Method for generating force operation plan and electronic device

Abstract

The present disclosure provides a method for generating a troop operation plan and an electronic device. The method includes: obtaining relevant information of current incoming targets, allocating weapons to different incoming targets to obtain a weapon-target allocation result; obtaining a pre-constructed troop formation behavior tree, and traversing and executing the troop formation behavior tree based on the weapon-target allocation result to obtain a troop formation for assaulting the incoming targets. Among them, the root node of the troop formation behavior tree is connected to a sequence node, the sequence node is connected to a first action node and multiple first selection nodes, the first selection nodes correspond to the incoming targets, and a formation subtree is connected to each of the multiple first selection nodes. The formation subtree is used to determine a troop formation for assaulting the incoming targets based on the weapon-target allocation result and the available combat platform situation; generating a troop operation plan for the troop formation to assault the incoming targets. The troop operation plan provided by the present disclosure can meet the requirements of complex combat tasks.

Description

Technical Field

[0001] The present disclosure relates to the technical field of command and decision-making, and in particular, to a method for generating a force operation plan and an electronic device. Background Art

[0002] The revolution of science and technology has promoted the development of the military field. With the rapid development of artificial intelligence (AI) technology and the large-scale deployment of unmanned platforms, the combat patterns and winning mechanisms of future wars will undergo subversive changes. Promoting the integration of AI technology into combat planning and command and control fields, innovating intelligent combat command models, and realizing the intelligent generation of combat command decision-making plans have become research hotspots in the military field.

[0003] The generation of a command decision-making plan refers to rationally allocating limited combat resources and formulating an operation plan that meets combat tasks based on various factors such as the current combat intention, strike task requirements, battlefield situation, enemy and our situation, combat resource constraints, and the reliability of force completion indicators. However, in the related art, when generating a command decision-making plan, fewer factors are considered, and only a force operation plan that meets simple combat tasks can be generated. It cannot comprehensively consider the impact of various factors on command decision-making and cannot cope with the complex decision-making requirements under various factors.

[0004] Aiming at the problem that the related art cannot comprehensively consider the impact of various factors on command decision-making and cannot cope with complex combat tasks under various factors, no effective technical solution has been proposed yet. Summary of the Invention

[0005] The main purpose of the present disclosure is to provide a method for generating a force operation plan and an electronic device, so as to solve the problem that the related art cannot comprehensively consider the impact of various factors on command decision-making and cannot cope with the complex decision-making requirements under various factors.

[0006] To achieve the above object, a first aspect of the present disclosure provides a method for generating a force operation plan, including:

[0007] Obtain relevant information of current incoming targets, allocate weapons to different incoming targets, and obtain a weapon-target allocation result;

[0008] Obtain a pre - constructed force - grouping behavior tree. Based on the weapon - target assignment result, traverse and execute the force - grouping behavior tree to obtain a force grouping for assaulting the incoming target. Among them, a sequence node is connected under the root node of the force - grouping behavior tree. A first action node and multiple first selection nodes are sequentially connected under this sequence node. Each first selection node corresponds to an incoming target, and a grouping subtree is connected under each of the multiple first selection nodes. The grouping subtree is used to determine the force grouping for assaulting the incoming target based on the weapon - target assignment result and the available combat platform situation; and

[0009] Generate a force action plan for the force grouping to assault the incoming target;

[0010] Among them, generating a force action plan for the force grouping to assault the incoming target includes:

[0011] Obtain a pre - constructed force - action behavior tree;

[0012] Traverse and execute the force - action behavior tree to generate a force action plan for the force grouping to assault the incoming target;

[0013] A sequence node is connected under the root node of the force - action behavior tree. Under this sequence node, a fifth condition node, a fifth action node, a sixth condition node, a sixth action node, a first action subtree, and a second action subtree are sequentially connected in descending order of priority;

[0014] Among them, traversing and executing the force - action behavior tree to generate a force action plan for the force grouping to assault the incoming target includes:

[0015] Execute the sequence node connected under the root node of the force - action behavior tree and all the sub - nodes and sub - trees connected under this sequence node. When the execution results of all sub - nodes and sub - trees are successful executions, return a successful execution result to the root node;

[0016] Execute the fifth condition node to determine whether the position, course, and speed information of the incoming target can be obtained; if it can be obtained, trigger the fifth action node to determine the position of the incoming target at the reference assault moment based on the obtained position, course, and speed information of the incoming target;

[0017] Execute the sixth condition node to determine whether combat platform information, weapon information, and the position of the enemy interference source can be obtained; if it can be obtained, trigger the sixth action node to determine the assault position of the force grouping based on the obtained combat platform information, weapon information, and the position of the enemy interference source;

[0018] Execute the first action subtree to select from multiple candidate formations and configure the selected formation for the force grouping;

[0019] Execute the second action subtree, and when the force formation reaches the assault position and the incoming target reaches the position at the reference assault time, instruct the force formation to stop maneuvering and launch a fire attack at the reference assault time. Optionally, based on the weapon-target allocation result, traverse the force formation behavior tree to obtain the force formation for assaulting the incoming target, including:

[0020] Execute the sequence node connected to the root node of the force organization behavior tree and the first action node connected to the sequence node, and evaluate the threat level of the incoming target through the first action node;

[0021] The incoming targets are matched to the multiple first selection nodes one by one in the order of threat level from high to low, and the multiple first selection nodes and the grouping subtrees connected under the first selection nodes are executed in sequence;

[0022] When the execution results of all the first selected nodes are successful, the execution results of successful execution are returned to the root node.

[0023] Further, the top layer of the marshaling subtree is provided with a first sequence node and a second sequence node in the order of execution, the first condition node and the second selection node are sequentially connected under the first sequence node in the order of execution, and the fourth condition node and the fourth action node are sequentially connected under the second sequence node in the order of execution;

[0024] The execution of the first selection node and the grouping subtree connected under the first selection node includes:

[0025] Select from the first sequence node and the second sequence node connected in sequence under the first selection node in order from high to low priority, and when the execution result of the selected first sequence node or the second sequence node is successful, end the selection operation and return the successful execution result; wherein the first sequence node is used to indicate that the currently available combat platform has a fighter, and the second sequence node is used to indicate that the currently available combat platform has only a destroyer;

[0026] When executing the first sequence node, the first condition node and the second selection node connected to the first sequence node are executed in sequence, and the first condition node is used to determine whether the currently available combat platform has a fighter jet; if it is determined that the currently available combat platform has a fighter jet, the second selection node is triggered to determine the attack method for the incoming target and the force formation for assaulting the incoming target, and the execution result of successful execution is returned; if it is determined that the currently available combat platform has no fighter jet, the execution result of failed execution is directly returned;

[0027] When executing the second-order node, the fourth conditional node and the fourth action node connected under the second-order node are executed in sequence. The fourth conditional node is used to judge whether the number of weapons assigned to the incoming target in the weapon-target assignment result meets the strike requirement, and whether the number of destroyers in the currently available combat platforms meets the ammunition-carrying requirement; if both are met, the fourth action node is triggered to determine the force formation for assaulting the incoming target, and an execution result of successful execution is returned; if either is not met, an execution result of failed execution is directly returned.

[0028] Furthermore, triggering the second selection node to determine the strike method against the incoming target and the force formation for assaulting the incoming target includes:

[0029] Select from the third-order node and the fourth-order node connected in sequence under the second selection node in descending order of priority. When the execution result of the selected third-order node or fourth-order node is successful execution, the selection operation is ended and an execution result of successful execution is returned; among them, the strike method corresponding to the third-order node is single air strike, and the strike method corresponding to the fourth-order node is air-sea combined strike;

[0030] When executing the third-order node, the second conditional node and the second action node connected under the third-order node are executed in sequence. The second conditional node is used to judge whether the number of weapons assigned to the incoming target in the weapon-target assignment result meets the single air strike requirement, and whether the number of fighter jets in the currently available combat platforms meets the ammunition-carrying requirement; if both are met, the second action node is triggered to determine the force formation for assaulting the incoming target; if either is not met, an execution result of failed execution is directly returned;

[0031] When executing the fourth-order node, the third conditional node and the third action node connected under the fourth-order node are executed in sequence. The third conditional node is used to judge whether the number of weapons assigned to the incoming target in the weapon-target assignment result meets the air-sea combined strike requirement, and whether the number of fighter jets and destroyers in the currently available combat platforms meets their respective ammunition-carrying requirements; if both are met, the third action node is triggered to determine the force formation for assaulting the incoming target; if either is not met, an execution result of failed execution is directly returned.

[0032] Furthermore, the first action subtree includes a third selection node. Multiple fifth-order nodes are connected under the third selection node. Each fifth-order node corresponds to a candidate formation. The seventh conditional node and the seventh action node are connected under the fifth-order node in the order of execution sequence;

[0033] Among them, when executing the first action subtree, select from multiple candidate formations and configure the selected formation for the force formation, including:

[0034] Execute the third selection node, select from multiple fifth-order nodes connected below the third selection node in order of decreasing priority. When the execution result of any one of the fifth-order nodes is successful execution, end the selection operation and return the execution result of successful execution;

[0035] When executing the fifth-order node, first execute the seventh conditional node to determine whether the force formation meets the configuration conditions of the corresponding candidate formation; if it meets, trigger the seventh action node to configure the candidate formation for the force formation and return the execution result of successful execution; if it does not meet, directly return the execution result of failed execution.

[0036] Furthermore, use the third selection node to configure the formation for the current force formation;

[0037] The multiple candidate formations include, in order of decreasing priority: diamond, chevron, trapezoid, and wedge.

[0038] Furthermore, determine whether the current force formation meets the configuration conditions of the diamond candidate formation defined in the seventh conditional node. If it meets, trigger the seventh action node on the right to configure the diamond candidate formation for the current force formation;

[0039] Determine whether the current force formation meets the configuration conditions of the chevron candidate formation defined in the seventh conditional node. If it meets, trigger the seventh action node on the right to configure the chevron candidate formation for the current force formation;

[0040] Determine whether the current force formation meets the configuration conditions of the trapezoid candidate formation defined in the seventh conditional node. If it meets, trigger the seventh action node on the right to configure the trapezoid candidate formation for the current force formation;

[0041] Determine whether the current force formation meets the configuration conditions of the wedge candidate formation defined in the seventh conditional node. If it meets, trigger the seventh action node on the right to configure the wedge candidate formation for the current force formation.

[0042] Furthermore, the second action subtree includes a fourth selection node. The fourth selection node is connected to a sixth-order node and an eighth action node in order of execution sequence. The sixth-order node is connected to an eighth conditional node and a ninth action node in order of execution sequence;

[0043] Among them, when executing the second action subtree, when the force formation reaches the assault position and the incoming target reaches the position at the reference assault moment, command the force formation to stop maneuvering and conduct a fire strike at the reference assault moment, including:

[0044] Execute the fourth selection node, select from the sixth sequential node and the eighth action node connected below the fourth selection node, and when the execution result of the selected sixth sequential node or eighth action node is successful execution, end the selection operation and return the execution result of successful execution;

[0045] When executing the sixth sequential node, first execute the eighth conditional node to judge whether the force formation has reached the assault position and whether the incoming target has reached the position at the reference assault moment; if both have reached, trigger the ninth action node to order the force formation to stop maneuvering and conduct a fire strike at the reference assault moment, and return the execution result of successful execution;

[0046] When executing the eighth action node, use the Monte Carlo tree search algorithm to generate a force action plan for the force formation and return the execution result of successful execution.

[0047] Furthermore, using the Monte Carlo tree search algorithm to generate a force action plan for the force formation includes:

[0048] Based on a pre-designed reward function, obtain the score values of the corresponding array positions on the path selected by the combat platform, where the path is the path of the combat platform from the current position to the assault position;

[0049] Use the Monte Carlo tree search algorithm to continuously simulate and determine the force action plan corresponding to the array position with the highest score value.

[0050] The second aspect of the present disclosure provides a computer-readable storage medium, and the computer-readable storage medium stores computer instructions for causing a computer to execute the method for generating a force action plan provided in any item of the first aspect.

[0051] The third aspect of the present disclosure provides an electronic device, and the electronic device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to cause the at least one processor to execute the method for generating a force action plan provided in any item of the first aspect.

[0052] In the method for generating a combat force operation plan provided by the embodiments of the present disclosure, the planning process of the command decision-making assault plan is divided into several tasks: weapon-target assignment, force grouping, and force operation. The logical structure of the behavior tree and prior knowledge are used to hierarchically divide the process of generating the command decision-making assault plan. Based on the weapon-target assignment results, the pre-constructed force grouping behavior tree is traversed and executed to obtain the force grouping for assaulting the incoming targets. Then, a force operation plan for assaulting the incoming targets is generated for the force grouping. By sequentially executing several different tasks and considering the influence of different factors on the force operation plan, the present disclosure finally obtains a command decision-making assault plan composed of weapon-target assignment results, force grouping plans, and force operation plans, which can meet the requirements of complex combat tasks and solve the problems in the related art that cannot comprehensively consider the influence of multiple factors on command decision-making and cannot cope with the complex decision-making requirements under multiple factors. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the specific embodiments of the present disclosure or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present disclosure. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0054] Figure 1 It is a node type diagram of the behavior tree provided by the embodiments of the present disclosure;

[0055] Figure 2 It is a schematic flowchart of the method for generating a command decision-making assault plan provided by the embodiments of the present disclosure;

[0056] Figure 3 It is a schematic flowchart of the method for obtaining weapon-target assignment results provided by the embodiments of the present disclosure;

[0057] Figure 4 It is a schematic structural diagram of the weapon-target assignment behavior tree provided by the embodiments of the present disclosure;

[0058] Figure 5 It is a schematic flowchart of the method for obtaining force grouping provided by the embodiments of the present disclosure;

[0059] Figure 6 It is a schematic structural diagram of the force grouping behavior tree provided by the embodiments of the present disclosure;

[0060] Figure 7 It is a schematic flowchart of the method for generating a force operation plan provided by the embodiments of the present disclosure;

[0061] Figure 8Structural schematic diagram of the military force operation behavior tree provided by the embodiments of the present disclosure;

[0062] Figure 9 Block diagram of the electronic device provided by the embodiments of the present disclosure. Detailed implementation manners

[0063] In order to enable those skilled in the art to better understand the solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present disclosure.

[0064] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances for the embodiments of the present disclosure described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units does not have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0065] In the present disclosure, the orientation or positional relationship indicated by the terms "upper", "lower", "left", "right", "front", "rear", "top", "bottom", "inner", "outer", "middle", "vertical", "horizontal", "lateral", "longitudinal", etc. is based on the orientation or positional relationship shown in the accompanying drawings. These terms are mainly used to better describe the present disclosure and its embodiments, and are not used to limit that the indicated device, element or component must have a specific orientation or be constructed and operated in a specific orientation.

[0066] It should be noted that, without conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other. The present disclosure will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0067] The generation of command decision-making plans refers to rationally allocating limited combat resources and formulating action plans that meet combat tasks based on factors such as the current combat intent, strike task requirements, battlefield situation, enemy and our situation, combat resource constraints, and the reliability of the force completion indicators. However, in related technologies, fewer factors are considered when generating command decision-making plans, and only force action plans that meet simple combat tasks can be generated. It is unable to comprehensively consider the impact of various factors on command decision-making and cannot cope with the complex decision-making requirements under multiple factors. Regarding the problem that related technologies cannot comprehensively consider the impact of various factors on command decision-making and cannot cope with complex combat tasks under multiple factors, no effective technical solutions have been proposed yet.

[0068] To solve the above problems, the embodiments of the present disclosure provide a method for generating a command decision-making assault plan. This method is based on military operations research and supported by behavior trees and reinforcement learning. Considering the characteristics and objectives of the assault task and combining the command decision-making process, the assault plan is decomposed into weapon-target allocation, force grouping generation, and force action planning, and command decision-making is carried out by constructing a behavior tree. The final command decision-making assault plan consists of weapon-target allocation results, force grouping plans, and force action plans.

[0069] The multiple nodes of the behavior tree include composite nodes, decorator nodes, and leaf nodes. The composite nodes include sequence nodes, selector nodes, parallel nodes, and random nodes. The leaf nodes include condition nodes and action nodes. The logical structure of the behavior tree is implemented by the composite nodes, and the leaf nodes are used to help it play its actual role. The leaf nodes can be regarded as the interface between the logical strategy and the entity. The node type diagram of the behavior tree provided by the present disclosure is as Figure 1 shown.

[0070] Among them, the behavior tree can be defined as an organized hierarchical directed node tree composed of edges and nodes, which can be expressed as a binary tuple BT = <V, E>, where V represents the set of nodes of the behavior tree, and E represents the set of edges connecting each node in the behavior tree. When executing the logic of the behavior tree, there is a hierarchical division between nodes. The root node is the starting point of the execution of the entire tree, and then the behavior tree is traversed in a depth-first order. Moreover, the child nodes of the root node can only be sequence nodes or selector nodes. During the traversal process, the next execution node is determined by the composite nodes and decorator nodes, and decisions are made through the action nodes to execute predefined atomic actions. When the leaf nodes finish execution, the execution results will be fed back to the parent nodes. Among them, the execution results include: Success, Failure, and Running.

[0071] The execution logic of each node in the behavior tree will be described below.

[0072] Sequence Node: It can be regarded as a logical "AND" relationship. Each child node of the sequence node is executed in the order of decreasing priority or from left to right. If the return value is Success, the execution continues; if the return value is Failure, the execution stops at that node. Among them, the leftmost child node has the highest priority for execution.

[0073] Selector Node: It can be regarded as a logical "OR" relationship. The child nodes of the selector node represent the sub-tasks or executable actions that can be selected when completing the task of the selector node. The leftmost child node has the highest priority for being selected or executed.

[0074] Parallel Node: Regardless of the return values of the child nodes of the parallel node, all child nodes are executed in parallel in order from left to right.

[0075] Random Node: It has a similar logical structure to the selector node. The difference is that the execution order of the child nodes under the random node is random, and each child node is executed only once.

[0076] Decorator Node: It can be regarded as a prefix of other node types, used to change the behavior of the node and allow custom logic.

[0077] Condition Node: It is equivalent to a conditional judgment statement and returns a corresponding value according to the judgment result.

[0078] Action Node: It is used to execute specific tasks or actions related to the node and takes the execution result as the return value.

[0079] Based on the execution logic of each node in the above behavior tree, the traversal and execution logic of the behavior tree is as follows: starting from the execution starting point, traversing in the order of depth first, and determining the next execution node in the node through the composite nodes and decorator nodes in the behavior tree; when the execution node is a leaf node, execute the predefined atomic action through the action node in the leaf node; after the leaf node is executed, generate the execution result and feedback the execution result to the parent node.

[0080] During the process of constructing the behavior tree, the structure of the behavior tree can be restricted by the following conditions to avoid invalid combinations: (1) The composite nodes in the behavior tree appear in an alternating hierarchical structure of sequence nodes and selector nodes; (2) Any composite node has at least two child nodes.

[0081] The flowchart of the method for generating the command decision-making assault plan provided by the embodiments of the present disclosure is as Figure 2 shown, and the method includes the following steps S101 to step S103:

[0082] Step S101: Obtain relevant information of the current incoming targets, allocate weapons to different incoming targets, and obtain the weapon-target allocation result; among them, the relevant information of the current incoming targets may include the number of current incoming targets and the target types to which the current incoming targets belong.

[0083] This step is used to solve the Weapon Target Assignment (WTA) problem. As one of the main problems to be solved in the field of command and control decision-making, its core is to allocate weapons with different damage capabilities and economic values to different targets under the constraint conditions of known information such as the number of weapons, maximize the weapon utilization rate and combat effectiveness, and at the same time minimize the consumption of combat resources as much as possible, so as to optimize the entire fire strike system. The weapon-target allocation result obtained in this step, that is, the allocation result of allocating weapons to incoming targets.

[0084] In actual implementation, relevant technologies can be adopted in step S101 to complete the weapon-target allocation and obtain the weapon-target allocation result. For example, the method of simulation modeling or the method of generating solutions based on cases can be adopted.

[0085] Optionally, a weapon-target allocation behavior tree can also be pre-constructed, and the weapon-target allocation is completed by traversing and executing the weapon-target allocation behavior tree to obtain the weapon-target allocation result, where the weapon-target allocation behavior tree is a behavior tree constructed based on weapons and incoming targets and used to allocate weapons to incoming targets. In this way, the flow schematic diagram of the method for obtaining the weapon-target allocation result in step S101 is as Figure 3 shown, specifically including the following steps S1011 to S1013:

[0086] Step S1011: Obtain the pre-constructed weapon-target allocation behavior tree, where a sequence node is connected under the root node of the weapon-target allocation behavior tree, and a plurality of fifth selection nodes are connected under the sequence node. Each fifth selection node corresponds to a target, and a first allocation subtree is connected under each fifth selection node; the first allocation subtree is used to determine the target type to which the target belongs, select an assault strike method for the target, and allocate weapons to the target according to the selected assault strike method;

[0087] Allocate our weapons to enemy targets through the pre-constructed weapon-target allocation behavior tree so that the weapons can cause damage to the targets, where the weapon-target allocation behavior tree is a behavior tree used to solve the WTA problem.

[0088] Step S1012: According to the current battlefield situation, obtain the relevant information of all incoming targets and establish the corresponding relationship between the incoming targets and the fifth selection nodes;

[0089] The current battlefield situation includes current incoming targets. The relevant information of the current incoming targets includes the number of current incoming targets and the target type to which the targets belong. The current incoming targets are corresponded to the fifth selection nodes in the weapon-target assignment behavior tree one by one, and the current incoming targets are mainly destroyers.

[0090] Step S1013: Traverse and execute the weapon-target assignment behavior tree to assign weapons to all incoming targets, and obtain the weapon-target assignment result.

[0091] By associating weapons and incoming targets in the way of a behavior tree, hierarchically dividing the weapon-target assignment task by using the logical structure of the behavior tree and prior knowledge, and the subtrees in the behavior tree support reuse. In a dynamic environment, only by executing the weapon-target assignment behavior tree can the WTA problem be flexibly solved to obtain the weapon-target assignment result, while meeting the weapon consumption requirements and target damage requirements, and improving the flexibility and adaptability to the dynamic environment.

[0092] Specifically, the traversing and executing the weapon-target assignment behavior tree in step S1013 to assign weapons to all incoming targets specifically includes:

[0093] Start from the root node and traverse and execute the sequence nodes connected under the root node in the way of depth-first traversal, sequentially execute the multiple fifth selection nodes connected under the sequence node in the order from high to low in priority, and execute the first assignment subtree connected under the fifth selection node;

[0094] When the execution results of all the fifth selection nodes are all successful executions, return the execution result of successful execution to the root node.

[0095] Sequentially execute the multiple fifth selection nodes in the order from high to low in priority corresponding to the targets of each fifth selection node, and when the execution results of all the fifth selection nodes are all successful executions, return successful execution to the execution starting point.

[0096] The structural schematic diagram of the weapon-target assignment behavior tree provided by the embodiments of the present disclosure is as Figure 4 shown. Among them, the child node of the root node is a sequence node, and N fifth selection nodes are connected under the sequence node as child nodes. N represents the number of current incoming targets obtained according to the intelligence. In Figure 4 it, N is 3, and each fifth selection node corresponds to a target;

[0097] Execute N fifth selection nodes in order from the highest to the lowest priority. When the execution results returned by all the fifth selection nodes are all successful, return success to the root node. The priority order of these N fifth selection nodes can be randomly sorted or set according to the evaluation and decision-making preferences of the modeler for the feasible solutions. Figure 4 The parent node and the child nodes have an up-down logical structure. Multiple fifth selection nodes are arranged in a left-right form, and N fifth selection nodes are executed in order from left to right.

[0098] Specifically, multiple seventh order nodes are set at the top layer of the first distribution subtree. Each seventh order node corresponds to a target type. Under the seventh order node, a ninth condition node and a sixth selection node are connected in sequence according to the execution sequence. The ninth condition node is used to limit the corresponding target type, and a second distribution subtree is connected under the sixth selection node; the second distribution subtree is used to select a strike method that can meet the damage requirements for the target and allocate weapons to the target according to the selected strike method.

[0099] Among them, the execution of the fifth selection node and the first distribution subtree connected under the fifth selection node specifically includes:

[0100] Select from multiple seventh order nodes connected under the fifth selection node in order from the highest to the lowest priority. When the execution result of a seventh order node is successful execution, end the selection operation and return the execution result of successful execution;

[0101] When executing the seventh order node, judge whether the type of the incoming target conforms to the target type limited in the ninth condition node; if it conforms, continue to execute the sixth selection node and the second distribution subtree connected under it, and return the execution result of successful execution; if it does not conform, directly return the execution result of failed execution.

[0102] A first distribution subtree is connected under each fifth selection node. Each fifth selection node reuses the first distribution subtree. The first distribution subtree is used to represent the allocation of weapons for the target corresponding to each fifth selection node. Reuse means the module reuse of the behavior tree. Through the reuse of the first distribution subtree, the strong reuse advantage of the weapon-target allocation behavior tree is reflected, effectively improving the flexibility and adaptability of command and decision-making.

[0103] Among them, multiple seventh order nodes are executed in order from the highest to the lowest priority according to the priority of various target types. Among them, the target type is the type of destroyer, and the type of destroyer includes, in order from the highest to the lowest priority: the first level, the second level, and the third level.

[0104] For example, in Figure 4Among them, the first distribution subtree is connected to the N fifth selection nodes. There are A seventh-order nodes at the top layer of the first distribution subtree. A represents the number of target types. Each seventh-order node corresponds to a target type, and the A seventh-order nodes are sorted according to the priorities of the corresponding target types. In the embodiments of the present disclosure, the currently incoming targets considered are mainly destroyers. The current types of destroyers mainly include the first level, the second level, and the third level. Figure 4 Taking A = 3 as an example for illustration, these 3 seventh-order nodes are sorted in the priority order of the first level > the second level > the third level.

[0105] Two child nodes are connected to each seventh-order node. They are a ninth conditional node and a sixth selection node in order from left to right. And the sixth selection node is connected to the second distribution subtree; among them, the ninth conditional node is used to define the corresponding target type, determine whether the currently incoming target conforms to the target type defined in the ninth conditional node, and determine whether to continue executing the subsequent corresponding sixth selection node according to the judgment result. Figure 4 Among them, the rhombus represents the conditional node; moreover, by reusing the second distribution subtree through each sixth selection node, the strong reusability advantage of the behavior tree can be reflected, effectively improving the flexibility and adaptability of command and decision-making.

[0106] When executing the fifth selection node SL corresponding to the mth (m = 1, 2,..., N) incoming target m select from the A seventh-order nodes of the first distribution subtree in order from left to right. After receiving the execution result of the first seventh-order node that returns successfully, immediately return successfully to the parent node and end the selection operation; among them, when executing a certain seventh-order node whose parent node is the fifth selection node SL m determine whether the type of the mth incoming target conforms to the target type defined in the ninth conditional node. If it conforms, continue to execute the sixth selection node on the right, and return successfully to the parent node after successful execution. If it does not conform, directly return failure to the parent node.

[0107] Furthermore, there are multiple eighth-order nodes at the top layer of the second distribution subtree. Each eighth-order node corresponds to a strike method. The tenth conditional node and the action node are connected to the eighth-order node in the order of execution. The tenth conditional node is used to define the corresponding strike method.

[0108] Among them, executing the sixth selection node and the second distribution subtree connected thereto specifically includes:

[0109] Select from the multiple eighth-order nodes connected to the sixth selection node in the order of priority from high to low. When the execution result of an eighth-order node is successful execution, end the selection operation and return the execution result of successful execution.

[0110] When executing the eighth sequential node, it is judged whether the damage requirement can be met by adopting the strike method defined in the tenth conditional node; if it can be met, the subsequent action nodes are triggered to allocate weapons to the incoming target according to the strike method defined in the tenth conditional node, and the execution result of successful execution is returned; if it cannot be met, the execution result of failed execution is directly returned.

[0111] Take Figure 4 as an example. There are two eighth sequential nodes set at the top layer in the second allocation subtree. Here, the strike methods are divided into two categories: single air strike and air-sea mixed strike. With the priority order of single air strike > air-sea mixed strike, the left eighth sequential node corresponds to the single air strike method, and the right eighth sequential node corresponds to the air-sea mixed strike method; the weapons for a single air strike on the incoming destroyer target include air-to-ship missiles, and the weapons for an air-sea mixed strike on the incoming destroyer target include air-to-ship missiles and ship-to-ship missiles.

[0112] When executing the second allocation subtree, select from the two eighth sequential nodes at the top layer in the order from left to right. Immediately return success to the parent node after receiving the execution result of the first eighth sequential node that returns success, and end the selection operation; among them, when executing a certain eighth sequential node, according to the quantity and type of our weapons, the target hit probability and damage probability, it is judged whether the damage requirement can be met by adopting the strike method defined in the tenth conditional node. If it is met, the corresponding action node on the right is triggered to complete the weapon-target allocation based on the minimum consumption of resource value, and the execution result of successful execution is returned; if it is not met, the execution result of failed execution is directly returned to the parent node.

[0113] Furthermore, the target type is the type of destroyer, and the types of destroyers include, in descending order of priority: the first level, the second level, and the third level;

[0114] The strike methods include, in descending order of priority: single air strike and air-sea mixed strike.

[0115] Specifically, the following formula (1) is established with the minimum consumption of resource value F, and weapons are allocated to the incoming target according to the following formula (1):

[0116]

[0117] where x ij represents the quantity of the j-th type of weapon allocated to the i-th target, v j represents the comprehensive value attribute of the j-th type of weapon, and the comprehensive value attribute includes economic value, scarcity value, transportation consumption, etc., q ij represents the damage value of the j-th type of weapon to the i-th target, Qi Indicates the damage requirement for the i-th target, M j Indicates the number of the j-th type of weapon, M represents the total number of weapons, K represents the number of weapon types, N represents the number of targets, and F represents the value of consumed resources.

[0118] Under the assumption that each weapon can only attack one target, without considering the firing order of weapons, and the strike effects between different weapons and different targets do not affect each other, based on the number of our weapons M, the number of weapon types K, and the number of targets N, establish the objective function and constraint conditions of weapon-target allocation to complete the combat mission with the minimum consumed resource value F, that is, formula (1); by executing the objective function and its constraint conditions shown in formula (1), the optimal solution of weapon-target allocation can be obtained, and the weapon-target allocation result that meets the weapon consumption requirement and target damage requirement can be obtained. On the premise of meeting the target damage requirement, the consumption of combat resources is minimized, the weapon utilization rate is maximized, and the entire firepower strike system is optimized.

[0119] Specifically, when executing the eighth sequential node corresponding to a single air strike method, according to the number of weapons, the type of weapons, the hit probability and damage probability to the target, judge whether the single air strike method defined in the tenth conditional node can meet the damage requirement. If it meets, trigger the corresponding action node on the right, and complete the weapon-target allocation based on the minimum consumed resource value, that is, calculate the weapon-target allocation result according to formula (1), and return the execution result of successful execution; if it does not meet, directly return the execution result of failure to the parent node.

[0120] When executing the eighth sequential node corresponding to the air-sea mixed strike method, according to the number of weapons, the type of weapons, the hit probability and damage probability to the target, judge whether the air-sea mixed strike method defined in the tenth conditional node can meet the damage requirement. If it meets, trigger the corresponding action node on the right, and complete the weapon-target allocation based on the minimum consumed resource value, that is, calculate the weapon-target allocation result according to formula (1), and return the execution result of successful execution; if it does not meet, directly return the execution result of failure to the parent node.

[0121] Furthermore, since the incoming targets are mainly destroyers, and the weapons for single air strikes against the incoming destroyer targets are mainly air-to-ship missiles, and the weapons for air-sea mixed strikes against the incoming destroyer targets are mainly air-to-ship missiles and ship-to-ship missiles; therefore, if the single air strike method can meet the damage requirement, the number of weapon types when calculating the weapon-target allocation result according to formula (1) can be 1, that is, air-to-ship missiles; if the air-sea mixed strike method can meet the damage requirement, the number of weapon types when calculating the weapon-target allocation result according to formula (1) can be 2, that is, air-to-ship missiles and ship-to-ship missiles.

[0122] Through the traversal execution process of the above-mentioned weapon-target assignment behavior tree, the weapon-target assignment result can be finally obtained, which can meet the target damage requirement and reduce the consumption of weapon resources at the same time, so as to maximize the weapon utilization rate. Moreover, by reusing the first assignment subtree and the second assignment subtree in the weapon-target assignment behavior tree, it can adapt to the decision-making requirements in a dynamic environment and improve the flexibility and adaptability of command and decision-making.

[0123] Step S102: Obtain the pre-constructed force grouping behavior tree. Based on the weapon-target assignment result, traverse and execute the force grouping behavior tree to obtain the force grouping for assaulting the incoming targets. Among them, a sequence node is connected under the root node of the force grouping behavior tree, and a first action node and multiple first selection nodes are sequentially connected under the sequence node. Each first selection node corresponds to an incoming target, and a grouping subtree is connected under multiple first selection nodes. The grouping subtree is used to determine the force grouping for assaulting the incoming targets based on the weapon-target assignment result and the available combat platform situation. The force grouping includes combat platforms carrying weapons, and the force grouping behavior tree is a behavior tree for determining combat platforms for incoming targets.

[0124] This step is used to obtain the force grouping for assaulting different incoming targets based on the weapon-target assignment result obtained in step S101. The force grouping is the main content of the combat plan planning by the commander and an important link in command and decision-making. According to the requirements and characteristics of the combat mission, quantitatively judge the existing combat platforms and the required combat power, and flexibly and skillfully formulate the grouping plan to achieve the purpose of completing the combat mission. The combat power should be reasonably allocated among the groups to meet combat requirements such as efficient command and coordination, strong mobility, ability to approach the enemy covertly, and strong firepower assault. The force grouping can be defined as first solving the assault mission. By inputting the superior's instructions and the weapon-target assignment result, generate the combat platforms in each group and the number of missiles carried by each of them. A force grouping is the set of all our combat platforms attacking a certain incoming target.

[0125] Specifically, the schematic flow chart of the method for obtaining the force grouping provided by the embodiments of the present disclosure is as Figure 5 shown. Traversing and executing the force grouping behavior tree based on the weapon-target assignment result in step S102 to obtain the force grouping for assaulting the incoming targets includes:

[0126] Step S1021: Execute the sequence node connected under the root node of the force grouping behavior tree and the first action node connected under the sequence node, and evaluate the threat level of the incoming targets through the first action node. Among them, the incoming targets include destroyers.

[0127] The schematic structural diagram of the force grouping behavior tree provided by the embodiments of the present disclosure is asFigure 6 As shown, the child node of the root node is a sequential node, and a first action node and N ( Figure 6 where N = 3 in this example) first selection nodes are connected in sequence from left to right as child nodes.

[0128] Step S1022: Correspond the incoming targets to the multiple first selection nodes one by one in descending order of threat level, and sequentially execute the multiple first selection nodes and the grouped subtrees connected under the first selection nodes; wherein, each first selection node corresponds to an incoming target, each first selection node reuses the grouped subtrees, and the same grouped subtree is connected under the N first selection nodes. The grouped subtree is used to determine the force grouping for assaulting the incoming target based on the weapon type and quantity assigned to the incoming target in the allocation result;

[0129] Starting from the root node, traverse in depth-first order. When traversing to the sequential node connected under the root node, execute the step of threat assessment and sorting of the N incoming targets in descending order of threat level in the leftmost first action node, and establish the correspondence between the N first selection nodes and the N incoming targets according to the sorting result. Then, sequentially execute the N first selection nodes from left to right, and when the execution results returned by the N first selection nodes are all successful executions, return success to the root node.

[0130] Obviously, the leftmost node (i.e., the node with the highest priority) among the N first selection nodes corresponds to the incoming target with the highest threat level, and so on. The rightmost node (i.e., the node with the lowest priority) corresponds to the incoming target with the lowest threat level.

[0131] The grouped subtree of the force grouping behavior tree is used to judge whether the quantity of weapons assigned to the incoming target in the allocation result meets the strike requirements for the incoming target, and judge whether the quantity of currently available combat platforms meets the ammunition-carrying requirements. If both are met, determine the force grouping for assaulting the incoming target; wherein, the combat platforms include fighter jets and destroyers, the weapon types include air-to-ship missiles and ship-to-ship missiles, the weapon type carried by the fighter jet is air-to-ship missiles, and the weapon type carried by the destroyer is ship-to-ship missiles.

[0132] Step S1023: When the execution results of all the first selection nodes are all successful executions, return the execution result of successful execution to the root node.

[0133] The embodiments of the present disclosure hierarchically divide the tasks through the logical structure of the force grouping behavior tree and prior knowledge, and the subtrees in the force grouping behavior tree support reuse, effectively improving the flexibility and adaptability of command and decision-making.

[0134] Specifically, the top layer of the grouping subtree is provided with a first sequential node and a second sequential node in the order of execution. Under the first sequential node, a first conditional node and a second selection node are sequentially connected in the order of execution. Under the second sequential node, a fourth conditional node and a fourth action node are sequentially connected in the order of execution. The top layer of the grouping subtree is provided with two sequential nodes. The left first sequential node represents the situation where the fighter is available as the combat platform for our side currently. The right second sequential node represents the situation where only the destroyer is available as the combat platform for our side currently.

[0135] The execution of the first selection node and the grouping subtree connected under the first selection node in step S1022 includes:

[0136] Select from the first sequential node and the second sequential node sequentially connected under the first selection node in the order of decreasing priority. When the execution result of the selected first sequential node or second sequential node is successful execution, end the selection operation and return the execution result of successful execution. Among them, the first sequential node is used to represent that the fighter is available as the current combat platform, and the second sequential node is used to represent that only the destroyer is available as the current combat platform.

[0137] When executing the first selection node corresponding to the nth (n = 1, 2,..., N) incoming target, select from the first sequential node and the second sequential node of the grouping subtree in the order from left to right. Immediately return success to the parent node after receiving the execution result of the first returned successful sequential node, and end the selection operation.

[0138] When executing the first sequential node, sequentially execute the first conditional node and the second selection node connected under the first sequential node, and judge whether the fighter is available as the current combat platform through the first conditional node. If it is judged that the fighter is available as the current combat platform, trigger the second selection node to determine the strike method for the incoming target and the force grouping for assaulting the incoming target, and return the execution result of successful execution. If it is judged that the fighter is not available as the current combat platform, directly return the execution result of failure.

[0139] When executing the second sequential node, sequentially execute the fourth conditional node and the fourth action node connected under the second sequential node, and judge whether the quantity of the weapon assigned to the incoming target in the weapon - target assignment result meets the strike requirement, and judge whether the quantity of the destroyers in the current available combat platforms meets the ammunition - carrying requirement. If both are met, trigger the fourth action node to determine the force grouping for assaulting the incoming target, and return the execution result of successful execution. If any one is not met, directly return the execution result of failure.

[0140] Further, trigger the second selection node to determine the strike method against the incoming target and the force formation for assaulting the incoming target, including:

[0141] Select from the third-order node and the fourth-order node connected in sequence under the second selection node in descending order of priority. When the execution result of the selected third-order node or fourth-order node is successful execution, end the selection operation and return the execution result of successful execution; among them, the strike method corresponding to the third-order node is single air strike, the strike method corresponding to the fourth-order node is air-sea combined strike, the weapon types corresponding to the single air strike include air-to-ship missiles, and the weapon types corresponding to the air-sea combined strike include air-to-ship missiles and ship-to-ship missiles;

[0142] The child nodes of the second selection node are two sequential nodes. The third-order node on the left corresponds to the single air strike method, while the fourth-order node on the right corresponds to the air-sea combined strike method; further connected to the condition node and the action node under each sequential node, and the subsequent action node can be triggered only after the conditions defined by the condition node are met.

[0143] When executing the third-order node, sequentially execute the second condition node and the second action node connected under the third-order node. Determine through the second condition node whether the number of weapons allocated to the incoming target in the weapon-target allocation result meets the requirements of single air strike, and determine whether the number of fighter jets among the currently available combat platforms meets the ammunition-carrying requirements; if both are met, trigger the second action node to determine the force formation for assaulting the incoming target; if either is not met, directly return the execution result of failure.

[0144] When executing the third-order node, judge according to the second condition node whether the number of weapons (the number of air-to-ship missiles) allocated to the nth incoming target in the weapon-target allocation result obtained in the above step S101 meets the requirements of single air strike, and judge whether the number of currently available fighter jets meets the ammunition-carrying requirements; if both are met, trigger the second action node to calculate the force formation for attacking the nth incoming target and return the success status to the third-order node; if either is not met, directly return failure to the third-order node. Among them, the strike methods include single air strike and air-sea combined strike, the combat platforms include fighter jets and destroyers, the weapon types include air-to-ship missiles and ship-to-ship missiles. When the single air strike method is selected, the corresponding combat platform is the fighter jet, and the weapon type carried by the fighter jet is the air-to-ship missile; when the air-sea combined strike method is selected, the corresponding combat platforms are the fighter jet and the destroyer, and the weapon type carried by the destroyer is the ship-to-ship missile. Therefore, the weapon for a single air strike against the incoming target (mainly the destroyer) is the air-to-ship missile, and the weapons for an air-sea combined strike against the destroyer are the air-to-ship missile and the ship-to-ship missile.

[0145] When calculating the force formation, the weapon loading is carried out according to the weapon - target assignment result. Under the constraint of meeting the maximum ammunition load, as few combat platforms as possible should be selected.

[0146] When executing the fourth - order node, the third - order conditional node and the third - order action node connected under the fourth - order node are executed in sequence. The third - order conditional node is used to judge whether the number of weapons assigned to the incoming target in the weapon - target assignment result meets the requirements of the air - sea mixed strike, and whether the numbers of fighter jets and destroyers among the currently available combat platforms meet their respective ammunition load requirements; if both are met, the third - order action node is triggered to determine the force formation for assaulting the incoming target; if either is not met, the execution result of failure is directly returned.

[0147] When executing the fourth - order node, it is judged according to the third - order conditional node whether the number of weapons (the number of air - to - ship missiles and ship - to - ship missiles) assigned to the nth incoming target in the weapon - target assignment result obtained in the above step S101 meets the requirements of the air - sea mixed strike, and whether the numbers of currently available fighter jets and destroyers meet their respective ammunition load requirements; if both are met, the third - order action node is executed to calculate the force formation for assaulting the nth incoming target, and the success status is returned to the fourth - order node; if either is not met, failure is directly returned to the fourth - order node.

[0148] Furthermore, the formation subtree of the force formation behavior tree connected under the first - order selection node also includes:

[0149] When executing the second - order node, the fourth - order conditional node and the fourth - order action node connected under the second - order node are executed in sequence. The fourth - order conditional node is used to judge whether the number of ship - to - ship missiles among the weapons assigned to the incoming target in the weapon - target assignment result meets the strike requirements, and whether the number of destroyers among the currently available combat platforms meets the ammunition load requirements; if both are met, the fourth - order action node is triggered to determine the force formation for assaulting the incoming target; if either is not met, the execution result of failure is directly returned.

[0150] The second - order node is connected to the fourth - order conditional node and the fourth - order action node. Only after meeting the conditions defined by the fourth - order conditional node can the subsequent fourth - order action node be triggered for execution.

[0151] When executing the second-order node, according to the fourth conditional node, it is judged whether the number of weapons (the number of ship-to-ship missiles) assigned to the nth incoming target in the weapon-target allocation result obtained in the above step S101 meets the strike requirements, and whether the number of currently available destroyers meets the ammunition load requirements; if both are satisfied, the fourth action node is triggered to calculate the force formation for attacking the nth incoming target, and the success status is returned to the second-order node; if either one is not satisfied, a failure is directly returned to the second-order node.

[0152] Step S103: Generate a force operation plan for the force formation to conduct a surprise attack on the incoming target.

[0153] This step is used to generate a force operation plan for each force formation to conduct a surprise attack on the incoming target based on the force formation obtained in the above step S102.

[0154] The flowchart of the method for generating a force operation plan provided by the embodiments of the present disclosure is as Figure 7 shown, and step S103 includes the following steps S1031 to S1032:

[0155] Step S1031: Obtain a pre-constructed force operation behavior tree; wherein, the force operation behavior tree is a behavior tree used to generate a force operation plan for the force formation;

[0156] Step S1032: Traverse and execute the force operation behavior tree to generate a force operation plan for the force formation to conduct a surprise attack on the incoming target.

[0157] The present disclosure hierarchically divides the process of generating a command decision surprise attack plan by using the logical structure of the behavior tree and prior knowledge. By sequentially traversing and executing the pre-constructed weapon-target allocation behavior tree, force formation behavior tree, and force operation behavior tree, and considering the influence of different factors on the force operation plan in different behavior trees, a command decision surprise attack plan composed of a weapon-target allocation result, a force formation plan, and a force operation plan is finally obtained, which can meet the requirements of complex combat tasks and solves the problem in the related art that the influence of multiple factors on command decisions cannot be comprehensively considered and the complex decision-making requirements under multiple factors cannot be addressed.

[0158] The structural schematic diagram of the force operation behavior tree provided by the embodiments of the present disclosure is as Figure 8As shown, a sequence node is connected under the root node of the military force operation behavior tree. Under this sequence node, a fifth conditional node, a fifth action node, a sixth conditional node, a sixth action node, a first action subtree, and a second action subtree are connected in order from the highest to the lowest priority. By traversing and executing the military force operation behavior tree, a military force operation plan is generated for the military force formation that assaults the k-th (k = 1, 2,..., N) incoming target. When traversing and executing the military force operation behavior tree, traversal is performed in the depth-first order starting from the root node.

[0159] Among them, step S1032 includes:

[0160] Execute the sequence node connected under the root node of the military force operation behavior tree and all the child nodes and subtrees connected under this sequence node. When the execution results of all the child nodes and subtrees are all successful executions, return a successful execution result to the root node;

[0161] Execute the fifth conditional node to determine whether the position, course, and speed information of the incoming target can be obtained. If it can be obtained, trigger the fifth action node to determine the position of the incoming target at the reference assault moment according to the obtained position, course, and speed information of the incoming target, where the reference assault moment is the moment when the combat platform assaults the incoming target;

[0162] Optionally, it is possible to judge whether the position, course, and speed information of the incoming target can be obtained according to the battlefield environment and the situation of both sides of the enemy and us. First, trigger the fifth conditional node connected under the sequence node in the order from left to right to judge whether the position, course, and speed information of the k-th incoming target can be obtained according to the battlefield environment and the situation of both sides of the enemy and us. If it can be obtained, trigger the first action node to calculate the position information of the k-th incoming target of the enemy at the reference assault moment, that is, the strike position, according to the obtained position, course, and speed information.

[0163] Execute the sixth conditional node to judge whether the combat platform information, weapon information, and the position of the enemy interference source can be obtained according to the military force formation, battlefield environment, and the situation of both sides of the enemy and us. If it can be obtained, trigger the sixth action node to determine the assault position of the military force formation according to the obtained combat platform information, weapon information, and the position of the enemy interference source, that is, the precise assault position of our military force formation;

[0164] Execute the first action subtree to select from multiple candidate formations and configure the selected formation for the military force formation; configure the formation for the military force formation by executing the first action subtree;

[0165] Execute the second action subtree. When the military force formation reaches the assault position and the incoming target reaches the position at the reference assault moment, command the military force formation to stop maneuvering and conduct a fire strike at the reference assault moment.

[0166] Further, the first action subtree includes a third selection node, under which are connected multiple fifth-order nodes, each fifth-order node corresponding to a candidate formation. Under each fifth-order node are connected a seventh conditional node and a seventh action node in the order of execution priority.

[0167] Among them, when executing the first action subtree, selecting from multiple candidate formations and configuring the selected formation for the force grouping includes:

[0168] Execute the third selection node, and select from multiple fifth-order nodes connected under the third selection node in the order of priority from high to low. When the execution result of any one fifth-order node is successful execution, end the selection operation and return the execution result of successful execution.

[0169] Optionally, use the third selection node to configure the formation for the current force grouping to achieve the purpose of enhancing its strike ability and defense ability; the child nodes of the third selection node are B fifth-order nodes, where B represents the number of candidate formations, and the B fifth-order nodes correspond one-to-one to B candidate formations. Figure 8 Among them, B = 4, and the 4 candidate formations include: diamond, chevron, trapezoid, and wedge. The 4 fifth-order nodes connected under the third selection node are sorted in the priority order of diamond > chevron > trapezoid > wedge, and the child nodes connected under each of the 4 fifth-order nodes are a seventh conditional node and a seventh action node.

[0170] When executing the fifth-order node, first execute the seventh conditional node to determine whether the force grouping meets the configuration conditions of the corresponding candidate formation; if it meets, trigger the seventh action node to configure this candidate formation for the force grouping and return the execution result of successful execution; if it does not meet, directly return the execution result of failure. Among them, each fifth-order node corresponds to a candidate formation of the force grouping, and the multiple candidate formations include, in the order of priority from high to low: diamond, chevron, trapezoid, and wedge.

[0171] When executing the first action subtree, select from the B fifth-order nodes connected under the third selection node in the order from left to right. After receiving the execution result of the first fifth-order node that returns successfully, immediately return success to the parent node and end the selection operation. Among them, when executing a certain fifth-order node among the B fifth-order nodes, first determine whether the current force grouping meets the configuration conditions of the candidate formation defined in the seventh conditional node. If it meets, trigger the seventh action node on the right to configure this candidate formation for the current force grouping and return success to the third selection node.

[0172] Optionally, determine whether the current troop formation meets the configuration conditions of the diamond-shaped candidate formation defined in the seventh conditional node. If it meets, trigger the seventh action node on the right to configure the diamond-shaped candidate formation for the current troop formation;

[0173] Determine whether the current troop formation meets the configuration conditions of the chevron-shaped candidate formation defined in the seventh conditional node. If it meets, trigger the seventh action node on the right to configure the chevron-shaped candidate formation for the current troop formation;

[0174] Determine whether the current troop formation meets the configuration conditions of the trapezoidal candidate formation defined in the seventh conditional node. If it meets, trigger the seventh action node on the right to configure the trapezoidal candidate formation for the current troop formation;

[0175] Determine whether the current troop formation meets the configuration conditions of the wedge-shaped candidate formation defined in the seventh conditional node. If it meets, trigger the seventh action node on the right to configure the wedge-shaped candidate formation for the current troop formation.

[0176] Furthermore, the second action subtree includes a fourth selection node. Under the fourth selection node, a sixth sequence node and an eighth action node are connected in sequence according to the execution order. Under the sixth sequence node, an eighth conditional node and a ninth action node are connected in sequence according to the execution order;

[0177] Among them, when executing the second action subtree, when the troop formation reaches the assault position and the incoming target reaches the position at the reference assault moment, instruct the troop formation to stop maneuvering and conduct a fire strike at the reference assault moment, including:

[0178] Execute the fourth selection node, select from the sixth sequence node and the eighth action node connected under the fourth selection node. When the execution result of the selected sixth sequence node or eighth action node is successful execution, end the selection operation and return the execution result of successful execution;

[0179] Optionally, when executing the fourth selection node, you can select from the sixth sequence node and the eighth action node connected in sequence under the fourth selection node in the order of decreasing priority. After receiving the execution result of the first node that returns successfully, immediately return success to the parent node and end the selection operation. Use the fourth selection node to determine the next action for the troop formation. The child nodes of the fourth selection node are the sixth sequence node and the eighth action node. The child nodes of the sixth sequence node are the eighth conditional node and the ninth action node. In the Figure 8 In the shown troop action behavior tree, two eighth conditional nodes and two ninth action nodes are connected under the fourth selection node. Obviously, in the actual implementation process, there can also be only one eighth conditional node and only one ninth action node. The embodiments of the present disclosure do not make limitations on this.

[0180] When executing the sixth sequential node, first execute the eighth conditional node to determine whether the force formation has reached the assault position and whether the incoming target has reached the position at the reference assault time; if both conditions are met, trigger the ninth action node to command the force formation to stop maneuvering and conduct a fire strike at the reference assault time, and return the execution result of successful execution; otherwise, directly return failure to the sixth sequential node.

[0181] When executing the eighth action node, use the Monte Carlo tree search algorithm to generate a force action plan for the force formation and return the execution result of successful execution. When executing the eighth action node, the Monte Carlo tree search algorithm in the reinforcement learning method can be used to generate a force action plan for the force formation and return the execution result of successful execution to the sixth sequential node.

[0182] Among them, using the Monte Carlo tree search algorithm to generate a force action plan for the force formation includes:

[0183] Based on the pre-designed reward function, obtain the score value of the corresponding position point on the path selected by the combat platform, where the path is the path of the combat platform from the current position to the assault position;

[0184] Use the Monte Carlo tree search algorithm to continuously simulate and determine the force action plan corresponding to the position point with the highest score value.

[0185] Specifically, the steps to generate the force action plan are as follows: give the score value of the combat platform selecting the corresponding position point through the reward function design, and then continuously simulate through the Monte Carlo tree search algorithm to find the best force action plan; among them, the position point is the path planning data set for the combat platform to move from the current position to the precise assault position. By traversing and executing the force action behavior tree based on the learning ability of the reinforcement learning algorithm, the ability to handle complex problems is improved, the learning and adaptability of the force action plan are effectively enhanced, and the problem in the related technology that the influence of multiple factors on command decision-making cannot be comprehensively considered and the complex combat tasks under multiple factors cannot be coped with is solved.

[0186] Furthermore, the Monte Carlo tree search algorithm includes a selection part, an expansion part, a simulation part, and a backtracking part, where:

[0187] Selection part: Assume that the current is the t-th decision moment, and the battlefield situation is x(t), which is obtained based on information such as the position information of both sides, weapon information, incoming target information, and force formation information. The maneuver of the combat platform is a(t). In this disclosure, the action space is divided by the maneuver direction, and each 45° represents a maneuver action, and the number of actions is 8; make a decision on the maneuver action at the t+1 moment. First, select different a(t), and then calculate the state values {S(t)|a(t)=c i}, i ∈ 1, 2, ..., 8, where S(t) and c i respectively represent the angle and action state at the next moment after performing the maneuver at the current moment; if the Monte Carlo tree has not reached the termination condition and there are unexplored nodes, then execute the steps in the expansion part, otherwise execute the steps in the backtracking part.

[0188] Expansion part: Continue to select maneuvers downward, list all possible actions, select the first node as the new node for the next step, and add it to the Monte Carlo tree.

[0189] Simulation part: Simulate the newly added child node until the outcome is determined.

[0190] Backtracking part: Feed the simulation results upward along the new node to the root node and update all parent nodes.

[0191] According to the force action plan obtained by the above Monte Carlo tree search algorithm, determine each battle position point for the action of each force formation.

[0192] This disclosure modularly uses reinforcement learning as the leaf node of the force action behavior tree, continuously simulates using the Monte Carlo tree search algorithm in reinforcement learning to find the best force action plan, and realizes the force action planning task in the assault plan generation by combining the logical rule method and the learning method. The problems of slow convergence speed and difficult reward function design in the learning method are solved through the logical rules and prior constraints of the behavior tree, and the ability to handle complex problems is improved based on the learning ability of reinforcement learning, effectively improving the learning and adaptability of force action planning.

[0193] Specifically, in the process of constructing the weapon - target assignment behavior tree, the force formation behavior tree, or the force action behavior tree, this method further includes:

[0194] Establish constraint conditions for the behavior tree structure, where the constraint conditions include: the composite nodes in the behavior tree appear in a hierarchical structure alternating with sequence nodes and selection nodes, and any composite node has at least two child nodes;

[0195] Construct the structure of the weapon - target assignment behavior tree, the force formation behavior tree, or the force action behavior tree through the above - mentioned constraint conditions. Constraining the structures of the weapon - target assignment behavior tree, the force formation behavior tree, or the force action behavior tree through the constraint conditions can avoid invalid combinations.

[0196] From the above description, it can be seen that this disclosure has achieved the following technical effects:

[0197] The present disclosure divides the planning process of the command decision-making assault plan into several tasks: weapon-target assignment, force grouping, and force operation. Behavior trees are constructed for each task respectively. Through the logical structure of the behavior trees and prior knowledge, the tasks are hierarchically divided, and the subtrees in the behavior trees support reuse, effectively improving the flexibility and adaptability of command decision-making;

[0198] The present disclosure uses reinforcement learning as the leaf node of the force operation behavior tree for modular use, and continuously simulates using the Monte Carlo tree search algorithm in reinforcement learning to find the best force operation plan. The force operation planning task in the generation of the assault plan is realized by combining the logical rule method and the learning method; through the logical rules and prior constraints of the behavior tree, the problems of slow convergence speed and difficult reward function design in the learning method are solved, and the ability to handle complex problems is improved based on the learning ability of reinforcement learning, effectively improving the learning and adaptability of force operation planning, and solving the problem in the related technology that the influence of multiple factors on command decision-making cannot be comprehensively considered and complex combat tasks under multiple factors cannot be handled;

[0199] The present disclosure hierarchically divides the process of generating the command decision-making assault plan by using the logical structure of the behavior tree and prior knowledge. By sequentially traversing and executing the pre-constructed weapon-target assignment behavior tree, force grouping behavior tree, and force operation behavior tree, different factors affecting the command decision-making assault plan are considered in different behavior trees, and finally, a command decision-making assault plan composed of weapon-target assignment results, force grouping plans, and force operation plans is obtained, which can meet the requirements of complex combat tasks;

[0200] By using constraint conditions to constrain the structures of the pre-constructed weapon-target assignment behavior tree, force grouping behavior tree, or force operation behavior tree, invalid combinations can be avoided.

[0201] It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order from here.

[0202] The embodiment of the present disclosure also provides an electronic device, such as Figure 9 shown, the electronic device includes one or more processors 91 and a memory 92, Figure 9 Taking one processor 91 as an example in

[0203] The controller may further include: an input device 93 and an output device 94.

[0204] The processor 91, the memory 92, the input device 93, and the output device 94 can be connected through a bus or other means, Figure 9Take the bus connection as an example.

[0205] The processor 91 can be a central processing unit (CPU for short), and the processor 91 can also be other general-purpose processors, digital signal processors (DSP for short), application specific integrated circuits (ASIC for short), field-programmable gate arrays (FPGA for short), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. chips, or a combination of the above types of chips. The general-purpose processor can be a microprocessor or any conventional processor.

[0206] As a non-transitory computer-readable storage medium, the memory 92 can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the control method in the embodiments of the present disclosure. The processor 91 executes various functional applications and data processing of the server by running the non-transitory software programs, instructions, and modules stored in the memory 92, that is, implements the method for generating the command decision-making assault plan in the above method embodiments.

[0207] The memory 92 can include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the processing device of the server operation, etc. In addition, the memory 92 can include high-speed random access memory, and can also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory 92 optionally includes a memory remotely set relative to the processor 91, and these remote memories can be connected to the network connection device through a network. Examples of the above networks include, but are not limited to, the Internet, enterprise intranets, local area networks, mobile communication networks, and combinations thereof.

[0208] The input device 93 can receive input digital or character information, and generate key signal inputs related to the user settings and function controls of the processing device of the server. The output device 94 can include display devices such as a display screen.

[0209] One or more modules are stored in the memory 92 and, when executed by one or more processors 91, execute as Figure 2 shown in the method.

[0210] Those skilled in the art can understand that to implement all or part of the processes in the methods of the above embodiments, it can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above various motor control methods. Among them, the storage medium can be a magnetic disk, an optical disc, a read-only memory (ROM), a random access memory (RAM), a flash memory (FM), a hard disk drive (HDD), or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above types of memories.

[0211] Although the embodiments of the present disclosure have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present disclosure, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for generating a military force operation plan, characterized in that Including: Obtaining relevant information of the current incoming targets, allocating weapons to different incoming targets, and obtaining a weapon-target allocation result; Obtaining a pre-constructed force formation behavior tree, traversing and executing the force formation behavior tree based on the weapon-target allocation result to obtain a force formation for assaulting the incoming targets. Among them, a sequence node is connected under the root node of the force formation behavior tree, and a first action node and multiple first selection nodes are sequentially connected under this sequence node. Each first selection node corresponds to an incoming target, and a formation subtree is connected under each of the multiple first selection nodes. The formation subtree is used to determine a force formation for assaulting the incoming target based on the weapon-target allocation result and the available combat platform situation; and Generating a force operation plan for the force formation to assault the incoming targets; Among them, generating a force operation plan for the force formation to assault the incoming targets includes: Obtaining a pre-constructed force operation behavior tree; Traversing and executing the force operation behavior tree to generate a force operation plan for the force formation to assault the incoming targets; A sequence node is connected under the root node of the force operation behavior tree, and a fifth condition node, a fifth action node, a sixth condition node, a sixth action node, a first action subtree, and a second action subtree are sequentially connected under this sequence node in order of decreasing priority; Among them, traversing and executing the force operation behavior tree to generate a force operation plan for the force formation to assault the incoming targets includes: Executing the sequence node connected under the root node of the force operation behavior tree and all the child nodes and subtrees connected under this sequence node. When the execution results of all the child nodes and subtrees are execution successes, returning an execution success result to the root node; Executing the fifth condition node to determine whether the position, course, and speed information of the incoming target can be obtained; if it can be obtained, triggering the fifth action node to determine the position of the incoming target at the reference assault moment according to the obtained position, course, and speed information of the incoming target; Executing the sixth condition node to determine whether the combat platform information, weapon information, and enemy jammer position can be obtained; if it can be obtained, triggering the sixth action node to determine the assault position of the force formation according to the obtained combat platform information, weapon information, and enemy jammer position; Executing the first action subtree to select from multiple candidate formations and configure the selected formation for the force formation; Executing the second action subtree to command the force formation to stop maneuvering and conduct a fire strike at the reference assault moment when the force formation reaches the assault position and the incoming target reaches the position at the reference assault moment; 2. The method according to claim 1, characterized in that, Traversing and executing the force formation behavior tree based on the weapon-target allocation result to obtain a force formation for assaulting the incoming targets includes: Execute the sequence node connected to the root node of the force organization behavior tree and the first action node connected to the sequence node, and evaluate the threat level of the incoming target through the first action node; Matching the incoming targets to the multiple first selection nodes one by one in the order of threat level from high to low, and sequentially executing the multiple first selection nodes and the grouping subtrees connected under the first selection nodes; When the execution results of all the first selected nodes are successful, the execution results of successful execution are returned to the root node.

3. The method according to claim 2, wherein The top layer of the grouping subtree is provided with a first sequence node and a second sequence node in the order of execution, the first condition node and the second selection node are sequentially connected under the first sequence node in the order of execution, and the fourth condition node and the fourth action node are sequentially connected under the second sequence node in the order of execution; Wherein, executing the first selection node and the grouping subtree connected under the first selection node includes: Selecting from the first sequence node and the second sequence node sequentially connected to the first selection node in order from high to low priority, and when the execution result of the selected first sequence node or the second sequence node is successful, ending the selection operation and returning the successful execution result; wherein the first sequence node is used to indicate that the currently available combat platform has a fighter, and the second sequence node is used to indicate that the currently available combat platform has only a destroyer; When executing the first sequence node, the first condition node and the second selection node connected to the first sequence node are executed in sequence, and the first condition node is used to determine whether the currently available combat platform has a fighter jet; if it is determined that the currently available combat platform has a fighter jet, the second selection node is triggered to determine the attack method for the incoming target and the force formation for assaulting the incoming target, and an execution result of successful execution is returned; if it is determined that the currently available combat platform has no fighter jet, an execution result of failed execution is directly returned; When executing the second sequence node, the fourth condition node and the fourth action node connected to the second sequence node are executed in sequence, and the fourth condition node is used to determine whether the number of weapons allocated to the incoming target in the weapon-target allocation result meets the strike requirement, and whether the number of destroyers in the currently available combat platform meets the ammunition requirement; if both are satisfied, the fourth action node is triggered to determine the force formation for assaulting the incoming target, and the execution result of successful execution is returned; if any one is not satisfied, the execution result of failed execution is directly returned.

4. The method according to claim 3, wherein Triggering the second selection node to determine a strike method for the incoming target and a force formation for assaulting the incoming target includes: Select from the third-order node and the fourth-order node connected in sequence under the second selection node in the order of decreasing priority. When the execution result of the selected third-order node or fourth-order node is successful execution, end the selection operation and return the execution result of successful execution; wherein, the strike method corresponding to the third-order node is single air strike, and the strike method corresponding to the fourth-order node is air-sea combined strike; When executing the third-order node, sequentially execute the second conditional node and the second action node connected under the third-order node. Determine whether the number of weapons allocated to the incoming target in the weapon-target allocation result meets the requirements of single air strike through the second conditional node, and determine whether the number of fighter jets in the currently available combat platforms meets the ammunition-carrying requirements; if both are met, trigger the second action node to determine the force formation for assaulting the incoming target; if either is not met, directly return the execution result of failed execution; When executing the fourth-order node, sequentially execute the third conditional node and the third action node connected under the fourth-order node. Determine whether the number of weapons allocated to the incoming target in the weapon-target allocation result meets the requirements of air-sea combined strike through the third conditional node, and determine whether the numbers of fighter jets and destroyers in the currently available combat platforms meet their respective ammunition-carrying requirements; if both are met, trigger the third action node to determine the force formation for assaulting the incoming target; if either is not met, directly return the execution result of failed execution.

5. The method according to claim 1, wherein The first action subtree includes a third selection node, and multiple fifth-order nodes are connected under the third selection node. Each fifth-order node corresponds to a candidate formation, and a seventh conditional node and a seventh action node are connected in sequence under the fifth-order node according to the execution sequence; Among them, executing the first action subtree to select from multiple candidate formations and configure the selected formation for the force formation includes: Execute the third selection node, and select from the multiple fifth-order nodes connected under the third selection node in the order of decreasing priority. When the execution result of any one of the fifth-order nodes is successful execution, end the selection operation and return the execution result of successful execution; When executing the fifth-order node, first execute the seventh conditional node to determine whether the force formation meets the configuration conditions of the corresponding candidate formation; if it meets, trigger the seventh action node to configure the candidate formation for the force formation and return the execution result of successful execution; if it does not meet, directly return the execution result of failed execution.

6. The method according to claim 5, wherein Use the third selection node to configure the formation for the current force formation; The multiple candidate formations include, in order of decreasing priority: diamond, herringbone, trapezoid, and wedge.

7. The method according to claim 6, wherein Determine whether the current force formation meets the configuration conditions of the diamond-shaped candidate formation defined in the seventh conditional node. If it meets, trigger the seventh action node on the right to configure the diamond-shaped candidate formation for the current force formation; Determine whether the current force formation meets the configuration conditions of the chevron-shaped candidate formation defined in the seventh conditional node. If it meets, trigger the seventh action node on the right to configure the chevron-shaped candidate formation for the current force formation; Determine whether the current force formation meets the configuration conditions of the trapezoidal candidate formation defined in the seventh conditional node. If it meets, trigger the seventh action node on the right to configure the trapezoidal candidate formation for the current force formation; Determine whether the current force formation meets the configuration conditions of the wedge-shaped candidate formation defined in the seventh conditional node. If it meets, trigger the seventh action node on the right to configure the wedge-shaped candidate formation for the current force formation.

8. The method according to claim 1, wherein The second action subtree includes a fourth selection node. Under the fourth selection node, a sixth sequence node and an eighth action node are connected in sequence according to the execution order. Under the sixth sequence node, an eighth conditional node and a ninth action node are connected in sequence according to the execution order; Among them, when executing the second action subtree, when the force formation reaches the assault position and the incoming target reaches the position at the reference assault moment, instruct the force formation to stop maneuvering and conduct a fire strike at the reference assault moment, including: Execute the fourth selection node, select from the sixth sequence node and the eighth action node connected under the fourth selection node. When the execution result of the selected sixth sequence node or the eighth action node is successful execution, end the selection operation and return the execution result of successful execution; When executing the sixth sequence node, first execute the eighth conditional node to determine whether the force formation has reached the assault position and whether the incoming target has reached the position at the reference assault moment; if both have reached, trigger the ninth action node to instruct the force formation to stop maneuvering and conduct a fire strike at the reference assault moment, and return the execution result of successful execution; When executing the eighth action node, use the Monte Carlo tree search algorithm to generate a force action plan for the force formation and return the execution result of successful execution.

9. The method according to claim 8, wherein The method of using the Monte Carlo tree search algorithm to generate a force action plan for the force formation includes: Based on a pre-designed reward function, obtain the score values of the corresponding array points on the path selected by the combat platform, where the path is the path of the combat platform from the current position to the assault position; Use the Monte Carlo tree search algorithm to continuously simulate and determine the force action plan corresponding to the array point with the highest score value.

10. An electronic device, characterized in that, The electronic device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor executes the method for generating the force action plan according to any one of claims 1-9.

Images