Remote unmanned aerial vehicle supervisory control system and control method thereof

Abstract

The present application relates to the technical field of autonomous cooperative control of unmanned aerial vehicle cluster, in particular to a remote unmanned aerial vehicle supervision control system and a control method thereof, comprising: collecting environmental data of the unmanned aerial vehicle and collecting state vectors of the unmanned aerial vehicle; calculating a task contract deviation index based on the environmental data and the state vectors; in response to the task contract deviation index being greater than a preset deviation threshold, triggering task contract reconstruction, including: generating a candidate reconstruction scheme; calculating utility values of each adjacent node for the candidate reconstruction scheme; determining a preliminary reconstruction decision based on the utility values; solving a task intention fidelity for the preliminary reconstruction decision; if the task intention fidelity is greater than or equal to a preset fidelity threshold, confirming the preliminary reconstruction decision as a final task contract; if the task intention fidelity is less than the preset fidelity threshold, rejecting the preliminary reconstruction decision; the present application realizes a technical leap from passive response to active prediction, greatly improving the sensitivity and accuracy of risk warning.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of autonomous cooperative control of unmanned aerial vehicle (UAV) clusters, and in particular to a remote UAV supervisory control system and a control method thereof. BACKGROUND

[0002] With the wide application of UAV cluster technology, the dynamicity and uncertainty of its task execution environment are increasing. When a UAV cluster executes a preset task, unexpected situations such as dynamic obstacles, communication signal interference, or abnormal energy state of the UAV cluster may occur, which may result in that the original task planning is no longer applicable.

[0003] Currently, the response to these unexpected situations usually relies on remote manual intervention. When a UAV encounters an abnormal situation, the system reports the problem to a remote monitoring center, and an operator analyzes, decides, and issues a new instruction. This centralized control mode has a long response chain and poor real-time performance. In particular, in a large-scale cluster task, the burden of manual monitoring is huge, and it is difficult to timely handle the unexpected situations of each node. In addition, traditional autonomous obstacle avoidance or fault response strategies usually only focus on the safety of individual UAVs, lack of evaluation and constraint mechanisms for whether local decision-making behavior will affect the global task goal, and may result in that a locally optimal decision damages the final completion of the overall task.

[0004] Therefore, how to establish a decentralized autonomous decision-making mechanism to enable the UAV cluster to evaluate task risks in real time without continuous manual intervention, quickly respond to dynamic changes through local coordination and negotiation, and ensure that all local adjustment behaviors always serve the ultimate intention of the global task to improve the robustness and success rate of task execution has become a technical problem urgently to be solved in the field. SUMMARY

[0005] To solve the above technical problems, the present application provides a remote UAV supervisory control system and a control method thereof. Specifically, the technical solution of the present application is as follows:

[0006] A remote UAV supervisory control method, comprising:

[0007] collecting environment data of the UAV and collecting a state vector of the UAV;

[0008] calculating a task contract deviation index based on the environment data and the state vector;

[0009] in response to the task contract deviation index being greater than a preset deviation threshold, triggering task contract reconstruction, comprising:

[0010] generating a candidate reconstruction scheme;

[0011] calculating utility values of each neighboring node for the candidate reconstruction scheme;

[0012] Based on utility values, a preliminary restructuring decision was determined;

[0013] For the initial reconstruction resolution, calculate the fidelity of the task intent;

[0014] If the fidelity of the task intent is greater than or equal to the preset fidelity threshold, then the preliminary reconstruction decision is confirmed as the final task contract.

[0015] If the fidelity of the task intent is less than the preset fidelity threshold, the initial reconstruction decision is rejected.

[0016] Preferably, the calculation of the task contract deviation index includes:

[0017] Identify path risk factors, communication quality factors, and energy mismatch factors;

[0018] Based on path risk factors, communication quality factors, and energy mismatch factors, a task contract deviation index is generated through weighted summation.

[0019] Preferably, the calculation of utility value includes:

[0020] Determine the task gain factor, and combine risk factors and additional cost factors;

[0021] The utility value is generated based on the task gain factor, combined with the risk factor and additional cost factor.

[0022] Preferably, a preliminary restructuring decision is determined, including:

[0023] The utility values ​​of each neighboring node are aggregated to generate the collective total utility of each candidate reconstruction scheme;

[0024] The candidate restructuring scheme with the highest collective total utility is selected as the initial restructuring decision.

[0025] Preferably, the fidelity of the task intent is calculated, including:

[0026] Determine the degree of deviation from the target and the degree of violation of the core constraints;

[0027] Based on the deviation from the target and the violation of the core constraints, the fidelity of the task intent is generated through penalty weighting.

[0028] Preferably, the state vector includes: the UAV's position information, velocity information, remaining energy information, and current communication quality information.

[0029] Preferably, the additional overhead factor is determined based on the estimated additional computational overhead and estimated additional energy overhead of executing the candidate refactoring scheme.

[0030] A remote unmanned aerial vehicle (UAV) monitoring and control system includes:

[0031] a data collection module configured to collect environmental data of the UAV and collect a state vector of the UAV;

[0032] a deviation quantification module configured to calculate a task contract deviation index based on the environmental data and the state vector;

[0033] a contract reconstruction module triggered in response to the task contract deviation index being greater than a preset deviation threshold, and configured to perform task contract reconstruction, the contract reconstruction module comprising:

[0034] a scheme generation unit configured to generate a candidate reconstruction scheme;

[0035] a utility evaluation unit configured to calculate utility values of each adjacent node for the candidate reconstruction scheme;

[0036] a consensus decision unit configured to determine a preliminary reconstruction decision based on the utility values;

[0037] a fidelity verification unit configured to calculate a task intent fidelity for the preliminary reconstruction decision;

[0038] a closed-loop control module configured to: if the task intent fidelity is greater than or equal to a preset fidelity threshold, confirm the preliminary reconstruction decision as a final task contract; and if the task intent fidelity is less than the preset fidelity threshold, reject the preliminary reconstruction decision.

[0039] Compared with the prior art, the present application has the following beneficial effects:

[0040] 1. The technical solution collects the environmental data and the state vector of the UAV, and combines the path risk factor, the communication quality factor and the energy mismatch factor to generate a comprehensive task contract deviation index through weighted sum. This method breaks through the limitation of traditional schemes that rely on a single indicator for post-response, realizes multi-dimensional and customizable proactive prediction of potential task failure risks, and compares the deviation index with a preset threshold. The present solution can more accurately identify risks such as entering a communication blind area or running out of energy, thereby realizing a technical leap from passive response to proactive prediction, and greatly improving the sensitivity and accuracy of risk warning.

[0041] 2.The method introduces the utility theory of microeconomics into the decentralized negotiation process of the UAV cluster, provides a unified quantitative evaluation standard for decision-making, and particularly, creatively includes the estimated calculation overhead and energy overhead into the additional overhead factor for evaluation, solves the problem of ignoring internal resource consumption in the prior art, and ensures that the made decision not only considers the adaptability of the external environment, but also takes into account the sustainability of the internal resources of the cluster, avoids the selfish decision-making of a single UAV based on one-sided information which may damage the collective interests, and improves the collaboration and overall task execution capability of the cluster;

[0042] 3.The method designs a task intention fidelity verification mechanism to ensure that the local autonomous decision-making does not violate the core task target of the upper layer, quantifies the target deviation degree and the core constraint violation degree, and generates the task intention fidelity in combination with the penalty weight, sets an insurmountable safety constraint for the distributed autonomous decision-making of the UAV, effectively solves the problem that the local optimum does not equal the global optimum which generally exists in the distributed intelligent agent, verifies the fidelity of the preliminary reconstruction decision, and only confirms the execution when the requirement is met, so that the method ensures that the UAV is flexible in response to the complex environment, and the behavior of the UAV is always constrained within the framework of meeting the global task target, thereby guaranteeing the final success of the task. BRIEF DESCRIPTION OF DRAWINGS

[0043] The application will be further explained below in combination with the drawings and embodiments.

[0044] Figure 1 is a flowchart of the method of the application;

[0045] Figure 2 is a structural diagram of the system of the application. DETAILED DESCRIPTION

[0046] To make the purpose, technical scheme and advantages of the application clearer and more apparent, the application will be further described in detail below in combination with specific embodiments.

[0047] Embodiment 1

[0048] Please refer to Figure 1 A remote UAV supervision control method, characterized in that it comprises:

[0049] collecting environmental data of the UAV and collecting a state vector of the UAV;

[0050] calculating a task contract deviation index based on the environmental data and the state vector;

[0051] in response to the task contract deviation index being greater than the preset deviation threshold, triggering task contract reconstruction, including:

[0052] generating a candidate reconstruction scheme;

[0053] for the candidate reconstruction scheme, calculating the utility value of each adjacent node;

[0054] based on the utility value, determining a preliminary reconstruction decision;

[0055] for the preliminary reconstruction decision, solving the task intention fidelity;

[0056] if the task intention fidelity is greater than or equal to the preset fidelity threshold, confirming the preliminary reconstruction decision as the final task contract;

[0057] if the task intention fidelity is less than the preset fidelity threshold, rejecting the preliminary reconstruction decision.

[0058] The embodiment of the application provides a remote unmanned aerial vehicle supervision control method, aiming to solve the problem of low task execution efficiency and high failure risk of an unmanned aerial vehicle cluster in executing a preset task, i.e., a task contract, due to encountering a dynamically changing environment or an abnormal state of the unmanned aerial vehicle cluster; the method constructs a closed-loop control logic of monitoring-negotiation-verification-execution, so that the unmanned aerial vehicle can make distributed local autonomous decision-making with the global task intention as the boundary, thereby significantly improving the resilience and success rate of the task;

[0059] In a specific application scenario, an unmanned aerial vehicle cluster is assigned to execute a task contract containing a specific flight path planning and a reconnaissance target; during task execution, the method flow described in the embodiment is as follows:

[0060] environmental data of the unmanned aerial vehicle is collected, and a state vector of the unmanned aerial vehicle is collected; the purpose of this step is to provide real-time and accurate data input for subsequent risk assessment and decision-making; in this embodiment, the environmental data refers to unstructured data about the physical environment around the unmanned aerial vehicle, such as the position and speed of dynamic obstacles, the range of signal blind areas, etc., which is obtained in real time by the airborne sensors such as laser radar and camera carried by the unmanned aerial vehicle; these data are used to construct a local environment model of the unmanned aerial vehicle; the state vector refers to a data set containing key performance indicators of the unmanned aerial vehicle, which is used to comprehensively represent the current running state of the unmanned aerial vehicle; in this embodiment, each unmanned aerial vehicle node periodically broadcasts its state vector to its adjacent nodes through an adjacent communication protocol;

[0061] Based on the environmental data and the state vector, a task contract deviation index is calculated; the purpose of this step is to quantify the degree of inconsistency between the current state of the UAV and the preset task contract, generating a clear criterion that can be used to trigger the subsequent decision-making process; this index is a comprehensive risk warning indicator, rather than a post-response after the task has completely failed;

[0062] In response to the task contract deviation index being greater than the preset deviation threshold, the task contract reconstruction is triggered; the preset deviation threshold, i.e. , refers to a pre-set critical value for determining whether the current task deviation has reached an unacceptable level, which serves as a risk warning criterion; the determination method of the threshold can be: based on the statistical analysis of the deviation index in the normal execution state of a large amount of historical task data, the 95% quantile is taken as the threshold, to ensure that only when a statistically significant anomaly occurs, the reconstruction is started, avoiding excessive consumption of system resources; when the calculated task contract deviation index exceeds this threshold, it indicates that the risk of the current UAV node continuing to execute the task according to the original plan is too high or the efficiency is too low, and the dynamic adjustment process of the task contract must be started;

[0063] A candidate reconstruction scheme is generated: the UAV that triggered the reconstruction generates one or more local task plan amendments that aim to avoid risks or improve efficiency based on its perceived local environmental model; for example, a set of alternative waypoints that can bypass dynamic obstacles is generated;

[0064] The utility value of each adjacent node is calculated for the candidate reconstruction scheme: the negotiation initiator broadcasts the generated candidate scheme to its adjacent nodes; each adjacent node independently evaluates each scheme and calculates a utility value; the purpose of the utility value is to quantitatively evaluate the benefit-cost ratio of the candidate scheme from the perspective of each relevant node, providing a unified and quantitative evaluation standard for subsequent consensus decision-making;

[0065] Based on the utility value, a preliminary reconstruction resolution is determined: the negotiation initiator collects the utility values fed back by all adjacent nodes and collectively evaluates each candidate scheme, selecting an optimal scheme as the preliminary reconstruction resolution; the purpose of this step is to reach a consensus within a local range and select a scheme that is most beneficial to the overall local cluster;

[0066] The task intention fidelity is calculated for the preliminary reconstruction resolution; the purpose of this step is to check whether the above local optimal solution conflicts with the global task goal; the task intention fidelity refers to a quantitative evaluation of how much the preliminary reconstruction resolution remains faithful to the original task core goal; this step is one of the key innovations of the invention, which sets an insurmountable autonomous decision-making safety constraint for the distributed autonomous decision-making of the UAV, ensuring that local optimization behavior does not harm the final achievement of the global task;

[0067] If the task intention fidelity is greater than or equal to a preset fidelity threshold, the preliminary reconstruction decision is confirmed as the final task contract; the preset fidelity threshold refers to a minimum acceptable standard for determining whether the task intention is sufficiently maintained; its setting logic depends on the criticality of the task, for example, for high-precision mapping tasks, the threshold can be set to 0.95 to strictly limit the deviation of the final position; if the fidelity meets the requirements, the preliminary reconstruction decision is confirmed as a new and valid task contract, and is broadcast by the initiator to the relevant nodes to start execution;

[0068] If the task intention fidelity is less than the preset fidelity threshold, the preliminary reconstruction decision is rejected; this means that the local optimal solution seriously damages the global task goal, so it must be rejected; at this time, the system can choose to return to the step of generating candidate reconstruction schemes to regenerate the scheme, or report the decision conflict to the remote monitoring center after multiple failures and request manual intervention;

[0069] The present application establishes a set of remote unmanned aerial vehicle adaptive supervision control mechanism through the above complete closed-loop control method; it can quantitatively measure the task risk in real time, and through distributed negotiation and decision-making process with global intention verification, the unmanned aerial vehicle cluster can safely and efficiently cope with complex dynamic environment without continuous manual intervention, significantly enhancing the autonomy, robustness and success rate of task execution.

[0070] Embodiment 2:

[0071] Calculate the task contract deviation index, including:

[0072] Determine the path risk factor, the communication quality factor and the energy mismatch factor;

[0073] Based on the path risk factor, the communication quality factor and the energy mismatch factor, the task contract deviation index is generated by weighted sum.

[0074] This embodiment is a specific implementation of the technical feature of calculating the task contract deviation index; in order to establish a quantitative trigger mechanism that can predict task risk, this embodiment uses the weighted sum model in the multi-criteria decision-making theory to comprehensively consider multiple key factors affecting task execution;

[0075] Determine the path risk factor, the communication quality factor and the energy mismatch factor; these three factors are core indicators extracted from the three most critical dimensions affecting task execution: safety, communication stability and energy sustainability;

[0076] Based on the path risk factor, the communication quality factor and the energy mismatch factor, the task contract deviation index is generated by weighted sum; the calculation is performed by the deviation quantification module, and the specific calculation formula is as follows:​

[0077] ;

[0078] wherein, : mission contract deviation index, is the comprehensive deviation index of the UAV node , which is a dimensionless scalar, whose calculation result will be compared with the preset deviation threshold ;

[0079] : path risk factor, which quantifies the physical safety risk on the preset route; in this embodiment, a feasible calculation method is to define a cylindrical safety channel body with a radius of around the preset route, is calculated as the ratio of the volume occupied by dynamic obstacles or signal blind areas in the channel body to the total volume of the channel body; wherein, the data of dynamic obstacles or signal blind areas is derived from the on-board environmental sensor; the value is a dimensionless value in the interval , the higher the value, the greater the path risk;

[0080] : communication quality factor, which quantifies the stability of the communication link between the UAV and the cluster or the ground station; in this embodiment, the factor is calculated based on the received signal strength indication RSSI:

[0081] ;

[0082] wherein and are the preset best and worst thresholds of signal strength; the RSSI value here is collected in real time by the UAV communication module, and the value is a dimensionless value in the interval , the higher the value, the worse the communication quality;

[0083] : energy mismatch factor, which predicts whether the remaining energy of the UAV is sufficient to complete the subsequent task; in this embodiment, the factor is calculated according to the current remaining energy and the task model, the ratio of the estimated energy consumption for executing the remaining task to the current available energy; the current available energy value is provided by the on-board battery management system; the value is a dimensionless value, and when its value is greater than 1, it indicates that the energy is insufficient;

[0084] : weight coefficient, which is a dimensionless preset parameter corresponding to the above three factors respectively, and the sum is 1; its source is the senior task planning stage, which is artificially configured according to the characteristics of different tasks, or the output of a machine learning model based on historical task data;

[0085] Compared with the traditional method which only relies on a single indicator, the deviation index calculation method proposed in this embodiment provides a multi-dimensional and customizable comprehensive risk assessment model. It can identify potential task failure risks earlier and more accurately, realizes the technical leap from passive response to active prediction, and greatly improves the sensitivity and accuracy of risk warning.

[0086] Embodiment 3:

[0087] Calculate the utility value, including:

[0088] Determine the task gain factor, the comprehensive risk factor and the additional overhead factor;

[0089] Generate the utility value based on the task gain factor, the comprehensive risk factor and the additional overhead factor;

[0090] The additional overhead factor is determined according to the estimated additional computational overhead and the estimated additional energy overhead of the candidate reconstruction scheme.

[0091] This embodiment is a specific implementation of the technical feature of calculating the utility value of each adjacent node, and further limits the sub-feature of the additional overhead factor. This embodiment draws on the utility theory in microeconomics, aiming to provide a unified and quantitative evaluation standard for the decentralized negotiation process and guide the cluster to make local optimization;

[0092] Determine the task gain factor, the comprehensive risk factor and the additional overhead factor; these three factors respectively evaluate a candidate reconstruction scheme from the perspectives of revenue, risk and cost;

[0093] Generate the utility value based on the task gain factor, the comprehensive risk factor and the additional overhead factor; this calculation is performed by the utility evaluation unit, and the specific calculation formula is as follows:

[0094] ;

[0095] The smoothing processing of adding 1 to the numerator and the denominator in the formula is to ensure that the basic utility value is 1 and to avoid the denominator being zero;

[0096] Wherein, : utility value, is the utility value of the adjacent node to the candidate scheme ; the utility value evaluated by the candidate scheme is a dimensionless scalar, and the value will be fed back to the negotiation initiator;

[0097] : task gain factor, its role is to quantify the additional task revenue that the scheme may bring, its value range is ; a feasible calculation method is: , wherein , , and The estimated value is based on simulation and deduction of the path and task model of scheme p; if the scheme only avoids risks and has no additional task gains, then ;

[0098] Comprehensive risk factors serve to quantify the implementation plan. The associated comprehensive risk; a feasible calculation method is as a weighted sum of different risk probabilities: ,in It is the probability of occurrence of the k-th risk predicted by combining the trajectory of scheme p with the environmental model. This is the corresponding risk weight; this value is... Dimensionless values ​​of the interval;

[0099] Additional overhead factor, which is used to quantify the execution plan. The required marginal resource consumption is determined based on the estimated additional computational overhead and estimated additional energy overhead of executing the candidate reconfiguration scheme; its innovation lies in incorporating computational resources into the decision-making considerations; in this embodiment, the calculation formula is as follows:

[0100] ;

[0101] in and This is a projected estimate obtained by analyzing the computational complexity and energy consumption model of the algorithm required for execution plan p. It is the maximum computing load. It is residual energy. and These are preset weighting coefficients, derived from the characteristics of the UAV hardware platform to determine the relative importance of computing and energy resources.

[0102] The adjustment coefficient is a dimensionless preset parameter for risk and cost, which is derived from the global strategy configuration of the task and is used to adjust the risk preference of the decision model.

[0103] This utility value calculation method provides a refined decision-making model for distributed negotiation among UAVs. It goes beyond the simple logic of choosing the lesser of two evils and instead achieves a quantitative trade-off between benefits, risks, and costs. It creatively incorporates the previously overlooked internal resource consumption of computing overhead and energy overhead into the evaluation system, so that the decision not only focuses on the adaptability to the external environment but also takes into account the sustainability of internal resources, thereby generating more feasible and efficient reconfiguration solutions.

[0104] Example 4:

[0105] determining a preliminary reconstruction decision, comprising:

[0106] aggregating the utility values of the neighboring nodes to generate a collective total utility of each candidate reconstruction scheme;

[0107] selecting the candidate reconstruction scheme with the highest collective total utility as the preliminary reconstruction decision.

[0108] This embodiment is a specific implementation of the technical feature of determining a preliminary reconstruction decision; the purpose of this method is to form a consensus decision that can maximize the local collective benefit based on the independent evaluation results of each neighboring node on the candidate scheme;

[0109] aggregating the utility values of the neighboring nodes to generate a collective total utility of each candidate reconstruction scheme; the negotiation initiator node for each candidate scheme the utility value of the feedback After that, the consensus decision unit built-in generates the collective total utility of each scheme ; in this embodiment, the calculation method is direct summation:

[0110] ;

[0111] selecting the candidate reconstruction scheme with the highest collective total utility as the preliminary reconstruction decision; the consensus decision unit sorts the collective total utilities of all candidate schemes, and determines the scheme with the highest value as the preliminary reconstruction decision;

[0112] This embodiment realizes a simple and efficient distributed consensus mechanism by aggregating the utility values and selecting the scheme with the highest total utility; this collective decision-making mode avoids the possibility of a single UAV making a selfish decision that may harm the interests of neighboring nodes based on its own one-sided information; it ensures that the selected preliminary reconstruction decision is the most beneficial scheme for the entire neighboring cluster within the scope of local information, thereby promoting the synergy and stability of UAV cluster action and improving the overall survival and task execution capability of the local cluster.

[0113] Embodiment 5:

[0114] solving the task intention fidelity, comprising:

[0115] determining the target deviation degree and the core constraint violation degree;

[0116] based on the target deviation degree and the core constraint violation degree, generating the task intention fidelity through penalty weight processing.

[0117] ​This embodiment is a specific implementation of the technical feature of solving the task intention fidelity; this method aims to provide a top-level constraint for distributed autonomous decision-making, ensuring that any local, bottom-up task adjustment will not violate the high-level, top-down core task intention, which is the core constraint verification mechanism to ensure the final success of the task;

[0118] Determine the target deviation degree and the core constraint violation degree; these two indicators respectively review the preliminary reconstruction decision from the two dimensions of whether to deviate from the ultimate goal and whether to violate the key prohibition;

[0119] Based on the target deviation degree and the core constraint violation degree, the task intention fidelity is generated through the penalty weight processing; this calculation is performed by the fidelity verification unit, which uses a self-defined constraint satisfaction degree evaluation model, and the formula is as follows:

[0120] ;

[0121] Among them, : task intention fidelity, is a dimensionless value in the interval [0, 1], and its calculation result is used for comparison with the minimum fidelity threshold ; represents complete fidelity;

[0122] : target deviation degree, which quantifies the deviation between the final state of the new scheme and the original task target; for end position type targets, a feasible calculation method is: , where is the Euclidean distance between the new scheme end point and the original task end point, which is obtained by geometric operation on the waypoint data of the new and old schemes, is the preset maximum allowed deviation distance and is constrained to be a positive value. In the limit case where approaches 0, any non-zero will cause to tend to infinity, and in actual calculation, a very large penalty value can be set;

[0123] : core constraint violation degree, which quantifies the degree of violation of the new scheme to the core task constraint; for the spatial no-fly zone constraint, a feasible calculation method is: , where is the length of the segment that falls into the no-fly zone in the new scheme path, which is obtained by overlaying the new path with the preset no-fly zone geographic information, is the total length of the segment, and in calculation, if is 0, then the corresponding must be 0, and ​Should be recorded as 0 to avoid division by zero error, if not violated, then ;

[0124] : penalty weight, dimensionless preset parameter, its source is the priority ranking of different task targets in task planning; for example, if reaching the specified terminal point is the highest priority task, then A larger value will be assigned to define the severity of violating the corresponding intention;

[0125] The task intention fidelity verification mechanism proposed in this embodiment establishes a key security layer in the distributed autonomous system of the UAV cluster; it effectively solves the problem that the local optimum does not equal the global optimum in the distributed intelligent agent; through the quantitative punishment of target deviation and core constraint violation, this method ensures that the UAV enjoys the flexibility of local autonomous decision-making while its behavior is always constrained within the framework of meeting the global task target, thereby fundamentally ensuring the success of the entire task.

[0126] Embodiment 6:

[0127] The state vector includes: the position information, speed information, remaining energy information and current communication quality information of the UAV.

[0128] This embodiment is a specific content of the technical feature of the state vector; a well-defined and comprehensive state vector is the data basis for the effective operation of the entire supervision control method;

[0129] In this embodiment, the state vector is a structured data packet, which includes: the position information, speed information, remaining energy information and current communication quality information of the UAV;

[0130] Position information: usually three-dimensional coordinates, which is the basis for path planning, obstacle avoidance and cooperative positioning;

[0131] Speed information: usually a three-dimensional velocity vector, used to predict future trajectory and assess dynamic risk;

[0132] Remaining energy information: usually the percentage of battery remaining capacity or estimated remaining flight time, which is the key input for calculating and evaluating the energy mismatch factor of the reconstruction scheme;

[0133] Current communication quality information: usually the RSSI value with the neighboring nodes or base station, which is the direct basis for calculating and evaluating the network topology stability of the communication quality factor;

[0134] By combining the four specific information into a state vector, the application ensures that the most core and dynamic data required by the decision system can be acquired and shared in time and completeness. The combination of this information forms a direct data support relationship with the needs of the aforementioned deviation quantification, utility evaluation and other algorithm modules: position and speed information support path risk assessment, remaining energy information directly input energy mismatch factor calculation, and communication quality information directly input communication quality factor calculation. This highly relevant parameter selection makes every step of the entire control method calculation have a solid data foundation, thereby greatly improving the accuracy and timeliness of decision-making.

[0135] Embodiment 7:

[0136] Please refer to Figure 2 A remote unmanned aerial vehicle supervision control system comprises:

[0137] A data acquisition module is configured to acquire environmental data of the unmanned aerial vehicle and acquire a state vector of the unmanned aerial vehicle.

[0138] A deviation quantification module is configured to calculate a task contract deviation index based on the environmental data and the state vector.

[0139] A contract reconstruction module is triggered in response to the task contract deviation index being greater than a preset deviation threshold, and is configured to perform task contract reconstruction. The contract reconstruction module comprises:

[0140] A scheme generation unit is configured to generate a candidate reconstruction scheme.

[0141] A utility evaluation unit is configured to calculate utility values of each adjacent node for the candidate reconstruction scheme.

[0142] A consensus decision unit is configured to determine a preliminary reconstruction resolution based on the utility values.

[0143] A fidelity verification unit is configured to calculate a task intention fidelity for the preliminary reconstruction resolution.

[0144] A closed-loop control module is configured to: if the task intention fidelity is greater than or equal to a preset fidelity threshold, confirm the preliminary reconstruction resolution as a final task contract; and if the task intention fidelity is less than the preset fidelity threshold, veto the preliminary reconstruction resolution.

[0145] The application also provides a remote unmanned aerial vehicle supervision control system, the internal structure and function modules of which correspond to the method steps one by one.

[0146] The remote unmanned aerial vehicle supervision control system comprises:

[0147] The data acquisition module can be physically implemented as a set of software drivers and interfaces integrated into the UAV flight control system, used to connect the onboard GPS, IMU, battery management system (BMS), wireless communication module, and various environmental sensors; this module is configured to perform the data acquisition steps in the aforementioned method, namely, to acquire the UAV's environmental data and the UAV's state vector.

[0148] The deviation quantification module, physically implemented as a dedicated algorithm process running on the UAV's main control chip, receives data from the data acquisition module and is used to calculate the mission contract deviation index based on environmental data and state vectors. ;

[0149] The contract restructuring module, which is the core of the system's decision-making process, is triggered when the task contract deviation index exceeds a preset deviation threshold and is used to perform task contract restructuring. Logically, it consists of the following units:

[0150] The scheme generation unit is used to generate candidate reconstruction schemes based on the local environment model;

[0151] The utility evaluation unit calculates the utility value of each neighboring node for each candidate reconstruction scheme. ;

[0152] The consensus decision-making unit, based on utility values, determines the initial restructuring resolution. ;

[0153] The fidelity verification unit calculates the fidelity of the task intent based on the initial reconstruction decision. ;

[0154] The closed-loop control module is responsible for decision execution and process closure; it receives fidelity data from the fidelity verification unit. And it is used for: if the mission intent fidelity is greater than or equal to the preset fidelity threshold, the preliminary reconstruction decision is confirmed as the final mission contract and sent to the flight control module for execution; if the mission intent fidelity is less than the preset fidelity threshold, the preliminary reconstruction decision is rejected and the contract reconstruction module is triggered to renegotiate or report to the superior.

[0155] The system provided in this embodiment offers a clear and feasible hardware and software architecture for realizing the aforementioned remote UAV supervision and control method through its modular structural design. Each module has a clear responsibility and is highly decoupled, which not only ensures the complete implementation of the entire monitoring-negotiation-verification-execution closed-loop logic, but also facilitates the deployment, upgrading and maintenance of the system. The system transforms complex methodologies into concrete, engineerable entities, providing solid technical support for improving the autonomous control level of UAV swarms.

[0156] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced, without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. A remote unmanned aerial vehicle (UAV) monitoring and control method, characterized in that, include: Collect environmental data from the drone and collect the drone's state vector; Calculate the task contract deviation index based on environmental data and state vectors; If the task contract deviation index exceeds a preset deviation threshold, a task contract refactoring is triggered, including: Generate candidate reconstruction schemes; For each candidate reconstruction scheme, calculate the utility value of each neighboring node; Based on utility values, a preliminary restructuring decision was determined; For the initial reconstruction resolution, calculate the fidelity of the task intent; If the fidelity of the task intent is greater than or equal to the preset fidelity threshold, then the preliminary reconstruction decision is confirmed as the final task contract. If the fidelity of the task intent is less than the preset fidelity threshold, the initial reconstruction decision is rejected. Calculate the task contract deviation index, including: Identify path risk factors, communication quality factors, and energy mismatch factors; Based on path risk factors, communication quality factors, and energy mismatch factors, a task contract deviation index is generated through weighted summation. The path risk factor is calculated as the ratio of the volume occupied by dynamic obstacles or signal blind spots within the channel to the total volume of the channel. Communication quality factor Calculated based on Received Signal Strength Indicator (RSSI): ； in and The preset optimal and worst threshold values ​​for signal strength; The energy mismatch factor is calculated based on the current remaining energy and the task model, and is the ratio of the estimated energy consumption for executing the remaining tasks to the currently available energy. Calculating utility values ​​includes: Determine the task gain factor, and combine risk factors and additional cost factors; Additional overhead factor The calculation formula is: ； in and This is a projected estimate obtained by analyzing the computational complexity and energy consumption model of the algorithm required for execution plan p. It is the maximum computing load. It is residual energy. and These are preset weighting coefficients; Based on the task gain factor, combined with the risk factor and additional cost factor, a utility value is generated. The specific calculation formula is as follows: ； The formula adds 1 to both the numerator and denominator for smoothing, in order to ensure that the basic utility value is 1 and to avoid the denominator being zero; in, Utility value is the value of neighboring nodes. For candidate solutions The assessed utility value is a dimensionless scalar, which will be fed back to the negotiation initiator. Task gain factor, its function is to quantify the scheme. The potential additional task benefits have a value range of [value range missing]. The calculation method is as follows: ,in , , and The estimated value is based on simulation and deduction of the path and task model of scheme p; if the scheme only avoids risks and has no additional task gains, then ; Comprehensive risk factors serve to quantify the implementation plan. The associated comprehensive risk; the calculation method is as a weighted sum of different risk probabilities: ,in It is the probability of occurrence of the k-th risk predicted by combining the trajectory of scheme p with the environmental model. This is the corresponding risk weight; this value is... Dimensionless values ​​of the interval; Additional overhead factor; The fidelity of the task intent is calculated, including: Determine the degree of deviation from the target and the degree of violation of the core constraints; Based on the degree of deviation from the target and the degree of violation of the core constraints, the fidelity of the task intent is generated by processing the penalty weights. The formula is as follows: ； in, Task intent fidelity is a The dimensionless value of the interval, the result of which is used to compare with the minimum fidelity threshold. Compare; Indicates complete authenticity; Target deviation, its function is to quantify the deviation between the final state of the new solution and the original task objective; for endpoint location-type objectives, the calculation method is as follows: ,in It is the Euclidean distance between the endpoint of the new plan and the endpoint of the original mission, obtained by performing geometric calculations on the waypoint data of the old and new plans. It is the preset maximum allowable deviation distance and Constrained to a positive value; in In the limiting case approaching 0, any non-zero This will lead to Approaching infinity, an extremely large penalty value can be set in actual calculations; Core constraint violation degree: Its function is to quantify the degree to which the new scheme violates the core mission constraints; for the no-fly zone constraint, the calculation method is as follows: ,in This refers to the length of the no-fly zone segment within the new route, obtained by overlaying and analyzing the new route with pre-defined geographical information about the no-fly zone. It is the total length of the path segment. When calculating, if... If it is 0, then its corresponding It must be 0 at this time. It should be recorded as 0 to avoid division by zero errors; if this is not violated, then... ; The penalty weight is a dimensionless preset parameter derived from the priority ranking of different task objectives during task planning.

2. The remote unmanned aerial vehicle (UAV) monitoring and control method according to claim 1, characterized in that, The preliminary restructuring resolutions include: The utility values ​​of each neighboring node are aggregated to generate the collective total utility of each candidate reconstruction scheme; The candidate restructuring scheme with the highest collective total utility is selected as the initial restructuring decision.

3. The remote unmanned aerial vehicle (UAV) monitoring and control method according to claim 1, characterized in that, The state vector includes: the UAV's position information, velocity information, remaining energy information, and current communication quality information.

4. The remote unmanned aerial vehicle (UAV) monitoring and control method according to claim 1, characterized in that, The additional overhead factor is determined based on the estimated additional computational overhead and estimated additional energy overhead of executing candidate refactoring schemes.

5. A remote unmanned aerial vehicle (UAV) monitoring and control system, applied to the remote UAV monitoring and control method according to any one of claims 1-4, characterized in that, include: The data acquisition module is used to collect environmental data of the UAV and the UAV's state vector. The deviation metric module is used to calculate the task contract deviation index based on environmental data and state vectors; The contract restructuring module is triggered when the task contract deviation index exceeds a preset deviation threshold. It is used to perform task contract restructuring and includes: The scheme generation unit is used to generate candidate reconstruction schemes; The utility evaluation unit is used to calculate the utility value of each neighboring node for a candidate reconstruction scheme. Consensus decision-making unit, used to determine the initial restructuring resolution based on utility value; The fidelity verification unit is used to calculate the fidelity of the task intent for the initial reconstruction decision. The closed-loop control module is used to: confirm the preliminary reconstruction decision as the final task contract if the task intent fidelity is greater than or equal to the preset fidelity threshold; and reject the preliminary reconstruction decision if the task intent fidelity is less than the preset fidelity threshold.

Images