CN110222406A - Unmanned aerial vehicle autonomous capacity assessment method based on task stage complexity - Google Patents

本发明提供一种基于任务阶段复杂性的无人机自主能力评估方法，包括：建立环境复杂性评估体系，对无人机系统在执行任务过程中的环境复杂度进行计算；建立任务复杂度评估体系，获取无人机系统在不同阶段的任务复杂度；建立无人机自主能力评估模型，将环境复杂度与任务复杂度输入无人机自主能力评估模型，无人机自主能力评估模型根据环境复杂度、任务复杂度与人机自主等级划分标准中的自主等级分类输出无人机自主等级。利用无人机自主能力评估模型以人机自主等级划分标准为基准结合环境复杂度和任务复杂度对无人机自主等级作出评估，使无人机系统的操作模式可根据自主等级评估结果动态调整，改变操作员控制权限。本发明应用于无人机系统技术领域。

The invention provides a method for evaluating the autonomous capability of UAVs based on the complexity of the task stage, including: establishing an environment complexity evaluation system to calculate the environmental complexity of the UAV system in the process of executing tasks; establishing task complexity evaluation system to obtain the task complexity of the UAV system at different stages; establish a UAV autonomous capability evaluation model, input the environment complexity and task complexity into the UAV autonomy evaluation model, and the UAV autonomy evaluation model is based on the environment Complexity, task complexity, and the autonomy level classification in the classification standard of human-machine autonomy level output the autonomy level of UAV. Use the UAV autonomy evaluation model to evaluate the UAV autonomy level based on the human-machine autonomy level classification standard combined with environmental complexity and task complexity, so that the operation mode of the UAV system can be dynamically adjusted according to the autonomy level evaluation results , to change the operator control permissions. The invention is applied to the technical field of unmanned aerial vehicle systems.

一种基于任务阶段复杂性的无人机自主能力评估方法A method for evaluating autonomous capability of unmanned aerial vehicles based on the complexity of mission phases

技术领域technical field

本发明涉及无人机系统技术领域，尤其涉及一种基于任务阶段复杂性的无人机自主能力评估方法。The invention relates to the technical field of unmanned aerial vehicles, in particular to a method for evaluating the autonomous capability of unmanned aerial vehicles based on the complexity of task phases.

背景技术Background technique

从应用上看，无人机系统任务复杂性也具有较高研究价值，对于一个无人机系统任务的复杂度评价结果可用于指导选择执行该任务的无人机平台。通过综合评价模型对特定任务进行评价，得到其复杂度，任务复杂性评估结果可与环境复杂性评估结果以及人机交互评估结果相结合组成最终无人机系统自主等级评估结果，用于指导无人机自主作战时控制权限的分配。如何确定人在无人机系统任务执行过程中的角色及参与程度，以辅助无人机在实战中发挥最大自主作战效能成为无人机研究的一个方向。人的介入程度与无人机系统的自主程度在一定情况下是紧密相关的，故无人机系统的自主性评估对于指导无人机在实际应用中的控制权限分配具有重要意义。From the perspective of application, the complexity of UAV system tasks also has high research value. The evaluation results of the complexity of a UAV system task can be used to guide the selection of UAV platforms to perform this task. Evaluate the specific task through the comprehensive evaluation model to obtain its complexity. The task complexity evaluation result can be combined with the environmental complexity evaluation result and the human-computer interaction evaluation result to form the final autonomous level evaluation result of the UAV system, which is used to guide the UAV system. Allocation of control authority for man-machine autonomous combat. How to determine the role and degree of participation of human beings in the execution of UAV system tasks, so as to assist UAVs to maximize their autonomous combat effectiveness in actual combat has become a direction of UAV research. The degree of human intervention is closely related to the degree of autonomy of the UAV system under certain circumstances, so the evaluation of the autonomy of the UAV system is of great significance for guiding the distribution of control authority of the UAV in practical applications.

当前国内外针对无人机系统任务复杂性的研究较少，未系统地提出任务复杂性的评估模型。一些学者对于复杂任务的评估提出理论方法，如周艳美、李伟华等人对于复杂环境中任务方案的评价给出了一种层次化的指标体系，一些研究是针对广泛的无人系统，一些研究未具体规定任务执行的主体，孙扬以无人驾驶车辆为研究对象，建立了由环境复杂度、任务复杂度、人工干预程度组成的无人驾驶车辆评测模型。但该研究对任务复杂性的研究是基于无人驾驶车辆对该任务的执行表现，与根据任务参数对任务复杂度进行预先评价仍存在差别。At present, there are few studies on the task complexity of UAV systems at home and abroad, and no evaluation model for task complexity has been systematically proposed. Some scholars have proposed theoretical methods for the evaluation of complex tasks. For example, Zhou Yanmei, Li Weihua and others have given a hierarchical index system for the evaluation of task programs in complex environments. Some studies are aimed at a wide range of unmanned systems, and some studies are not specific. As the subject of task execution, Sun Yang took unmanned vehicles as the research object and established an unmanned vehicle evaluation model consisting of environmental complexity, task complexity, and human intervention. However, the study of task complexity in this study is based on the performance of unmanned vehicles on the task, which is still different from the pre-evaluation of task complexity based on task parameters.

发明内容Contents of the invention

针对现有技术中的不足，本发明的目的是提供一种基于任务阶段复杂性的无人机自主能力评估方法。Aiming at the deficiencies in the prior art, the purpose of the present invention is to provide a method for evaluating the autonomy of UAVs based on the complexity of the task phase.

其采用的技术方案是：The technical solutions it adopts are:

一种基于任务阶段复杂性的无人机自主能力评估方法，包括以下步骤：A method for evaluating autonomous capabilities of UAVs based on the complexity of mission phases, including the following steps:

步骤101，建立环境复杂性评估体系，对无人机系统在执行任务过程中的环境复杂度进行计算；Step 101, establishing an environmental complexity evaluation system, and calculating the environmental complexity of the UAV system during mission execution;

步骤102，建立任务复杂度评估体系，获取无人机系统在不同阶段的任务复杂度；Step 102, establishing a task complexity evaluation system to obtain the task complexity of the UAV system at different stages;

步骤103，建立无人机自主能力评估模型，将环境复杂度与任务复杂度输入无人机自主能力评估模型，无人机自主能力评估模型根据环境复杂度、任务复杂度与人机自主等级划分标准中的自主等级分类输出无人机自主等级。Step 103, establish the UAV autonomous capability assessment model, input the environment complexity and task complexity into the UAV autonomy assessment model, and the UAV autonomy assessment model is divided according to the environment complexity, task complexity and human-machine autonomy level The autonomy level classification in the standard outputs the drone autonomy level.

作为上述技术方案的进一步改进，步骤101中，所述环境复杂度包括地形复杂度、气象复杂度、通信复杂度、目标识别复杂度、威胁复杂度。As a further improvement of the above technical solution, in step 101, the environment complexity includes terrain complexity, meteorological complexity, communication complexity, target identification complexity, and threat complexity.

作为上述技术方案的进一步改进，所述地形复杂度的计算过程为：As a further improvement of the above technical solution, the calculation process of the terrain complexity is:

步骤201，获取实时地形图片；Step 201, obtaining real-time terrain pictures;

步骤202，计算实时地形图片的图像熵与灰度共生矩阵的反差值；Step 202, calculating the image entropy of the real-time terrain image and the contrast value of the gray level co-occurrence matrix;

步骤203，对实时地形图片的图像熵与灰度共生矩阵的反差值分别进行归一化处理，获得图像熵值的归一化值，以及灰度共生矩阵的反差值的归一化值；Step 203, performing normalization processing on the image entropy of the real-time terrain image and the contrast value of the gray level co-occurrence matrix respectively, to obtain the normalized value of the image entropy value and the normalized value of the contrast value of the gray level co-occurrence matrix;

步骤204，根据图像熵值的归一化值与灰度共生矩阵的反差值的归一化值计算地形复杂度。Step 204, calculating the terrain complexity according to the normalized value of the image entropy value and the normalized value of the contrast value of the gray level co-occurrence matrix.

作为上述技术方案的进一步改进，所述威胁复杂度的计算过程为：As a further improvement of the above technical solution, the calculation process of the threat complexity is:

步骤301，针对雷达、高炮和地空导弹三种战场防空火力建立了杀伤区域模型；In step 301, a kill zone model is established for the three battlefield air defense firepowers of radar, antiaircraft gun and surface-to-air missile;

步骤302，根据威胁点的位置在杀伤区域模型中绘制战场防空火力分布图；Step 302, drawing a battlefield air defense firepower distribution map in the kill zone model according to the position of the threat point;

步骤303，计算战场防空火力分布图中的安全区域比例，进而获得威胁复杂度。Step 303, calculating the proportion of the safe area in the battlefield air defense firepower distribution map, and then obtaining the threat complexity.

作为上述技术方案的进一步改进，As a further improvement of the above technical solution,

所述气象复杂度的计算过程为：The calculation process of the meteorological complexity is:

步骤401，选取风切变、风力等级、雷暴天气和降雨天气进行模糊综合评价；Step 401, selecting wind shear, wind level, thunderstorm weather and rainfall weather for fuzzy comprehensive evaluation;

步骤402，针对评价结果出现的不符合常理的情况，采取隶属度次大的趋于评价等级较差的评价等级作为评价的结果，进而获得气象复杂度；Step 402, in view of the unreasonable situation in the evaluation results, take the evaluation grade with the second highest degree of membership tending to be inferior to the evaluation grade as the evaluation result, and then obtain the meteorological complexity;

所述通信复杂度的计算过程为：The calculation process of the communication complexity is:

步骤501，选取包括丢包率，误码率，时间延迟和中断的因素集合和评语集；Step 501, selecting factor sets and comment sets including packet loss rate, bit error rate, time delay and interruption;

步骤502，根据模糊综合评价的方法进行综合评价,进而获得通信复杂度；Step 502, perform comprehensive evaluation according to the method of fuzzy comprehensive evaluation, and then obtain communication complexity;

所述目标识别复杂度的计算过程为：The calculation process of the target recognition complexity is:

步骤601，针对目标混淆采取生成目标结构特征空间的方法进行度量；Step 601, taking the method of generating the feature space of the target structure to measure the target confusion;

步骤602，针对目标遮掩采取计算目标与局部背景对比度的方法进行度量；Step 602, measure the target mask by calculating the contrast between the target and the local background;

步骤603，计算得到目标识别复杂度。Step 603, calculating the target recognition complexity.

作为上述技术方案的进一步改进，步骤102中，任务复杂度的计算过程为：As a further improvement of the above technical solution, in step 102, the calculation process of task complexity is:

步骤701，将任务复杂度的评估指标量化为任务及战术行为、协同与协作、规划与分析、态势感知；Step 701, quantifying the evaluation indicators of task complexity into tasks and tactical behaviors, coordination and collaboration, planning and analysis, and situational awareness;

步骤702，求取步骤701中每一评估指标的主观复杂度与客观复杂度，通过加权计算主观复杂度与客观复杂度得到每一评估指标的综合复杂度；Step 702, obtaining the subjective complexity and objective complexity of each evaluation index in step 701, and obtaining the comprehensive complexity of each evaluation index by weighting the subjective complexity and objective complexity;

步骤703，综合四种评估指标的综合复杂度进而得到任务复杂度。Step 703, combining the comprehensive complexities of the four evaluation indicators to obtain the task complexity.

作为上述技术方案的进一步改进，步骤103中，所述人机自主等级划分标准中的自主等级分类包括：机为主模式、机主人辅模式、人主机辅模式与人为主模式。As a further improvement of the above technical solution, in step 103, the classification of autonomy levels in the human-machine autonomy classification standard includes: machine-based mode, machine-master-assisted mode, human-master-assisted mode, and human-based mode.

作为上述技术方案的进一步改进，步骤103中，采用无监督学习中的自适应共振网络模型ART作为无人机自主能力评估模型，所述无人机自主等级的求取过程为:As a further improvement of the above-mentioned technical solution, in step 103, the self-adaptive resonance network model ART in unsupervised learning is adopted as the autonomous capability evaluation model of the UAV, and the process of obtaining the autonomous level of the UAV is:

步骤801，根据人机自主等级划分标准中的自主等级分类对自适应共振网络模型ART中识别层的每个神经元进行分类；Step 801, classify each neuron of the identification layer in the adaptive resonance network model ART according to the classification of autonomous levels in the standard of human-machine autonomous classification;

步骤802，根据环境复杂度、任务复杂度计算自适应共振网络模型ART的最大激活值；Step 802, calculating the maximum activation value of the adaptive resonance network model ART according to the environment complexity and task complexity;

步骤803，将最大激活值识别层的每个神经元进行比较，距离最大激活值最近的神经元对应的自主等级分类即为无人机自主等级。In step 803, each neuron in the maximum activation value recognition layer is compared, and the autonomous level classification corresponding to the neuron closest to the maximum activation value is the autonomous level of the drone.

作为上述技术方案的进一步改进，步骤802中，所述最大激活值的求取过程为：As a further improvement of the above technical solution, in step 802, the calculation process of the maximum activation value is:

式中，S为最大激活值；g为神经元非线性激励函数；x_i为神经元输入，n＝9，其中(x₁,x₂,x₃,x₄,x₅)为环境复杂度、(x₆,x₇,x₈,x₉)为任务复杂度；ω_i为前向传播权重。In the formula, S is the maximum activation value; g is the nonlinear activation function of the neuron; x _i is the input of the neuron, n=9, where (x ₁ , x ₂ , x ₃ , x ₄ , x ₅ ) is the complexity of the environment , (x ₆ , x ₇ , x ₈ , x ₉ ) are the task complexity; ω _i is the forward propagation weight.

一种基于任务阶段复杂性的无人机自主能力评估系统，包括存储器和处理器，所述存储器存储有计算机程序，所述处理器执行所述计算机程序时实现上述方法的步骤。A system for evaluating the autonomous capability of an unmanned aerial vehicle based on the complexity of a mission stage includes a memory and a processor, the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

本发明的有益技术效果：Beneficial technical effect of the present invention:

本发明通过建立建立环境复杂性评估体系与任务复杂度评估体系来获取无人机系统在执行任务过程中的环境复杂度和任务复杂度，最终利用无人机自主能力评估模型以人机自主等级划分标准为基准结合环境复杂度和任务复杂度对无人机自主等级作出评估，使得无人机系统的操作模式可以根据自主等级评估结果动态调整，从而改变操作员控制权限。The present invention acquires the environmental complexity and task complexity of the UAV system in the process of performing tasks by establishing an environment complexity evaluation system and a task complexity evaluation system, and finally uses the UAV autonomous capability evaluation model to determine the human-machine autonomy level The division standard is used as the benchmark to evaluate the autonomy level of the UAV in combination with the complexity of the environment and the complexity of the task, so that the operation mode of the UAV system can be dynamically adjusted according to the evaluation results of the autonomy level, thereby changing the operator's control authority.

附图说明Description of drawings

图1是本发明中基于任务阶段复杂性的无人机自主能力评估方法的流程示意图；Fig. 1 is the schematic flow chart of the autonomous capability evaluation method of the unmanned aerial vehicle based on the task stage complexity among the present invention;

图2是本发明中环境复杂性评估体系的评估结果示例图；Fig. 2 is an example diagram of the evaluation results of the environmental complexity evaluation system in the present invention;

图3是本发明中任务复杂性评估体系的评估结果示例图；Fig. 3 is the evaluation result example figure of task complexity evaluation system in the present invention;

图4是本发明中地形复杂度的获取流程示意图；Fig. 4 is a schematic diagram of the acquisition process of terrain complexity in the present invention;

图5是本发明中威胁复杂度的获取流程示意图；Fig. 5 is a schematic diagram of the acquisition process of threat complexity in the present invention;

图6是本发明中气象复杂度的获取流程示意图；Fig. 6 is a schematic diagram of the acquisition process of meteorological complexity in the present invention;

图7是本发明中通信复杂度的获取流程示意图；Fig. 7 is a schematic diagram of the acquisition process of communication complexity in the present invention;

图8是本发明中目标识别复杂度的获取流程示意图；Fig. 8 is a schematic diagram of the acquisition process of target recognition complexity in the present invention;

图9是本发明中任务复杂度的获取流程示意图；Fig. 9 is a schematic diagram of the acquisition process of task complexity in the present invention;

图10是本发明中各阶段任务复杂度示例图；Fig. 10 is an example diagram of task complexity at each stage in the present invention;

图11是本发明中人机自主等级划分标准的示意图；Fig. 11 is a schematic diagram of the human-machine autonomy classification standard in the present invention;

图12是本发明中无人机自主等级的获取流程示意图。Fig. 12 is a schematic diagram of the acquisition process of the autonomy level of the drone in the present invention.

具体实施方式Detailed ways

为了使本公开的目的、技术方案和优点更加清楚明白，下结合具体实施例，并根据附图，对本发明进一步详细说明。需要说明的是，在附图或说明书描述中，未描述的内容以及部分英文简写为所属技术领域中普通技术人员所熟知的内容。本实施例中给定的一些特定参数仅作为示范，在不同的实施方式中该值可以相应地改变为合适的值。In order to make the purpose, technical solutions and advantages of the present disclosure clearer, the present invention will be further described in detail below in conjunction with specific embodiments and according to the accompanying drawings. It should be noted that, in the drawings or descriptions in the specification, the content not described and some English abbreviations are the content well known to those of ordinary skill in the art. Some specific parameters given in this embodiment are only for demonstration, and the values can be correspondingly changed to appropriate values in different implementations.

如图1所示的一种基于任务阶段复杂性的无人机自主能力评估方法，包括以下步骤：As shown in Figure 1, a method for evaluating the autonomous capability of UAVs based on the complexity of the mission phase includes the following steps:

步骤101，建立环境复杂性评估体系，对无人机系统在执行任务过程中的环境复杂度进行计算；环境复杂性评估体系主要针对影响飞行与任务效能两方面对环境复杂度进行评估，因此，环境复杂度具体包括影响飞行安全的地形复杂度与气象复杂度，以及以及影响任务效能的通信复杂度、目标识别复杂度与威胁复杂度，即如图2所示。Step 101, establishing an environmental complexity evaluation system to calculate the environmental complexity of the UAV system during the mission execution process; the environmental complexity evaluation system mainly evaluates the environmental complexity in terms of affecting flight and mission effectiveness. Therefore, Environmental complexity specifically includes terrain complexity and meteorological complexity that affect flight safety, as well as communication complexity, target recognition complexity, and threat complexity that affect mission effectiveness, as shown in Figure 2.

步骤102，建立任务复杂度评估体系，获取无人机系统在不同阶段的任务复杂度；首先将将无人机系统的任务分解量化，然后对分解后的任务进行复杂度评估。任务分解方法采用基于有限状态机的分解方法，将无人机系统中的每个无人机以及无人机上的各个功能每个模块视为一个有限状态机，将任务执行过程转化为一系列有限状态机的状态转换过程，进而进行子任务的复杂度评估。本实施例中将任务复杂度的评估指标量化为任务及战术行为、协同与协作、规划与分析、态势感知，即如图3所示。Step 102, establishing a task complexity evaluation system to obtain the task complexity of the UAV system at different stages; first, decompose and quantify the tasks of the UAV system, and then evaluate the complexity of the decomposed tasks. The task decomposition method adopts the decomposition method based on the finite state machine, regards each UAV in the UAV system and each module of each function on the UAV as a finite state machine, and transforms the task execution process into a series of finite The state transition process of the state machine, and then evaluate the complexity of the subtasks. In this embodiment, the evaluation indicators of task complexity are quantified into tasks and tactical behaviors, coordination and cooperation, planning and analysis, and situational awareness, as shown in FIG. 3 .

步骤103，建立无人机自主能力评估模型，将环境复杂度与任务复杂度输入无人机自主能力评估模型，无人机自主能力评估模型根据环境复杂度、任务复杂度与人机自主等级划分标准中的自主等级分类输出无人机自主等级。即以无人机自主能力评估模型作为分类器，人机自主等级划分标准中的自主等级分类作为无人机自主能力评估模型的分类标准，在无人机自主能力评估模型经过多次训练后输入当前无人机系统的环境复杂度与任务复杂度即能输出该无人机系统的无人机自主等级。Step 103, establish the UAV autonomous capability assessment model, input the environment complexity and task complexity into the UAV autonomy assessment model, and the UAV autonomy assessment model is divided according to the environment complexity, task complexity and human-machine autonomy level The autonomy level classification in the standard outputs the drone autonomy level. That is, the UAV autonomy evaluation model is used as the classifier, and the autonomy level classification in the human-machine autonomy classification standard is used as the classification standard of the UAV autonomy evaluation model. After multiple trainings, the UAV autonomy evaluation model is input The environment complexity and task complexity of the current UAV system can output the UAV autonomy level of the UAV system.

参考图4，步骤101中，地形复杂度的计算过程为：With reference to Fig. 4, in step 101, the calculation process of terrain complexity is:

步骤201，获取实时地形图片；Step 201, obtaining real-time terrain pictures;

步骤202，计算实时地形图片的图像熵与灰度共生矩阵的反差值；Step 202, calculating the image entropy of the real-time terrain image and the contrast value of the gray level co-occurrence matrix;

步骤203，对实时地形图片的图像熵与灰度共生矩阵的反差值分别进行归一化处理，获得图像熵值的归一化值，以及灰度共生矩阵的反差值的归一化值；Step 203, performing normalization processing on the image entropy of the real-time terrain image and the contrast value of the gray level co-occurrence matrix respectively, to obtain the normalized value of the image entropy value and the normalized value of the contrast value of the gray level co-occurrence matrix;

步骤204，根据图像熵值的归一化值与灰度共生矩阵的反差值的归一化值计算地形复杂度：Land＝0.2×Entropy+0.8×Contrast，其中，Land表示地形复杂度，Entropy表示图像熵值的归一化值，Contrast表示灰度共生矩阵的反差值的归一化值。Step 204, calculate the terrain complexity according to the normalized value of the image entropy value and the normalized value of the contrast value of the gray level co-occurrence matrix: Land=0.2×Entropy+0.8×Contrast, where Land represents the terrain complexity, and Entropy represents The normalized value of the image entropy value, Contrast represents the normalized value of the contrast value of the gray level co-occurrence matrix.

参考图5，威胁复杂度的计算过程为：Referring to Figure 5, the calculation process of threat complexity is:

步骤301，针对雷达、高炮和地空导弹三种战场防空火力建立了杀伤区域模型；In step 301, a kill zone model is established for the three battlefield air defense firepowers of radar, antiaircraft gun and surface-to-air missile;

步骤302，根据威胁点的位置在杀伤区域模型中绘制战场防空火力分布图；Step 302, drawing a battlefield air defense firepower distribution map in the kill zone model according to the position of the threat point;

步骤303，计算战场防空火力分布图中的安全区域比例，进而获得威胁复杂度。Step 303, calculating the proportion of the safe area in the battlefield air defense firepower distribution map, and then obtaining the threat complexity.

参考图6，气象复杂度的计算过程为：Referring to Figure 6, the calculation process of meteorological complexity is:

步骤401，选取风切变，风力等级，雷暴天气和降雨天气进行模糊综合评价。综合评价方法为：Step 401, select wind shear, wind level, thunderstorm weather and rainfall weather for fuzzy comprehensive evaluation. The comprehensive evaluation method is:

T_R＝F(U)→F(V)T _R ＝F(U)→F(V)

其中，F(*)为模糊变化，U为因素集，包括信息尽可能不交叉的评价指标；V为评语集，由各种不同评判等级形成的集合：Among them, F(*) is a fuzzy change, U is a factor set, including evaluation indicators whose information is not intersected as much as possible; V is a comment set, a set formed by various evaluation levels:

U＝{风，风切变，雷暴，降水}U = {wind, wind shear, thunderstorm, precipitation}

V＝{好，较好，中等，较差，很差}V={good, good, average, poor, very poor}

步骤402，针对评价结果出现的不符合常理的情况，采取隶属度次大的趋于评价等级较差的评价等级作为评价的结果，最终模糊综合评价的结果即为气象复杂度。Step 402 , in view of the unreasonable situation in the evaluation results, the evaluation grade with the second highest degree of membership tending to be inferior to the evaluation grade is used as the evaluation result, and the final fuzzy comprehensive evaluation result is the meteorological complexity.

参考图7，通信复杂度的计算过程为：Referring to Figure 7, the calculation process of communication complexity is:

步骤501，选取能够全面反映通信环境优劣的因素集和评语集。Step 501, selecting a factor set and a comment set that can fully reflect the quality of the communication environment.

这里因素集U和评语集V为：Here the factor set U and the comment set V are:

U＝{丢包率，误码率，时间延迟，中断}U={packet loss rate, bit error rate, time delay, interruption}

V＝{好，较好，中等，较差，差}V={good, better, average, poor, poor}

其中误码率计算方法为：The bit error rate calculation method is:

丢包率计算方法为：The calculation method of packet loss rate is:

步骤502，根据模糊综合评价的方法进行综合评价,最终模糊综合评价的评价结果即为通信复杂度。In step 502, a comprehensive evaluation is performed according to the fuzzy comprehensive evaluation method, and the final evaluation result of the fuzzy comprehensive evaluation is the communication complexity.

参考图8，目标识别复杂度的计算过程为：Referring to Figure 8, the calculation process of target recognition complexity is:

步骤601，针对目标混淆采取生成目标结构特征空间的方法进行度量；Step 601, taking the method of generating the feature space of the target structure to measure the target confusion;

目标混淆度RSS计算方法为：The calculation method of target confusion RSS is:

RSS＝[(μ_T-μ_B)²+σ_T ²]^1/2 RSS＝[(μ _T -μ _B ) ² +σ _T ² ] ^1/2

其中，σ_T为目标的灰度标准差，μ_T为目标的灰度均值，μ_B为背景的灰度均值；Among them, σ _T is the gray standard deviation of the target, μ _T is the gray mean value of the target, and μ _B is the gray level mean value of the background;

步骤602，针对目标遮掩采取计算目标与局部背景对比度的方法进行度量；Step 602, measure the target mask by calculating the contrast between the target and the local background;

目标遮掩度DTS计算方法为：The calculation method of target concealment degree DTS is as follows:

其中S_被遮挡为被遮挡面积，S_目标为目标面积。Among them, S _{is occluded} as the area to be occluded, and S is the _target area.

步骤603，根据目标混淆度和目标遮掩率计算得到目标识别复杂度，目标复杂度Target计算方法为：In step 603, the target recognition complexity is calculated according to the target confusion degree and the target concealment rate, and the calculation method of the target complexity Target is:

其中RSS为目标混消度，DTS为目标遮掩度，DFA为目标虚警度，目标虚警度计算方法为：Among them, RSS is the degree of target confusion, DTS is the degree of target concealment, DFA is the degree of false alarm of the target, and the calculation method of the degree of false alarm of the target is:

其中，n为疑似目标的数量。Among them, n is the number of suspected targets.

参考图9，步骤102中，任务复杂度的计算过程为：With reference to Fig. 9, in step 102, the calculation process of task complexity is:

步骤,701，将任务复杂度的评估指标量化为任务及战术行为、协同与协作、规划与分析、态势感知；Step 701, quantifying the evaluation indicators of task complexity into tasks and tactical behaviors, coordination and collaboration, planning and analysis, and situational awareness;

步骤702，求取步骤401中每一评估指标的主观复杂度与客观复杂度，通过加权计算主观复杂度与客观复杂度得到每一评估指标的综合复杂度；Step 702, obtain the subjective complexity and objective complexity of each evaluation index in step 401, and obtain the comprehensive complexity of each evaluation index by weighting the subjective complexity and objective complexity;

步骤703，综合四种评估指标的综合复杂度进而得到任务复杂度。Step 703, combining the comprehensive complexities of the four evaluation indicators to obtain the task complexity.

在步骤702中，每一评估指标的主观复杂度通过主观定权法来求取，客观复杂度复杂度通过客观定权法来求取，本实施中，主观定权法为可拓展层次分析法，客观定权法为熵权法，因此，对于任意一种评估指标的综合复杂度为：In step 702, the subjective complexity of each evaluation index is obtained through the subjective weighting method, and the objective complexity is obtained through the objective weighting method. In this implementation, the subjective weighting method is the scalable analytic hierarchy process , the objective weighting method is the entropy weighting method, therefore, the comprehensive complexity for any evaluation index is:

W_n＝α·W_1n+β·W_2n W _n =α·W _1n +β·W _2n

α+β＝1,α＞0,β＞0α+β=1, α>0, β>0

式中，W_n表示一种评估指标的综合复杂度；W_1n表示该评估指标通过可拓展层次分析法得到的主观复杂度；W_2n表示该评估指标通过熵权法得到的客观复杂度；In the formula, W _n represents the comprehensive complexity of an evaluation index; W _1n represents the subjective complexity of the evaluation index obtained through the scalable AHP; W _2n represents the objective complexity of the evaluation index obtained through the entropy weight method;

参考图10，本实施例中将无人机系统执行任务的过程分为八个阶段，通过上述的计算方法针对每种阶段进行任务复杂度的求取，其中EAHP指的即为可拓展层次分析法。Referring to Fig. 10, in this embodiment, the process of performing tasks by the UAV system is divided into eight stages, and the task complexity is calculated for each stage through the above-mentioned calculation method, where EAHP refers to Extensible Hierarchy Analysis Law.

步骤103中，人机自主等级划分标准中的自主等级分类包括：机为主模式、机主人辅模式、人主机辅模式与人为主模式。参考图11，R为机为主模式，指的是在有人机监督的前提下，无人机系统自主控制，实现分析态势、识别目标、决策与轨迹规划等工作；RH为机主人辅模式，指的是由有人机进行目标识别、授权决策请求、航机规划与任务控制工作，由无人机自主进行态势感知、决策请求、机载载重规划工作，为半自动控制模式；HR为人主机辅模式，指的是由无人机进行提示态势与目标、提供决策支持。离线重规划与自主飞控工作，由有人机进行态势感知、目标识别、任务决策与分配以及航迹规划工作；H为人为主模式，指的由有人机发布分析态势、识别目标、决策与轨迹规划的指令，由无人机来具体执行。其中，人为主模式自主等级为最低，需要有人机来实现目标识别、任务决策、航迹规划、底层控制等功能；人主机辅模式自主等级次之，目标识别需要有人机来做，任务决策和航迹规划为有人机和无人机共同来完成，飞行控制由无人机完成；机主人辅自主等级较高，由无人机进行自主飞行和态势感知，无人机做出的任务决策由有人机来授权，有人机来规划初始航迹，无人机做航迹重规划。机为主模式的自主等级最高，整个态势感知、目标识别、自主决策、航迹规划、飞行控制都有无人机完成。In step 103, the autonomy level classification in the human-machine autonomy level classification standard includes: machine-based mode, machine-master-assisted mode, human-master-assisted mode, and human-based mode. Referring to Figure 11, R is the aircraft-based mode, which means that under the premise of human-machine supervision, the UAV system is autonomously controlled to achieve situation analysis, target recognition, decision-making, and trajectory planning; RH is the aircraft-master-assistant mode, It refers to the semi-automatic control mode where the manned machine performs target recognition, authorized decision-making request, aircraft planning and mission control, and the unmanned aerial vehicle independently performs situation awareness, decision-making request, and airborne load planning. , which means that the UAV is used to prompt the situation and target, and provide decision support. Offline re-planning and autonomous flight control work, the situation awareness, target recognition, mission decision and assignment, and track planning are performed by manned machines; H is the human-based mode, which means that manned machines release and analyze the situation, identify targets, make decisions, and track The planned instructions are executed by the drone. Among them, the autonomy level of the human-master mode is the lowest, requiring manned machines to realize functions such as target recognition, mission decision-making, track planning, and bottom-level control; The trajectory planning is completed by manned aircraft and unmanned aerial vehicle, and the flight control is completed by the unmanned aerial vehicle. Authorization is done by the manned machine, the initial trajectory is planned by the manned machine, and the trajectory re-planned by the unmanned aerial vehicle. The machine-based mode has the highest level of autonomy, and the entire situation awareness, target recognition, autonomous decision-making, trajectory planning, and flight control are all completed by drones.

步骤103中，采用无监督学习中的自适应共振网络模型ART作为无人机自主能力评估模型，参考图12，无人机自主等级的求取过程为:In step 103, the self-adaptive resonance network model ART in unsupervised learning is used as the autonomous capability evaluation model of the UAV. Referring to Figure 12, the process of obtaining the autonomous level of the UAV is as follows:

步骤801，根据人机自主等级划分标准中的自主等级分类对自适应共振网络模型ART中识别层的每个神经元进行分类；采用训练分类器的方式对自适应共振网络模型ART中识别层的每个神经元进行分类。Step 801: Classify each neuron in the identification layer in the adaptive resonance network model ART according to the autonomous level classification in the human-computer autonomous level division standard; Each neuron performs classification.

步骤802，根据环境复杂度、任务复杂度计算自适应共振网络模型ART的最大激活值，最大激活值的求取过程为：Step 802, calculate the maximum activation value of the adaptive resonance network model ART according to the environment complexity and task complexity, the calculation process of the maximum activation value is:

式中，S为最大激活值；g为神经元非线性激励函数；x_i为神经元输入，n＝9，其中(x₁,x₂,x₃,x₄,x₅)为环境复杂度、(x₆,x₇,x₈,x₉)为任务复杂度；ω_i为前向传播权重。In the formula, S is the maximum activation value; g is the nonlinear activation function of the neuron; x _i is the input of the neuron, n=9, where (x ₁ , x ₂ , x ₃ , x ₄ , x ₅ ) is the complexity of the environment , (x ₆ , x ₇ , x ₈ , x ₉ ) are the task complexity; ω _i is the forward propagation weight.

步骤803，将最大激活值识别层的每个神经元进行比较，距离最大激活值最近的神经元对应的自主等级分类即为无人机自主等级。In step 803, each neuron in the maximum activation value recognition layer is compared, and the autonomous level classification corresponding to the neuron closest to the maximum activation value is the autonomous level of the drone.

以上包含了本发明优选实施例的说明，这是为了详细说明本发明的技术特征，并不是想要将发明内容限制在实施例所描述的具体形式中，依据本发明内容主旨进行的其他修改和变型也受本专利保护。本发明内容的主旨是由权利要求书所界定，而非由实施例的具体描述所界定。The description of the preferred embodiment of the present invention is included above, which is to describe the technical characteristics of the present invention in detail, and is not intended to limit the content of the invention to the specific form described in the embodiment. Other modifications and Variations are also protected by this patent. The gist of the present invention is defined by the claims rather than by the detailed description of the embodiments.