Power station water regime prediction method and system based on multi-source data fusion

Abstract

The application provides a power station water regime prediction method and system based on multi-source data fusion, relates to the technical field of power station water regime prediction, first acquires a multi-source data set related to the power station water regime, then sets an influence transmission path according to a water regime influence transmission relationship with each data unit as a link node, forms a multi-source data dynamic correlation link structure, and performs progressive fusion processing based on the same, generates a progressive fusion feature set, generates a water regime influence factor link according to the link correlation relationship of the progressive fusion feature set, each influence factor has a link transmission identifier, and finally performs water regime trend deduction processing based on the time period transmission characteristics of the water regime influence factor link, combines the collection time period information to generate a power station water regime prediction result containing different deduction time periods, contains water regime state description information corresponding to each deduction time period, realizes dynamic and accurate prediction of the power station water regime, and effectively improves the accuracy and reliability of the power station water regime prediction.

Description

Technical Field

[0001] This invention relates to the field of hydrological prediction technology for power plants, and more specifically, to a method and system for hydrological prediction of power plants based on multi-source data fusion. Background Technology

[0002] In power plant operation and management, hydrological forecasting is directly related to the safe and stable operation of the power plant and the rational allocation of water resources. Traditional power plant hydrological forecasting methods mainly rely on a single type of data, such as forecasting rainfall based solely on meteorological observation data, or inferring water level changes solely based on hydrological monitoring data.

[0003] However, hydrological conditions at power stations are influenced by a combination of factors, making it difficult for a single data source to comprehensively and accurately reflect dynamic changes in hydrological conditions. Meteorological factors affect rainfall and evaporation, thus influencing reservoir inflow; hydrological conditions determine river flow, velocity, and water level changes; geological environmental factors may trigger landslides, debris flows, and other disasters, impacting the reservoir; and the power station's own operational data, such as power generation water consumption and gate opening status, also directly affect hydrological conditions. Existing forecasting methods, lacking comprehensive analysis and fusion of multi-source data, cannot fully explore the inherent connections and mutual influences between various data sources, resulting in low accuracy and reliability of forecast results, making it difficult to meet the needs of refined power station management. Summary of the Invention

[0004] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, the present invention provides a method for hydrological prediction of power plants based on multi-source data fusion, the method comprising:

[0005] Acquire a multi-source data set related to the hydrological conditions of the power station. The multi-source data set includes meteorological observation data units, hydrological monitoring data units, geological environment data units, and power station operation data units. Each data unit is marked with a collection period identifier.

[0006] Using each data unit in the multi-source data set as a link node, the influence transmission path between nodes is set according to the water situation influence transmission relationship between different data units, and the transmission direction and transmission range of each influence transmission path are marked to form a multi-source data dynamic association link structure that includes node association relationships.

[0007] Based on the multi-source data dynamic association link structure, a progressive fusion process is performed. Starting from the first link node of the multi-source data dynamic association link structure, the fusion features obtained from the previous level of fusion processing are sequentially fused with the data units corresponding to the subsequent link nodes to generate a progressive fusion feature set containing the association information of each link node.

[0008] Hydrological influence factor links are generated based on the link association relationship of the progressive fusion feature set. Each influence factor in the hydrological influence factor link is associated with the fusion feature of the corresponding link node in the progressive fusion feature set, and each influence factor has a corresponding link transmission identifier.

[0009] Based on the time-period transmission characteristics of the hydrological influencing factor links, hydrological trend extrapolation processing is performed. Combining the collection time period information corresponding to the link transmission identifier of each influencing factor, hydrological prediction results of the power station containing different extrapolation time periods are generated. The hydrological prediction results of the power station contain hydrological status description information corresponding to each extrapolation time period.

[0010] In another aspect, the present invention also provides a hydrological prediction system for power plants based on multi-source data fusion, including a processor and a machine-readable storage medium connected to the processor. The machine-readable storage medium is used to store programs, instructions or code, and the processor is used to execute the programs, instructions or code in the machine-readable storage medium to implement the above-mentioned method.

[0011] Based on the above, by comprehensively acquiring multi-source data sets related to the power station's hydrological conditions, including meteorological observations, hydrological monitoring, geological environment, and power station operation, a dynamic correlation link structure is constructed with each data unit as a link node. This clearly reveals the hydrological influence transmission relationship between different data units. Based on this dynamic correlation link structure, progressive fusion processing is performed to generate a progressive fusion feature set containing the correlation information of each link node. This fully explores the deep-seated connections and interactions between multi-source data, improving the quality and effectiveness of data fusion. Hydrological influence factor links are generated based on the progressive fusion feature set, and hydrological trend extrapolation is performed based on their time-period transmission characteristics. Combined with the collected time period information, power station hydrological prediction results containing different extrapolation time periods are generated, achieving dynamic and accurate prediction of the power station's hydrological conditions and effectively improving the accuracy and reliability of power station hydrological prediction. Attached Figure Description

[0012] Figure 1 This is a schematic diagram of the execution flow of the hydrological prediction method for power plants based on multi-source data fusion provided in an embodiment of the present invention.

[0013] Figure 2 This is a schematic diagram of exemplary hardware and software components of a power station hydrological prediction system based on multi-source data fusion provided in an embodiment of the present invention. Detailed Implementation

[0014] The present invention will now be described in detail with reference to the accompanying drawings. Figure 1This is a flowchart illustrating a hydrological prediction method for power plants based on multi-source data fusion, provided in one embodiment of the present invention. The following is a detailed description of this hydrological prediction method for power plants based on multi-source data fusion.

[0015] Step S110: Obtain a multi-source data set related to the hydrological conditions of the power station. The multi-source data set includes meteorological observation data units, hydrological monitoring data units, geological environment data units, and power station operation data units. Each data unit is marked with a collection period identifier.

[0016] This embodiment uses a hydropower station in a river basin as an application scenario. The river basin where the hydropower station is located belongs to the subtropical monsoon climate zone, with a large basin area and complex meteorological, hydrological, and geological conditions. To achieve accurate prediction of the hydropower station's water conditions, it is necessary to first acquire comprehensive multi-source data. The meteorological observation data unit collects data through multiple automatic weather stations deployed within and around the river basin. These weather stations are distributed in the upper, middle, and lower reaches of the river basin, as well as near major mountain ranges. The data collection frequency is once per hour, and the data content covers various meteorological elements related to water conditions. Each data point corresponds to a unique collection time period identifier, which is accurate to the minute and formatted as "year-month-day-hour-minute".

[0017] The hydrological monitoring data unit is provided by hydrological monitoring stations located at different locations along the main rivers, tributaries, and reservoir dams within the basin. Data is collected every ten minutes and includes monitoring data on water level, flow, and water quality. Each data record is also marked with a corresponding collection time period. The geological environment data unit is derived from geological survey data of the area where the hydropower station is located, including data on soil, topography, and rock strata. The collection time period is marked with the specific survey time.

[0018] The power station operation data unit is provided in real time by the hydropower station's operation monitoring system. The data is collected at a high frequency, once per second, and includes various parameters related to unit operation, gate control, and water conveyance. Each data point also carries a precise identification of the collection period. Through this method, data from different sources and of different types are integrated to form a multi-source data set for hydrological forecasting.

[0019] Step S120: Using each data unit in the multi-source data set as a link node, set the influence transmission path between nodes according to the water situation influence transmission relationship between different data units, mark the transmission direction and transmission range of each influence transmission path, and form a multi-source data dynamic association link structure containing node association relationships.

[0020] After obtaining the multi-source dataset, it is necessary to clarify the relationships between the data units. Meteorological observation data units, hydrological monitoring data units, geological environment data units, and power plant operation data units are each treated as independent link nodes. Next, the hydrological influence transmission relationships between these nodes are analyzed. For example, precipitation data in the meteorological observation data unit directly affects water level data in the hydrological monitoring data unit. This influence is unidirectional; changes in precipitation data lead to changes in water level data, but changes in water level data typically do not conversely affect precipitation data. Based on this unidirectional influence relationship, an influence transmission path can be established from the meteorological observation data unit node to the hydrological monitoring data unit node. Simultaneously, the transmission direction of this path should be marked as from the meteorological observation data unit node to the hydrological monitoring data unit node, and the transmission range is determined based on the collection period identifiers of the two data units, covering all time periods where both data units have data records. Following a similar method, the influence transmission relationships between other data units are analyzed, corresponding influence transmission paths are established, and the transmission direction and range of each path are marked, ultimately forming a multi-source data dynamic correlation link structure that clearly demonstrates the relationships between the nodes.

[0021] Step S121: Extract the water situation impact transmission attribute of each data unit in the multi-source data set. The water situation impact transmission attribute is the attribute information that the data unit can have a water situation impact on other data units. The water situation impact transmission attribute of different types of data units contains different impact dimension information.

[0022] To accurately define the impact transmission paths between nodes, it is necessary to first extract the hydrological impact transmission attributes for each data unit. For meteorological observation data units, their hydrological impact transmission attributes refer to those attributes within that data unit that can influence the hydrological conditions of other data units. Different types of data units contain different impact dimensions in their hydrological impact transmission attributes. For example, the impact dimensions of meteorological observation data units may be related to precipitation, wind force, and temperature, while the impact dimensions of hydrological monitoring data units may be related to water level, water flow, and water quality.

[0023] Step S1211: Extract precipitation characteristics, wind force characteristics, and temperature characteristics from the meteorological observation data unit. The precipitation characteristics are attribute information related to precipitation in the meteorological observation data unit, the wind force characteristics are attribute information related to wind force in the meteorological observation data unit, and the temperature characteristics are attribute information related to temperature in the meteorological observation data unit.

[0024] For meteorological observation data units, specific precipitation, wind, and temperature characteristics are extracted. Precipitation characteristics are attribute information related to precipitation, including precipitation type (e.g., light rain, moderate rain, heavy rain, torrential rain), duration, and intensity variations. Wind characteristics involve wind direction and speed; wind direction can be represented by azimuth, while wind speed includes indicators such as average wind speed and instantaneous maximum wind speed. Temperature characteristics include air temperature, surface temperature, and temperature variations such as diurnal temperature range and temperature trends.

[0025] Step S1212: The precipitation characteristics, wind characteristics, and temperature characteristics are used as the water situation influence transmission attributes of the meteorological observation data unit, wherein the precipitation characteristics correspond to the precipitation influence dimension, the wind characteristics correspond to the wind influence dimension, and the temperature characteristics correspond to the temperature influence dimension.

[0026] The extracted precipitation, wind, and temperature characteristics were identified as the water situation impact transmission attributes of the meteorological observation data units. Precipitation characteristics correspond to the precipitation impact dimension, which primarily reflects the direct influence of precipitation on water conditions; for example, the amount of precipitation directly affects river water levels. Wind characteristics correspond to the wind impact dimension; the magnitude and direction of wind affect the evaporation rate from the water surface, thus influencing the total water volume, and may also affect the speed and direction of water flow. Temperature characteristics correspond to the temperature impact dimension; temperature affects the evaporation rate of water and also influences chemical reactions and biological activities in the water, thereby indirectly affecting water conditions.

[0027] Step S1213: Extract water level features, flow features and water quality features from the hydrological monitoring data unit. The water level features are attribute information related to water level in the hydrological monitoring data unit. The flow features are attribute information related to flow in the hydrological monitoring data unit. The water quality features are attribute information related to water quality in the hydrological monitoring data unit.

[0028] For hydrological monitoring data units, water level characteristics, flow characteristics, and water quality characteristics are extracted. Water level characteristics are attribute information related to water level, including water level height at different monitoring points, the range of water level changes, and the trend of water level rise and fall. Flow characteristics involve water velocity, flow rate, and flow pulsation, which reflect the movement state of the water body. Water quality characteristics include dissolved oxygen content, pH value, turbidity, and the concentration of various ions in the water; these parameters characterize the quality status of the water body.

[0029] Step S1214: Use the water level feature, the water flow feature, and the water quality feature as the water situation impact transmission attributes of the hydrological monitoring data unit, wherein the water level feature corresponds to the water level impact dimension, the water flow feature corresponds to the water flow impact dimension, and the water quality feature corresponds to the water quality impact dimension.

[0030] Water level characteristics, flow characteristics, and water quality characteristics are defined as the hydrological impact transmission attributes of the hydrological monitoring data unit. Water level characteristics correspond to the water level impact dimension; the water level directly affects the water storage and power generation capacity of a hydropower station, and also impacts the surrounding geological environment. Flow characteristics correspond to the flow impact dimension; the velocity and flow rate of the water flow affect riverbed scouring and sedimentation, and also influence the operation of the hydropower station's generating units. Water quality characteristics correspond to the water quality impact dimension; the quality of water not only affects the ecological environment but may also cause corrosion and other impacts on the hydropower station's equipment.

[0031] Step S1215: Extract soil features, topographic features, and rock strata features from the geological environment data unit. The soil features are attribute information related to soil in the geological environment data unit, the topographic features are attribute information related to topography in the geological environment data unit, and the rock strata features are attribute information related to rock strata in the geological environment data unit.

[0032] The extraction of water impact transmission attributes for geological environmental data units includes soil characteristics, topographic features, and rock strata characteristics. Soil characteristics are soil-related attributes, such as soil type (clay, loam, sand, etc.), soil structure, soil porosity, and soil moisture content. Topographic features involve the terrain type of the area (mountains, plains, hills, etc.), altitude, slope, and aspect, which affect water collection and flow paths. Rock strata characteristics include rock type (sedimentary rocks, igneous rocks, metamorphic rocks, etc.), stratum thickness, strike and dip angle, and the degree of fissure development in the strata. This information is closely related to groundwater movement and storage, as well as reservoir leakage.

[0033] Step S1216: The soil features, the terrain features, and the rock strata features are used as the water situation influence transfer attributes of the geological environment data unit, wherein the soil features correspond to the soil influence dimension, the terrain features correspond to the terrain influence dimension, and the rock strata features correspond to the rock strata influence dimension.

[0034] Soil characteristics, topographic features, and rock strata characteristics are used as water situation impact transmission attributes for geological environmental data units. Soil characteristics correspond to the soil impact dimension; soil properties determine its water permeability and retention capacity, thus influencing the formation of surface and groundwater runoff. Topographic features correspond to the topographic impact dimension; different topographic conditions lead to differences in water flow velocity and direction, affecting water situation changes. Rock strata characteristics correspond to the rock strata impact dimension; the permeability and stability of rock strata affect groundwater recharge and the safe operation of reservoirs, thereby impacting water situation.

[0035] Step S1217: Extract the unit operation characteristics, gate control characteristics, and water conveyance characteristics from the power plant operation data unit. The unit operation characteristics are attribute information related to unit operation in the power plant operation data unit. The gate control characteristics are attribute information related to gate control in the power plant operation data unit. The water conveyance characteristics are attribute information related to water conveyance in the power plant operation data unit.

[0036] The extraction of hydrological impact transmission attributes for the power station operation data unit covers unit operation characteristics, gate control characteristics, and water conveyance characteristics. Unit operation characteristics are attribute information related to unit operation, including operating parameters such as unit output, speed, vibration, and temperature. These parameters reflect the unit's operating status, which affects the hydropower station's water consumption and drainage. Gate control characteristics involve gate opening degree, opening / closing status, and actuation time; gate operation directly controls the discharge flow and reservoir water level. Water conveyance characteristics include the flow rate, pressure, and losses during water conveyance, information relevant to the hydropower station's water resource allocation and utilization.

[0037] Step S1218: The unit operation characteristics, the gate control characteristics, and the water conveyance characteristics are used as the water situation influence transmission attributes of the power station operation data unit, wherein the unit operation characteristics correspond to the unit influence dimension, the gate control characteristics correspond to the gate influence dimension, and the water conveyance characteristics correspond to the water conveyance influence dimension.

[0038] The characteristics of generating units, gate control, and water conveyance are defined as the hydrological impact transmission attributes of the power station's operational data unit. Generating unit operating characteristics correspond to the unit impact dimension; the operating status of the generating units directly affects the hydropower station's water consumption and discharge, thus altering the hydrological situation. Gate control characteristics correspond to the gate impact dimension; gate control is a crucial means of regulating hydrological conditions. By controlling the gate opening, the downstream flow rate can be altered, thereby affecting downstream water levels and flow conditions. Water conveyance characteristics correspond to the water conveyance impact dimension; parameters such as flow rate and pressure during water conveyance affect water resource allocation, indirectly influencing the hydrological situation.

[0039] Step S122: Analyze the matching relationship of hydrological influence transmission attributes between the meteorological observation data unit and the hydrological monitoring data unit, determine the effectiveness of direct influence transmission between the two, and when the effectiveness of direct influence transmission meets the preset transmission conditions, set the first influence transmission path from the meteorological observation data unit node to the hydrological monitoring data unit node.

[0040] After extracting the hydrological impact transmission attributes of each data unit, it is necessary to analyze the attribute matching relationship between the meteorological observation data unit and the hydrological monitoring data unit. Specifically, this involves examining whether there are inherent connections and interactions between the precipitation, wind, and temperature impact dimensions of the meteorological observation data unit and the water level, flow, and water quality impact dimensions of the hydrological monitoring data unit. For example, there might be a direct causal relationship between the precipitation characteristics (precipitation impact dimension) of the meteorological observation data unit and the water level characteristics (water level impact dimension) of the hydrological monitoring data unit; increased precipitation leads to a rise in water level. By analyzing the above attribute matching relationships, the effectiveness of the direct impact transmission between the two can be determined. If the above effectiveness meets the preset transmission conditions, such as the existence of significant impact relationships between multiple attribute dimensions and a large degree of influence, then a first impact transmission path from the meteorological observation data unit node to the hydrological monitoring data unit node can be set.

[0041] Step S123: Mark the transmission direction of the first influence transmission path as from the meteorological observation data unit node to the hydrological monitoring data unit node, and mark the transmission range of the first influence transmission path as the time range corresponding to all collection time period identifiers of the meteorological observation data unit and the hydrological monitoring data unit.

[0042] For a pre-defined first impact propagation path, its direction and range must be clearly marked. The propagation direction is from the meteorological observation data unit node to the hydrological monitoring data unit node, indicating that changes in the meteorological observation data unit will affect the hydrological monitoring data unit. The propagation range covers the entire time period corresponding to the data collection period identifiers of both the meteorological observation data unit and the hydrological monitoring data unit. In other words, the impact propagation path is effective for all time periods during which data is recorded in both data units, and changes in meteorological data during these time periods may affect the hydrological data through this path.

[0043] Step S124: Analyze the matching relationship of hydrological influence transmission attributes between the hydrological monitoring data unit and the geological environment data unit, determine the effectiveness of direct influence transmission between the two, and when the effectiveness of direct influence transmission meets the preset transmission conditions, set a second influence transmission path from the hydrological monitoring data unit node to the geological environment data unit node.

[0044] Next, the matching relationship of water situation impact transmission attributes between hydrological monitoring data units and geological environment data units is analyzed. Water level characteristics (water level impact dimension) of hydrological monitoring data units may affect soil characteristics (soil impact dimension) of geological environment data units; for example, prolonged high water levels may lead to soil soaking and softening. Water flow characteristics (water flow impact dimension) may cause erosion or siltation of topographic features (topographic impact dimension). Water quality characteristics (water quality impact dimension) may chemically react with rock strata characteristics (rock strata impact dimension), affecting the properties of the rock strata. By analyzing these matching relationships, the effectiveness of direct impact transmission between the two is determined. When this effectiveness meets the preset transmission conditions, a second impact transmission path is set from the hydrological monitoring data unit node to the geological environment data unit node.

[0045] Step S125: Mark the transmission direction of the second influence transmission path as from the hydrological monitoring data unit node to the geological environment data unit node, and mark the transmission range of the second influence transmission path as covering the time range corresponding to all collection time period identifiers of the hydrological monitoring data unit and the geological environment data unit.

[0046] The second influence propagation path is marked with a propagation direction from the hydrological monitoring data unit node to the geological environment data unit node, indicating that changes in the hydrological monitoring data unit will affect the geological environment data unit. The propagation range covers the time period corresponding to all data collection time periods of both the hydrological monitoring data unit and the geological environment data unit. That is, within all time periods when data is recorded in these two data units, this second influence propagation path may be in effect, and changes in hydrological data will affect the geological environment data through this path.

[0047] Step S126: Analyze the matching relationship of hydrological influence transmission attributes between the geological environment data unit and the power station operation data unit, determine the effectiveness of direct influence transmission between the two, and when the effectiveness of direct influence transmission meets the preset transmission conditions, set a third influence transmission path from the geological environment data unit node to the power station operation data unit node.

[0048] Then, the matching relationship of hydrological influence transmission attributes between the geological environment data unit and the power plant operation data unit is analyzed. Soil characteristics (soil influence dimension) of the geological environment data unit may affect the foundation stability of the power plant, thus indirectly affecting the unit operation characteristics (unit influence dimension). Topographic features (topographic influence dimension) may affect the site selection of the power plant and the layout of the water conveyance line, thereby affecting the water conveyance characteristics (water conveyance influence dimension). The stability of rock strata characteristics (rock strata influence dimension) may affect the foundation structure of the gate, thus being related to the gate control characteristics (gate influence dimension). Through the analysis of the above attribute matching relationships, the effectiveness of the direct influence transmission between the two is determined. When the effectiveness meets the preset transmission conditions, a third influence transmission path is set from the geological environment data unit node to the power plant operation data unit node.

[0049] Step S127: Mark the transmission direction of the third influence transmission path as from the geological environment data unit node to the power plant operation data unit node, and mark the transmission range of the third influence transmission path as the time range corresponding to all collection time period identifiers of the geological environment data unit and the power plant operation data unit.

[0050] The third impact transmission path follows a direction from the geological environment data unit node to the power plant operation data unit node, indicating that the condition of the geological environment data unit will affect the power plant operation data unit. The transmission range covers the entire time period corresponding to the data collection time period identifiers of both the geological environment data unit and the power plant operation data unit. That is, within all time periods when data is recorded in these two data units, changes in the geological environment data may affect the power plant operation data through this path.

[0051] Step S128: Analyze the indirect hydrological impact transmission attribute matching relationship from the meteorological observation data unit to the geological environment data unit through the hydrological monitoring data unit. Combine the transmission effectiveness of the first impact transmission path and the second impact transmission path to determine the indirect impact transmission effectiveness. When the indirect impact transmission effectiveness meets the preset transmission conditions, set the fourth impact transmission path from the meteorological observation data unit node to the geological environment data unit node through the hydrological monitoring data unit node.

[0052] Besides direct influence transmission paths, indirect influence transmission paths also exist. This study analyzes the attribute matching relationships where meteorological observation data units indirectly influence geological environmental data units through hydrological monitoring data units. For example, the precipitation characteristics (precipitation influence dimension) of the meteorological observation data unit first influence the water level characteristics (water level influence dimension) of the hydrological monitoring data unit through a first influence transmission path. Then, the water level characteristics of the hydrological monitoring data unit further influence the soil characteristics (soil influence dimension) of the geological environmental data unit through a second influence transmission path, thus forming an indirect influence transmission path. The effectiveness of the above indirect influence transmission is comprehensively determined by combining the transmission effectiveness of the first and second influence transmission paths. If the effectiveness of the indirect influence transmission meets the preset transmission conditions, a fourth influence transmission path is established, linking the meteorological observation data unit node to the geological environmental data unit node through the hydrological monitoring data unit node.

[0053] Step S129: Mark the transmission direction of the fourth impact transmission path as from the meteorological observation data unit node through the hydrological monitoring data unit node to the geological environment data unit node, and mark the transmission range of the fourth impact transmission path as the intersection time range corresponding to all collection time period identifiers of the meteorological observation data unit, the hydrological monitoring data unit, and the geological environment data unit.

[0054] The fourth influence transmission path is marked with its transmission direction from the meteorological observation data unit node through the hydrological monitoring data unit node to the geological environment data unit node, clearly indicating the intermediate process of influence transmission. The transmission range covers the intersection time range corresponding to all data collection time period identifiers of the meteorological observation data unit, hydrological monitoring data unit, and geological environment data unit. That is, this indirect influence transmission path is only effective within the time period when data is recorded in all three data units, because only within the above-mentioned time period can meteorological data sequentially affect geological environment data through hydrological data.

[0055] Step S1210: Integrate the first influence transmission path, the second influence transmission path, the third influence transmission path, and the fourth influence transmission path to form a multi-source data dynamic association link structure that includes node association relationships. Each influence transmission path in the multi-source data dynamic association link structure has a unique path identifier and corresponding transmission direction and transmission range information.

[0056] The previously established first, second, third, and fourth influence propagation paths are integrated. Each path is assigned a unique path identifier for easy differentiation and management. Simultaneously, each path retains its corresponding propagation direction and range information. Through this integration, a complete multi-source data dynamic association link structure is formed, which can demonstrate the association relationships formed between various data unit nodes through different influence propagation paths.

[0057] Step S130: Based on the multi-source data dynamic association link structure, perform progressive fusion processing. Starting from the first link node of the multi-source data dynamic association link structure, sequentially perform feature interaction fusion with the data unit corresponding to the previous link node to generate a progressive fusion feature set containing the association information of each link node.

[0058] After forming a dynamic association link structure for multi-source data, a progressive fusion process is performed based on this structure. First, the first link node of the link structure is determined; in this embodiment, the first link node is the meteorological observation data unit node. Starting from this node, its corresponding original data features are used as initial fusion features. Then, according to the order of influence transmission paths in the link structure, the fusion features obtained from the previous level of fusion processing are sequentially fused with the data units corresponding to subsequent link nodes. For example, the initial fusion feature (features of the meteorological observation data unit) is fused with the original features of the hydrological monitoring data unit through the first influence transmission path to obtain a first-level fusion feature; then, the first-level fusion feature is fused with the original features of the geological environment data unit through the second influence transmission path to obtain a second-level fusion feature; and so on. Simultaneously, cross-level influence transmission paths are also considered, such as the fourth influence transmission path, where the initial fusion feature is fused with the original features of the geological environment data unit after passing through the hydrological monitoring data unit node to obtain a cross-level fusion feature. Through the above progressive fusion process, a progressive fusion feature set containing the association information of each link node is generated.

[0059] Step S131: Determine the first link node of the multi-source data dynamic association link structure as the meteorological observation data unit node, extract the original features of the meteorological observation data unit corresponding to the meteorological observation data unit node, and use the original features of the meteorological observation data unit as the initial fusion features.

[0060] When performing progressive fusion processing, the first step is to determine the leading link node in the dynamic association structure of multi-source data. Based on the preceding analysis and the link structure setup, the leading link node is determined to be the meteorological observation data unit node. Then, the original features of the meteorological observation data unit corresponding to this node are extracted. These original features are the previously extracted precipitation, wind, and temperature features, including their specific attribute information. These original features are combined together as the initial fusion features for progressive fusion processing; subsequent fusion processes will be based on these initial fusion features.

[0061] Step S132: Invoke the first influence transmission path in the multi-source data dynamic association link structure, input the initial fusion feature into the first influence transmission path, and perform the first feature interaction fusion with the original feature of the hydrological monitoring data unit corresponding to the hydrological monitoring data unit node pointed to by the first influence transmission path.

[0062] The first influence propagation path in the multi-source data dynamic association link structure is invoked. This first influence propagation path points from the meteorological observation data unit node to the hydrological monitoring data unit node. Initial fusion features (the original features of the meteorological observation data unit) are input into this path, enabling initial feature interaction fusion with the original features of the hydrological monitoring data unit corresponding to the node pointed to by the first influence propagation path. The original features of the hydrological monitoring data unit include water level features, flow features, and water quality features. These features will interact with the original features of the meteorological observation data unit to achieve preliminary data fusion.

[0063] Step S1321: During the initial feature interaction fusion process, the precipitation feature in the initial fused features is interactively mapped with the water level feature in the original features of the hydrological monitoring data unit to generate a first interactive feature; the wind force feature in the initial fused features is interactively mapped with the water flow feature in the original features of the hydrological monitoring data unit to generate a second interactive feature; and the temperature feature in the initial fused features is interactively mapped with the water quality feature in the original features of the hydrological monitoring data unit to generate a third interactive feature.

[0064] In the initial feature interaction fusion process, it is necessary to perform interactive mapping between each feature in the initial fused features and the corresponding features in the original features of the hydrological monitoring data unit. Specifically, the precipitation features in the initial fused features are interactively mapped to the water level features in the original features of the hydrological monitoring data unit. This includes analyzing how changes in precipitation features cause changes in water level features, such as how an increase in precipitation intensity leads to a faster rise in water level, and how the duration of precipitation affects the cumulative rise in water level, etc., generating the first interactive feature through the above interactive mapping. The wind features in the initial fused features are interactively mapped to the water flow features in the original features of the hydrological monitoring data unit, considering the influence of wind direction and speed on water flow velocity and direction, such as how headwinds may slow down water flow velocity, and how strong winds may cause fluctuations in water flow, thus generating the second interactive feature. The temperature features in the initial fused features are interactively mapped to the water quality features in the original features of the hydrological monitoring data unit, as changes in temperature may affect the dissolved oxygen content and chemical reaction rate in the water, thereby affecting the water quality features, generating the third interactive feature through the above mapping.

[0065] Step S13211: Extract the collection period identifier corresponding to the precipitation feature in the initial fusion features, extract the collection period identifier corresponding to the water level feature in the original features of the hydrological monitoring data unit, and select the time period with the same collection period identifier as the first interaction time period.

[0066] When interactively mapping precipitation features to water level features, it is essential to first ensure their temporal correspondence. The data collection period identifiers corresponding to the precipitation features in the initial fused features and the water level features in the original features of the hydrological monitoring data units are extracted. These collection period identifiers are then compared, and identical periods are selected as the first interactive period. Within this first interactive period, precipitation and water level data coexist, allowing for effective interactive analysis.

[0067] Step S13212: During the first interaction period, extract the trend information of the precipitation feature in each period, extract the trend information of the water level feature in each period, and establish the correspondence between the trend information of the precipitation feature and the trend information of the water level feature.

[0068] Within the first interactive time period, the changing trends of precipitation and water level characteristics are extracted for each time period. For precipitation characteristics, the analysis examines whether they exhibit an upward, downward, or stable trend within each time period, as well as the rate of trend change. Similarly, for water level characteristics, the analysis examines their changing trends within each time period, such as whether the water level is rising, falling, or stable, and the magnitude of any rise or fall. Then, a correspondence is established between the changing trends of precipitation and water level characteristics. For example, for periods when precipitation characteristics show a rapid upward trend, does the corresponding water level characteristic also show a rapid upward trend, or is there a certain lag?

[0069] Step S13213: Based on the correspondence, map the changing trend information of the precipitation characteristics to the dimensional space of the changing trend information of the water level characteristics, and generate the mapped first trend association information.

[0070] Based on the established correspondence between precipitation characteristic change trends and water level characteristic change trends, the precipitation characteristic change trends are mapped into the dimensional space of the water level characteristic change trends. This means that the change trends of precipitation characteristics are represented using the same dimension as the change trends of water level characteristics, facilitating comparison and correlation analysis. Through the above mapping, first trend correlation information is generated, which reflects the inherent connection between precipitation characteristics and water level characteristics in terms of change trends.

[0071] Step S13214: Integrate the changing trend information of the precipitation characteristics, the changing trend information of the water level characteristics, and the mapped first trend association information to generate the first interactive feature.

[0072] The information on the changing trends of precipitation characteristics, water level characteristics, and the mapped first trend correlation information is integrated. During this integration process, these three parts of information need to be organically combined to form a comprehensive feature that fully reflects the interaction between precipitation and water level characteristics—the first interaction feature. This first interaction feature includes the temporal correspondence between the two, the correlation of their changing trends, and the mapped trend correlation information.

[0073] Step S13215: Extract the collection period identifier corresponding to the wind force feature in the initial fusion feature, extract the collection period identifier corresponding to the water flow feature in the original feature of the hydrological monitoring data unit, and select the time period with the same collection period identifier as the second interaction time period.

[0074] For the interactive mapping between wind and water flow features, it is also necessary to first determine the corresponding time periods. The data collection time period identifiers corresponding to the wind features in the initial fused features and the water flow features in the original features of the hydrological monitoring data units are extracted, and the time periods where both are the same are selected as the second interactive time period. Within the second interactive time period, the data for both wind and water flow features coexist, allowing for interactive analysis.

[0075] Step S13216: During the second interaction period, extract the variation range information of the wind force feature in each period, extract the variation range information of the water flow feature in each period, and establish the correspondence between the variation range information of the wind force feature and the variation range information of the water flow feature.

[0076] During the second interactive time period, the variation amplitude information of wind force characteristics and water flow characteristics in each time period is extracted. The variation amplitude of wind force characteristics can be represented by the change in wind speed, such as the difference in wind speed between adjacent time periods. The variation amplitude of water flow characteristics can be measured by the change in water flow velocity or flow rate. Then, a correspondence between the variation amplitude information of wind force characteristics and the variation amplitude information of water flow characteristics is established, and the influence of the magnitude of wind force variation amplitude on the magnitude of water flow variation amplitude is analyzed. For example, does the variation amplitude of water flow amplitude also increase during periods of greater wind force variation amplitude?

[0077] Step S13217: Based on the correspondence, map the variation amplitude information of the wind force feature to the dimensional space of the variation amplitude information of the water flow feature to generate the first amplitude association information after mapping.

[0078] Based on the correspondence between wind force variation amplitude information and water flow variation amplitude information, the wind force variation amplitude information is mapped to the dimensional space of water flow variation amplitude information. Through this mapping, the wind force variation amplitude is represented using the same dimension as the water flow variation amplitude, which facilitates the analysis of the amplitude correlation between the two and generates the first amplitude correlation information.

[0079] Step S13218: Integrate the variation amplitude information of the wind force feature, the variation amplitude information of the water flow feature, and the mapped first amplitude correlation information to generate a second interactive feature.

[0080] The variation amplitude information of wind force characteristics, the variation amplitude information of water flow characteristics, and the mapped first amplitude correlation information are integrated to form a second interactive feature. This second interactive feature comprehensively reflects the interaction relationship between wind force characteristics and water flow characteristics in terms of variation amplitude, including the magnitude of the variation amplitude of both and the correlation information between them.

[0081] Step S13219: Extract the collection period identifier corresponding to the temperature feature in the initial fusion features, extract the collection period identifier corresponding to the water quality feature in the original features of the hydrological monitoring data unit, and select the time period with the same collection period identifier as the third interaction time period.

[0082] For the interactive mapping between temperature and water quality features, the data collection period identifiers corresponding to the temperature features in the initial fused features and the data collection period identifiers corresponding to the water quality features in the original features of the hydrological monitoring data units are extracted. The same time periods are selected as the third interactive time period. Within the third interactive time period, the data for both temperature and water quality features coexist, allowing for interactive analysis.

[0083] Step S132110: During the third interaction period, extract the stable state information of the temperature feature in each period, extract the stable state information of the water quality feature in each period, and establish the correspondence between the stable state information of the temperature feature and the stable state information of the water quality feature.

[0084] Within the third interaction period, the stability information of temperature and water quality characteristics at each time point is extracted. The stability information of temperature characteristics can be represented by the temperature fluctuation range within the time period; a smaller fluctuation range indicates a more stable temperature. The stability information of water quality characteristics can be measured by the fluctuation of various water quality indicators (such as pH value and dissolved oxygen content) within the time period. A correspondence between the stability information of temperature characteristics and the stability information of water quality characteristics is established to analyze how the stability of temperature affects the stability of water quality; for example, whether water quality indicators are also relatively stable during periods of stable temperature.

[0085] Step S132111: Based on the correspondence, map the stable state information of the temperature feature to the dimensional space of the stable state information of the water quality feature to generate the first stable association information after mapping.

[0086] Based on the correspondence between the stable state information of temperature features and the stable state information of water quality features, the stable state information of temperature features is mapped to the dimensional space of the stable state information of water quality features to generate the first stable correlation information, which reflects the correlation between temperature and water quality in the stable state.

[0087] Step S132112: Integrate the stable state information of the temperature feature, the stable state information of the water quality feature, and the mapped first stable correlation information to generate a third interactive feature.

[0088] The steady-state information of temperature features, the steady-state information of water quality features, and the mapped first steady-state correlation information are integrated to form a third interactive feature, which comprehensively reflects the interactive relationship between temperature features and water quality features in terms of steady state.

[0089] Step S1322: Integrate the first interaction feature, the second interaction feature and the third interaction feature to generate a first-level fusion feature. The first-level fusion feature carries the path identifier of the first influence transmission path and the node identifiers of the corresponding meteorological observation data unit node and hydrological monitoring data unit node.

[0090] After generating the first, second, and third interactive features, these three interactive features are integrated. During integration, the information they contain needs to be merged to form a unified first-level fused feature. Simultaneously, for ease of traceability and management, the first-level fused feature should include the path identifier of the first impact transmission path, as well as the node identifiers of the corresponding meteorological observation data unit nodes and hydrological monitoring data unit nodes. This clearly identifies the source of the fused feature and the linked nodes.

[0091] Step S133: Invoke the second influence transmission path in the multi-source data dynamic association link structure, input the first-level fusion feature into the second influence transmission path, and perform secondary feature interaction fusion with the original feature of the geological environment data unit corresponding to the geological environment data unit node pointed to by the second influence transmission path.

[0092] After generating the primary fusion features, the second influence transmission path in the multi-source data dynamic association link structure is invoked. This second influence transmission path points from the hydrological monitoring data unit node to the geological environment data unit node. The primary fusion features are input into this path, enabling secondary feature interaction fusion with the original features of the geological environment data unit corresponding to the geological environment data unit node pointed to by the second influence transmission path. The original features of the geological environment data unit include soil features, topographic features, and rock strata features. The primary fusion features will interact with these features to obtain deeper association information.

[0093] Step S1331: In the secondary feature interaction fusion process, the first interactive feature in the primary fusion feature is interactively mapped with the soil feature in the original feature of the geological environment data unit to generate the fourth interactive feature; the second interactive feature in the primary fusion feature is interactively mapped with the topographic feature in the original feature of the geological environment data unit to generate the fifth interactive feature; the third interactive feature in the primary fusion feature is interactively mapped with the rock strata feature in the original feature of the geological environment data unit to generate the sixth interactive feature.

[0094] In the secondary feature interaction fusion process, the interaction features in the primary fusion features are interactively mapped to the corresponding features in the original features of the geological environment data unit. The first interaction feature in the primary fusion features (the interaction result of precipitation and water level features) is interactively mapped to the soil features in the original features of the geological environment data unit. The analysis examines how water level changes, reflected by the first interaction feature, affect soil characteristics such as water content and porosity; for example, a prolonged high water level may lead to soil saturation, thus generating the fourth interaction feature. The second interaction feature in the primary fusion features (the interaction result of wind and water flow features) is interactively mapped to the topographic features in the original features of the geological environment data unit, considering the erosion and deposition effects of changes in water flow features; for example, an increase in water flow velocity may exacerbate erosion of riverbank topography, generating the fifth interaction feature. The third interaction feature in the primary fusion features (the interaction result of temperature and water quality features) is interactively mapped to the rock strata features in the original features of the geological environment data unit, analyzing how changes in water quality features (such as changes in pH) chemically react with the rock strata, affecting their structure and properties, thus generating the sixth interaction feature.

[0095] Step S1332: Integrate the fourth interaction feature, the fifth interaction feature and the sixth interaction feature to generate a secondary fusion feature. The secondary fusion feature carries the path identifier of the second influence transmission path and the node identifiers of the corresponding hydrological monitoring data unit node and geological environment data unit node.

[0096] The generated fourth, fifth, and sixth interaction features are integrated to form a secondary fusion feature. This secondary fusion feature also carries the path identifier of the second influence transmission path, as well as the node identifiers of the corresponding hydrological monitoring data unit nodes and geological environment data unit nodes, to clearly identify its source and the involved link nodes.

[0097] Step S134: Invoke the third influence transmission path in the multi-source data dynamic association link structure, input the secondary fusion feature into the third influence transmission path, and perform three feature interaction fusions with the original features of the power plant operation data unit corresponding to the power plant operation data unit node pointed to by the third influence transmission path.

[0098] Next, the third influence transmission path in the multi-source data dynamic association link structure is invoked. This third influence transmission path points from the geological environment data unit node to the power plant operation data unit node. Secondary fusion features are input into this path, and a three-stage feature interaction fusion is performed with the original features of the power plant operation data unit corresponding to the power plant operation data unit node pointed to by the third influence transmission path. The original features of the power plant operation data unit include unit operation features, gate control features, and water conveyance features. The secondary fusion features will interact and fuse with these features to further explore the correlations between the data.

[0099] Step S1341: During the three-stage feature interaction fusion process, the fourth interaction feature in the secondary fusion feature is interactively mapped with the unit operation feature in the original feature of the power plant operation data unit to generate the seventh interaction feature; the fifth interaction feature in the secondary fusion feature is interactively mapped with the gate control feature in the original feature of the power plant operation data unit to generate the eighth interaction feature; and the sixth interaction feature in the secondary fusion feature is interactively mapped with the water conveyance feature in the original feature of the power plant operation data unit to generate the ninth interaction feature.

[0100] In the three-stage feature interaction fusion process, each interaction feature in the secondary fusion feature is interactively mapped to the corresponding feature in the original feature of the power plant operation data unit. The fourth interaction feature in the secondary fusion feature (the interaction result of the first interaction feature and soil features) is interactively mapped to the unit operation feature in the original feature of the power plant operation data unit. This analyzes how changes in soil features affect unit operation; for example, soil instability may affect the stability of the unit foundation, thus affecting the unit's vibration, output, and other operating characteristics, generating the seventh interaction feature. The fifth interaction feature in the secondary fusion feature (the interaction result of the second interaction feature and terrain features) is interactively mapped to the gate control feature in the original feature of the power plant operation data unit, considering the impact of terrain changes on gate operation; for example, riverbed siltation may change the gate's stress conditions, affecting gate opening control, generating the eighth interaction feature. The sixth interaction feature in the secondary fusion feature (the interaction result of the third interaction feature and rock strata features) is interactively mapped to the water conveyance feature in the original feature of the power plant operation data unit, analyzing how changes in rock strata features (such as changes in permeability) affect the flow and pressure of the water conveyance pipeline, generating the ninth interaction feature.

[0101] Step S1342: Integrate the seventh interaction feature, the eighth interaction feature, and the ninth interaction feature to generate a three-level fusion feature. The three-level fusion feature carries the path identifier of the third influence transmission path and the node identifiers of the corresponding geological environment data unit node and power plant operation data unit node.

[0102] The seventh, eighth, and ninth interaction features are integrated to generate a three-level fused feature. This three-level fused feature includes the path identifier of the third influence transmission path, as well as the node identifiers of the corresponding geological environment data unit node and power plant operation data unit node.

[0103] Step S135: Call the fourth influence transmission path in the multi-source data dynamic association link structure, input the initial fusion feature into the fourth influence transmission path, and perform cross-level feature interaction fusion with the original features of the geological environment data unit corresponding to the hydrological monitoring data unit node after passing through the hydrological monitoring data unit node.

[0104] In addition to the step-by-step fusion described above, it is also necessary to consider the cross-level impact transmission path. The fourth impact transmission path in the multi-source data dynamic association link structure is invoked. This fourth impact transmission path leads from the meteorological observation data unit node, through the hydrological monitoring data unit node, to the geological environment data unit node. The initial fusion features are input into this path, and after passing through the hydrological monitoring data unit node, cross-level feature interaction fusion is performed with the original features of the corresponding geological environment data unit.

[0105] Step S1351: In the cross-level feature interaction fusion process, the precipitation feature in the initial fusion feature is transformed through the water level feature of the hydrological monitoring data unit and then interactively mapped with the soil feature of the geological environment data unit to generate the tenth interactive feature; the wind force feature in the initial fusion feature is transformed through the water flow feature of the hydrological monitoring data unit and then interactively mapped with the topographic feature of the geological environment data unit to generate the eleventh interactive feature; the temperature feature in the initial fusion feature is transformed through the water quality feature of the hydrological monitoring data unit and then interactively mapped with the rock strata feature of the geological environment data unit to generate the twelfth interactive feature.

[0106] In the cross-level feature interaction fusion process, each feature in the initial fusion features needs to be transitioned through the corresponding features of the hydrological monitoring data unit before being interactively mapped with the features of the geological environment data unit. For example, the precipitation feature in the initial fusion features is transitioned through the water level feature of the hydrological monitoring data unit and then interactively mapped with the soil feature of the geological environment data unit. That is, first, the influence of precipitation features on water level features is analyzed, then the influence of water level features on soil features is analyzed, and the combined effects of these two processes generate the tenth interactive feature. Similarly, the wind force feature in the initial fusion features is transitioned through the water flow feature of the hydrological monitoring data unit and then interactively mapped with the topographic features of the geological environment data unit. First, the influence of wind force features on water flow features is analyzed, then the influence of water flow features on topographic features is analyzed, generating the eleventh interactive feature. Finally, the temperature feature in the initial fusion features is transitioned through the water quality feature of the hydrological monitoring data unit and then interactively mapped with the rock strata feature of the geological environment data unit. First, the influence of temperature features on water quality features is analyzed, then the influence of water quality features on rock strata features is analyzed, generating the twelfth interactive feature.

[0107] Step S1352: Integrate the tenth interaction feature, the eleventh interaction feature and the twelfth interaction feature to generate a cross-level fusion feature. The cross-level fusion feature carries the path identifier of the fourth influence transmission path and the node identifiers of the corresponding meteorological observation data unit node, hydrological monitoring data unit node and geological environment data unit node.

[0108] The tenth, eleventh, and twelfth interaction features are integrated to generate a cross-level fusion feature. The cross-level fusion feature carries the path identifier of the fourth influence transmission path, as well as the node identifiers of the corresponding meteorological observation data unit node, hydrological monitoring data unit node, and geological environment data unit node.

[0109] Step S136: Collect the first-level fusion features, the second-level fusion features, the third-level fusion features, and the cross-level fusion features, and arrange them in the order of the influence transmission path corresponding to each fusion feature in the dynamic association link structure of the multi-source data to form a progressive fusion feature set containing the association information of each link node. Each fusion feature in the progressive fusion feature set retains the corresponding interaction mapping relationship record.

[0110] Collect the previously generated first-level, second-level, third-level, and cross-level fusion features. Then, arrange them according to the order of the influence propagation paths corresponding to these fusion features in the dynamic association link structure of multi-source data. For example, the first-level fusion features corresponding to the first influence propagation path are listed first, followed by the second-level fusion features corresponding to the second influence propagation path, then the third-level fusion features corresponding to the third influence propagation path, and finally the cross-level fusion features corresponding to the fourth influence propagation path. Through this arrangement, a progressive set of fusion features containing the association information of each link node is formed. In this set, each fusion feature retains a corresponding interaction mapping relationship record for subsequent analysis and tracing.

[0111] Step S140: Generate hydrological influencing factor links based on the link association relationship of the progressive fusion feature set. Each influencing factor in the hydrological influencing factor link is associated with the fusion feature of the corresponding link node in the progressive fusion feature set, and each influencing factor has a corresponding link transmission identifier.

[0112] Based on a progressive fusion feature set, the link relationships between the various fusion features are analyzed. Each fusion feature corresponds to a specific impact transmission path and link node. According to these relationships, the fusion features are converted into impact factors that reflect the hydrological impact. Each impact factor is associated with the fusion feature of the corresponding link node in the progressive fusion feature set and carries a corresponding link transmission identifier. This link transmission identifier can be a path identifier of the impact transmission path, used to indicate the source and transmission path of the impact factor. Connecting these impact factors in a certain order forms a hydrological impact factor link, which can clearly show the transmission relationship between the various impact factors and their comprehensive impact on the hydrological situation.

[0113] Step S141: Analyze the link association relationship between the first-level fusion feature in the progressive fusion feature set and the first influence transmission path of the multi-source data dynamic association link structure, and determine the link transmission element corresponding to the first-level fusion feature. The link transmission element includes the path identifier of the first influence transmission path, the corresponding node identifier, and the interaction mapping relationship record.

[0114] First, we analyze the link association relationship between the first-level fusion features and the first influence transmission path of the dynamic association link structure of multi-source data in the progressive fusion feature set. The first-level fusion features are obtained through fusion via the first influence transmission path; therefore, their link transmission elements include the path identifier of the first influence transmission path, and the node identifiers of the corresponding meteorological observation data unit nodes and hydrological monitoring data unit nodes. They also include records of interactive mapping relationships during the generation process of the first-level fusion features, such as the mapping relationships between the first interactive features, the second interactive features, and the third interactive features. Through the analysis of the above association relationships, we determine the complete link transmission elements corresponding to the first-level fusion features.

[0115] Step S142: Based on the link transmission elements of the first-level fusion features, extract the impact contribution information of each interactive feature in the first-level fusion features on the hydrological changes, and convert the impact contribution information into a first hydrological influence factor. The first hydrological influence factor carries the path identifier of the first influence transmission path as a link transmission identifier.

[0116] Based on the link transmission elements with primary fusion features, the contribution information of each interaction feature to hydrological changes is extracted. For example, the magnitude of the role of the first interaction feature (the interaction between precipitation and water level) in hydrological changes is analyzed, i.e., the contribution information; similarly, the contribution information of the second and third interaction features is analyzed. Then, the above contribution information is comprehensively processed and converted into a primary hydrological influence factor that can quantify its impact on hydrological conditions. The primary hydrological influence factor carries the path identifier of the primary influence transmission path as a link transmission identifier to indicate its source path.

[0117] Step S1421: Extract the interaction mapping relationship record in the link transmission element of the first-level fusion feature, and determine the first interaction feature, second interaction feature and third interaction feature included in the first-level fusion feature.

[0118] Specifically, the interaction mapping relationship records in the first-level fusion feature link transmission elements are extracted, and the first, second, and third interaction features included in the first-level fusion features are identified from these records. This is the basis for subsequent analysis of the influence and contribution information of each interaction feature.

[0119] Step S1422: Retrieve historical hydrological change records, which include hydrological status change information within historical time periods and corresponding historical multi-source data fusion feature information.

[0120] The historical hydrological records of the hydropower station were retrieved. These records contain information on changes in hydrological conditions over a period of time, such as rises and falls in water level and changes in flow rate. They also contain historical multi-source data fusion feature information corresponding to these changes in hydrological conditions, that is, how similar multi-source data fusion features in history have affected changes in hydrological conditions.

[0121] Step S1423: In the historical hydrological change records, select historical first interactive features that have similar interactive mapping relationships with the first interactive feature, and count the number of times the historical first interactive features play a role in the historical hydrological state change process.

[0122] In historical hydrological change records, based on the similarity of interaction mapping relationships, historical first interaction features with similar interaction mapping relationships to the first interaction feature in the current first-level fusion features are selected. Then, the number of times these historical first interaction features played a role in the historical hydrological state changes is counted, that is, the number of times the hydrological state changed accordingly when they appeared.

[0123] Step S1424: Based on the number of interactions, calculate the percentage of influence of the historical first interaction feature on the historical hydrological status changes, and use the percentage of influence as the first contribution information of the first interaction feature to the hydrological status changes.

[0124] Based on the statistically obtained number of interactions, the proportion of influence of the historical first interaction feature in the historical hydrological state change process is calculated. For example, among all factors leading to hydrological state changes, the proportion of influence of the historical first interaction feature is used as the first contribution information of the current first interaction feature to hydrological state changes.

[0125] Step S1425: In the historical hydrological change records, select historical second interactive features that have similar interactive mapping relationships with the second interactive feature, and count the number of times the historical second interactive features play a role in the historical hydrological state change process.

[0126] Similarly, historical second interactive features with similar interactive mapping relationships to the second interactive features were selected from historical hydrological change records, and the number of times they played a role in the historical hydrological state change process was counted.

[0127] Step S1426: Calculate the proportion of the impact of the historical second interaction feature on the historical hydrological status change based on the number of interactions, and use the proportion of the impact as the second impact contribution information of the second interaction feature on the hydrological status change.

[0128] Based on the number of times the second interaction feature has been used in history, its proportion of influence on changes in historical hydrological conditions is calculated, which serves as the second contribution information of the second interaction feature to changes in hydrological conditions.

[0129] Step S1427: In the historical hydrological change records, select historical third interactive features that have similar interactive mapping relationships with the third interactive feature, and count the number of times the historical third interactive feature plays a role in the historical hydrological state change process.

[0130] Historical third interaction features with similar interaction mapping relationships to the third interaction feature were selected from historical water condition change records, and the number of times they were applied was counted.

[0131] Step S1428: Based on the number of interactions, calculate the proportion of the impact of the historical third interaction feature on the historical hydrological status changes, and use the proportion of the impact as the third impact contribution information of the third interaction feature on the hydrological status changes.

[0132] Based on the number of times the third interaction feature has been used in history, its proportion of influence on changes in historical hydrological conditions is calculated, which serves as the third contribution information of the third interaction feature to changes in hydrological conditions.

[0133] Step S1429: Integrate the first impact contribution information, the second impact contribution information, and the third impact contribution information to form the comprehensive impact contribution information corresponding to the first-level fusion feature.

[0134] The first, second, and third impact contribution information are integrated, and their respective impact proportions are comprehensively considered to form the comprehensive impact contribution information corresponding to the first-level fusion feature. This comprehensive impact contribution information fully reflects the overall impact contribution of the first-level fusion feature on hydrological changes.

[0135] Step S14210: Associate and bind the comprehensive impact contribution information with the path identifier and node identifier in the link transmission element of the first-level fusion feature to generate a first hydrological impact factor. The first hydrological impact factor is used to reflect the impact of the first-level fusion feature on hydrological changes through the first impact transmission path.

[0136] By associating and binding the comprehensive impact contribution information with the path identifiers and node identifiers in the first-level fusion feature link transmission elements, the first hydrological impact factor not only includes impact contribution information but can also be traced back to its corresponding path and node. The first hydrological impact factor generated in this way can accurately reflect the impact of the first-level fusion features on hydrological changes through the first impact transmission path.

[0137] Step S143: Analyze the link association relationship between the secondary fusion features in the progressive fusion feature set and the second influence transmission path of the multi-source data dynamic association link structure, and determine the link transmission elements corresponding to the secondary fusion features. The link transmission elements include the path identifier of the second influence transmission path, the corresponding node identifier, and the interaction mapping relationship record.

[0138] Following a similar approach, the link association relationship between the second-level fusion features and the dynamic association link structure of multi-source data in the progressive fusion feature set is analyzed to determine the link transmission elements corresponding to the second-level fusion features, including the path identifier of the second-level influence transmission path, the node identifiers of the corresponding hydrological monitoring data unit nodes and geological environment data unit nodes, and the interactive mapping relationship records in the generation process of the second-level fusion features.

[0139] Step S144: Based on the link transmission elements of the secondary fusion features, extract the impact contribution information of each interactive feature in the secondary fusion features on the hydrological changes, and convert the impact contribution information into a second hydrological influence factor. The second hydrological influence factor carries the path identifier of the second influence transmission path as a link transmission identifier.

[0140] Based on the link transmission elements of the second-level fusion features, the impact contribution information of each interaction feature (fourth interaction feature, fifth interaction feature, and sixth interaction feature) on hydrological changes is extracted, and this information is integrated and converted into a second hydrological influence factor. This second hydrological influence factor carries the path identifier of the second influence transmission path as the link transmission identifier.

[0141] Step S145: Analyze the link association relationship between the three-level fusion features in the progressive fusion feature set and the third influence transmission path of the multi-source data dynamic association link structure, and determine the link transmission elements corresponding to the three-level fusion features. The link transmission elements include the path identifier of the third influence transmission path, the corresponding node identifier, and the interaction mapping relationship record.

[0142] The link association between the three-level fusion features and the third impact transmission path in the progressive fusion feature set is analyzed to determine the link transmission elements corresponding to the three-level fusion features, including the path identifier of the third impact transmission path, the node identifiers of the corresponding geological environment data unit nodes and power plant operation data unit nodes, and the interactive mapping relationship records.

[0143] Step S146: Based on the link transmission elements of the three-level fusion features, extract the impact contribution information of each interactive feature in the three-level fusion features on the hydrological changes, and convert the impact contribution information into a third hydrological influence factor. The third hydrological influence factor carries the path identifier of the third influence transmission path as a link transmission identifier.

[0144] Based on the link transmission elements with three-level fusion characteristics, the impact contribution information of each interaction feature (seventh interaction feature, eighth interaction feature, and ninth interaction feature) is extracted, integrated and converted into the third hydrological influence factor, and the path identifier with the third influence transmission path is used as the link transmission identifier.

[0145] Step S147: Analyze the link association relationship between the cross-level fusion features in the progressive fusion feature set and the fourth influence transmission path of the multi-source data dynamic association link structure, and determine the link transmission elements corresponding to the cross-level fusion features. The link transmission elements include the path identifier of the fourth influence transmission path, the corresponding node identifier, and the interaction mapping relationship record.

[0146] The link relationship between cross-level fusion features and the fourth impact transmission path in the progressive fusion feature set is analyzed to determine the link transmission elements corresponding to the cross-level fusion features, including the path identifier of the fourth impact transmission path, the node identifiers of the corresponding meteorological observation data unit nodes, hydrological monitoring data unit nodes and geological environment data unit nodes, and the interactive mapping relationship records.

[0147] Step S148: Based on the link transmission elements of the cross-level fusion features, extract the impact contribution information of each interactive feature in the cross-level fusion features on the hydrological changes, and convert the impact contribution information into a fourth hydrological influence factor. The fourth hydrological influence factor carries the path identifier of the fourth influence transmission path as a link transmission identifier.

[0148] Based on the cross-level fusion characteristics of the link transmission elements, the impact contribution information of each interaction feature (tenth interaction feature, eleventh interaction feature, and twelfth interaction feature) is extracted, integrated and converted into the fourth hydrological influence factor, and the path identifier with the fourth influence transmission path is used as the link transmission identifier.

[0149] Step S149: According to the transmission order of each influence transmission path in the multi-source data dynamic association link structure, the first hydrological influence factor, the second hydrological influence factor, the third hydrological influence factor and the fourth hydrological influence factor are connected in sequence to form a hydrological influence factor link. Adjacent influence factors in the hydrological influence factor link establish an association relationship through the corresponding link transmission identifier, and each influence factor retains the corresponding fusion feature interaction mapping relationship record.

[0150] Following the transmission order of each influence path in the multi-source data dynamic association link structure—namely, the first influence path, the second influence path, the third influence path, and the fourth influence path—the first, second, third, and fourth hydrological influence factors are sequentially connected. During the connection process, adjacent influence factors establish associations through corresponding link transmission identifiers. For example, the link transmission identifier of the first hydrological influence factor (the path identifier of the first influence path) is associated with the link transmission identifier of the second hydrological influence factor (the path identifier of the second influence path), indicating their transmission relationship. Simultaneously, each influence factor retains a corresponding fusion feature interaction mapping record, forming a complete hydrological influence factor link.

[0151] Step S150: Based on the time period transmission characteristics of the hydrological influencing factor links, perform hydrological trend extrapolation processing, and combine the collection time period information corresponding to the link transmission identifier of each influencing factor to generate hydrological prediction results for power plants containing different extrapolation time periods. The hydrological prediction results for power plants contain hydrological status description information corresponding to each extrapolation time period.

[0152] Hydrological trend extrapolation is performed using the hydrological influencing factor chain. First, the transmission characteristics of each influencing factor in the chain across different time periods are analyzed, i.e., time-period transmission characteristics. Combining the data collection time period information corresponding to the transmission identifiers of each influencing factor chain, the entire prediction time range is divided into different extrapolation periods. For each extrapolation period, the time-period transmission characteristics of the influencing factors and the interrelationships between them are analyzed, such as synergistic relationships, continuity relationships, and cumulative relationships. Based on these relationships, hydrological trend characteristics for each extrapolation period are generated. Finally, hydrological state description information is extracted from the hydrological trend characteristics and integrated to form hydrological prediction results for the power station encompassing different extrapolation periods.

[0153] Step S151: Extract the link transmission identifier of each influencing factor in the hydrological influencing factor link, retrieve the transmission range information of the corresponding influence transmission path according to the link transmission identifier, and determine the collection time period range corresponding to each influencing factor.

[0154] First, the link transmission identifiers of each influencing factor in the hydrological influencing factor chain are extracted. For example, the link transmission identifier of the first hydrological influencing factor is the path identifier of the first influence transmission path. Based on these link transmission identifiers, the transmission range information of the corresponding influence transmission path in the dynamic association link structure of multi-source data is retrieved, thereby determining the collection period range corresponding to each influencing factor, that is, within which time periods the influencing factor is effective.

[0155] Step S152: Based on the effect period of each influencing factor and the continuous length of the collection period, the collection period is divided into three consecutive projection periods, namely the first projection period, the second projection period, and the third projection period. The first projection period corresponds to the collection period in which the effect period of the influencing factor matches the early period of the collection period. The second projection period corresponds to the collection period in which the effect period of the influencing factor matches the middle period of the collection period. The third projection period corresponds to the collection period in which the effect period of the influencing factor matches the late period of the collection period. Each projection period corresponds to a continuous collection period.

[0156] Based on the period of influence of each influencing factor (i.e., the duration of its impact on hydrological conditions) and the continuous length of the data collection period, the data collection period is divided into three consecutive projection periods: the first projection period, the second projection period, and the third projection period. The first projection period corresponds to the early to middle stage of the data collection period, during which the period of influence of the factors matches that of the earlier stage. The second projection period corresponds to the middle stage of the data collection period, and the third projection period corresponds to the later stage. Each projection period is a continuous data collection period used to project hydrological trends for different periods.

[0157] Step S153: For the first simulation period, extract the time-period transmission characteristics of the first and fourth hydrological influencing factors in the hydrological influencing factor link during the first simulation period. The time-period transmission characteristics are the influence change trend information of the influencing factors during the period.

[0158] For the first projection period, the time-period transmission characteristics of the first and fourth hydrological influencing factors in the hydrological influencing factor chain are extracted within this period. Time-period transmission characteristics refer to the changing trend information of the influencing factors' impact on hydrological conditions within this period, such as whether the impact gradually increases, gradually decreases, or remains stable.

[0159] Step S1531: Based on the time range of the first simulation period, extract the influence contribution information of the first hydrological influence factor at each moment within the time range to form a first influence sequence. Each element in the first influence sequence corresponds to the influence contribution information at a moment within the first simulation period.

[0160] Specifically, based on the time range of the first projection period, the impact contribution information of the first hydrological influencing factor at each moment within that period is extracted. This information is then arranged chronologically to form the first impact sequence. Each element in the first impact sequence corresponds to the impact contribution information at a specific moment within the first projection period.

[0161] Step S1532: Extract the influence contribution information of the fourth hydrological influence factor at each moment within the first simulation period to form the fourth influence sequence. Each element in the fourth influence sequence corresponds to the influence contribution information at a moment within the first simulation period.

[0162] Similarly, the impact contribution information of the fourth hydrological influencing factor at each time point within the first projection period is extracted to form the fourth influence sequence.

[0163] Step S1533: Perform trend analysis on the first impact sequence, extract the rising period, falling period and stable period of the first impact sequence, determine the direction of change trend of the first impact sequence, and use the direction of change trend and the corresponding period distribution as the period transmission characteristics of the first hydrological impact factor.

[0164] Trend analysis is performed on the first impact sequence to identify rising periods (gradually increasing impact contribution information), falling periods (gradually decreasing impact contribution information), and stable periods (little changes in impact contribution information). The overall trend direction of the first impact sequence is determined, such as overall rising, overall falling, or overall stable, and the trend direction and the distribution of rising, falling, and stable periods are used as the time-period transmission characteristics of the first hydrological impact factor.

[0165] Step S1534: Perform trend analysis on the fourth influence sequence, extract the rising period, falling period and stable period of the fourth influence sequence, determine the direction of change trend of the fourth influence sequence, and use the direction of change trend and the corresponding period distribution as the period transmission characteristics of the fourth hydrological influence factor.

[0166] A similar trend analysis was performed on the fourth impact sequence to extract its rising period, falling period, and stable period, and to determine the direction of the changing trend, which serves as the time-period transmission characteristics of the fourth hydrological impact factor.

[0167] Step S154: Perform a correlation analysis between the time-period transmission characteristics of the first hydrological influencing factor and the time-period transmission characteristics of the fourth hydrological influencing factor to determine the synergistic relationship of their influence during the first projection period, and generate the first hydrological trend characteristics based on the synergistic relationship.

[0168] A correlation analysis was conducted on the time-period transmission characteristics of the first and fourth hydrological influencing factors. The overlap of their rising, falling, and stable periods, and the consistency of their changing trends, were compared to determine the synergistic relationship between the two factors within the first projection period. This synergy was categorized as unidirectional synergy (same changing trend), inverse synergy (opposite changing trends), or no significant synergy. Based on this synergistic relationship, a first hydrological trend characteristic was generated, reflecting the overall changing trend of the hydrological situation within the first projection period.

[0169] Step S1541: Compare the rising period in the time-transmission characteristics of the first hydrological influencing factor with the rising period in the time-transmission characteristics of the fourth hydrological influencing factor, determine the length of the overlapping period, and calculate the proportion of the overlapping period length to the total length of the first extrapolated period.

[0170] Specifically, by comparing the rising periods in the time-transmission characteristics of the first and fourth hydrological influencing factors, the overlapping periods are identified, and the length of the overlapping periods is calculated. Then, the length of the overlapping periods is divided by the total length of the first projected time period to obtain the proportion of the overlapping period length to the total length of the first projected time period.

[0171] Step S1542: Compare the decreasing period in the time-transmission characteristics of the first hydrological influencing factor with the decreasing period in the time-transmission characteristics of the fourth hydrological influencing factor, determine the length of the overlapping period, and calculate the proportion of the overlapping period length to the total length of the first extrapolated period.

[0172] Similarly, by comparing the descent periods of the two, the length of the overlapping period and its proportion to the total length of the first extrapolated period are calculated.

[0173] Step S1543: Compare the stable period in the time-transmission characteristics of the first hydrological influencing factor with the stable period in the time-transmission characteristics of the fourth hydrological influencing factor, determine the length of the overlapping period, and calculate the proportion of the overlapping period length to the total length of the first extrapolated period.

[0174] By comparing the stable periods of the two, the length of the overlapping period and its proportion to the total length of the first extrapolated period are calculated.

[0175] Step S1544: Based on the length ratio of the three overlapping periods, determine the degree of synergy between the first and fourth hydrological influencing factors in the first projection period. When the overlap ratio of the rising period is the highest, the synergy relationship is determined to be rising synergy in the same direction; when the overlap ratio of the falling period is the highest, the synergy relationship is determined to be falling synergy in the same direction; when the overlap ratio of the stable period is the highest, the synergy relationship is determined to be stable synergy in the same direction.

[0176] Based on the calculated proportions of the three overlapping periods (overlap ratio of the rising period, the falling period, and the stable period), the degree of synergy between the two influencing factors is determined. If the overlap ratio of the rising period is the highest, it indicates the strongest synergy between the two in the upward trend, and the synergistic relationship is determined to be upward synergy; if the overlap ratio of the falling period is the highest, it is downward synergy; if the overlap ratio of the stable period is the highest, it is stable synergy.

[0177] Step S1545: Based on the determined synergistic relationship, and combining the influence contribution information of the first hydrological influence factor and the fourth hydrological influence factor, generate the first hydrological trend feature, which includes information on the synergistic direction, synergistic period and synergistic influence intensity.

[0178] Based on the established synergistic relationships, and combining the contribution information of the first and fourth hydrological influencing factors, a first hydrological trend feature is generated. This first hydrological trend feature includes the synergistic direction (e.g., rising, falling, or stable), the synergistic period (overlapping periods), and the synergistic influence intensity information (determined based on the magnitude of the influence contribution information).

[0179] Step S155: For the second simulation period, extract the time-period transmission characteristics of the second hydrological influencing factor in the hydrological influencing factor link within the second simulation period, and combine them with the residual time-period transmission characteristics of the first hydrological influencing factor within the second simulation period to determine the influence continuity relationship between the two within the second simulation period; generate the second hydrological trend characteristics based on the influence continuity relationship, which are used to reflect the changing trend of hydrological conditions under the continued influence of previous influencing factors within the second simulation period.

[0180] For the second projected period, the temporal transmission characteristics of the second hydrological influencing factor in the hydrological influencing factor chain are extracted within this period. Simultaneously, the residual effects of the first hydrological influencing factor that may still exist in the second projected period are considered, i.e., residual temporal transmission characteristics. The influence continuity relationship between the temporal transmission characteristics of the second hydrological influencing factor and the residual temporal transmission characteristics of the first hydrological influencing factor is analyzed; for example, how the residual effects affect the effectiveness of the current influencing factor. Based on the above influence continuity relationship, a second hydrological trend feature is generated, which reflects the changing trend of hydrological conditions under the continued influence of previous influencing factors within the second projected period.

[0181] Step S156: For the third simulation period, extract the time-period transmission characteristics of the third hydrological influencing factor in the hydrological influencing factor link during the third simulation period, and combine them with the residual time-period transmission characteristics of the second hydrological influencing factor during the third simulation period to determine the cumulative influence relationship between the two during the third simulation period; generate the third hydrological trend characteristics based on the cumulative influence relationship, which are used to reflect the changing trend of hydrological conditions under the cumulative effect of previous influencing factors during the third simulation period.

[0182] For the third projection period, the temporal transmission characteristics of the third hydrological influencing factor and the residual temporal transmission characteristics of the second hydrological influencing factor within this period are extracted. The cumulative impact relationship between the two is analyzed, i.e., how the residual impact of previous influencing factors and the impact of current influencing factors cumulatively affect the hydrological situation. Based on the above cumulative impact relationship, third hydrological trend characteristics are generated, reflecting the changing trend of the hydrological situation under the cumulative effect of previous influencing factors during the third projection period.

[0183] Step S157: Collect the first hydrological trend feature, the second hydrological trend feature, and the third hydrological trend feature, and extract the corresponding hydrological status description information from each trend feature. The hydrological status description information includes the direction of hydrological change, change characteristics, and description of the effects of influencing factors within the simulation period. In the order of the first simulation period, the second simulation period, and the third simulation period, integrate the hydrological status description information corresponding to each simulation period to form a hydrological prediction result for the power station containing different simulation periods. The hydrological status description information for each simulation period in the hydrological prediction result of the power station carries a corresponding influencing factor link association identifier.

[0184] The hydrological trend features of the first, second, and third periods are collected, and hydrological status description information is extracted from each trend feature. This description includes the direction of hydrological change (rising, falling, stable), change characteristics (such as rate of change, fluctuations, etc.), and explanations of the effects of influencing factors (which factors play a major role and how they act). Then, the hydrological status description information for each period is integrated according to the chronological order of the first, second, and third periods to form hydrological prediction results for the power station encompassing different periods. Each period's hydrological status description information is accompanied by a corresponding influencing factor link association identifier to indicate the source of that influencing factor link.

[0185] For example, in step S1571: extract the coordinating direction from the first hydrological trend features as the first hydrological change direction, extract the hydrological change pattern corresponding to the coordinating period as the first hydrological change feature, and extract the synergistic effect description of the first hydrological influencing factor and the fourth hydrological influencing factor as the effect description of the first influencing factor; integrate the first hydrological change direction, the first hydrological change feature and the effect description of the first influencing factor to form the first hydrological state description information, which carries the time period identifier of the first projection period and the corresponding influencing factor link association identifier.

[0186] The coordinated direction is extracted from the first hydrological trend characteristics as the first hydrological change direction. For example, if the coordinated direction is upward, then the first hydrological change direction is upward. The hydrological change pattern corresponding to the coordinated period is extracted; if the hydrological level shows an accelerating upward trend within the coordinated period, this is taken as the first hydrological change characteristic. The synergistic effect description between the first and fourth hydrological influencing factors is extracted; if the combined effect of both leads to an increase in the hydrological level, this is taken as the effect description of the first influencing factor. The above information is integrated to form the first hydrological state description information, including the time period identifier of the first projection period and the corresponding influencing factor link association identifier.

[0187] Step S1572: Extract the direction of water level change corresponding to the influence continuity relationship from the second water level trend characteristics as the second water level change direction; extract the water level change pattern in the influence continuity process as the second water level change feature; extract the description of the continuity effect of the second water level influence factor and the first water level influence factor as the description of the effect of the second influence factor; integrate the second water level change direction, the second water level change feature and the description of the effect of the second influence factor to form the second water level status description information, which carries the time period identifier of the second projection period and the corresponding influence factor link association identifier.

[0188] The direction of hydrological change corresponding to the continuity of influence is extracted from the characteristics of the second hydrological trend as the direction of hydrological change in the second hydrological situation. The pattern of hydrological change in the process of the continuity of influence is extracted as the characteristics of hydrological change in the second hydrological situation. The description of the continuity of relevant influencing factors is extracted as the description of the effect of the second influencing factors. The information is integrated to form the description of the state of the second hydrological situation, with the time period identifier of the second projection period and the link association identifier of the influencing factors.

[0189] Step S1573: Extract the direction of water level change corresponding to the cumulative relationship of the third water level trend features as the direction of water level change of the third water level; extract the pattern of water level change in the cumulative process of the influence as the feature of water level change of the third water level; extract the description of the cumulative effect of the third water level influence factor and the second water level influence factor as the description of the effect of the third influence factor; integrate the direction of water level change of the third water level, the feature of water level change of the third water level, and the description of the effect of the third influence factor to form the description information of the state of the third water level. The description information of the state of the third water level carries the time period identifier of the third projection period and the corresponding influence factor link association identifier.

[0190] The direction of water level change corresponding to the cumulative relationship of the third water level trend characteristics is extracted as the direction of water level change of the third water level. The pattern of water level change in the process of cumulative influence is extracted as the feature of water level change of the third water level. The cumulative effect description of relevant influencing factors is extracted as the effect description of the third influencing factors. The information is integrated to form the description of the state of the third water level, with the time period identifier of the third projection period and the link association identifier of the influencing factors.

[0191] Step S1574: Arrange the first hydrological status description information, the second hydrological status description information, and the third hydrological status description information in chronological order according to the first simulation period, the second simulation period, and the third simulation period; compare the end part of the first hydrological status description information with the beginning part of the second hydrological status description information to analyze whether the influence transmission relationship between the two is consistent; compare the end part of the second hydrological status description information with the beginning part of the third hydrological status description information to analyze whether the influence transmission relationship between the two is consistent; on the basis of maintaining the consistency of the influence transmission relationship, supplement the association description of the influence factor link association identifier corresponding to each hydrological status description information to form the hydrological prediction results of the power station containing different simulation periods. The hydrological prediction results of the power station are used to reflect the hydrological changes in each simulation period and the corresponding multi-source data fusion influence process.

[0192] The first, second, and third hydrological status descriptions are arranged chronologically. The beginning and end of the hydrological status descriptions for adjacent time periods are compared to analyze the consistency of influence transmission relationships. For example, can the hydrological trend at the end of the first hydrological status description naturally transition to the trend at the beginning of the second hydrological status description? While ensuring consistency in influence transmission relationships, the association descriptions of the influence factor links corresponding to each hydrological status description are supplemented, such as the transmission order and interaction between each influence factor link. This ultimately forms the hydrological prediction results for the power station, encompassing different projection time periods. These prediction results comprehensively reflect the hydrological changes in each projection time period and the corresponding multi-source data fusion impact process.

[0193] Figure 2 The illustration shows exemplary hardware and software components of a power plant hydrological prediction system 100 based on multi-source data fusion, which can implement the ideas of this application, according to some embodiments of this application. For example, a processor 120 can be used in the power plant hydrological prediction system 100 based on multi-source data fusion and to perform the functions in this application.

[0194] For example, a power plant hydrological prediction system 100 based on multi-source data fusion may include a network port 110 connected to a network, one or more processors 120 for executing program instructions, a communication bus 130, and various forms of storage media 140, such as a disk, ROM, or RAM, or any combination thereof. Exemplarily, the power plant hydrological prediction system 100 based on multi-source data fusion may also include program instructions stored in ROM, RAM, or other types of non-transitory storage media, or any combination thereof. The methods of this application can be implemented according to these program instructions. The power plant hydrological prediction system 100 based on multi-source data fusion also includes an I / O interface 150 between the computer and other input / output devices.

[0195] Furthermore, this embodiment of the invention also provides a readable storage medium, wherein computer-executable instructions are preset in the readable storage medium, and when the processor executes the computer-executable instructions, the above-mentioned power station hydrological prediction method based on multi-source data fusion is implemented.

[0196] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. A hydrological prediction method for power plants based on multi-source data fusion, characterized in that, The method includes: Acquire a multi-source data set related to the hydrological conditions of the power station. The multi-source data set includes meteorological observation data units, hydrological monitoring data units, geological environment data units, and power station operation data units. Each data unit is marked with a collection period identifier. Using each data unit in the multi-source data set as a link node, the influence transmission path between nodes is set according to the water situation influence transmission relationship between different data units, and the transmission direction and transmission range of each influence transmission path are marked to form a multi-source data dynamic association link structure that includes node association relationships. Based on the multi-source data dynamic association link structure, a progressive fusion process is performed. Starting from the first link node of the multi-source data dynamic association link structure, the fusion features obtained from the previous level of fusion processing are sequentially fused with the data units corresponding to the subsequent link nodes to generate a progressive fusion feature set containing the association information of each link node. Hydrological influence factor links are generated based on the link association relationship of the progressive fusion feature set. Each influence factor in the hydrological influence factor link is associated with the fusion feature of the corresponding link node in the progressive fusion feature set, and each influence factor has a corresponding link transmission identifier. Based on the time-period transmission characteristics of the hydrological influencing factor links, hydrological trend extrapolation processing is performed. Combining the collection time period information corresponding to the link transmission identifier of each influencing factor, hydrological prediction results of the power station containing different extrapolation time periods are generated. The hydrological prediction results of the power station contain hydrological status description information corresponding to each extrapolation time period.

2. The hydrological prediction method for power plants based on multi-source data fusion according to claim 1, characterized in that, The process involves using each data unit in the multi-source data set as a link node, setting influence transmission paths between nodes based on the hydrological influence transmission relationships between different data units, and marking the transmission direction and range of each influence transmission path to form a dynamic multi-source data association link structure that includes node association relationships. Extract the hydrological impact transmission attribute of each data unit in the multi-source data set. The hydrological impact transmission attribute is the attribute information that the data unit can have a hydrological impact on other data units. The hydrological impact transmission attribute of different types of data units contains different impact dimension information. Analyze the matching relationship of hydrological impact transmission attributes between the meteorological observation data unit and the hydrological monitoring data unit, determine the effectiveness of direct impact transmission between the two, and when the effectiveness of direct impact transmission meets the preset transmission conditions, set the first impact transmission path from the meteorological observation data unit node to the hydrological monitoring data unit node. The direction of the first influence transmission path is marked as from the meteorological observation data unit node to the hydrological monitoring data unit node, and the transmission range of the first influence transmission path is marked as the time range corresponding to all collection time period identifiers of the meteorological observation data unit and the hydrological monitoring data unit. Analyze the matching relationship of hydrological impact transmission attributes between the hydrological monitoring data unit and the geological environment data unit to determine the effectiveness of direct impact transmission between them. When the effectiveness of direct impact transmission meets the preset transmission conditions, set a second impact transmission path from the hydrological monitoring data unit node to the geological environment data unit node. The direction of the second impact transmission path is marked as from the hydrological monitoring data unit node to the geological environment data unit node, and the transmission range of the second impact transmission path is marked as covering the time range corresponding to all collection time period identifiers of the hydrological monitoring data unit and the geological environment data unit. Analyze the matching relationship of hydrological influence transmission attributes between the geological environment data unit and the power station operation data unit to determine the effectiveness of direct influence transmission between the two. When the effectiveness of direct influence transmission meets the preset transmission conditions, set a third influence transmission path from the geological environment data unit node to the power station operation data unit node. The direction of the transmission of the third influence transmission path is marked as from the geological environment data unit node to the power plant operation data unit node, and the transmission range of the third influence transmission path is the time range corresponding to all collection time period identifiers of the geological environment data unit and the power plant operation data unit. The matching relationship of the indirect hydrological impact transmission attributes from the meteorological observation data unit to the geological environment data unit through the hydrological monitoring data unit is analyzed. The transmission effectiveness of the first and second impact transmission paths is combined to determine the indirect impact transmission effectiveness. When the indirect impact transmission effectiveness meets the preset transmission conditions, a fourth impact transmission path is set from the meteorological observation data unit node to the geological environment data unit node through the hydrological monitoring data unit node. The direction of the fourth impact transmission path is marked as from the meteorological observation data unit node through the hydrological monitoring data unit node to the geological environment data unit node, and the transmission range of the fourth impact transmission path is the intersection time range corresponding to all collection time period identifiers of the meteorological observation data unit, the hydrological monitoring data unit, and the geological environment data unit. The first influence transmission path, the second influence transmission path, the third influence transmission path, and the fourth influence transmission path are integrated to form a multi-source data dynamic association link structure that includes node association relationships. Each influence transmission path in the multi-source data dynamic association link structure has a unique path identifier and corresponding transmission direction and transmission range information.

3. The hydrological prediction method for power plants based on multi-source data fusion according to claim 2, characterized in that, The extraction of the hydrological impact transmission attributes of each data unit in the multi-source dataset includes: The precipitation characteristics, wind force characteristics, and temperature characteristics are extracted from the meteorological observation data unit. The precipitation characteristics are attribute information related to precipitation in the meteorological observation data unit, the wind force characteristics are attribute information related to wind force in the meteorological observation data unit, and the temperature characteristics are attribute information related to temperature in the meteorological observation data unit. The precipitation characteristics, wind force characteristics, and temperature characteristics are used as the water situation impact transmission attributes of the meteorological observation data unit, wherein the precipitation characteristics correspond to the precipitation impact dimension, the wind force characteristics correspond to the wind force impact dimension, and the temperature characteristics correspond to the temperature impact dimension. Extract water level features, flow features, and water quality features from the hydrological monitoring data unit. The water level features are attribute information related to water level in the hydrological monitoring data unit, the flow features are attribute information related to flow in the hydrological monitoring data unit, and the water quality features are attribute information related to water quality in the hydrological monitoring data unit. The water level feature, the water flow feature, and the water quality feature are used as the water situation impact transmission attributes of the hydrological monitoring data unit, wherein the water level feature corresponds to the water level impact dimension, the water flow feature corresponds to the water flow impact dimension, and the water quality feature corresponds to the water quality impact dimension. Soil features, topographic features, and rock strata features are extracted from the geological environment data unit. The soil features are attribute information related to soil in the geological environment data unit, the topographic features are attribute information related to topography in the geological environment data unit, and the rock strata features are attribute information related to rock strata in the geological environment data unit. The soil features, the topographic features, and the rock strata features are used as the water situation impact transmission attributes of the geological environment data unit, where the soil features correspond to the soil impact dimension, the topographic features correspond to the topographic impact dimension, and the rock strata features correspond to the rock strata impact dimension. Extract the unit operation characteristics, gate control characteristics, and water conveyance characteristics from the power plant operation data unit. The unit operation characteristics are attribute information related to unit operation in the power plant operation data unit. The gate control characteristics are attribute information related to gate control in the power plant operation data unit. The water conveyance characteristics are attribute information related to water conveyance in the power plant operation data unit. The unit operation characteristics, the gate control characteristics, and the water conveyance characteristics are used as the water situation impact transmission attributes of the power station operation data unit, wherein the unit operation characteristics correspond to the unit impact dimension, the gate control characteristics correspond to the gate impact dimension, and the water conveyance characteristics correspond to the water conveyance impact dimension.

4. The hydrological prediction method for power plants based on multi-source data fusion according to claim 1, characterized in that, The progressive fusion processing based on the multi-source data dynamic association link structure involves, starting from the first link node of the multi-source data dynamic association link structure, sequentially performing feature interaction fusion with the data units corresponding to the subsequent link nodes to generate a progressive fusion feature set containing the association information of each link node, including: The first link node of the multi-source data dynamic association link structure is determined to be the meteorological observation data unit node. The original features of the meteorological observation data unit corresponding to the meteorological observation data unit node are extracted and the original features of the meteorological observation data unit are used as the initial fusion features. The first influence transmission path in the multi-source data dynamic association link structure is invoked, the initial fusion feature is input into the first influence transmission path, and the original feature interaction fusion is performed with the original feature of the hydrological monitoring data unit corresponding to the hydrological monitoring data unit node pointed to by the first influence transmission path. During the initial feature interaction fusion process, the precipitation feature in the initial fused features is interactively mapped with the water level feature in the original features of the hydrological monitoring data unit to generate a first interactive feature; the wind force feature in the initial fused features is interactively mapped with the water flow feature in the original features of the hydrological monitoring data unit to generate a second interactive feature; and the temperature feature in the initial fused features is interactively mapped with the water quality feature in the original features of the hydrological monitoring data unit to generate a third interactive feature. The first interaction feature, the second interaction feature, and the third interaction feature are integrated to generate a first-level fusion feature. The first-level fusion feature carries the path identifier of the first influence transmission path and the node identifiers of the corresponding meteorological observation data unit node and hydrological monitoring data unit node. The second influence transmission path in the multi-source data dynamic association link structure is invoked, the first-level fusion feature is input into the second influence transmission path, and a second feature interaction fusion is performed with the original feature of the geological environment data unit corresponding to the geological environment data unit node pointed to by the second influence transmission path. In the secondary feature interaction fusion process, the first interactive feature in the primary fusion feature is interactively mapped with the soil feature in the original feature of the geological environment data unit to generate the fourth interactive feature; the second interactive feature in the primary fusion feature is interactively mapped with the topographic feature in the original feature of the geological environment data unit to generate the fifth interactive feature; and the third interactive feature in the primary fusion feature is interactively mapped with the rock strata feature in the original feature of the geological environment data unit to generate the sixth interactive feature. The fourth interaction feature, the fifth interaction feature, and the sixth interaction feature are integrated to generate a secondary fusion feature. The secondary fusion feature carries the path identifier of the second influence transmission path and the node identifiers of the corresponding hydrological monitoring data unit node and geological environment data unit node. The third influence transmission path in the multi-source data dynamic association link structure is invoked, and the secondary fusion feature is input into the third influence transmission path. The original features of the power plant operation data unit corresponding to the power plant operation data unit node pointed to by the third influence transmission path are fused three times. During the three-stage feature interaction fusion process, the fourth interaction feature in the secondary fusion feature is interactively mapped with the unit operation feature in the original feature of the power plant operation data unit to generate the seventh interaction feature; the fifth interaction feature in the secondary fusion feature is interactively mapped with the gate control feature in the original feature of the power plant operation data unit to generate the eighth interaction feature; and the sixth interaction feature in the secondary fusion feature is interactively mapped with the water conveyance feature in the original feature of the power plant operation data unit to generate the ninth interaction feature. The seventh interaction feature, the eighth interaction feature, and the ninth interaction feature are integrated to generate a three-level fusion feature. The three-level fusion feature carries the path identifier of the third influence transmission path and the node identifiers of the corresponding geological environment data unit node and power plant operation data unit node. The fourth influence transmission path in the multi-source data dynamic association link structure is invoked, and the initial fusion feature is input into the fourth influence transmission path. After passing through the hydrological monitoring data unit node, it performs cross-level feature interaction fusion with the original features of the geological environment data unit corresponding to the geological environment data unit node. In the cross-level feature interaction fusion process, the precipitation feature in the initial fusion feature is transformed through the water level feature of the hydrological monitoring data unit and then interactively mapped with the soil feature of the geological environment data unit to generate the tenth interactive feature; the wind force feature in the initial fusion feature is transformed through the water flow feature of the hydrological monitoring data unit and then interactively mapped with the topographic feature of the geological environment data unit to generate the eleventh interactive feature; the temperature feature in the initial fusion feature is transformed through the water quality feature of the hydrological monitoring data unit and then interactively mapped with the rock strata feature of the geological environment data unit to generate the twelfth interactive feature. Integrate the tenth interaction feature, the eleventh interaction feature and the twelfth interaction feature to generate a cross-level fusion feature. The cross-level fusion feature carries the path identifier of the fourth influence transmission path and the node identifiers of the corresponding meteorological observation data unit node, hydrological monitoring data unit node and geological environment data unit node. The first-level fusion features, the second-level fusion features, the third-level fusion features, and the cross-level fusion features are collected and arranged in the order of the influence transmission path corresponding to each fusion feature in the dynamic association link structure of the multi-source data to form a progressive fusion feature set containing the association information of each link node. Each fusion feature in the progressive fusion feature set retains the corresponding interaction mapping relationship record.

5. The hydrological prediction method for power plants based on multi-source data fusion according to claim 4, characterized in that, In the initial feature interaction fusion process, the precipitation feature in the initial fused features is interactively mapped with the water level feature in the original features of the hydrological monitoring data unit to generate a first interactive feature; the wind force feature in the initial fused features is interactively mapped with the water flow feature in the original features of the hydrological monitoring data unit to generate a second interactive feature; and the temperature feature in the initial fused features is interactively mapped with the water quality feature in the original features of the hydrological monitoring data unit to generate a third interactive feature, including: Extract the collection period identifier corresponding to the precipitation feature in the initial fusion feature, extract the collection period identifier corresponding to the water level feature in the original feature of the hydrological monitoring data unit, and select the period with the same collection period identifier as the first interaction period. During the first interaction period, the change trend information of the precipitation feature in each period is extracted, the change trend information of the water level feature in each period is extracted, and the correspondence between the change trend information of the precipitation feature and the change trend information of the water level feature is established. Based on the correspondence, the changing trend information of the precipitation characteristics is mapped to the dimensional space of the changing trend information of the water level characteristics to generate the mapped first trend association information; By integrating the changing trend information of the precipitation characteristics, the changing trend information of the water level characteristics, and the mapped first trend correlation information, a first interactive feature is generated; Extract the collection period identifier corresponding to the wind force feature in the initial fusion feature, extract the collection period identifier corresponding to the water flow feature in the original feature of the hydrological monitoring data unit, and select the period with the same collection period identifier as the second interaction period. During the second interaction period, the variation range information of the wind force feature in each period is extracted, the variation range information of the water flow feature in each period is extracted, and the correspondence between the variation range information of the wind force feature and the variation range information of the water flow feature is established. Based on the correspondence, the variation amplitude information of the wind force feature is mapped to the dimension space of the variation amplitude information of the water flow feature to generate the first amplitude association information after mapping; By integrating the variation amplitude information of the wind force feature, the variation amplitude information of the water flow feature, and the mapped first amplitude correlation information, a second interactive feature is generated; Extract the collection period identifier corresponding to the temperature feature in the initial fusion features, extract the collection period identifier corresponding to the water quality feature in the original features of the hydrological monitoring data unit, and select the period with the same collection period identifier as the third interaction period. During the third interaction period, the stable state information of the temperature feature in each period is extracted, the stable state information of the water quality feature in each period is extracted, and the correspondence between the stable state information of the temperature feature and the stable state information of the water quality feature is established. Based on the correspondence, the stable state information of the temperature feature is mapped to the dimensional space of the stable state information of the water quality feature to generate the first stable association information after mapping. By integrating the stable state information of the temperature feature, the stable state information of the water quality feature, and the mapped first stable correlation information, a third interactive feature is generated.

6. The hydrological prediction method for power plants based on multi-source data fusion according to claim 1, characterized in that, The step of generating hydrological influencing factor links based on the link correlation of the progressive fusion feature set includes: Analyze the link association relationship between the first-level fusion feature in the progressive fusion feature set and the first influence transmission path of the multi-source data dynamic association link structure, and determine the link transmission elements corresponding to the first-level fusion feature. The link transmission elements include the path identifier of the first influence transmission path, the corresponding node identifier, and the interaction mapping relationship record. Based on the link transmission elements of the first-level fusion features, the influence contribution information of each interaction feature in the first-level fusion features on the hydrological changes is extracted, and the influence contribution information is converted into a first hydrological influence factor. The first hydrological influence factor carries the path identifier of the first influence transmission path as a link transmission identifier. Analyze the link association relationship between the secondary fusion features in the progressive fusion feature set and the second influence transmission path of the multi-source data dynamic association link structure, and determine the link transmission elements corresponding to the secondary fusion features. The link transmission elements include the path identifier of the second influence transmission path, the corresponding node identifier, and the interaction mapping relationship record. Based on the link transmission elements of the second-level fusion features, the influence contribution information of each interaction feature in the second-level fusion features on the hydrological changes is extracted, and the influence contribution information is converted into a second hydrological influence factor. The second hydrological influence factor carries the path identifier of the second influence transmission path as a link transmission identifier. Analyze the link association relationship between the three-level fusion features in the progressive fusion feature set and the third influence transmission path of the multi-source data dynamic association link structure, and determine the link transmission elements corresponding to the three-level fusion features. The link transmission elements include the path identifier of the third influence transmission path, the corresponding node identifier, and the interaction mapping relationship record. Based on the link transmission elements of the three-level fusion features, the influence contribution information of each interaction feature in the three-level fusion features on the hydrological changes is extracted, and the influence contribution information is converted into a third hydrological influence factor. The third hydrological influence factor carries the path identifier of the third influence transmission path as a link transmission identifier. Analyze the link association relationship between the cross-level fusion features in the progressive fusion feature set and the fourth influence transmission path of the multi-source data dynamic association link structure, and determine the link transmission elements corresponding to the cross-level fusion features. The link transmission elements include the path identifier of the fourth influence transmission path, the corresponding node identifier, and the interaction mapping relationship record. Based on the link transmission elements of the cross-level fusion features, the influence contribution information of each interaction feature in the cross-level fusion features on the hydrological changes is extracted, and the influence contribution information is converted into a fourth hydrological influence factor. The fourth hydrological influence factor carries the path identifier of the fourth influence transmission path as a link transmission identifier. According to the transmission order of each influence transmission path in the multi-source data dynamic association link structure, the first hydrological influence factor, the second hydrological influence factor, the third hydrological influence factor and the fourth hydrological influence factor are connected in sequence to form a hydrological influence factor link. Adjacent influence factors in the hydrological influence factor link establish an association relationship through the corresponding link transmission identifier, and each influence factor retains the corresponding fusion feature interaction mapping relationship record.

7. The hydrological prediction method for power plants based on multi-source data fusion according to claim 6, characterized in that, The link transmission elements based on the first-level fusion features extract the impact contribution information of each interaction feature in the first-level fusion features on hydrological changes, and convert the impact contribution information into a first hydrological influence factor, including: Extract the interaction mapping relationship records from the link transmission elements of the first-level fusion feature, and determine the first interaction feature, second interaction feature, and third interaction feature included in the first-level fusion feature; Retrieve historical hydrological change records, which include information on changes in hydrological status during historical periods and corresponding historical multi-source data fusion feature information; In the historical hydrological change records, historical first interaction features that have similar interaction mapping relationships with the first interaction feature are selected, and the number of times the historical first interaction feature plays a role in the historical hydrological state change process is counted. Based on the number of interactions, the proportion of the impact of the historical first interaction feature on the historical hydrological status change is calculated, and the proportion of the impact is used as the first impact contribution information of the first interaction feature on the hydrological status change. In the historical hydrological change records, historical second interactive features that have similar interactive mapping relationships with the second interactive feature are selected, and the number of times the historical second interactive features play a role in the historical hydrological state change process is counted. Based on the number of interactions, the proportion of the impact of the historical second interaction feature on the historical hydrological status change is calculated, and the proportion of the impact is used as the second impact contribution information of the second interaction feature on the hydrological status change. In the historical hydrological change records, historical third interactive features that have similar interactive mapping relationships with the third interactive feature are selected, and the number of times the historical third interactive feature plays a role in the historical hydrological state change process is counted. Based on the number of interactions, the proportion of the impact of the historical third interaction feature on the historical hydrological status changes is calculated, and the proportion of the impact is used as the third impact contribution information of the third interaction feature on the hydrological status changes. Integrate the first impact contribution information, the second impact contribution information, and the third impact contribution information to form the comprehensive impact contribution information corresponding to the first-level fusion feature; The comprehensive impact contribution information is associated and bound with the path identifier and node identifier in the link transmission element of the first-level fusion feature to generate a first hydrological impact factor. The first hydrological impact factor is used to reflect the impact of the first-level fusion feature on hydrological changes through the first impact transmission path.

8. The hydrological prediction method for power plants based on multi-source data fusion according to claim 1, characterized in that, The hydrological trend extrapolation process, based on the time-period transmission characteristics of the hydrological influencing factor links, combines the data collection time period information corresponding to the link transmission identifiers of each influencing factor to generate hydrological prediction results for power stations containing different extrapolation time periods, including: Extract the link transmission identifier of each influencing factor in the hydrological influencing factor link, retrieve the transmission range information of the corresponding influence transmission path based on the link transmission identifier, and determine the collection period range corresponding to each influencing factor; Based on the duration of each influencing factor and the continuous length of the data collection period, the data collection period is divided into three consecutive projection periods: the first projection period, the second projection period, and the third projection period. The first projection period corresponds to the data collection period in which the duration of the influencing factors matches the early period of the data collection period; the second projection period corresponds to the data collection period in which the duration of the influencing factors matches the middle period of the data collection period; and the third projection period corresponds to the data collection period in which the duration of the influencing factors matches the later period of the data collection period. Each projection period corresponds to a continuous data collection period. For the first simulation period, the time-period transmission characteristics of the first and fourth hydrological influencing factors in the hydrological influencing factor chain are extracted during the first simulation period. The time-period transmission characteristics are the influence change trend information of the influencing factors during the time period. The time-period transmission characteristics of the first hydrological influencing factor and the time-period transmission characteristics of the fourth hydrological influencing factor are correlated to determine the synergistic relationship of their influence in the first projection period, and the trend characteristics of the first hydrological condition are generated based on the synergistic relationship. For the second simulation period, the time-period transmission characteristics of the second hydrological influence factor in the hydrological influence factor link are extracted during the second simulation period. Combined with the residual time-period transmission characteristics of the first hydrological influence factor during the second simulation period, the influence continuity relationship between the two during the second simulation period is determined. Based on the aforementioned influence continuity relationship, a second hydrological trend feature is generated. The second hydrological trend feature is used to reflect the changing trend of hydrological conditions under the continued influence of previous influencing factors during the second simulation period. For the third simulation period, the time-period transmission characteristics of the third hydrological influence factor in the hydrological influence factor link are extracted during the third simulation period. Combined with the residual time-period transmission characteristics of the second hydrological influence factor during the third simulation period, the cumulative influence relationship between the two during the third simulation period is determined. Based on the aforementioned cumulative influence relationship, a third hydrological trend feature is generated. This third hydrological trend feature is used to reflect the changing trend of hydrological conditions under the cumulative effect of previous influencing factors during the third simulation period. Collect the first hydrological trend features, the second hydrological trend features, and the third hydrological trend features, and extract the corresponding hydrological status description information from each trend feature. The hydrological status description information includes the direction of hydrological change, change characteristics, and description of the effects of influencing factors during the simulation period. Following the order of the first, second, and third simulation periods, the hydrological status description information corresponding to each simulation period is integrated to form a hydrological prediction result for the power station containing different simulation periods. The hydrological status description information for each simulation period in the hydrological prediction result of the power station carries a corresponding influence factor link association identifier.

9. The hydrological prediction method for power plants based on multi-source data fusion according to claim 8, characterized in that, For the first simulation period, the time-period transmission characteristics of the first and fourth hydrological influencing factors in the hydrological influencing factor chain are extracted within the first simulation period. These time-period transmission characteristics represent the trend information of the influencing factors' influence changes within that time period. A correlation analysis is performed between the time-period transmission characteristics of the first and fourth hydrological influencing factors to determine their synergistic influence relationship within the first simulation period. Based on this synergistic influence relationship, a first hydrological trend feature is generated, including: Based on the time range of the first simulation period, the influence contribution information of the first hydrological influence factor at each moment within the time range is extracted to form the first influence sequence. Each element in the first influence sequence corresponds to the influence contribution information at a moment within the first simulation period. The influence contribution information of the fourth hydrological influencing factor at each moment within the first simulation period is extracted to form the fourth influence sequence. Each element in the fourth influence sequence corresponds to the influence contribution information at a moment within the first simulation period. Trend analysis is performed on the first impact sequence to extract the rising period, falling period and stable period of the first impact sequence, and the direction of change trend of the first impact sequence is determined. The direction of change trend and the corresponding time period distribution are used as the time period transmission characteristics of the first hydrological impact factor. Trend analysis is performed on the fourth impact sequence to extract the rising period, falling period and stable period of the fourth impact sequence, and the direction of the change trend of the fourth impact sequence is determined. The direction of the change trend and the corresponding period distribution are used as the period transmission characteristics of the fourth hydrological impact factor. By comparing the rising period in the time-transmission characteristics of the first hydrological influencing factor with the rising period in the time-transmission characteristics of the fourth hydrological influencing factor, the length of the overlapping period is determined, and the proportion of the overlapping period length to the total length of the first extrapolated period is calculated. By comparing the decreasing period in the time-transmission characteristics of the first hydrological influencing factor with the decreasing period in the time-transmission characteristics of the fourth hydrological influencing factor, the length of the overlapping period is determined, and the proportion of the overlapping period length to the total length of the first extrapolated period is calculated. By comparing the stable periods in the time-transmission characteristics of the first hydrological influencing factor with the stable periods in the time-transmission characteristics of the fourth hydrological influencing factor, the length of the overlapping period is determined, and the proportion of the overlapping period length to the total length of the first extrapolated period is calculated. Based on the proportion of the length of the three overlapping periods, the degree of synergy between the first and fourth hydrological influencing factors in the first projection period is determined. When the overlap ratio of the rising period is the highest, the synergy relationship is determined to be rising synergy in the same direction; when the overlap ratio of the falling period is the highest, the synergy relationship is determined to be falling synergy in the same direction; when the overlap ratio of the stable period is the highest, the synergy relationship is determined to be stable synergy in the same direction. Based on the established synergistic relationship, and combining the influence contribution information of the first and fourth hydrological factors, a first hydrological trend feature is generated. The first hydrological trend feature includes information on the synergistic direction, synergistic period, and synergistic influence intensity.

10. A hydrological prediction system for power plants based on multi-source data fusion, characterized in that, The hydrological prediction system for power plants based on multi-source data fusion includes a processor and a memory. The memory and the processor are connected. The memory is used to store programs, instructions, or code. The processor is used to execute the programs, instructions, or code in the memory to implement the hydrological prediction method for power plants based on multi-source data fusion as described in any one of claims 1-9.

Images