CN-122024528-A - Unmanned aerial vehicle and unmanned aerial vehicle cooperative flight early warning method and system

CN122024528ACN 122024528 ACN122024528 ACN 122024528ACN-122024528-A

Abstract

The invention provides a method and a system for pre-warning cooperative flight of a man-machine and an unmanned aerial vehicle, which relate to the technical field of aviation flight, and the method and the system are used for scientifically dividing and distributing task areas by acquiring the route of the man-machine and constructing a clear task dependency graph so as to clearly define the relation between each unmanned aerial vehicle and the task areas; the unmanned aerial vehicle automatic collaborative network system comprises a unmanned aerial vehicle, a safety evaluation model, a dynamic and flexible unmanned aerial vehicle task weight adjustment, a real-time monitoring, reassignment and continuous updating of safety level and early warning information, wherein the unmanned aerial vehicle sampling data is used for accurately dividing the environmental safety level through the safety evaluation model, providing a reliable basis for subsequent decisions, the unmanned aerial vehicle task weight is dynamically and flexibly adjusted based on the safety level, the effective task switching decision is realized by a switching value quantization index, and finally, an autonomous collaborative network mechanism which is continuously optimized from the beginning of a task to the end and adapts to the collaborative flight early warning of the unmanned aerial vehicle and the unmanned aerial vehicle is formed, the early warning response speed and the resource utilization efficiency under the complex environment are remarkably improved, and the safety of collaborative flight and the task completion quality are effectively improved.

Inventors

GE ZHIBIN
WANG LAI
CHEN RONG
WANG YUMING
ZHAO ZHIYUAN
WANG SHUAI
PAN JUN
LIU XINQIAO
ZHANG HAO
ZENG LI
ZHENG QUAN

Assignees

中国民用航空飞行学院
四川捷途宇博航空科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260304

Claims (8)

1. The method for pre-warning the cooperative flight of the unmanned aerial vehicle and the man-machine is characterized by comprising the following steps of: s1, acquiring a current planned route path of a unmanned aerial vehicle, dividing task areas based on the current planned route path, and initially distributing each unmanned aerial vehicle to the corresponding task area to obtain an initial distribution result of the unmanned aerial vehicle; S2, constructing a task dependency graph with the unmanned aerial vehicle as a node and the association relation between task areas as edges based on an initial allocation result of the unmanned aerial vehicle, and setting initial task weights for the unmanned aerial vehicles; s3, carrying out data sampling on the task areas through the initially allocated unmanned aerial vehicles to obtain current environment data of the task areas, and analyzing the current environment data through a pre-trained safety evaluation model to divide the safety level of the current environment of the task areas; s4, according to the safety level of the current environment of each task area, combining a task dependency graph of each unmanned aerial vehicle, automatically adjusting initial task weights for each unmanned aerial vehicle, and obtaining task weights after adjustment of each unmanned aerial vehicle; S5, calculating a switching value of task switching of each unmanned aerial vehicle based on the task weight adjusted by each unmanned aerial vehicle and the current environmental security level of each task area; And S6, when judging that the switching value of task switching of each unmanned aerial vehicle exceeds a preset switching threshold, reassigning each unmanned aerial vehicle to a corresponding task area, and repartitioning the current environmental security level for early warning.
2. The method for collaborative flight pre-warning of an unmanned aerial vehicle and an unmanned aerial vehicle according to claim 1, wherein the specific steps of step S1 include: s101, acquiring a planned route path planned by the man-machine in real time through a flight control system of the man-machine; S102, dividing a space range covered by the man-machine flight path into a plurality of task areas which are mutually related according to a planned route path planned by the man-machine at present; S103, according to the preset number of unmanned aerial vehicles and each task area, adopting an average allocation mode to initially allocate each unmanned aerial vehicle to the corresponding task area so as to obtain an initial allocation result of the unmanned aerial vehicle.
3. The method for collaborative flight pre-warning of an unmanned aerial vehicle and an unmanned aerial vehicle according to claim 1, wherein the specific step of step S2 comprises: s201, constructing a task dependency graph by taking each unmanned aerial vehicle as a node and taking the association relation between task areas as edges based on the obtained initial allocation result of the unmanned aerial vehicle; S202, setting initial task weights corresponding to the performance of the unmanned aerial vehicles according to the performance of the unmanned aerial vehicles, and recording the initial task weights in node attributes of a task dependency graph, wherein the performance of the unmanned aerial vehicles comprises flying speed and cruising ability.
4. The method for collaborative flight pre-warning of an unmanned aerial vehicle and an unmanned aerial vehicle according to claim 1, wherein the specific step of step S3 comprises: s301, each unmanned aerial vehicle performs data sampling in a corresponding task area according to a preset sampling frequency and through each sensor to obtain current environment data of each task area, wherein the environment data comprise topography and meteorological condition data; S302, collecting historical environment data and corresponding historical evaluation scores of unmanned aerial vehicle sampling, extracting features from the historical environment data as a training set, inputting the training set into an adopted convolutional neural network model, training the convolutional neural network model by taking the historical evaluation scores as output, and obtaining a safety evaluation model after training; S303, extracting features of the current environment data of each task area, inputting the features into a security evaluation model, analyzing and outputting evaluation scores through the security evaluation model, and judging preset security level threshold sections to which the evaluation scores specifically belong to so as to divide the security level of the current environment of each task area, wherein the security level threshold sections comprise threshold sections corresponding to high security levels, medium security levels and low security levels in sequence.
5. The method for collaborative flight pre-warning of an unmanned aerial vehicle and an unmanned aerial vehicle according to claim 1, wherein the specific step of step S4 comprises: s401, associating the security level of the current environment of each task area to the node attribute of the corresponding unmanned aerial vehicle in the task dependency graph, and obtaining an association result of the security level and the task dependency graph; S402, setting a weight adjustment mechanism to automatically adjust initial task weights of all unmanned aerial vehicles based on the association result of the security level and the task dependency graph, wherein the set weight adjustment mechanism specifically comprises that for the unmanned aerial vehicle corresponding to the task area with the lowered security level, the task weight of the unmanned aerial vehicle is increased, and meanwhile, the weight of the adjacent unmanned aerial vehicle is increased according to the association strength proportion through the task dependency graph.
6. The method for collaborative flight pre-warning of an unmanned aerial vehicle and an unmanned aerial vehicle according to claim 1, wherein in step S5, the specific step of calculating the switching value of task switching of each unmanned aerial vehicle comprises: and for each unmanned aerial vehicle, calculating the product of the current task weight and the average safety level of the adjacent task area, and taking the numerical value of the product as a switching value.
7. The method for collaborative flight pre-warning of an unmanned aerial vehicle and an unmanned aerial vehicle according to claim 1, wherein the specific step of step S6 comprises: S601, monitoring the switching value of each unmanned aerial vehicle in real time, and if the switching value exceeds a preset switching threshold, reassigning each unmanned aerial vehicle to a task area with the highest switching value according to the priority of the switching value to obtain a reassignment result; And S602, based on the reassignment result, executing a new round of data sampling by the reassigned unmanned aerial vehicle, and iteratively updating the security level and the early warning information until the task is finished.
8. A co-flight pre-warning system for an unmanned aerial vehicle and an unmanned aerial vehicle, applying the co-flight pre-warning method for an unmanned aerial vehicle and an unmanned aerial vehicle according to any one of claims 1 to 7, wherein the pre-warning system comprises: The route distribution module is used for acquiring a current planned route path of the unmanned aerial vehicle, dividing task areas based on the current planned route path, and initially distributing each unmanned aerial vehicle to the corresponding task area to obtain an initial distribution result of the unmanned aerial vehicle; the dependency graph construction module is used for constructing a task dependency graph taking the unmanned aerial vehicle as a node and the association relation between task areas as edges based on the initial allocation result of the unmanned aerial vehicle, and setting initial task weights for all unmanned aerial vehicles; The environment sensing module is used for carrying out data sampling on the task areas through the initially distributed unmanned aerial vehicles so as to acquire current environment data of the task areas, and analyzing the current environment data through a pre-trained safety evaluation model so as to divide the safety level of the current environment of the task areas; The weight adjustment module is used for automatically adjusting initial task weights for the unmanned aerial vehicles according to the security level of the current environment of each task area and combining the task dependency graph of the unmanned aerial vehicle, and obtaining the task weights after adjustment of the unmanned aerial vehicles; the switching decision module is used for calculating a switching value of task switching of each unmanned aerial vehicle based on the task weight adjusted by each unmanned aerial vehicle and the current environmental security level of each task area; And the cooperative early warning module is used for re-distributing each unmanned aerial vehicle to a corresponding task area when judging that the switching value of task switching of each unmanned aerial vehicle exceeds a preset switching threshold value, and re-dividing the current environment security level for early warning.

Description

Unmanned aerial vehicle and unmanned aerial vehicle cooperative flight early warning method and system Technical Field The invention relates to the technical field of aviation flight, in particular to a method and a system for collaborative flight early warning of a man-machine and an unmanned plane. Background In the modern aviation field, the application of the collaborative flight operation of the man-machine and the unmanned aerial vehicle is increasingly wide, and a plurality of key fields such as military reconnaissance, environment monitoring, disaster relief, logistics transportation and the like are covered. The cooperative operation mode can fully exert the advantages of decision command of the unmanned aerial vehicle, the characteristics of flexibility, high efficiency, low cost and the like of the unmanned aerial vehicle, realize complementary advantages and greatly expand the range and capability of aviation operation. However, in the actual collaborative flight, there are the following technical problems: In the traditional collaborative flight task of the man-machine and the unmanned plane, the division of task areas often lacks scientific and reasonable planning basis. The division is usually simply based on geographical areas or general task areas, and the planned route of the man-machine is not fully considered. The division mode may cause mismatching of the task area and the flight track of the unmanned aerial vehicle, so that the unmanned aerial vehicle cannot execute tasks in the effective cooperative range of the unmanned aerial vehicle, and the efficiency of cooperative operation is reduced. For example, in an environmental monitoring task, if the task area division is not combined with a human-machine flight path, an oversubscription of a part of the area may occur, while an underserve of other critical areas may occur; When unmanned aerial vehicles are distributed to task areas, the existing methods mostly adopt fixed or simple distribution strategies, and lack flexibility. The allocation is generally completed once before the task starts, and the allocation is rarely adjusted according to actual conditions in the task execution process. This allocation is difficult to accommodate for dynamically changing task environments and emergency situations. In the cooperative flight process of the unmanned aerial vehicle and the unmanned aerial vehicle, the environmental safety level of the task area can be changed continuously along with time and task progress. However, the existing method lacks a dynamic adjustment mechanism for the task weight of the unmanned aerial vehicle, and is difficult to adjust the task allocation of the unmanned aerial vehicle in time according to the change of the environmental security level. Therefore, it is necessary to provide a method and a system for collaborative flight warning of a man-machine and an unmanned aerial vehicle, which solve the above technical problems. Disclosure of Invention In order to solve the technical problems, the invention provides a method and a system for collaborative flight early warning of a man-machine and an unmanned aerial vehicle, which are used for solving the problems of unreasonable task planning, inflexible allocation mechanism and poor task scheduling flexibility of the unmanned aerial vehicle in the prior art. The invention provides a method for pre-warning cooperative flight of a man-machine and an unmanned aerial vehicle, which comprises the following steps: s1, acquiring a current planned route path of a unmanned aerial vehicle, dividing task areas based on the current planned route path, and initially distributing each unmanned aerial vehicle to the corresponding task area to obtain an initial distribution result of the unmanned aerial vehicle; S2, constructing a task dependency graph with the unmanned aerial vehicle as a node and the association relation between task areas as edges based on an initial allocation result of the unmanned aerial vehicle, and setting initial task weights for the unmanned aerial vehicles; s3, carrying out data sampling on the task areas through the initially allocated unmanned aerial vehicles to obtain current environment data of the task areas, and analyzing the current environment data through a pre-trained safety evaluation model to divide the safety level of the current environment of the task areas; s4, according to the safety level of the current environment of each task area, combining a task dependency graph of each unmanned aerial vehicle, automatically adjusting initial task weights for each unmanned aerial vehicle, and obtaining task weights after adjustment of each unmanned aerial vehicle; S5, calculating a switching value of task switching of each unmanned aerial vehicle based on the task weight adjusted by each unmanned aerial vehicle and the current environmental security level of each task area; And S6, when judging that the switching value of task switching of each unma