Report data processing system and method of unmanned aerial vehicle management platform based on AI

Abstract

The invention relates to the field of intelligent processing, and particularly discloses an AI-based report data processing system and method for an unmanned aerial vehicle management platform. Flight parameters, battery states and environment data information of an unmanned aerial vehicle are collected through an unmanned aerial vehicle sensor; and time sequence analysis and correlation are carried out on flight parameters, battery states and environment data information of the unmanned aerial vehicle based on a deep learning technology, so that a real-time flight log report is intelligently obtained. In this way, the system can collect the data of the unmanned aerial vehicle in real time and automatically generate the real-time flight log report, the workload of manual recording is reduced, the report generation efficiency and accuracy are improved, and therefore a basis is provided for flight parameter adjustment and performance optimization.

Description

Technical Field

[0001] The present application relates to the field of intelligent processing, and more specifically, to a report data processing system and method for an AI-based drone management platform. Background Art

[0002] The report data processing of the drone management platform refers to the process of collecting, analyzing and presenting key data during drone flight missions. Report data can help users understand drone performance, monitor flight activities, optimize operational processes, provide decision support, and comply with regulatory requirements.

[0003] However, since traditional methods usually involve manual collection and processing of drone data, they are unable to provide real-time data processing and report generation capabilities, which may lead to delays in data processing and report generation, and such delays may affect the immediate understanding of the drone’s flight status and decision-making.

[0004] Therefore, a report data processing system and method for an AI-based drone management platform is expected. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a report data processing system and method for an AI-based drone management platform, which collects the flight parameters, battery status and environmental data information of the drone through the drone sensor, and uses deep learning technology to perform time series analysis and correlation on the flight parameters, battery status and environmental data information of the drone, so as to intelligently obtain real-time flight log reports. In this way, the system can collect drone data in real time and automatically generate real-time flight log reports, reduce the workload of manual recording, and improve the efficiency and accuracy of report generation, thereby providing a basis for flight parameter adjustment and performance optimization.

[0006] According to one aspect of the present application, a report data processing system for an AI-based drone management platform is provided, which includes:

[0007] The UAV flight data acquisition module is used to collect the flight parameters, battery status and environmental data information of the UAV through the UAV sensor to obtain the time series of the flight parameters, the time series of the battery status and the time series of the environmental data;

[0008] A data regularization module, used for regularizing the time series of the flight parameters, the time series of the battery status, and the time series of the environmental data to obtain a sequence of flight parameter timing input vectors, a sequence of battery status timing input vectors, and a sequence of environmental data timing input vectors;

[0009] A timing feature extraction module, used to extract timing features from the sequence of the flight parameter timing input vector, the sequence of the battery status timing input vector, and the sequence of the environmental data timing input vector, respectively, to obtain a sequence of flight parameter timing feature vectors, a sequence of battery status timing feature vectors, and a sequence of environmental data timing feature vectors;

[0010] A parameter arrangement module, used to arrange the sequence of the flight parameter timing feature vectors, the sequence of the battery status timing feature vectors and the sequence of the environment data timing feature vectors into a flight parameter timing feature matrix, a battery status timing feature matrix and an environment data timing feature matrix;

[0011] A timing feature association module, used to perform timing feature association on the flight parameter timing feature matrix, the battery status timing feature matrix and the environmental data timing feature matrix to obtain a flight parameter timing association feature matrix, a battery status timing association feature matrix and an environmental data timing association feature matrix;

[0012] A global timing interdependence capture module is used to perform feature fusion on the flight parameter timing correlation feature matrix, the battery status timing correlation feature matrix and the environmental data timing correlation feature matrix to obtain a UAV flight process timing feature diagram, and then perform global timing interdependence capture to obtain a UAV flight process global timing dependency feature diagram;

[0013] The flight log report generation module is used to obtain a real-time flight log report based on the global time-series dependency feature diagram of the UAV flight process.

[0014] According to another aspect of the present application, a report data processing method for an AI-based drone management platform is provided, comprising:

[0015] The flight parameters, battery status and environmental data information of the drone are collected through the drone sensor to obtain a time series of the flight parameters, a time series of the battery status and a time series of the environmental data;

[0016] Regularizing the time series of the flight parameters, the time series of the battery status, and the time series of the environmental data to obtain a sequence of flight parameter timing input vectors, a sequence of battery status timing input vectors, and a sequence of environmental data timing input vectors;

[0017] Performing time series feature extraction on the sequence of the flight parameter time series input vectors, the sequence of the battery status time series input vectors, and the sequence of the environment data time series input vectors, respectively, to obtain a sequence of flight parameter time series feature vectors, a sequence of battery status time series feature vectors, and a sequence of environment data time series feature vectors;

[0018] Arranging the sequence of the flight parameter timing feature vectors, the sequence of the battery status timing feature vectors and the sequence of the environment data timing feature vectors into a flight parameter timing feature matrix, a battery status timing feature matrix and an environment data timing feature matrix;

[0019] Performing timing feature association on the flight parameter timing feature matrix, the battery status timing feature matrix, and the environmental data timing feature matrix to obtain a flight parameter timing association feature matrix, a battery status timing association feature matrix, and an environmental data timing association feature matrix;

[0020] The flight parameter timing correlation feature matrix, the battery status timing correlation feature matrix and the environment data timing correlation feature matrix are subjected to feature fusion to obtain a UAV flight process timing feature diagram, and then global timing interdependence capture is performed to obtain a UAV flight process global timing dependency feature diagram;

[0021] Based on the global time-series dependency feature graph of the UAV flight process, a real-time flight log report is obtained.

[0022] Compared with the prior art, the report data processing system and method of the AI-based drone management platform provided in this application collects the flight parameters, battery status and environmental data information of the drone through the drone sensor, and uses deep learning technology to perform time series analysis and correlation on the flight parameters, battery status and environmental data information of the drone, so as to intelligently obtain real-time flight log reports. In this way, the system can collect drone data in real time and automatically generate real-time flight log reports, reduce the workload of manual recording, improve the efficiency and accuracy of report generation, and provide a basis for flight parameter adjustment and performance optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0024] Figure 1 This is a block diagram of a report data processing system for an AI-based drone management platform according to an embodiment of the present application.

[0025] Figure 2 This is a schematic diagram of the architecture of a report data processing system for an AI-based drone management platform according to an embodiment of the present application.

[0026] Figure 3It is a block diagram of a global timing interdependence capture module in a report data processing system of an AI-based drone management platform according to an embodiment of the present application.

[0027] Figure 4 This is a block diagram of a flight log report generation module in a report data processing system of an AI-based drone management platform according to an embodiment of the present application.

[0028] Figure 5 This is a flowchart of a report data processing method for an AI-based drone management platform according to an embodiment of the present application.

[0029] The reference numerals are as follows:

[0030] 100. Report data processing system of AI-based drone management platform; 110. Drone flight data acquisition module; 120. Data regularization module; 130. Time series feature extraction module; 140. Parameter arrangement module; 150. Time series feature association module; 160. Global time series interdependence capture module; 161. Feature fusion unit; 162. Global time series capture unit; 170. Flight log report generation module; 171. Feature flattening unit; 172. Feature optimization unit; 173. Log report generation unit. DETAILED DESCRIPTION

[0031] The embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, but rather these embodiments are provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0032] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0033] It should be noted that the terms "first\second\third" involved in the embodiments of the present application are only used to distinguish similar objects, and do not represent a specific order for the objects. It is understandable that the specific order or sequence of "first\second\third" can be interchanged where permitted. It should be understood that the objects distinguished by "first\second\third" can be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein.

[0034] As shown in the present disclosure and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0035] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0036] The report data processing of the drone management platform refers to the process of collecting, analyzing and presenting key data during drone flight missions. Report data can help users understand drone performance, monitor flight activities, optimize operational processes, provide decision support, and comply with regulatory requirements.

[0037] However, since traditional methods usually involve manual collection and processing of drone data, they are unable to provide real-time data processing and report generation capabilities, which may lead to delays in data processing and report generation, and such delays may affect the immediate understanding of the drone’s flight status and decision-making.

[0038] Therefore, a report data processing system for an AI-based drone management platform is desired, which collects the flight parameters, battery status, and environmental data information of the drone through the drone sensor, and uses deep learning technology to perform time series analysis and correlation on the flight parameters, battery status, and environmental data information of the drone, so as to intelligently obtain real-time flight log reports. In this way, the system can collect drone data in real time and automatically generate real-time flight log reports, reduce the workload of manual recording, and improve the efficiency and accuracy of report generation, thereby providing a basis for flight parameter adjustment and performance optimization.

[0039] Figure 1 This is a block diagram of a report data processing system for an AI-based drone management platform according to an embodiment of the present application. Figure 2 FIG. 1 is a schematic diagram of the architecture of a report data processing system of an AI-based drone management platform according to an embodiment of the present application. Figure 1 and Figure 2As shown, according to the report data processing system 100 of the AI-based drone management platform of the embodiment of the present application, it includes: a drone flight data acquisition module 110, which is used to collect the flight parameters, battery status and environmental data information of the drone through the drone sensor to obtain the time series of the flight parameters, the time series of the battery status and the time series of the environmental data; a data regularization module 120, which is used to regularize the time series of the above-mentioned flight parameters, the time series of the above-mentioned battery status and the time series of the above-mentioned environmental data to obtain a sequence of flight parameter timing input vectors, a sequence of battery status timing input vectors and a sequence of environmental data timing input vectors; a timing feature extraction module 130, which is used to extract timing features from the sequence of the above-mentioned flight parameter timing input vectors, the sequence of the above-mentioned battery status timing input vectors and the sequence of the above-mentioned environmental data timing input vectors respectively to obtain a sequence of flight parameter timing feature vectors, a sequence of battery status timing feature vectors and a sequence of environmental data timing feature vectors; a parameter arrangement module 140, which is used to regularize the above-mentioned flight parameters. The sequence of parameter timing feature vectors, the sequence of the battery status timing feature vectors and the sequence of the environmental data timing feature vectors are arranged into a flight parameter timing feature matrix, a battery status timing feature matrix and an environmental data timing feature matrix; a timing feature association module 150 is used to perform timing feature association on the flight parameter timing feature matrix, the battery status timing feature matrix and the environmental data timing feature matrix to obtain a flight parameter timing association feature matrix, a battery status timing association feature matrix and an environmental data timing association feature matrix; a global timing interdependence capture module 160 is used to perform feature fusion on the flight parameter timing association feature matrix, the battery status timing association feature matrix and the environmental data timing association feature matrix to obtain a UAV flight process timing feature diagram, and then perform global timing interdependence capture to obtain a UAV flight process global timing dependency feature diagram; and a flight log report generation module 170 is used to obtain a real-time flight log report based on the UAV flight process global timing dependency feature diagram.

[0040] In the embodiment of the present application, the above-mentioned UAV flight data acquisition module 110 is used to collect the flight parameters, battery status and environmental data information of the UAV through the UAV sensor to obtain the time series of flight parameters, the time series of battery status and the time series of environmental data. It should be understood that the flight parameters, battery status and environmental data information of the UAV are key data for the operation and performance management monitoring of the UAV. In detail, the above-mentioned flight parameters include position, speed, flight direction, etc. The above-mentioned battery status includes battery voltage and current. The above-mentioned environmental data includes air temperature, humidity and pressure. Based on this, in the technical solution of the present application, the flight parameters, battery status and environmental data information of the UAV are collected through the UAV sensor to obtain the time series of flight parameters, the time series of battery status and the time series of environmental data, and these time series information are processed and analyzed in time series, which can be used for the generation of flight logs and historical data analysis, so that flight missions can be better planned and flight strategies can be adjusted, so that UAV operators and management systems can more effectively monitor the flight status of UAVs, optimize flight missions, ensure flight safety, and improve the performance and efficiency of UAVs.

[0041] In an embodiment of the present application, the data regularization module 120 is used to regularize the time series of the flight parameters, the time series of the battery status, and the time series of the environmental data to obtain a sequence of flight parameter time series input vectors, a sequence of battery status time series input vectors, and a sequence of environmental data time series input vectors. Specifically, in an embodiment of the present application, the data regularization module is used to regularize the time series of the flight parameters, the time series of the battery status, and the time series of the environmental data according to the time dimension and the sample dimension to obtain a sequence of flight parameter time series input vectors, a sequence of battery status time series input vectors, and a sequence of environmental data time series input vectors. Accordingly, considering that each parameter in the flight parameters, the battery status, and the environmental data has time series characteristic information on a time scale, and each parameter has different time series characteristics in time series, at the same time, there is also a correlation relationship between each parameter in time series. Therefore, in order to separately analyze and process the timing characteristic information of each parameter and the correlation between them, in the technical solution of the present application, the time series of the above-mentioned flight parameters, the time series of the above-mentioned battery status and the time series of the above-mentioned environmental data are regularized according to the time dimension and the sample dimension to obtain the sequence of the above-mentioned flight parameter timing input vectors, the sequence of the above-mentioned battery status timing input vectors and the sequence of the above-mentioned environmental data timing input vectors. This can ensure that all time series data are synchronized on the time axis, facilitate the analysis of the state and changes of the data at the same time point or time period, and the regularized data is easier to extract features, such as calculation speed, acceleration, battery charge and discharge rate, temperature and humidity changes, etc., so that the drone management platform can more effectively process and analyze the flight data of the drone, and improve the safety, efficiency and reliability of drone operations.

[0042] In the embodiment of the present application, the above-mentioned timing feature extraction module 130 is used to extract timing features from the sequence of the above-mentioned flight parameter timing input vector, the sequence of the above-mentioned battery status timing input vector and the sequence of the above-mentioned environmental data timing input vector respectively to obtain the sequence of flight parameter timing feature vectors, the sequence of battery status timing feature vectors and the sequence of environmental data timing feature vectors. Specifically, in the embodiment of the present application, the above-mentioned timing feature extraction module is used to: input the sequence of the above-mentioned flight parameter timing input vector, the sequence of the above-mentioned battery status timing input vector and the sequence of the above-mentioned environmental data timing input vector into the timing feature extractor based on the deep time convolution network to obtain the sequence of the above-mentioned flight parameter timing feature vector, the sequence of the above-mentioned battery status timing feature vector and the sequence of the above-mentioned environmental data timing feature vector. It should be understood that each timing input vector of the sequence of the above-mentioned flight parameter timing input vector, the sequence of the above-mentioned battery status timing input vector and the sequence of the above-mentioned environmental data timing input vector contains the timing feature information and dynamic changes about each parameter. Therefore, in order to extract and mine the timing feature information implied in each of the above-mentioned timing input vectors, in the technical solution of the present application, the sequence of the above-mentioned flight parameter timing input vectors, the sequence of the above-mentioned battery status timing input vectors, and the sequence of the above-mentioned environmental data timing input vectors are input into a timing feature extractor based on a deep time convolutional network to obtain the sequence of the above-mentioned flight parameter timing feature vectors, the sequence of the above-mentioned battery status timing feature vectors, and the sequence of the above-mentioned environmental data timing feature vectors. It is worth mentioning that the above-mentioned deep time convolutional network is specially designed to process time series data and can capture the timing dynamic characteristics in the data. That is, through the above-mentioned timing feature extractor based on the deep time convolutional network, valuable timing information can be extracted from the sequences of the timing input vectors of the flight parameters, battery status, and environmental data of the drone, respectively, providing richer input for subsequent understanding and analysis.

[0043] In an embodiment of the present application, the parameter arrangement module 140 is used to arrange the sequence of the flight parameter timing feature vector, the sequence of the battery state timing feature vector and the sequence of the environmental data timing feature vector into a flight parameter timing feature matrix, a battery state timing feature matrix and an environmental data timing feature matrix. Accordingly, considering that each parameter in the sequence of the flight parameter timing feature vector, the sequence of the battery state timing feature vector and the sequence of the environmental data timing feature vector has a timing correlation relationship of each data dimension. Based on this, in order to integrate the timing feature information belonging to each dimension so as to conduct a more in-depth analysis and processing later, in the technical solution of the present application, the sequence of the flight parameter timing feature vector, the sequence of the battery state timing feature vector and the sequence of the environmental data timing feature vector are arranged into a flight parameter timing feature matrix, a battery state timing feature matrix and an environmental data timing feature matrix, so that the data of each dimension can be fully integrated, thereby improving the timing correlation between each parameter in each data dimension.

[0044] In an embodiment of the present application, the above-mentioned timing feature association module 150 is used to perform timing feature association on the above-mentioned flight parameter timing feature matrix, the above-mentioned battery status timing feature matrix and the above-mentioned environmental data timing feature matrix to obtain the flight parameter timing association feature matrix, the battery status timing association feature matrix and the environmental data timing association feature matrix. Specifically, in an embodiment of the present application, the above-mentioned timing feature association module is used to: input the above-mentioned flight parameter timing feature matrix, the above-mentioned battery status timing feature matrix and the above-mentioned environmental data timing feature matrix into the timing feature association module based on the two-way attention mechanism to obtain the above-mentioned flight parameter timing association feature matrix, the above-mentioned battery status timing association feature matrix and the above-mentioned environmental data timing association feature matrix. It should be understood that considering that the above-mentioned flight parameter timing feature matrix, the above-mentioned battery status timing feature matrix and the above-mentioned environmental data timing feature matrix contain the mutual correlation relationship of each data dimension on the time scale, and some of these relationships are important feature information, and some are irrelevant. At the same time, considering that the two-way attention mechanism can simultaneously consider the context information in two directions, so as to capture the long-term and short-term dependencies in the time series data. Therefore, in the technical solution of the present application, the above-mentioned flight parameter timing feature matrix, the above-mentioned battery status timing feature matrix and the above-mentioned environmental data timing feature matrix are respectively input into the timing feature association module based on the bidirectional attention mechanism to respectively capture and mine out the more significant and relevant timing features in the above-mentioned flight parameter timing feature matrix, the above-mentioned battery status timing feature matrix and the above-mentioned environmental data timing feature matrix, thereby obtaining the above-mentioned flight parameter timing association feature matrix, the above-mentioned battery status timing association feature matrix and the above-mentioned environmental data timing association feature matrix.

[0045] In an embodiment of the present application, the global timing interdependence capture module 160 is used to perform feature fusion on the flight parameter timing correlation feature matrix, the battery status timing correlation feature matrix and the environmental data timing correlation feature matrix to obtain a timing feature diagram of the UAV flight process, and then perform global timing interdependence capture to obtain a global timing dependency feature diagram of the UAV flight process.

[0046] Figure 3 : is a block diagram of a global timing interdependence capture module in a report data processing system of an AI-based drone management platform according to an embodiment of the present application. Specifically, in the embodiment of the present application, Figure 3 As shown, the above-mentioned global timing interdependence capture module 160 includes: a feature fusion unit 161, which is used to perform feature fusion on the above-mentioned flight parameter timing correlation feature matrix, the above-mentioned battery status timing correlation feature matrix and the above-mentioned environmental data timing correlation feature matrix along the channel dimension to obtain the above-mentioned UAV flight process timing feature diagram; and a global timing capture unit 162, which is used to input the above-mentioned UAV flight process timing feature diagram into the global timing interdependence module based on the non-local neural model to obtain the above-mentioned UAV flight process global timing dependency feature diagram.

[0047] Specifically, the feature fusion unit 161 is used to perform feature fusion along the channel dimension on the flight parameter time series correlation feature matrix, the battery state time series correlation feature matrix and the environmental data time series correlation feature matrix to obtain the above-mentioned UAV flight process time series feature diagram. Accordingly, considering that the flight parameter time series correlation feature matrix, the battery state time series correlation feature matrix and the environmental data time series correlation feature matrix are all important parameter information during the UAV flight process. Therefore, in order to fully and comprehensively display the flight state of the UAV during flight, in the technical solution of the present application, the flight parameter time series correlation feature matrix, the battery state time series correlation feature matrix and the environmental data time series correlation feature matrix are feature fused along the channel dimension to obtain the above-mentioned UAV flight process time series feature diagram. It should be understood that the fusion of the channel dimension allows data from different sources (flight parameters, battery status, environmental data) to be integrated into a unified feature representation, which helps the model to understand the flight state of the UAV more comprehensively, thereby improving the model's ability to understand the UAV flight process.

[0048] Specifically, the global timing capture unit 162 is used to input the above-mentioned UAV flight process timing feature graph into the global timing interdependence module based on the non-local neural model to obtain the above-mentioned UAV flight process global timing dependency feature graph. It should be understood that considering that the above-mentioned UAV flight process timing feature graph is obtained by fusing the above-mentioned flight parameter timing correlation feature matrix, the above-mentioned battery state timing correlation feature matrix and the above-mentioned environmental data timing correlation feature matrix along the channel dimension, therefore, in the technical solution of the present application, in order to more clearly understand and analyze the global characteristics of the entire flight process, in the technical solution of the present application, the above-mentioned UAV flight process timing feature graph is input into the global timing interdependence module based on the non-local neural model to obtain the above-mentioned UAV flight process global timing dependency feature graph. That is to say, the above-mentioned non-local neural model is good at capturing long-distance dependencies in time series data, which is crucial for understanding the global patterns and trends in the flight process, and by analyzing the entire flight process, the non-local model can provide a global understanding of the flight behavior, thereby obtaining a UAV flight process global timing dependency feature graph containing more global timing features.

[0049] In an embodiment of the present application, the flight log report generating module 170 is used to obtain a real-time flight log report based on the global timing dependency feature graph of the UAV flight process. Figure 4 : is a block diagram of a flight log report generation module in a report data processing system of an AI-based drone management platform according to an embodiment of the present application. Specifically, in the embodiment of the present application, Figure 4 As shown, the flight log report generation module 170 includes: a feature flattening unit 171, used to perform feature flattening on the global timing dependency feature graph of the UAV flight process to obtain a global timing dependency feature vector of the UAV flight process; a feature optimization unit 172, used to perform feature optimization on the global timing dependency feature vector of the UAV flight process to obtain an optimized global timing dependency feature vector of the UAV flight process; and a log report generation unit 173, used to input the optimized global timing dependency feature vector of the UAV flight process into a generator-based flight log report processor to obtain the real-time flight log report.

[0050] Specifically, the feature optimization unit 172 is used to perform feature optimization on the global time-dependent feature vector of the UAV flight process to obtain the optimized global time-dependent feature vector of the UAV flight process. Specifically, in the embodiment of the present application, the feature optimization unit is used to: perform backward compatibility perception based on the generated regression domain parameter space on the global time-dependent feature vector of the UAV flight process to obtain the optimized global time-dependent feature vector of the UAV flight process. In particular, in the technical solution of the present application, considering that various noises and interferences may exist in actual data collection, these factors will affect the accuracy and consistency of the data, resulting in the extracted global time-dependent feature vector of the UAV flight process having poor fluency in the time dimension. In addition, the design of modules such as the deep time convolution network and the bidirectional attention mechanism may not be ideal, and the key features and time-dependent relationships in the flight data cannot be well captured, resulting in the extraction of the global time-dependent feature vector of the UAV flight process being not accurate and coherent. That is, the manifold monotonicity of the overall feature distribution of the global time-dependent feature vector of the UAV flight process is poor. Due to the poor flow monotonicity of the global time-dependent feature vector of the UAV flight process, the generator lacks sufficient constraints when performing generative regression, that is, the generator cannot make good use of the input global time-dependent feature vector of the UAV flight process to accurately generate flight log reports, because the generator cannot obtain enough information from the global time-dependent feature vector of the UAV flight process to correctly predict the output. At the same time, due to the poor flow monotonicity of the overall feature distribution of the global time-dependent feature vector of the UAV flight process, the generated global time-dependent feature vector may lack stability and consistency, which will lead to the lack of accuracy and reliability of the generated flight log report. Based on this, the global time-dependent feature vector of the UAV flight process is backward-compatible perceived based on the parameter space of the generative regression domain to obtain the optimized global time-dependent feature vector of the UAV flight process.

[0051] More specifically, in an embodiment of the present application, the feature optimization unit is used to: multiply the global timing-dependent feature vector of the UAV flight process by the generated weight matrix of the generator to obtain an intermediate feature vector; splice the global timing-dependent feature vector of the UAV flight process with the intermediate feature vector to obtain a spliced ​​feature vector; multiply the spliced ​​feature vector by the first weight matrix and then add the first bias vector to obtain a coded spliced ​​feature vector; activate the coded spliced ​​feature vector through a sigmoid function to obtain an activation value; subtract the activation value from the intermediate feature vector and then multiply it by position to obtain weighted intermediate feature vector; multiplying the above activation value by the above UAV flight process global timing dependency feature vector by position to obtain a weighted UAV flight process global timing dependency feature vector; subtracting the above intermediate feature vector from the above UAV flight process global timing dependency feature vector by position to obtain a difference feature vector; adding the above weighted intermediate feature vector to the above weighted UAV flight process global timing dependency feature vector and then dividing it by the above difference feature vector to obtain a second intermediate feature vector; and passing the above second intermediate feature vector through the ReLU function to obtain an optimized UAV flight process global timing dependency feature vector.

[0052] In the embodiment of the present application, specifically, the feature optimization unit is used to: perform backward compatibility perception based on the generated regression domain parameter space on the global timing dependency feature vector of the UAV flight process using the following optimization formula to obtain the global timing dependency feature vector of the optimized UAV flight process; wherein the optimization formula is:

[0053]

[0054] Among them, v c represents the global time-dependent feature vector of the above-mentioned UAV flight process, M is the generation weight matrix of the generator, concat represents the concatenation function, W1 represents the first weight matrix, b1 represents the first bias vector, Represents matrix multiplication, sigmoid represents the S-shaped activation function, τ represents the activation value, ReLU represents the linear rectification function, and V′ represents the global timing dependency feature vector of the above-mentioned optimized UAV flight process.

[0055] In the technical solution of the present application, the manifold monotonicity of the overall feature distribution of the global timing-dependent feature vector of the above-mentioned UAV flight process is poor, resulting in insufficient generative regression constraints relative to the overall feature distribution of the above-mentioned generator when it is generated and regressed through the generator, thereby affecting the accuracy of the generated results.

[0056] Therefore, in the technical solution of the present application, the global timing-dependent feature vector of the UAV flight process is perceived for backward compatibility based on the parameter space of the generative regression domain, and the geometric characteristics of the feature manifold of the global timing-dependent feature vector of the UAV flight process are auxiliary analyzed and described by the generation weight matrix of the generator, and the difference between the auxiliary description analysis result and the feature manifold of the original global timing-dependent feature vector of the UAV flight process is used as a differential factor to support the descriptiveness of the improved global timing-dependent feature vector of the UAV flight process for the predetermined generation result of the generator, and finally an activation mechanism is added for activation to maintain the reinforcement of the distribution dependency with a forward description. In this way, the manifold monotonicity of the overall feature layout of the global timing-dependent feature vector of the UAV flight process is significantly improved, thereby enhancing its generative regression constraint relative to the overall feature layout of the generator and improving the accuracy of the generation result.

[0057] Specifically, the log report generation unit 173 is used to input the global timing dependency feature vector of the optimized UAV flight process into the generator-based flight log report processor to obtain the real-time flight log report. That is, the real-time flight log report is intelligently obtained by classifying and processing the global timing dependency feature vector of the optimized UAV flight process. In this way, the system can collect UAV data in real time and automatically generate real-time flight log reports, reduce the workload of manual recording, improve the efficiency and accuracy of report generation, and provide a basis for flight parameter adjustment and performance optimization.

[0058] In summary, the report data processing system 100 of the AI-based drone management platform based on the embodiment of the present application is explained, which collects the flight parameters, battery status and environmental data information of the drone through the drone sensor, and uses the deep learning technology to perform time series analysis and correlation on the flight parameters, battery status and environmental data information of the above-mentioned drone, so as to intelligently obtain real-time flight log reports. In this way, the system can collect drone data in real time and automatically generate real-time flight log reports, reduce the workload of manual recording, improve the efficiency and accuracy of report generation, and provide a basis for flight parameter adjustment and performance optimization.

[0059] Figure 5 Flow chart of the report data processing method of the AI-based drone management platform according to the embodiment of the present application. Figure 5As shown, according to the report data processing method of the AI-based drone management platform of the embodiment of the present application, the method includes: S110, collecting the flight parameters, battery status and environmental data information of the drone through the drone sensor to obtain the time series of the flight parameters, the time series of the battery status and the time series of the environmental data; S120, performing data regularization on the time series of the flight parameters, the time series of the battery status and the time series of the environmental data to obtain a sequence of flight parameter timing input vectors, a sequence of battery status timing input vectors and a sequence of environmental data timing input vectors; S130, performing timing feature extraction on the sequence of the flight parameter timing input vectors, the sequence of the battery status timing input vectors and the sequence of the environmental data timing input vectors respectively to obtain a sequence of flight parameter timing feature vectors, a sequence of battery status timing feature vectors and a sequence of environmental data timing feature vectors; S140, performing data regularization on the flight parameters, the time series of the battery status and the time series of the environmental data to obtain a sequence of flight parameter timing feature vectors, a sequence of battery status timing feature vectors and a sequence of environmental data timing feature vectors; The sequence of timing feature vectors, the sequence of the above-mentioned battery status timing feature vectors and the sequence of the above-mentioned environmental data timing feature vectors are arranged into a flight parameter timing feature matrix, a battery status timing feature matrix and an environmental data timing feature matrix; S150, respectively perform timing feature association on the above-mentioned flight parameter timing feature matrix, the above-mentioned battery status timing feature matrix and the above-mentioned environmental data timing feature matrix to obtain a flight parameter timing association feature matrix, a battery status timing association feature matrix and an environmental data timing association feature matrix; S160, perform feature fusion on the above-mentioned flight parameter timing association feature matrix, the above-mentioned battery status timing association feature matrix and the above-mentioned environmental data timing association feature matrix to obtain a UAV flight process timing feature diagram, and then perform global timing interdependence capture to obtain a UAV flight process global timing dependency feature diagram; and, S170, obtain a real-time flight log report based on the above-mentioned UAV flight process global timing dependency feature diagram.

[0060] Here, those skilled in the art can understand that the specific operations of each step in the report data processing method of the above-mentioned AI-based drone management platform have been referred to above. Figures 1 to 4 The description of the report data processing system of the AI-based drone management platform has been introduced in detail, and therefore, its repeated description will be omitted.

[0061] Although the subject matter has been described in language specific to structural features and / or methodological logical actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Instead, the specific features and actions described above are merely example forms of implementing the claims.

Claims

1. A report data processing system for an AI-based drone management platform, characterized in that: include: The UAV flight data acquisition module is used to collect the flight parameters, battery status and environmental data information of the UAV through the UAV sensor to obtain the time series of the flight parameters, the time series of the battery status and the time series of the environmental data; A data regularization module, used for regularizing the time series of the flight parameters, the time series of the battery status, and the time series of the environmental data to obtain a sequence of flight parameter timing input vectors, a sequence of battery status timing input vectors, and a sequence of environmental data timing input vectors; A timing feature extraction module, used to extract timing features from the sequence of the flight parameter timing input vector, the sequence of the battery status timing input vector, and the sequence of the environmental data timing input vector, respectively, to obtain a sequence of flight parameter timing feature vectors, a sequence of battery status timing feature vectors, and a sequence of environmental data timing feature vectors; A parameter arrangement module, used to arrange the sequence of the flight parameter timing feature vectors, the sequence of the battery status timing feature vectors and the sequence of the environment data timing feature vectors into a flight parameter timing feature matrix, a battery status timing feature matrix and an environment data timing feature matrix; A timing feature association module, used to perform timing feature association on the flight parameter timing feature matrix, the battery status timing feature matrix and the environmental data timing feature matrix to obtain a flight parameter timing association feature matrix, a battery status timing association feature matrix and an environmental data timing association feature matrix; A global timing interdependence capture module is used to perform feature fusion on the flight parameter timing correlation feature matrix, the battery status timing correlation feature matrix and the environmental data timing correlation feature matrix to obtain a UAV flight process timing feature diagram, and then perform global timing interdependence capture to obtain a UAV flight process global timing dependency feature diagram; The flight log report generation module is used to obtain a real-time flight log report based on the global time-series dependency feature diagram of the UAV flight process.

2. The report data processing system of the AI-based drone management platform according to claim 1 is characterized in that: The data regularization module is used to: The time series of the flight parameters, the time series of the battery status and the time series of the environmental data are regularized according to the time dimension and the sample dimension to obtain a sequence of the flight parameter timing input vectors, a sequence of the battery status timing input vectors and a sequence of the environmental data timing input vectors.

3. The report data processing system of the AI-based drone management platform according to claim 2 is characterized in that: The temporal feature extraction module is used to: The sequence of the flight parameter timing input vectors, the sequence of the battery status timing input vectors and the sequence of the environmental data timing input vectors are input into a timing feature extractor based on a deep temporal convolutional network to obtain a sequence of the flight parameter timing feature vectors, a sequence of the battery status timing feature vectors and a sequence of the environmental data timing feature vectors.

4. The report data processing system of the AI-based drone management platform according to claim 3 is characterized in that: The time series feature association module is used to: The flight parameter timing feature matrix, the battery status timing feature matrix and the environmental data timing feature matrix are respectively input into a timing feature association module based on a bidirectional attention mechanism to obtain the flight parameter timing association feature matrix, the battery status timing association feature matrix and the environmental data timing association feature matrix.

5. The report data processing system of the AI-based drone management platform according to claim 4 is characterized in that: The global timing interdependence capturing module comprises: A feature fusion unit, used for performing feature fusion along the channel dimension on the flight parameter time series correlation feature matrix, the battery status time series correlation feature matrix and the environmental data time series correlation feature matrix to obtain a time series feature graph of the UAV flight process; The global timing capture unit is used to input the timing feature diagram of the UAV flight process into a global timing interdependence module based on a non-local neural model to obtain a global timing dependency feature diagram of the UAV flight process.

6. The report data processing system of the AI-based drone management platform according to claim 5 is characterized in that: The flight log report generation module includes: A feature flattening unit, used for flattening the feature of the global time-series dependency feature graph of the UAV flight process to obtain a global time-series dependency feature vector of the UAV flight process; A feature optimization unit, used for performing feature optimization on the global timing dependency feature vector of the UAV flight process to obtain an optimized global timing dependency feature vector of the UAV flight process; The log report generating unit is used to input the global timing dependency feature vector of the optimized UAV flight process into the generator-based flight log report processor to obtain the real-time flight log report.

7. The report data processing system of the AI-based drone management platform according to claim 6 is characterized in that: The feature optimization unit is used to: The global timing dependency feature vector of the UAV flight process is subjected to backward compatibility perception based on the generated regression domain parameter space to obtain the global timing dependency feature vector of the optimized UAV flight process.

8. The report data processing system of the AI-based drone management platform according to claim 7 is characterized in that: The feature optimization unit is used to: Multiplying the global time-dependent feature vector of the UAV flight process by the generation weight matrix of the generator to obtain an intermediate feature vector; Splicing the global time-dependent feature vector of the UAV flight process with the intermediate feature vector to obtain a spliced ​​feature vector; Multiplying the concatenated feature vector by the first weight matrix and then adding the first bias vector to obtain a coded concatenated feature vector; Activating the encoded concatenated feature vector through a sigmoid function to obtain an activation value; Subtracting the activation value from one and then multiplying the intermediate feature vector by position to obtain a weighted intermediate feature vector; Multiplying the activation value by the global time-series dependency feature vector of the UAV flight process by position to obtain a weighted global time-series dependency feature vector of the UAV flight process; Subtracting the intermediate feature vector from the global time-dependent feature vector of the UAV flight process by position to obtain a difference feature vector; Adding the weighted intermediate feature vector and the weighted UAV flight process global time-series dependency feature vector and then dividing by the difference feature vector to obtain a second intermediate feature vector; The second intermediate feature vector is passed through the ReLU function to obtain the global timing dependency feature vector of the optimized UAV flight process.

9. A report data processing method for an AI-based drone management platform, characterized in that: include: The flight parameters, battery status and environmental data information of the drone are collected through the drone sensor to obtain a time series of the flight parameters, a time series of the battery status and a time series of the environmental data; Regularizing the time series of the flight parameters, the time series of the battery status, and the time series of the environmental data to obtain a sequence of flight parameter timing input vectors, a sequence of battery status timing input vectors, and a sequence of environmental data timing input vectors; Performing time series feature extraction on the sequence of the flight parameter time series input vectors, the sequence of the battery status time series input vectors, and the sequence of the environment data time series input vectors, respectively, to obtain a sequence of flight parameter time series feature vectors, a sequence of battery status time series feature vectors, and a sequence of environment data time series feature vectors; Arranging the sequence of the flight parameter timing feature vectors, the sequence of the battery status timing feature vectors and the sequence of the environment data timing feature vectors into a flight parameter timing feature matrix, a battery status timing feature matrix and an environment data timing feature matrix; Performing timing feature association on the flight parameter timing feature matrix, the battery status timing feature matrix, and the environmental data timing feature matrix to obtain a flight parameter timing association feature matrix, a battery status timing association feature matrix, and an environmental data timing association feature matrix; The flight parameter timing correlation feature matrix, the battery status timing correlation feature matrix and the environment data timing correlation feature matrix are subjected to feature fusion to obtain a UAV flight process timing feature diagram, and then global timing interdependence capture is performed to obtain a UAV flight process global timing dependency feature diagram; Based on the global time-series dependency feature graph of the UAV flight process, a real-time flight log report is obtained.

10. The report data processing method of the AI-based drone management platform according to claim 9 is characterized in that: The time series of the flight parameters, the time series of the battery status, and the time series of the environmental data are data regularized to obtain a sequence of flight parameter timing input vectors, a sequence of battery status timing input vectors, and a sequence of environmental data timing input vectors, including: The time series of the flight parameters, the time series of the battery status and the time series of the environmental data are regularized according to the time dimension and the sample dimension to obtain a sequence of the flight parameter timing input vectors, a sequence of the battery status timing input vectors and a sequence of the environmental data timing input vectors.

Images