Multi-source heterogeneous data fusion and data preprocessing method and system based on cloud computing

Abstract

The invention provides a multi-source heterogeneous data fusion and data preprocessing method and system based on cloud computing, belongs to the field of water resource digital twinhole wild river basin construction, and comprises the steps of data collection, data processing, data fusion, data backup and optimization, data safety management and the like. In the data processing process, the integrity, accuracy and timeliness of the data are checked and evaluated, and data cleaning and data conversion are carried out. For data fusion, a graph theory method is adopted, a data relation graph is constructed, and graph convolution operation is carried out. In the aspect of data management, backup, optimization and safety management of data are realized. And finally, completing data storage by establishing a water conservancy thematic database. On this basis, a plurality of sub-databases are established according to different data attributes. According to the system, elastic and extensible computing resources are provided through cloud computing, large-scale data fusion and preprocessing are achieved, and data quality and processing efficiency are improved.

Description

Technical Field

[0001] The present invention belongs to the field of water resources digital twin Kuye River Basin construction, and more specifically to a multi-source heterogeneous data fusion and data preprocessing method and system based on cloud computing. Background Art

[0002] In the field of water conservancy, a large amount of data of various types has emerged, including geospatial data, basic data, monitoring data, business management data, and external shared data. Due to the complex environmental background, diversified data types, and ultra-large data volumes, the processing, fusion, and preprocessing of multi-source heterogeneous data pose a major technical challenge in the field of water conservancy.

[0003] In existing technologies, data fusion and preprocessing are mainly performed manually or semi-automatically, which is time-consuming, labor-intensive, error-prone, and cannot meet the needs of the big data era. Moreover, when it comes to large-scale heterogeneous data, existing data processing methods often become infeasible due to limitations in computing power and storage capacity.

[0004] Cloud computing-based solutions are considered to be an effective way to process large-scale heterogeneous data. Cloud computing provides elastic and scalable computing resources and services that can process and store large amounts of data. However, how to effectively utilize the advantages of cloud computing to design and implement a method and system that can process, fuse and preprocess multi-source heterogeneous data in the field of water conservancy is still a technical problem to be solved. Summary of the invention

[0005] The present invention proposes a multi-source heterogeneous data fusion and data preprocessing method and system based on cloud computing, aiming to achieve effective processing and fusion of large-scale, multi-source heterogeneous data, so as to improve the quality and effectiveness of data and meet the needs of modern water conservancy business management and services. Gather data and build a database covering geospatial data, basic data, monitoring data, business management data and external shared data, thereby realizing map services, data resource directory services, data sharing services and data management and control services.

[0006] In order to achieve the above object, the present invention is implemented by adopting the following technical scheme: the method comprises:

[0007] Data collection: automatically collect geospatial data, basic data, monitoring data, business management data and external shared data through APIs and Webhooks interface technology;

[0008] Data processing, data quality assessment, checking data completeness, accuracy and timeliness, data cleaning and data conversion;

[0009] Data fusion, integrating data from multiple sources;

[0010] Database construction and data storage, establishment of water conservancy thematic database and data management strategy.

[0011] In one embodiment, the geospatial data includes DOM digital orthophotos, DEM digital elevation models, oblique photography models, underwater terrain, and BIM building information models;

[0012] Basic data include main attribute data and spatial data of watersheds, rivers, lakes, water conservancy projects, and hydrological objects;

[0013] Monitoring data include rainfall, water conditions, construction conditions, water quality, sediment, disaster conditions, groundwater level, water intake, soil moisture conditions, water conservancy project safety operation monitoring, and video data;

[0014] Business management data include basin flood and drought disaster management business application data;

[0015] Externally shared data include economic, social, land use, ecological environment, meteorological, and remote sensing related data that need to be shared from other industry sectors.

[0016] The time reference of the data is Beijing time, the space reference is the 2000 National Geodetic Coordinate System (CGCS2000), and the elevation reference is the 1985 National Elevation Datum.

[0017] In one embodiment, the data processing includes:

[0018] Data quality assessment: For spatial data, evaluate the accuracy of its spatial reference system; for monitoring data, check the frequency and accuracy of data collection to determine its timeliness and accuracy.

[0019] Data completeness = (1-proportion of missing values ​​in the data) * 100%;

[0020] Missing value ratio = number of missing values ​​in the data set / total number of data points in the data set;

[0021] Data cleaning: For text data, perform stemming and stop word processing; for time series data, process missing values ​​and outliers;

[0022] Data conversion: spatial data is transformed to match a unified spatial reference system; monitoring data is normalized to eliminate the impact of dimension; external shared data can be converted into data format through APIs.

[0023] In one solution, the data fusion constructs a data relationship graph and uses graph theory to perform data fusion. The specific steps are as follows:

[0024] Build a graph structure: Build a graph based on the data. Each node can represent a specific management business object, such as a disaster event or a location. Each edge represents the relationship between two business objects, such as the time sequence or the distance between locations.

[0025] Initialize node representation: Then initialize the representation of each node. This node representation can be some attribute values ​​related to the node.

[0026] Then, a graph convolution operation is performed to update the representation of a node to the weighted average of its own representation and the representations of all its neighbors;

[0027] Data fusion: After performing several graph convolutions, the final representation of each node is obtained. This representation contains both the characteristics of the node itself and the relationship information with other nodes. These nodes are then used for data fusion.

[0028] In one embodiment, the data management strategy includes:

[0029] Data backup: a data backup is performed once a day and a full backup is performed once a week. The backup data is stored on redundant cloud storage devices to prevent primary storage failures.

[0030] Data optimization: based on the access frequency and access speed, frequently accessed data is stored on fast but expensive SSD hard disks, while infrequently accessed data is stored on slower but cheaper HDD hard disks.

[0031] Data security management includes encrypting stored and transmitted data, verifying the identity and permissions of users accessing data, and conducting regular security audits of the system.

[0032] On the other hand, a multi-source heterogeneous data fusion and data preprocessing system based on cloud computing is applicable to the method described above, and the system includes:

[0033] Data collection module, used to automatically acquire geospatial data, basic data, monitoring data, business management data and external shared data, and unify them into Beijing time and CGCS2000 national geodetic coordinate system as the standard;

[0034] Data cleaning module, used to check the integrity, accuracy and timeliness of data, and remove redundant, repeated or invalid information in the data;

[0035] Data fusion module, used to fuse data from various heterogeneous data sources, including data conversion and data integration;

[0036] Data service module, used to provide data query, display, sharing and management services;

[0037] Water conservancy database, storing structured and unstructured data;

[0038] The basin basic spatial database adopts a hierarchical database architecture, which divides the data into the original data layer, basic data layer and application data layer.

[0039] In one embodiment, the water conservancy database includes: water resources database, flood database, water conservancy project database, land use and soil and water conservation database, hydrological and meteorological database and water resources economic database.

[0040] In one embodiment, the system further includes: a metadata management module that provides metadata definition, collection, entry, query and retrieval services.

[0041] Beneficial effects of the present invention:

[0042] The multi-source heterogeneous data fusion and data preprocessing method and system based on cloud computing of the present invention can effectively realize the fusion of multi-source heterogeneous data, reduce data redundancy, and improve the application value of data. By implementing data preprocessing, including data cleaning and data verification, not only the quality of data is improved, the accuracy and integrity of data are enhanced, but also a reliable data basis is provided for data analysis and decision-making. At the same time, by using cloud computing technology, large-scale parallel processing of computing and storage tasks can be carried out, which greatly improves data processing efficiency and saves processing time. In addition, data fusion and preprocessing based on cloud computing can also realize rapid data sharing and collaboration, improve the work efficiency among various departments, and strengthen the management and decision-making of water resources, floods, ecological environment, etc. Finally, the invention method and system provide important support for big data applications in the field of water conservancy, such as flood disaster warning, water resources allocation, environmental protection, etc., which helps to promote the progress of water conservancy science and technology and social and economic development. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 is a flow chart of the method of the present invention;

[0044] Figure 2 This is a data fusion flow chart of the present invention;

[0045] Figure 3 This is a system block diagram of the present invention. DETAILED DESCRIPTION

[0046] In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. Typical embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described in the present invention. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0047] Unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as those understood by those skilled in the art of the present invention. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0048] Embodiment 1:

[0049] like Figure 1 As shown in FIG. 1 , the multi-source heterogeneous data fusion and data preprocessing method based on cloud computing includes:

[0050] S1. Data collection: Automatically collect geospatial data, basic data, monitoring data, business management data and external shared data through APIs and Webhooks interface technologies.

[0051] The data include geospatial data, basic data, monitoring data, business management data and external shared data. Among them, geospatial data is the focus of the construction of data twin basins, mainly including DOM (Digital Orthophoto Map), DEM (Digital Elevation Model), oblique photography model, underwater terrain, BIM (Building Information Model) and other contents; basic data include the main attribute data and spatial data of various water conservancy objects such as basins, rivers, lakes, water conservancy projects, hydrology, etc.; monitoring data include rainfall conditions, water conditions, construction conditions, water quality, sediment, disaster conditions, groundwater levels, water use, soil moisture conditions, water conservancy project safety operation monitoring, videos and other data; business management data include basin water and drought disaster management business application data; external shared data include economic and social, land use, ecological environment, meteorology, remote sensing and other related data that need to be shared from other industry departments. The time reference of the data should be Beijing time, the spatial reference should be the 2000 National Geodetic Coordinate System (CGCS2000), and the elevation reference should be the 1985 National Elevation Datum.

[0052] The main methods involved in the data collection process are as follows: With the help of various data interface technologies such as APIs, Webhooks, etc., data collection and access can be automated.

[0053] S2. Data processing: conduct data quality assessment, check data completeness, accuracy and timeliness, perform data cleaning and data conversion.

[0054] Data quality assessment: For spatial data, evaluate the accuracy of its spatial reference system; for monitoring data, check the frequency and accuracy of data collection to determine its timeliness and accuracy.

[0055] Data completeness = (1-proportion of missing values ​​in the data) * 100%;

[0056] Missing value ratio = number of missing values ​​in the dataset / total number of data points in the dataset.

[0057] Data cleaning: Heterogeneous data requires cleaning techniques. For example, for text data, stemming and stopword processing are required; for time series data, missing values ​​and outliers need to be processed.

[0058] Data conversion: spatial data is transformed to match a unified spatial reference system; monitoring data is normalized to eliminate the impact of dimension; external shared data can be converted into data format through APIs.

[0059] S3, data fusion, integrates multi-source data together. The basic idea of ​​using graph theory fusion is to view data as a graph structure, where each node represents a data object and each edge represents the relationship between two data objects. Then, through graph theory methods, graph convolutional networks (GCNs), we can obtain the embedded representation of each node. This embedded representation will take into account the characteristics of the node itself and its relationship with other nodes, and then use these embedded representations to perform data fusion. Figure 2 As shown, the following are the specific implementation steps:

[0060] S301. Construct a graph structure: First, construct a graph based on the business management data and the watershed flood and drought disaster management business application data. Each node can represent a specific management business object (such as a disaster event, a location, etc.), and each edge can represent the relationship between two business objects (such as the time sequence, the distance relationship of the location, etc.).

[0061] S302, initializing node representation: then initializing the representation of each node, and this node representation may be some attribute values ​​related to the node.

[0062] Geospatial data: including DOM digital orthophotos, DEM digital elevation models, BIM building information models, etc. These data can be used to define the geometric structure and spatial properties of the map.

[0063] Basic data: includes main attribute data and spatial data such as watersheds, rivers, lakes, and water conservancy projects. These data can be used to define the nodes and node attributes of the graph.

[0064] Monitoring data: including rainfall, water conditions, construction conditions, water quality, sediment, disaster conditions, groundwater level, water use, soil moisture and other data. These data can further enrich the attributes of graph nodes or be used to define the weights of edges between nodes.

[0065] Business management data: basin flood and drought disaster management business application data. This data can be used to define the weights of graph edges or to mark specific nodes and edges.

[0066] External shared data: data related to economy and society, land use, ecological environment, meteorology, and remote sensing. This data can be used to further enrich the attributes of nodes and edges.

[0067] S303, then perform a graph convolution operation. The basic idea of ​​graph convolution is to update the representation of a node to the weighted average of its own representation and the representations of all its neighbors (the specific weighting method can be in many ways). In this way, the new representation of each node will take into account the information of other nodes that are related to it. In the graph convolution network, the graph convolution operation can be expressed as:

[0068] H {(l+1)} =ReLU(D {-1 / 2} AD {-1 / 2} H {(l)} W {(l)} )

[0069] Where A is the adjacency matrix of the graph, D is the diagonal matrix formed by the sum of the elements in each row of A, and D {-1 / 2} AD {-1 / 2} It is to normalize the original graph adjacency matrix A to eliminate the influence of node degree. {(l)} is the feature representation of each node at layer l, W {(l)} is the weight matrix for layer l. ReLU is an activation function that aims to give nonlinearity in the convolution calculation of each layer.

[0070] After such calculation operation, we will get the new node feature representation H {(l+1)} , that is, a graph convolution operation is performed.

[0071] Through the above steps, data fusion is achieved. The idea of ​​graph theory takes into account the complex relationship between data, and through graph convolution, the new expression of each node gathers information about its surrounding environment, achieving the goal of using complex relationships to identify and fuse multi-source data. The final representation of each node is the fused data.

[0072] The data fusion method based on graph theory has the following advantages in the present invention:

[0073] Graph fusion method truly considers the intrinsic connection and relationship between different data. The concepts of nodes and edges can naturally express the dependency and interaction between multi-source data.

[0074] When faced with various types of complex data including geospatial data, basic data, monitoring data, business management data and external shared data, graph fusion methods can handle them better without losing the characteristics of the data source itself.

[0075] Through methods such as graph convolutional networks (GCN), the new expression of each node incorporates information about its surrounding environment, obtaining a richer feature expression than the original feature. This makes the fused data more representative and valuable for reference.

[0076] Traditional data fusion methods mostly adopt simple fusion methods, such as mean, maximum and minimum, voting, etc., while the present invention is a brand-new method by introducing graph theory and using graph convolutional networks for fusion. It is innovative to fuse large-scale heterogeneous data covering multiple fields such as geography, monitoring, and management. In the process of data fusion, economic and social data shared from other industries and sectors are taken into consideration. This mode of data intercommunication and sharing is advanced. Graph convolutional networks can capture and utilize complex dependencies between data, which plays an important role in improving the prediction and classification performance of the model. For such complex heterogeneous data, graph theory can be used to discover and utilize these relationships.

[0077] S4. Database construction and data storage: establish a water conservancy database and data management strategy.

[0078] Data backup: a data backup is performed once a day and a full backup is performed once a week. The backup data is stored on redundant cloud storage devices to prevent primary storage failures. Data optimization: based on the access frequency and access speed, frequently accessed data is stored on fast but expensive SSD hard drives, and infrequently accessed data is stored on slower but cheaper HDD hard drives. Data security management: encryption of stored and transmitted data, identity and authority verification of users accessing data, and regular security audits of the system.

[0079] Embodiment 2:

[0080] Based on the first embodiment, a multi-source heterogeneous data fusion and data preprocessing system based on cloud computing is constructed.

[0081] like Figure 3 As shown, the system comprises:

[0082] The data collection module is used to automatically acquire geospatial data, basic data, monitoring data, business management data and external shared data, and unify them into Beijing time and CGCS2000 national geodetic coordinate system as the standard.

[0083] The data cleaning module is used to check the integrity, accuracy and timeliness of the data and remove redundant, repeated or invalid information from the data.

[0084] The data fusion module is used to fuse data from various heterogeneous data sources, including data conversion and data integration.

[0085] The data service module is used to provide data query, display, sharing and management services.

[0086] The data sharing service aims to promote data sharing and exchange among water administration departments at all levels and realize cross-departmental and cross-field data applications. Its implementation plan includes: formulating data sharing policies, clarifying the scope, authority and usage rules of data sharing; realizing cross-departmental data sharing, establishing a data exchange mechanism, and ensuring the security and accuracy of data transmission; setting data access rights and controls to protect data privacy and confidentiality. By formulating unified data exchange standards and interfaces, data sharing services can achieve seamless docking and sharing of data. Water administration departments at all levels can report and share data in a timely and accurate manner, promote the circulation and sharing of information, and improve the scientific nature and timeliness of decision-making.

[0087] Data control services are key links in ensuring data security and quality, including data authority management, security management and data quality control. That is, adopting log records and auditing measures to monitor data usage, discover and prevent data abuse and leakage, ensure that only authorized users or departments can access and use data, and avoid illegal access and abuse of data; clean, integrate, verify and correct data to ensure data accuracy and integrity.

[0088] The water conservancy database stores structured data and unstructured data. The water conservancy database includes: water resources database, flood database, water conservancy project database, land use and soil and water conservation database, hydrological and meteorological database and water resources economic database.

[0089] Water resources database: including hydrological data such as water quantity, water quality, water level, precipitation, evaporation, and spatial data of water bodies such as reservoirs, rivers, and lakes. These data are used in fields such as water resources evaluation, water quantity scheduling, and water resources management.

[0090] Flood database: including historical flood data, flood process data, flood warning data, etc., as well as spatial data such as flood disaster areas and flood inundation ranges. These data are used in flood warning, flood prevention and control, and disaster risk assessment.

[0091] Water conservancy project database: including the design, construction and operation data of water conservancy projects such as reservoirs, levees, sluices, channels, and spatial data of water conservancy projects. These data are used in the fields of water conservancy project management, water conservancy project design, operation and maintenance.

[0092] Land use and soil and water conservation database: including data on land use types, land cover changes, land use evaluation, and spatial data on soil and water conservation projects, vegetation coverage, etc. These data are used in the fields of land resource management, soil and water conservation planning, and ecological restoration.

[0093] Hydrometeorological database: including meteorological observation data, climate change data, hydrological model input data, and spatial data such as meteorological stations and hydrological stations. These data are used in fields such as meteorological forecasting, hydrological simulation and climate change research.

[0094] Water resources economic database: including water price, water fee, water resource utilization efficiency and other water resources economic data, as well as water resources economic zoning, water resources allocation and other spatial data. These data are used in water resources economic analysis, water resources policy formulation and water resources allocation optimization.

[0095] The architectural design of the basin basic spatial database is the key to ensuring the performance and functionality of the database. A hierarchical database architecture can be used to divide the data into the original data layer, the basic data layer, and the application data layer. The original data layer includes the original data from different sources, the basic data layer includes the cleaned, integrated and fused data, and the application data layer includes the application data of various water conservancy databases.

[0096] The metadata management module provides metadata definition, collection, entry, query and retrieval services.

[0097] Metadata definition and standardization. Metadatabase is a central database that manages metadata of various databases. When establishing a metadatabase, it is necessary to define the content and format of metadata and standardize them. The definition and standardization of metadata can ensure the consistency and understandability of various databases and improve the interoperability and discoverability of data.

[0098] Metadata collection and entry, metadata collection and entry are important links in establishing a metadata database. Metadata can be obtained through manual entry and automatic extraction, and then classified, sorted and verified to ensure the accuracy and completeness of metadata. Metadata collection and entry can be achieved through metadata management tools or database management systems.

[0099] Metadata query and retrieval: After the metadata database is established, metadata query and retrieval functions need to be provided to facilitate users to find and use relevant data. Metadata can be quickly retrieved by establishing indexes and using keyword searches. At the same time, metadata browsing and visual display can be provided to help users understand the structure and content of the database.

[0100] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM).

[0101] It should be understood that the detailed description of the technical solutions of the present invention by means of the preferred embodiments is illustrative rather than restrictive. A person skilled in the art may modify the technical solutions described in the embodiments, or replace some of the technical features by equivalents, based on reading the specification of the present invention; and 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 embodiments of the present invention.

Claims

1. A multi-source heterogeneous data fusion and data preprocessing method based on cloud computing, characterized by: The method includes: Data collection, automatically collect geospatial data, basic data, monitoring data, business management data and external shared data through APIs and Webhooks interface technology; Data processing, data quality assessment, checking data completeness, accuracy and timeliness, data cleaning and data conversion; Data fusion, integrating data from multiple sources; Database construction and data storage, establishment of water conservancy thematic database and data management strategy.

2. The multi-source heterogeneous data fusion and data preprocessing method based on cloud computing according to claim 1 is characterized by: The geospatial data include DOM digital orthophotos, DEM digital elevation models, oblique photography models, underwater topography, and BIM building information models; Basic data include main attribute data and spatial data of watersheds, rivers, lakes, water conservancy projects, and hydrological objects; Monitoring data include rainfall, water conditions, construction conditions, water quality, sediment, disaster conditions, groundwater level, water intake, soil moisture conditions, water conservancy project safety operation monitoring, and video data; Business management data include basin flood and drought disaster management business application data; Externally shared data include economic, social, land use, ecological environment, meteorological, and remote sensing related data that need to be shared from other industry sectors. The time reference of the data is Beijing time, the space reference is the 2000 National Geodetic Coordinate System (CGCS2000), and the elevation reference is the 1985 National Elevation Datum.

3. The multi-source heterogeneous data fusion and data preprocessing method based on cloud computing according to claim 1 is characterized by: The data processing includes: Data quality assessment: For spatial data, evaluate the accuracy of its spatial reference system; for monitoring data, check the frequency and accuracy of data collection to determine its timeliness and accuracy. Data completeness = (1-proportion of missing values ​​in the data) * 100%; Missing value ratio = number of missing values ​​in the data set / total number of data points in the data set; Data cleaning: For text data, perform stemming and stop word processing; for time series data, process missing values ​​and outliers; Data conversion: spatial data is transformed to match a unified spatial reference system; monitoring data is normalized to eliminate the impact of dimension; external shared data can be converted into data format through APIs.

4. The multi-source heterogeneous data fusion and data preprocessing method based on cloud computing according to claim 1 is characterized by: The data fusion described above constructs a data relationship graph and uses graph theory to perform data fusion. The specific steps are as follows: Build a graph structure: Build a graph based on the data. Each node can represent a specific management business object, such as a disaster event or a location. Each edge represents the relationship between two business objects, such as the time sequence or the distance between locations. Initialize node representation: Then initialize the representation of each node. This node representation can be some attribute values ​​related to the node. Then, a graph convolution operation is performed to update the representation of a node to the weighted average of its own representation and the representations of all its neighbors; Data fusion: After performing several graph convolutions, the final representation of each node is obtained. This representation contains both the characteristics of the node itself and the relationship information with other nodes. These nodes are then used for data fusion.

5. The multi-source heterogeneous data fusion and data preprocessing method based on cloud computing according to claim 1 is characterized by: The data management strategy includes: Data backup: a data backup is performed once a day and a full backup is performed once a week. The backup data is stored on redundant cloud storage devices to prevent primary storage failures. Data optimization: based on the access frequency and access speed, frequently accessed data is stored on fast but expensive SSD hard disks, while infrequently accessed data is stored on slower but cheaper HDD hard disks. Data security management includes encrypting stored and transmitted data, verifying the identity and permissions of users accessing data, and conducting regular security audits of the system.

6. A multi-source heterogeneous data fusion and data preprocessing system based on cloud computing, wherein the system is applicable to the method according to any one of claims 1 to 5, characterized in that: The system comprises: Data collection module, used to automatically acquire geospatial data, basic data, monitoring data, business management data and external shared data, and unify them into Beijing time and CGCS2000 national geodetic coordinate system as the standard; Data cleaning module, used to check the integrity, accuracy and timeliness of data, and remove redundant, repeated or invalid information in the data; Data fusion module, used to fuse data from various heterogeneous data sources, including data conversion and data integration; Data service module, used to provide data query, display, sharing and management services; Water conservancy database, storing structured and unstructured data; The basin basic spatial database adopts a hierarchical database architecture, which divides the data into the original data layer, basic data layer and application data layer.

7. The multi-source heterogeneous data fusion and data preprocessing system based on cloud computing according to claim 6 is characterized by: The water conservancy database includes: water resources database, flood database, water conservancy project database, land use and soil and water conservation database, hydrological and meteorological database and water resources economic database.

8. The multi-source heterogeneous data fusion and data preprocessing system based on cloud computing according to claim 6 is characterized by: The system also includes: a metadata management module, which provides metadata definition, collection, entry, query and retrieval services.

Images