Multilevel data security and storage management platform in hydropower industry

Abstract

The invention discloses a multilevel data security and storage management platform in the hydropower industry, and relates to the technical field of data management in the hydropower industry, the management platform performs hierarchical division on a hydropower station according to regions and functions through a node division unit, and each node represents a multi-dimensional data set; data of each generator set can be independently collected, the problems of redundancy and insufficient precision are reduced, the storage hierarchical management module carries out classification through an access heat index Rh of a node, data which are frequently accessed are preferentially stored in a high-performance storage device, and cold data are stored in a low-cost storage medium. Through a hierarchical storage strategy, the storage cost is reduced; the disaster backup strategy module enhances the disaster response capability and optimizes the data backup and recovery strategy; and the network interruption security policy module ensures the data security during the network interruption, provides a flexible data synchronization scheme, and ensures the accurate synchronization and recovery of the data after the network is recovered.

Description

Technical Field

[0001] The present invention relates to the technical field of data management in the hydropower industry, and particularly to a multi-level data security and storage management platform for the hydropower industry. Background Art

[0002] With the continuous advancement of the informatization process in the hydropower industry, a large amount of multi-dimensional data generated during the production and management of hydropower stations plays an important role in ensuring power generation efficiency, improving management levels, and optimizing dispatching decisions. However, various equipment and facilities in hydropower stations are widely distributed and operate in complex environments. How to ensure the security, reliability, and efficient storage of this data has become the main challenge faced by the hydropower industry.

[0003] Currently, traditional hydropower data storage management mostly relies on simple databases or cloud storage solutions, often unable to meet the data storage requirements and the complexity of data access in different regions and different functional nodes of the hydropower industry. For example, data at certain nodes may be wasted in storage space due to frequent changes or data aging as they are not accessed for a long time, while important node data in other parts requires high-frequency and fast access, demanding higher-performance storage devices. In addition, hydropower stations also face the risk of being affected by natural disasters (such as earthquakes, landslides, floods, etc.), which may pose a serious threat to data security. Therefore, the hydropower industry urgently needs a multi-level data security and storage management platform for the hydropower industry to solve these problems. Summary of the Invention

[0004] In view of the deficiencies of the prior art, the present invention provides a multi-level data security and storage management platform for the hydropower industry to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: A multi-level data security and storage management platform for the hydropower industry, comprising:

[0006] A first data acquisition module for acquiring a multi-dimensional data set of each node in the hydropower station;

[0007] A second data acquisition module for real-time monitoring of the access behavior of the multi-dimensional data set of each node, capturing access events, and recording all access logs through a database trigger or API Hook mechanism to establish an access data set;

[0008] A storage hierarchical management module for analyzing based on the access data set, constructing an access heat index Rh for each node for evaluation and classification, and performing storage management according to the corresponding hierarchical strategy:

[0009] The disaster backup strategy module is used to collect real-time geological disaster data, construct and evaluate the disaster risk index Gr to obtain the disaster assessment result, and optimize the hierarchical strategy based on the disaster assessment result to form a hierarchical backup plan.

[0010] Preferably, the first data collection module includes a node division unit and a first collection unit;

[0011] The node division unit is used to divide the hydropower station according to regions and functions, and each node corresponds to a multi-dimensional data set of a region;

[0012] The specific steps for dividing nodes are as follows:

[0013] First, divide the hydropower station into first nodes according to the upstream, middle stream, and downstream, collect the generator set data within each first node, then perform a second node division on each generator set, and mark them as generator sets A1, A2,..., A n , where n represents the number of generator sets;

[0014] The first collection unit is used to collect the multi-dimensional data set of the second node, and the multi-dimensional data set includes the equipment operation data of each generator set, the hydrological data connected to each generator set, and the environmental data of the region where each hydropower unit is located.

[0015] Preferably, the second data collection module includes a second collection unit and a data processing unit;

[0016] The second collection unit is used to monitor the access behavior of the multi-dimensional data set of each second node in real time, capture access events, and record all access logs through the database trigger or APIHook mechanism to establish an access data set;

[0017] The steps for establishing the access data set are as follows:

[0018] S11. Used to count the access volume from the log_access_table access log according to the time window to obtain the access frequency F v :

[0019] S12. And extract the timestamp t of each access of the multi-dimensional data set of each second node from the log_access_table access log i , and calculate the difference t cz : t cz =t i+1 -t i ;

[0020] Calculate and obtain the average access interval T through the following formula a :

[0021]

[0022] wherein, the difference t between adjacent timestamps cz represents the time difference between the i-th and the (i + 1)-th access, and t i represents the timestamp of the i-th access, and m represents the total number of accesses recorded in the log table;

[0023] S13. Extract the total access time length T of the multi-dimensional data set of each second node data from the query_execution_time or data_transfer_size performance monitoring tool r and the amount of data accessed S r :

[0024] The amount of data accessed S r represents the amount of data actually read and written for each second node data in a single access;

[0025] S14. Extract the initial timestamp T and the current timestamp T of the multi-dimensional data set of each second node data generated from the log_access_table access log g and calculate the data decay coefficient I through the following formula C : d :

[0026] ΔT = T C - T g ;

[0027]

[0028] I d = e -λ*ΔT ;

[0029] wherein, ΔT represents the time span from data creation to the current time, λ represents the data aging coefficient, with one month set as the unit, λ is a constant, and is obtained through derivation by the decay model based on historical access data experience using the least squares regression analysis method; log 10 represents the common logarithm with base 10 of the natural number, and e represents the base of the exponential function; F v represents the access frequency;

[0030] S15. Extract the actual number of access users U of each second node data from the log_access_table access log a and the total number of users U in the management platform T , and calculate the access complexity U r :

[0031]

[0032] S16. Count the access frequency F obtained from S11-S15 v , average access interval T a , Total access time length T r , access data volume S r , data attenuation coefficient I d And the access complexity U r ,Preprocessing is performed through the data processing unit, the data denoising algorithm is used to remove abnormal data, and the data is processed uniformly using standardization technology. After obtaining normalized data through dimensionless technology, the access data set is established.

[0033] Preferably, the storage hierarchical management module includes a data analysis unit, a data classification unit and a storage strategy unit;

[0034] The data analysis unit is used to extract the access frequency F in the access data set. v , average access interval T a , Total access time length T r , access data volume S r , data attenuation coefficient I d And the access complexity U r , the access heat index Rh is calculated by the following associated formula;

[0035]

[0036] Wherein, k1, k2, k3 and k4 represent weight coefficients, and the sum of weight coefficients is 1;

[0037] The data classification unit is used to preset a high heat threshold H1, a medium heat threshold H2 and a low heat threshold H3, and the high heat threshold H1> the medium heat threshold H2> the low heat threshold H3, and compare the access heat index Rh with the high heat threshold H1, the medium heat threshold H2 and the low heat threshold H3 respectively to obtain a classification result, including:

[0038] If the access heat index Rh> the high heat threshold H1, a first-level label is generated, indicating that the access frequency and relevance of the multidimensional data set of the current second node are very high, and it is core or key data;

[0039] If the medium heat threshold H2≤access heat index Rh≤high heat threshold H1, a second-level label is generated, indicating that the access frequency and relevance of the multidimensional data set of the current second node are in the medium range, which is neither the most frequently accessed data nor the unpopular data, but active data;

[0040] If the low heat threshold H3 ≤ access heat index Rh < medium heat threshold H2, generate a third-level label, indicating a relatively low access frequency and relevance, belonging to data with less frequent access;

[0041] If the access heat index Rh < low heat threshold H3, generate a fourth-level label, indicating that the access frequency and relevance of the current data set are very low, belonging to cold data.

[0042] Preferably, the storage strategy unit is used to generate a hierarchical storage strategy according to the classification result, including:

[0043] Generate a first hierarchical strategy based on the first-level label. Specifically: store the multi-dimensional data set of the second node with the first-level label in the solid-state drive SSD;

[0044] Generate a second hierarchical strategy based on the second-level label. Specifically: store the multi-dimensional data set of the second node with the second-level label in the traditional hard disk HDD or the network storage device NAS;

[0045] Generate a third hierarchical strategy based on the third-level label. Specifically: store the multi-dimensional data set of the second node with the third-level label in the cold storage cloud service, tape storage or optical disc storage;

[0046] Generate a fourth hierarchical strategy based on the third-level label. Specifically: archive the multi-dimensional data set of the second node with the fourth-level label. Due to the extremely low reading requirement, focus on the storage cost rather than the access speed.

[0047] Preferably, the disaster backup strategy module includes a third collection unit, a disaster factor calculation unit and a disaster prediction unit; element;

[0048] The third collection unit is used to collect disaster data sets in the upstream, middle and downstream areas of the first node;

[0049] The disaster data set includes earthquake magnitude ZJ, focal depth Sd, seismic wave propagation speed zsd, maximum reservoir water level Sw, water flow speed Ssd, landslide and debris flow volume Lsh, landslide slope Pd, wind speed Fs and hail diameter BL;

[0050] The earthquake magnitude ZJ and focal depth Sd are collected and obtained through a three-component seismograph;

[0051] The seismic wave propagation speed zsd is collected and obtained through a seismic wave propagation sensor;

[0052] The maximum reservoir water level Sw is collected and obtained through a water level gauge;

[0053] The water flow speed Ssd is collected and obtained through a flow meter

[0054] The landslide and debris flow volume Lsh is obtained by lidar scanning the upstream, midstream, and downstream areas of the first node to acquire three-dimensional terrain data. Before and after the landslide or debris flow event, lidar data is collected twice. By comparing the data from the two scans and using ground elevation difference analysis, the ground settlement or accumulation areas caused by the landslide or debris flow are identified to obtain the difference areas, and the difference areas are calculated using three-dimensional geometric algorithms, including the cross-section method or the volume extrapolation method.

[0055] The landslide slope Pd is obtained by analyzing the difference area to obtain the height information of each pixel in the area. The landslide slope is equal to the ratio of the height change to the horizontal distance.

[0056] The wind speed Fs is obtained by collecting with an anemometer.

[0057] The hail particle size BL is obtained by detecting hail particles in the meteorological phenomenon through radar scanner scanning of the meteorological cloud layer and collecting through the reflected signal, or by collecting with a laser particle size analyzer.

[0058] Preferably, the disaster factor calculation unit is used to remove abnormal data from the disaster data set by adopting a data denoising algorithm, and uniformly process the data using standardization technology. After obtaining the normalized data through dimensionless technology, the earthquake magnitude ZJ, focal depth Sd, seismic wave propagation velocity zsd, maximum reservoir water level Sw, water flow velocity Ssd, landslide and debris flow volume Lsh, landslide slope Pd, wind speed Fs, and hail particle size BL are extracted, and the disaster risk index Gr is calculated through the following formula:

[0059] F eq1 = W1*Sd + W2*Sd + W3*zsd;

[0060] F eq2 = W4*Sw + W5*Ssd;

[0061] F eq3 = W6*Lsh + W7*Pd;

[0062] F eq4 = W8*Fs + W9*BL;

[0063] Gr = α*F eq1 + β*F eq2 + γ*F eq3 + δ*F eq4 ;

[0064] In the formula, F eq1 represents the earthquake disaster factor, F eq2 represents the flood disaster factor, F eq3Denote the mountain disaster factor as F eq4 Denote the meteorological disaster factors as W1, W2, and W3 which are the weight coefficients of earthquake magnitude ZJ, focal depth Sd, and seismic wave propagation velocity zsd respectively; W4 and W5 are the weight coefficients of the maximum reservoir water level Sw and water flow velocity Ssd respectively; W6 and W7 are the weight coefficients of landslide and debris flow volume Lsh and landslide slope Pd respectively; W8 and W9 are the weight coefficients of wind speed Fs and hail diameter BL respectively; α, β, γ, and δ are the weight coefficients of earthquake disaster factor F eq1 , flood disaster factor F eq2 , mountain disaster factor F eq3 , and meteorological disaster factor F eq4 , and the sum of the weight coefficients is 1.

[0065] Preferably, the disaster prediction unit is used to preset a disaster risk threshold Q, and compare the disaster risk index Gr with the disaster risk threshold Q to obtain a disaster prediction result, including:

[0066] When the disaster risk index Gr ≥ the disaster risk threshold Q, it indicates that there is a disaster risk in the first node area, identify whether the current first node area is upstream, middle stream or downstream, and generate corresponding backup strategies;

[0067] When the disaster risk index Gr < the disaster risk threshold Q, it indicates that there is no disaster risk in the first node area;

[0068] The corresponding backup strategies include:

[0069] Identify the first node area. If it is upstream, implement the first backup strategy, including: on the basis of the hierarchical storage strategy, increase the backup redundancy ratio of the multi-dimensional data set of the second node in the first node in the upstream by more than 80%, because the upstream area is more affected by disasters such as landslides and debris flows;

[0070] Identify the first node area. If it is middle stream, implement the second backup strategy, including: on the basis of the hierarchical storage strategy, increase the backup redundancy ratio of the multi-dimensional data set of the second node in the first node in the middle stream by 50%-80%; compared with the upstream area, the middle stream area is relatively less affected by disasters, but data backup still needs to consider water level changes and meteorological conditions;

[0071] Identify the first node area. If it is downstream, implement the third backup strategy, including: increase the backup redundancy ratio of the multi-dimensional data set of the second node in the first node in the upstream by less than 40%. The downstream area is usually affected by water level, flood, and climate change, etc., and the disaster risk is relatively low, but data backup for water flow changes is still required. The backup strategy is simplified, but it still needs to be uploaded to the cloud platform or backed up remotely regularly to prevent data loss caused by sudden disasters.

[0072] Preferably, it further includes a network disconnection security policy module, and the network disconnection security policy module includes an identification unit, a recovery unit, and a synchronization policy unit;

[0073] The identification unit is used for the network card status monitoring device to monitor the network interface status. If the network card is in the "no connection" or "disconnected" state, it can be judged as a network disconnection. When the network disconnection state is identified, the local data security storage mechanism is enabled, and the multi-dimensional data set of the second node is backed up in real time through the local area network, and the backup database is collected;

[0074] And based on the backup database, the local storage integrity index Cl is generated through the following formula:

[0075]

[0076] In the formula, Wc s represents the amount of valid data successfully backed up and stored during the network disconnection period, and Wc y represents all the data that should be stored locally during the network disconnection period;

[0077] The recovery unit is used for when the network is restored, to synchronize and update data based on the local storage integrity index Cl and the remote database, and generate a synchronization consistency coefficient Sc through the following formula:

[0078]

[0079] In the formula, Bc s represents the amount of data that has not been successfully synchronized between the local storage and the remote database; Zc y represents all the data that should be synchronized. The closer the value of the synchronization consistency coefficient Sc is to 1, the higher the reliability of the synchronization, and the smaller the error or delay of the data synchronization.

[0080] Preferably, the synchronization policy unit is used to further evaluate the synchronization consistency coefficient Sc, including:

[0081] When 0.8 < synchronization consistency coefficient Sc < 1, trigger the manual synchronization method, trigger the manual synchronization process, and resynchronize the data that has not been successfully synchronized;

[0082] If the synchronization consistency coefficient Sc ≤ 0.8, trigger the incremental synchronization method, and only synchronize the data that has changed since the network disconnection, so as to reduce the synchronization time and pressure.

[0083] The present invention provides a multi-level data security and storage management platform for the hydropower industry. It has the following beneficial effects:

[0084] (1) The multi-level data security and storage management platform for the hydropower industry hierarchically divides hydropower stations by region (upstream, midstream, downstream) and function through the node division unit, ensuring more targeted data collection. Each node represents a multi-dimensional data set within a region, clearly reflecting the specific data characteristics within each region. This zoning division can effectively avoid problems such as data redundancy or insufficient collection accuracy caused by overly broad or irrelevant data. Each generator set is managed and collected in detail through the second node division. This refined management method enables data collection not only at the overall hydropower station level but also independent monitoring of each generator set. This means that the equipment operation data, connected hydrological data, and environmental data of each generator set can be independently collected and analyzed, providing a more accurate data basis for subsequent data processing and storage. After dividing the data by nodes, it is possible to store and analyze according to the data characteristics of each node. This not only improves the storage efficiency but also helps the management platform optimize storage and processing strategies according to the specific needs of different nodes. For example, high-frequency data can be stored in higher-performance storage devices, while infrequently used data can adopt low-cost storage methods.

[0085] (2) The multi-level data security and storage management platform for the hydropower industry classifies data according to the access heat index Rh of the nodes. The system can identify which data is most frequently accessed and which is less accessed. In this way, hot data can be preferentially stored on high-performance storage media, while cold data can be stored on more economical storage media, thereby reducing storage costs and improving access efficiency. The hierarchical storage strategy ensures that frequently accessed data can be quickly responded to, optimizing the overall performance of the system and avoiding the problem of cold data with frequent access slowing down the system response speed. Storing cold data on lower-cost storage media can reduce the pressure on high-performance storage devices, extend the service life of these devices, and avoid resource waste.

[0086] (3) The multi-level data security and storage management platform for the hydropower industry collects real-time geological disaster data and constructs a disaster risk index Gr. The system can timely evaluate the disaster risk and make backup preparations in advance to ensure that critical data can be effectively protected during disasters and reduce the risk of data loss. Based on the disaster risk assessment, the system will dynamically adjust the backup strategy and set different backup redundancy ratios for regions with different risk levels. For example, the upstream region may be greatly affected by natural disasters such as landslides and debris flows, and the system will increase the backup redundancy ratio to ensure data security during disasters.

[0087] (4) The multi-level data security and storage management platform for the hydropower industry. When the network is interrupted, the system will enable the local data security storage mechanism and perform real-time backup of key data through the local area network to prevent data loss. By calculating the local storage integrity index Cl, the effectiveness of the local backup can be evaluated in real time, providing a basis for subsequent data recovery and synchronization. The recovery unit monitors the data synchronization process through the synchronization consistency coefficient Sc to ensure data consistency between the local storage and the remote database after the network is restored. If the synchronization consistency coefficient is low, the system will automatically trigger incremental synchronization to ensure that only the changed data is synchronized, reducing the synchronization time and system load. According to the synchronization consistency coefficient, the system can intelligently select the most suitable synchronization method (manual synchronization or incremental synchronization). In this way, the most appropriate synchronization strategy can be selected according to the actual situation to ensure the reliability of data recovery and minimize resource consumption during the synchronization process. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] Figure 1 It is a schematic block diagram flow chart of a multi-level data security and storage management platform for the hydropower industry according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0089] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0090] Embodiment 1

[0091] Please refer to Figure 1 , the present invention provides a multi-level data security and storage management platform for the hydropower industry, including:

[0092] The first data acquisition module is used to acquire the multi-dimensional data set of each node in the hydropower station;

[0093] The second data acquisition module is used to monitor the access behavior of the multi-dimensional data set of each node in real time, capture access events, and record all access logs through the database trigger or APIHook mechanism to establish an access data set;

[0094] The storage hierarchical management module is used to analyze according to the access data set, construct the access heat index Rh of each node for evaluation and classification, and perform storage management according to the corresponding hierarchical strategy:

[0095] The disaster backup strategy module is used to collect real-time geological disaster data, construct and evaluate the disaster risk index Gr to obtain the disaster assessment result, and optimize the hierarchical strategy based on the disaster assessment result to form a hierarchical backup plan.

[0096] In this embodiment, through the storage hierarchical management module, according to the heat of node data access (evaluated by the access heat index Rh), the data is classified and stored, optimizing the use of storage space. In this way, data that is not accessed for a long time will not occupy too much storage resources, while for frequently accessed data, higher-performance storage support is provided. Multi-level storage management is adopted to ensure that high-frequency nodes of data access can use higher-performance storage devices, while infrequently used data can be stored on devices with lower costs. This hierarchical storage strategy improves the overall performance of the management platform while reducing the storage cost. The disaster backup strategy module collects geological disaster data in real time (such as earthquakes, landslides, floods, etc.) and constructs the disaster risk index Gr to evaluate the disaster risk. On this basis, the storage strategy can be optimized to ensure that critical data can be backed up in a timely manner during a disaster, preventing data loss or damage and ensuring the data security of the hydropower station. Through the flexible multi-level data storage and disaster risk management mechanism, this platform can adapt to the special needs of the hydropower industry, process large-scale and multi-dimensional real-time data, and can be customized according to different regions, nodes, and actual access behaviors. With the continuous expansion of the hydropower station network, the scalability of the platform can effectively handle more storage requirements.

[0097] In the event of natural disasters and other emergencies, through the disaster backup strategy module and the optimized hierarchical backup plan, the platform can effectively protect critical data and improve the disaster recovery ability of the hydropower industry. Whether it is the high or low frequency of data access or the risk assessment of disasters, the management platform can make dynamic adjustments to ensure that the data storage and recovery capabilities meet the actual needs. Traditional storage management platforms cannot dynamically classify and optimize the access frequency and importance of data. However, through the analysis of the access data set and the hierarchical storage strategy of this invention, while ensuring data security, the storage efficiency and performance are improved. Traditional systems are prone to irreversible damage to data due to natural disasters, while this platform introduces a disaster backup strategy module to evaluate the disaster risk in real time and optimize the backup, effectively coping with the impact of natural disasters.

[0098] Embodiment 2

[0099] This embodiment is an explanatory description based on Embodiment 1. Please refer to Figure 1 , specifically, the first data collection module includes a node division unit and a first collection unit;

[0100] The node division unit is used to divide the hydropower station according to regions and functions, and each node corresponds to a multi-dimensional data set of a region;

[0101] The specific steps for dividing nodes are as follows:

[0102] First, divide the hydropower station into first nodes according to the upper, middle, and lower reaches. Then, collect the generator set data within each first node. Next, perform a second node division on each generator set and label them as generator sets A1, A2,..., A n , where n represents the number of generator sets;

[0103] The first acquisition unit is used to acquire the multi-dimensional data set of the second node. The multi-dimensional data set includes the equipment operation data of each generator set, the hydrological data connected to each generator set, and the environmental data of the region where each hydropower unit is located.

[0104] In this embodiment, the node division unit hierarchically divides the hydropower station according to regions (upper, middle, and lower reaches) and functions, ensuring more targeted data acquisition. Each node represents a multi-dimensional data set within a region, which can clearly reflect the specific data characteristics within each region. This zoning division can effectively avoid problems such as data redundancy or insufficient acquisition accuracy caused by overly broad or irrelevant data. Each generator set is managed and acquired in detail through the second node division (such as A1, A2,..., A n ). This refined management method enables data acquisition not only at the overall hydropower station level but also independent monitoring of each generator set. This means that the equipment operation data, connected hydrological data, and environmental data of each generator set can be independently acquired and analyzed, providing a more accurate data basis for subsequent data processing and storage. After dividing the data according to nodes, it is possible to store and analyze the data characteristics of each node. This not only improves the storage efficiency but also helps the management platform optimize the storage and processing strategies according to the specific needs of different nodes. For example, high-frequency data can be stored in higher-performance storage devices, while infrequently used data can be stored in a low-cost manner.

[0105] Embodiment 3

[0106] This embodiment is an explanatory description based on Embodiment 1. Please refer to Figure 1 , specifically, the second data acquisition module includes a second acquisition unit and a data processing unit;

[0107] The second acquisition unit is used to monitor in real time the access behavior of the multi-dimensional data set of each second node, capture access events, and record all access logs through a database trigger or APIHook mechanism to establish an access data set;

[0108] The steps for establishing the access dataset are as follows:

[0109] S11. To statistically count the access volume from the log_access_table access log by time window, and obtain the access frequency F v : By statistically counting the access volume by time window, the access frequency of each node at different time periods can be accurately understood. This provides a necessary basis for subsequent data storage and access policy formulation. For example, data with high access frequency can be preferentially stored in more efficient storage devices to meet the requirements of fast access. This step helps to identify the popularity of nodes and provides decision support for hierarchical storage and data backup.

[0110] S12. And extract the timestamp t of each access of the multi-dimensional data set of each second node from the log_access_table access log i , and calculate the difference t between adjacent timestamps cz : t cz = t i+1 - t i ;

[0111] Calculate and obtain the average access interval T through the following formula a :

[0112]

[0113] In the formula, the difference t between adjacent timestamps cz represents the time difference between the i-th and (i + 1)-th accesses, t i represents the timestamp of the i-th access, and m represents the total number of access records in the log table; by calculating the difference between adjacent timestamps and obtaining the average access interval, the access frequency and pattern of each data set can be reflected. This indicator helps to distinguish frequently accessed nodes from nodes that are not accessed for a long time. With this information, storage management can be optimized to avoid excessive resource occupation by infrequently accessed data and improve the utilization efficiency of storage space.

[0114] S13. Extract the total access time length T and the accessed data volume S of the multi-dimensional data set of each second node data from the query_execution_time or data_transfer_size performance monitoring tool r and the accessed data volume S r :

[0115] The accessed data volume S rIndicates the amount of data actually read and written for each access to the second node data; by monitoring the total length of the access time and the amount of accessed data, the performance bottleneck of data access can be deeply understood. For example, the access to certain nodes may cause long delays or require a large amount of data transmission, which may affect the overall performance of the management platform. By collecting this data, it can help optimize the allocation of storage resources and optimize the data storage or caching strategy on nodes with heavy access.

[0116] S14. Extract the initial timestamp T of the multi-dimensional data set of each second node data from the log_access_table access log g and the current timestamp T C , and calculate and obtain the data decay coefficient I through the following formula d :

[0117] ΔT = T C - T g ;

[0118]

[0119] I d = e -λ*ΔT ;

[0120] In the formula, ΔT represents the time span from data creation to the current time, λ represents the data aging coefficient. Setting one month as the unit time, λ is a constant, which is obtained by derivation from the decay model based on historical access data experience through the least squares regression analysis method; log 10 represents the common logarithm with base 10 of natural numbers, e represents the base of the exponential function; F v represents the access frequency; by calculating the data decay coefficient, the life cycle and aging degree of data can be effectively measured, and then it can be determined whether it is necessary to archive or delete the aging data. The calculation of the data decay coefficient enables the management platform to automatically adjust the storage strategy according to the usage frequency and time span of data, thereby reducing the waste of storage space and improving the storage efficiency.

[0121] S15. Extract the actual number of access users U of each second node data from the log_access_table access log a and the total number of users U in the management platform T , and calculate the access complexity U r :

[0122]

[0123] By calculating the access complexity U r, the access complexity of each dataset and its requirements for the management platform resources can be understood. For example, high-complexity access may require more computing and storage resources, so it is necessary to allocate the management platform resources reasonably according to the access complexity. This step helps to identify the data and nodes that consume a large amount of resources in the management platform, and adjust the storage strategy in a timely manner to ensure the efficient use of resources.

[0124] S16. Statistically analyze the access frequency F obtained in S11 - S15 v , the average access interval T a , the total access time length T r , the amount of accessed data S r , the data attenuation coefficient I d and the access complexity U r , perform preprocessing through the data processing unit, use the data denoising algorithm to remove abnormal data, and uniformly process the data using the standardization technology. After obtaining the normalized data through the dimensionless technology, establish an access dataset.

[0125] In this embodiment, through steps S11 - S16, the second data acquisition module can deeply analyze the data access behavior and performance metrics of each node, helping the management platform to make more accurate and efficient decisions in data storage management. Specifically, through the collection and analysis of factors such as access frequency, amount of accessed data, and data attenuation in each step, the management platform can more intelligently optimize the storage strategy, improve data access efficiency, reduce storage costs, and provide necessary support for disaster backup and data protection.

[0126] Embodiment 4

[0127] This embodiment is an explanatory description based on Embodiment 1. Please refer to Figure 1 , specifically, the storage hierarchical management module includes a data analysis unit, a data classification unit, and a storage strategy unit;

[0128] The data analysis unit is used to extract the access frequency F in the access dataset v , the average access interval T a , the total access time length T r , the amount of accessed data S r , the data attenuation coefficient I d and the access complexity U r , and calculate the access heat index Rh through the following related formulas;

[0129]

[0130] Wherein, k1, k2, k3, and k4 represent weight coefficients, and the sum of the weight coefficients is 1; the access heat index can accurately evaluate the "heat" of data, that is, the access frequency and storage relevance of the data. Through a comprehensive analysis of the access heat, the management platform can better understand the importance of data in actual use, thereby providing a scientific basis for subsequent data hierarchical storage strategies.

[0131] The data classification unit is used to preset a high heat threshold H1, a medium heat threshold H2, and a low heat threshold H3, and the high heat threshold H1 > the medium heat threshold H2 > the low heat threshold H3; according to changes in business requirements, users can adjust these thresholds to more precisely manage data storage. For example, during certain periods, the access frequency of certain data may suddenly increase. By adjusting the heat threshold, it is possible to dynamically adapt to changes in the data access pattern.

[0132] And compare the access heat index Rh with the high heat threshold H1, the medium heat threshold H2, and the low heat threshold H3 respectively to obtain classification results, including:

[0133] If the access heat index Rh > the high heat threshold H1, generate a first-level label, indicating that the access frequency and relevance of the multi-dimensional data set of the current second node are very high, and it is core or key data;

[0134] If the medium heat threshold H2 ≤ the access heat index Rh ≤ the high heat threshold H1, generate a second-level label, indicating that the access frequency and relevance of the multi-dimensional data set of the current second node are within a medium range, neither the most frequently accessed data nor cold data, and belong to active data;

[0135] If the low heat threshold H3 ≤ the access heat index Rh < the medium heat threshold H2, generate a third-level label, indicating that the access frequency and relevance are relatively low, and it belongs to data that is not frequently accessed;

[0136] If the access heat index Rh < the low heat threshold H3, generate a fourth-level label, indicating that the access frequency and relevance of the current data set are very low, and it belongs to cold data.

[0137] The data classification unit classifies according to the access heat index Rh and the set heat thresholds (high heat threshold H1, medium heat threshold H2, and low heat threshold H3), and divides the data set into four levels: core data, active data, data that is not frequently accessed (infrequently used data), and cold data.

[0138] The storage strategy unit is used to generate a hierarchical storage strategy according to the classification result, including:

[0139] Generate the first stratification strategy based on the first-level label. Specifically: Store the multi-dimensional data set of the second node of the first-level label in a solid-state drive (SSD). This storage solution can ensure the fast reading and writing of frequently accessed data and improve the overall performance of the management platform.

[0140] Generate the second stratification strategy based on the second-level label. Specifically: Store the multi-dimensional data set of the second node of the second-level label in a traditional hard disk (HDD) or a network attached storage (NAS). This storage solution is suitable for data with a moderate access frequency, which can not only ensure a relatively fast access speed but also control the storage cost well.

[0141] Generate the third stratification strategy based on the third-level label. Specifically: Store the multi-dimensional data set of the second node of the third-level label in a cold storage cloud service, tape storage, or optical disc storage. For data with low data access requirements, using this storage solution can greatly reduce the storage cost while ensuring the long-term preservation of data.

[0142] Generate the fourth stratification strategy based on the fourth-level label. Specifically: Archive the multi-dimensional data set of the second node of the fourth-level label. Since the read requirements are extremely low, the storage cost rather than the access speed is mainly considered.

[0143] In this embodiment, during actual operation, the access pattern of data may change. The storage strategy unit can dynamically adjust the storage strategy according to the new access popularity and classification results. For example, if the access frequency of certain data suddenly increases, the management platform can migrate it from low-frequency storage to high-performance storage to ensure the efficiency of data access and the flexibility of the management platform. Through the access popularity index and the set threshold, the data can be accurately classified and the storage strategy can be determined. According to the changes in data access behavior, the storage strategy can be flexibly adjusted to ensure the long-term and efficient operation of the management platform. Using low-cost storage for low-frequency accessed data improves the overall storage efficiency.

[0144] Embodiment 5

[0145] This embodiment is an explanatory description based on Embodiment 1. Please refer to Figure 1 , specifically, the disaster backup strategy module includes a third collection unit, a disaster factor calculation unit, and a disaster prediction unit;

[0146] The third collection unit is used to collect disaster data sets from the upstream, midstream, and downstream regions of the first node;

[0147] The disaster data sets include earthquake magnitude (ZJ), focal depth (Sd), earthquake wave propagation speed (zsd), maximum reservoir water level (Sw), water flow speed (Ssd), landslide and debris flow volume (Lsh), landslide slope (Pd), wind speed (Fs), and hail diameter (BL);

[0148] The earthquake magnitude ZJ and the focal depth Sd are acquired through collection by a three-component seismograph;

[0149] The earthquake wave propagation velocity zsd is acquired through collection by an earthquake wave propagation sensor;

[0150] The maximum water level Sw of the reservoir is acquired through collection by a water level gauge;

[0151] The water flow velocity Ssd is acquired through collection by a current meter

[0152] The landslide and debris flow volume Lsh is obtained by scanning the upstream, midstream, and downstream areas of the first node with lidar to obtain three-dimensional terrain data. Before and after a landslide or debris flow event, lidar data is collected twice; by comparing the data from the two scans; using ground elevation difference analysis to identify the ground settlement or accumulation areas caused by landslides or debris flows to obtain the difference areas, and using three-dimensional geometric algorithms, including the cross-section method or the volume extrapolation method, to calculate and obtain the difference areas;

[0153] The landslide slope Pd is obtained by analyzing the difference area to obtain the height information of each pixel in the area. The landslide slope is equal to the ratio of the height change to the horizontal distance, and the landslide slope is calculated and obtained;

[0154] The wind speed Fs is acquired through collection by an anemometer;

[0155] The hail particle size BL is acquired by scanning the meteorological cloud layer with a radar scanner to detect hail particles in the meteorological phenomenon and collecting through the reflected signal, or by using a laser particle size analyzer for collection.

[0156] The disaster factor calculation unit is used to remove abnormal data from the disaster data set by adopting a data denoising algorithm, and uniformly process the data by using a standardization technique. After obtaining the normalized data through a dimensionless technique, the earthquake magnitude ZJ, the focal depth Sd, the earthquake wave propagation velocity zsd, the maximum water level Sw of the reservoir, the water flow velocity Ssd, the landslide and debris flow volume Lsh, the landslide slope Pd, the wind speed Fs, and the hail particle size BL are extracted, and the disaster risk index Gr is calculated and obtained through the following formula:

[0157] F eq1 = W1*Sd + W2*Sd + W3*zsd;

[0158] F eq2 = W4*Sw + W5*Ssd;

[0159] F eq3 = W6*Lsh + W7*Pd;

[0160] F eq4 = W8*Fs + W9*BL;

[0161] Gr = α * F eq1 + β * F eq2 + γ * F eq3 + δ * F eq4 ;

[0162] Wherein, F eq1 represents the earthquake disaster factor, F eq2 represents the flood disaster factor, F eq3 represents the mountain disaster factor, F eq4 represents the meteorological disaster factor, W1, W2, and W3 respectively represent the weight coefficients of the earthquake magnitude ZJ, the focal depth Sd, and the seismic wave propagation velocity zsd; W4 and W5 respectively represent the weight coefficients of the maximum reservoir water level Sw and the water flow velocity Ssd; W6 and W7 respectively represent the weight coefficients of the landslide and debris flow volume Lsh and the landslide slope Pd; W8 and W9 respectively represent the weight coefficients of the wind speed Fs and the hail diameter BL; α, β, γ, and δ respectively represent the earthquake disaster factor F eq1 , the flood disaster factor F eq2 , the mountain disaster factor F eq3 , and the meteorological disaster factor F eq4 , and the sum of the weight coefficients is 1. The calculation of the disaster factors (such as earthquakes, floods, mountain disasters, meteorological disasters) comprehensively considers the weights of multiple factors, and has strong flexibility and adaptability. By adjusting the weighting coefficients, the disaster risk assessment model can be customized according to different regions and conditions. The calculation of the disaster risk index Gr provides a quantitative basis for disaster prediction, enabling the management platform to dynamically adjust the backup strategy based on real-time data.

[0163] The disaster prediction unit is used to preset a disaster risk threshold Q, and compare the disaster risk index Gr with the disaster risk threshold Q to obtain a disaster prediction result, including:

[0164] When the disaster risk index Gr ≥ the disaster risk threshold Q, it indicates that there is a disaster risk in the first node area, and identify whether the current first node area is upstream, midstream, or downstream, and generate a corresponding backup strategy;

[0165] When the disaster risk index Gr < the disaster risk threshold Q, it indicates that there is no disaster risk in the first node area;

[0166] The corresponding backup strategy includes:

[0167] Identify the first node area. If it is upstream, implement the first backup strategy, including: on the basis of the hierarchical storage strategy, increase the backup redundancy ratio of the multi-dimensional data set of the second node in the first node upstream by more than 80% to ensure data security because the upstream area is more affected by disasters such as landslides and debris flows;

[0168] Identify the first node area. If it is the middle reaches, implement the second backup strategy, including: on the basis of the hierarchical storage strategy, increase the backup redundancy ratio of the multi-dimensional data set of the second node in the first node in the middle reaches to 50%-80%; compared with the upper reaches area, the middle reaches area is relatively less affected by disasters, but data backup still needs to be considered in light of water level changes and meteorological conditions;

[0169] Identify the first node area. If it is the lower reaches, implement the third backup strategy, including: increase the backup redundancy ratio of the multi-dimensional data set of the second node in the first node in the upper reaches to less than 40%. The lower reaches area is usually affected by water level, floods, and climate change, etc., and the disaster risk is relatively low, but data backup for water flow changes still needs to be carried out. The backup strategy is simplified, but it still needs to be regularly uploaded to the cloud platform or backed up remotely to prevent data loss caused by sudden disasters.

[0170] In this embodiment, according to the results of the disaster risk assessment, the management platform divides the first node (area) into three areas: the upper reaches, the middle reaches, and the lower reaches according to the degree of disaster risk, and sets different backup redundancy ratios for each area. This differentiated backup strategy helps to reasonably allocate resources, ensure more protection for data in high-risk areas, and at the same time reduce the storage and operation and maintenance costs of low-risk areas. The platform has strong adaptability and can adjust the backup strategy in real time according to changes in disaster risks to ensure the response capabilities in different disaster types and different regions. The management platform can identify disaster risks in advance and generate emergency backup strategies, greatly improving the efficiency and flexibility in dealing with disaster events. Especially when disasters such as earthquakes, floods, and landslides occur, making data backups in advance can prevent disasters from damaging or losing data and ensure the recoverability of key data.

[0171] Embodiment 6

[0172] This embodiment is an explanatory description based on Embodiment 1. Please refer to Figure 1 Specifically, it further includes an offline security policy module, and the offline security policy module includes an identification unit, a recovery unit, and a synchronization policy unit;

[0173] The identification unit is used for the network card status monitoring device to monitor the network interface status. If the network card is in the "disconnected" or "off" state, it can be judged as offline. When the offline state is identified, enable the local data security storage mechanism, perform real-time backup of the multi-dimensional data set of the second node through the local area network, and collect the backup database; enable the local data security storage mechanism. This design ensures that data can be immediately backed up during network disconnection to prevent data loss.

[0174] And based on the backup database, generate the local storage integrity index Cl through the following formula:

[0175]

[0176] Wherein, Wc s represents the amount of valid data successfully backed up and stored during the network outage, and Wc y represents all the data that should be stored locally during the network outage; by defining the local storage integrity index (Cl), the system can measure the ratio between the amount of data successfully backed up during the network outage and all the data that should be stored locally. This calculation formula can quantify the integrity of local storage, thus helping to evaluate the secure storage of data during network interruption.

[0177] The recovery unit is used to synchronize and update data with the remote database based on the local storage integrity index Cl when the network is restored, and generate a synchronization consistency coefficient Sc through the following formula:

[0178]

[0179] Wherein, Bc s represents the amount of data that has not been successfully synchronized between local storage and the remote database; Zc y represents all the data that should be synchronized. The closer the value of the synchronization consistency coefficient Sc is to 1, the higher the reliability of synchronization, and the smaller the error or delay of data synchronization. After the network is restored, the recovery unit will coordinate the synchronization update with the remote database according to the local storage integrity index (Cl). The calculation formula of the synchronization consistency coefficient (Sc) measures the amount of data that has not been synchronized between local storage and the remote database, thus evaluating the reliability of synchronization.

[0180] The synchronization strategy unit is used to further evaluate the synchronization consistency coefficient Sc, including:

[0181] When 0.8 < synchronization consistency coefficient Sc < 1, trigger the manual synchronization method, trigger the manual synchronization process, and resynchronize the data that has not been successfully synchronized; this solution provides a flexible synchronization management method, which is suitable for high-reliability requirements of data synchronization.

[0182] If the synchronization consistency coefficient Sc ≤ 0.8, trigger the incremental synchronization method, and only synchronize the data that has changed since the network outage, so as to reduce the synchronization time and pressure. Only synchronize the data that has changed since the network outage. This strategy reduces the time and resource pressure during the synchronization process, and at the same time ensures the efficiency and accuracy of data synchronization, especially when the data volume is large.

[0183] In this embodiment, the incremental synchronization mode reduces the amount of data synchronization, avoids unnecessary redundant data transmission, reduces the bandwidth consumption and storage burden of the system, and improves the synchronization efficiency. Through flexible synchronization methods (manual synchronization and incremental synchronization), the system can select the most suitable synchronization strategy under different network recovery conditions, ensuring the stability of the system and the high efficiency of data synchronization. The design of this module provides strong fault tolerance. In the case of network disconnection, the system can automatically enable local backup to ensure data security; after the network is restored, the flexible synchronization strategy can ensure the consistency of data under different network recovery conditions and avoid losing key information due to synchronization failure.

[0184] The setting of the threshold value is for the convenience of comparison. Regarding the size of the threshold value, it depends on the amount of sample data and the base quantity set by those skilled in the art for each group of sample data; as long as the proportional relationship between the parameter and the quantized value is not affected.

[0185] The above formulas are all obtained by collecting a large amount of data for software simulation and selecting a formula close to the true value. The coefficients in the formulas are set by those skilled in the art according to the actual situation. As mentioned above, the above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution of the present invention and its inventive concept, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. A multi-level data security and storage management platform for the hydropower industry, characterized in that: include: The first data collection module is used to collect multi-dimensional data sets of each node in the hydropower station; The second data collection module is used to monitor the access behavior of the multidimensional data set of each node in real time, capture access events, and record all access logs through database triggers or APIHook mechanisms to establish access data sets; The storage tier management module is used to analyze the access data set, build the access heat index Rh of each node, classify it after evaluation, and perform storage management according to the corresponding tier strategy: The disaster backup strategy module is used to collect real-time geological disaster data, construct and evaluate the disaster risk index Gr to obtain disaster assessment results, and optimize the stratified strategy based on the disaster assessment results to form a stratified backup plan.

2. A multi-level data security and storage management platform for the hydropower industry according to claim 1, characterized in that: The first data acquisition module includes a node division unit and a first acquisition unit; The node division unit is used to divide the hydropower station according to region and function, and each node corresponds to a multidimensional data set of a region; The specific steps for dividing nodes are: First, the hydropower station is divided into the upstream, midstream and downstream nodes, and the data of the generator sets in each first node are collected. Then, each generator set is divided into the second node and marked as generator sets A1, A2, ..., A n , n represents the number of generating sets; The first acquisition unit is used to acquire a multidimensional data set of the second node, wherein the multidimensional data set includes equipment operation data of each generator set, hydrological data connected to each generator set, and environmental data of the area where each hydropower unit is located.

3. A multi-level data security and storage management platform for the hydropower industry according to claim 1, characterized in that: The second data acquisition module includes a second acquisition unit and a data processing unit; The second acquisition unit is used to monitor the access behavior of the multidimensional data set of each second node in real time, capture access events, and record all access logs through database triggers or APIHook mechanisms to establish access data sets; The steps for establishing the access data set are as follows: S11, used to obtain the access frequency F from the log_access_table access log by time window and count the access volume v : S12, extracting the timestamp t of each access to the multidimensional data set of each second node from the log_access_table access log i , calculate the difference t between adjacent timestamps cz :t cz =t i+1 -t i ; The average access interval T is calculated by the following formula a : In the formula, the difference between adjacent timestamps is cz represents the time difference between the i-th and i+1-th visits, t i represents the timestamp of the i-th access, and m represents the total number of accesses recorded in the log table; S13, extracting the total access time length T of the multidimensional data set of each second node data from the query_execution_time or data_transfer_size performance monitoring tool r And the amount of data accessed S r : Access data volume S r Indicates the amount of data actually read and written in a single access to each second node data; S14: extracting the initial timestamp T generated by the multidimensional data set of each second node data from the log_access_table access log g and the current timestamp T C The data attenuation coefficient I is calculated by the following formula d : ΔT=T C -T g ; I d =e -λ*ΔT ; In the formula, ΔT represents the time span from data creation to the current time, λ represents the data aging coefficient, and the time unit is set to one month. λ is a constant, and the decay model is derived based on the historical access data experience through the least squares regression analysis method; log 10 represents the common logarithm with natural number 10 as base, e represents the base of exponential function; F v Indicates the frequency of access; S15: Extract the number of users who actually accessed each second node data from the log_access_table access log U. a And the total number of users U in the management platform T , calculate the access complexity U r : S16. Count the access frequency F obtained from S11-S15 v , average access interval T a , Total access time length T r , access data volume S r , data attenuation coefficient I d And the access complexity U r ,Preprocessing is performed through the data processing unit, the data denoising algorithm is used to remove abnormal data, and the data is processed uniformly using standardization technology. After obtaining normalized data through dimensionless technology, the access data set is established.

4. A multi-level data security and storage management platform for the hydropower industry according to claim 1, characterized in that: The storage hierarchical management module includes a data analysis unit, a data classification unit and a storage strategy unit; The data analysis unit is used to extract the access frequency F in the access data set. v , average access interval T a , Total access time length T r , access data volume S r , data attenuation coefficient I d And the access complexity U r , the access heat index Rh is calculated by the following associated formula; Wherein, k1, k2, k3 and k4 represent weight coefficients, and the sum of weight coefficients is 1; The data classification unit is used to preset a high heat threshold H1, a medium heat threshold H2 and a low heat threshold H3, and the high heat threshold H1> the medium heat threshold H2> the low heat threshold H3, and compare the access heat index Rh with the high heat threshold H1, the medium heat threshold H2 and the low heat threshold H3 respectively to obtain a classification result, including: If the access heat index Rh> the high heat threshold H1, a first-level label is generated, indicating that the access frequency and relevance of the multidimensional data set of the current second node are very high, and it is core or key data; If the medium heat threshold H2≤access heat index Rh≤high heat threshold H1, a second-level label is generated, indicating that the access frequency and relevance of the multidimensional data set of the current second node are in the medium range, which is neither the most frequently accessed data nor the unpopular data, but active data; If the low heat threshold H3 ≤ access heat index Rh < medium heat threshold H2, a third-level label is generated, indicating that the access frequency and relevance are low, and the data is not frequently accessed; If the access heat index Rh is less than the low heat threshold H3, a fourth level label is generated, indicating that the access frequency and relevance of the current data set are very low and belong to cold data.

5. A multi-level data security and storage management platform for the hydropower industry according to claim 4, characterized in that: The storage strategy unit is used to generate a hierarchical storage strategy according to the classification result, including: Generating a first stratification strategy according to the first level label, specifically: storing the multidimensional data set of the second node of the first level label in a solid state drive SSD; Generating a second tiering strategy according to the second level label, specifically: storing the multidimensional data set of the second node of the second level label in a traditional hard disk HDD or a network storage device NAS; Generating a third tiering strategy according to the third level label, specifically: storing the multidimensional data set of the second node of the third level label in a cold storage cloud service, tape storage or optical disk storage; The fourth tiering strategy is generated based on the third level label, specifically: the multidimensional data set of the second node of the fourth level label is archived. Since the reading demand is extremely low, the focus is on storage cost rather than access speed.

6. A multi-level data security and storage management platform for the hydropower industry according to claim 5, characterized in that: The disaster backup strategy module includes a third acquisition unit, a disaster factor calculation unit and a disaster prediction unit; The third collection unit is used to collect disaster data sets for the upstream, midstream and downstream areas of the first node; The disaster data set includes earthquake magnitude ZJ, focal depth Sd, seismic wave propagation speed zsd, maximum reservoir water level Sw, water flow speed Ssd, landslide and debris flow volume Lsh, landslide slope Pd, wind speed Fs and hail particle size BL; The earthquake magnitude ZJ and focal depth Sd are acquired by three-component seismograph; The seismic wave propagation velocity zsd is acquired by collecting through a seismic wave propagation sensor; The maximum water level Sw of the reservoir is obtained by collecting data through a water level meter; The water flow velocity Ssd is acquired by a flow meter The landslide and debris flow volume Lsh is obtained by scanning the upstream, midstream and downstream areas of the first node by laser radar, and three-dimensional terrain data is obtained. Before and after the landslide or debris flow event, laser radar data are collected twice respectively; by comparing the two scanned data; using ground elevation difference analysis, the ground settlement or accumulation area caused by the landslide or debris flow is identified to obtain the difference area, and the difference area is calculated and obtained using a three-dimensional geometric algorithm, including a cross-section method or a volume extrapolation method; The landslide slope Pd is obtained by analyzing the difference area to obtain the height information of each pixel in the area. The landslide slope is equal to the ratio of the height change to the horizontal distance, and the landslide slope is calculated; The wind speed Fs is acquired by an anemometer; The hail particle size BL is obtained by detecting hail particles in meteorological phenomena through a radar scanner of meteorological clouds and collecting reflected signals, or by using a laser particle size meter.

7. A multi-level data security and storage management platform for the hydropower industry according to claim 6, characterized in that: The disaster factor calculation unit is used to remove abnormal data from the disaster data set using a data denoising algorithm, and to uniformly process the data using a standardization technique. After obtaining normalized data using a dimensionless technique, the earthquake magnitude ZJ, focal depth Sd, seismic wave propagation velocity zsd, maximum reservoir water level Sw, water flow velocity Ssd, landslide and debris flow volume Lsh, landslide slope Pd, wind speed Fs and hail particle size BL are extracted, and the disaster risk index Gr is calculated using the following formula: <h2 style=";text-align:left;direction:ltr">F<h2 style=";text-align:left;direction:ltr"> eq1 <h2 style=";text-align:left;direction:ltr"> (W1*Sd+W2*Sd+W3*zsd) F eq2 =W4*Sw+W5*Ssd; <h2 style=";text-align:left;direction:ltr">F<h2 style=";text-align:left;direction:ltr"> eq3 <h2 style=";text-align:left;direction:ltr"> (W6*Lsh+W7*Pd) F eq4 =W8*Fs+W9*BL; Gr=α*F eq1 +β*F eq2 +γ*F eq3 +δ*F eq4 ; In the formula, F eq1 represents the earthquake hazard factor, F eq2 represents the flood hazard factor, F eq3 represents the mountain disaster factor, F eq4 represents meteorological disaster factors, W1, W2 and W3 represent the weight coefficients of earthquake magnitude ZJ, focal depth Sd and seismic wave propagation speed zsd respectively; W4 and W5 represent the weight coefficients of maximum reservoir water level Sw and water flow speed Ssd respectively; W6 and W7 represent the weight coefficients of landslide and debris flow volume Lsh and landslide slope Pd respectively; W8 and W9 represent the weight coefficients of wind speed Fs and hail particle size BL respectively; α, β, γ and δ represent earthquake disaster factors F eq1 , flood disaster factor F eq2 , Mountain disaster factor F eq3 and meteorological disaster factor F eq4 The weight coefficients of , and the sum of the weight coefficients is 1.

8. A multi-level data security and storage management platform for the hydropower industry according to claim 7, characterized in that: The disaster prediction unit is used to preset a disaster risk threshold Q, and compare the disaster risk index Gr with the disaster risk threshold Q to obtain a disaster prediction result, including: When the disaster risk index Gr ≥ the disaster risk threshold Q, it means that there is a disaster risk in the first node area, and the current first node area is identified as upstream, midstream or downstream, and the corresponding backup strategy is generated; When the disaster risk index Gr is less than the disaster risk threshold Q, it means that there is no disaster risk in the first node area; The corresponding backup strategies include: Identify the first node area, and if it is upstream, implement the first backup strategy, including: based on the hierarchical storage strategy, increase the backup redundancy ratio of the multidimensional data set of the second node in the upstream first node to more than 80%; Identify the first node area, and if it is midstream, implement a second backup strategy, including: based on the hierarchical storage strategy, increase the backup redundancy ratio of the multidimensional data set of the second node in the first node of the midstream to 50%-80%; The first node area is identified, and if it is downstream, a third backup strategy is implemented, including: increasing the backup redundancy ratio of the multidimensional data set of the second node in the upstream first node to less than 40%.

9. A multi-level data security and storage management platform for the hydropower industry according to claim 1, characterized in that: It also includes a network disconnection security policy module, which includes an identification unit, a recovery unit and a synchronization policy unit; The identification unit is used for the network card status monitoring device to monitor the network interface status. If the network card is in a "no connection" or "disconnected" state, it can be determined that the network is disconnected. When the network disconnection state is identified, the local data security storage mechanism is enabled, and the multi-dimensional data set of the second node is backed up in real time through the local area network to collect the backup database; Based on the backup database, the local storage integrity index Cl is generated by the following formula: Where Wc s Indicates the amount of valid data successfully backed up and stored during the network outage, Wc y Indicates the amount of data that should be stored locally during network outage; The recovery unit is used to synchronize and update data with the remote database based on the local storage integrity index Cl when the network is restored, and generate a synchronization consistency coefficient Sc by the following formula: In the formula, Bc s Indicates the amount of data that was not successfully synchronized between the local storage and the remote database; Zc y Indicates the total amount of data that should be synchronized.

10. A multi-level data security and storage management platform for the hydropower industry according to claim 9, characterized in that: The synchronization strategy unit is used to further evaluate the synchronization consistency coefficient Sc, including: When 0.8<synchronization consistency coefficient Sc<1, the manual synchronization mode is triggered, triggering the manual synchronization process by resynchronizing the data that has not been successfully synchronized; If the synchronization consistency coefficient Sc≤0.8, the incremental synchronization method is triggered, and only the data that has changed since the network disconnection is synchronized.

Images