An intelligent dispatching management system for water conservancy projects

Abstract

The application discloses a kind of water conservancy engineering intelligent scheduling management systems, it is related to water conservancy intelligent scheduling technical field, including data acquisition module, feature extraction module, conflict measurement module and parallel optimization module;Extract time series scheduling dataset, according to scheduling time period to time series scheduling dataset is aggregated and is handled calculation, obtains water storage margin characteristic value and drainage urgency characteristic value, then the state classification of each time period reservoir is obtained, and time series characteristic vector set is obtained;Parallel reservoir adjacency matrix is constructed and calculated to obtain total inflow, the path conflict coefficient of reservoir;According to time series characteristic vector set and path conflict coefficient, scheduling classification reservoir label is set, and target reservoir is adjusted according to scheduling classification reservoir label Gate opening, generates scheduling adjustment instruction, effectively matches the timely demand of different reservoirs and the safety constraint of whole reservoir network, realizes the fine generation of control instruction.

Description

Technical Field

[0001] This invention relates to the field of intelligent water conservancy scheduling technology, and in particular to an intelligent scheduling and management system for water conservancy projects. Background Technology

[0002] With the increasing contradiction between water supply and demand, traditional manual experience-based scheduling models are no longer sufficient to meet the multi-dimensional needs of modern water conservancy projects for water supply and demand security and ecological protection. Existing scheduling systems mostly rely on static rules or single-objective optimization, often neglecting the spatiotemporal coupling effects between cascade reservoirs and the dynamic game under complex constraints.

[0003] Currently, a Chinese invention with publication number CN119886497B discloses an intelligent scheduling method and system for water conservancy projects. This method collects water conveyance routes and reservoir water transfer records, then identifies water-demanding and water-supplying reservoirs, calculates their matching index, establishes a scheme library for each water-demanding party based on the matching index, and adds water-supplying parties to the scheme library. It then plans the optimal path based on the scheme library, generates a scheduling plan, pushes it to each reservoir for execution, and obtains feedback. However, this scheme always considers the maximum water transfer volume between two adjacent reservoirs as a reference when scheduling reservoirs. But in reality, the maximum water transfer volume does not necessarily represent the highest energy efficiency, and it fails to consider the sufficiency and urgency of the maximum water transfer volume. It also lacks coordination and handling of water conservancy conflicts and predictions among parallel reservoirs, making it difficult to generate dynamic adjustment and optimization strategies for differentiated reservoirs. Summary of the Invention

[0004] The technical problem solved by this invention is that the existing technology only considers the maximum water transfer volume record between two adjacent reservoirs as a reference. However, the maximum water transfer volume in reality does not necessarily represent the optimal scheduling. Furthermore, it fails to consider the sufficiency and urgency of the maximum water transfer volume, lacks coordination and handling of water conservancy conflicts and predictions in parallel reservoirs, and makes it difficult to generate dynamic regulation optimization strategies for differentiated reservoirs.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an intelligent scheduling and management system for water conservancy projects, comprising a data acquisition module, a feature extraction module, a conflict measurement module, and a parallel optimization module;

[0006] The data acquisition module is used to preprocess historical scheduling data to obtain a time-series scheduling dataset;

[0007] The feature extraction module is used to aggregate and calculate the time-series scheduling dataset according to the scheduling period to obtain the water storage margin feature value and the drainage urgency feature value. The reservoir in each time period is classified according to the water storage margin feature value and the drainage urgency feature value to obtain the time-series feature vector set.

[0008] The conflict measurement module is used to construct a parallel reservoir adjacency matrix to calculate the total inflow and the path conflict coefficient of the reservoir.

[0009] The parallel optimization module is used to set scheduling classification reservoir labels based on the time-series feature vector set and path conflict coefficient, dynamically adjust the valve opening of the target reservoir according to the scheduling classification reservoir labels, and generate scheduling adjustment instructions.

[0010] Preferably, the data acquisition module is used to preprocess historical scheduling data to obtain a time-series scheduling dataset;

[0011] Historical scheduling data is collected by deploying sensors in reservoirs, rivers, pumping stations, and meteorological stations. The historical scheduling data includes hydrological data, engineering status data, and external environmental data.

[0012] The hydrological data includes the real-time water level, inflow, outflow and downstream river flow of each reservoir.

[0013] The engineering status data includes the current reservoir capacity, gate opening degree, and pump station operation status of each reservoir. The pump station operation status includes the start-up status, number of start-ups and shutdowns, instantaneous water pumping flow rate, and power consumption value.

[0014] The external environmental data includes precipitation and evaporation.

[0015] Historical scheduling data is cleaned, normalized, and aligned according to a unified time step to obtain a time-series scheduling dataset. The time-series scheduling dataset includes reservoir number, scheduling period, hydrological data, engineering status data, and external environment data. The scheduling period is one time step.

[0016] Preferably, the feature extraction module includes a temporal vector construction unit, a storage and sorting feature extraction unit, and a feature vector construction unit;

[0017] The time-series vector construction unit is used to aggregate the time-series scheduling dataset according to the scheduling period to obtain a one-dimensional time-series vector group.

[0018] Extract the time-series scheduling dataset, and aggregate the dataset according to scheduling periods, using the reservoir number as an index:

[0019] The total outflow of each reservoir during each scheduling period is calculated as the first outflow. Using the pump station number as an index, the total total water lifting flow of each pump station during each scheduling period is calculated as the first water lifting volume.

[0020] The first outflow and the first water withdrawal are concatenated with the current water level, reservoir capacity, inflow, valve opening, precipitation, and evaporation to obtain a one-dimensional time-series vector.

[0021] Preferably, the water storage and drainage feature extraction unit is used to calculate the water storage margin feature value and the drainage urgency feature value through a one-dimensional time-series vector group;

[0022] In a one-dimensional time-series vector group, for each scheduling period of each reservoir, the water storage margin characteristic value and the drainage urgency characteristic value are calculated. The processing logic for the water storage margin characteristic value and the drainage urgency characteristic value is as follows:

[0023] By comparing the current reservoir capacity with the minimum, maximum, and operational control water level capacities, the adequacy of the current water storage space is calculated to generate a water storage leeway characteristic value:

[0024] If the minimum reservoir capacity is less than the current reservoir capacity and less than the operating control water level capacity, it is in a low water level stage. The difference between the current reservoir capacity and the minimum reservoir capacity is taken as the first difference value, and the difference between the operating control water level capacity and the current capacity is taken as the second difference value. The first margin is obtained by subtracting the ratio of the second difference value to the sum of the first and second differences values ​​from 1.

[0025] If the operating control water level capacity is less than the current reservoir capacity and less than the maximum reservoir capacity, it is in a high water level stage. The difference between the maximum reservoir capacity and the current reservoir capacity is taken as the third difference value, and the difference between the maximum reservoir capacity and the operating control water level capacity is taken as the fourth difference value. The ratio of the third difference value and the fourth difference value is calculated to obtain the second margin.

[0026] The first and second margins are used as characteristic values ​​of water storage margin;

[0027] By fusing real-time water level, inflow, outflow, downstream river flow, and external environmental data using a flow stress rule, drainage urgency characteristic values ​​are obtained. The flow stress rule is as follows:

[0028] Calculate the difference between the real-time water level and the preset lower limit, and the difference between the preset upper limit and the preset lower limit. Use the ratio of the two differences as the water level urgency. Calculate the ratio of the average hourly inflow rate within a day to the historical maximum inflow rate per unit time as the inflow stress. Calculate the ratio of the difference between the downstream river flow and the current outflow rate to the downstream river flow as the downstream stress.

[0029] Using the water level urgency, inflow stress, and downstream stress as stress coupling factors, the drainage urgency characteristic value is calculated by analyzing these three factors. The expression for calculating the drainage urgency characteristic value is as follows:

[0030] ;

[0031] in, This is a characteristic value indicating the urgency of drainage. Due to the urgency of the water level, As to the degree of coercion in order to enter the warehouse, The degree of downstream stress.

[0032] Preferably, the feature vector construction unit is used to classify the state of the reservoir in each time period by the water storage margin feature value and the drainage urgency feature value to obtain a state label and correspond one-to-one with the one-dimensional time series vector group to obtain a time series feature vector set;

[0033] By comparing the values ​​of water storage leeway and drainage urgency characteristics using preset water storage and drainage thresholds, the reservoir is classified into states for each time period to obtain state labels. The processing logic for state classification is as follows:

[0034] When the water storage margin characteristic value is higher than the preset water storage threshold, and the water storage margin characteristic value is higher than the drainage urgency characteristic value, the time period is marked as the water storage peak of the reservoir.

[0035] When the drainage urgency characteristic value is higher than the preset drainage threshold and the drainage urgency characteristic value is higher than the water storage margin characteristic value, the time period is marked as the drainage valley of the reservoir.

[0036] When the absolute difference between the water storage margin characteristic value and the drainage urgency characteristic value is less than the preset minimum stress threshold, the time period is marked as the equilibrium time period of the reservoir.

[0037] The status labels are feature-encoded to obtain label codes. The water storage margin feature values, drainage urgency feature values, and label codes are then mapped to a one-dimensional time-series vector group according to the scheduling period to obtain a time-series feature vector set.

[0038] Preferably, the conflict measurement module includes a topology mapping unit, a parallel traffic statistics unit, and a conflict assessment unit;

[0039] The topology mapping unit is used to obtain the reservoir topology map and construct a parallel reservoir adjacency matrix;

[0040] Obtain a reservoir topology map, which includes reservoir number, reservoir location, inlet location, and outlet location. The reservoirs are connected by water conveyance channels. Each water conveyance channel includes an initial node, a tail node, and a key node. The initial node is the starting reservoir, the tail node is the ending reservoir, and the key node includes control valves, water distribution hubs, and cross-sections.

[0041] Iterate through the reservoir numbers, randomly select any two reservoirs as the first and second reservoirs, and construct an adjacency matrix. Matrix elements To establish a direct water transfer relationship from reservoir i to reservoir j, identify whether the first reservoir and the second reservoir are directly connected through a water transfer channel. If yes, record the adjacency relationship between the first reservoir and the second reservoir as 1 in the adjacency matrix; otherwise, record the adjacency relationship between the first reservoir and the second reservoir as 0 in the adjacency matrix.

[0042] Obtain the parameters of the water conveyance channel, calculate the hydraulic transfer coefficients of the first and second reservoirs based on the parameters, obtain the instantaneous contribution rate, save the instantaneous contribution rate as an instantaneous contribution matrix according to the adjacency relationship between the reservoirs, and use the instantaneous contribution matrix and the adjacency matrix as the parallel reservoir adjacency matrix.

[0043] Preferably, the parallel flow statistics unit is used to aggregate the total flow into the reservoir in real time through a parallel reservoir adjacency matrix;

[0044] Extract the time-series feature vector set, obtain the outflow of all reservoirs in the current scheduling period t, and obtain the outflow vector. , where n is the reservoir number;

[0045] Selecting the current reservoir j, traverse the adjacency matrix to find all upstream adjacent reservoirs. Calculate the total inflow using the outflow from all upstream adjacent reservoirs and the instantaneous contribution rate of the current reservoir j. The expression for calculating the total inflow is:

[0046] ;

[0047] in, Let k be the total inflow to reservoir j, and k be the number of all upstream adjacent reservoirs of reservoir j. The instantaneous contribution rate of the reservoir. Let be the set of reservoirs that are directly upstream of reservoir j in the adjacency matrix. This refers to the outflow from the upstream reservoir.

[0048] Preferably, the conflict assessment unit is used to quantify the path conflict coefficient of the current reservoir by using the total inflow and channel;

[0049] The maximum safe inflow rate of reservoir j in the current time period is obtained. This maximum safe inflow rate is calculated by subtracting the maximum capacity and current capacity of reservoir j. The ratio of the total inflow to the maximum safe inflow rate is calculated as the flow overload ratio. Based on this flow overload ratio, a path conflict coefficient is calculated. The expression for the path conflict coefficient is as follows:

[0050] ;

[0051] Let be the path conflict coefficient that causes the upstream reservoir i to cause the downstream reservoir j to overload during the current time period t. Let i be the instantaneous contribution rate from the i-th upstream reservoir to the current reservoir j. The outflow from upstream reservoir i during the scheduling period t. To maximize the safe flow of traffic, The percentage of traffic overload;

[0052] Calculate the path conflict coefficients for all reservoirs.

[0053] Preferably, the parallel optimization module includes a scheduling target construction unit and a scheduling instruction unit;

[0054] The scheduling target construction unit is used to construct the target reservoir based on the time-series feature vector set;

[0055] A target reservoir selection strategy is constructed using the state labels, water storage margin feature values, and drainage urgency feature values ​​of the time-series feature vector set. The processing logic of the target reservoir selection strategy is as follows:

[0056] The threshold for continuous scheduling periods is set to m. When a reservoir is marked as a drainage valley for more than m consecutive scheduling periods and the drainage urgency characteristic value is higher than the preset drainage threshold, the current reservoir is determined to be a priority drainage scheduling reservoir.

[0057] When a reservoir is marked as a peak water storage period for more than m consecutive scheduling periods, and the water storage margin characteristic value is higher than the preset water storage threshold, the current reservoir is determined to be a priority water storage scheduling reservoir.

[0058] Reservoirs that are in a period of equilibrium are used as reserve reservoirs for scheduling.

[0059] Extract the reservoirs for priority drainage scheduling, arrange them from largest to smallest according to their drainage urgency characteristic value, and match them one-to-one with the path conflict coefficient and flow overload ratio output by the conflict measurement module to obtain the first arrangement of reservoir groups;

[0060] Extract the priority reservoirs for water storage scheduling, arrange them from largest to smallest according to their water storage margin characteristic values, and match them one-to-one with the path conflict coefficient and flow overload ratio output by the conflict measurement module to obtain the second arrangement of reservoir groups;

[0061] The regulation target type of the first and second row of reservoir groups is determined to obtain the reservoir labels for scheduling classification:

[0062] Iterate through the path conflict coefficients of all reservoirs in the current time period, calculate the average of the path conflict coefficients, and use it as the first coefficient. When the urgency of drainage is higher than the water storage margin characteristic value and the path conflict coefficient in the current time period is greater than the first coefficient, set the current reservoir's scheduling classification reservoir label as a high priority drainage target reservoir.

[0063] When the water storage margin characteristic value is higher than the drainage urgency characteristic value and the path conflict coefficient is less than the first coefficient, the current reservoir's scheduling classification reservoir label is set as the target reservoir for increasing water storage.

[0064] When a reservoir is in a balanced period, its current reservoir scheduling classification reservoir label is set as the regulation target reservoir.

[0065] Preferably, the dispatching instruction unit is used to dynamically adjust the valve opening of the reservoir according to the dispatching classification reservoir label, and generate dispatching and adjustment instructions:

[0066] Calculate the difference between the current valve opening and the maximum valve opening of the high-priority drainage target reservoir, take 50% of the current difference as the basic adjustment valve opening amount, and set the upper limit of the valve opening for a single adjustment to 10%;

[0067] Calculate the difference between the current valve opening and the minimum allowable opening of the target reservoir for water storage, take 40% of this difference as the basic adjustment closure amount, and set the upper limit of valve closure for a single adjustment to 8%;

[0068] Calculate the difference between the current opening degree and the historical average opening degree of the target reservoir. If the difference is less than 5%, adjust according to the current difference; if the difference is greater than 5%, adjust only by 5%.

[0069] The beneficial effects of this invention are as follows: By constructing a dual-feature quantitative model of water storage margin and drainage urgency, a precise characterization of reservoir operation status is achieved. This method breaks through the limitations of traditional single water level thresholds, integrating multi-dimensional information such as reservoir capacity, inflow pressure, and downstream demand, making the state description more comprehensive and effective. Secondly, by constructing a parallel reservoir adjacency matrix that integrates topological structure and hydraulic transmission characteristics, and designing a path conflict coefficient algorithm, a forward assessment of potential risk characteristics of multi-reservoir linkage scheduling in complex water systems is achieved, making previously difficult-to-detect implicit conflicts explicit, providing a core basis for scheduling coordination. Finally, the reservoir state classification and conflict assessment results are systematically integrated to generate control target labels, and drive a valve dynamic adjustment strategy with adjustable differential parameters to adaptively match the timely needs of different reservoirs and the safety constraints of the entire network, realizing the refined generation of control commands.

[0070] This invention effectively solves key challenges in parallel scheduling of water conservancy projects. By using dual-feature value calculation, it elevates the scheduling basis from qualitative judgment to quantitative analysis, addressing the issues of scheduling decisions relying on human experience and lacking unified quantitative standards. Secondly, this invention solves the problem of difficulty in predicting and coordinating hydraulic conflicts when multiple reservoirs operate in parallel. The conflict measurement module is used to provide early warning of downstream overload risks, improving the overall safety of complex water network operations. Furthermore, this invention changes the traditional extensive scheduling model. Based on a dynamic mechanism labeling based on real-time status classification, the system allows for customized control strategies for each reservoir's time period, significantly improving the level of refined water resource management and utilization efficiency. Finally, by optimizing valve control strategies, it effectively prevents equipment impacts and drastic changes in water flow while ensuring rapid response, balancing scheduling efficiency and project safety. Attached Figure Description

[0071] Figure 1This is a basic flowchart of an intelligent scheduling and management system for water conservancy projects provided in one embodiment of the present invention. Detailed Implementation

[0072] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0073] Example, refer to Figure 1 A smart scheduling and management system for water conservancy projects is provided, characterized by including a data acquisition module, a feature extraction module, a conflict measurement module, and a parallel optimization module;

[0074] The data acquisition module is used to preprocess historical scheduling data to obtain a time-series scheduling dataset;

[0075] The feature extraction module is used to aggregate and calculate the time-series scheduling dataset according to the scheduling period to obtain the water storage margin feature value and the drainage urgency feature value. The reservoir status is classified for each time period by the water storage margin feature value and the drainage urgency feature value to obtain the time-series feature vector set.

[0076] The conflict measurement module is used to construct a parallel reservoir adjacency matrix to calculate the total inflow and the path conflict coefficient of the reservoir.

[0077] The parallel optimization module is used to set the scheduling classification reservoir labels based on the time-series feature vector set and path conflict coefficient, and dynamically adjust the valve opening of the target reservoir according to the scheduling classification reservoir labels to generate scheduling adjustment instructions.

[0078] This embodiment effectively solves key challenges in parallel scheduling of water conservancy projects. By using dual feature value calculation, the scheduling basis is upgraded from qualitative judgment to quantitative analysis, addressing the issues of scheduling decisions relying on human experience and lacking unified quantitative standards. Secondly, this invention solves the problem of difficulty in predicting and coordinating hydraulic conflicts when multiple reservoirs operate in parallel. The conflict measurement module is used to provide early warning of downstream overload risks, improving the overall safety of complex water network operations. Furthermore, this invention changes the traditional extensive scheduling model. Based on a dynamic mechanism labeling based on real-time status classification, the system can customize the control strategy for each reservoir for its specific time period, significantly improving the level of refined water resource management and utilization efficiency. Finally, by optimizing valve control strategies, while ensuring rapid response, it effectively prevents equipment impacts and drastic changes in water flow, balancing scheduling efficiency and project safety.

[0079] The data acquisition module is used to preprocess historical scheduling data to obtain a time-series scheduling dataset;

[0080] The acquisition unit includes:

[0081] Historical scheduling data is collected by deploying sensors in reservoirs, rivers, pumping stations, and meteorological stations. This historical scheduling data includes hydrological data, engineering status data, and external environmental data.

[0082] Hydrological data includes real-time water levels, inflows, outflows, and downstream river flows for each reservoir.

[0083] The project status data includes the current reservoir capacity, gate opening degree, and pump station operation status of each reservoir. The pump station operation status includes the start-up status, number of start-ups and shutdowns, instantaneous water pumping flow rate, and power consumption value.

[0084] External environmental data include precipitation and evaporation;

[0085] Historical scheduling data is cleaned, normalized, and aligned according to a unified time step to obtain a time-series scheduling dataset. The time-series scheduling dataset includes reservoir number, scheduling period, hydrological data, engineering status data, and external environment data. The scheduling period is one time step.

[0086] Hydrological sensors (such as water level gauges and flow meters), engineering status sensors (gate opening sensors and pump station operation monitoring devices), and meteorological sensors (rain gauges and evaporation pans) are installed in reservoirs, rivers, and pumping stations. All sensors collect data via the Industrial Internet of Things (IIoT) at 5 to 15-minute intervals and upload it to the cloud. This acquires hydrological data, engineering status data, and external environmental data as historical scheduling data, providing raw data for subsequent analysis. The collected historical scheduling data is cleaned and normalized. Cleaning includes handling missing values ​​and detecting anomalies. Missing values ​​are filled using the interpolation method, and detected anomalies are corrected based on the three-standard-deviation criterion. The cleaned historical scheduling data is then standardized using z-scores. Finally, using the scheduling period as a fixed window, a unified time step of 1 hour is set for the fixed window. The cumulative precipitation value within 1 hour is calculated according to the unified time step. Valve opening is taken as the end value of the period, and flow rate is taken as the average value of the period. Time alignment enables the system to analyze all reservoirs at the same time granularity, providing a unified timestamp for coordinated scheduling and obtaining a time-series scheduling dataset.

[0087] The feature extraction module includes a temporal vector construction unit, a storage and sorting feature extraction unit, and a feature vector construction unit;

[0088] The time-series vector construction unit is used to aggregate the time-series scheduling dataset according to the scheduling period to obtain a one-dimensional time-series vector group;

[0089] Extract the time-series scheduling dataset, and aggregate the dataset according to scheduling periods, using the reservoir number as an index:

[0090] The total outflow of each reservoir during each scheduling period is calculated as the first outflow. Using the pump station number as an index, the total total water lifting flow of each pump station during each scheduling period is calculated as the first water lifting volume.

[0091] The first outflow and the first water withdrawal are concatenated with the current water level, reservoir capacity, inflow, valve opening, precipitation, and evaporation to obtain a one-dimensional time-series vector.

[0092] In this embodiment, the reservoirs are grouped according to their numbers and scheduling periods. The first outflow and the first pumping volume are calculated for each reservoir and each scheduling period. The first outflow is calculated by summing all outflow flows for each reservoir at each scheduling time. If the data is aligned to a single recorded value that includes the total outflow flow, that value is used directly as the first outflow flow. If the group contains multiple records, the outflow flow of each record is accumulated. The instantaneous pumping flow of each pumping station in that time period is accumulated and summed to obtain the first pumping volume. This method separately calculates the pumping volume, enabling the system to distinguish outflow water for different purposes.

[0093] The water storage and drainage feature extraction unit is used to calculate the water storage margin feature value and the drainage urgency feature value through a one-dimensional time series vector group;

[0094] In a one-dimensional time-series vector group, for each scheduling period of each reservoir, the water storage margin characteristic value and the drainage urgency characteristic value are calculated. The processing logic for the water storage margin characteristic value and the drainage urgency characteristic value is as follows:

[0095] By comparing the current reservoir capacity with the minimum, maximum, and operational control water level capacities, the adequacy of the current water storage space is calculated to generate a water storage leeway characteristic value:

[0096] If the minimum reservoir capacity is less than the current reservoir capacity and less than the operating control water level capacity, it is in a low water level stage. The difference between the current reservoir capacity and the minimum reservoir capacity is taken as the first difference value, and the difference between the operating control water level capacity and the current capacity is taken as the second difference value. The first margin is obtained by subtracting the ratio of the second difference value to the sum of the first and second differences values ​​from 1.

[0097] If the operating control water level capacity is less than the current reservoir capacity and less than the maximum reservoir capacity, it is in a high water level stage. The difference between the maximum reservoir capacity and the current reservoir capacity is taken as the third difference value, and the difference between the maximum reservoir capacity and the operating control water level capacity is taken as the fourth difference value. The ratio of the third difference value and the fourth difference value is calculated to obtain the second margin.

[0098] The first and second margins are used as characteristic values ​​of water storage margin;

[0099] By fusing real-time water level, inflow, outflow, downstream river flow, and external environmental data using flow stress rules, drainage urgency characteristic values ​​are obtained. The flow stress rules are as follows:

[0100] Calculate the difference between the real-time water level and the preset lower limit, and the difference between the preset upper limit and the preset lower limit. Use the ratio of the two differences as the water level urgency. Calculate the ratio of the average hourly inflow rate within a day to the historical maximum inflow rate per unit time as the inflow stress. Calculate the ratio of the difference between the downstream river flow and the current outflow rate to the downstream river flow as the downstream stress.

[0101] Using the water level urgency, inflow stress, and downstream stress as stress coupling factors, the drainage urgency characteristic value is calculated based on these three factors. The expression for calculating the drainage urgency characteristic value is as follows:

[0102] ;

[0103] in, This is a characteristic value indicating the urgency of drainage. Due to the urgency of the water level, As to the degree of coercion in order to enter the warehouse, The degree of downstream stress.

[0104] The minimum reservoir capacity corresponding to the dead water level is determined as the minimum reservoir capacity at which the reservoir can operate normally. The operating control water level and flow rate are the upper limit of the reservoir's daily operation corresponding to the normal water storage position. The maximum reservoir capacity is the limit capacity of the reservoir corresponding to the flood or check flood level. The calculated first margin and second margin respectively represent the low water level stage and the high water level stage, and are used as water storage margin characteristic values. This embodiment is used to quantify the remaining water storage capacity at the current moment. The higher the water storage margin characteristic value, that is, the closer it is to 1, the more capable the reservoir is of storing water. The lower the water storage margin characteristic value, that is, the closer it is to 0, the more strained the water storage space is. This provides a core quantitative indicator for reservoir status classification.

[0105] Calculate the three stress coupling factors:

[0106] Water level urgency: The preset lower limit of the water level is the dead water level, and the preset upper limit of the water level is the flood level (or check flood level). Calculate the difference between the current water level and the dead water level, and calculate the difference between the flood level (or check flood level) and the dead water level. The ratio of the differences between the two is used as the water level urgency.

[0107] Inbound stress level: The ratio of the average inbound flow over the past 24 hours to the maximum inbound flow over the past 7 days is used as the inbound stress level.

[0108] Downstream stress level: Calculate the difference between the downstream river flow and the current outflow, and use the ratio of the difference to the downstream river flow as the downstream stress level;

[0109] Based on the three stress coupling factors obtained above—water level urgency, inflow stress, and downstream stress—a drainage urgency characteristic value is calculated by integrating them. The molecules reflect the dual pressures of high water levels and rapid water inflow or downstream water shortage; The denominator is used for normalization to prevent data overflow. Ensure the result is within the interval [0, 1].

[0110] This embodiment comprehensively assesses the pressure from three aspects: the reservoir's own water level, upstream inflow, and downstream water demand, and quantifies the overall urgency of the reservoir needing to increase its outflow.

[0111] The feature vector construction unit is used to classify the state of the reservoir in each time period by the water storage margin feature value and the drainage urgency feature value to obtain the state label and correspond one-to-one with the one-dimensional time series vector group to obtain the time series feature vector set;

[0112] By comparing the values ​​of water storage leeway and drainage urgency characteristics with preset water storage and drainage thresholds, the reservoir is classified into states for each time period, and a state label is obtained. The processing logic for state classification is as follows:

[0113] When the water storage margin characteristic value is higher than the preset water storage threshold, and the water storage margin characteristic value is higher than the drainage urgency characteristic value, the time period is marked as the water storage peak of the reservoir.

[0114] When the drainage urgency characteristic value is higher than the preset drainage threshold and the drainage urgency characteristic value is higher than the water storage margin characteristic value, the time period is marked as the drainage valley of the reservoir.

[0115] When the absolute difference between the water storage margin characteristic value and the drainage urgency characteristic value is less than the preset minimum stress threshold, the time period is marked as the equilibrium time period of the reservoir.

[0116] The status labels are feature-encoded to obtain label codes. The water storage margin feature values, drainage urgency feature values, and label codes are then mapped to a one-dimensional time-series vector group according to the scheduling period to obtain a time-series feature vector set.

[0117] The preset water storage threshold is 0.6, the preset drainage threshold is 0.5, and the preset minimum stress threshold is 0.2.

[0118] When the water storage margin characteristic value (S) is higher than 0.5, the reservoir has significant water storage demand or capacity. When the drainage urgency characteristic value (U) is higher than 0.5, the reservoir faces significant drainage pressure. When the absolute difference between the water storage margin characteristic value and the drainage urgency characteristic value is less than 0.2, the reservoir is in a state of equilibrium and stalemate during the current period.

[0119] Peak water storage period is used to indicate the corresponding reservoir period in which water storage should be maintained or increased; valley water storage period is used to indicate that the water level should be lowered by increasing discharge or parallel drainage; balance period is used to indicate that the reservoir should mainly operate smoothly and moderately during the current period.

[0120] This embodiment avoids meaningless frequent state switching in areas where the characteristic value is close to 0 by setting preset water storage thresholds and preset drainage thresholds. For example, S=0.1 (almost no water storage space) and U=0.2 (slight drainage pressure). Although U>S, neither has reached the preset threshold that requires action. The system can classify it as a balance period for fine-tuning instead of initiating large-scale flood discharge in the drainage valley, thus enhancing stability at the decision boundary.

[0121] The conflict measurement module includes a topology mapping unit, a parallel traffic statistics unit, and a conflict assessment unit;

[0122] The topology mapping unit is used to obtain the reservoir topology graph and construct a parallel reservoir adjacency matrix;

[0123] Obtain the reservoir topology map, which includes reservoir numbers, locations, inlet locations, and outlet locations. The reservoirs are connected by water conveyance channels. Each water conveyance channel includes an initial node, a tail node, and key nodes. The initial node is the starting reservoir, the tail node is the ending reservoir, and the key nodes include control valves, water distribution hubs, and cross-sections.

[0124] Iterate through the reservoir numbers, randomly select any two reservoirs as the first and second reservoirs, and construct an adjacency matrix. Matrix elements To establish a direct water transfer relationship from reservoir i to reservoir j, identify whether the first reservoir and the second reservoir are directly connected through a water transfer channel. If yes, record the adjacency relationship between the first reservoir and the second reservoir as 1 in the adjacency matrix; otherwise, record the adjacency relationship between the first reservoir and the second reservoir as 0 in the adjacency matrix.

[0125] Obtain the parameters of the water conveyance channel, calculate the hydraulic transfer coefficients of the first and second reservoirs based on the parameters, obtain the instantaneous contribution rate, save the instantaneous contribution rate as an instantaneous contribution matrix according to the adjacency relationship between the reservoirs, and use the instantaneous contribution matrix and the adjacency matrix as the parallel reservoir adjacency matrix.

[0126] Read the reservoir topology map to obtain the number, location, inlet, outlet, and water conveyance channel of each reservoir. The water conveyance channel includes the starting node, the tail node, and the key node. Assuming there are N reservoirs, construct an adjacency matrix M (N*N) and initialize it to all 0. Traverse all water conveyance channels. For each water conveyance channel, obtain the starting node (starting reservoir) and the tail node (ending reservoir). When the water conveyance channel is a directed channel, pointing from the starting reservoir to the ending reservoir, set the corresponding element in the adjacency matrix M to 1, indicating that there is a direct water conveyance channel from reservoir i to reservoir j.

[0127] Assume there are 4 reservoirs forming a network:

[0128] R1→R2 (direct water transfer);

[0129] R1→R3 (direct water transfer);

[0130] R2→R4 (direct water transfer);

[0131] R3→R4 (direct water transfer);

[0132] Then the adjacency matrix is The rows are labeled R1, R2, R3, and R4 from left to right, and the columns are labeled R1, R2, R3, and R4 from top to bottom. The adjacency matrix defines the relationship between upstream and downstream, serves as the basis for the parallel flow statistics unit to identify upstream adjacent reservoirs, and also determines the subsequent conflict propagation path.

[0133] Calculate the hydraulic degree transfer factor:

[0134] Obtain water conveyance channel parameters, including channel length, bottom width, side slope, roughness coefficient, bottom slope, and control rules for key nodes;

[0135] Assume the current outflow from the upstream reservoir is The flow rate reaching the downstream reservoir through the water conveyance channel is calculated using hydraulic formulas under steady flow conditions. For example, the Manning formula can be used to calculate the flow rate for open channels, while the Darcy-Weisbach formula can be used for pressurized pipelines. If there is a diversion along the route, the flow rate reaching the downstream will be reduced. The diversion may be controlled by a diversion hub, which diverts a portion of the water according to a fixed proportion or a fixed flow rate. The instantaneous contribution rate is then calculated. , among them It is the outflow from upstream reservoir i. This refers to the actual flow rate reaching the downstream reservoir. For example, if the design flow transmission efficiency of a water conveyance channel is 0.9 (i.e., 10% water loss or diversion), then the instantaneous contribution rate is taken as 0.9.

[0136] Assume that the instantaneous contribution rate of reservoir R1 to R2 is 0.9, the instantaneous contribution rate of reservoir R1 to reservoir R3 is 0.7, the instantaneous contribution rate of reservoir R2 to reservoir R4 is 0.8, and the instantaneous contribution rate of reservoir R3 to reservoir R4 is 0.6. Based on the above, store the data as a sparse matrix and save it as an instantaneous contribution matrix.

[0137] This embodiment constructs a dual matrix structure, capturing both structure and function, which helps save storage memory, improves computational efficiency, and establishes an effective topological relationship between reservoirs.

[0138] The parallel flow statistics unit is used to aggregate the total flow into the reservoir in real time through the parallel reservoir adjacency matrix;

[0139] Extract the time-series feature vector set, obtain the outflow of all reservoirs in the current scheduling period t, and obtain the outflow vector. , where n is the reservoir number;

[0140] Selecting the current reservoir j, we traverse the adjacency matrix to find all upstream adjacent reservoirs. Using the outflow from all upstream adjacent reservoirs and the instantaneous contribution rate of the current reservoir j, we calculate the total inflow. The expression for calculating the total inflow is:

[0141] ;

[0142] in, Let k be the total inflow to reservoir j, and k be the number of all upstream adjacent reservoirs of reservoir j. The instantaneous contribution rate of the reservoir. Let be the set of reservoirs that are directly upstream of reservoir j in the adjacency matrix. This refers to the outflow from the upstream reservoir.

[0143] The calculation of total inflow is essentially a linear propagation model. This represents the portion of the outflow from upstream reservoir k that actually reaches downstream reservoir j. The method solves for the superposition effect of multiple upstream sources and then calculates the total inflow based on the instantaneous contribution rate. Traditional methods only consider the water volume relationship between each pair of reservoirs, lacking a global perspective. This method treats all reservoirs in the network as a system, uniformly calculating all inflows to avoid global conflicts caused by local optimization. By using real-time outflow and dynamic contribution rate calculations, it improves the ability to calculate real-time dynamic flow, which is beneficial for intelligent flow prediction and conflict early warning.

[0144] The conflict assessment unit is used to quantify the path conflict coefficient of the current reservoir by using the total inflow and channel;

[0145] The maximum safe inflow rate of reservoir j in the current time period is obtained. The maximum safe inflow rate is calculated by subtracting the maximum capacity and the current capacity of reservoir j. The ratio of the total inflow to the maximum safe inflow rate is calculated as the flow overload ratio. Based on the flow overload ratio, the path conflict coefficient is calculated. The expression for the path conflict coefficient is as follows:

[0146] ;

[0147] Let be the path conflict coefficient that causes the upstream reservoir i to cause the downstream reservoir j to overload during the current time period t. Let i be the instantaneous contribution rate from the i-th upstream reservoir to the current reservoir j. The outflow from upstream reservoir i during the scheduling period t. To maximize the safe flow of traffic, The percentage of traffic overload;

[0148] Calculate the path conflict coefficients for all reservoirs.

[0149] In the expression for calculating the path conflict coefficient, This represents the portion of the outflow from upstream reservoir i that actually contributes to the conflict with downstream reservoir j, used to represent the tracing of the source from the overall conflict to individual responsibility. This is the amplification factor that triggers the amplification; when the amplification factor is greater than 1, this term is positive. A value greater than 0 indicates that a conflict has actually occurred, and The larger the overload, the more severe the overload, and the more the path conflict coefficient is amplified proportionally, with a higher degree of significance. ...) are used to ensure that the path conflict coefficient is non-negative;

[0150] In this embodiment, the path conflict coefficient enables the source of conflict responsibility, allowing the scheduling system to prioritize the scheduling of upstream reservoirs that contribute significantly to the conflict, rather than restricting all upstream reservoirs. This improves the fairness of scheduling efficiency. Through the dynamic calculation of the maximum safe acceptance flow and the flow overload ratio, the system can perceive the reservoir's acceptance capacity and marginal changes in real time, provide early warnings when reservoir capacity is tight, and quantify risks, significantly improving the overall safety of the water conveyance channel operation.

[0151] The parallel optimization module includes a scheduling target construction unit and a scheduling instruction unit;

[0152] The scheduling target construction unit is used to construct the target reservoir based on the time-series feature vector set;

[0153] A target reservoir selection strategy is constructed using the state labels, water storage margin feature values, and drainage urgency feature values ​​of the time-series feature vector set. The processing logic of the target reservoir selection strategy is as follows:

[0154] The threshold for continuous scheduling periods is set to m. When a reservoir is marked as a drainage valley for more than m consecutive scheduling periods and the drainage urgency characteristic value is higher than the preset drainage threshold, the current reservoir is determined to be a priority drainage scheduling reservoir.

[0155] When a reservoir is marked as a peak water storage period for more than m consecutive scheduling periods, and the water storage margin characteristic value is higher than the preset water storage threshold, the current reservoir is determined to be a priority water storage scheduling reservoir.

[0156] Reservoirs that are in a period of equilibrium are used as reserve reservoirs for scheduling.

[0157] Extract the reservoirs for priority drainage scheduling, arrange them from largest to smallest according to their drainage urgency characteristic value, and match them one-to-one with the path conflict coefficient and flow overload ratio output by the conflict measurement module to obtain the first arrangement of reservoir groups;

[0158] Extract the priority reservoirs for water storage scheduling, arrange them from largest to smallest according to their water storage margin characteristic values, and match them one-to-one with the path conflict coefficient and flow overload ratio output by the conflict measurement module to obtain the second arrangement of reservoir groups;

[0159] The regulation target type of the first and second row of reservoir groups is determined to obtain the reservoir labels for scheduling classification:

[0160] Iterate through the path conflict coefficients of all reservoirs in the current time period, calculate the average of the path conflict coefficients, and use it as the first coefficient. When the urgency of drainage is higher than the water storage margin characteristic value and the path conflict coefficient in the current time period is greater than the first coefficient, set the current reservoir's scheduling classification reservoir label as a high priority drainage target reservoir.

[0161] When the water storage margin characteristic value is higher than the drainage urgency characteristic value and the path conflict coefficient is less than the first coefficient, the current reservoir's scheduling classification reservoir label is set as the target reservoir for increasing water storage.

[0162] When a reservoir is in a balanced period, its current reservoir scheduling classification reservoir label is set as the regulation target reservoir.

[0163] By introducing a first coefficient, this method ensures that the scheduling label of each reservoir is no longer determined in isolation, but rather after weighing the factors across the entire reservoir network. This effectively guides the system toward overall optimal behavior, alleviates conflicts, and realizes a coordinated strategy for the reservoir group.

[0164] The dispatching instruction unit is used to dynamically adjust the valve opening of the reservoir according to the reservoir's dispatching classification label, and generate dispatching and adjustment instructions:

[0165] Calculate the difference between the current valve opening and the maximum valve opening of the high-priority drainage target reservoir, take 50% of the current difference as the basic adjustment valve opening amount, and set the upper limit of the valve opening for a single adjustment to 10%;

[0166] Calculate the difference between the current valve opening and the minimum allowable opening of the target reservoir for water storage, take 40% of this difference as the basic adjustment closure amount, and set the upper limit of valve closure for a single adjustment to 8%;

[0167] Calculate the difference between the current opening degree and the historical average opening degree of the target reservoir. If the difference is less than 5%, adjust according to the current difference. If the difference is greater than 5%, adjust only by 5%.

[0168] By breaking down the overall scheduling objective into small-step adjustments over multiple cycles, the changes in reservoir water level and downstream river flow become smooth scheduling tasks, avoiding sudden impacts on the ecosystem and downstream water users, improving the environmental acceptability of reservoir scheduling, and preventing out-of-control situations caused by large-scale errors.

[0169] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0170] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the protection scope of the present invention.

Claims

1. A smart scheduling and management system for water conservancy projects, characterized in that, It includes a data acquisition module, a feature extraction module, a conflict measurement module, and a parallel optimization module; The data acquisition module is used to preprocess historical scheduling data to obtain a time-series scheduling dataset; The feature extraction module is used to aggregate and calculate the time-series scheduling dataset according to the scheduling period to obtain the water storage margin feature value and the drainage urgency feature value. The reservoir in each time period is classified according to the water storage margin feature value and the drainage urgency feature value to obtain the time-series feature vector set. The conflict measurement module is used to construct a parallel reservoir adjacency matrix to calculate the total inflow and the path conflict coefficient of the reservoir. The parallel optimization module is used to set scheduling classification reservoir labels according to the time-series feature vector set and path conflict coefficient, dynamically adjust the valve opening of the target reservoir according to the scheduling classification reservoir labels, and generate scheduling adjustment instructions. The feature extraction module includes a temporal vector construction unit, a storage and sorting feature extraction unit, and a feature vector construction unit; The time-series vector construction unit is used to aggregate the time-series scheduling dataset according to the scheduling period to obtain a one-dimensional time-series vector group. Extract the time-series scheduling dataset, and aggregate the dataset according to scheduling periods, using the reservoir number as an index: The total outflow of each reservoir during each scheduling period is calculated as the first outflow. Using the pump station number as an index, the total total water lifting flow of each pump station during each scheduling period is calculated as the first water lifting volume. The first outflow and the first water withdrawal are concatenated with the current water level, reservoir capacity, inflow, valve opening, precipitation, and evaporation to obtain a one-dimensional time-series vector set; The water storage and drainage feature extraction unit is used to calculate the water storage margin feature value and the drainage urgency feature value through a one-dimensional time-series vector group. In a one-dimensional time-series vector group, for each scheduling period of each reservoir, the water storage margin characteristic value and the drainage urgency characteristic value are calculated. The processing logic for the water storage margin characteristic value and the drainage urgency characteristic value is as follows: By comparing the current reservoir capacity with the minimum, maximum, and operational control water level capacities, the adequacy of the current water storage space is calculated to generate a water storage leeway characteristic value: If the minimum reservoir capacity is less than the current reservoir capacity and less than the operating control water level capacity, it is in a low water level stage. The difference between the current reservoir capacity and the minimum reservoir capacity is taken as the first difference value, and the difference between the operating control water level capacity and the current capacity is taken as the second difference value. The first margin is obtained by subtracting the ratio of the second difference value to the sum of the first and second differences values ​​from 1. If the operating control water level capacity is less than the current reservoir capacity and less than the maximum reservoir capacity, it is in a high water level stage. The difference between the maximum reservoir capacity and the current reservoir capacity is taken as the third difference value, and the difference between the maximum reservoir capacity and the operating control water level capacity is taken as the fourth difference value. The ratio of the third difference value and the fourth difference value is calculated to obtain the second margin. The first and second margins are used as characteristic values ​​of water storage margin; By fusing real-time water level, inflow, outflow, downstream river flow, and external environmental data using a flow stress rule, drainage urgency characteristic values ​​are obtained. The flow stress rule is as follows: Calculate the difference between the real-time water level and the preset lower limit, and the difference between the preset upper limit and the preset lower limit. Use the ratio of the two differences as the water level urgency. Calculate the ratio of the average hourly inflow rate within a day to the historical maximum inflow rate per unit time as the inflow stress. Calculate the ratio of the difference between the downstream river flow and the current outflow rate to the downstream river flow as the downstream stress. Using the water level urgency, inflow stress, and downstream stress as stress coupling factors, the drainage urgency characteristic value is calculated by analyzing these three factors. The expression for calculating the drainage urgency characteristic value is as follows: ； in, This is a characteristic value indicating the urgency of drainage. Due to the urgency of the water level, As to the degree of coercion in order to enter the warehouse, The degree of downstream coercion; The conflict measurement module includes a topology mapping unit, a parallel traffic statistics unit, and a conflict assessment unit. The parallel flow statistics unit is used to aggregate the total flow into the reservoir in real time through the parallel reservoir adjacency matrix; Extract the time-series feature vector set, obtain the outflow of all reservoirs in the current scheduling period t, and obtain the outflow vector. , where n is the reservoir number; Selecting the current reservoir j, traverse the adjacency matrix to find all upstream adjacent reservoirs. Calculate the total inflow using the outflow from all upstream adjacent reservoirs and the instantaneous contribution rate of the current reservoir j. The expression for calculating the total inflow is: ； in, Let k be the total inflow to reservoir j, and k be the number of all upstream adjacent reservoirs of reservoir j. The instantaneous contribution rate of the reservoir. Let be the set of reservoirs that are directly upstream of reservoir j in the adjacency matrix. This refers to the outflow from the upstream reservoir. The conflict assessment unit is used to quantify the path conflict coefficient of the current reservoir by using the total inflow and channel; The maximum safe inflow rate of reservoir j in the current time period is obtained. This maximum safe inflow rate is calculated by subtracting the maximum capacity and current capacity of reservoir j. The ratio of the total inflow to the maximum safe inflow rate is calculated as the flow overload ratio. Based on this flow overload ratio, a path conflict coefficient is calculated. The expression for the path conflict coefficient is as follows: ； Let be the path conflict coefficient that causes the upstream reservoir i to cause the downstream reservoir j to overload during the current time period t. Let i be the instantaneous contribution rate from the i-th upstream reservoir to the current reservoir j. The outflow from upstream reservoir i during the scheduling period t. To maximize the safe flow of traffic, The percentage of traffic overload; Calculate the path conflict coefficients for all reservoirs.

2. The intelligent scheduling and management system for water conservancy projects as described in claim 1, characterized in that, The data acquisition module is used to preprocess historical scheduling data to obtain a time-series scheduling dataset; Historical scheduling data is collected by deploying sensors in reservoirs, rivers, pumping stations, and meteorological stations. The historical scheduling data includes hydrological data, engineering status data, and external environmental data. The hydrological data includes the real-time water level, inflow, outflow and downstream river flow of each reservoir. The engineering status data includes the current reservoir capacity, gate opening degree, and pump station operation status of each reservoir. The pump station operation status includes the start-up status, number of start-ups and shutdowns, instantaneous water pumping flow rate, and power consumption value. The external environmental data includes precipitation and evaporation. Historical scheduling data is cleaned, normalized, and aligned according to a unified time step to obtain a time-series scheduling dataset. The time-series scheduling dataset includes reservoir number, scheduling period, hydrological data, engineering status data, and external environment data. The scheduling period is one time step.

3. The intelligent scheduling and management system for water conservancy projects as described in claim 1, characterized in that, The feature vector construction unit is used to classify the state of the reservoir in each time period by the water storage margin feature value and the drainage urgency feature value to obtain the state label and correspond one-to-one with the one-dimensional time series vector group to obtain the time series feature vector set. By comparing the values ​​of water storage leeway and drainage urgency characteristics using preset water storage and drainage thresholds, the reservoir is classified into states for each time period to obtain state labels. The processing logic for state classification is as follows: When the water storage margin characteristic value is higher than the preset water storage threshold, and the water storage margin characteristic value is higher than the drainage urgency characteristic value, the time period is marked as the water storage peak of the reservoir. When the drainage urgency characteristic value is higher than the preset drainage threshold and the drainage urgency characteristic value is higher than the water storage margin characteristic value, the time period is marked as the drainage valley of the reservoir. When the absolute difference between the water storage margin characteristic value and the drainage urgency characteristic value is less than the preset minimum stress threshold, the time period is marked as the equilibrium time period of the reservoir. The status labels are feature-encoded to obtain label codes. The water storage margin feature values, drainage urgency feature values, and label codes are then mapped to a one-dimensional time-series vector group according to the scheduling period to obtain a time-series feature vector set.

4. The intelligent scheduling and management system for water conservancy projects as described in claim 1, characterized in that, The topology mapping unit is used to obtain the reservoir topology map and construct a parallel reservoir adjacency matrix; Obtain a reservoir topology map, which includes reservoir number, reservoir location, inlet location, and outlet location. The reservoirs are connected by water conveyance channels. Each water conveyance channel includes an initial node, a tail node, and a key node. The initial node is the starting reservoir, the tail node is the ending reservoir, and the key node includes control valves, water distribution hubs, and cross-sections. Iterate through the reservoir numbers, randomly select any two reservoirs as the first and second reservoirs, and construct an adjacency matrix. Matrix elements To establish a direct water transfer relationship from reservoir i to reservoir j, identify whether the first reservoir and the second reservoir are directly connected through a water transfer channel. If yes, record the adjacency relationship between the first reservoir and the second reservoir as 1 in the adjacency matrix; otherwise, record the adjacency relationship between the first reservoir and the second reservoir as 0 in the adjacency matrix. Obtain the parameters of the water conveyance channel, calculate the hydraulic transfer coefficients of the first and second reservoirs based on the parameters, obtain the instantaneous contribution rate, save the instantaneous contribution rate as an instantaneous contribution matrix according to the adjacency relationship between the reservoirs, and use the instantaneous contribution matrix and the adjacency matrix as the parallel reservoir adjacency matrix.

5. The intelligent scheduling and management system for water conservancy projects as described in claim 1, characterized in that, The parallel optimization module includes a scheduling target construction unit and a scheduling instruction unit; The scheduling target construction unit is used to construct the target reservoir based on the time-series feature vector set; A target reservoir selection strategy is constructed using the state labels, water storage margin feature values, and drainage urgency feature values ​​of the time-series feature vector set. The processing logic of the target reservoir selection strategy is as follows: The threshold for continuous scheduling periods is set to m. When a reservoir is marked as a drainage valley for more than m consecutive scheduling periods and the drainage urgency characteristic value is higher than the preset drainage threshold, the current reservoir is determined to be a priority drainage scheduling reservoir. When a reservoir is marked as a peak water storage period for more than m consecutive scheduling periods, and the water storage margin characteristic value is higher than the preset water storage threshold, the current reservoir is determined to be a priority water storage scheduling reservoir. Reservoirs that are in a period of equilibrium are used as reserve reservoirs for scheduling. Extract the priority drainage scheduling reservoirs, arrange them from largest to smallest according to the drainage urgency characteristic value, and match them one-to-one with the path conflict coefficient and flow overload ratio output by the conflict measurement module to obtain the first arrangement of reservoir groups; Extract the priority reservoirs for water storage scheduling, arrange them from largest to smallest according to their water storage margin characteristic values, and match them one-to-one with the path conflict coefficient and flow overload ratio output by the conflict measurement module to obtain the second arrangement of reservoir groups; The regulation target type of the first and second row of reservoir groups is determined to obtain the reservoir labels for scheduling classification: Iterate through the path conflict coefficients of all reservoirs in the current time period, calculate the average of the path conflict coefficients, and use it as the first coefficient. When the urgency of drainage is higher than the water storage margin characteristic value and the path conflict coefficient in the current time period is greater than the first coefficient, set the current reservoir's scheduling classification reservoir label as a high priority drainage target reservoir. When the water storage margin characteristic value is higher than the drainage urgency characteristic value and the path conflict coefficient is less than the first coefficient, the current reservoir's scheduling classification reservoir label is set as the target reservoir for increasing water storage. When a reservoir is in a balanced period, its current reservoir scheduling classification reservoir label is set as the regulation target reservoir.

6. The intelligent scheduling and management system for water conservancy projects as described in claim 5, characterized in that, The dispatching instruction unit is used to dynamically adjust the valve opening of the reservoir according to the reservoir's dispatching classification label, and generate dispatching and adjustment instructions: Calculate the difference between the current valve opening and the maximum valve opening of the high-priority drainage target reservoir, take 50% of the current difference as the basic adjustment valve opening amount, and set the upper limit of the valve opening for a single adjustment to 10%; Calculate the difference between the current valve opening and the minimum allowable opening of the target reservoir for water storage, take 40% of this difference as the basic adjustment closure amount, and set the upper limit of valve closure for a single adjustment to 8%; Calculate the difference between the current opening degree and the historical average opening degree of the target reservoir. If the difference is less than 5%, adjust according to the current difference. If the difference is greater than 5%, adjust only by 5%.

Images