Self-adaptive detection and detection system for low-altitude flight inspection

Abstract

The embodiment of the invention provides a self-adaptive detection and detection system for low-altitude flight inspection, and aims to improve the efficiency and safety of unmanned aerial vehicle inspection operation. Firstly, a plurality of unmanned aerial vehicle description information sets, unmanned aerial vehicle cluster description information and flight inspection task description information are acquired; the information covers performance parameters and task requirements of the unmanned aerial vehicles in different inspection scenes, then the information is loaded into a target adaptive inspection decision model for knowledge learning in advance, and target estimation confidence is generated by analyzing the information so as to evaluate the success probability of the unmanned aerial vehicle cluster in executing a specific inspection task. And when the target estimation confidence meets the set requirement, the corresponding unmanned aerial vehicle cluster is intelligently scheduled to execute the inspection task, so that efficient completion of the task is ensured, and the adaptability and flexibility of the unmanned aerial vehicle inspection operation are effectively improved through intelligent evaluation and dynamic scheduling.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles, and in particular to an adaptive detection and detection system for low-altitude flight inspection. Background Art

[0002] With the rapid development of drone technology, low-altitude flight inspections have been widely used in power line inspections, environmental monitoring, disaster assessment and other fields. However, in actual applications, drone inspections face many challenges, such as complex and changeable inspection environments, differences in drone performance, and diverse mission requirements. Traditional drone inspection scheduling methods often rely on manual experience and fixed rules, and are difficult to dynamically adjust according to the specific requirements of the inspection tasks and the real-time status of the drones, resulting in low inspection efficiency and increased safety risks.

[0003] In order to solve the above problems, relevant technologies have begun to explore data-driven UAV inspection scheduling methods. These methods collect and analyze UAV performance data, inspection environment information, and task requirements, and use advanced algorithms such as machine learning and deep learning to intelligently schedule and optimize UAV inspection tasks. However, most existing methods focus on the performance evaluation and task allocation of a single UAV, ignoring the synergy between UAV clusters and the impact of different inspection scenarios on UAV performance, making it difficult to achieve the optimal scheduling results. Summary of the invention

[0004] In order to at least overcome the above-mentioned deficiencies in the prior art, an object of the embodiments of the present application is to provide an adaptive detection and detection system for low-altitude flight inspections.

[0005] According to one aspect of the present application, there is provided an adaptive detection method for low-altitude flight inspection, the method comprising:

[0006] Obtain X drone description information sets, cluster description information, and flight inspection task description information, wherein the X drone description information sets include Y knowledge field description vectors of each drone in the X drones, the X drones are drones that are dispatched to a flight inspection task in a target inspection execution scenario in a target low-altitude inspection operation, the Y knowledge field description vectors of each drone include description vectors of each drone in multiple inspection execution scenarios, the multiple inspection execution scenarios are inspection execution scenarios with full authority authentication in the target low-altitude inspection operation, the cluster description information includes description vectors of each drone cluster in the Z drone clusters in the multiple inspection execution scenarios, the flight inspection task description information includes description vectors of the target inspection execution scenario, the Z drone clusters are multiple drone clusters generated by clustering the X drones according to the current cluster division strategy, and X, Y, and Z are all not less than 2;

[0007] The X drone description information sets, cluster description information and flight inspection task description information are loaded into a target adaptive inspection decision model learned in advance, and a target estimation confidence generated by the target adaptive inspection decision model is generated, wherein the target estimation confidence represents the confidence that the Z drone clusters corresponding to the Z drone clusters successfully complete the low-altitude flight inspection task in the flight inspection task under the condition that the Z drone clusters are used as the respective drone clusters scheduled to the flight inspection task;

[0008] Under the condition that the target estimation confidence meets the set requirements, the Z drone clusters are dispatched as the drone clusters in the one flight inspection task to obtain the adaptive detection results before the low-altitude flight inspection.

[0009] According to another aspect of the present application, a detection system is provided, which includes a processor and a machine-readable storage medium, wherein the machine-readable storage medium stores machine-executable instructions, and the machine-executable instructions are loaded and executed by the processor to implement the above-mentioned adaptive detection method for low-altitude flight inspection.

[0010] In the technical solutions provided in some embodiments of the present application, the embodiments of the present application integrate and analyze the multi-dimensional description information of individual drones, drone clusters, and flight inspection tasks, so as to realize the accurate evaluation of the ability of drone clusters to perform inspection tasks, and use the target adaptive inspection decision model learned by prior knowledge to generate the target estimation confidence to quantify the possibility of the drone cluster successfully completing the low-altitude flight inspection in a specific inspection task. Under the condition that the target estimation confidence meets the set requirements, the appropriate drone cluster can be intelligently dispatched to the inspection task, thereby ensuring the efficiency and safety of the inspection operation. That is, by intelligently evaluating the matching degree between the drone cluster and the inspection task, it can ensure that the dispatched drone cluster has the ability to complete the task, avoiding the delay or failure of the task caused by insufficient drone performance or improper cluster configuration, thereby significantly improving the efficiency of the inspection operation. Before dispatching the drone cluster, the ability of the cluster is comprehensively evaluated through the target adaptive inspection decision model, which effectively avoids the safety risks that may be caused by the performance defects of the drone or cluster, and enhances the safety and reliability of the inspection task. In addition, it can intelligently select the most suitable drone cluster for scheduling according to the specific needs of the inspection task, realize the optimal configuration of drone resources, and improve resource utilization efficiency. By introducing the target adaptive inspection decision model with pre-knowledge learning, it realizes the intelligent evaluation and decision-making of the inspection capability of the drone cluster, improves the intelligent level of inspection task scheduling, and reduces the possibility of manual intervention and decision-making errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be extracted in combination with these drawings without creative work.

[0012] Figure 1 A schematic diagram of a flow chart of an adaptive detection method for low-altitude flight inspection provided in an embodiment of the present application;

[0013] Figure 2 A schematic block diagram of the architecture of the detection system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0014] The following description is intended to enable one of ordinary skill in the art to implement and combine the present application, and the description is provided in the context of a specific application scenario and its requirements. It will be apparent to one of ordinary skill in the art that various changes may be made to the disclosed embodiments, and that the general principles defined in the present application may be applied to other embodiments and application scenarios without departing from the principles and scope of the present application. Therefore, the present application is not limited to the described embodiments, but should be given the broadest scope consistent with the claims.

[0015] Figure 1 It is a flow chart of an adaptive detection method for low-altitude flight inspection provided by an embodiment of the present application. The adaptive detection method for low-altitude flight inspection is introduced in detail below.

[0016] Step S110, obtaining X drone description information sets, cluster description information and flight inspection task description information, wherein the X drone description information sets include Y knowledge field description vectors of each drone in the X drones, the X drones are drones scheduled to a flight inspection task in a target inspection execution scenario in a target low-altitude inspection operation, the Y knowledge field description vectors of each drone include description vectors of each drone in multiple inspection execution scenarios, the multiple inspection execution scenarios are inspection execution scenarios with full authority authentication in the target low-altitude inspection operation, the cluster description information includes description vectors of each drone cluster in the Z drone clusters in the multiple inspection execution scenarios, the flight inspection task description information includes description vectors of the target inspection execution scenario, the Z drone clusters are multiple drone clusters generated by clustering the X drones according to the current cluster division strategy, and X, Y and Z are all not less than 2.

[0017] In detail, the drone description information set contains detailed description information of X drones participating in a flight inspection mission. Each drone has Y knowledge field description vectors, which cover the performance, status, historical data and other information of the drone in multiple inspection execution scenarios. The inspection execution scenario refers to the specific environment or conditions in which the drone performs the inspection mission. Different inspection scenarios may have different performance requirements for drones. The cluster description information contains the overall performance description of Z drone clusters in multiple inspection execution scenarios. Each drone cluster is composed of multiple drones, which can cooperate with each other when performing tasks. The drone cluster refers to a team composed of multiple drones, which can cooperate with each other and complete the task together when performing tasks. Drone clusters can improve the efficiency and success rate of inspection tasks.

[0018] The inspection task description information includes a specific description of the target inspection execution scenario, such as geographical features, task objectives, time limits, etc., which is crucial for evaluating whether the drone cluster is suitable for performing the task.

[0019] For example, in this embodiment, for a low-altitude inspection operation of a power line, there are multiple inspection execution scenarios in this low-altitude inspection operation of the power line, and all of them have been fully authenticated, such as mountain inspection scenarios, urban suburban inspection scenarios, and cross-river area inspection scenarios.

[0020] Now there is a flight inspection mission about to be executed, and the server needs to obtain relevant description information. Assume that there are X = 5 drones dispatched to this flight inspection mission under the target inspection execution scenario (assuming that this target inspection execution scenario is a mountain inspection scenario) in this target low-altitude inspection operation. For each drone, there are Y = 3 knowledge field description vectors. These 3 knowledge field description vectors contain description vectors of each drone in multiple inspection execution scenarios (mountain inspection scenario, urban suburban inspection scenario, and cross-river area inspection scenario).

[0021] Take one of the drones as an example. In the mountain inspection scenario, one of its knowledge field description vectors may contain information about its flight altitude adaptability. For example, in previous mountain inspections, this drone was able to fly stably at an average flight altitude of 100-300 meters, and the data related to this altitude range will be recorded in this knowledge field description vector. In the suburban inspection scenario, another knowledge field description vector may record its ability to deal with interference from small buildings, such as the probability of accurately avoiding areas with many low buildings when flying over them. In the inspection scenario across river areas, the third knowledge field description vector may contain data on its imaging capabilities when facing interference from water surface reflections.

[0022] At the same time, according to the current cluster division strategy, these 5 drones are divided into Z = 3 drone clusters. The server needs to obtain cluster description information, which includes the description vectors of each drone cluster in the three drone clusters in multiple inspection execution scenarios. For example, for one of the drone clusters, in the mountain inspection scenario, the description vector of this drone cluster may contain the overall flight speed stability data of the drones in the drone cluster. If the average flight speed of the drones in this cluster fluctuates little in previous mountain inspections, then this data will be recorded. In the urban suburban inspection scenario, it may include information such as the collaborative ability of drones in the drone cluster to deal with signal interference. In the inspection scenario across the river area, it may include an overall evaluation of the adaptability of drones in the drone cluster to humid environments.

[0023] Finally, the server also needs to obtain the flight inspection task description information, which is mainly the description vector of the target inspection execution scenario (mountain inspection scenario). This description vector may contain the geographical features of the mission area, such as the altitude range of the mountain area, the direction of the mountain range, and other information. It may also contain mission target-related features, such as the location of the power tower that needs to be inspected, the direction of the power line, and other information. It may also contain mission time limit-related features, such as this inspection mission needs to be completed between 8 am and 5 pm.

[0024] Step S120, loading the X drone description information sets, cluster description information and flight inspection task description information into the target adaptive inspection decision model learned in advance, generating a target estimation confidence generated by the target adaptive inspection decision model, wherein the target estimation confidence represents the confidence of the Z drone clusters in successfully completing the low-altitude flight inspection task in the flight inspection task under the condition that the Z drone clusters are used as the respective drone clusters scheduled to the flight inspection task.

[0025] In detail, the target adaptive inspection decision model is a pre-trained model that can generate a target estimation confidence based on the input drone description information set, cluster description information and flight inspection task description information to evaluate the success probability of the drone cluster when performing the inspection task. The target adaptive inspection decision model may be built based on deep learning, machine learning and other technologies. After inputting the description information of 5 drones, the description information of 3 drone clusters and the description information of the mountain inspection task, the target adaptive inspection decision model will output a target estimation confidence, such as 0.85, indicating that the drone cluster has an 85% probability of successfully completing the mountain inspection task.

[0026] For example, in this embodiment, after all the above information is obtained, the information is loaded into the target adaptive inspection decision model of pre-knowledge learning. First, the five drone description information sets are encoded and represented to generate the corresponding five drone description coding vectors. For example, the data in the three knowledge field description vectors of each drone are encoded according to certain rules. For the aforementioned information on flight altitude adaptability in mountain inspection scenarios, if the altitude range is represented by numbers, it may be encoded into a specific coding segment. At the same time, the cluster description information is encoded and represented to generate the corresponding cluster description coding vector. For example, the various capability data of the cluster in different inspection scenarios are encoded into a cluster description coding vector. And the flight inspection task description information is encoded and represented to generate the corresponding flight inspection task description coding vector.

[0027] Then, the description coding vectors of these five drones are analyzed for correlation with the description coding vectors of the flight inspection task. Take one of the drones as an example, assuming it is the first drone, and interact its description coding vector with the description coding vector of the flight inspection task. First, the description coding vector of the first drone is decomposed into multiple different first feature groups, such as the flight capability-related feature group (including the flight altitude adaptability mentioned above), the device status-related feature group (such as the coding corresponding to the information such as the drone battery power and device temperature), and the task history-related feature group (the performance data coding in previous similar inspection tasks). At the same time, the flight inspection task description coding vector is decomposed into multiple different second feature groups related to task scenario requirements, such as the task area geographical feature group (geographic data coding in mountainous areas), the task target-related feature group (coding related to power towers and lines), and the task time limit-related feature group (coding for the time period from 8 am to 5 pm).

[0028] For each first feature group, based on semantic understanding and logical relationships, preliminary association analysis is performed one by one with all second feature groups. For example, the flight altitude adaptability in the flight capability-related feature group is preliminarily associated with the mountain altitude range in the mission area geographic feature group. If the flight altitude adaptability of this drone can work well within the mountain altitude range, then the degree of association will be judged to be high, and an association relationship description document will be established to record the feature group pair consisting of the associated first feature group and the second feature group, as well as the basis of the association (such as the adaptability principle of the flight altitude within the mountain altitude range) and the preliminary judgment of the degree of association. For feature groups with a high degree of association, deep interaction operations are performed to generate coded segments after deep interaction. For example, after deep interaction between the flight altitude and the mountain altitude, a new coded segment may be obtained to represent the high adaptation advantage of this drone in the mountain inspection scene. For feature groups with a medium degree of association, partial interaction is performed to generate coded segments after partial interaction. For feature groups with a low degree of association, identification records are performed to generate identification record data. Finally, the coded segments after deep interaction and partial interaction are reassembled according to the set logical order (for example, considering geographical factors first and then target factors in the task execution process), and the identification record data is attached to the reassembled coded segments in a predefined coding form to generate the first drone-related coding vector. The other four drone-related coding vectors are generated in the same way.

[0029] Next, the target estimation confidence is determined based on the five drone association coding vectors, cluster description coding vectors, and flight inspection task description coding vectors. First, the five drone association coding vectors, cluster description coding vectors, and flight inspection task description coding vectors are aggregated to generate an aggregated coding vector. Then, the aggregated coding vector is fully connected and mapped based on the fully connected mapping layer. In this process, the model calculates a confidence that the three drone clusters will successfully complete the low-altitude flight inspection task in this flight inspection task based on the previously learned knowledge and algorithms, that is, the target estimation confidence.

[0030] Step S130, under the condition that the target estimation confidence meets the set requirements, the Z drone clusters are dispatched as the drone clusters in the one flight inspection task to obtain the adaptive detection results before the low-altitude flight inspection.

[0031] In detail, the set requirement refers to a standard or threshold used to determine whether the target estimation confidence meets the requirements. Only when the target estimation confidence meets the set requirement will the corresponding drone cluster be scheduled to the inspection task. For example, suppose the set requirement is that the target estimation confidence is not less than 0.8. If the target estimation confidence output by the model is 0.85, then it meets the set requirement and the corresponding drone cluster can be scheduled to the inspection task.

[0032] The adaptive detection result refers to the evaluation result of the matching degree between the cluster and the task by the target adaptive inspection decision model before the drone cluster is dispatched to the inspection task. If the evaluation result meets the set requirements, the adaptive detection result is considered good and the inspection task can be performed. For example, assuming that the target estimation confidence of the model output is 0.85, which meets the set requirements (not less than 0.8), then the adaptive detection result can be considered good, and the corresponding drone cluster can be dispatched to the mountain inspection task to perform the inspection operation.

[0033] For another example, in this embodiment, it is assumed that the set requirement is that the deviation between the target estimation confidence and the threshold confidence is less than 0.1. If the target estimation confidence calculated by the server meets this requirement, for example, the target estimation confidence is 0.85, the threshold confidence is 0.8, and their deviation is 0.05, which is less than 0.1, then the server will dispatch these three drone clusters as the various drone clusters in this flight inspection mission. This completes the acquisition of the adaptive detection results before the low-altitude flight inspection. This adaptive detection result indicates that the current cluster division and task allocation are more reasonable, and these three drone clusters have a higher probability of successfully completing the low-altitude flight inspection mission in the mountainous area. If the set requirements are not met, it may be necessary to readjust the cluster division strategy or further optimize the relevant parameters of the drone to ensure that the mission requirements can be met and the confidence of mission success can be improved.

[0034] In another case, it is assumed that the server obtains three estimated confidences, which are estimated confidences determined by the target adaptive inspection decision model based on the drone description information, cluster description information and flight inspection task description information of each drone cluster division data in three different drone cluster division data. These three different cluster division strategies include the current cluster division strategy, and the number of clusters of the multiple drone clusters included in each drone cluster division data is the same, which is 3. If the deviation between the current target estimated confidence and the threshold confidence is the smallest among the three estimated confidences, then the server will also dispatch these three drone clusters as the drone clusters in this flight inspection task.

[0035] There is another situation, that is, among these three estimated confidences, the deviation between the target estimated confidence and the threshold confidence is the smallest and this deviation is less than 0.1. The server will also dispatch these three drone clusters as the various drone clusters in this flight inspection task to obtain the adaptive detection results before the low-altitude flight inspection. This adaptive detection result is of great significance for the efficient and accurate execution of the entire low-altitude inspection operation. It can evaluate and optimize the feasibility and success probability of the task before the flight inspection, thereby improving the quality and efficiency of the entire inspection operation.

[0036] Based on the above steps, the embodiment of the present application integrates and analyzes the multi-dimensional description information of individual drones, drone clusters, and flight inspection tasks, so as to achieve an accurate assessment of the ability of drone clusters to perform inspection tasks, and uses the target adaptive inspection decision model learned by prior knowledge to generate the target estimation confidence to quantify the possibility of the drone cluster successfully completing the low-altitude flight inspection in a specific inspection task. Under the condition that the target estimation confidence meets the set requirements, the appropriate drone cluster can be intelligently dispatched to the inspection task, thereby ensuring the efficiency and safety of the inspection operation. That is, by intelligently evaluating the matching degree between the drone cluster and the inspection task, it can ensure that the dispatched drone cluster has the ability to complete the task, avoiding task delays or failures caused by insufficient drone performance or improper cluster configuration, thereby significantly improving the efficiency of the inspection operation. Before dispatching the drone cluster, the target adaptive inspection decision model is used to comprehensively evaluate the cluster's capabilities, effectively avoiding the safety risks that may be caused by drone or cluster performance defects, and enhancing the safety and reliability of the inspection task. In addition, according to the specific needs of the inspection task, the most suitable drone cluster can be intelligently selected for dispatching, realizing the optimal configuration of drone resources and improving resource utilization efficiency. By introducing a target-adaptive inspection decision model with prior knowledge learning, intelligent evaluation and decision-making of the inspection capabilities of drone clusters are achieved, the intelligence level of inspection task scheduling is improved, and the possibility of human intervention and decision-making errors is reduced.

[0037] In a possible implementation, step S120 includes:

[0038] Step S121, encode and represent the X drone description information sets to generate corresponding X drone description encoding vectors, encode and represent the cluster description information to generate corresponding cluster description encoding vectors, and encode and represent the flight inspection task description information to generate corresponding flight inspection task description encoding vectors, wherein the X drone description encoding vectors correspond to the X drones respectively.

[0039] Step S122: performing correlation analysis on the X drone description coding vectors and the flight inspection task description coding vectors to generate X drone associated coding vectors, wherein the X drone associated coding vectors correspond to the X drones respectively.

[0040] Step S123: determining the target estimation confidence based on the X drone association coding vectors, the cluster description coding vector and the flight inspection task description coding vector.

[0041] Based on the previous example of low-altitude inspection of power lines, after obtaining X drone description information sets, cluster description information and flight inspection task description information, this information is loaded into the pre-learned target adaptive inspection decision model to generate target estimation confidence. First, the server must encode the X drone description information sets to generate the corresponding X drone description encoding vectors, encode the cluster description information to obtain the cluster description encoding vector, and encode the flight inspection task description information to generate the flight inspection task description encoding vector. For example, in the scenario mentioned above where 5 drones (X=5) participate in the power line inspection task, for each drone description information set, it contains knowledge field description vectors in inspection scenarios such as mountainous areas, urban suburbs, and cross-river areas. The server converts this information into drone description encoding vectors according to predetermined encoding rules. In terms of cluster description information, the description vectors of the capabilities of the three (Z=3) drone clusters in each inspection scenario are encoded to form a cluster description coding vector. The geographical features, mission objectives, time limits and other information related to the mountain inspection scenario in the flight inspection task description information are also encoded into a flight inspection task description coding vector, and these five drone description coding vectors correspond one-to-one to the five drones.

[0042] Next, the X (5) drone description coding vectors are analyzed for correlation with the flight inspection task description coding vectors to generate X (5) drone correlation coding vectors. Taking one of the drones as an example, as mentioned earlier, the drone with good flight altitude adaptability in the mountain inspection scenario, its description coding vector is decomposed into different first feature groups such as flight capability related feature group (such as flight altitude adaptability), equipment status related feature group (battery power, equipment temperature, etc.) and task history related feature group (previous similar task performance), and the flight inspection task description coding vector is decomposed into task area geographical feature group (mountain altitude, mountain direction, etc.), task target related feature group (power tower and line location direction) and task time limit related feature group (8 am-5 pm) and other second feature groups. Then, based on semantic understanding and logical relationship, the first feature group and the second feature group are subjected to preliminary correlation analysis one by one. For example, the flight altitude adaptability in the flight capability related feature group is analyzed with the mountain altitude range in the mission area geographic feature group. If the flight altitude of this drone is adapted to the mountain altitude range and the correlation is high, a correlation relationship description document is established to record the feature group pair, correlation basis and correlation degree. For the high correlation degree, a deep interaction operation is performed, such as the flight altitude and mountain altitude are deeply interacted to obtain the coding segment representing the high adaptation advantage. The correlation degree is partially interacted to obtain the corresponding coding segment. The low correlation degree is identified and recorded to obtain the identification record data. Finally, the coding segments after deep interaction and partial interaction are recombined according to the order of the task execution process and the identification record data is attached to form the drone association coding vector of this drone. The other four drones are operated in the same way to obtain five drone association coding vectors, which correspond to the five drones respectively.

[0043] Finally, the server determines the target estimation confidence based on the X (5) drone association coding vectors, cluster description coding vectors and flight inspection task description coding vectors. The server aggregates the 5 drone association coding vectors, cluster description coding vectors and flight inspection task description coding vectors to generate an aggregated coding vector. Then, the fully connected mapping layer in the target adaptive inspection decision model is used to perform fully connected mapping processing on the aggregated coding vector. In this process, the model comprehensively analyzes and calculates various information in the aggregated coding vector based on the previously learned knowledge and algorithms. For example, the drone flight capability, equipment status and other information in the drone association coding vector, the overall performance information of the cluster in the cluster description coding vector, and the task requirement information in the flight inspection task description coding vector are all comprehensively considered. Through the complex calculation process inside the model, a value is finally obtained. This value is the target estimation confidence, which indicates the confidence of the three (Z=3) drone clusters in successfully completing the low-altitude flight inspection task in this flight inspection task under the condition that the three (Z=3) drone clusters are dispatched to each drone cluster in this flight inspection task.

[0044] In a possible implementation, step S122 includes:

[0045] Based on the following steps, the ath drone description coding vector among the X drone description coding vectors is analyzed for correlation with the flight inspection task description coding vector to generate the ath drone association coding vector among the X drone association coding vectors, where a is a positive integer not less than 1 and not greater than X:

[0046] Step S1221: interact the a-th UAV description coding vector with the flight inspection task description coding vector to generate the a-th blending coding vector.

[0047] Step S1222: The aggregation result of the a-th UAV description coding vector and the a-th blending coding vector is used as the a-th UAV associated coding vector. Alternatively, the a-th blending coding vector is fused with the a-th influencing factor defined previously to generate the a-th fused coding vector, and the aggregation result of the a-th UAV description coding vector and the a-th fused coding vector is used as the a-th UAV associated coding vector.

[0048] Wherein, step S1221 includes:

[0049] Step S1221-1, decompose the a-th drone description coding vector into multiple different first feature groups, where the first feature groups are divided based on different attributes of the drone, specifically including a flight capability-related feature group, a device status-related feature group, and a mission history-related feature group.

[0050] And, step S1221-2, decomposing the flight inspection task description coding vector into a plurality of second feature groups related to different task scenario requirements, specifically including a task area geographical feature group, a task target related feature group and a task time limit related feature group.

[0051] Step S1221-3, for each first feature group, based on semantic understanding and logical relationships, preliminary association analysis is performed one by one with all second feature groups, and an association relationship description document is established for each preliminary association analysis result, wherein the association relationship description document is used to record the feature group pairs formed by the associated first feature group and the second feature group, as well as the basis for the association and the preliminary judgment of the degree of association.

[0052] Step S1221-4, according to the association relationship description document, deep interaction operations are performed on feature pairs with a high degree of association to generate coded segments after deep interaction, partial interaction is performed on feature pairs with a medium degree of association to generate coded segments after partial interaction, and identification records are performed on feature pairs with a low degree of association to generate identification record data.

[0053] Step S1221-5, recombining the coded segments after deep interaction and partial interaction according to a set logical order, wherein the logical order is determined based on the task execution process or the importance of the data, and appending the identification record data to the recombined coded segments in a predefined coding form to generate the ath blended coding vector.

[0054] In this embodiment, this part of the content is described in detail based on the previous example of low-altitude inspection of power lines.

[0055] In the previous steps, X drone description coding vectors and flight inspection task description coding vectors have been obtained. Now we need to perform correlation analysis on them to generate X drone association coding vectors. Take the ath drone among the X drones (here we assume a=3, in the scenario where 5 drones participated in the inspection task) as an example to illustrate this process.

[0056] First, the third drone description coding vector is interacted with the flight inspection task description coding vector to generate the third fusion coding vector. First, the third drone description coding vector is decomposed into multiple different first feature groups, including flight capability-related feature groups, equipment status-related feature groups, and task history-related feature groups. For the flight capability-related feature group, in this power line inspection scenario, it may include coding information such as the maximum flight speed, minimum turning radius, and flight stability of the drone during mountain inspections; the equipment status-related feature group may include coding information such as the current battery power percentage of the drone, the operating temperature of each key device (such as motors, sensors, etc.), and the remaining service life of the device; the task history-related feature group may record the number of successes and failures of the drone in previous similar inspection tasks, and the response performance when encountering special situations (such as strong wind interference, short signal interruption, etc.).

[0057] At the same time, the flight inspection task description coding vector is decomposed into a plurality of second feature groups related to different task scenario requirements, specifically, the task area geographical feature group, the task target related feature group and the task time limit related feature group. In the mountain power line inspection scenario, the task area geographical feature group includes coding information such as the average altitude of the mountains, the undulation of the mountains, the distribution of valleys and peaks, the distribution of rivers and roads and their potential impact on the flight; the task target related feature group includes coding information such as the specific coordinates of the power towers to be inspected, the direction of the power lines (whether along the valley or ridge, whether crossing the river, etc.), the height and type of the power towers; the task time limit related feature group includes the start time (such as 8 am), end time (such as 5 pm) specified for this inspection task, and special requirements for inspection speed or accuracy in different time periods (for example, when the strong sunlight at noon may affect the imaging quality, the flight speed needs to be reduced to improve the inspection accuracy) and other coding information.

[0058] Next, for each first feature group, based on semantic understanding and logical relationships, preliminary association analysis is performed one by one with all second feature groups, and an association description document is established for each preliminary association analysis result. For example, for the maximum flight speed in the flight capability-related feature group, a preliminary association analysis is performed with the degree of mountain undulation in the mission area geographical feature group. If the mountain undulation is large, the drone needs to have a high flight speed to efficiently complete the inspection task, then it will be judged that their degree of association is high, and the feature group pair consisting of the maximum flight speed in the flight capability-related feature group and the degree of mountain undulation in the mission area geographical feature group is recorded in the association description document, and the basis for the association (i.e., high speed is required for large mountain undulations to ensure inspection efficiency) and the preliminary judgment of the degree of association (high) are recorded at the same time. A preliminary association analysis is performed on the battery power percentage in the device status-related feature group and the task duration in the task time limit-related feature group. If the task duration is long and the battery power percentage is low, the degree of association may be judged to be medium and recorded in the association description document. For example, the performance in dealing with special situations in the task history-related feature group is preliminarily associated with the power tower type in the task goal-related feature group. If a certain type of power tower has never encountered special situations in previous inspections, then the degree of association may be low, which is also recorded in the association description document.

[0059] According to the association relationship description document, for the feature pairs with high association, deep interaction operations are performed to generate deep interaction code segments. For example, for the previously mentioned high association feature pair of maximum flight speed and mountain undulation, the deep interaction operation may be to accurately calculate the optimal flight speed range that meets efficient inspection based on the mountain undulation, and merge the code related to this speed range with the original maximum flight speed code to generate deep interaction code segments. For the feature pairs with medium association, partial interaction is performed to generate partial interaction code segments. For example, for the feature pairs with medium association, such as battery power percentage and task duration, partial interaction may be to estimate the battery power consumption based on the task duration, and partially merge the estimated result code with the original battery power percentage code to obtain partial interaction code segments. For the feature pairs with low association, identification records are performed to generate identification record data. For example, for the low association feature pair of performance in dealing with special situations and power tower type, the identification record data may simply record this low association situation and the corresponding feature group content.

[0060] Then, the coded segments after deep interaction and partial interaction are reassembled in a set logical order, which is determined based on the task execution process or the importance of the data. In this power line inspection scenario, according to the task execution process, the coded segments related to the geographical features of the task area should be considered first, because it will directly affect the flight path planning of the drone, so the coded segments after deep interaction or partial interaction with geographical features such as the undulation of the mountain range are placed first. Next is the coded segment related to the task goal, because the inspection goal is the core of the entire task, and then the coded segments related to the task goals such as the type and location of the power tower are placed. Finally, the coded segments related to the task time limit are placed, because it is more of a time constraint on the entire inspection process. After the coded segments are reassembled in this order, the previously generated identification record data is attached to the reassembled coded segments in a predefined coding form, so that the third fusion coding vector is generated.

[0061] After the third interleaved coding vector is generated, there are two ways to obtain the third drone-associated coding vector. One way is to directly aggregate the third drone description coding vector with the third interleaved coding vector, merge or calculate the corresponding codes in the two vectors according to certain rules (such as simple addition or weighted addition), and the result is the third drone-associated coding vector. Another way is to fuse the third interleaved coding vector with the third previously defined influencing factor, which is a numerical value or coding form predetermined based on previous experience or a specific algorithm. For example, this influencing factor may be a quantitative value of the impact of external factors such as the current season and weather conditions on drone inspection. The third interleaved coding vector is fused and calculated with this influencing factor, such as multiplication or calculation according to a certain functional relationship, to generate the third fused coding vector, and then the third drone description coding vector is aggregated with the third fused coding vector (also according to the aforementioned merging or calculation rules), and finally the third drone-associated coding vector is obtained. In the same way, for other drones (from the 1st to the Xth), the corresponding drone association coding vectors can be generated respectively, thus completing the entire correlation analysis and generating the X drone association coding vectors. This process fully considers the various characteristics of the drone itself and the various requirements of the flight inspection mission. Through the above correlation analysis and interactive operations, it lays the foundation for the subsequent accurate determination of the target estimation confidence.

[0062] In a possible implementation, step S123 includes:

[0063] Step S1231: Aggregate the X drone association coding vectors, the cluster description coding vector, and the flight inspection task description coding vector to generate an aggregated coding vector.

[0064] Step S1232: Perform a fully connected mapping process on the aggregated coding vector based on a fully connected mapping layer to generate the target estimation confidence.

[0065] In this embodiment, in the previous operation, X drone association coding vectors, cluster description coding vectors and flight inspection task description coding vectors have been obtained respectively, and now the target estimation confidence is determined based on these coding vectors.

[0066] First, the server aggregates X drone-associated coding vectors, cluster description coding vectors, and flight inspection task description coding vectors to generate an aggregated coding vector. In the previous example, assuming that there are 5 (X=5) drones participating in the inspection task, then these 5 drone-associated coding vectors contain information about each drone and the flight inspection task after multiple aspects of the associated analysis. For example, each drone-associated coding vector contains coding information about the interaction between the drone's flight capability and the geographical characteristics of the mission area, the mission objectives, and the mission time limit, the coding information about the interaction between the device status and these mission requirements, and the coding information about the interaction between the mission history and the mission requirements. The cluster description coding vector contains coding information about the description vectors of the overall performance of the 3 (Z=3) drone clusters in multiple inspection scenarios, such as the coordination ability of the drones in the cluster to cope with different geographical environments, and the overall stability in different mission scenarios. The flight inspection task description coding vector contains the coding of information such as the regional geographical characteristics (mountain direction, altitude, etc.), mission objectives (power tower and line location, etc.), and mission time limit (8 am to 5 pm) of this mountain power line inspection task.

[0067] When these code vectors are collected, they can be combined into a new collection code vector according to certain rules. This collection code vector is like a large collection that integrates all relevant information, and each part of the code represents a factor related to the success of the inspection task. For example, the UAV flight stability-related code in the UAV association code vector, the cluster overall flight stability-related code in the cluster description code vector, and the task area geographical features (such as mountain undulations affecting flight stability)-related code in the flight inspection task description code vector are interrelated in the collection code vector, and together reflect the comprehensive situation of the entire inspection task in terms of flight stability.

[0068] Next, the server performs a fully connected mapping process on the pooled coding vector based on the fully connected mapping layer to generate the target estimation confidence. The fully connected mapping layer is an important part of the target adaptive inspection decision model. It has pre-set connection weights and algorithms. When performing the fully connected mapping process, each element in the pooled coding vector is connected to the neurons in the fully connected mapping layer and calculated according to the set weights.

[0069] Taking the UAV flight capability-related codes in the pooled coding vector as an example, in the fully connected mapping layer, the neurons related to flight capability will process this part of the code according to the pre-learned weights. If it is found in the previous learning process that the high flight speed of the UAV is an important success factor in the mountain inspection mission, then during the fully connected mapping process, the coding elements related to the flight speed will be given a higher weight for calculation. Similarly, for the cluster coordination capability-related codes in the cluster description coding vector, the corresponding neurons in the fully connected mapping layer will determine the weight and calculate it according to the importance of cluster coordination capability to the success of the inspection mission. For the task time limit-related codes in the flight inspection task description coding vector, the fully connected mapping layer will also determine the weight and calculate it according to the degree of influence of the time limit on the task.

[0070] In this complex calculation process, the fully connected mapping layer will comprehensively consider all the information in the aggregated encoding vector, not only the individual flight capabilities, cluster coordination capabilities or mission time limits, but also the relationships and interactions between them. For example, the relationship between the drone's flight capability and device status, the relationship between cluster coordination capabilities and mission objectives, etc. will all be included in the calculation scope.

[0071] After a series of calculations in the fully connected mapping layer, a value is finally obtained, which is the target estimation confidence. This target estimation confidence represents the confidence of the three previously divided drone clusters in successfully completing the low-altitude flight inspection task in this flight inspection task under the condition that the three drone clusters are dispatched to the various drone clusters in this flight inspection task. For example, if the target estimation confidence is 0.8, it means that based on the existing drone individual information, cluster information, and flight inspection task information, the server believes that there is an 80% probability that these three drone clusters will successfully complete the low-altitude flight inspection task of the mountain power lines. This target estimation confidence provides an important basis for subsequent decision-making, such as whether to perform the inspection task according to the current cluster division, or whether certain parameters need to be adjusted to improve the confidence.

[0072] In a possible implementation, step S110 includes:

[0073] Step S111, obtaining flight trajectory description information, immediate status description information and long-term status description information of each drone among the X drones, and generating a description information set of the X drones, wherein the flight trajectory description information of each drone includes the flight trajectory information of each drone in the previous b flight inspection tasks, the long-term status description information of each drone includes the record information of each drone in each of the multiple inspection execution scenarios within a previous set time period, and the immediate status description information of each drone includes the current record information of each drone in each of the multiple inspection execution scenarios.

[0074] Step S112: Obtain average performance statistics of each drone cluster in the Z drone clusters in each inspection execution scenario in the multiple inspection execution scenarios.

[0075] Step S113, obtaining the scene label index of the target inspection execution scene, or obtaining the scene label index of the target inspection execution scene and the timing node scheduled to the one-time flight inspection task.

[0076] In this embodiment, in the scenario of power line inspection, the server needs to obtain the flight trajectory description information, immediate state description information and long-term state description information of each of the X drones to generate X drone description information sets.

[0077] For the flight trajectory description information of each drone, suppose X = 5, and these 5 drones have participated in b = 3 flight inspection missions before. Take one of the drones as an example. In these 3 flight inspection missions, its flight trajectory information contains a lot of content. In the first inspection mission, the drone took off from the base and flew to the power line inspection area in the mountainous area according to the set route. Its flight trajectory may be to fly along a valley first and then cross a small hill. During this process, the server records its flight altitude, speed, turning point and other information. For example, when it flies in the valley, in order to avoid some tall trees, the flight altitude is maintained at about 80 meters and the speed is 30 meters / second. When crossing the hill, the flight altitude is increased to 120 meters and the speed is slightly reduced to 25 meters / second. The second inspection mission may be to inspect power lines in the suburbs of the city. The flight trajectory is different. It needs to avoid high-rise buildings and densely populated areas. The flight altitude may be constantly adjusted between 100-150 meters, and the speed also fluctuates between 20-30 meters per second according to the complexity of the surrounding environment. The server records the change points and change time of these trajectories in detail. If the third inspection mission is to inspect across a river area, when the drone approaches the river, it may be affected by the reflection of the water surface and the airflow, and the flight trajectory will be adjusted accordingly, such as maintaining a high safety distance from the river surface. The flight altitude may be maintained at about 130 meters, and the speed will also be adjusted according to the wind direction and wind force. All this flight trajectory information is fully recorded.

[0078] The instant status description information of each drone includes the current record information of each drone in multiple inspection execution scenarios. Taking this drone as an example, in the current mountain inspection scenario, its instant status description information includes the current battery power, assuming it is 80%, the operating temperature of each device, such as the motor temperature is 40 degrees Celsius, the sensor temperature is 35 degrees Celsius, and also includes the current flight altitude of 90 meters, the flight speed is 28 meters / second, and the current communication signal strength with the control center is good. In the suburban inspection scenario, the instant status may be 70% battery power (because signal interference in the city may increase power consumption), slightly increased device temperature, flight altitude of 120 meters, flight speed of 25 meters / second, communication signal is interfered by some buildings but still stable, etc. In the cross-river area inspection scenario, the instant status may be 75% battery power, normal device temperature, flight altitude of 135 meters, flight speed of 27 meters / second, good communication signal, etc.

[0079] The long-term status description information of each drone includes the record information of each inspection execution scenario of the drone in multiple inspection execution scenarios during the previous set period. Assuming that the set period is the past month, during this month, the long-term status record of this drone in the mountain inspection scenario may show that it participated in a total of 5 inspection tasks, one of which was caused by strong wind weather, causing the flight trajectory to deviate from the scheduled route, but it was corrected in time through the automatic adjustment system. In the urban suburban inspection scenario, it had 3 inspection tasks in this month, and once there was a brief communication interruption due to the signal being blocked by a building, but it was quickly restored. In the cross-river area inspection scenario, there were 2 inspection tasks, and the overall performance was good, but in one inspection, it was found that the imaging quality of the imaging device was slightly reduced when the reflection on the water surface was strong. These long-term status information were collected by the server. By obtaining these three aspects of information (flight trajectory description information, immediate status description information, and long-term status description information), the server generated corresponding drone description information sets for these 5 drones.

[0080] The next step is to obtain cluster description information. In this power line inspection operation, according to the current cluster division strategy, 5 drones are divided into Z = 3 drone clusters. The server needs to obtain the average performance statistics of each drone cluster in these 3 drone clusters in each inspection execution scenario in multiple inspection execution scenarios. Taking one of the drone clusters as an example, in the mountain inspection scenario, its average performance statistics contain a lot of content. The average flight speed of the drones in this cluster is 28 meters per second, the average flight altitude is 95 meters, the average battery power consumption is 60% in a complete mountain inspection mission, and the average communication signal strength between drones in the cluster is at a good level. When dealing with special weather in mountainous areas (such as strong winds), the overall stability of the drones in the cluster is good, and the average deviation from the scheduled route is within an acceptable range, such as no more than 5 meters. In the urban suburban inspection scenario, the average performance statistics of this cluster show that the average flight speed is 25 meters per second, the average flight altitude is 110 meters, and the average battery consumption is 65% (due to more signal interference). The communication coordination ability between the drones in the cluster is manifested in that when encountering buildings blocking the signal, the communication adjustment mechanism within the cluster can maintain more than 90% of the information transmission success rate. In the cross-river area inspection scenario, the average flight speed of this cluster is 27 meters per second, the average flight altitude is 130 meters, and the average battery consumption is 62%. In terms of dealing with the reflection of the water surface affecting the imaging quality, the overall response ability of the drones in the cluster is to adjust the flight angle and imaging parameters to make more than 80% of the power lines and power towers clear. The server obtains such average performance statistics for each cluster in different inspection scenarios to obtain cluster description information.

[0081] Finally, the description information of the flight inspection task is obtained. In this example, the server needs to obtain the scene label index of the target inspection execution scene, or obtain the scene label index of the target inspection execution scene and the timing node that is scheduled to a flight inspection task. Assuming that the target inspection execution scene this time is a mountain inspection scene, its scene label index may be a specific code, such as "001" for a mountain inspection scene. This scene label index contains some basic feature information of the mountain inspection scene, such as geographical environment characteristics (mountain direction, altitude range, etc.), possible weather types (windy, foggy, etc.), distribution density of power lines and power towers, and other information. If you also need to obtain the timing node that is scheduled to a flight inspection task, assuming that this timing node is to start executing this flight inspection task at 9 am, this timing node information is of great significance to the planning and execution of the entire inspection task. For example, weather conditions at 9 a.m. (such as light intensity, wind direction and wind speed) will affect the flight safety and inspection results of drones. This time is also related to factors such as the peak period of power line use. It may be necessary to focus on the status of certain key power equipment during this period. The server obtains these flight inspection task description information, which provides an important basis for subsequent decision-making and analysis.

[0082] In a possible implementation, step S130 includes:

[0083] Step S131, under the condition that the deviation between the target estimation confidence and the threshold confidence is less than a set value, the Z drone clusters are used as the drone clusters scheduled to the one flight inspection task.

[0084] Step S132: Under the condition that the deviation between the target estimated confidence and the threshold confidence in the acquired A estimated confidences is the smallest, the Z drone clusters are used as the drone clusters scheduled to the flight inspection task. The A estimated confidences include the estimated confidences determined by the target adaptive inspection decision model based on the drone description information, cluster description information and flight inspection task description information of each drone cluster division data in the A drone cluster division data. Each drone cluster division data in the A drone cluster division data includes a plurality of drone clusters generated by clustering the X drones according to a corresponding cluster division strategy among A different cluster division strategies. The A different cluster division strategies include the current cluster division strategy. The number of clusters of the plurality of drone clusters included in each drone cluster division data is the same, and A is not less than 2. Or

[0085] Step S133, under the condition that the deviation between the target estimated confidence and the threshold confidence in the A estimated confidences is the smallest and the deviation is less than the set value, the Z drone clusters are dispatched as the respective drone clusters in the one flight inspection mission.

[0086] In this embodiment, in the previous steps, the target estimation confidence has been obtained through the target adaptive inspection decision model. This target estimation confidence represents the possibility that the Z drone clusters (assuming Z=3) currently set as the drone clusters dispatched to this flight inspection task can successfully complete this low-altitude flight inspection task (such as the mountain power line inspection task mentioned above). Now, it is necessary to decide whether to dispatch these Z drone clusters to this flight inspection task based on whether the target estimation confidence meets the set requirements.

[0087] First, consider the case where Z drone clusters are scheduled as individual drone clusters in a flight inspection mission under the condition that the deviation between the target estimate confidence and the threshold confidence is less than the set value. Assume that the threshold confidence is set to 0.7 and the set value is 0.1. If the target estimate confidence calculated by the server is 0.75, then the deviation between the target estimate confidence and the threshold confidence is 0.05, which is less than the set value of 0.1. This means that the current three drone clusters are very close to the ideal confidence (threshold confidence) for successful completion of the mission under the evaluation of the model.

[0088] From the perspective of the actual inspection mission, the comprehensive capabilities of the three drone clusters in the previous information analysis (including the flight trajectory, immediate status, long-term status of each drone, and the average performance statistics of the cluster, etc.) are matched and evaluated with the requirements of this flight inspection mission (such as the geographical characteristics of the mission area, mission objectives, mission time limits, etc.), and the result of this close to threshold confidence indicates that they have sufficient capabilities to perform the mission. For example, the flight trajectory adaptability, equipment status stability, and ability to cope with various emergencies (such as weather changes, geographical environment interference, etc.) of the drones in the three clusters in the previous mountain inspection mission, combined with the specific requirements of this mission (such as specific inspection routes, inspection time, etc.), are considered by the model to have a high probability of success and a small deviation after comprehensive evaluation, so the server can dispatch these three drone clusters as the various drone clusters in this flight inspection mission.

[0089] Next, let's look at the situation where Z drone clusters are used as each drone cluster scheduled to a flight inspection task under the condition that the deviation between the target estimated confidence and the threshold confidence in the A estimated confidences obtained is the smallest. Assume that A = 3, that is, the server determines three estimated confidences based on the drone description information, cluster description information, and flight inspection task description information of each drone cluster partition data in the three different drone cluster partition data through the target adaptive inspection decision model. These three different cluster partition strategies all cluster X drones (assuming X = 5) to generate multiple drone clusters, and the number of clusters of the multiple drone clusters generated in each cluster partition data is the same (all 3).

[0090] Under the first clustering strategy, the estimated confidence obtained by model calculation is 0.65; under the second clustering strategy (that is, the current clustering strategy), the target estimated confidence is 0.75; and under the third clustering strategy, the estimated confidence is 0.6. It can be seen that among these three estimated confidences, the deviation between the target estimated confidence of 0.75 obtained under the current clustering strategy and the threshold confidence (assuming it is still 0.7) is the smallest.

[0091] From the actual situation, different cluster division strategies will lead to different cluster combinations. The drones in each cluster are matched differently, and their overall performance and ability to cope with tasks will also be different. Under the first cluster division strategy, the estimated confidence may be low due to factors such as poor coordination between some drones or unreasonable resource allocation within the cluster. Under the current cluster division strategy, the performance of the three drone clusters is more balanced and excellent in all aspects. For example, in a certain cluster, the flight capabilities, equipment status, and mission history of the drones complement each other, making this cluster more adaptable in dealing with the complex geographical environment of the mountainous area, the distribution of power lines, and the time limit of the mission. In the model evaluation, the target estimated confidence with the smallest deviation from the threshold confidence is obtained, so the server will choose these three drone clusters (under the current cluster division strategy) as the drone clusters scheduled to this flight inspection mission.

[0092] Finally, consider the case where Z drone clusters are dispatched to each drone cluster in a flight inspection task under the condition that the deviation between the target estimated confidence and the threshold confidence in A estimated confidences is the smallest and the deviation is less than the set value. Assume that A=3, the threshold confidence is 0.7, and the set value is 0.1. The estimated confidence obtained under the first clustering strategy is 0.6, which deviates from the threshold confidence by 0.1; the target estimated confidence obtained under the second clustering strategy (current clustering strategy) is 0.75, which deviates from the threshold confidence by 0.05; the estimated confidence obtained under the third clustering strategy is 0.68, which deviates from the threshold confidence by 0.02. It can be seen that the deviation between the target estimated confidence of 0.75 and the threshold confidence under the current clustering strategy is the smallest, and this deviation of 0.05 is less than the set value of 0.1.

[0093] From the perspective of the actual inspection task execution, this shows that the current three drone clusters not only have the smallest deviation from the threshold confidence, but also this smallest deviation is within the acceptable setting range. This means that this cluster division scheme is the most likely to successfully complete this flight inspection task after considering multiple factors such as the individual capabilities of drones, the overall performance of the cluster, and the task requirements. For example, among the current three drone clusters, the experience and capabilities accumulated by each cluster in previous different inspection scenarios (such as coping with strong winds in mountainous areas, signal interference in urban suburbs, and water surface reflections in cross-river areas, etc.), combined with the specific requirements of this mountain power line inspection task (such as specific power tower locations, inspection time limits, etc.), are combined in an optimal way under the evaluation of the model, which not only meets the requirements of the degree of proximity to the threshold confidence, but also is within the acceptable deviation range. Therefore, the server will dispatch these three drone clusters as the various drone clusters in this flight inspection task, thereby providing a reliable cluster scheduling scheme for low-altitude flight inspection tasks, which helps to improve the success rate and efficiency of inspection tasks.

[0094] In a possible implementation, the method further includes:

[0095] Step A110, obtaining a sample learning data sequence and a sample confidence sequence, wherein the sample learning data sequence includes B sample learning data, each sample learning data includes X sample drone cluster description information, sample cluster description information and sample flight inspection task description information, the X sample drone cluster description information includes Y sample knowledge field description vectors of each sample drone in the X sample drones, the X sample drones are drones dispatched to a sample flight inspection task in the sample inspection execution scenario in the target low-altitude inspection operation, and the Y sample knowledge field description vectors of each sample drone The segment description vector includes the description vector of each example drone in the multiple inspection execution scenarios, the example cluster description information includes the description vector of each example drone cluster in the Z example drone clusters in the multiple inspection execution scenarios, the example flight inspection task description information includes the description vector of the example inspection execution scenario, the Z example drone clusters are multiple example drone clusters generated by clustering the X example drones according to the example cluster division strategy, and the sample confidence sequence includes B sample confidences corresponding to the B sample learning data respectively, and B is not less than 2.

[0096] Step A120, model parameter learning is performed on the initialized adaptive patrol decision model according to the sample learning data sequence until the cost value of the target training cost function corresponding to the adaptive patrol decision model meets the previously defined training termination requirements, the training is terminated, and the adaptive patrol decision model at the time of termination of the training is used as the target adaptive patrol decision model, and the neuron weight coefficients in the adaptive patrol decision model are updated under the condition that the cost value of the target training cost function corresponding to the adaptive patrol decision model does not meet the previously defined training termination requirements. The cost value of the target training cost function is a loss value determined and generated based on the example estimated confidence generated by the adaptive patrol decision model and the corresponding sample confidence of the B sample confidences, and the example estimated confidence is the estimated confidence determined and generated by the adaptive patrol decision model based on the X example drone cluster description information, the example cluster description information and the example flight inspection task description information.

[0097] In this embodiment, under the general framework of low-altitude inspection of power lines, the server needs to train the model in order to obtain a target adaptive inspection decision model that can accurately make inspection decisions.

[0098] First, the server needs to obtain the sample learning data sequence and the sample confidence sequence. Assume that B = 3, that is, the sample learning data sequence contains 3 sample learning data.

[0099] For the first sample learning data, it contains description information of X sample drone clusters. Assume that X = 5, these 5 sample drones are the drones that are dispatched to a sample flight inspection mission in the sample inspection execution scenario in the target low-altitude inspection operation. Each sample drone has Y sample knowledge field description vectors, assuming that Y = 3. These 3 sample knowledge field description vectors cover the description vectors of each sample drone in multiple inspection execution scenarios (such as mountain inspection scenarios, urban suburban inspection scenarios, and cross-river area inspection scenarios).

[0100] Taking one of the example drones as an example, in a mountain inspection scenario, one of its sample knowledge field description vectors may contain data related to its flight altitude adaptability, such as its optimal flight altitude range when flying at different altitudes in the mountainous area, and its altitude adjustment strategy when flying near slopes of different slopes. In an urban suburban inspection scenario, another sample knowledge field description vector may record its flight path planning capabilities when dealing with densely built areas, such as how to bypass many buildings in the shortest path and ensure effective inspection of power lines. In a cross-river area inspection scenario, the third sample knowledge field description vector may contain relevant data such as its ability to resist airflow interference when flying above the water surface and the imaging device's ability to adapt to water surface reflections.

[0101] The example cluster description information includes the description vectors of each example drone cluster in the Z example drone clusters in multiple inspection execution scenarios. Assuming Z = 3, for one of the example drone clusters, in the mountain inspection scenario, its description vector may contain the overall flight speed coordination data of the drones in the cluster. For example, in the complex terrain of the mountainous area, how do the drones in the cluster adjust their respective speeds to maintain a relatively stable spacing and flight sequence, so as to efficiently complete the inspection task. In the suburban inspection scenario, it may include information such as the collaborative communication capabilities of the drones in the cluster when dealing with frequent switching of communication signals (due to building obstructions). In the cross-river area inspection scenario, it may include relevant data such as the overall response strategy of the drones in the cluster to the impact of humid environments on equipment performance.

[0102] The sample flight inspection task description information includes the description vector of the sample inspection execution scenario. If the sample inspection execution scenario is a mountain inspection scenario, this description vector may contain detailed geographical features of the mission area, such as the specific altitude difference range of the mountains, the distribution density of valleys and peaks, the intersection of rivers and power lines, etc. It also includes mission target-related features, such as the specific type of power tower that needs to be inspected (high-voltage tower or ordinary voltage tower), the special direction of the power line (whether it is along the ridge and close to dangerous areas such as cliffs), etc. It also includes mission time limit-related features, such as whether the sample inspection task needs to complete a specific part of the inspection work in a specific time period (such as the early morning when the light is dim but the power load is low).

[0103] At the same time, the sample confidence sequence includes three sample confidences corresponding to the three sample learning data. These sample confidences are obtained by other methods (such as manual evaluation or statistical results of some verified data) before, indicating the confidence that the sample drone cluster successfully completes the sample flight inspection task under the corresponding sample learning data.

[0104] Next, the server learns the model parameters of the initialized adaptive inspection decision model based on the sample learning data sequence.

[0105] The initialized adaptive inspection decision model is like an intelligent evaluation system in its initial state, and its internal neuron weight coefficients are set randomly or according to some default rules. The server loads the first sample learning data into this initialized model. In the model, a series of processing is performed based on the X example drone cluster description information, the example cluster description information, and the example flight inspection task description information. For example, the model will analyze the description vectors of each example knowledge field of the example drone to identify the factors that have an important impact on the success of the inspection task. It will also analyze the overall cluster performance factors in the example cluster description information and the task requirement factors in the example flight inspection task description information.

[0106] Then, the model determines the confidence level of the generated example estimate based on this information. This example estimate confidence level is the model's assessment of the likelihood that the example drone cluster will successfully complete the task in the sample flight inspection task based on its current parameter state (initial neuron weight coefficients, etc.).

[0107] Based on this example estimation confidence and the corresponding sample confidence, determine the cost value of the target training cost function. This cost value is actually a loss value, which reflects the gap between the model's example estimation confidence and the known sample confidence. If this cost value is large, it means that the current parameter setting of the model makes its evaluation result differ greatly from the actual sample confidence, and the model needs to be adjusted.

[0108] If the cost value of the target training cost function does not meet the previously defined training termination requirements (for example, the training termination requirement may be that the cost value is less than a certain set threshold, assuming that this threshold is 0.1), then the server will update the neuron weight coefficients in the adaptive inspection decision model. The update method can be based on some optimization algorithms, such as the gradient descent algorithm. By adjusting the neuron weight coefficients, the model will reprocess the sample learning data, generate the example estimation confidence again, and recalculate the cost value of the target training cost function.

[0109] The server will process the second and third sample learning data in the same way. Each time a sample learning data is processed, the cost value of the target training cost function is calculated based on the generated example estimation confidence and the corresponding sample confidence, and whether to update the model's neuron weight coefficients is determined based on whether the cost value meets the training termination requirements.

[0110] This process will continue until the cost value of the target training cost function corresponding to the adaptive patrol decision model meets the previously defined training termination requirements. For example, after multiple processing of sample learning data and adjustment of model parameters, the cost value calculated by the model for the difference between the example estimation confidence obtained after processing the third sample learning data and the corresponding sample confidence is 0.08, which is less than the set threshold of 0.1. At this time, the server will terminate the training and use the adaptive patrol decision model at the time of termination of training as the target adaptive patrol decision model. This target adaptive patrol decision model has been adjusted to a relatively accurate state after training with sample learning data, and can generate target estimation confidence more accurately based on the input drone description information, cluster description information, and flight inspection task description information, thereby providing a reliable basis for the actual drone inspection task scheduling decision.

[0111] In a possible implementation, step A120 includes:

[0112] Perform model optimization for the a-th model optimization stage based on the following steps, where a is not less than 2:

[0113] Step A121, load the ath sample learning data among the B sample learning data into the adaptive inspection decision model generated in the a-1th model optimization stage, the ath sample learning data corresponds to the ath sample confidence among the B sample confidences, the ath sample learning data includes Xa example drone cluster description information, the ath example cluster description information and the ath example flight inspection task description information, Xa=X.

[0114] Step A122, in the adaptive inspection decision model generated in the a-1th model optimization stage, the Xa example drone description information sets are encoded and represented to generate corresponding Xa example drone description encoding vectors, the ath example cluster description information is encoded and represented to generate the corresponding ath example cluster description encoding vector, and the ath example flight inspection task description information is encoded and represented to generate the corresponding ath example flight inspection task description encoding vector, and the Xa example drone description encoding vectors correspond to the Xa example drones in the ath sample learning data, respectively.

[0115] Step A123, in the adaptive inspection decision model generated in the a-1th model optimization stage, the Xa example drone description coding vectors are respectively analyzed for correlation with the ath example cluster description coding vector to generate Xa example drone associated coding vectors, and the Xa example drone associated coding vectors correspond to the Xa example drones in the ath sample learning data respectively.

[0116] Step A124, in the adaptive inspection decision model generated in the a-1th model optimization stage, determine the ath estimated confidence based on the Xa example drone association coding vectors, the ath example cluster description coding vector and the ath example flight inspection task description coding vector.

[0117] Step A125, determining the ath cost value of the target training cost function based on the ath estimated confidence and the ath sample confidence.

[0118] Step A126, under the condition that the ath cost value of the target training cost function meets the previously defined training termination requirement, terminate the training, and use the adaptive patrol decision model when the training is terminated as the target adaptive patrol decision model. Under the condition that the ath cost value of the target training cost function does not meet the previously defined training termination requirement, update the neuron weight coefficient of the adaptive patrol decision model generated in the a-1th model optimization stage to obtain the adaptive patrol decision model generated in the ath model optimization stage.

[0119] In this embodiment, in the entire model training process, in order to enable the adaptive inspection decision model to accurately evaluate and make decisions on drone inspection tasks, it is necessary to gradually optimize the model based on the sample learning data sequence. Here, it is assumed that B = 3, and the model optimization of the ath model optimization stage is started, where a is not less than 2.

[0120] First, the model optimization phase with a=2 is performed. The server loads the second sample learning data among the B sample learning data into the adaptive inspection decision model generated in the first model optimization phase. This second sample learning data corresponds to the second sample confidence among the B sample confidences. The second sample learning data includes Xa example drone cluster description information, the second example cluster description information, and the second example flight inspection task description information, where Xa=X, assuming X=5.

[0121] In the adaptive inspection decision model generated in the first model optimization stage, data processing begins. First, the description information sets of the five example drones are encoded and represented to generate the corresponding five example drone description encoding vectors. Taking one of the example drones as an example, this drone has various information in different inspection scenarios before, such as flight altitude adaptability and ability to cope with complex terrain in mountain inspection scenarios, signal interference response capabilities in urban suburban inspection scenarios, and water surface reflection processing capabilities in cross-river area inspection scenarios. This information is encoded according to the encoding rules set by the model to form an example drone description encoding vector. The same operation is performed on the other four example drones, and finally five example drone description encoding vectors are obtained, and these encoding vectors correspond to the five example drones in the second sample learning data.

[0122] Next, the second example cluster description information is encoded to generate the corresponding second example cluster description coding vector. This example cluster description information contains various performance-related information of the example drone cluster in multiple inspection scenarios. For example, in the mountain inspection scenario, the coordinated flight altitude control capability of the drones in the cluster, the overall stability in strong winds, and other information; in the urban suburban inspection scenario, the communication coordination capability of the drones in the cluster in response to interference from a large number of buildings, and in the cross-river area inspection scenario, the collaborative management capability of the drones in the cluster for equipment protection in humid environments, and other information. Encode this information into an example cluster description coding vector according to the coding rules.

[0123] At the same time, the description information of the second example flight inspection task is encoded to generate the corresponding second example flight inspection task description encoding vector. If this example flight inspection task is in a mountain inspection scenario, the task description information may include the detailed terrain structure of the mountains in the task area, such as valley depth, peak height distribution and other information, the specific layout of the power lines to be inspected, such as line direction, power tower spacing and other information, and information related to the time limit of the task, such as whether the inspection of a specific area needs to be completed within a specific time period. Encode this information into an example flight inspection task description encoding vector.

[0124] Then, in the adaptive inspection decision model generated in the first model optimization stage, the description coding vectors of the five example drones are analyzed for correlation with the description coding vectors of the second example cluster, and five example drone correlation coding vectors are generated. Taking one of the example drones as an example, the description coding vector of the example drone is analyzed for correlation with the description coding vector of the example cluster. The description coding vector of the example drone is first decomposed into different feature groups, such as the flight capability-related feature group (including flight altitude adaptability, speed control capability, etc.), the device status-related feature group (including device health status, battery power, etc.) and the task history-related feature group (including the number of successes in previous inspection tasks, the results of handling special situations, etc.). At the same time, the example cluster description coding vector can also be regarded as containing cluster flight capability-related features (such as the overall flight altitude coordination of the cluster, speed adjustment capability, etc.), cluster device status-related features (such as the average health status of the equipment in the cluster, etc.) and cluster task history-related features (such as the overall performance of the cluster in performing inspection tasks before, etc.).

[0125] Based on semantic understanding and logical relationships, a preliminary correlation analysis is performed on the flight capability-related feature group of the example UAV and the flight capability-related features of the example cluster. For example, if the flight altitude adaptability of the example UAV is good, and the example cluster has a specific flight altitude coordination strategy in the mountain inspection scenario, their correlation degree is determined. If the correlation degree is high, the basis of the correlation (such as the UAV's altitude adaptability helps to meet the cluster's flight altitude coordination strategy) and the preliminary judgment of the correlation degree are recorded, and a deep interaction operation is performed to generate a coding segment after deep interaction. If the correlation degree is medium, a partial interaction is performed to obtain a coding segment after partial interaction. If the correlation degree is low, an identification record is performed to obtain identification record data. In the same way, an association analysis is performed on the device status-related feature group and the task history-related feature group. Finally, the coding segments after deep interaction and partial interaction are recombined according to the set logical order (such as the order of importance of the impact on the success of the inspection task), and the identification record data is attached to the recombined coding segment in a predefined coding form to obtain the example UAV association coding vector of this example UAV. The other four example drones are operated in the same way, and finally five example drone associated coding vectors are generated, which correspond to the five example drones in the second sample learning data respectively.

[0126] Then, in the adaptive inspection decision model generated in the first model optimization stage, the second estimated confidence is determined based on the five example drone association coding vectors, the second example cluster description coding vector, and the second example flight inspection task description coding vector. This process is like the model comprehensively considering the relationship between the example drone individual and the cluster and the task requirements, and then evaluating the possibility of the example drone cluster successfully completing the task in this example flight inspection task. For example, the model will consider the performance of the individual drone capabilities in the example drone association coding vector in the example cluster, as well as the degree of match between the overall cluster performance in the example cluster description coding vector and the task requirements in the example flight inspection task description coding vector, and other factors, and obtain the second estimated confidence through complex internal calculations.

[0127] Based on the second estimated confidence and the second sample confidence, the second cost value of the target training cost function is determined. This cost value reflects the gap between the estimated confidence obtained by the model and the known sample confidence. If the second sample confidence is obtained through actual experience or accurate evaluation, indicating the true confidence of the example drone cluster in successfully completing the task in this example flight inspection task, then this cost value measures the degree of deviation between the current evaluation result of the model and the actual situation.

[0128] If the second cost value of the target training cost function meets the previously defined training termination requirements (for example, the training termination requirement is that the cost value is less than 0.1), the server will terminate the training and use the adaptive inspection decision model at this time as the target adaptive inspection decision model. This target adaptive inspection decision model can be used for actual drone inspection task decisions. However, if the second cost value does not meet the training termination requirements, the server will update the neuron weight coefficients of the adaptive inspection decision model generated in the first model optimization phase to obtain the adaptive inspection decision model generated in the second model optimization phase. The process of updating the neuron weight coefficients is like adjusting the evaluation logic inside the model, so that the model can more accurately evaluate the success probability of the example drone cluster in the example flight inspection task.

[0129] Next, the model optimization phase with a=3 is carried out, and the process is similar to a=2. The server loads the third sample learning data into the adaptive inspection decision model generated in the second model optimization phase, and then encodes the example drone description information set, the example cluster description information, and the example flight inspection task description information. Then, the correlation analysis between the example drone description encoding vector and the example cluster description encoding vector is performed to determine the third estimated confidence. Based on the third estimated confidence and the third sample confidence, the third cost value of the target training cost function is determined. Finally, according to whether this cost value meets the training termination requirements, it is decided whether to terminate the training or continue to update the model's neuron weight coefficients to obtain a more accurate adaptive inspection decision model. This step-by-step optimization process continuously adjusts the parameters of the model so that it can better adapt to different example learning data, and finally obtains a target adaptive inspection decision model that can accurately evaluate drone inspection tasks.

[0130] In a possible implementation, step S120 includes:

[0131] The X drone description information sets, cluster description information and flight inspection task description information are loaded into the target adaptive inspection decision model learned in advance, and a first estimated confidence generated by the target adaptive inspection decision model is generated, wherein the target estimated confidence includes the first estimated confidence, and the first estimated confidence represents the confidence of a previously defined drone cluster in the Z drone clusters in successfully completing the low-altitude flight inspection task in the flight inspection task under the condition that the Z drone clusters are used as the drone clusters scheduled to the flight inspection task. Or

[0132] The X drone description information sets, cluster description information and flight inspection task description information are loaded into the target adaptive inspection decision model learned in advance, and a second estimated confidence generated by the target adaptive inspection decision model is generated, wherein the target estimated confidence includes the second estimated confidence, and the second estimated confidence represents the confidence that each drone cluster in the Z drone clusters successfully completes the low-altitude flight inspection task in the flight inspection task under the condition that the Z drone clusters are used as each drone cluster scheduled to the flight inspection task.

[0133] In this embodiment, in the low-altitude inspection operation scenario of power lines, after the server obtains X drone description information sets, cluster description information and flight inspection task description information, it loads this information into the target adaptive inspection decision model learned in advance to generate target estimation confidence. The target estimation confidence here includes two cases, namely the first estimation confidence and the second estimation confidence.

[0134] First, let's look at the situation of generating the first estimated confidence. Assume that X = 5 drones are dispatched to a flight inspection task under the target inspection execution scenario (such as mountain inspection scenario) in this low-altitude inspection of power lines, and these 5 drones are divided into Z = 3 drone clusters according to the current cluster division strategy. The server loads the X drone description information sets, cluster description information, and flight inspection task description information into the target adaptive inspection decision model.

[0135] The description information set of each drone contains information about each drone in multiple inspection execution scenarios (mountainous areas, urban suburbs, and cross-river areas). Taking one of the drones as an example, in the mountain inspection scenario, its information may include the flight trajectory in the complex terrain of the mountain during the previous inspection, the device status during flight (such as battery power consumption, device temperature, etc.), and the historical records of dealing with special situations in the mountain (such as strong wind interference). In the urban suburban inspection scenario, there may be information such as route adjustment when flying in densely built areas and the degree of signal interference. In the cross-river area inspection scenario, there is relevant information such as dealing with water surface reflection imaging problems and the impact of humid environments on equipment. The cluster description information contains the description vectors of the three drone clusters in multiple inspection scenarios, such as the cooperative flight capability and overall stability in dealing with strong winds between drones in a cluster in the mountain inspection scenario; the communication cooperation capability of drones in the cluster in dealing with signal interference in the urban suburban inspection scenario; and the overall performance of drones in the cluster in adapting to humid environments in the cross-river area inspection scenario. The flight inspection task description information includes relevant information about the target inspection execution scenario (mountainous area inspection scenario), such as the topographical features of the mountainous area (mountain direction, altitude changes, etc.), the distribution of power lines and power towers that need to be inspected, and the time limit for this inspection task (such as completion between 8 a.m. and 5 p.m.).

[0136] When this information is loaded into the target adaptive inspection decision model, the model will process this information comprehensively to generate the first estimated confidence. This first estimated confidence represents the confidence that one of the three drone clusters defined previously will successfully complete the low-altitude flight inspection task in this flight inspection task, under the condition that the three drone clusters are used as the drone clusters scheduled to the flight inspection task. For example, assuming that a previously defined drone cluster is a cluster containing two drones, the model will focus on the matching degree between the capability information of the individual drones in this cluster and the requirements of the flight inspection task, as well as the synergy between this cluster and other clusters in the overall inspection task (although the focus here is on this specific cluster). For the individual drones in this cluster, the model will consider whether their flight capabilities (such as flight altitude adaptability, speed stability, etc.) in the mountain inspection scenario can meet the current mountain terrain and task requirements (such as whether they can accurately inspect power lines within the specified flight altitude range, whether they can maintain a stable flight speed in complex terrain, etc.). At the same time, the model will also consider whether the equipment is in good condition, such as whether the battery power is sufficient to support the entire inspection task process, whether the equipment temperature is within the normal range to avoid affecting the performance of the equipment, etc. From the perspective of the cluster, the model will evaluate the collaborative ability of this cluster in the mountain inspection scenario, such as whether the flight distance between the two drones and the division of labor and collaboration (such as one responsible for the main inspection and the other responsible for assistance and data backup) are reasonable and effective. Combining these factors and the adaptability of the cluster to the entire mission scenario (including the collaborative relationship with other clusters, etc.), the model uses complex internal calculations and analysis to derive the first estimated confidence level of this specific cluster's successful completion of the mission in this flight inspection mission.

[0137] Let's look at the situation of generating the second estimated confidence. Similarly, the X drone description information sets, cluster description information, and flight inspection task description information are loaded into the target adaptive inspection decision model learned in advance. The model will still conduct a comprehensive and in-depth analysis of this information. At this time, the generated second estimated confidence represents the confidence of each of the three drone clusters in successfully completing the low-altitude flight inspection task in this flight inspection task, under the condition that these three drone clusters are used as the drone clusters scheduled to the flight inspection task.

[0138] In this process, the model will conduct a similar but more comprehensive evaluation of each drone cluster. Taking the three drone clusters as an example, for the first drone cluster, the model will consider the flight capabilities, equipment status, mission history and other information of the individual drones in the cluster and the matching degree with the flight inspection task requirements, just like the previous analysis of specific clusters. For example, in the mountain inspection scenario, whether the flight altitude, speed and other parameters of the drones in the cluster can adapt to the mountain terrain, whether the equipment can operate stably, and whether the performance in the previous mountain inspection is good. At the same time, it will also focus on analyzing the synergy between this cluster and the other two clusters, such as whether their flight routes cooperate with each other in the entire inspection task, and whether the division of tasks is reasonable (for example, one cluster is responsible for the inspection of the eastern section of the mountain area, another cluster is responsible for the inspection of the western section, and the third cluster is responsible for the inspection of the middle area or key areas, etc.). For the second and third drone clusters, the model will also conduct the same comprehensive analysis, including the situation of the individual drones in each of them and the synergy with other clusters. By comprehensively considering all these factors, the model evaluates the possibility of each of the three drone clusters successfully completing the task in this flight inspection task, and finally obtains the second estimated confidence. This second estimated confidence level can comprehensively reflect the confidence level of the three drone clusters in successfully completing the flight inspection mission as a whole, and provide an important basis for subsequent decision-making (such as whether to schedule inspection tasks according to the current cluster division).

[0139] Figure 2 The detection system 100 shown includes: a processor 1001 and a memory 1003. The processor 1001 and the memory 1003 are connected, such as through a bus 1002. Optionally, the detection system 100 may also include a transceiver 1004, which may be used for data interaction between the server and other servers, such as data transmission and / or data reception. It should be noted that the transceiver 1004 is not limited to one in actual scheduling, and the structure of the detection system 100 does not constitute a limitation on the embodiments of the present application.

[0140] Processor 1001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of this application. Processor 1001 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0141] The bus 1002 may include a path to transmit information between the above components. The bus 1002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 1002 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 2 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0142] The memory 1003 can be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), magnetic disk storage medium, other magnetic storage devices, or any other medium that can be used to carry or store program code and can be read by a computer, without limitation herein.

[0143] The memory 1003 is used to store program codes for executing the embodiments of the present application, and the execution is controlled by the processor 1001. The processor 1001 is used to execute the program codes stored in the memory 1003 to implement the steps shown in the above method embodiments.

[0144] In addition, an embodiment of the present application also provides a readable storage medium, in which computer executable instructions are pre-installed. When the processor executes the computer executable instructions, the above-mentioned adaptive detection method for low-altitude flight inspection is implemented.

[0145] Similarly, it should be noted that in order to simplify the description disclosed in this application, thereby helping to understand one or more embodiments of the invention, in the foregoing description of the embodiments of the present application, multiple features are sometimes combined into one embodiment, drawings, or descriptions thereof. Similarly, it should be noted that in order to simplify the description disclosed in this application, thereby helping to understand one or more embodiments of the invention, in the foregoing description of the embodiments of the present application, multiple features are sometimes combined into one embodiment, drawings, or descriptions thereof.

Claims

1. An adaptive detection method for low-altitude flight inspection, characterized in that: The method comprises: Obtain X drone description information sets, cluster description information, and flight inspection task description information, wherein the X drone description information sets include Y knowledge field description vectors of each drone in the X drones, the X drones are drones that are dispatched to a flight inspection task in a target inspection execution scenario in a target low-altitude inspection operation, the Y knowledge field description vectors of each drone include description vectors of each drone in multiple inspection execution scenarios, the multiple inspection execution scenarios are inspection execution scenarios with full authority authentication in the target low-altitude inspection operation, the cluster description information includes description vectors of each drone cluster in the Z drone clusters in the multiple inspection execution scenarios, the flight inspection task description information includes description vectors of the target inspection execution scenario, the Z drone clusters are multiple drone clusters generated by clustering the X drones according to the current cluster division strategy, and X, Y, and Z are all not less than 2; The X drone description information sets, cluster description information and flight inspection task description information are loaded into a target adaptive inspection decision model learned in advance, and a target estimation confidence generated by the target adaptive inspection decision model is generated, wherein the target estimation confidence represents the confidence that the Z drone clusters corresponding to the Z drone clusters successfully complete the low-altitude flight inspection task in the flight inspection task under the condition that the Z drone clusters are used as the respective drone clusters scheduled to the flight inspection task; Under the condition that the target estimation confidence meets the set requirements, the Z drone clusters are dispatched as the drone clusters in the one flight inspection task to obtain the adaptive detection results before the low-altitude flight inspection.

2. The adaptive detection method for low-altitude flight inspection according to claim 1 is characterized in that: The step of loading the X drone description information sets, cluster description information, and flight inspection task description information into a target adaptive inspection decision model learned in advance, and generating a target estimation confidence generated by the target adaptive inspection decision model, comprises: Encode and represent the X drone description information sets to generate corresponding X drone description encoding vectors, encode and represent the cluster description information to generate corresponding cluster description encoding vectors, and encode and represent the flight inspection task description information to generate corresponding flight inspection task description encoding vectors, wherein the X drone description encoding vectors correspond to the X drones respectively; Performing correlation analysis on the X drone description coding vectors and the flight inspection task description coding vectors to generate X drone association coding vectors, where the X drone association coding vectors correspond to the X drones respectively; The target estimation confidence is determined based on the X drone association coding vectors, the cluster description coding vector and the flight inspection task description coding vector.

3. The adaptive detection method for low-altitude flight inspection according to claim 2 is characterized in that: The step of performing correlation analysis on the X drone description coding vectors and the flight inspection task description coding vectors to generate X drone association coding vectors includes: Based on the following steps, the ath drone description coding vector among the X drone description coding vectors is analyzed for correlation with the flight inspection task description coding vector to generate the ath drone association coding vector among the X drone association coding vectors, where a is a positive integer not less than 1 and not greater than X: Interact the a-th UAV description coding vector with the flight inspection task description coding vector to generate an a-th blending coding vector; The aggregation result of the a-th UAV description coding vector and the a-th blending coding vector is used as the a-th UAV associated coding vector; or, the a-th blending coding vector is fused with the a-th influencing factor defined previously to generate the a-th fused coding vector, and the aggregation result of the a-th UAV description coding vector and the a-th fused coding vector is used as the a-th UAV associated coding vector; The step of interacting the a-th UAV description coding vector with the flight inspection task description coding vector to generate the a-th blended coding vector includes: Decomposing the a-th drone description coding vector into a plurality of different first feature groups, wherein the first feature groups are divided based on different attributes of the drone, specifically including a flight capability-related feature group, a device status-related feature group, and a mission history-related feature group; And, decomposing the flight inspection task description coding vector into a plurality of second feature groups related to different task scenario requirements, specifically including a task area geographical feature group, a task target related feature group and a task time limit related feature group; For each first feature group, based on semantic understanding and logical relationship, preliminary association analysis is performed one by one with all second feature groups, and an association relationship description document is established for each preliminary association analysis result, wherein the association relationship description document is used to record the feature group pairs formed by the associated first feature group and the second feature group, as well as the basis for the association and the preliminary determined degree of association; According to the association relationship description document, a deep interaction operation is performed on a feature pair with a high association degree to generate a coded segment after the deep interaction, and a partial interaction is performed on a feature pair with a medium association degree to generate a coded segment after the partial interaction, and an identification record is performed on a feature pair with a low association degree to generate identification record data; The coding segments after deep interaction and partial interaction are recombined according to a set logical order, where the logical order is determined based on the task execution process or the importance of the data, and the identification record data is attached to the recombined coding segments in a predefined coding form to generate the ath interleaved coding vector.

4. The adaptive detection method for low-altitude flight inspection according to claim 2 is characterized in that: The determining the target estimation confidence based on the X drone association coding vectors, the cluster description coding vector and the flight inspection task description coding vector includes: Aggregate the X drone association coding vectors, the cluster description coding vector, and the flight inspection task description coding vector to generate an aggregated coding vector; The aggregated coding vector is subjected to full connection mapping processing based on a full connection mapping layer to generate the target estimation confidence.

5. The adaptive detection method for low-altitude flight inspection according to claim 1 is characterized in that: The step of obtaining X sets of drone description information includes: Acquire flight trajectory description information, immediate state description information, and long-term state description information of each drone among the X drones, and generate a description information set of the X drones, wherein the flight trajectory description information of each drone includes the flight trajectory information of each drone in the previous b flight inspection tasks, the long-term state description information of each drone includes the record information of each drone in each of the multiple inspection execution scenarios within a previous set time period, and the immediate state description information of each drone includes the current record information of each drone in each of the multiple inspection execution scenarios; And, get the cluster description information, including: Obtain average performance statistics of each drone cluster in the Z drone clusters in each inspection execution scenario in the multiple inspection execution scenarios; And obtain the flight inspection task description information, including: Obtain the scenario label index of the target inspection execution scenario, or obtain the scenario label index of the target inspection execution scenario and the timing node scheduled to the one-time flight inspection task.

6. The adaptive detection method for low-altitude flight inspection according to claim 1 is characterized in that: The method of dispatching the Z drone clusters as the respective drone clusters in the one flight inspection task under the condition that the target estimation confidence meets the set requirements includes: Under the condition that the deviation between the target estimation confidence and the threshold confidence is less than a set value, the Z drone clusters are used as the respective drone clusters scheduled to the one flight inspection mission; or Under the condition that the deviation between the target estimated confidence and the threshold confidence in the acquired A estimated confidences is minimal, the Z drone clusters are used as the drone clusters scheduled to the flight inspection task, the A estimated confidences include the estimated confidences determined by the target adaptive inspection decision model based on the drone description information, cluster description information and flight inspection task description information of each drone cluster division data in the A drone cluster division data, each drone cluster division data in the A drone cluster division data includes a plurality of drone clusters generated by clustering the X drones according to a corresponding cluster division strategy among A different cluster division strategies, the A different cluster division strategies include the current cluster division strategy, the number of clusters of the plurality of drone clusters included in each drone cluster division data is the same, and A is not less than 2; or Under the condition that the deviation between the target estimated confidence and the threshold confidence among the A estimated confidences is the smallest and the deviation is less than the set value, the Z drone clusters are dispatched as the respective drone clusters in the one flight inspection mission.

7. The adaptive detection method for low-altitude flight inspection according to any one of claims 1 to 6, characterized in that: The method further comprises: Obtain a sample learning data sequence and a sample confidence sequence, wherein the sample learning data sequence includes B sample learning data, each sample learning data includes X sample drone cluster description information, sample cluster description information, and sample flight inspection task description information, the X sample drone cluster description information includes Y sample knowledge field description vectors of each sample drone in the X sample drones, the X sample drones are drones that are dispatched to a sample flight inspection task in the sample inspection execution scenario in the target low-altitude inspection operation, and the Y sample knowledge field description vectors of each sample drone The quantity includes description vectors of each example drone in the multiple inspection execution scenarios, the example cluster description information includes description vectors of each example drone cluster in the Z example drone clusters in the multiple inspection execution scenarios, the example flight inspection task description information includes description vectors of the example inspection execution scenarios, the Z example drone clusters are multiple example drone clusters generated by clustering the X example drones according to the example cluster division strategy, the sample confidence sequence includes B sample confidences corresponding to the B sample learning data, and B is not less than 2; Model parameter learning is performed on the initialized adaptive patrol decision model according to the sample learning data sequence until the cost value of the target training cost function corresponding to the adaptive patrol decision model meets the previously defined training termination requirement, the training is terminated, and the adaptive patrol decision model at the time of the termination of the training is used as the target adaptive patrol decision model. Under the condition that the cost value of the target training cost function corresponding to the adaptive patrol decision model does not meet the previously defined training termination requirement, the neuron weight coefficient in the adaptive patrol decision model is updated; the cost value of the target training cost function is a loss value determined based on the example estimated confidence generated by the adaptive patrol decision model and the corresponding sample confidence of the B sample confidences, and the example estimated confidence is an estimated confidence determined by the adaptive patrol decision model based on the X example drone cluster description information, the example cluster description information and the example flight inspection task description information.

8. The adaptive detection method for low-altitude flight inspection according to claim 7 is characterized in that: The performing model parameter learning on the initialized adaptive inspection decision model according to the sample learning data sequence includes: Perform model optimization for the a-th model optimization stage based on the following steps, where a is not less than 2: Loading the ath sample learning data among the B sample learning data into the adaptive inspection decision model generated in the a-1th model optimization stage, wherein the ath sample learning data corresponds to the ath sample confidence among the B sample confidences, and the ath sample learning data includes Xa example drone cluster description information, the ath example cluster description information and the ath example flight inspection task description information, Xa=X; In the adaptive inspection decision model generated in the a-1th model optimization stage, the Xa example drone description information sets are encoded to generate corresponding Xa example drone description encoding vectors, the ath example cluster description information is encoded to generate the corresponding ath example cluster description encoding vector, and the ath example flight inspection task description information is encoded to generate the corresponding ath example flight inspection task description encoding vector, and the Xa example drone description encoding vectors correspond to the Xa example drones in the ath example learning data, respectively; In the adaptive inspection decision model generated in the a-1th model optimization stage, the Xa example drone description coding vectors are respectively analyzed for correlation with the ath example cluster description coding vector to generate Xa example drone association coding vectors, and the Xa example drone association coding vectors correspond to the Xa example drones in the ath sample learning data respectively; In the adaptive inspection decision model generated in the a-1th model optimization stage, determining an ath estimated confidence based on the Xa example drone association coding vectors, the ath example cluster description coding vector, and the ath example flight inspection task description coding vector; Determining an ath cost value of the target training cost function based on the ath estimated confidence and the ath sample confidence; Under the condition that the ath cost value of the target training cost function meets the previously defined training termination requirements, the training is terminated, and the adaptive patrol decision model at the time of termination of the training is used as the target adaptive patrol decision model; under the condition that the ath cost value of the target training cost function does not meet the previously defined training termination requirements, the neuron weight coefficients of the adaptive patrol decision model generated in the a-1th model optimization stage are updated to obtain the adaptive patrol decision model generated in the ath model optimization stage.

9. The adaptive detection method for low-altitude flight inspection according to any one of claims 1 to 6, characterized in that: The step of loading the X drone description information sets, cluster description information, and flight inspection task description information into a target adaptive inspection decision model learned in advance, and generating a target estimation confidence generated by the target adaptive inspection decision model, comprises: Loading the X drone description information sets, cluster description information, and flight inspection task description information into the target adaptive inspection decision model learned in advance, generating a first estimated confidence generated by the target adaptive inspection decision model, wherein the target estimated confidence includes the first estimated confidence, and the first estimated confidence represents the confidence that a previously defined drone cluster among the Z drone clusters successfully completes the low-altitude flight inspection task in the flight inspection task under the condition that the Z drone clusters are used as the drone clusters scheduled to the flight inspection task; or The X drone description information sets, cluster description information and flight inspection task description information are loaded into the target adaptive inspection decision model learned in advance, and a second estimated confidence generated by the target adaptive inspection decision model is generated, wherein the target estimated confidence includes the second estimated confidence, and the second estimated confidence represents the confidence that each drone cluster in the Z drone clusters successfully completes the low-altitude flight inspection task in the flight inspection task under the condition that the Z drone clusters are used as each drone cluster scheduled to the flight inspection task.

10. A detection system, characterized in that: The detection system includes a processor and a machine-readable storage medium, wherein the machine-readable storage medium stores machine-executable instructions, and the machine-executable instructions are loaded and executed by the processor to implement the adaptive detection method for low-altitude flight inspection as described in any one of claims 1-9.

Images