Distributed multi-modal data information collaborative management method

Abstract

The invention discloses a distributed multi-mode data information collaborative management method, and relates to the technical field of distributed systems, and the method comprises the steps: setting an air collection end to carry out the information collection of an unmanned aerial vehicle, and obtaining multi-mode operation data; setting a modal memory layer to perform signal conversion on the multi-mode operation data to obtain an input multi-mode signal sequence, performing scale decomposition on a set initial change cardinal number to obtain a mother wave parameter, and performing convolution on the mother wave parameter and the input multi-mode signal sequence to obtain a multi-mode coefficient; filtering extraction is carried out on the obtained multi-mode coefficient to obtain a characteristic multi-mode coefficient, and down-sampling capture and sampling supplement are carried out on the obtained characteristic multi-mode coefficient to obtain a comprehensive multi-mode coefficient; generating a comprehensive multi-mode spectrogram according to the obtained comprehensive multi-mode coefficient, setting a cooperative detection point, and performing cooperative verification on the unmanned aerial vehicle at the cooperative detection point according to the comprehensive multi-mode spectrogram to obtain an abnormal node; the data processing efficiency is greatly improved, the system collaboration of the unmanned aerial vehicle is enhanced, and the working efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of distributed systems, and in particular to a distributed multi-modal data information collaborative management method. Background Art

[0002] A distributed system is a computer network architecture that uses multiple computing nodes interconnected through a network to work together to complete a task. A distributed system can achieve high availability and flexibility of tasks, and can effectively utilize network resources to improve task processing efficiency and performance.

[0003] Multimodal data information collaboration refers to combining data from different sources, different formats or different modes, and jointly processing and analyzing them to achieve richer, more comprehensive information understanding and more accurate data interpretation. UAVs generate a large amount of multimodal data, such as images, videos, sounds, etc., in tasks such as aerial photography, monitoring, and surveying, making distributed multimodal data collaboration have a wide range of application value in the field of UAVs; but at the same time, the application of multimodal data collaboration in the field of UAVs also faces challenges such as data fusion and privacy protection, and requires comprehensive consideration of factors such as data processing, communication, computing, and security; therefore, it is of great theoretical and practical significance to realize the intelligence and autonomy of UAV systems by collaboratively managing data from different sources and different modes.

[0004] How to use distributed system technology to build a modal memory layer to perform signal conversion on multi-mode operation data, obtain the input multi-mode signal sequence, and convolve it with the mother wave parameters after scale decomposition to obtain multi-mode coefficients; filter and extract the multi-mode coefficients, downsample and capture, and sample and supplement to obtain comprehensive multi-mode coefficients; generate a comprehensive multi-mode spectrum diagram based on the comprehensive multi-mode coefficients, set collaborative detection points, and perform collaborative verification of drones based on the comprehensive multi-mode spectrum diagram to obtain abnormal nodes. This is the problem we need to solve. To this end, a distributed multi-modal data information collaborative management method is now provided. Summary of the invention

[0005] The purpose of the present invention can be achieved by the following technical solutions:

[0006] A distributed multi-modal data information collaborative management method comprises the following steps:

[0007] Step S1: Setting an air collection terminal and an operation node, collecting information from the operation node of the UAV through the air collection terminal, and obtaining multi-mode operation data and air collection time;

[0008] Step S2: setting an idle sampling period to sort the multimode operation data to obtain a multimode state sequence, setting a modal memory layer according to the multimode state sequence, setting a conversion window, an action window and an output window in the modal memory layer, performing signal conversion on the multimode state sequence through the conversion window to obtain an input multimode signal sequence, setting a starting change base, performing scale decomposition and signal decomposition on the input multimode signal sequence and the starting change base, obtaining a multimode sequence segment and a mother wave parameter, and uploading the multimode sequence segment and the mother wave parameter to the action window for convolution combination to obtain a multimode coefficient;

[0009] Step S3: uploading the obtained multimode coefficients to the output window for filtering and extraction to obtain characteristic multimode coefficients, performing downsampling capture and sampling supplement on the obtained characteristic multimode coefficients to obtain comprehensive multimode coefficients;

[0010] Step S4: Generate a comprehensive multimode spectrum diagram according to the obtained comprehensive multimode coefficients, set collaborative detection points, and perform collaborative verification on the drones at the collaborative detection points according to the comprehensive multimode spectrum diagram to obtain abnormal nodes.

[0011] The process of obtaining the input multimode signal sequence includes:

[0012] Setting an idle mining cycle according to the obtained idle mining time, sorting the multi-mode operation data according to the set idle mining cycle, obtaining a multi-mode state sequence, and associating the obtained multi-mode state sequence with a corresponding operation node;

[0013] A modal memory layer is set according to the obtained multi-modal state sequence, and a conversion window, an action window and an output window are set in the modal memory layer;

[0014] The obtained multi-mode state sequence is marked as an input state sequence, and the obtained input state sequence is uploaded to a conversion window, and the obtained input state sequence is converted into an electrical signal to obtain an input multi-mode signal sequence.

[0015] The process of obtaining the multimode sequence segments and mother wave parameters includes:

[0016] Setting the starting variation base number according to the obtained empty sampling period and the input multi-mode signal sequence;

[0017] Setting a transformation basis, and setting the number of transformations according to the obtained transformation basis and the multimode signal sequence;

[0018] Decompose the initial change base according to the obtained transformation base to obtain the mother wave parameters;

[0019] The input multimode signal sequence is decomposed according to the obtained number of transformations to obtain multimode sequence segments.

[0020] The process of obtaining the multimodal coefficients includes:

[0021] Upload the obtained multi-mode sequence segments and mother wave parameters to the action window;

[0022] Performing piecewise convolution on the obtained multimode sequence segments and the mother wave parameters to obtain multimode piecewise coefficients;

[0023] The multimode segment coefficients are combined according to the obtained input multimode signal sequence to obtain multimode coefficients.

[0024] The process of obtaining characteristic multimodal coefficients includes:

[0025] Upload the obtained multi-mode coefficients to the output window;

[0026] Setting a first filtering endpoint and a second filtering endpoint in the output window, performing low-pass filtering extraction on the multi-mode coefficients through the first filtering endpoint to obtain segmented multi-mode coefficients;

[0027] The segmented multimode coefficients are subjected to high-pass filtering extraction through the second filtering endpoint to obtain characteristic multimode coefficients.

[0028] The process of obtaining the comprehensive multimode coefficients includes:

[0029] Set a sampling factor, and perform downsampling capture on the characteristic multi-mode coefficient according to the obtained sampling factor to obtain the captured multi-mode coefficient;

[0030] A target sampling factor is set, a preset sampling point is set according to the obtained captured multi-mode coefficient, and the characteristic multi-mode coefficient is sampled and supplemented according to the obtained preset sampling point and the target sampling factor to obtain a comprehensive multi-mode coefficient.

[0031] At the collaborative detection point, the drone is collaboratively verified based on the comprehensive multi-mode spectrum diagram, and the process of obtaining abnormal nodes includes:

[0032] generating a comprehensive multimode spectrum diagram according to the obtained comprehensive multimode coefficients;

[0033] Uploading the obtained integrated multi-mode spectrum graph to the operation node associated with the corresponding multi-mode state sequence;

[0034] Set up collaborative detection points based on running nodes;

[0035] Set a shutdown command, execute the shutdown command through the coordinated detection point, collect information from the remaining operating nodes through the empty collection terminal, and mark the integrated multi-mode spectrum graph corresponding to the collected multi-mode operating data as a comprehensive change spectrum graph;

[0036] At the running node, the obtained integrated multi-mode spectrum graph is overlapped and matched with the integrated change spectrum graph to obtain a non-overlapping spectrum graph, and the running node corresponding to the non-overlapping spectrum graph is marked as an abnormal node.

[0037] Compared with the prior art, the present invention has the following beneficial effects: the modal memory layer includes a conversion window, an action window and an output window, and the multi-mode operation data are processed in the conversion window, the action window and the output window respectively, and the comprehensive multi-mode coefficient is finally obtained, which is conducive to systematically processing the collected multi-mode operation data, and processing the multi-mode operation data distributed on different drones into the comprehensive multi-mode coefficient in a unified data form, which is convenient for the subsequent collaborative management of the multi-mode data of the drone;

[0038] A comprehensive multimode spectrum diagram is generated according to the comprehensive multimode coefficients, and the UAV is collaboratively checked according to the comprehensive multimode spectrum diagram at the collaborative detection point to obtain abnormal nodes; the visualization effect of multimode operation data is improved to facilitate collaborative detection of nodes, increase data processing speed, and improve the task processing efficiency of UAVs. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 It is a schematic diagram of the present invention. DETAILED DESCRIPTION

[0041] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] like Figure 1 As shown, a distributed multimodal data information collaborative management method includes the following steps:

[0043] Step S1: Setting an air collection terminal and an operation node, collecting information from the operation node of the UAV through the air collection terminal, and obtaining multi-mode operation data and air collection time;

[0044] Step S2: setting an idle sampling period to sort the multimode operation data to obtain a multimode state sequence, setting a modal memory layer according to the multimode state sequence, setting a conversion window, an action window and an output window in the modal memory layer, performing signal conversion on the multimode state sequence through the conversion window to obtain an input multimode signal sequence, setting a starting change base, performing scale decomposition and signal decomposition on the input multimode signal sequence and the starting change base, obtaining a multimode sequence segment and a mother wave parameter, and uploading the multimode sequence segment and the mother wave parameter to the action window for convolution combination to obtain a multimode coefficient;

[0045] Step S3: uploading the obtained multimode coefficients to the output window for filtering and extraction to obtain characteristic multimode coefficients, performing downsampling capture and sampling supplement on the obtained characteristic multimode coefficients to obtain comprehensive multimode coefficients;

[0046] Step S4: Generate a comprehensive multimode spectrum diagram according to the obtained comprehensive multimode coefficients, set collaborative detection points, and perform collaborative verification on the drones at the collaborative detection points according to the comprehensive multimode spectrum diagram to obtain abnormal nodes.

[0047] It should be further explained that, in the specific implementation process, the process of collecting multi-mode operation data of drones includes:

[0048] Set up the air collection terminal and running node;

[0049] The air collection terminal is connected to different sensors and devices of the drone respectively. Accordingly, the data information collected by different air collection terminals are from sensors and devices of different drones. The operation node is a node in the drone that collects data through the air collection terminal, and the set operation node includes all locations in the drone that can generate data;

[0050] Mark the running node as j, where j=1, 2, 3, ..., v1, v1 is a positive integer;

[0051] The air acquisition terminal collects information from the UAV at the operation node to obtain multi-mode operation data, and marks the time of collecting the multi-mode operation data as the air acquisition time;

[0052] The obtained multi-mode operation data is marked as ;

[0053] It should be further explained that, in the specific implementation process, the multi-mode operation data includes image and video data, sensor data, position and speed data, meteorological data, audio data and remote sensing data.

[0054] Set the empty mining cycle according to the obtained empty mining time;

[0055] Based on the idle mining cycle, the multi-mode operation data is sorted according to the time sequence of the obtained idle mining time to obtain a multi-mode state sequence, and the obtained multi-mode state sequence is associated with the corresponding operation node;

[0056] A modal memory layer is set according to the obtained multi-modal state sequence, and a conversion window, an action window and an output window are set in the modal memory layer;

[0057] The conversion window is used to perform signal conversion on the multimode state sequence to obtain an input multimode signal sequence;

[0058] The action window is used to perform scale decomposition and signal decomposition on the input multimode signal sequence, obtain multimode sequence segments and mother wave parameters, and perform convolution combination to obtain multimode coefficients;

[0059] The output window is used to filter and extract the multimode coefficients to obtain characteristic multimode coefficients, and to downsample and capture the obtained characteristic multimode coefficients and perform sampling supplementation to obtain comprehensive multimode coefficients;

[0060] Mark the obtained multi-mode state sequence as the input state sequence;

[0061] The conversion window performs signal conversion on the multimode state sequence to obtain an input multimode signal sequence, which includes:

[0062] Uploading the obtained input state sequence to the conversion window;

[0063] Performing electrical signal conversion on the obtained input state sequence to obtain an input multi-mode signal sequence, and marking signal sampling points of the input multi-mode signal sequence;

[0064] It should be further explained that, in the specific implementation process, according to the image and video data, sensor data, position and speed data, meteorological data, audio data and remote sensing data included in the multi-mode operation data, the converted input multi-mode signal sequence also includes an image and video data signal sequence, a sensor data signal sequence, a position and speed data signal sequence, a meteorological data signal sequence, an audio data signal sequence and a remote sensing data signal sequence;

[0065] The starting variation base is set according to the obtained idle sampling period and the input multi-mode signal sequence; further, the starting variation base is expressed in a function form;

[0066] Setting a transformation base, wherein the transformation base is used to control the translation and expansion of the starting variation base, that is, to control the translation and expansion of the starting variation base in the time component and the frequency component;

[0067] Setting the number of transformations according to the obtained transformation base and the multimode signal sequence;

[0068] Furthermore, the number of transformations is obtained according to the length of the multimode signal sequence and the transformation base, that is, the number of transformations = ;

[0069] Decompose the initial change base according to the obtained transformation base to obtain the mother wave parameters;

[0070] It should be further explained that, in the specific implementation process, the scale transformation is to perform scale transformation and translation operations on the initial change base according to the set transformation base, that is, to perform scale transformation and translation operations on the time component and the frequency component to obtain mother wave parameters of different scales and positions;

[0071] Decomposing the input multimode signal sequence according to the obtained number of transformations to obtain multimode sequence segments;

[0072] Furthermore, the signal decomposition is to equally decompose the multimode signal sequence according to the number of transformations, so that the length of each multimode sequence segment is equal;

[0073] The obtained multimode sequence segments and mother wave parameters are uploaded to the action window; the action window is used to perform scale decomposition and signal decomposition on the input multimode signal sequence, obtain the multimode sequence segments and mother wave parameters and perform convolution combination to obtain the multimode coefficients. The specific process includes:

[0074] Performing piecewise convolution on the obtained multimode sequence segments and the mother wave parameters to obtain multimode piecewise coefficients;

[0075] Combining the multimode segment coefficients according to the obtained input multimode signal sequence to obtain multimode coefficients;

[0076] It should be further explained that, in the specific implementation process, the segmented convolution is to multiply the multimode sequence segments with the elements in the mother wave parameters one by one and then perform convolution to obtain multimode segmented coefficients, and combine the multimode segmented coefficients according to the decomposition order when the input multimode signal sequence is decomposed to obtain multimode coefficients; in particular, the multimode coefficients contain the amplitude and phase information of the input multimode signal sequence at each frequency component, which is used to extract the characteristics of the signal.

[0077] The obtained multimode coefficients are uploaded to an output window; the output window is used to filter and extract the multimode coefficients to obtain characteristic multimode coefficients, and downsample and capture and sample the obtained characteristic multimode coefficients to obtain comprehensive multimode coefficients. The specific process includes:

[0078] Setting a first filtering endpoint and a second filtering endpoint in the output window, performing low-pass filtering extraction on the multi-mode coefficients through the first filtering endpoint to obtain segmented multi-mode coefficients;

[0079] Performing high-pass filtering on the segmented multimode coefficients through the second filtering endpoint to obtain characteristic multimode coefficients;

[0080] It should be further explained that, in the specific implementation process, in this embodiment, the first filtering endpoint is a low-pass filter, and the second filtering endpoint is a high-pass filter. Furthermore, the type of the low-pass filter is determined according to the frequency range, signal components, signal bandwidth, noise level and signal resolution of the input multimode signal sequence, and the cutoff frequency is set according to the obtained multimode coefficient to determine the frequency at which the low-pass filter starts and stops filtering; similarly, the type of the high-pass filter at the second filtering endpoint is determined by the frequency range, signal components, signal bandwidth, noise level and signal resolution of the input multimode signal sequence, and the cutoff frequency is set;

[0081] Set a sampling factor, where the sampling factor means removing one signal sampling point every a signal sampling points. For example, if the sampling factor is 3, one signal sampling point is removed every 3 signal sampling points.

[0082] Down-sampling and capturing the characteristic multi-mode coefficients according to the obtained sampling factors to obtain captured multi-mode coefficients;

[0083] Furthermore, the downsampling capture is to remove the signal sampling points of the characteristic multi-mode coefficients according to the sampling factor to obtain the captured multi-mode coefficients with reduced signal sampling points;

[0084] Set the target sampling factor, which depends on the original sampling rate of the characteristic multi-mode coefficients and the target sampling rate that the multi-mode coefficients are to be restored. The target sampling factor = ;

[0085] Furthermore, the target sampling rate is set before downsampling capture;

[0086] A preset sampling point is set according to the obtained captured multi-mode coefficient, and the characteristic multi-mode coefficient is sampled and supplemented according to the obtained preset sampling point and the target sampling factor to obtain a comprehensive multi-mode coefficient;

[0087] Furthermore, the sampling supplementation is to insert the target sampling factor into the preset sampling point to supplement the missing signal sampling points after downsampling capture, so as to obtain the comprehensive multi-mode coefficient;

[0088] It needs to be further explained that, in the specific implementation process, after downsampling capture and sampling supplementation, the characteristic multimode coefficients will be aliased. It is necessary to set an anti-aliasing filter for the comprehensive multimode coefficients and filter the comprehensive multimode coefficients through the anti-aliasing filter to reduce the impact of aliasing.

[0089] generating a comprehensive multimode spectrum diagram according to the obtained comprehensive multimode coefficients;

[0090] Uploading the obtained integrated multi-mode spectrum graph to the operation node associated with the corresponding multi-mode state sequence;

[0091] Select any running node in the drone and mark the selected running node as a collaborative detection point;

[0092] Perform collaborative verification on drones at collaborative detection points to obtain abnormal nodes;

[0093] The process of collaboratively verifying the UAV at the collaborative detection point includes:

[0094] Set a shutdown command, execute the shutdown command through the coordinated detection point, and collect information from the remaining operating nodes through the empty collection terminal, mark the collected multi-mode operating data as variable multi-mode data, and mark the comprehensive multi-mode spectrum diagram obtained according to the variable multi-mode data as a comprehensive variable spectrum diagram;

[0095] Furthermore, the shutdown instruction indicates stopping all work at the coordinated detection point, and ensures that only the coordinated detection point is in a shutdown state, so that the air collection terminal cannot collect any data information at the coordinated detection point;

[0096] At the running node, the obtained integrated multi-mode spectrum graph is overlapped and matched with the integrated change spectrum graph to obtain a non-overlapping spectrum graph, and the running node corresponding to the non-overlapping spectrum graph is marked as an abnormal node;

[0097] It should be further explained that, in the specific implementation process, the overlap matching is to compare the original spectrum diagram of the same operating node, that is, the comprehensive multi-mode spectrum diagram, with the changed spectrum diagram after the shutdown instruction is executed, that is, the comprehensive changed spectrum diagram, and map the obtained comprehensive multi-mode spectrum diagram and the comprehensive changed spectrum diagram to the same spectrum diagram for comparison. If the two spectrum curves completely overlap in the same spectrum diagram, the match is successful, indicating that any one of the operating nodes of the drone is not working, which does not affect the working effect of the remaining operating nodes; if the two spectrum curves do not completely overlap in the same spectrum diagram, the match is unsuccessful, indicating that any one of the operating nodes of the drone is not working, which will affect the working effect of the remaining operating nodes, and the operating node that failed to match is marked as an abnormal node;

[0098] Select any running node from the remaining running nodes and mark it as a collaborative detection point, repeat the process of obtaining abnormal nodes until all running nodes are selected as collaborative detection points, and perform collaborative repair on the obtained abnormal nodes;

[0099] It needs to be further explained that, in the specific implementation process, collaborative repair means that it is necessary to notify the drone management personnel to perform node repair management on the node, repair or adjust the operating status of the abnormal node, and perform collaborative verification on the operating node after the collaborative repair to detect whether the operating node after the collaborative repair is a non-overlapping spectrum graph. If not, it means that the repair is successful. If it is still a non-overlapping spectrum graph, continue to repair until there is no non-overlapping spectrum graph.

[0100] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A distributed multimodal data information collaborative management method, characterized in that: The following steps are involved: Step S1: Setting an air collection terminal and an operation node, collecting information from the operation node of the UAV through the air collection terminal, and obtaining multi-mode operation data and air collection time; Step S2: setting an idle sampling period to sort the multimode operation data to obtain a multimode state sequence, setting a modal memory layer according to the multimode state sequence, setting a conversion window, an action window and an output window in the modal memory layer, performing signal conversion on the multimode state sequence through the conversion window to obtain an input multimode signal sequence, setting a starting change base, performing scale decomposition and signal decomposition on the input multimode signal sequence and the starting change base, obtaining a multimode sequence segment and a mother wave parameter, and uploading the multimode sequence segment and the mother wave parameter to the action window for convolution combination to obtain a multimode coefficient; Step S3: uploading the obtained multimode coefficients to the output window for filtering and extraction to obtain characteristic multimode coefficients, performing downsampling capture and sampling supplement on the obtained characteristic multimode coefficients to obtain comprehensive multimode coefficients; Step S4: Generate a comprehensive multimode spectrum diagram according to the obtained comprehensive multimode coefficients, set collaborative detection points, and perform collaborative verification on the drones at the collaborative detection points according to the comprehensive multimode spectrum diagram to obtain abnormal nodes.

2. The method for collaborative management of distributed multimodal data information according to claim 1, characterized in that: The process of obtaining the input multimode signal sequence includes: Setting an idle mining cycle according to the obtained idle mining time, sorting the multi-mode operation data according to the set idle mining cycle, obtaining a multi-mode state sequence, and associating the obtained multi-mode state sequence with a corresponding operation node; A modal memory layer is set according to the obtained multi-modal state sequence, and a conversion window, an action window and an output window are set in the modal memory layer; The obtained multi-mode state sequence is marked as an input state sequence, and the obtained input state sequence is uploaded to a conversion window, and the obtained input state sequence is converted into an electrical signal to obtain an input multi-mode signal sequence.

3. The distributed multimodal data information collaborative management method according to claim 2 is characterized in that: The process of obtaining the multimode sequence segments and mother wave parameters includes: Setting the starting variation base number according to the obtained empty sampling period and the input multi-mode signal sequence; Setting a transformation basis, and setting the number of transformations according to the obtained transformation basis and the multimode signal sequence; Decompose the initial change base according to the obtained transformation base to obtain the mother wave parameters; The input multimode signal sequence is decomposed according to the obtained number of transformations to obtain multimode sequence segments.

4. The distributed multimodal data information collaborative management method according to claim 3 is characterized in that: The process of obtaining the multimodal coefficients includes: Upload the obtained multi-mode sequence segments and mother wave parameters to the action window; Performing piecewise convolution on the obtained multimode sequence segments and the mother wave parameters to obtain multimode piecewise coefficients; The multimode segment coefficients are combined according to the obtained input multimode signal sequence to obtain multimode coefficients.

5. The distributed multimodal data information collaborative management method according to claim 4 is characterized in that: The process of obtaining characteristic multimodal coefficients includes: Upload the obtained multi-mode coefficients to the output window; Setting a first filtering endpoint and a second filtering endpoint in the output window, performing low-pass filtering extraction on the multi-mode coefficients through the first filtering endpoint to obtain segmented multi-mode coefficients; The segmented multimode coefficients are subjected to high-pass filtering extraction through the second filtering endpoint to obtain characteristic multimode coefficients.

6. A distributed multimodal data information collaborative management method according to claim 5, characterized in that: The process of obtaining the comprehensive multimode coefficients includes: Set a sampling factor, and perform downsampling capture on the characteristic multi-mode coefficient according to the obtained sampling factor to obtain the captured multi-mode coefficient; A target sampling factor is set, a preset sampling point is set according to the obtained captured multi-mode coefficient, and the characteristic multi-mode coefficient is sampled and supplemented according to the obtained preset sampling point and the target sampling factor to obtain a comprehensive multi-mode coefficient.

7. A distributed multimodal data information collaborative management method according to claim 6, characterized in that: At the collaborative detection point, the drone is collaboratively verified based on the comprehensive multi-mode spectrum diagram, and the process of obtaining abnormal nodes includes: generating a comprehensive multimode spectrum diagram according to the obtained comprehensive multimode coefficients; Uploading the obtained integrated multi-mode spectrum graph to the operation node associated with the corresponding multi-mode state sequence; Set up collaborative detection points based on running nodes; Set a shutdown command, execute the shutdown command through the coordinated detection point, collect information from the remaining operating nodes through the empty collection terminal, and mark the integrated multi-mode spectrum graph corresponding to the collected multi-mode operating data as a comprehensive change spectrum graph; At the running node, the obtained integrated multi-mode spectrum graph is overlapped and matched with the integrated change spectrum graph to obtain a non-overlapping spectrum graph, and the running node corresponding to the non-overlapping spectrum graph is marked as an abnormal node.

Images