Multi-dimensional hydrological data fusion and analysis method and device based on intelligent hydrological rod

Abstract

The invention discloses a multi-dimensional hydrological data fusion and analysis method and device based on an intelligent hydrological rod, and the method and device can synchronously obtain multi-dimensional hydrological detection data, such as water level height, water velocity, water temperature and water quality, through a hydrological sensor and a video collection device which are disposed on the intelligent hydrological rod. Visual video monitoring information such as river surface fluctuation, river bank change and river channel floating objects can be monitored; efficient data fusion and cross validation can be performed on the data based on an artificial intelligence algorithm, so that the accuracy and integrity of the data can be improved; in addition, deep analysis can be carried out on the data based on an artificial intelligence algorithm, and potential hydrological change rules and hidden risks are mined; besides, based on the data processing result, an accurate early warning, maintenance or emergency event processing decision can be generated through a preset decision coping algorithm, and the emergency response capability of hydrological monitoring is improved while the data utilization rate of the data processing result is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent data processing, and in particular to a multi-dimensional hydrological data fusion and analysis method and device based on a smart hydrological pole. Background Art

[0002] In the traditional hydrological monitoring system, the collection of hydrological data mainly relies on dispersed water level meters, flow meters, water quality monitoring stations and other equipment. These data often exist in isolation and lack a unified integration and analysis platform, resulting in serious information islands and difficulty in forming a comprehensive understanding of the hydrological environment. At the same time, although the video surveillance system can provide intuitive on-site images, its application value has not been fully explored due to the lack of effective association with hydrological data. In addition, after collecting the hydrological data and video data, the data processing and analysis methods set for these data are relatively backward, mostly relying on manual judgment or simple statistical analysis, and the corresponding data analysis efficiency is low and the accuracy of data analysis is not high.

[0003] At the same time, with the rapid development of technologies such as the Internet of Things, big data, and artificial intelligence, the field of hydrological monitoring is gradually transforming towards intelligence and integration. However, the existing intelligent hydrological monitoring system has a low utilization rate of data and focuses on special data monitoring, which makes it difficult to fully reflect the complexity and dynamics of the hydrological environment. Summary of the invention

[0004] The present invention provides a multi-dimensional hydrological data fusion and analysis method and device based on a smart hydrological pole, which can be beneficial to improving the intelligence level and emergency response capability of hydrological monitoring, while improving the thoroughness and accuracy of hydrological data analysis.

[0005] In order to solve the above technical problems, the first aspect of the present invention discloses a multi-dimensional hydrological data fusion and analysis method based on a smart hydrological pole, the method comprising:

[0006] Acquire hydrological detection data collected by a hydrological sensor, wherein the hydrological sensor is configured on the smart hydrological pole, and the hydrological detection data includes at least one of the following data: water level height, water flow velocity, water temperature, and water quality of a river;

[0007] Acquire video data collected by a video acquisition device for the river channel, and determine river channel monitoring information based on the video data, wherein the river channel monitoring information includes at least one of river surface fluctuation information, river bank change information, and increase or decrease information of floating objects in the river channel;

[0008] Determine the hydrological detection data and the video data as target analysis data, and perform target data processing on the target analysis data according to a preset artificial intelligence algorithm to obtain a data processing result corresponding to the target analysis data; the target data processing at least includes data fusion, data cross-validation and in-depth analysis;

[0009] According to a preset decision response algorithm, a response decision is generated for the data processing result; the response decision includes at least one type of decision from among early warning decisions, river maintenance decisions, and emergency event decisions.

[0010] As an optional implementation manner, in the first aspect of the present invention, performing target data processing on the target analysis data according to a preset artificial intelligence algorithm to obtain a data processing result corresponding to the target analysis data includes:

[0011] Performing data preprocessing on the target analysis data to obtain a data preprocessing result corresponding to the target analysis data, wherein the data preprocessing includes at least one of data cleaning, time dimension synchronization, and format conversion;

[0012] According to a preset data fusion strategy, data fusion is performed on the data preprocessing result to obtain a data fusion result corresponding to the data preprocessing result, wherein the data fusion includes feature extraction and feature fusion based on a multi-source data fusion algorithm;

[0013] Performing cross validation on the data fusion result to obtain a cross validation result corresponding to the data fusion result;

[0014] After determining that the cross-validation result indicates that the validation is passed, a deep analysis is performed on the data fusion result according to a preset deep learning model to obtain a deep analysis result corresponding to the data fusion result as a data processing result corresponding to the target analysis data.

[0015] As an optional implementation, in the first aspect of the present invention, performing data fusion on the data preprocessing result according to a preset data fusion strategy to obtain a data fusion result corresponding to the data preprocessing result includes:

[0016] Determine a plurality of first characteristic parameters corresponding to the hydrological detection data and a plurality of second characteristic parameters corresponding to the video data; the plurality of first characteristic parameters include at least two of the water level change rate, the peak value of the water flow velocity, and the change trend of the water quality index; the plurality of second characteristic parameters include at least two of the river surface fluctuation characteristics, the signs of river bank erosion, and the types and quantities of floating objects;

[0017] Extracting first feature data matching each of the first feature parameters from the hydrological detection data in the data preprocessing result, and extracting second feature data matching each of the second feature parameters from the video data in the data preprocessing result;

[0018] According to a preset data fusion algorithm, feature fusion is performed on all the first feature data and all the second feature data to obtain hydrological-video feature fusion data, and the hydrological-video feature fusion data includes a feature fusion vector.

[0019] As an optional implementation, in the first aspect of the present invention, when the data preprocessing includes the data cleaning, the data cleaning includes denoising and deduplication processing for the hydrological detection data and frame rate adjustment, de-jittering, and color correction for the video data;

[0020] The time dimension synchronization is used to perform time dimension synchronization on the hydrological detection data and / or the video data;

[0021] The format conversion is used to convert the hydrological detection data and / or the video data into an image sequence or feature vector adapted to machine learning requirements.

[0022] As an optional implementation manner, in the first aspect of the present invention, performing cross-validation on the data fusion result to obtain a cross-validation result corresponding to the data fusion result includes:

[0023] Acquire historical analysis data corresponding to the target analysis data, wherein the historical verification data is obtained by sequentially performing the data preprocessing and data fusion on the historical hydrological detection data corresponding to the hydrological detection data and the historical video data corresponding to the video data;

[0024] Performing cross validation on the historical analysis data and the data fusion result to obtain a cross validation result corresponding to the historical analysis data and the data fusion result, wherein the cross validation includes at least one of difference calculation, correlation coefficient calculation, and cluster analysis;

[0025] Among them, the difference calculation includes at least one of mean calculation, variance calculation, standard deviation calculation, and cross entropy loss calculation; the correlation coefficient calculation is used to determine the data correlation between the historical analysis data and the data fusion result; the cluster analysis is used to characterize the fusion effect of the historical analysis data and the data fusion result.

[0026] As an optional implementation manner, in the first aspect of the present invention, performing cross-validation on the data fusion result to obtain a cross-validation result corresponding to the data fusion result further includes:

[0027] Before acquiring the historical analysis data corresponding to the target analysis data, performing integrity check and outlier detection on the data fusion result to obtain a review result corresponding to the data fusion result; the integrity check is used to detect whether there is data loss or omission in the data fusion result; the outlier detection detects outliers in the data fusion result by using a statistical algorithm or a machine learning algorithm;

[0028] When the review result indicates that the data fusion result contains unqualified data that does not meet expectations, the previously executed operation is backtracked and located according to the unqualified data to obtain a location node to be corrected for the previously executed operation; the previously executed operation at least includes the data preprocessing and the data fusion;

[0029] Execute node correction on the positioning node and re-execute the node operation corresponding to the positioning node; and trigger execution of the operation corresponding to obtaining the historical analysis data corresponding to the target analysis data;

[0030] Among them, when the positioning node is located in the operation process of the data preprocessing, the node correction includes at least one of data re-cleaning, re-synchronization of the time dimension, and format verification; when the positioning node is located in the operation process of the data fusion, the node correction includes at least one of feature extraction algorithm adjustment, data fusion algorithm optimization or feature parameter adjustment.

[0031] As an optional implementation, in the first aspect of the present invention, the preset decision response algorithm includes a classification algorithm and a regression algorithm; the generating a response decision for the data processing result according to the preset decision response algorithm includes:

[0032] Acquire abnormal record information set for the river channel, and determine whether the data processing result indicates the existence of abnormal matters according to the decision-making response algorithm and the abnormal record information;

[0033] When it is determined that the data processing result indicates the presence of the abnormal event, determining the abnormal type of the abnormal event according to the classification algorithm;

[0034] Analyze the abnormal events and the abnormal types according to the regression algorithm to obtain the predicted change trend of the abnormal events;

[0035] According to the strategy library of the decision response algorithm, combined with the abnormal event, the abnormal type and the predicted change trend, a response decision for the abnormal event is generated.

[0036] The second aspect of the present invention discloses a multi-dimensional hydrological data fusion and analysis device based on a smart hydrological pole, the device comprising:

[0037] A data acquisition module, used to acquire hydrological detection data collected by a hydrological sensor, wherein the hydrological sensor is arranged on the smart hydrological pole, and the hydrological detection data includes at least one of the following data: water level height, water flow velocity, water temperature, and water quality of a river;

[0038] The data acquisition module is further used to acquire video data collected by a video acquisition device for the river channel;

[0039] A determination module, configured to determine river channel monitoring information according to the video data, wherein the river channel monitoring information includes at least one of river surface fluctuation information, river bank change information, and increase or decrease information of floating objects in the river channel;

[0040] The determination module is further used to determine the hydrological detection data and the video data as target analysis data;

[0041] A data processing module, used to perform target data processing on the target analysis data according to a preset artificial intelligence algorithm to obtain a data processing result corresponding to the target analysis data; the target data processing at least includes data fusion, data cross-validation and in-depth analysis;

[0042] The decision generation module is used to generate a response decision for the data processing result according to a preset decision response algorithm; the response decision includes at least one type of decision among early warning decisions, river maintenance decisions, and emergency event decisions.

[0043] As an optional implementation, in the second aspect of the present invention, the data processing module performs target data processing on the target analysis data according to a preset artificial intelligence algorithm, and a method of obtaining a data processing result corresponding to the target analysis data specifically includes:

[0044] Performing data preprocessing on the target analysis data to obtain a data preprocessing result corresponding to the target analysis data, wherein the data preprocessing includes at least one of data cleaning, time dimension synchronization, and format conversion;

[0045] According to a preset data fusion strategy, data fusion is performed on the data preprocessing result to obtain a data fusion result corresponding to the data preprocessing result, wherein the data fusion includes feature extraction and feature fusion based on a multi-source data fusion algorithm;

[0046] Performing cross validation on the data fusion result to obtain a cross validation result corresponding to the data fusion result;

[0047] After determining that the cross-validation result indicates that the validation is passed, a deep analysis is performed on the data fusion result according to a preset deep learning model to obtain a deep analysis result corresponding to the data fusion result as a data processing result corresponding to the target analysis data.

[0048] As an optional implementation, in the second aspect of the present invention, the data processing module performs data fusion on the data preprocessing result according to a preset data fusion strategy, and a method of obtaining a data fusion result corresponding to the data preprocessing result specifically includes:

[0049] Determine a plurality of first characteristic parameters corresponding to the hydrological detection data and a plurality of second characteristic parameters corresponding to the video data; the plurality of first characteristic parameters include at least two of the water level change rate, the peak value of the water flow velocity, and the change trend of the water quality index; the plurality of second characteristic parameters include at least two of the river surface fluctuation characteristics, the signs of river bank erosion, and the types and quantities of floating objects;

[0050] Extracting first feature data matching each of the first feature parameters from the hydrological detection data in the data preprocessing result, and extracting second feature data matching each of the second feature parameters from the video data in the data preprocessing result;

[0051] According to a preset data fusion algorithm, feature fusion is performed on all the first feature data and all the second feature data to obtain hydrological-video feature fusion data, and the hydrological-video feature fusion data includes a feature fusion vector.

[0052] As an optional implementation, in the second aspect of the present invention, when the data preprocessing includes the data cleaning, the data cleaning includes denoising and deduplication processing for the hydrological detection data and frame rate adjustment, de-jittering, and color correction for the video data;

[0053] The time dimension synchronization is used to perform time dimension synchronization on the hydrological detection data and / or the video data;

[0054] The format conversion is used to convert the hydrological detection data and / or the video data into an image sequence or feature vector adapted to machine learning requirements.

[0055] As an optional implementation, in the second aspect of the present invention, the data processing module performs cross-validation on the data fusion result, and a method of obtaining a cross-validation result corresponding to the data fusion result specifically includes:

[0056] Acquire historical analysis data corresponding to the target analysis data, wherein the historical verification data is obtained by sequentially performing the data preprocessing and data fusion on the historical hydrological detection data corresponding to the hydrological detection data and the historical video data corresponding to the video data;

[0057] Performing cross validation on the historical analysis data and the data fusion result to obtain a cross validation result corresponding to the historical analysis data and the data fusion result, wherein the cross validation includes at least one of difference calculation, correlation coefficient calculation, and cluster analysis;

[0058] Among them, the difference calculation includes at least one of mean calculation, variance calculation, standard deviation calculation, and cross entropy loss calculation; the correlation coefficient calculation is used to determine the data correlation between the historical analysis data and the data fusion result; the cluster analysis is used to characterize the fusion effect of the historical analysis data and the data fusion result.

[0059] As an optional implementation, in the second aspect of the present invention, the data processing module performs cross-validation on the data fusion result, and a method of obtaining a cross-validation result corresponding to the data fusion result specifically includes:

[0060] Before acquiring the historical analysis data corresponding to the target analysis data, performing integrity check and outlier detection on the data fusion result to obtain a review result corresponding to the data fusion result; the integrity check is used to detect whether there is data loss or omission in the data fusion result; the outlier detection detects outliers in the data fusion result by using a statistical algorithm or a machine learning algorithm;

[0061] When the review result indicates that the data fusion result contains unqualified data that does not meet expectations, the previously executed operation is backtracked and located according to the unqualified data to obtain a location node to be corrected for the previously executed operation; the previously executed operation at least includes the data preprocessing and the data fusion;

[0062] Execute node correction on the positioning node and re-execute the node operation corresponding to the positioning node; and trigger execution of the operation corresponding to obtaining the historical analysis data corresponding to the target analysis data;

[0063] Among them, when the positioning node is located in the operation process of the data preprocessing, the node correction includes at least one of data re-cleaning, re-synchronization of the time dimension, and format verification; when the positioning node is located in the operation process of the data fusion, the node correction includes at least one of feature extraction algorithm adjustment, data fusion algorithm optimization or feature parameter adjustment.

[0064] As an optional implementation, in the second aspect of the present invention, the preset decision response algorithm includes a classification algorithm and a regression algorithm; the decision generation module generates a response decision for the data processing result according to the preset decision response algorithm, specifically including:

[0065] Acquire abnormal record information set for the river channel, and determine whether the data processing result indicates the existence of abnormal matters according to the decision-making response algorithm and the abnormal record information;

[0066] When it is determined that the data processing result indicates the presence of the abnormal event, determining the abnormal type of the abnormal event according to the classification algorithm;

[0067] Analyze the abnormal events and the abnormal types according to the regression algorithm to obtain the predicted change trend of the abnormal events;

[0068] According to the strategy library of the decision response algorithm, combined with the abnormal event, the abnormal type and the predicted change trend, a response decision for the abnormal event is generated.

[0069] The third aspect of the present invention discloses another multi-dimensional hydrological data fusion and analysis device based on a smart hydrological pole, the device comprising:

[0070] A memory storing executable program code;

[0071] a processor coupled to the memory;

[0072] The processor calls the executable program code stored in the memory to execute the multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole disclosed in the first aspect of the present invention.

[0073] The fourth aspect of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole disclosed in the first aspect of the present invention.

[0074] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0075] In an embodiment of the present invention, a multi-dimensional hydrological data fusion and analysis method based on a smart hydrological pole is provided, the method comprising: obtaining hydrological detection data collected by a hydrological sensor, the hydrological sensor being arranged on the smart hydrological pole, the hydrological detection data comprising at least one of the following data: water level height, water flow velocity, water temperature, and water quality of a river channel; obtaining video data collected by a video acquisition device for a river channel, and determining river channel monitoring information based on the video data, the river channel monitoring information comprising at least one of the following information: river surface fluctuation information, river bank change information, and increase or decrease information of floating objects in the river channel; determining the hydrological detection data and the video data as target analysis data, and performing target data processing on the target analysis data according to a preset artificial intelligence algorithm to obtain a data processing result corresponding to the target analysis data; the target data processing comprises at least data fusion, data cross-validation, and in-depth analysis; generating a response decision for the data processing result according to a preset decision response algorithm; the response decision comprises at least one of early warning decisions, river channel maintenance decisions, and emergency decisions. It can be seen that the implementation of the present invention, through the integration of various hydrological sensors and video acquisition equipment configured on the smart hydrological pole, can synchronously obtain multi-dimensional hydrological detection data such as water level height, water flow velocity, water temperature, water quality, and intuitive video monitoring information such as river surface fluctuations, river bank changes, and floating objects in the river; further, it can also perform efficient data fusion and cross-validation on these data based on artificial intelligence algorithms, which is beneficial to improving the accuracy and completeness of the data; and, it can also perform in-depth analysis on the data based on artificial intelligence algorithms to explore potential hydrological change patterns and potential risks; in addition, it can also be based on the comprehensive analysis results of the data (the above-mentioned data processing results), through a preset decision-making response algorithm, automatically generate accurate early warning, maintenance or emergency event handling decisions, while improving the utilization rate of the data corresponding to the data processing results, It can also be beneficial to improve the timeliness of hydrological monitoring and emergency response capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. 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 creative work.

[0077] Figure 1 It is a flow chart of a multi-dimensional hydrological data fusion and analysis method based on a smart hydrological pole disclosed in an embodiment of the present invention;

[0078] Figure 2 It is a flow chart of another multi-dimensional hydrological data fusion and analysis method based on a smart hydrological pole disclosed in an embodiment of the present invention;

[0079] Figure 3 It is a structural schematic diagram of a multi-dimensional hydrological data fusion and analysis device based on a smart hydrological pole disclosed in an embodiment of the present invention;

[0080] Figure 4 It is a structural schematic diagram of another multi-dimensional hydrological data fusion and analysis device based on a smart hydrological pole disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0081] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not 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.

[0082] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or end including a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or ends.

[0083] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0084] The present invention discloses a multi-dimensional hydrological data fusion and analysis method and device based on a smart hydrological pole. By integrating a variety of hydrological sensors and video acquisition equipment configured on the smart hydrological pole, it is possible to synchronously obtain multi-dimensional hydrological detection data such as water level height, water flow velocity, water temperature, water quality, and intuitive video monitoring information such as river surface fluctuations, river bank changes, and river channel floating objects; further, it is also possible to perform efficient data fusion and cross-validation on these data based on artificial intelligence algorithms, which is conducive to improving the accuracy and completeness of the data; and, it is also possible to perform in-depth analysis on the data based on artificial intelligence algorithms to dig out potential hydrological change laws and risk hazards; in addition, it is also possible to automatically generate accurate early warning, maintenance or emergency event handling decisions based on the comprehensive analysis results of the data (the above-mentioned data processing results) through a preset decision-making response algorithm, while improving the utilization rate of the data corresponding to the data processing results, it is also conducive to improving the timeliness and emergency response capabilities of hydrological monitoring. The following are detailed descriptions.

[0085] Embodiment 1

[0086] See also Figure 1 , Figure 1 1 is a flow chart of a multi-dimensional hydrological data fusion and analysis method based on a smart hydrological pole disclosed in an embodiment of the present invention. Figure 1 The multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole described above can be applied to a multi-dimensional hydrological data fusion and analysis device based on the smart hydrological pole, and the embodiment of the present invention does not limit this. Figure 1 As shown, the multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole may include the following operations:

[0087] 101. Obtain hydrological detection data collected by a hydrological sensor, where the hydrological sensor is configured on a smart hydrological pole. The hydrological detection data includes at least one of the following data: water level height, water flow velocity, water temperature, and water quality of the river.

[0088] 102. Obtain video data collected by a video acquisition device for a river channel, and determine river channel monitoring information based on the video data, where the river channel monitoring information includes at least one of river surface fluctuation information, river bank change information, and increase or decrease information of floating objects in the river channel.

[0089] 103. Determine the hydrological detection data and the video data as target analysis data, and perform target data processing on the target analysis data according to a preset artificial intelligence algorithm to obtain a data processing result corresponding to the target analysis data.

[0090] In the embodiment of the present invention, the target data processing includes at least data fusion, data cross-validation and in-depth analysis.

[0091] 104. Generate a response decision for the data processing result according to a preset decision response algorithm; the response decision includes at least one of the following: early warning decision, river maintenance decision, and emergency decision.

[0092] It can be seen that the implementation Figure 1 The described multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole can synchronously obtain multi-dimensional hydrological detection data such as water level height, water flow velocity, water temperature, water quality, etc., as well as intuitive video monitoring information such as river surface fluctuations, river bank changes, and floating objects in the river channel by integrating a variety of hydrological sensors and video acquisition equipment configured on the smart hydrological pole; further, it can also perform efficient data fusion and cross-validation on these data based on artificial intelligence algorithms, which is conducive to improving the accuracy and completeness of the data; and, it can also perform in-depth analysis on the data based on artificial intelligence algorithms to explore potential hydrological change patterns and risk hazards; in addition, it can also automatically generate accurate early warning, maintenance or emergency event handling decisions based on the comprehensive analysis results of the data (the above-mentioned data processing results) through a preset decision-making response algorithm, which can improve the utilization rate of the data corresponding to the data processing results while also helping to improve the timeliness and emergency response capabilities of hydrological monitoring.

[0093] In an optional embodiment, the above step 103 performs target data processing on the target analysis data according to a preset artificial intelligence algorithm, and the method of obtaining the data processing result corresponding to the target analysis data specifically includes:

[0094] Performing data preprocessing on the target analysis data to obtain a data preprocessing result corresponding to the target analysis data, wherein the data preprocessing includes at least one of data cleaning, time dimension synchronization, and format conversion;

[0095] According to the preset data fusion strategy, data fusion is performed on the data preprocessing results to obtain data fusion results corresponding to the data preprocessing results. The data fusion includes feature extraction and feature fusion based on a multi-source data fusion algorithm;

[0096] Perform cross validation on the data fusion result to obtain a cross validation result corresponding to the data fusion result;

[0097] After determining that the cross-validation result indicates that the validation has passed, a deep analysis is performed on the data fusion result according to a preset deep learning model to obtain a deep analysis result corresponding to the data fusion result as a data processing result corresponding to the target analysis data.

[0098] In this optional embodiment, when the data preprocessing includes data cleaning, the data cleaning includes denoising and deduplication processing for the hydrological detection data and frame rate adjustment, de-jittering, and color correction for the video data;

[0099] The time dimension synchronization is used to perform time dimension synchronization on the hydrological detection data and / or the video data;

[0100] Format conversion is used to convert hydrological detection data and / or video data into image sequences or feature vectors that meet the requirements of machine learning.

[0101] It can be seen that in this optional embodiment, first, by performing data preprocessing on the target analysis data, including key steps such as data cleaning, time dimension synchronization and format conversion, the noise, inconsistency and format mismatch problems that may exist in the original data are effectively solved, and the accuracy and reliability of the data used for subsequent data fusion and analysis are improved; secondly, a preset data fusion strategy is adopted to perform feature extraction on the data preprocessing results and feature fusion based on a multi-source data fusion algorithm, so as to achieve deep integration and complementary advantages of multi-dimensional hydrological data. This step further improves the comprehensiveness and accuracy of the data. Then, by performing cross-validation, the fused data is strictly verified and screened to ensure the reliability of the data and the accuracy of the analysis, and further enhance the stability and credibility of the obtained data processing results. Finally, after determining that the cross-validation result indicates that the verification is passed, the preset deep learning model is used to perform in-depth analysis on the data fusion results to dig out the deep laws and characteristics hidden behind the data. The setting of this step further improves the accuracy and depth of data analysis, and at the same time can improve the data utilization rate of the data fusion results.

[0102] In another optional embodiment, the method of performing data fusion on the data preprocessing result according to the preset data fusion strategy to obtain the data fusion result corresponding to the data preprocessing result specifically includes:

[0103] Determine a plurality of first characteristic parameters corresponding to the hydrological detection data and a plurality of second characteristic parameters corresponding to the video data; the plurality of first characteristic parameters include at least two of the water level change rate, the peak value of the water flow velocity, and the change trend of the water quality index; the plurality of second characteristic parameters include at least two of the river surface fluctuation characteristics, the signs of river bank erosion, and the types and quantities of floating objects;

[0104] Extracting first feature data matching each first feature parameter from the hydrological detection data in the data preprocessing result, and extracting second feature data matching each second feature parameter from the video data in the data preprocessing result;

[0105] According to a preset data fusion algorithm, feature fusion is performed on all the first feature data and all the second feature data to obtain hydrological-video feature fusion data, and the hydrological-video feature fusion data includes a feature fusion vector.

[0106] In this optional embodiment, the selection of characteristic parameters fully considers the actual needs of hydrological monitoring and the diversity of data; among them, the first characteristic parameter covers key hydrological information such as water level change rate, peak water flow velocity, and change trend of water quality indicators, which can intuitively reflect the dynamic changes and potential risks of water bodies. At the same time, the second characteristic parameter includes video monitoring information such as river surface fluctuation characteristics, signs of river bank erosion, types and quantities of floating objects, etc., which are of great significance for evaluating river status and warning of natural disasters.

[0107] Secondly, in the extraction of feature data, the accuracy and completeness of the data are ensured by accurately extracting the data matching each feature parameter from the data preprocessing results. This step not only improves the efficiency of data processing, but also provides a reliable data foundation for subsequent feature fusion.

[0108] In the feature fusion stage, a preset data fusion algorithm is used to deeply integrate all the first feature data and the second feature data. Through this step, not only the seamless connection of hydrological data and video data is achieved, but also hydrological-video feature fusion data containing rich information is generated. This feature fusion data is presented in the form of feature fusion vectors, which not only facilitates subsequent data analysis and processing, but also greatly improves the data utilization value and decision support capabilities.

[0109] It can be seen that in this optional embodiment, through the refined data fusion step, not only the intelligent level of hydrological monitoring is improved, but also the comprehensiveness and accuracy of the data fusion results obtained are improved; through deep mining and integration of hydrological data and video data, it is beneficial to improve the efficiency and accuracy of subsequent data analysis based on the data fusion results.

[0110] Embodiment 2

[0111] See also Figure 2 , Figure 2 1 is a flow chart of another multi-dimensional hydrological data fusion and analysis method based on a smart hydrological pole disclosed in an embodiment of the present invention. Figure 2 The multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole described above can be applied to a multi-dimensional hydrological data fusion and analysis device based on the smart hydrological pole, and the embodiment of the present invention does not limit this. Figure 2 As shown, the multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole may include the following operations:

[0112] 201. Obtain hydrological detection data collected by a hydrological sensor, where the hydrological sensor is configured on a smart hydrological pole. The hydrological detection data includes at least one of the following data: water level height, water flow velocity, water temperature, and water quality of a river.

[0113] 202. Obtain video data collected by a video acquisition device for a river channel, and determine river channel monitoring information based on the video data, where the river channel monitoring information includes at least one of river surface fluctuation information, river bank change information, and increase or decrease information of floating objects in the river channel.

[0114] 203. Determine the hydrological detection data and the video data as target analysis data, and perform target data processing on the target analysis data according to a preset artificial intelligence algorithm to obtain a data processing result corresponding to the target analysis data.

[0115] In the embodiment of the present invention, the pre-set decision-making response algorithm includes a classification algorithm and a regression algorithm.

[0116] 204. Obtain abnormal record information set for the river channel, and determine whether the data processing result indicates the existence of abnormal events based on the decision-making response algorithm and the abnormal record information.

[0117] 205. When it is determined that the data processing result indicates the existence of an abnormal event, the abnormal type of the abnormal event is determined according to a classification algorithm.

[0118] 206. Analyze abnormal events and abnormal types based on regression algorithms to obtain the predicted change trend of abnormal events.

[0119] 207. Based on the strategy library of the decision response algorithm, combined with abnormal events, abnormal types and predicted change trends, response decisions for abnormal events are generated.

[0120] In the embodiment of the present invention, for other descriptions of step 201 to step 203, please refer to other specific descriptions of step 101 to step 103 in embodiment 1, and the embodiment of the present invention will not be repeated here.

[0121] It can be seen that the implementation Figure 2The described multi-dimensional hydrological data fusion and analysis method based on smart hydrological poles, firstly, in the identification of abnormal events, by obtaining the abnormal record information of the river channel setting, and combining the decision-making response algorithm to intelligently analyze the data processing results, it is possible to quickly and accurately determine whether there are abnormal events. This step not only improves the efficiency of identifying abnormal events, but also avoids potential hydrological disaster risks and improves the discovery speed and processing speed of abnormal events; secondly, in the determination of the abnormal type, a classification algorithm is adopted, which can accurately determine the type of the abnormal event according to its specific characteristics. The setting of this step is conducive to improving the accuracy and pertinence of the subsequent processing of the abnormal event; thirdly, in the analysis of the predicted change trend of abnormal events, the abnormal events and their types are deeply analyzed through the regression algorithm, and their possible development trends are predicted, which is conducive to enhancing the scientificity and accuracy of the response decisions generated subsequently; finally, in the generation of response decisions, it can automatically generate targeted response decisions based on the strategy library of the decision-making response algorithm, combined with abnormal events, abnormal types and predicted change trends. This step not only simplifies the decision-making process and improves the decision-making efficiency, but also ensures the scientificity and effectiveness of the decision.

[0122] In an optional embodiment, the method of performing cross-validation on the data fusion result to obtain a cross-validation result corresponding to the data fusion result specifically includes:

[0123] Acquire historical analysis data corresponding to the target analysis data, and the historical verification data is obtained by sequentially performing data preprocessing and data fusion on the historical hydrological detection data corresponding to the hydrological detection data and the historical video data corresponding to the video data;

[0124] Performing cross validation on the historical analysis data and the data fusion result to obtain a cross validation result corresponding to the historical analysis data and the data fusion result, wherein the cross validation includes at least one of difference calculation, correlation coefficient calculation, and cluster analysis;

[0125] Among them, the difference calculation includes at least one of the mean calculation, variance calculation, standard deviation calculation, and cross entropy loss calculation; the correlation coefficient calculation is used to determine the data correlation between the historical analysis data and the data fusion results; the cluster analysis is used to characterize the fusion effect of the historical analysis data and the data fusion results.

[0126] In this optional embodiment, in the data preparation stage of cross-validation, historical analysis data corresponding to the target analysis data is obtained by sequentially performing data preprocessing and data fusion on historical hydrological detection data and historical video data. This step ensures the comparability of historical analysis data with current data, and provides reliable data support for subsequent cross-validation.

[0127] In this optional embodiment, in the specific implementation of cross-validation, a variety of validation methods are used, among which the difference calculation quantifies the difference between the historical analysis data and the data fusion results by calculating statistical quantities such as mean, variance, standard deviation and cross entropy loss, thereby intuitively reflecting the accuracy of the data fusion results; the correlation coefficient calculation is used to determine the data correlation between the historical analysis data and the data fusion results, further verifying the reliability of the data fusion; the clustering analysis evaluates the performance of the data fusion in retaining the data features and structure by comparing the clustering effects of the historical analysis data and the data fusion results;

[0128] It can be seen that in this optional embodiment, by setting up a comprehensive application of multiple cross-validation strategies, it is beneficial to improve the comprehensiveness and accuracy of cross-validation; and, by setting up a cross-validation mechanism for data fusion results, it is possible to promptly discover possible problems in the data fusion process, thereby ensuring the accuracy and reliability of the data fusion results.

[0129] In another optional embodiment, the method of performing cross-validation on the data fusion result to obtain a cross-validation result corresponding to the data fusion result specifically further includes:

[0130] Before obtaining the historical analysis data corresponding to the target analysis data, perform integrity check and outlier detection on the data fusion results to obtain the review results corresponding to the data fusion results; the integrity check is used to detect whether there is data loss or omission in the data fusion results; outlier detection detects outliers in the data fusion results through statistical algorithms or machine learning algorithms;

[0131] When the review result indicates that the data fusion result contains unqualified data that does not meet expectations, the previously executed operation is backtracked and located based on the unqualified data to obtain a location node to be corrected for the previously executed operation; the previously executed operation at least includes data preprocessing and data fusion;

[0132] Execute node correction on the positioning node and re-execute the node operation corresponding to the positioning node; and trigger the execution of the above-mentioned operation corresponding to obtaining the historical analysis data corresponding to the target analysis data;

[0133] Among them, when the positioning node is located in the operation process of data preprocessing, the node correction includes at least one of data re-cleaning, resynchronization of time dimension, and format verification; when the positioning node is located in the operation process of data fusion, the node correction includes at least one of feature extraction algorithm adjustment, data fusion algorithm optimization or feature parameter adjustment.

[0134] In this optional embodiment, the above-mentioned method of performing integrity check on the data fusion result specifically includes:

[0135] Compare the data set sizes, the number of records, and the number of unique values ​​of specific fields of the data sets before and after data fusion, and obtain data comparison results between the data sets before and after data fusion;

[0136] Obtaining a record log for the data fusion result, and determining whether the record log contains a record of error information and / or warning information, and obtaining a determination result for the record log;

[0137] The data comparison result and the judgment result are determined as integrity check results corresponding to the data fusion result, and are added to the review result.

[0138] And, in this optional embodiment, the above-mentioned method of performing outlier detection on the data fusion result specifically includes:

[0139] According to a preset outlier detection algorithm, outlier calculation is performed on the data fusion result to obtain a target outlier value corresponding to the data fusion result;

[0140] The outlier detection algorithm includes any one of the 3σ rule, box plot, and Z-score algorithm.

[0141] In this optional embodiment, regarding the 3σ rule: when the data follows a normal distribution, sample values ​​less than μ-3σ or greater than μ+3σ are considered outliers, where μ is the sample mean and σ is the sample standard deviation. Regarding the boxplot: the range of outliers is defined according to the quartiles (Q1 and Q3) and interquartile range (IQR) of the data. Sample values ​​less than Q1-1.5IQR or greater than Q3+1.5IQR are generally considered outliers. Regarding the Z-score: the distance between the data point and the mean is measured. If the data point differs from the mean by more than a set multiple of the standard deviation (such as 3 times), it is considered an outlier.

[0142] In this optional embodiment, the accuracy and relevance of the data fusion results can be fully evaluated through the cross-validation mechanism to ensure that the obtained data can truly reflect the hydrological conditions. This step provides a reliable benchmark for subsequent integrity checks and outlier detection.

[0143] In this optional embodiment, integrity check and outlier detection, as key links in data quality control, further ensure the accuracy and availability of data fusion results. Integrity check can timely detect data loss or omission problems, avoiding interference of information loss on subsequent analysis; while outlier detection uses advanced statistical algorithms or machine learning algorithms to accurately identify and mark outliers in data fusion results, effectively preventing data errors from misleading analysis results.

[0144] In this optional embodiment, when the review results show that there are unqualified data in the data fusion results, the operation backtracking and positioning mechanism can be quickly started to accurately identify the source of the problem. This step not only improves the efficiency of problem solving, but also provides a clear direction for subsequent node corrections by comprehensively combing the previously executed operations.

[0145] In this optional embodiment, in the node correction stage, targeted correction measures can be taken according to the specific location of the positioning node. When the positioning node is located in the data preprocessing stage, the problems in the data preprocessing process are effectively solved by means of data re-cleaning, resynchronization of the time dimension, and format verification; and when the positioning node is located in the data fusion stage, the accuracy and reliability of the data fusion results are ensured by adjusting the feature extraction algorithm, optimizing the data fusion algorithm, or adjusting the feature parameters.

[0146] It can be seen that in this optional embodiment, multiple links such as cross-validation, integrity check, outlier detection, and operation backtracking and correction are integrated to form an efficient, accurate and self-correcting data processing closed loop; it not only further improves the accuracy and reliability of the data fusion results corresponding to the previous steps, but also enhances the self-diagnosis and repair capabilities of the hydrological related data management system, further improving the practicality of the data fusion results.

[0147] Embodiment 3

[0148] See also Figure 3 , Figure 3 It is a structural diagram of a multi-dimensional hydrological data fusion and analysis device based on a smart hydrological pole disclosed in an embodiment of the present invention. The multi-dimensional hydrological data fusion and analysis device based on a smart hydrological pole can be a multi-dimensional hydrological data fusion and analysis terminal, equipment, system or server based on a smart hydrological pole. The server can be a local server, a remote server, or a cloud server (also called a cloud server). When the server is a non-cloud server, the non-cloud server can communicate with the cloud server, which is not limited in the embodiment of the present invention. Figure 3 As shown, the multi-dimensional hydrological data fusion and analysis device based on the smart hydrological pole may include a data acquisition module 301, a determination module 302, a data processing module 303 and a decision generation module 304, wherein:

[0149] The data acquisition module 301 is used to acquire the hydrological detection data collected by the hydrological sensor, the hydrological sensor is configured on the smart hydrological pole, and the hydrological detection data includes at least one of the water level height, water flow velocity, water temperature and water quality of the river;

[0150] The data acquisition module 301 is also used to acquire video data collected by the video acquisition device for the river channel;

[0151] A determination module 302 is used to determine river monitoring information according to the video data, where the river monitoring information includes at least one of river surface fluctuation information, river bank change information, and increase or decrease information of floating objects in the river;

[0152] The determination module 302 is further used to determine the hydrological detection data and the video data as target analysis data;

[0153] The data processing module 303 is used to perform target data processing on the target analysis data according to a preset artificial intelligence algorithm to obtain a data processing result corresponding to the target analysis data; the target data processing at least includes data fusion, data cross-validation and in-depth analysis;

[0154] The decision generation module 304 is used to generate a response decision for the data processing result according to a preset decision response algorithm; the response decision includes at least one of the following decisions: early warning decision, river maintenance decision, and emergency decision.

[0155] It can be seen that the implementation Figure 3 The described multi-dimensional hydrological data fusion and analysis device based on the smart hydrological pole can synchronously obtain multi-dimensional hydrological detection data such as water level height, water flow velocity, water temperature, water quality, etc., as well as intuitive video monitoring information such as river surface fluctuations, river bank changes, and floating objects in the river channel, by integrating a variety of hydrological sensors and video acquisition equipment configured on the smart hydrological pole; further, it can also perform efficient data fusion and cross-validation on these data based on artificial intelligence algorithms, which is conducive to improving the accuracy and completeness of the data; and, it can also perform in-depth analysis of the data based on artificial intelligence algorithms to explore potential hydrological change patterns and risk hazards; in addition, it can also automatically generate accurate early warning, maintenance or emergency event handling decisions based on the comprehensive analysis results of the data (the above-mentioned data processing results) through a preset decision-making response algorithm, which can improve the utilization rate of the data corresponding to the data processing results while also helping to improve the timeliness and emergency response capabilities of hydrological monitoring.

[0156] In an optional embodiment, the data processing module 303 performs target data processing on the target analysis data according to a preset artificial intelligence algorithm, and the method of obtaining the data processing result corresponding to the target analysis data specifically includes:

[0157] Performing data preprocessing on the target analysis data to obtain a data preprocessing result corresponding to the target analysis data, wherein the data preprocessing includes at least one of data cleaning, time dimension synchronization, and format conversion;

[0158] According to the preset data fusion strategy, data fusion is performed on the data preprocessing results to obtain data fusion results corresponding to the data preprocessing results. The data fusion includes feature extraction and feature fusion based on a multi-source data fusion algorithm;

[0159] Perform cross validation on the data fusion result to obtain a cross validation result corresponding to the data fusion result;

[0160] After determining that the cross-validation result indicates that the validation has passed, a deep analysis is performed on the data fusion result according to a preset deep learning model to obtain a deep analysis result corresponding to the data fusion result as a data processing result corresponding to the target analysis data.

[0161] In this optional embodiment, when the data preprocessing includes data cleaning, the data cleaning includes denoising and deduplication processing for the hydrological detection data and frame rate adjustment, de-jittering, and color correction for the video data;

[0162] The time dimension synchronization is used to perform time dimension synchronization on the hydrological detection data and / or the video data;

[0163] Format conversion is used to convert hydrological detection data and / or video data into image sequences or feature vectors that meet the requirements of machine learning.

[0164] It can be seen that in this optional embodiment, first, by performing data preprocessing on the target analysis data, including key steps such as data cleaning, time dimension synchronization and format conversion, the noise, inconsistency and format mismatch problems that may exist in the original data are effectively solved, and the accuracy and reliability of the data used for subsequent data fusion and analysis are improved; secondly, a preset data fusion strategy is adopted to perform feature extraction on the data preprocessing results and feature fusion based on a multi-source data fusion algorithm, so as to achieve deep integration and complementary advantages of multi-dimensional hydrological data. This step further improves the comprehensiveness and accuracy of the data. Then, by performing cross-validation, the fused data is strictly verified and screened to ensure the reliability of the data and the accuracy of the analysis, and further enhance the stability and credibility of the obtained data processing results. Finally, after determining that the cross-validation result indicates that the verification is passed, the preset deep learning model is used to perform in-depth analysis on the data fusion results to dig out the deep laws and characteristics hidden behind the data. The setting of this step further improves the accuracy and depth of data analysis, and at the same time can improve the data utilization rate of the data fusion results.

[0165] In another optional embodiment, the data processing module 303 performs data fusion on the data preprocessing result according to a preset data fusion strategy, and the method of obtaining the data fusion result corresponding to the data preprocessing result specifically includes:

[0166] Determine a plurality of first characteristic parameters corresponding to the hydrological detection data and a plurality of second characteristic parameters corresponding to the video data; the plurality of first characteristic parameters include at least two of the water level change rate, the peak value of the water flow velocity, and the change trend of the water quality index; the plurality of second characteristic parameters include at least two of the river surface fluctuation characteristics, the signs of river bank erosion, and the types and quantities of floating objects;

[0167] Extracting first feature data matching each first feature parameter from the hydrological detection data in the data preprocessing result, and extracting second feature data matching each second feature parameter from the video data in the data preprocessing result;

[0168] According to a preset data fusion algorithm, feature fusion is performed on all the first feature data and all the second feature data to obtain hydrological-video feature fusion data, and the hydrological-video feature fusion data includes a feature fusion vector.

[0169] It can be seen that in this optional embodiment, through the refined data fusion step, not only the intelligent level of hydrological monitoring is improved, but also the comprehensiveness and accuracy of the data fusion results obtained are improved; through deep mining and integration of hydrological data and video data, it is beneficial to improve the efficiency and accuracy of subsequent data analysis based on the data fusion results.

[0170] In another optional embodiment, the data processing module 303 performs cross-validation on the data fusion result, and the method of obtaining the cross-validation result corresponding to the data fusion result specifically includes:

[0171] Acquire historical analysis data corresponding to the target analysis data, and the historical verification data is obtained by sequentially performing data preprocessing and data fusion on the historical hydrological detection data corresponding to the hydrological detection data and the historical video data corresponding to the video data;

[0172] Performing cross validation on the historical analysis data and the data fusion result to obtain a cross validation result corresponding to the historical analysis data and the data fusion result, wherein the cross validation includes at least one of difference calculation, correlation coefficient calculation, and cluster analysis;

[0173] Among them, the difference calculation includes at least one of the mean calculation, variance calculation, standard deviation calculation, and cross entropy loss calculation; the correlation coefficient calculation is used to determine the data correlation between the historical analysis data and the data fusion results; the cluster analysis is used to characterize the fusion effect of the historical analysis data and the data fusion results.

[0174] It can be seen that in this optional embodiment, by setting up a comprehensive application of multiple cross-validation strategies, it is beneficial to improve the comprehensiveness and accuracy of cross-validation; and, by setting up a cross-validation mechanism for data fusion results, it is possible to promptly discover possible problems in the data fusion process, thereby ensuring the accuracy and reliability of the data fusion results.

[0175] In another optional embodiment, the data processing module 303 performs cross-validation on the data fusion result, and the method of obtaining the cross-validation result corresponding to the data fusion result specifically includes:

[0176] Before obtaining the historical analysis data corresponding to the target analysis data, perform integrity check and outlier detection on the data fusion results to obtain the review results corresponding to the data fusion results; the integrity check is used to detect whether there is data loss or omission in the data fusion results; outlier detection detects outliers in the data fusion results through statistical algorithms or machine learning algorithms;

[0177] When the review result indicates that the data fusion result contains unqualified data that does not meet expectations, the previously executed operation is backtracked and located based on the unqualified data to obtain a location node to be corrected for the previously executed operation; the previously executed operation at least includes data preprocessing and data fusion;

[0178] Execute node correction on the positioning node and re-execute the node operation corresponding to the positioning node; and trigger the execution of the above-mentioned operation corresponding to obtaining the historical analysis data corresponding to the target analysis data;

[0179] Among them, when the positioning node is located in the operation process of data preprocessing, the node correction includes at least one of data re-cleaning, resynchronization of time dimension, and format verification; when the positioning node is located in the operation process of data fusion, the node correction includes at least one of feature extraction algorithm adjustment, data fusion algorithm optimization or feature parameter adjustment.

[0180] It can be seen that in this optional embodiment, multiple links such as cross-validation, integrity check, outlier detection, and operation backtracking and correction are integrated to form an efficient, accurate and self-correcting data processing closed loop; it not only further improves the accuracy and reliability of the data fusion results corresponding to the previous steps, but also enhances the self-diagnosis and repair capabilities of the hydrological related data management system, further improving the practicality of the data fusion results.

[0181] In yet another optional embodiment, the preset decision response algorithm includes a classification algorithm and a regression algorithm;

[0182] The decision generation module 304 generates a response decision for the data processing result according to a preset decision response algorithm, specifically including:

[0183] Obtain the abnormal record information set for the river channel, and determine whether the data processing result indicates the existence of abnormal matters based on the decision-making response algorithm and the abnormal record information;

[0184] When it is determined that the data processing result indicates that an abnormal event exists, the abnormal type of the abnormal event is determined according to the classification algorithm;

[0185] Analyze abnormal events and abnormal types according to regression algorithms to obtain the predicted change trend of abnormal events;

[0186] Based on the strategy library of the decision response algorithm, combined with abnormal events, abnormal types and predicted change trends, response decisions for abnormal events are generated.

[0187] It can be seen that in this optional embodiment, firstly, in the identification of abnormal events, by obtaining the abnormal record information of the river channel setting and combining the decision-making response algorithm to intelligently analyze the data processing results, it is possible to quickly and accurately determine whether there are abnormal events. This step not only improves the efficiency of identifying abnormal events, but also avoids potential hydrological disaster risks and improves the discovery speed and processing speed of abnormal events; secondly, in determining the type of abnormality, a classification algorithm is adopted, which can accurately determine the type of abnormal event based on its specific characteristics. The setting of this step is conducive to improving the accuracy and pertinence of subsequent processing of the abnormal event; thirdly, in the analysis of the predicted change trend of abnormal events, the abnormal events and their types are deeply analyzed through the regression algorithm to predict their possible development trends, which is conducive to enhancing the scientificity and accuracy of the response decisions generated subsequently; finally, in the generation of response decisions, it is possible to automatically generate targeted response decisions based on the strategy library of the decision-making response algorithm, combined with abnormal events, abnormal types and predicted change trends. This step not only simplifies the decision-making process and improves the decision-making efficiency, but also ensures the scientificity and effectiveness of the decision.

[0188] Embodiment 4

[0189] See also Figure 4 , Figure 4 FIG. 1 is a schematic diagram of the structure of another multi-dimensional hydrological data fusion and analysis device based on a smart hydrological pole disclosed in an embodiment of the present invention. Figure 4 As shown, the multi-dimensional hydrological data fusion and analysis device based on the smart hydrological pole may include:

[0190] A memory 401 storing executable program codes;

[0191] a processor 402 coupled to the memory 401;

[0192] The processor 402 calls the executable program code stored in the memory 401 to execute the steps in the multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole described in the first embodiment of the present invention or the second embodiment of the present invention.

[0193] Embodiment 5

[0194] An embodiment of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the steps in the multidimensional hydrological data fusion and analysis method based on smart hydrological poles described in Embodiment 1 or Embodiment 2 of the present invention.

[0195] Embodiment 6

[0196] An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer storage medium storing a computer program, and the computer program can be operated to enable a computer to execute the steps in the multi-dimensional hydrological data fusion and analysis method based on a smart hydrological pole described in Example 1 or Example 2.

[0197] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0198] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or partly contributed to the prior art in the form of a software product, and the computer software product can be stored in a computer storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0199] Finally, it should be noted that the multi-dimensional hydrological data fusion and analysis method and device based on the smart hydrological pole disclosed in the embodiment of the present invention only discloses the preferred embodiment of the present invention, which is only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions recorded in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A multi-dimensional hydrological data fusion and analysis method based on a smart hydrological pole, characterized in that: The method comprises: Acquire hydrological detection data collected by a hydrological sensor, wherein the hydrological sensor is configured on the smart hydrological pole, and the hydrological detection data includes at least one of the following data: water level height, water flow velocity, water temperature, and water quality of a river; Acquire video data collected by a video acquisition device for the river channel, and determine river channel monitoring information based on the video data, wherein the river channel monitoring information includes at least one of river surface fluctuation information, river bank change information, and increase or decrease information of floating objects in the river channel; Determine the hydrological detection data and the video data as target analysis data, and perform target data processing on the target analysis data according to a preset artificial intelligence algorithm to obtain a data processing result corresponding to the target analysis data; the target data processing at least includes data fusion, data cross-validation and in-depth analysis; According to a preset decision response algorithm, a response decision is generated for the data processing result; the response decision includes at least one type of decision from among early warning decisions, river maintenance decisions, and emergency event decisions.

2. The multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole according to claim 1 is characterized in that: The step of performing target data processing on the target analysis data according to a preset artificial intelligence algorithm to obtain a data processing result corresponding to the target analysis data includes: Performing data preprocessing on the target analysis data to obtain a data preprocessing result corresponding to the target analysis data, wherein the data preprocessing includes at least one of data cleaning, time dimension synchronization, and format conversion; According to a preset data fusion strategy, data fusion is performed on the data preprocessing result to obtain a data fusion result corresponding to the data preprocessing result, wherein the data fusion includes feature extraction and feature fusion based on a multi-source data fusion algorithm; Performing cross validation on the data fusion result to obtain a cross validation result corresponding to the data fusion result; After determining that the cross-validation result indicates that the validation is passed, a deep analysis is performed on the data fusion result according to a preset deep learning model to obtain a deep analysis result corresponding to the data fusion result as a data processing result corresponding to the target analysis data.

3. The multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole according to claim 2 is characterized in that: The performing data fusion on the data preprocessing result according to the preset data fusion strategy to obtain a data fusion result corresponding to the data preprocessing result includes: Determine a plurality of first characteristic parameters corresponding to the hydrological detection data and a plurality of second characteristic parameters corresponding to the video data; the plurality of first characteristic parameters include at least two of the water level change rate, the peak value of the water flow velocity, and the change trend of the water quality index; the plurality of second characteristic parameters include at least two of the river surface fluctuation characteristics, the signs of river bank erosion, and the types and quantities of floating objects; Extracting first feature data matching each of the first feature parameters from the hydrological detection data in the data preprocessing result, and extracting second feature data matching each of the second feature parameters from the video data in the data preprocessing result; According to a preset data fusion algorithm, feature fusion is performed on all the first feature data and all the second feature data to obtain hydrological-video feature fusion data, and the hydrological-video feature fusion data includes a feature fusion vector.

4. The multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole according to claim 2 or 3 is characterized in that: When the data preprocessing includes the data cleaning, the data cleaning includes denoising and deduplication processing for the hydrological detection data and frame rate adjustment, de-jittering, and color correction for the video data; The time dimension synchronization is used to perform time dimension synchronization on the hydrological detection data and / or the video data; The format conversion is used to convert the hydrological detection data and / or the video data into an image sequence or feature vector adapted to machine learning requirements.

5. The multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole according to claim 2 or 3 is characterized in that: The performing cross validation on the data fusion result to obtain a cross validation result corresponding to the data fusion result includes: Acquire historical analysis data corresponding to the target analysis data, wherein the historical verification data is obtained by sequentially performing the data preprocessing and data fusion on the historical hydrological detection data corresponding to the hydrological detection data and the historical video data corresponding to the video data; Performing cross validation on the historical analysis data and the data fusion result to obtain a cross validation result corresponding to the historical analysis data and the data fusion result, wherein the cross validation includes at least one of difference calculation, correlation coefficient calculation, and cluster analysis; Among them, the difference calculation includes at least one of mean calculation, variance calculation, standard deviation calculation, and cross entropy loss calculation; the correlation coefficient calculation is used to determine the data correlation between the historical analysis data and the data fusion result; the cluster analysis is used to characterize the fusion effect of the historical analysis data and the data fusion result.

6. The multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole according to claim 5 is characterized in that: The performing cross validation on the data fusion result to obtain a cross validation result corresponding to the data fusion result also includes: Before acquiring the historical analysis data corresponding to the target analysis data, performing integrity check and outlier detection on the data fusion result to obtain a review result corresponding to the data fusion result; the integrity check is used to detect whether there is data loss or omission in the data fusion result; the outlier detection detects outliers in the data fusion result by using a statistical algorithm or a machine learning algorithm; When the review result indicates that the data fusion result contains unqualified data that does not meet expectations, the previously executed operation is backtracked and located according to the unqualified data to obtain a location node to be corrected for the previously executed operation; the previously executed operation at least includes the data preprocessing and the data fusion; Execute node correction on the positioning node and re-execute the node operation corresponding to the positioning node; and trigger execution of the operation corresponding to obtaining the historical analysis data corresponding to the target analysis data; Among them, when the positioning node is located in the operation process of the data preprocessing, the node correction includes at least one of data re-cleaning, re-synchronization of the time dimension, and format verification; when the positioning node is located in the operation process of the data fusion, the node correction includes at least one of feature extraction algorithm adjustment, data fusion algorithm optimization or feature parameter adjustment.

7. The multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole according to claim 1, 2, 3 or 6, characterized in that: The preset decision-making response algorithm includes a classification algorithm and a regression algorithm; the generating a response decision for the data processing result according to the preset decision-making response algorithm includes: Acquire abnormal record information set for the river channel, and determine whether the data processing result indicates the existence of abnormal matters according to the decision-making response algorithm and the abnormal record information; When it is determined that the data processing result indicates the presence of the abnormal event, determining the abnormal type of the abnormal event according to the classification algorithm; Analyze the abnormal events and the abnormal types according to the regression algorithm to obtain the predicted change trend of the abnormal events; According to the strategy library of the decision response algorithm, combined with the abnormal event, the abnormal type and the predicted change trend, a response decision for the abnormal event is generated.

8. A multi-dimensional hydrological data fusion and analysis device based on a smart hydrological pole, characterized in that: The device comprises: A data acquisition module, used to acquire hydrological detection data collected by a hydrological sensor, wherein the hydrological sensor is arranged on the smart hydrological pole, and the hydrological detection data includes at least one of the following data: water level height, water flow velocity, water temperature, and water quality of a river; The data acquisition module is further used to acquire video data collected by a video acquisition device for the river channel; A determination module, configured to determine river channel monitoring information according to the video data, wherein the river channel monitoring information includes at least one of river surface fluctuation information, river bank change information, and increase or decrease information of floating objects in the river channel; The determination module is further used to determine the hydrological detection data and the video data as target analysis data; A data processing module, used to perform target data processing on the target analysis data according to a preset artificial intelligence algorithm to obtain a data processing result corresponding to the target analysis data; the target data processing at least includes data fusion, data cross-validation and in-depth analysis; The decision generation module is used to generate a response decision for the data processing result according to a preset decision response algorithm; the response decision includes at least one type of decision among early warning decisions, river maintenance decisions, and emergency event decisions.

9. A multi-dimensional hydrological data fusion and analysis device based on a smart hydrological pole, characterized in that: The device comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: The computer storage medium stores computer instructions, which, when called, are used to execute the multi-dimensional hydrological data fusion and analysis method based on the smart hydrological pole as described in any one of claims 1-7.

Images