the genetic algorithm to randomly generate a certain number of chromosomes as the initial population, representing a set of initial candidate allocation schemes. The population size is popsize, the crossover probability is P c , and the mutation probability is P m . The initial temperature is T0.

[0126] In a specific embodiment, according to the principle of decreasing threat level, sensor resources are randomly allocated to each target in turn and added to the set Y of allocated sensor resources.

[0127]

[0128] Y = Y ∪ C i ;

[0129] where C is the set of resources of the final scheme for a single target; Y is the set of allocated sensor resources; S i is the set of sensor resources available for the i-th target; size(S) is the number of elements in the set of sensor resources available for the target; x is the first sensor selected; y is the second sensor selected; is an empty set;

[0130] S532. Calculate the fitness value of each chromosome (each individual) in the current population.

[0131] Use l to represent the probability that the target can be successfully located. l > 0 indicates that it may be successfully located, and l = 0 indicates that it cannot be successfully located.

[0132]

[0133] The number of locations weighted by the threat level of the target:

[0134]

[0135] The average accuracy of multi-target location weighted by the threat level of the target:

[0136]

[0137] The fitness value is:

[0138] f = kf1+(1 - k)f2;

[0139] In the formula, w i is the threat level of each target, w i ∈{1, 2, 3}, and the higher the number, the higher the threat level; l i is the probability that the target can be successfully located; g(C i ) is the positioning accuracy of the i-th target; p + q is the gene position of the chromosome, that is, the total number of targets; k is a set value, k = 0.9.

[0140] S533. Select, cross, and mutate the initial population to obtain the next generation population.

[0141] The process of the selection operation is as follows:

[0142] Adopt the method of paired cross inheritance for the selection operation. The selection mechanism is the fitness proportional selection mechanism, (fitnessproportionalmode1), also called the roulette wheel or Monte Carlo selection. The selection probability of each chromosome is:

[0143]

[0144] In the formula, f i is; popsize is the population size.

[0145] The selection probability is proportional to its fitness value. The chromosome with a larger fitness value has a higher probability of being selected. At the same time, adopt the best individual preservation method (elitistmod.e1), and directly copy the chromosome with the largest fitness value in the population to the next generation without paired cross, so as to ensure that the maximum fitness value of the newly generated population is not less than that of the previous generation.

[0146] The process of the cross operation is as follows:

[0147] Randomly select two points in the selected two parent chromosomes to determine the matching cross gene string. After swapping the gene strings between the two points to obtain two new chromosomes, compare them with the original chromosomes respectively, and save the excellent individuals. If the offspring obtained after the cross operation do not meet the specified constraint conditions, corresponding adjustment strategies need to be further adopted.

[0148] In a specific embodiment, assume that the chromosomes of two parents are z1 and z2 respectively, and randomly initialize the position of the cross point as x. If the elements in z1 x belong to Y z2 , then the conflicting resources outside the cross region need to be adjusted.

[0149] The specific adjustment method is as follows: in the set C of available sensor resources of the target where the conflicting resources are located, reselect a sensor resource that has not been used by this chromosome to replace the conflicting resources. If no replaceable resource can be found, reselect the position of the crossover point for the crossover operation.

[0150] The process of the mutation operation is as follows:

[0151] Adopt the method based on substitution mutation, that is, randomly select a substitution position, and replace the gene information at the current position with the gene information that meets the constraints in the current individual, so as to obtain a new chromosome, and then compare it with the original chromosome to save the excellent individuals.

[0152] In a specific embodiment, assume that the chromosome for which the mutation operation is taken is z, and a mutation point x is randomly generated. To ensure that the newly generated chromosome after mutation meets the conditions, first check how many elements in the set C of available sensor resources of the target corresponding to this mutation point are not in the set Y z . If there are two or more, randomly select any two of them to perform the mutation operation on this chromosome; if there is only one, randomly select one of the elements in it to perform the mutation operation; if there is none, regenerate the mutation point and repeat the above work.

[0153] S534. Use the simulated annealing method to perturb the next-generation population obtained after the above steps (such as operations like swapping), to generate a new solution, calculate the fitness values of the solution before and after the perturbation respectively, and obtain the increment Δf = f j -f i ; f i is the fitness value of the initial solution; f j is the fitness value of the new solution.

[0154] S535. Apply the Meteopolis rule to determine whether to accept the new solution. If Δf < 0, accept the new solution and go to S536; otherwise, generate a random number ξ = U(0,1). If it satisfies then accept the new solution.

[0155] S536. Determine whether the thermal equilibrium is reached (reaching the pre-set number of inner loop times). If it is reached, go to the next step; otherwise, go back to S534.

[0156] S537. Perform the temperature reduction operation: when the temperature T at the i-th moment i is less than the set temperature threshold, the annealing process ends; otherwise, T i+1 = αT i , α ∈ (0,1), i = i + 1, and continue to go back to S534 for annealing.

[0157] S538. Determine whether the termination condition (maximum number of iterations) is met. If not, go to S532; otherwise, output the individual with the maximum fitness value in the current population as the optimal solution to the problem sought, representing the best allocation scheme for sensor allocation.

[0158] Embodiment 2:

[0159] The present invention also provides a reconnaissance task decomposition and sensor resource allocation system, which is used to implement the steps of the reconnaissance task decomposition and sensor resource allocation method, including:

[0160] Information acquisition module: used to acquire sensor resource parameters; acquire the task type issued by the superior, as well as the location, quantity, threat level of the target, and the type and frequency band of the radiation source signal carried by the target;

[0161] Database module: including a sensor resource information database and a task information database;

[0162] Evaluation module: used to evaluate the probability of a sensor detecting a target;

[0163] Decomposition module: decompose the task issued by the superior using a hierarchical task network according to the known information in the current scenario;

[0164] Allocation module: sequentially execute each atomic task according to the task decomposition result, and generate a task resource allocation scheme in combination with the existing sensor resources.

[0165] Embodiment 3

[0166] The present invention also provides a device, including a memory and a processor. When the processor executes the computer management program stored in the memory, it implements the steps of the reconnaissance task decomposition and sensor resource allocation method.

[0167] Embodiment 4

[0168] The present invention also provides a computer-readable storage medium, on which a computer management program is stored. When the computer management program is executed by a processor, it implements the steps of the reconnaissance task decomposition and sensor resource allocation method.

Claims

1. A method for decomposing reconnaissance tasks and allocating sensor resources, characterized in that, It includes the following steps: S1. Obtain the task type issued by the superior and various parameters of the target, and construct a task information database; S2. Obtain the sensor resource parameters and construct a sensor resource information database; S3. Evaluate the probability of the sensor detecting the target to prepare for subsequent sensor resource allocation tasks; S4. According to the known information in the current scenario, decompose the task issued by the superior using a hierarchical task network; it includes the following steps: S41. According to the task type, select a multi-target positioning task or a single-target high-precision positioning task; S42. Select the next task according to the task constraint condition method: In the multi-target positioning task: the condition for constraint condition method 0 is that the target signal type is only a communication signal, the condition for constraint condition method 1 is that the target signal type is only a radar signal, and the condition for constraint condition method 2 is that the target signal type includes a communication signal and a radar signal; In the single-target high-precision positioning task: the condition for constraint condition method 0 is that the target signal type is only a communication signal, and the condition for constraint condition method 1 is that the target signal type is only a radar signal; S43. Split the task into executable atomic tasks according to the constraint conditions and task information; the atomic tasks include the communication reconnaissance sensors available for each target, determine the number of communication reconnaissance sensors to be used, allocate communication reconnaissance sensors for multiple targets, and allocate communication reconnaissance sensors for a single target; S5: Execute each atomic task in turn according to the task decomposition result of S4, and generate a task resource allocation plan in combination with the existing sensor resources; it includes the following steps: S51. Determine the set of sensor resources available for each target; S52. Determine the number of sensor resources used for each target, and the number of targets is divided into single-target and multi-target; S53. Allocate sensor resources for each target; If the number of sensor resources used is the number of elements size(S) in the set of sensor resources available for the target, the single-target optimal allocation plan is to allocate all the resources in the set S of sensor resources available for the target; otherwise, the plan with the highest precision for simultaneously positioning num resources is taken as the optimal allocation plan; num is the number of sensor resources used for a single target; The allocation of multi-target sensor resources includes the following steps: S531. Based on the set of sensor resources available for each target, use the genetic algorithm to randomly generate a certain number of chromosomes as the initial population, and the chromosomes are the initial candidate allocation plans; S532. Calculate the fitness value of each chromosome in the current population; S533. Select, cross, and mutate the initial population to obtain the next generation population; S534. Perturb the next generation population obtained after the above steps using the simulated annealing method to generate a new solution, calculate the fitness values of the solutions before and after the perturbation respectively, and obtain the increment Δf between the initial solution and the new solution; S535. Determine whether to accept the new solution by applying the Meteopolis rule. If Δf < 0, accept the new solution and go to S536; otherwise, generate a random number ξ = U(0, 1). If it satisfies then accept the new solution; S536. Determine whether the thermal equilibrium is reached. If it is reached, go to the next step; otherwise, go to S534; S537. Perform a cooling operation: When the temperature T i is less than the set temperature threshold, the annealing process ends; otherwise, T i61 = αT i , where α ∈ (0, 1), i = i + 1, and continue to go back to S534 for annealing; S538. Determine whether the termination condition is met. If not, go to S532; otherwise, output the individual with the maximum fitness value in the current population as the optimal solution to the problem sought, representing the best allocation plan for sensor allocation.

2. The method for decomposing a reconnaissance mission and allocating sensor resources according to claim 1, characterized in that In S1, the task types include multi-target positioning and single-target high-precision positioning; each parameter of the target includes the position, quantity, threat level, type and frequency band of the radiation source signal carried by the target; in S2, the sensor resource parameters include the type, quantity, position, operating frequency band, and detection range of the sensor resources.

3. The method for decomposing a reconnaissance mission and allocating sensor resources according to claim 1, wherein, In S3, the probability that the sensor discovers the target includes: The probability that a single sensor discovers the target: The probability that multiple sensors discover the target: Among them, Where, D is the detection range of the sensor; d indicates whether the target can be detected by the sensor; s c is the type of sensor resource, s c ∈{tz, lz}, tz i is the i-th communication reconnaissance sensor, lz j is the j-th mine reconnaissance sensor; i, j are the numbers of the sensors, i = 1, 2, …, m, j = 1, 2, …, n; s m is the type of the radiation source carried by the target, s m ∈{tz, lz}; r is the distance between the target and the sensor; f m is the frequency band of the target; f c is the operating frequency band of the sensor.

4. A method for decomposing a reconnaissance mission and allocating sensor resources according to claim 1, characterized in that In S52, The quantity of sensor resources used by the single target is the quantity of all available resources in the set of sensor resources available to the target; or, Calculate the highest precision macc for the simultaneous positioning of n resources n , where n ∈ [2, size(S) - 1]. When |macc n+1 - macc n | is less than the set value k, then num = m, where macc m = min(macc n , macc n+1 ); if the condition cannot be met, then num = m, where macc m = min(macc n ); Wherein, S is the set of sensor resources available to the target; size(S) is the number of elements in the set of sensor resources available to the target; num is the number of sensor resources used by a single target; m is the number of sensor resources that maximizes the accuracy; macc m is the highest accuracy at the set value k; The quantity of sensor resources used by the multi-target is: allocate 2 available sensor resources to each target; if the quantity of sensor resources available to a certain target is less than 2, then abandon the positioning task of that target and do not allocate sensor resources to it.

5. A method for decomposing a reconnaissance mission and allocating sensor resources according to claim 1, characterized in that The specific method of S531 is: allocate sensor resources to each target randomly in order from high to low threat level, and add them to the set of allocated sensor resources: Y = Y ∪ C i ; Where C is the resource set of the single-object final solution; Y is the allocated sensor resource set; S i is the set of sensor resources available for the i-th target; size(S) is the number of elements in the set of sensor resources available for the target; x is the first selected sensor; y is the second selected sensor; is an empty set; The specific method of S532 is: Calculate the probability that the target can be successfully located: l > 0 means it may be successfully located, l = 0 means it cannot be successfully located; Determine the positioning quantity weighted by the target threat level: Determine the average accuracy of multi-target positioning weighted by the target threat level: The fitness value is: f = kf1+(1 - k)f2; where, w i is the threat level of each target; l i is the probability that the target can be successfully located; g(C i ) is the positioning accuracy of the i-th target; p + q is the gene position of the chromosome, that is, the total number of targets; k is a set value.

6. The method for decomposing reconnaissance tasks and allocating sensor resources according to claim 5, wherein The specific method of S533 is: The process of the selection operation is as follows: Adopt the method of paired crossover inheritance for the selection operation, and the selection mechanism is the fitness proportion selection mechanism. The selection probability of each chromosome is: where f i is the fitness value of the i-th individual; popsize is the population size; The process of the crossover operation is as follows: Randomly select two points in the two selected parent chromosomes to determine the matching crossover gene strings. After swapping the gene strings between the two points to obtain two new chromosomes, compare them with the original chromosomes respectively, and save the excellent individuals; The process of the mutation operation is as follows: Adopt the method based on substitution mutation, that is, randomly select a substitution position, and replace the gene information at the current position with the gene information that meets the constraints in the current individual, so as to obtain a new chromosome, then compare it with the original chromosome, and save the excellent individuals.

7. A reconnaissance mission decomposition and sensor resource allocation system, characterized in that, The system is used to implement a method for reconnaissance task decomposition and sensor resource allocation according to any one of claims 1-6, including: Information acquisition module: used to acquire sensor resource parameters; acquire the task type issued by the superior, as well as the position, quantity, threat level, type and frequency band of the radiation source signal carried by the target; Database module: includes a sensor resource information database and a task information database; Evaluation module: used to evaluate the probability that the sensor discovers the target; Decomposition module: decompose the task issued by the superior by using a hierarchical task network according to the known information in the current scene; Allocation module: execute each atomic task in turn according to the task decomposition result, and generate a task resource allocation plan in combination with the existing sensor resources.

Images