A multi-dimensional task complexity evaluation method

Abstract

The application discloses a multi-dimension task complexity evaluation method, and particularly relates to the following steps: firstly, a task constitution, condition, completion requirement, risk and environment complexity evaluation system is established to evaluate the complexity of the task constitution, condition, completion requirement, risk and environment of the intelligent device; then, a multi-dimension comprehensive task complexity evaluation model of the task constitution, condition, completion requirement, risk and environment is established, and a fuzzy self-adaptive resonance neural network model is constructed; finally, the task complexity evaluation result is obtained by using the comprehensive task complexity evaluation model and the fuzzy self-adaptive resonance neural network model. The application improves the adaptability of the intelligent device to complex and changeable tasks, improves the scientificity, accuracy and coverage of the task complexity evaluation, and provides a theoretical basis for improving the sensing capability of the intelligent device.

Description

Technical Field

[0001] This invention relates to the field of equipment intelligence level assessment technology, and in particular to a multi-dimensional task complexity assessment method. Background Technology

[0002] In recent years, with the accelerated development of intelligent devices, they have been gradually developed and deployed in various fields. Objectively measuring the level of intelligence of these devices is an urgent requirement for relevant management and research departments to accurately grasp the starting point of intelligent device development, clearly set intelligent device development goals, and scientifically plan the path of intelligent device development. Different levels of device intelligence result in different levels of task complexity; therefore, task complexity is an important indicator for evaluating the level of device intelligence.

[0003] The role of intelligent devices in disaster relief and rescue operations is becoming increasingly important. When executing emergency missions, it's necessary to consider factors such as mission composition, conditions, completion requirements, risks, and environment. This necessitates that intelligent devices possess stronger perception capabilities, as well as collaborative and communication capabilities with similar and different types of devices. Faced with complex and ever-changing tasks, it's crucial to accurately determine whether a cluster of intelligent devices can complete the task. Therefore, task complexity is one of the effective metrics for evaluating dynamically changing tasks. Establishing scientific and comprehensive assessment indicators and models for task complexity will support the evaluation of device intelligence levels.

[0004] From a practical application perspective, the task complexity assessment of the intelligence level of urban rescue equipment is a typical case study in the research of task complexity assessment of equipment intelligence level, and it has high research value. The task complexity assessment results of urban rescue equipment can provide strategies for selecting urban rescue equipment platforms to perform the task. Using a multi-dimensional task complexity assessment model, the situational awareness capabilities of urban rescue equipment in completing tasks of different complexities, such as target acquisition, search and rescue operations, traffic and pedestrian monitoring, and emergency logistics delivery, are evaluated to obtain its intelligence level, thereby identifying its shortcomings and weaknesses to improve the intelligent perception capabilities of urban rescue equipment. How to assess the strength of the situational awareness capabilities of urban rescue equipment through influencing factors such as task composition, completion requirements, conditions, risks, and environment, so as to help urban rescue equipment maximize its situational awareness capabilities in actual combat, has become a direction for research on the level assessment of intelligent urban rescue equipment. Influencing factors such as task composition, conditions, completion requirements, risks, and environment are closely related to the perception level of urban rescue equipment systems under certain circumstances. Therefore, assessing the perception level of urban rescue equipment under different task complexities is of great significance for evaluating the intelligence level performance of urban rescue equipment.

[0005] Currently, there is limited research on task complexity for assessing the intelligence level of equipment, both domestically and internationally. Some scholars study equipment from the perspective of environmental complexity, such as meteorology and geography. For example, Chang Sha et al. proposed an assessment index for the complexity of simulation environments and established an environmental complexity assessment model and an environmental complexity classification model based on matrix-type grey relational theory and grey clustering method, respectively. Other scholars study the task complexity of intelligent driving. For instance, Yin Lu et al. designed a quantified assessment method for the complexity of scenario-based driving tasks by comprehensively using subjective load evaluation method and information entropy theory from three aspects: action composition, information perception, and judgment assessment. They established a quantified model for the complexity of scenario tasks applicable to seven types and fifty-four types of vehicle testing and verification. Zhu Bing established an assessment system for the importance of scenario elements based on a six-layer scenario model, analyzed the complexity mapping relationship between various elements of the test scenario and the perception, decision-making, and execution systems, and proposed a method for assessing the complexity of autonomous vehicle test scenarios.

[0006] However, existing task complexity assessments lack research on constructing task complexity indicators and models from multiple dimensions such as task composition, conditions, completion requirements, risks, and environment. There is an urgent need to invent a measurement model that can comprehensively consider task composition, task conditions, task completion requirements, task risks, and task environment complexity, so as to provide methods and technical means for assessing the intelligence level of equipment. Summary of the Invention

[0007] The purpose of this invention is to provide a scientific, comprehensive, accurate, and adaptable method for evaluating the task complexity of intelligent devices.

[0008] The technical solution to achieve the purpose of this invention is: a multi-dimensional task complexity evaluation method, comprising the following steps:

[0009] Step 1: Establish a task composition complexity assessment system to assess the task composition complexity of subtasks and their relationships in the execution of tasks by intelligent devices.

[0010] Step 2: Establish a task condition complexity assessment system to assess the task condition complexity of the number and speed of intelligent devices performing tasks.

[0011] Step 3: Establish a task completion requirement complexity assessment system to assess the task completion requirement complexity of intelligent devices in terms of completion time, completion effect, and task scope.

[0012] Step 4: Establish a task risk complexity assessment system to assess the task risk complexity of the target moving speed, the distance between the target and the intelligent device, and the number of targets when the intelligent device performs the task.

[0013] Step 5: Establish a task environment complexity assessment system to assess the task environment complexity of intelligent devices performing tasks, including meteorological and geographical environments.

[0014] Step 6: Establish a task composition, conditions, completion requirements, risks, and environmental complexity assessment model, and construct a fuzzy adaptive resonant neural network model to comprehensively obtain the task complexity assessment results.

[0015] Compared with the prior art, the present invention has the following significant advantages: (1) By establishing a complexity assessment system with five dimensions of intelligent device task composition, task conditions, task completion requirements, task risks, and task environment, the complexity of intelligent device system in the process of performing tasks is obtained, which improves the adaptability to complex and variable task assessment and improves the scientificity and coverage of intelligent device task complexity assessment; (2) Based on the intelligent perception level classification standard, combined with the complexity of five dimensions of task composition, conditions, completion requirements, risks, and environment, the fuzzy adaptive resonant neural network model is used to obtain the task complexity assessment results for equipment intelligence level testing, which improves the accuracy of intelligent device task complexity assessment; (3) Using a multi-dimensional task complexity assessment model, the complexity of urban rescue equipment with different intelligence levels in completing tasks is assessed, so as to indirectly verify its intelligence level and find its capability shortcomings and weaknesses, providing a theoretical basis for improving the intelligent perception capability of urban rescue equipment. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a multi-dimensional task complexity evaluation method according to the present invention.

[0017] Figure 2 This is a schematic diagram of the process for evaluating the complexity of tasks in intelligent devices in this invention.

[0018] Figure 3 This is a schematic diagram of the process for evaluating the task complexity of intelligent devices in this invention.

[0019] Figure 4 This is a schematic diagram of the process for evaluating the complexity of intelligent device task completion requirements in this invention.

[0020] Figure 5 This is a schematic diagram of the process for assessing the task risk complexity of intelligent devices in this invention.

[0021] Figure 6 This is a schematic diagram of the process for evaluating the task environment complexity of intelligent devices in this invention.

[0022] Figure 7 This is a schematic diagram of the process for evaluating the overall task complexity of intelligent devices in this invention. Detailed Implementation

[0023] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0024] like Figure 1 As shown, the present invention provides a multi-dimensional task complexity evaluation method, comprising the following steps:

[0025] Step 1: Establish a task composition complexity assessment system to assess the task composition complexity of subtasks and their relationships in the execution of tasks by intelligent devices.

[0026] Step 2: Establish a task condition complexity assessment system to assess the task condition complexity of the number and speed of intelligent devices performing tasks.

[0027] Step 3: Establish a task completion requirement complexity assessment system to assess the task completion requirement complexity of intelligent devices in terms of completion time, completion effect, and task scope.

[0028] Step 4: Establish a task risk complexity assessment system to assess the task risk complexity of the target moving speed, the distance between the target and the intelligent device, and the number of targets when the intelligent device performs the task.

[0029] Step 5: Establish a task environment complexity assessment system to assess the task environment complexity of intelligent devices performing tasks, including meteorological and geographical environments.

[0030] Step 6: Establish a task composition, conditions, completion requirements, risks, and environmental complexity assessment model, and construct a fuzzy adaptive resonant neural network model to comprehensively obtain the task complexity assessment results.

[0031] As a specific example, step 1 involves establishing a task composition complexity assessment system to evaluate the task composition complexity of sub-tasks and their relationships within a task performed by a smart device. Figure 2 As shown, the details are as follows:

[0032] Step 1.1: Break down the device intelligent sensing task into "subtask 1, subtask 2, subtask 3, ... subtask n", and obtain the number of subtasks;

[0033] Step 1.2: Based on the task structure network diagram and tree diagram, analyze the relationships between subtasks, and calculate the coupling quantity between subtasks using the network node coupling calculation formula;

[0034] Step 1.3: Determine the complexity assessment results of the intelligent device task composition based on the number of subtasks and their relationships.

[0035] As a specific example, step 2 involves establishing a task condition complexity evaluation system to assess the task condition complexity based on the number and speed of intelligent devices performing tasks. Figure 3 As shown, the details are as follows:

[0036] Step 2.1: Obtain the number of smart devices executing the task;

[0037] Step 2.2: Obtain the speed of the smart device executing the task;

[0038] Step 2.3: Based on the number and speed of the intelligent devices performing the task, obtain the evaluation results of the conditional complexity of the intelligent devices.

[0039] As a specific example, step 3 describes establishing a task completion requirement complexity assessment system to evaluate the task completion requirement complexity of intelligent devices based on the completion time, completion effect, and task scope. Figure 4 As shown, the details are as follows:

[0040] Step 3.1: Obtain the task completion time requirements for smart devices;

[0041] Step 3.2: Obtain the coverage requirements for the tasks performed by the smart device;

[0042] Step 3.3: Obtain the completion rate of tasks performed by smart devices;

[0043] Step 3.4: Based on the completion time, coverage, and completion rate of the intelligent device personnel, obtain the complexity assessment results of the intelligent device task completion requirements.

[0044] As a specific example, step 4 describes establishing a task risk complexity assessment system to evaluate the task risk complexity of the target moving speed, the distance between the target and the intelligent device, and the scale of the target quantity. Figure 5 As shown, the details are as follows:

[0045] Step 4.1: Obtain the target's moving speed and distance information between the target and the smart device during the task execution process;

[0046] Step 4.2: Obtain the target quantity and scale information during the execution of tasks by intelligent devices;

[0047] Step 4.3: Based on the target's moving speed, distance to the smart device, and target quantity during the smart device's task execution, obtain the smart device task risk complexity assessment result.

[0048] As a specific example, step 5 describes establishing a task environment complexity assessment system to evaluate the task environment complexity of intelligent devices performing tasks, including meteorological and geographical conditions. Figure 6 As shown, the details are as follows:

[0049] Step 5.1: Obtain visibility, wind force, and light intensity information for the day the smart device performs the task, and calculate meteorological environmental values ​​for fuzzy comprehensive evaluation;

[0050] Step 5.2: Obtain images of terrain undulations, vegetation cover, and building cover during the task execution process of homogeneous and heterogeneous intelligent devices, calculate the image entropy value and gray-level co-occurrence matrix, normalize the entropy value and matrix contrast value, and calculate the geographical complexity.

[0051] Step 5.3: Based on the complexity of the meteorological and geographical environment when the intelligent device performs the task, obtain the assessment result of the complexity of the intelligent device's task environment.

[0052] As a specific example, step 6 involves establishing a task composition, conditions, completion requirements, risks, and environmental complexity assessment model, and constructing a fuzzy adaptive resonant neural network model to comprehensively derive the task complexity assessment result, such as... Figure 7 As shown, the details are as follows:

[0053] Step 6.1: Decompose and quantify the five dimensions of task complexity—task composition, conditions, completion requirements, risks, and environmental complexity—to construct a three-level evaluation index system, as follows:

[0054] The subjective and objective complexity of each evaluation indicator in the task composition, task conditions, task completion requirements, task risks, and task environment complexity are calculated. The comprehensive complexity of each evaluation indicator is obtained by weighting the subjective and objective complexity.

[0055] The subjective complexity of each evaluation indicator is obtained using a subjective weighting method, and the objective complexity is obtained using an objective weighting method. In this embodiment, the subjective weighting method is the extensible analytic hierarchy process (AHP), and the objective weighting method is the entropy weighting method. Therefore, the comprehensive complexity for any evaluation indicator is:

[0056]

[0057]

[0058] in, This indicates the overall complexity of an evaluation metric. This indicates the subjective complexity of the evaluation indicator obtained through the extended analytic hierarchy process. This indicates the objective complexity of the evaluation index obtained through the entropy weight method;

[0059] Step 6.2: Establish the mapping relationship between task composition, conditions, completion requirements, risks, environmental complexity and overall task complexity, as well as the mapping relationship between overall task complexity and perception level classification criteria, and construct a fuzzy adaptive resonant neural network model.

[0060] Step 6.3: Based on the task composition, conditions, completion requirements, risks, and environmental complexity, use the fuzzy adaptive resonant neural network model to obtain the comprehensive task complexity evaluation result, as follows:

[0061] Based on the specific task, a fuzzy adaptive resonant neural network model is used to evaluate the task complexity. The activation function of the fuzzy adaptive resonant network model is obtained based on the complexity evaluation values ​​considering task composition, conditions, completion requirements, risks, and environment. The probability of different task complexities is then calculated using the following formula:

[0062]

[0063] Where y represents the probability of different task complexities; For nonlinear activation functions of neurons; For neuron input, where, These are respectively represented as task composition, task conditions, task completion requirements, task risks, and the complexity of the task environment; Forward propagation weights;

[0064] The activation function outputs the probability value of the task complexity level, and the one with the highest probability corresponds to the task complexity evaluation result.

[0065] Example

[0066] This embodiment provides a task complexity assessment method for evaluating the perception capabilities of urban rescue equipment, including the following steps:

[0067] Step 1: Establish a task composition complexity assessment system, which breaks down the rescue mission executed by the intelligent device system into four independent sub-tasks: information understanding, information acquisition, information processing, and information feedback. The task composition complexity assessment system mainly evaluates the complexity of the task structure. Therefore, the task composition complexity includes the number and type of sub-tasks, as well as the relationship between the sub-tasks.

[0068] Step 2: Establish a task condition complexity assessment system to calculate the condition complexity of urban rescue equipment performing tasks. The task condition complexity assessment mainly focuses on assessing the factors affecting whether urban rescue equipment can carry out tasks. Therefore, the task condition complexity includes the scale of rescue equipment, the number of operators, etc.

[0069] Step 3: Establish a task completion requirement complexity assessment system to calculate the conditional complexity of urban rescue equipment performing tasks. The task completion requirement complexity assessment mainly focuses on evaluating the factors affecting the task effectiveness of urban rescue equipment. Therefore, the task completion requirement complexity includes completion time, coverage area, and completion effect.

[0070] Step 4: Establish a task risk complexity assessment system to calculate the risk complexity of urban rescue equipment in performing tasks. The task risk complexity assessment mainly focuses on evaluating the indicators that affect the completion of urban rescue equipment tasks. Therefore, task risk complexity includes the target quantity, type, location, etc.

[0071] Step 5: Establish a mission environment complexity assessment system to calculate the environmental complexity of the rescue equipment system during mission execution. The environmental complexity assessment system primarily evaluates environmental complexity in terms of its impact on flight and mission effectiveness. Therefore, environmental complexity includes meteorological and geographical complexity affecting rescue distance, communication capabilities, and sensing capabilities. The specific process for establishing the mission environment complexity assessment system is as follows:

[0072] Step 5.1: Obtain visibility, wind speed, and light intensity information for the day the smart device performs the task, calculate meteorological environmental values, and conduct a fuzzy comprehensive evaluation, as follows:

[0073] Step 5.1.1: Select visibility, wind force level, thunderstorm weather, and precipitation weather for fuzzy comprehensive evaluation; the comprehensive evaluation method is as follows:

[0074]

[0075] in, For fuzzy changes, = {wind, wind shear, thunderstorm, precipitation} is the factor set, which includes evaluation indicators with information that does not overlap as much as possible; = {Good, Fairly Good, Average, Poor, Very Poor} is a set of comments, which is a collection of various different rating levels;

[0076] Step 5.1.2: Take the evaluation level with the highest membership degree as the evaluation result. The final result of the fuzzy comprehensive evaluation is the meteorological complexity.

[0077] Step 5.2: Obtain images of terrain undulations, vegetation cover, and building cover during the task execution process of homogeneous and heterogeneous intelligent devices. Calculate the image entropy value and gray-level co-occurrence matrix, and normalize the entropy value and matrix inverse value to calculate the geographical complexity, as follows:

[0078] Step 5.2.1: Obtain real-time terrain images;

[0079] Step 5.2.2: Calculate the inverse value between the image entropy and the gray-level co-occurrence matrix of the real-time terrain image;

[0080] Step 5.2.3: Normalize the image entropy and the contrast value of the gray-level co-occurrence matrix of the real-time terrain image to obtain the normalized value of the image entropy and the normalized value of the contrast value of the gray-level co-occurrence matrix.

[0081] Step 5.2.4: Calculate the terrain complexity based on the normalized value of the image entropy and the normalized value of the contrast of the gray-level co-occurrence matrix: L = 0.2 × E + 0.8 × C, where L represents the terrain complexity, E represents the normalized value of the image entropy, and C represents the normalized value of the contrast of the gray-level co-occurrence matrix.

[0082] Step 5.3: Based on the complexity of the meteorological and geographical environment when the intelligent device performs the task, obtain the assessment result of the complexity of the intelligent device's task environment.

[0083] Step 6: Establish a task composition, conditions, completion requirements, risks, and environmental complexity assessment model, and construct a fuzzy adaptive resonant neural network model to comprehensively obtain the task complexity assessment results, as follows:

[0084] Step 6.1: Establish a multi-dimensional task complexity assessment system to obtain the complexity level of tasks performed by urban rescue equipment, and decompose and quantify the indicators of task composition, task conditions, task completion requirements, task risks, and task environment when urban rescue equipment performs a specific task.

[0085] Step 6.2: Establish the mapping relationship between task composition, conditions, completion requirements, risks, environmental complexity, and overall task complexity, as well as the mapping relationship between overall task complexity and perception level classification criteria. Construct a fuzzy adaptive resonant neural network model. The input layer of the neural network contains the task complexity evaluation results across five dimensions, the output is the rescue task complexity level, and the hidden layer contains the mapping relationship between the five dimensions of complexity and the task complexity level. In this embodiment, the evaluation index of the perception task complexity of rescue equipment is quantified into five dimensions: task composition, conditions, completion requirements, risks, and environment, as detailed below:

[0086] Establish the mapping relationship between task composition, conditions, completion requirements, risks, environmental complexity and overall task complexity, as well as the mapping relationship between overall task complexity and perception level classification criteria. Construct a fuzzy adaptive resonant neural network model. The input layer of the neural network is the evaluation result of the five dimensions of task complexity, the output layer is the task complexity evaluation result, and the hidden layer is the mapping relationship between the five dimensions of complexity and the task complexity level.

[0087] The intelligent sensing level classification in the equipment intelligence level classification standard includes: machine-centric mode, machine-master-auxiliary mode, human-master-auxiliary mode, and human-centric mode.

[0088] The machine-centric mode is a mode in which intelligent rescue equipment is autonomously controlled under human supervision to acquire, process, or understand information about the rescue target.

[0089] The master-assisted mode is a semi-automatic control mode in which a man assists the rescue equipment in target identification, authorization decision-making requests, planning and control, while the rescue equipment semi-autonomously acquires, processes or understands the rescue target information.

[0090] In the human-machine-assisted mode, the rescue equipment provides target prompts and decision support, while the rescue equipment assists the human in acquiring, processing, or understanding information about the rescue target.

[0091] The human-centric model involves humans issuing instructions for situation analysis, target identification, decision-making, and trajectory planning, which are then executed by rescue equipment.

[0092] Among them, the human-centric mode has the lowest level of autonomy, requiring human intervention to perform functions such as target identification, task decision-making, trajectory planning, and low-level control; the human-host-assisted mode has the next lowest level of autonomy, requiring human intervention for target identification, and the acquisition, processing, or understanding of rescue target information is jointly completed by humans and rescue equipment; the aircraft-host-assisted mode has a relatively high level of autonomy, with the rescue equipment autonomously acquiring, processing, or understanding target information, and the task decisions made by the equipment being authorized by humans, with humans planning the initial decisions and the equipment making decisions and replanning; the aircraft-centric mode has the highest level of autonomy, with the entire acquisition, processing, or understanding of rescue target information being completed autonomously by the rescue equipment.

[0093] Step 6.3: Input the task composition, conditions, completion requirements, and risk complexity assessment results into the fuzzy self-resonant neural network model. Based on the mapping relationship between task composition, conditions, completion requirements, risk, and task environment complexity and task complexity classification level, output the rescue task complexity assessment result. That is, using the fuzzy self-resonant neural network model as a classifier, and the complexity assessment results of the five dimensions of the rescue task and the task complexity level as classification criteria, after the task complexity assessment model has been trained multiple times, inputting the current rescue task composition, conditions, completion requirements, risk, and task environment complexity will output the urban rescue task complexity assessment result, as detailed below:

[0094] Based on the complexity assessment results of five dimensions of urban rescue missions, the maximum activation value of the fuzzy adaptive resonant network model is selected. The formula for calculating the maximum activation value is as follows:

[0095]

[0096] in, Probabilities for different task complexities; For nonlinear activation functions of neurons; For neuron input, where These are respectively represented as task composition, task conditions, task completion requirements, task risks, and the complexity of the task environment; Forward propagation weights;

[0097] Each neuron in the maximum activation value recognition layer is compared, and the complexity level corresponding to the neuron closest to the maximum activation value is the result of the rescue mission complexity assessment.

[0098] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A multi-dimensional task complexity evaluation method, characterized in that, Includes the following steps: Step 1: Establish a task composition complexity assessment system to assess the task composition complexity of subtasks and their relationships in the execution of tasks by intelligent devices. Step 2: Establish a task condition complexity assessment system to assess the task condition complexity of the number and speed of intelligent devices performing tasks. Step 3: Establish a task completion requirement complexity assessment system to assess the task completion requirement complexity of intelligent devices in terms of completion time, completion effect, and task scope. Step 4: Establish a task risk complexity assessment system to assess the task risk complexity of the target moving speed, the distance between the target and the intelligent device, and the number of targets when the intelligent device performs the task. Step 5: Establish a task environment complexity assessment system to assess the task environment complexity of intelligent devices performing tasks, including meteorological and geographical environments. Step 6: Establish a task composition, conditions, completion requirements, risks, and environmental complexity assessment model, and construct a fuzzy adaptive resonant neural network model to comprehensively obtain the task complexity assessment results.

2. The multi-dimensional task complexity evaluation method according to claim 1, characterized in that, Step 1 describes establishing a task composition complexity assessment system to evaluate the task composition complexity of subtasks and their relationships in the execution of tasks by intelligent devices, as detailed below: Step 1.1: Break down the device intelligent sensing task into "subtask 1, subtask 2, subtask 3, ... subtask n", and obtain the number of subtasks; Step 1.2: Based on the task structure network diagram and tree diagram, analyze the relationships between subtasks, and calculate the coupling quantity between subtasks using the network node coupling calculation formula; Step 1.3: Determine the complexity assessment results of the intelligent device task composition based on the number of subtasks and their relationships.

3. The multi-dimensional task complexity evaluation method according to claim 1, characterized in that, Step 2 describes establishing a task condition complexity evaluation system to evaluate the task condition complexity based on the number and speed of intelligent devices executing tasks, as detailed below: Step 2.1: Obtain the number of smart devices executing the task; Step 2.2: Obtain the speed of the smart device executing the task; Step 2.3: Based on the number and speed of the intelligent devices performing the task, obtain the evaluation results of the conditional complexity of the intelligent devices.

4. The multi-dimensional task complexity evaluation method according to claim 1, characterized in that, Step 3 describes establishing a task completion requirement complexity assessment system to evaluate the task completion requirement complexity of intelligent devices based on the completion time, completion effect, and task scope. The specific details are as follows: Step 3.1: Obtain the task completion time requirements for smart devices; Step 3.2: Obtain the coverage requirements for the tasks performed by the smart device; Step 3.3: Obtain the completion rate of tasks performed by smart devices; Step 3.4: Based on the completion time, coverage, and completion rate of the intelligent device personnel, obtain the complexity assessment results of the intelligent device task completion requirements.

5. The multi-dimensional task complexity evaluation method according to claim 1, characterized in that, Step 4 describes establishing a task risk complexity assessment system to evaluate the target movement speed, distance to the intelligent device, and number of targets involved in the task, as detailed below: Step 4.1: Obtain the target's moving speed and distance information between the target and the smart device during the task execution process; Step 4.2: Obtain the target quantity and scale information during the execution of tasks by intelligent devices; Step 4.3: Based on the target's moving speed, distance to the smart device, and target quantity during the smart device's task execution, obtain the smart device task risk complexity assessment result.

6. The multi-dimensional task complexity evaluation method according to claim 1, characterized in that, Step 5 describes establishing a task environment complexity assessment system to evaluate the meteorological and geographical environment in which intelligent devices perform tasks. The specific details are as follows: Step 5.1: Obtain visibility, wind force, and light intensity information for the day the smart device performs the task, and calculate meteorological environmental values ​​for fuzzy comprehensive evaluation; Step 5.2: Obtain images of terrain undulations, vegetation cover, and building cover during the task execution process of homogeneous and heterogeneous intelligent devices, calculate the image entropy value and gray-level co-occurrence matrix, normalize the entropy value and matrix contrast value, and calculate the geographical complexity. Step 5.3: Based on the complexity of the meteorological and geographical environment when the intelligent device performs the task, obtain the assessment result of the complexity of the intelligent device's task environment.

7. The multi-dimensional task complexity evaluation method according to claim 1, characterized in that, Step 6 involves establishing a task composition, conditions, completion requirements, risks, and environmental complexity assessment model, and constructing a fuzzy adaptive resonant neural network model to comprehensively derive the task complexity assessment result, as detailed below: Step 6.1: Decompose and quantify the five dimensions of task complexity, including task composition, conditions, completion requirements, risks, and environmental complexity, and construct a three-level evaluation index system. Step 6.2: Establish the mapping relationship between task composition, conditions, completion requirements, risks, environmental complexity and overall task complexity, as well as the mapping relationship between overall task complexity and perception level classification criteria, and construct a fuzzy adaptive resonant neural network model. Step 6.3: Based on the task composition, conditions, completion requirements, risks, and environmental complexity, use the fuzzy adaptive resonant neural network model to obtain the comprehensive task complexity evaluation result.

8. The multi-dimensional task complexity evaluation method according to claim 7, characterized in that, Step 6.1 involves decomposing and quantifying the five dimensions of task complexity—task composition, conditions, completion requirements, risks, and environmental complexity—to construct a three-level evaluation index system, as detailed below: The subjective and objective complexity of each evaluation indicator in the task composition, task conditions, task completion requirements, task risks, and task environment complexity are calculated. The comprehensive complexity of each evaluation indicator is obtained by weighting the subjective and objective complexity.

9. The multi-dimensional task complexity evaluation method according to claim 7, characterized in that, Step 6.2 establishes the mapping relationship between task composition, conditions, completion requirements, risks, environmental complexity and overall task complexity, as well as the mapping relationship between overall task complexity and perception level classification criteria, and constructs a fuzzy adaptive resonant neural network model, as detailed below: Establish the mapping relationship between task composition, conditions, completion requirements, risks, environmental complexity and overall task complexity, as well as the mapping relationship between overall task complexity and perception level classification criteria. Construct a fuzzy adaptive resonant neural network model. The input layer of the neural network is the evaluation result of the five dimensions of task complexity, the output layer is the task complexity evaluation result, and the hidden layer is the mapping relationship between the five dimensions of complexity and the task complexity level.

10. The multi-dimensional task complexity evaluation method according to claim 7, characterized in that, Step 6.3 describes the use of a fuzzy adaptive resonant neural network model to obtain a comprehensive task complexity evaluation result based on the task composition, conditions, completion requirements, risks, and environmental complexity, as detailed below: Based on the specific task, a fuzzy adaptive resonant neural network model is used to evaluate the task complexity. The activation function of the fuzzy adaptive resonant network model is obtained based on the complexity evaluation values ​​of the task composition, conditions, completion requirements, risks, and environment. The probability of different task complexities is calculated using the following formula: Where y represents the probability of different task complexities; For nonlinear activation functions of neurons; For neuron input, where These are respectively represented as task composition, task conditions, task completion requirements, task risks, and the complexity of the task environment; Forward propagation weights; The activation function outputs the probability value of the task complexity level, and the one with the highest probability corresponds to the task complexity evaluation result.

Images