Civil heterogeneous unmanned cluster control method and system based on distributed control
Through the distributed control method, the task scheduling algorithm and digital twin control decision-making module are used to solve the problems of poor flexibility and strong dependence of the drone cluster under centralized control mode, and efficient collaborative operation of heterogeneous drone clusters is achieved and improve the effectiveness of task execution.
Patent Information
- Application Number
- CN202411052606.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-02
- Publication Date
- 2024-11-19
AI Technical Summary
The existing civil drone cluster control adopts a centralized method, resulting in strong dependence and poor flexibility in central nodes, making it difficult to fully utilize the characteristics and advantages of heterogeneous drones, resulting in poor task execution results.
The distributed control method is adopted to obtain the characteristic data of heterogeneous drones, task scheduling algorithms and digital twin control decision-making modules, and to realize task scheduling and control decision-making, reduce dependence on central nodes, and improve task execution efficiency and accuracy.
Through distributed control methods, efficient collaborative operations of heterogeneous drone clusters can be achieved, task execution effect is improved, dependence on central nodes is reduced, and task execution flexibility and reliability are improved.
Smart Images

Figure CN118982188A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of civilian unmanned aerial vehicle control, and in particular to a civilian heterogeneous unmanned aerial vehicle cluster control method and system based on distributed control. Background Art
[0002] With the rapid development of civil drone technology, the collaborative operation of civil drone clusters has received more and more attention. For example, tasks such as agricultural plant protection and power inspection often require multiple drones to work together. In the prior art, the control of civil drone clusters mainly adopts a centralized control method, that is, the central node uniformly dispatches and commands the actions of each drone. However, the centralized control method is highly dependent on the central node. Once the central node fails, the entire cluster will be difficult to work normally, and the robustness of the control system is poor. At the same time, centralized control lacks flexibility, making it difficult to give full play to the characteristics and advantages of heterogeneous drones and achieve efficient use of resources. Therefore, when civil drone clusters in the prior art perform tasks, the task execution effect is poor, and it is difficult to meet the growing practical application needs. Summary of the invention
[0003] This application provides a civilian heterogeneous unmanned cluster control method and system based on distributed control, aiming to solve the technical problems in the prior art that the centralized control of civilian heterogeneous unmanned clusters has poor flexibility, strong dependence on central nodes, and leads to poor task execution effects.
[0004] In view of the above problems, the present application provides a civilian heterogeneous unmanned cluster control method and system based on distributed control.
[0005] The first aspect disclosed in the present application provides a civilian heterogeneous unmanned cluster control method based on distributed control, the method comprising: obtaining M civilian heterogeneous unmanned control units of a civilian heterogeneous unmanned cluster, wherein M is a positive integer greater than 1; respectively collecting basic information and fixed civilian task information of the M civilian heterogeneous unmanned control units to obtain characteristic data of the M control units; obtaining a newly added civilian task scheduling instruction, wherein the newly added civilian task scheduling instruction includes newly added civilian task information and newly added civilian task time constraints; based on a civilian cluster task scheduling algorithm, performing task scheduling on the M civilian heterogeneous unmanned control units according to the characteristic data of the M control units, the newly added civilian task information and the newly added civilian task time constraints to obtain a civilian cluster task scheduling scheme, wherein the civilian cluster task scheduling scheme includes M unit allocation subtasks corresponding to the M civilian heterogeneous unmanned control units; based on a digital twin control decision module, performing control decision analysis on the M civilian heterogeneous unmanned control units according to the M unit allocation subtasks to obtain a distributed control scheme, wherein the distributed control scheme includes M unit task control decisions; and executing control of the M civilian heterogeneous unmanned control units according to the distributed control scheme.
[0006] Another aspect disclosed in the present application provides a civilian heterogeneous unmanned cluster control system based on distributed control, the system comprising: a control unit acquisition component for obtaining M civilian heterogeneous unmanned control units of a civilian heterogeneous unmanned cluster, wherein M is a positive integer greater than 1; an information collection component for respectively collecting basic information and fixed civilian task information of the M civilian heterogeneous unmanned control units to obtain characteristic data of the M control units; an instruction acquisition component for obtaining a newly added civilian task scheduling instruction, wherein the newly added civilian task scheduling instruction includes newly added civilian task information and newly added civilian task time constraints; a scheduling scheme acquisition component for, based on a civilian cluster task scheduling algorithm, obtaining a scheduling scheme according to the M civilian heterogeneous unmanned control units; a scheduling scheme acquisition component for obtaining a newly added civilian task scheduling instruction ... The control unit characteristic data, the newly added civilian task information and the newly added civilian task time constraints are used to schedule the M civilian heterogeneous unmanned control units to obtain a civilian cluster task scheduling plan, wherein the civilian cluster task scheduling plan includes M unit allocation subtasks corresponding to the M civilian heterogeneous unmanned control units; a control decision analysis component is used to perform control decision analysis on the M civilian heterogeneous unmanned control units according to the M unit allocation subtasks based on the digital twin control decision module to obtain a distributed control plan, wherein the distributed control plan includes M unit task control decisions; a control plan execution component is used to execute the control of the M civilian heterogeneous unmanned control units according to the distributed control plan.
[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:
[0008] By adopting M civil heterogeneous unmanned control units of civil heterogeneous unmanned clusters, the control object of the heterogeneous UAV cluster is obtained, which lays the foundation for subsequent distributed control; the basic information and fixed civil task information of M civil heterogeneous unmanned control units are collected respectively, and the characteristic data of M control units are obtained to provide a basis for subsequent task scheduling and control decision-making; the scheduling instructions of newly added civil tasks are obtained, and the dynamic and time constraints of tasks are considered to enhance the adaptability of the control method; based on the civil cluster task scheduling algorithm, the M civil heterogeneous unmanned control units are scheduled according to the characteristic data of M control units, the newly added civil task information and the newly added civil task time constraints, and the civil cluster task scheduling plan is obtained to realize the division of labor and cooperation of heterogeneous UAV clusters and improve the efficiency of task execution; based on the digital twin control decision module, According to the M units, subtasks are assigned to perform control decision analysis on M civil heterogeneous unmanned control units to obtain a distributed control solution. The feasibility of the control solution is verified through simulation to improve the accuracy and reliability of decision-making. At the same time, distributed control allows each drone to make autonomous decisions based on its own tasks and reduce dependence on central nodes. According to the distributed control solution, control of M civil heterogeneous unmanned control units is executed, and the distributed control solution is sent to each drone to collaboratively complete the assigned tasks to achieve a technical solution for distributed control of heterogeneous unmanned clusters. This solves the technical problems in the prior art that centralized control of civil heterogeneous unmanned clusters has poor flexibility, strong dependence on central nodes, and poor task execution effects, and achieves the technical effect of reducing dependence on central nodes and improving task execution effects through distributed control of civil heterogeneous unmanned clusters.
[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] Figure 1 A flow chart of a civilian heterogeneous unmanned cluster control method based on distributed control is provided for an embodiment of the present application;
[0011] Figure 2 A structural schematic diagram of a civilian heterogeneous unmanned cluster control system based on distributed control is provided for an embodiment of the present application.
[0012] Explanation of the reference numerals: control unit acquisition component 11, information collection component 12, instruction acquisition component 13, scheduling scheme acquisition component 14, control decision analysis component 15, control scheme execution component 16. DETAILED DESCRIPTION
[0013] The overall idea of the technical solution provided by this application is as follows:
[0014] The embodiments of the present application provide a civilian heterogeneous unmanned cluster control method and system based on distributed control. The tasks of the heterogeneous unmanned aerial vehicle cluster are reasonably allocated through task scheduling, and the control decision analysis of the allocation plan is performed using digital twin technology. The distributed control plan is then sent to each heterogeneous unmanned aerial vehicle for distributed control, thereby achieving efficient collaborative control of the heterogeneous unmanned aerial vehicle cluster.
[0015] Specifically, first, the control objects of the heterogeneous UAV cluster are obtained, the characteristic information of each UAV is collected, and new civil task scheduling instructions are introduced to adapt to dynamic task requirements. Then, the civil cluster task scheduling algorithm is used to combine multiple factors to assign tasks to heterogeneous UAVs, giving full play to the division of labor and cooperation advantages of heterogeneous clusters. Next, the digital twin control decision module is introduced to simulate and verify the task allocation and analyze the control decision to obtain a distributed control scheme to improve the reliability of the decision. Subsequently, the distributed control scheme is issued to each heterogeneous UAV, and the UAV makes autonomous decision-making control based on it, realizing decentralized distributed control and improving the task execution effect.
[0016] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically described below in conjunction with the drawings in the specification.
[0017] Embodiment 1
[0018] like Figure 1 As shown, the embodiment of the present application provides a civilian heterogeneous unmanned cluster control method based on distributed control, the method comprising:
[0019] Step S100: obtaining M civilian heterogeneous unmanned control units of a civilian heterogeneous unmanned cluster, where M is a positive integer greater than 1;
[0020] In the embodiment of the present application, after determining the civil heterogeneous unmanned cluster, first, the M civil heterogeneous unmanned control units of the civil heterogeneous unmanned cluster are obtained through the unique number of the civil heterogeneous unmanned control unit, where M is the number of civil heterogeneous unmanned control units, which is a positive integer greater than 1. Among them, the civil heterogeneous unmanned control units are M civil drones with different structures and functions.
[0021] By obtaining M civilian heterogeneous unmanned control units of a civilian heterogeneous unmanned cluster, the required basic data is provided for the subsequent cluster task scheduling and distributed control of these M civilian heterogeneous unmanned control units.
[0022] Step S200: respectively collecting basic information of the M civilian heterogeneous unmanned control units and fixed civilian task information to obtain characteristic data of the M control units;
[0023] In an embodiment of the present application, after M civilian heterogeneous unmanned control units are determined, basic information collection and inherent civilian mission information collection are performed on these M civilian heterogeneous unmanned control units, that is, M civilian drones of different models, so as to obtain relevant data describing the characteristics of these M control units and obtain M control unit characteristic data.
[0024] Specifically, first, basic information of M civil heterogeneous unmanned control units is collected. The collected basic information may include but is not limited to the model, weight, size, power system parameters, avionics system parameters, sensor configuration and other physical performance indicators of each UAV, so as to comprehensively characterize the basic attributes of each heterogeneous UAV. At the same time, the inherent civil mission information currently assigned and executed by these M civil heterogeneous unmanned control units is collected, such as the preset flight altitude, route, mission duration, load requirements, etc., to reflect the current mission status and usage constraints of each civil heterogeneous unmanned control unit.
[0025] By integrating and analyzing the above basic information and inherent task information, M control unit characteristic data describing the characteristics of M civil heterogeneous unmanned control units are obtained, which provides important information support for subsequent task scheduling and distributed collaborative control, thereby realizing efficient collaborative operation of civil UAV clusters.
[0026] Step S300: obtaining a newly added civil task scheduling instruction, wherein the newly added civil task scheduling instruction includes newly added civil task information and newly added civil task time constraints;
[0027] In an embodiment of the present application, after obtaining characteristic data of M control units, a newly added civil task scheduling instruction is further obtained. The newly added civil task scheduling instruction includes two parts: one is the newly added civil task information, and the other is the time constraint related to the newly added civil task information, that is, the newly added civil task time constraint.
[0028] A civil task scheduling instruction receiving module is set up for the civil heterogeneous unmanned cluster. This module receives the newly added civil task instructions from the upper system or command center in real time through the network or other communication methods. After the civil task scheduling instruction receiving module obtains the newly added civil task scheduling instruction, it parses and extracts it, and extracts the newly added civil task information and the newly added civil task time constraints respectively. Among them, the newly added civil task information describes the specific requirements of the newly added tasks to be assigned to the civil heterogeneous unmanned cluster, including but not limited to the task type, execution area range, required load capacity, estimated task execution time and other parameters, providing a basic basis for subsequent task scheduling and allocation; the newly added civil task time constraints put forward clear restrictions on the execution time, response speed and other aspects of the newly added civil task information, providing an important basis for judging the task scheduling algorithm of the civil cluster to find the optimal solution.
[0029] By obtaining the scheduling instructions for new civilian tasks, it not only includes the specific description of the new tasks, but also provides the matching time constraint requirements, laying the foundation for subsequent task allocation and distributed control decision-making based on the civilian cluster task scheduling algorithm.
[0030] Step S400: Based on the civilian cluster task scheduling algorithm, the M civilian heterogeneous unmanned control units are scheduled according to the characteristic data of the M control units, the newly added civilian task information and the time constraint of the newly added civilian task, to obtain a civilian cluster task scheduling scheme, wherein the civilian cluster task scheduling scheme includes M unit allocation subtasks corresponding to the M civilian heterogeneous unmanned control units;
[0031] Further, step S400 includes:
[0032] Step S410: taking the characteristic data of the M control units as task scheduling constraints, and taking the newly added civilian task information and the newly added civilian task time constraints as task scheduling targets;
[0033] Step S420: Based on the task scheduling constraint condition and the task scheduling target, perform multiple task scheduling on the M civilian heterogeneous unmanned control units to obtain a civilian cluster task scheduling solution space that satisfies a preset scheduling solution capacity constraint, wherein the civilian cluster task scheduling solution space includes multiple cluster task scheduling solutions, and each cluster task scheduling solution includes M unit randomly assigned subtasks corresponding to the M civilian heterogeneous unmanned control units;
[0034] Step S430: extracting a first cluster task scheduling solution according to the civilian cluster task scheduling solution space;
[0035] Step S440: performing fitness evaluation on the first cluster task scheduling solution according to the cluster task scheduling prediction and evaluation model, and determining a first solution task scheduling fitness coefficient;
[0036] Step S450: extracting a second cluster task scheduling solution according to the civilian cluster task scheduling solution space, and performing fitness evaluation on the second cluster task scheduling solution based on the cluster task scheduling prediction and evaluation model to obtain a second solution task scheduling fitness coefficient;
[0037] Step S460: determining whether the first solution task scheduling adaptation coefficient is less than the second solution task scheduling adaptation coefficient;
[0038] Step S470: if the first solution task scheduling adaptation coefficient is less than the second solution task scheduling adaptation coefficient, the second solution task scheduling adaptation coefficient is set as the current winning solution task scheduling adaptation coefficient, and the second cluster task scheduling solution is set as the current winning cluster task scheduling solution;
[0039] Step S480: if the first solution task scheduling adaptation coefficient is greater than / equal to the second solution task scheduling adaptation coefficient, the first solution task scheduling adaptation coefficient is set as the current winning solution task scheduling adaptation coefficient, and the first cluster task scheduling solution is set as the current winning cluster task scheduling solution;
[0040] Step S480: continuing to iteratively optimize the civilian cluster task scheduling solution space based on the current winning solution task scheduling adaptation coefficient and the current winning cluster task scheduling solution, and generating the civilian cluster task scheduling solution when the number of iterative optimizations meets the preset task scheduling optimization constraints.
[0041] Further, step S440 includes:
[0042] Step S441: inputting the first cluster task scheduling solution into the cluster task scheduling prediction and evaluation model to obtain a first scheduling prediction and evaluation result, wherein the first scheduling prediction and evaluation result includes a first execution efficiency prediction and evaluation coefficient, a first execution quality prediction and evaluation coefficient, and a first execution cost prediction and evaluation coefficient;
[0043] Step S442: normalizing the first scheduling prediction evaluation result to obtain a first standard scheduling prediction evaluation result;
[0044] Step S443: Perform weighted calculation on the first standard scheduling prediction evaluation result based on the fitness evaluation weight allocation condition to generate the first solution task scheduling adaptation coefficient.
[0045] In a feasible implementation, before executing the civil cluster task scheduling algorithm, first, the characteristic data of M control units are used as task scheduling constraints. At the same time, the newly added civil task information and the newly added civil task time constraints are used as task scheduling optimization targets to ensure that the civil cluster task scheduling scheme can fully meet the performance characteristics of each heterogeneous UAV and meet the needs of the newly added tasks to the maximum extent. Then, multiple rounds of task scheduling iterative calculations are performed on the M civil heterogeneous unmanned control units. In each round of scheduling calculations, attempts are made to assign the newly added tasks to the M civil heterogeneous unmanned control units to generate a feasible cluster task scheduling solution. This scheduling solution contains the subtask allocation schemes corresponding to each of the M civil heterogeneous unmanned control units, that is, the units randomly assign subtasks. Through multiple rounds of iterative calculations, a series of feasible cluster task scheduling solutions that meet the preset scheduling solution capacity constraints are obtained. Among them, the preset scheduling solution capacity constraint refers to limiting the number of generated cluster task scheduling solutions, which cannot exceed a certain preset threshold value, to ensure that the scale of the civil cluster task scheduling solution space is not too large. Then, multiple cluster task scheduling solutions are summarized to form a civil cluster task scheduling solution space.
[0046] Subsequently, multiple cluster task scheduling solutions in the civil cluster task scheduling solution space are traversed, and one cluster task scheduling solution is extracted each time as the first cluster task scheduling solution. Then, the first cluster task scheduling solution obtained is input into the pre-built cluster task scheduling prediction and evaluation model. The cluster task scheduling prediction and evaluation model is established based on the historical data of similar cluster operation tasks in the past and the experience accumulation of experts in related fields, and can simulate and predict the various performances of a given cluster task scheduling solution in the actual execution process. After the first cluster task scheduling solution is input into the cluster task scheduling prediction and evaluation model, the cluster task scheduling prediction and evaluation model outputs the corresponding first scheduling prediction and evaluation result, which includes the first execution efficiency prediction and evaluation coefficient, the first execution quality prediction and evaluation coefficient and the first execution cost prediction and evaluation coefficient. Among them, the first execution efficiency prediction and evaluation coefficient quantifies the performance of the first cluster task scheduling solution in terms of execution efficiency; the first execution quality prediction and evaluation coefficient quantifies the performance of the first cluster task scheduling solution in terms of execution quality; and the first execution cost prediction and evaluation coefficient quantifies the performance of the first cluster task scheduling solution in terms of execution cost. Through these three prediction and evaluation coefficients, the expected performance of the cluster task scheduling solution in terms of execution efficiency, execution quality and execution cost can be fully reflected. Next, the values of the first execution efficiency prediction evaluation coefficient, the first execution quality prediction evaluation coefficient, and the first execution cost prediction evaluation coefficient are respectively mapped to the standardized interval between 0 and 1, so as to realize the normalization processing of the first scheduling prediction evaluation result and obtain the first standard scheduling prediction evaluation result. Subsequently, for the first standard scheduling prediction evaluation result, corresponding weight factors are set for the three standardized indicators, and then weighted calculation is performed to obtain the first solution task scheduling adaptation coefficient.
[0047] After that, continue to traverse the space of civil cluster task scheduling solutions, extract another cluster task scheduling solution from it, as the second cluster task scheduling solution, input it into the cluster task scheduling prediction and evaluation model, evaluate and analyze its expected performance in terms of execution efficiency, execution quality and execution cost, and obtain the second execution efficiency prediction evaluation coefficient, the second execution quality prediction evaluation coefficient and the second execution cost prediction evaluation coefficient. After normalizing these three prediction evaluation coefficients, weight them to obtain the second solution task scheduling adaptation coefficient. Subsequently, compare the size of the first solution task scheduling adaptation coefficient and the second solution task scheduling adaptation coefficient. If the first solution task scheduling adaptation coefficient is less than the second solution task scheduling adaptation coefficient, it indicates that the comprehensive performance of the second cluster task scheduling solution is better than the first cluster task scheduling solution. At this time, the second solution task scheduling adaptation coefficient is set as the current winning solution task scheduling adaptation coefficient, and the second cluster task scheduling solution is set as the current winning cluster task scheduling solution. If the first solution task scheduling adaptation coefficient is greater than or equal to the second solution task scheduling adaptation coefficient, it indicates that the comprehensive performance of the first cluster task scheduling solution is not lower than or even better than the second cluster task scheduling solution. At this time, the first solution task scheduling adaptation coefficient is set as the current winning solution task scheduling adaptation coefficient, and the first cluster task scheduling solution is set as the current winning cluster task scheduling solution. By comparing the first solution task scheduling adaptation coefficient with the second solution task scheduling adaptation coefficient, it is ensured that the current winning solution task scheduling adaptation coefficient and the current winning cluster task scheduling solution are both currently optimal, laying the foundation for subsequent iterative optimization and solution selection.
[0048] Then, based on the current winning task scheduling adaptation coefficient and the current winning cluster task scheduling solution, the entire civil cluster task scheduling solution space is further iterated and optimized. The cluster task scheduling solution is extracted from the civil cluster task scheduling solution space, and the cluster task scheduling prediction and evaluation model is used to evaluate its fitness coefficient. The fitness coefficient of the extracted cluster task scheduling solution is compared with the current winning task scheduling adaptation coefficient. If there is a better one, the current winning task scheduling adaptation coefficient and the current winning cluster task scheduling solution are replaced. The iterative optimization process is continuously performed until the number of iterative optimizations meets the preset task scheduling optimization constraint. Among them, the preset task scheduling optimization constraint is that the number of iterations reaches the total number of multiple cluster task scheduling solutions. When the number of iterative optimizations meets the preset task scheduling optimization constraint, the current winning cluster task scheduling solution is extracted as the civil cluster task scheduling solution, which lays the foundation for distributed control decision-making. Among them, the civil cluster task scheduling solution includes M unit allocation subtasks corresponding to M civil heterogeneous unmanned control units.
[0049] Step S500: Based on the digital twin control decision module, control decision analysis is performed on the M civilian heterogeneous unmanned control units according to the M unit allocation subtasks to obtain a distributed control solution, wherein the distributed control solution includes M unit task control decisions;
[0050] Furthermore, step S500 includes:
[0051] Step S510: the digital twin control decision module includes a digital twin module and a mission control decision module, and the mission control decision module includes M mission control decision models corresponding to the M civil heterogeneous unmanned control units;
[0052] Step S520: inputting the M unit allocation subtasks into the M task control decision models to obtain M matching task control decisions;
[0053] Step S530: Based on the digital twin module, fault fitting optimization is performed on the M matching task control decisions respectively to generate the distributed control scheme.
[0054] Further, step S530 includes:
[0055] Step S531: extracting the mth matching task control decision corresponding to the mth civil heterogeneous unmanned control unit according to the M matching task control decisions, where m is a positive integer and m belongs to M;
[0056] Step S532: Based on the digital twin module, modeling is performed according to the mth civil heterogeneous unmanned control unit to obtain the mth civil heterogeneous unmanned control model;
[0057] Step S533: Based on the digital twin module, simulate and control the mth civil heterogeneous unmanned control model according to the mth matching task control decision to obtain an mth decision simulation working condition data set;
[0058] Step S534: performing fault prediction based on the mth decision simulation condition data set to obtain the mth decision fault prediction result;
[0059] Step S535: performing a fault impact assessment based on the mth decision fault prediction result to obtain an mth predicted fault impact assessment coefficient;
[0060] Step S536: if the m-th predicted fault impact evaluation coefficient satisfies the preset fault impact evaluation constraint, outputting the m-th matching task control decision as the m-th unit task control decision, and adding the m-th unit task control decision to the distributed control scheme;
[0061] Step S537: If the mth predicted fault impact evaluation coefficient does not satisfy the preset fault impact evaluation constraint, the mth matching task control decision is optimized and adjusted according to the mth decision fault prediction result to generate the mth unit task control decision.
[0062] In a feasible implementation, the digital twin control decision module includes two parts, a digital twin module and a task control decision module. Among them, the task control decision module contains M task control decision models corresponding to M civil heterogeneous unmanned control units. These task control decision models are established based on the analysis and modeling of various types of unmanned aerial vehicles, and can output corresponding control decision plans according to task requirements; the digital twin module uses simulation technology to establish a digital twin model for each civil heterogeneous unmanned control unit to accurately simulate and reproduce the various operating characteristics and working conditions of the actual unmanned aerial vehicle. Then, the civil cluster task scheduling plan is input into the task control decision model, and the task control decision model extracts M unit allocation subtasks from the civil cluster task scheduling plan, and inputs the M unit allocation subtasks into the corresponding M task control decision models respectively. Each model generates a matching task control decision plan according to its internal decision algorithm, and obtains M matching task control decisions.
[0063] Subsequently, from these M matching mission control decisions, the mth matching mission control decision corresponding to the mth civil heterogeneous unmanned control unit is extracted separately. Among them, m is a positive integer, and its value range is between 1 and M. In other words, from these M matching mission control decisions, the first, second, and finally the Mth matching mission control decisions are extracted in sequence. Then, using the digital twin module, the various performance parameters, working mechanisms, and external environmental factors of the mth civil heterogeneous unmanned control unit are deeply analyzed and mathematically modeled, including but not limited to the modeling of the UAV's body structure, power system, avionics system, sensor configuration, and other aspects, so as to obtain a high-fidelity mth civil heterogeneous unmanned control model. At the same time, the m-th matching mission control decision is taken as input and transmitted to the m-th civil heterogeneous unmanned control model. The simulation algorithm integrated in the digital twin module is used to simulate the execution of the m-th matching mission control decision in a digital environment, and a detailed operating data set of the m-th control decision scheme under various working conditions is obtained. The m-th decision simulation working condition data set is obtained, including various information such as the flight trajectory, power parameters, and sensor output of the UAV, which comprehensively reflects the execution effect of the m-th control decision scheme in the actual environment.
[0064] Subsequently, the fault prediction algorithm and machine learning technology are used to conduct in-depth mining and modeling analysis on the m-th decision simulation condition data set to achieve fault prediction analysis, thereby obtaining the m-th decision fault prediction result. Among them, the m-th decision fault prediction result not only includes the probability and degree of fault occurrence, but also involves more detailed information such as the cause analysis of the fault and the impact assessment. Then, considering the probability of fault occurrence, the impact of the fault on the key performance indicators of the UAV, the impact of the fault on the mission execution and other factors, the m-th decision simulation condition data set is evaluated based on the m-th decision fault prediction result to obtain a quantitative indicator reflecting the overall fault impact, namely the m-th prediction fault impact evaluation coefficient.
[0065] Subsequently, the mth predicted fault impact evaluation coefficient is compared with the preset fault impact evaluation constraint. Among them, the preset fault impact evaluation constraint is a threshold set in advance by the expert group for the acceptable fault impact degree. Only when the mth predicted fault impact evaluation coefficient is lower than this threshold, can the mth matching task control decision be considered reliable. If the mth predicted fault impact evaluation coefficient meets the preset fault impact evaluation constraint, the mth matching task control decision is directly output as the mth unit task control decision, and the mth unit task control decision is added to the entire distributed control scheme. If the mth predicted fault impact evaluation coefficient does not meet this preset constraint, that is, the mth matching task control decision has a large fault risk and cannot be directly output as the final unit task control decision. At this time, according to the obtained mth decision fault prediction result, the mth matching task control decision scheme is further optimized and adjusted in order to generate a more reliable mth unit task control decision. The specific optimization and adjustment methods include but are not limited to taking corresponding control strategy optimization according to the predicted fault type and impact degree, such as adjusting key parameters, adding redundant design, etc.; or improving the original decision scheme according to the simulation working condition data. When the mth predicted fault impact evaluation coefficient of the optimized and adjusted mth matching task control decision meets the preset fault impact evaluation constraint, the optimized and adjusted mth matching task control decision is added to the distributed control scheme as the mth unit task control decision.
[0066] Step S600: Execute control of the M civilian heterogeneous unmanned control units according to the distributed control scheme.
[0067] Furthermore, step S600 includes:
[0068] Step S610: performing real-time monitoring on the M civilian heterogeneous unmanned control units to obtain M unit monitoring data streams;
[0069] Step S620: performing anomaly identification and evaluation based on the M unit monitoring data streams to obtain M unit monitoring anomaly identification results and M unit monitoring anomaly evaluation coefficients;
[0070] Step S630: respectively determining whether the monitoring abnormality evaluation coefficients of the M units are greater than / equal to the monitoring abnormality evaluation threshold;
[0071] Step S640: if any one of the M unit monitoring abnormality evaluation coefficients is greater than / equal to the monitoring abnormality evaluation threshold, a monitoring abnormality control adjustment instruction is generated;
[0072] Step S650: Based on the monitoring abnormality control adjustment instruction, the M civilian heterogeneous unmanned control units and the M unit monitoring abnormality identification results are abnormally marked to generate marked unmanned control units and marked monitoring abnormality identification results;
[0073] Step S660: Based on the distributed control scheme, the identification unmanned control unit is subjected to abnormal compensation control according to the identification monitoring abnormality identification result.
[0074] In a feasible implementation, after obtaining the distributed control scheme, the M unit task control decisions are sent to the M civilian heterogeneous unmanned control units of the civilian heterogeneous unmanned cluster to execute the distributed control scheme. Then, a series of sensors are deployed on the interactive M unmanned control units to monitor the M unmanned control units in real time, and M unit monitoring data streams are obtained. Sensors include but are not limited to flight state sensors, power system sensors, load sensors, environmental sensors, etc. Among them, the flight state sensor is used to collect information such as the position, attitude, and speed of the drone; the power system sensor is used to collect data such as the thrust and battery power of the drone; the load sensor is used to collect the mission load currently carried by the drone; the environmental sensor is used to collect data such as the weather and wind field of the environment in which the drone is located. Subsequently, using the M unit monitoring data streams, the unit monitoring data streams of each civilian heterogeneous unmanned control unit are deeply analyzed to obtain M unit monitoring anomaly recognition results and M unit monitoring anomaly evaluation coefficients. For example, by learning the patterns of historical normal working data, a normal working state model of each civil heterogeneous unmanned control unit is established, and the real-time monitored unit monitoring data stream is compared and analyzed with the normal working state model. Once the data is found to deviate from the normal range, the civil heterogeneous unmanned control unit is judged to be in an abnormal state, and M unit monitoring anomaly identification results are obtained to identify which drones have abnormal conditions. At the same time, according to the degree of deviation between the abnormal state and the normal working state model, a quantitative evaluation of the abnormal state of the corresponding civil heterogeneous unmanned control unit is obtained, and M unit monitoring anomaly evaluation coefficients are obtained.
[0075] Then, check the M unit monitoring anomaly evaluation coefficients one by one to see if any of them is greater than or equal to the preset monitoring anomaly evaluation threshold. Among them, the monitoring anomaly evaluation threshold is a standard pre-set according to actual application requirements and experience, reflecting the acceptable abnormality level boundary. If any of the M unit monitoring anomaly evaluation coefficients is greater than or equal to the monitoring anomaly evaluation threshold, it means that a serious abnormal condition that needs to be dealt with immediately has occurred in a civilian heterogeneous unmanned control unit. In this case, a monitoring abnormality control adjustment instruction is immediately generated as a basis and guidance for subsequent control compensation for abnormal civilian heterogeneous unmanned control units. For example, the detection abnormality control adjustment instruction contains the number of the abnormal civilian heterogeneous unmanned control unit; descriptive diagnostic information on the specific fault type and severity of the abnormal civilian heterogeneous unmanned control unit, etc.
[0076] After receiving the monitoring abnormal control adjustment instruction, the abnormal civil heterogeneous unmanned control unit is identified according to the number of the abnormal civil heterogeneous unmanned control unit provided in the monitoring abnormal control adjustment instruction, and the identified abnormal civil heterogeneous unmanned control unit constitutes an identified unmanned control unit. At the same time, the unit monitoring abnormality identification result of the identified unmanned control unit is identified to obtain the identified monitoring abnormality identification result, so as to clearly determine which civil heterogeneous unmanned control units are in an abnormal state. Subsequently, according to the distributed control scheme, for those civil heterogeneous unmanned control units identified as abnormal, that is, the identified unmanned control unit, the corresponding unit task control decision is extracted. Then, the unit task control decision of the identified unmanned control unit is compared with the identified monitoring abnormality identification result, and the deviation between the current operating state of the identified unmanned control unit and its unit task control decision is extracted, and the abnormal identified unmanned control unit is adjusted and compensated based on the deviation. For example, the task reallocation of the identified unmanned control unit, the dynamic optimization adjustment of the control parameters, the rapid switching of fault isolation and backup plans, etc. Through precise compensation based on the distributed control scheme, the impact of the identified unmanned control unit can be alleviated, ensuring that the entire civil heterogeneous unmanned cluster can still operate stably according to the distributed control scheme and complete the established task objectives.
[0077] In summary, the civilian heterogeneous unmanned cluster control method based on distributed control provided by the embodiments of the present application has the following technical effects:
[0078] Obtain M civil heterogeneous unmanned control units of civil heterogeneous unmanned clusters to provide a basis for subsequent task allocation and control decision-making. Collect the basic information and fixed civil task information of M civil heterogeneous unmanned control units respectively, obtain the characteristic data of M control units, and provide a basis for subsequent task scheduling. Obtain the newly added civil task scheduling instructions, where the newly added civil task scheduling instructions include the newly added civil task information and the newly added civil task time constraints. Through the description of task information and time constraints, provide input for task scheduling and improve the ability to adapt to dynamic environments. Based on the civil cluster task scheduling algorithm, perform task scheduling on M civil heterogeneous unmanned control units according to the characteristic data of M control units, the newly added civil task information and the newly added civil task time constraints, obtain the civil cluster task scheduling plan, give play to the division of labor and cooperation advantages of heterogeneous clusters, realize the reasonable allocation of resources, and improve task execution efficiency. Based on the digital twin control decision module, perform control decision analysis on M civil heterogeneous unmanned control units according to the M unit allocation subtasks, and obtain a distributed control plan, where the distributed control plan includes M unit task control decisions, and provide an executable control decision sequence for each drone. The control of M civilian heterogeneous unmanned control units is executed according to the distributed control scheme, which reduces the dependence on the central node, efficiently completes the assigned tasks, and improves the task execution effect.
[0079] Embodiment 2
[0080] Based on the same inventive concept as the civilian heterogeneous unmanned cluster control method based on distributed control in the aforementioned embodiment, Figure 2 As shown, the embodiment of the present application provides a civilian heterogeneous unmanned cluster control system based on distributed control, the system comprising:
[0081] A control unit acquisition component 11 is used to obtain M civilian heterogeneous unmanned control units of a civilian heterogeneous unmanned cluster, where M is a positive integer greater than 1;
[0082] An information collection component 12 is used to collect basic information of the M civilian heterogeneous unmanned control units and fixed civilian task information respectively, and obtain characteristic data of the M control units;
[0083] The instruction acquisition component 13 is used to obtain the newly added civil task scheduling instruction, wherein the newly added civil task scheduling instruction includes the newly added civil task information and the newly added civil task time constraint;
[0084] The scheduling scheme acquisition component 14 is used to perform task scheduling on the M civilian heterogeneous unmanned control units based on the civilian cluster task scheduling algorithm according to the characteristic data of the M control units, the newly added civilian task information and the newly added civilian task time constraint, and obtain a civilian cluster task scheduling scheme, wherein the civilian cluster task scheduling scheme includes M unit allocation subtasks corresponding to the M civilian heterogeneous unmanned control units;
[0085] A control decision analysis component 15 is used to perform control decision analysis on the M civil heterogeneous unmanned control units according to the M unit allocation subtasks based on the digital twin control decision module to obtain a distributed control solution, wherein the distributed control solution includes M unit task control decisions;
[0086] The control scheme execution component 16 is used to execute the control of the M civilian heterogeneous unmanned control units according to the distributed control scheme.
[0087] Furthermore, the scheduling scheme acquisition component 14 includes the following execution steps:
[0088] Taking the characteristic data of the M control units as task scheduling constraints, and taking the newly added civilian task information and the newly added civilian task time constraints as task scheduling targets;
[0089] Based on the task scheduling constraint condition and the task scheduling target, multiple task scheduling is performed on the M civilian heterogeneous unmanned control units to obtain a civilian cluster task scheduling solution space that satisfies a preset scheduling solution capacity constraint, wherein the civilian cluster task scheduling solution space includes multiple cluster task scheduling solutions, and each cluster task scheduling solution includes M unit randomly assigned subtasks corresponding to the M civilian heterogeneous unmanned control units;
[0090] Extracting a first cluster task scheduling solution according to the civilian cluster task scheduling solution space;
[0091] Performing fitness evaluation on the first cluster task scheduling solution according to the cluster task scheduling prediction and evaluation model, and determining the first solution task scheduling fitness coefficient;
[0092] Extracting a second cluster task scheduling solution according to the civilian cluster task scheduling solution space, and performing fitness evaluation on the second cluster task scheduling solution based on the cluster task scheduling prediction and evaluation model to obtain a second solution task scheduling fitness coefficient;
[0093] Determining whether the first solution task scheduling adaptation coefficient is less than the second solution task scheduling adaptation coefficient;
[0094] If the task scheduling adaptation coefficient of the first solution is less than the task scheduling adaptation coefficient of the second solution, the task scheduling adaptation coefficient of the second solution is set as the task scheduling adaptation coefficient of the current winning solution, and the second cluster task scheduling solution is set as the current winning cluster task scheduling solution;
[0095] The civilian cluster task scheduling solution space is continuously iterated and optimized based on the task scheduling adaptation coefficient of the current winning solution and the current winning cluster task scheduling solution. When the number of iterative optimizations meets the preset task scheduling optimization constraints, the civilian cluster task scheduling solution is generated.
[0096] Furthermore, the scheduling scheme acquisition component 14 further includes the following execution steps:
[0097] Inputting the first cluster task scheduling solution into the cluster task scheduling prediction and evaluation model to obtain a first scheduling prediction and evaluation result, wherein the first scheduling prediction and evaluation result includes a first execution efficiency prediction and evaluation coefficient, a first execution quality prediction and evaluation coefficient, and a first execution cost prediction and evaluation coefficient;
[0098] Normalizing the first scheduling prediction evaluation result to obtain a first standard scheduling prediction evaluation result;
[0099] The first standard scheduling prediction evaluation result is weightedly calculated based on the fitness evaluation weight allocation condition to generate the first solution task scheduling adaptation coefficient.
[0100] Furthermore, the scheduling scheme acquisition component 14 further includes the following execution steps:
[0101] If the first solution task scheduling adaptation coefficient is greater than / equal to the second solution task scheduling adaptation coefficient, the first solution task scheduling adaptation coefficient is set as the current winning solution task scheduling adaptation coefficient, and the first cluster task scheduling solution is set as the current winning cluster task scheduling solution.
[0102] Furthermore, the control decision analysis component 15 includes the following execution steps:
[0103] The digital twin control decision module includes a digital twin module and a mission control decision module, and the mission control decision module includes M mission control decision models corresponding to the M civil heterogeneous unmanned control units;
[0104] Inputting the M unit allocation subtasks into the M task control decision models to obtain M matching task control decisions;
[0105] Based on the digital twin module, fault fitting optimization is performed on the M matching task control decisions respectively to generate the distributed control scheme.
[0106] Furthermore, the control decision analysis component 15 also includes the following execution steps:
[0107] Extracting the mth matching mission control decision corresponding to the mth civil heterogeneous unmanned control unit according to the M matching mission control decisions, wherein m is a positive integer and m belongs to M;
[0108] Based on the digital twin module, modeling is performed according to the mth civil heterogeneous unmanned control unit to obtain the mth civil heterogeneous unmanned control model;
[0109] Based on the digital twin module, simulate and control the mth civil heterogeneous unmanned control model according to the mth matching task control decision to obtain an mth decision simulation working condition data set;
[0110] Performing fault prediction based on the mth decision simulation condition data set to obtain an mth decision fault prediction result;
[0111] Performing a fault impact assessment based on the mth decision fault prediction result to obtain an mth predicted fault impact assessment coefficient;
[0112] If the m-th predicted fault impact evaluation coefficient satisfies the preset fault impact evaluation constraint, outputting the m-th matching task control decision as the m-th unit task control decision, and adding the m-th unit task control decision to the distributed control scheme;
[0113] If the mth predicted fault impact evaluation coefficient does not satisfy the preset fault impact evaluation constraint, the mth matching task control decision is optimized and adjusted according to the mth decision fault prediction result to generate the mth unit task control decision.
[0114] Furthermore, the control scheme execution component 16 includes the following execution steps:
[0115] Performing real-time monitoring on the M civilian heterogeneous unmanned control units to obtain M unit monitoring data streams;
[0116] Performing anomaly recognition and evaluation based on the M unit monitoring data streams to obtain M unit monitoring anomaly recognition results and M unit monitoring anomaly evaluation coefficients;
[0117] Determine whether the monitoring anomaly evaluation coefficients of the M units are greater than / equal to the monitoring anomaly evaluation thresholds;
[0118] If any one of the M unit monitoring abnormality evaluation coefficients is greater than / equal to the monitoring abnormality evaluation threshold, a monitoring abnormality control adjustment instruction is generated;
[0119] Based on the monitoring abnormality control adjustment instruction, the M civilian heterogeneous unmanned control units and the M unit monitoring abnormality identification results are abnormally marked to generate marked unmanned control units and marked monitoring abnormality identification results;
[0120] Based on the distributed control scheme, abnormal compensation control is performed on the identification unmanned control unit according to the identification monitoring abnormality recognition result.
[0121] Any step of the method described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present application, without any unnecessary restrictions.
[0122] Furthermore, the first or second mentioned above may not only represent an order relationship, but may also represent a specific concept, and / or refer to multiple elements that can be selected individually or in full. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these changes and variations.
Claims
1. A civil heterogeneous unmanned cluster control method based on distributed control, characterized in that: The method comprises: Obtain M civilian heterogeneous unmanned control units of a civilian heterogeneous unmanned cluster, where M is a positive integer greater than 1; respectively collecting basic information of the M civilian heterogeneous unmanned control units and fixed civilian mission information to obtain characteristic data of the M control units; Obtaining a newly added civil task scheduling instruction, wherein the newly added civil task scheduling instruction includes newly added civil task information and newly added civil task time constraints; Based on the civilian cluster task scheduling algorithm, the M civilian heterogeneous unmanned control units are task scheduled according to the characteristic data of the M control units, the newly added civilian task information and the time constraint of the newly added civilian task, to obtain a civilian cluster task scheduling scheme, wherein the civilian cluster task scheduling scheme includes M unit allocation subtasks corresponding to the M civilian heterogeneous unmanned control units; Based on the digital twin control decision module, control decision analysis is performed on the M civilian heterogeneous unmanned control units according to the M unit allocation subtasks to obtain a distributed control solution, wherein the distributed control solution includes M unit task control decisions; Control of the M civilian heterogeneous unmanned control units is performed according to the distributed control scheme.
2. The method according to claim 1, characterized in that Based on the civil cluster task scheduling algorithm, the M civil heterogeneous unmanned control units are scheduled according to the characteristic data of the M control units, the newly added civil task information and the time constraint of the newly added civil task, to obtain a civil cluster task scheduling plan, including: Taking the characteristic data of the M control units as task scheduling constraints, and taking the newly added civilian task information and the newly added civilian task time constraints as task scheduling targets; Based on the task scheduling constraint condition and the task scheduling target, multiple task scheduling is performed on the M civilian heterogeneous unmanned control units to obtain a civilian cluster task scheduling solution space that satisfies a preset scheduling solution capacity constraint, wherein the civilian cluster task scheduling solution space includes multiple cluster task scheduling solutions, and each cluster task scheduling solution includes M unit randomly assigned subtasks corresponding to the M civilian heterogeneous unmanned control units; Extracting a first cluster task scheduling solution according to the civilian cluster task scheduling solution space; Performing fitness evaluation on the first cluster task scheduling solution according to the cluster task scheduling prediction and evaluation model, and determining the first solution task scheduling fitness coefficient; Extracting a second cluster task scheduling solution according to the civilian cluster task scheduling solution space, and performing fitness evaluation on the second cluster task scheduling solution based on the cluster task scheduling prediction and evaluation model to obtain a second solution task scheduling fitness coefficient; Determining whether the first solution task scheduling adaptation coefficient is less than the second solution task scheduling adaptation coefficient; If the task scheduling adaptation coefficient of the first solution is less than the task scheduling adaptation coefficient of the second solution, the task scheduling adaptation coefficient of the second solution is set as the task scheduling adaptation coefficient of the current winning solution, and the second cluster task scheduling solution is set as the current winning cluster task scheduling solution; The civilian cluster task scheduling solution space is continuously iterated and optimized based on the task scheduling adaptation coefficient of the current winning solution and the current winning cluster task scheduling solution. When the number of iterative optimizations meets the preset task scheduling optimization constraints, the civilian cluster task scheduling solution is generated.
3. The method according to claim 2, characterized in that The first cluster task scheduling solution is evaluated for fitness according to the cluster task scheduling prediction and evaluation model to determine the first solution task scheduling fitness coefficient, including: Inputting the first cluster task scheduling solution into the cluster task scheduling prediction and evaluation model to obtain a first scheduling prediction and evaluation result, wherein the first scheduling prediction and evaluation result includes a first execution efficiency prediction and evaluation coefficient, a first execution quality prediction and evaluation coefficient, and a first execution cost prediction and evaluation coefficient; Normalizing the first scheduling prediction evaluation result to obtain a first standard scheduling prediction evaluation result; The first standard scheduling prediction evaluation result is weightedly calculated based on the fitness evaluation weight allocation condition to generate the first solution task scheduling adaptation coefficient.
4. The method according to claim 2, characterized in that If the first solution task scheduling adaptation coefficient is greater than / equal to the second solution task scheduling adaptation coefficient, the first solution task scheduling adaptation coefficient is set as the current winning solution task scheduling adaptation coefficient, and the first cluster task scheduling solution is set as the current winning cluster task scheduling solution.
5. The method according to claim 1, characterized in that Based on the digital twin control decision module, control decision analysis is performed on the M civil heterogeneous unmanned control units according to the M unit allocation subtasks to obtain a distributed control solution, including: The digital twin control decision module includes a digital twin module and a mission control decision module, and the mission control decision module includes M mission control decision models corresponding to the M civil heterogeneous unmanned control units; Inputting the M unit allocation subtasks into the M task control decision models to obtain M matching task control decisions; Based on the digital twin module, fault fitting optimization is performed on the M matching task control decisions respectively to generate the distributed control scheme.
6. The method according to claim 5, characterized in that Based on the digital twin module, the M matching task control decisions are respectively optimized for fault fitting to generate the distributed control scheme, including: Extracting the mth matching mission control decision corresponding to the mth civil heterogeneous unmanned control unit according to the M matching mission control decisions, wherein m is a positive integer and m belongs to M; Based on the digital twin module, modeling is performed according to the mth civil heterogeneous unmanned control unit to obtain the mth civil heterogeneous unmanned control model; Based on the digital twin module, simulate and control the mth civil heterogeneous unmanned control model according to the mth matching task control decision to obtain an mth decision simulation working condition data set; Performing fault prediction based on the mth decision simulation condition data set to obtain an mth decision fault prediction result; Performing a fault impact assessment based on the mth decision fault prediction result to obtain an mth predicted fault impact assessment coefficient; If the m-th predicted fault impact evaluation coefficient satisfies the preset fault impact evaluation constraint, outputting the m-th matching task control decision as the m-th unit task control decision, and adding the m-th unit task control decision to the distributed control scheme; If the mth predicted fault impact evaluation coefficient does not satisfy the preset fault impact evaluation constraint, the mth matching task control decision is optimized and adjusted according to the mth decision fault prediction result to generate the mth unit task control decision.
7. The method according to claim 1, characterized in that Executing control of the M civilian heterogeneous unmanned control units according to the distributed control scheme includes: Performing real-time monitoring on the M civilian heterogeneous unmanned control units to obtain M unit monitoring data streams; Performing anomaly recognition and evaluation based on the M unit monitoring data streams to obtain M unit monitoring anomaly recognition results and M unit monitoring anomaly evaluation coefficients; Determine whether the monitoring anomaly evaluation coefficients of the M units are greater than / equal to the monitoring anomaly evaluation thresholds; If any one of the M unit monitoring abnormality evaluation coefficients is greater than / equal to the monitoring abnormality evaluation threshold, a monitoring abnormality control adjustment instruction is generated; Based on the monitoring abnormality control adjustment instruction, the M civilian heterogeneous unmanned control units and the M unit monitoring abnormality identification results are abnormally marked to generate marked unmanned control units and marked monitoring abnormality identification results; Based on the distributed control scheme, abnormal compensation control is performed on the identification unmanned control unit according to the identification monitoring abnormality recognition result.
8. A civil heterogeneous unmanned cluster control system based on distributed control, characterized in that: A method for controlling a civilian heterogeneous unmanned cluster based on distributed control according to any one of claims 1 to 7, the system comprising: A control unit acquisition component, the control unit acquisition component is used to obtain M civilian heterogeneous unmanned control units of the civilian heterogeneous unmanned cluster, where M is a positive integer greater than 1; An information collection component, the information collection component is used to respectively collect basic information of the M civilian heterogeneous unmanned control units and fixed civilian task information to obtain characteristic data of the M control units; An instruction acquisition component, the instruction acquisition component is used to obtain a newly added civil task scheduling instruction, wherein the newly added civil task scheduling instruction includes newly added civil task information and newly added civil task time constraints; A scheduling scheme acquisition component, the scheduling scheme acquisition component is used to perform task scheduling on the M civilian heterogeneous unmanned control units based on the civilian cluster task scheduling algorithm, according to the characteristic data of the M control units, the newly added civilian task information and the newly added civilian task time constraints, to obtain a civilian cluster task scheduling scheme, wherein the civilian cluster task scheduling scheme includes M unit allocation subtasks corresponding to the M civilian heterogeneous unmanned control units; A control decision analysis component, the control decision analysis component is used to perform control decision analysis on the M civil heterogeneous unmanned control units according to the M unit allocation subtasks based on the digital twin control decision module to obtain a distributed control solution, wherein the distributed control solution includes M unit task control decisions; A control scheme execution component, wherein the control scheme execution component is used to execute control of the M civilian heterogeneous unmanned control units according to the distributed control scheme.
