A hydroelectric power plant alarm system, method and computer readable storage medium

Abstract

The present application relates to reservoir operation monitoring and risk early warning technical field, specifically to a kind of hydropower plant alarm system and method.The method first divides target reservoir into hydrology-hydrodynamic grid unit, and constructs the multidimensional parameter of each grid unit;Potential abnormal unit is identified based on the multidimensional parameter of the grid unit, and the abnormal unit is tracked along the runoff direction or reverse backwater direction, to construct a set of hydrodynamic evolution path;The reverse hydrodynamic chain index, effective storage correction coefficient and water level resonance index of each path in the set of hydrodynamic evolution path are calculated, and according to the calculation result, hierarchical early warning is carried out, to form a set of key risk units;Determine whether grid unit in the set of key risk units occurs water level pulse event and reverse flow event, determine whether reverse hydrodynamic event occurs, generate protection strategy corresponding to trigger early warning instruction.The present application improves the accuracy of risk early warning and the efficiency of dispatching response.

Description

Technical Field

[0001] This invention relates to the field of reservoir operation monitoring and risk early warning technology, specifically to a hydropower plant alarm system, method, and computer-readable storage medium. Background Technology

[0002] In the daily operation of hydropower plants, reservoirs not only undertake multiple tasks such as flood control, power generation, and water supply, but their operation and scheduling must also take into account reservoir capacity regulation, stable unit operation, and downstream flood safety. However, under the influence of complex hydrological conditions and sudden extreme weather events, the hydrodynamic distribution within the reservoir area often exhibits nonlinear evolution characteristics, such as local reverse flow, sudden changes in reservoir pressure, and resonance of reservoir bank water levels. These abnormal hydrodynamic events are often sudden and insidious; if they are not identified and warned of in a timely manner, they can easily lead to unstable unit operation, damage to reservoir bank structures, and even imbalance in reservoir area scheduling, thus bringing significant operational risks.

[0003] Although hydropower plants typically regulate the inflow and outflow of water into and out of the reservoir through valves, the control effect of these valves is mainly concentrated at the unit inlet or spillway, making it difficult to cover the hydrodynamic changes across the entire reservoir area. When valves are closed or partially open, backwater backflow or surge reflection effects may still occur. In areas where tributaries converge or during joint scheduling of multiple reservoirs, if the scheduling is not coordinated, it may even induce tributary water to be backflowed by the mainstream or even cause reverse backflow, resulting in a reversal of flow velocity in the reservoir area or downstream river section, forming a countercurrent.

[0004] Backflow risk is a typical anomaly in hydropower plant operation. Backflow not only weakens the effective storage capacity and regulation capacity of the reservoir area, but may also impact the water intake flow of the generating units, inducing vortices, cavitation, and even a decrease in unit efficiency. At the same time, once backflow occurs in the backwater section of the reservoir area, it can easily cause siltation, reservoir bank erosion, and damage to bank protection facilities, further increasing the scheduling risk.

[0005] In existing technologies, hydrodynamic monitoring methods in hydropower plants mainly focus on observing single parameters such as water level and flow rate, lacking systematic tracking of the hydrodynamic evolution process within the reservoir area and risk path analysis. This makes it difficult to formulate real-time early warning and dynamic protection measures against backflow risks. Consequently, when facing complex hydrodynamic evolution, scheduling strategies are often lagging behind and cannot effectively mitigate backflow risks. Summary of the Invention

[0006] The present invention aims to at least partially solve one of the technical problems in the related art.

[0007] Therefore, the first objective of this invention is to provide an alarm method for a hydropower plant, characterized by the following steps:

[0008] S1 divides the target reservoir area into hydrological-hydrodynamic grid units, and constructs the average flow velocity vector, slope stability coefficient, flow velocity reverse ratio and water level change rate of each grid unit as multidimensional parameters of the grid unit; S2, identify potential anomalous units based on the multidimensional parameters of the grid unit, and track the anomalous units along the runoff direction or the reverse backflow direction to construct a set of hydrodynamic evolution paths; S3, calculate the reverse hydrodynamic chain index, effective reservoir capacity correction coefficient and water level resonance index for each path in the set of hydrodynamic evolution paths, generate graded early warning instructions based on the calculation results and mark the alarm risk water body segments at the corresponding path endpoints to form a set of key risk units; S4. For each grid cell in the set of key risk units, determine whether a water level pulse event and a reverse flow event have occurred. When the conditions of the two types of events are met simultaneously at the same time in the same path, it is determined to be an actual reverse hydrodynamic event, and a protection strategy corresponding to the triggering warning command is generated.

[0009] In one embodiment of the present invention, S1 further includes: S11: Obtain the digital elevation model, water depth measurement data and dam structure monitoring data of the target reservoir area, and based on the digital elevation model, divide the target reservoir area into hydrological-hydrodynamic grid units using the planning grid division method. S12, construct a coordinate system based on the divided grid cells, calculate the boundary flow of the grid cells based on the instantaneous velocity distribution of the grid cells in the dam area structure monitoring data, use the boundary flow as a constraint, and obtain the average velocity vector of the grid cells by constructing a set of flow conservation equations; S13, based on the Mohr-Coulomb strength criterion, calculates the shear stress and effective normal stress using the shear strength, pore water pressure, and slope at monitoring points on the reservoir bank. Then, calculates the ratio of shear strength to actual shear stress to obtain the slope stability coefficient for the i-th grid element. S14. Determine the flow direction of the grid cells based on the digital elevation model, calculate the average velocity vector of the grid cells in that flow direction as the velocity vector along the mainstream direction, and statistically analyze the reverse flow portion of the grid cells in the time series to obtain the reverse flow ratio. S15: Collect water level time series data in grid cells and use the moving average calculation formula to obtain the water level change rate.

[0010] In one embodiment of the present invention, S12 further includes: Source and sink terms are constructed using the average water depth, x-direction average velocity component, and y-direction average velocity component of hydrological-hydrodynamic grid cells from water exploration measurement data. The instantaneous velocity distribution of grid cells in dam area structure monitoring data is integrated to obtain the boundary flow of the grid cell. The exchange flow between each grid cell and its adjacent grid cells is calculated using the average water depth, average velocity, and common side length between the grid cells. A continuous equation for the exchange flow is constructed using the boundary flow as a constraint. The water level increment in the grid cell is substituted into the continuous equation, and the source and sink terms are used for error correction. A set of flow conservation equations is constructed, and the average velocity vector of the grid cell is obtained by inversion.

[0011] In one embodiment of the present invention, S2 further includes: S21. When the water level change rate exceeds the change threshold, when the flow velocity reverse ratio exceeds the ratio threshold, or when the slope stability coefficient is lower than the stability threshold, the grid cell is marked as an abnormal cell. S22, for each anomalous unit, trace it along the runoff direction or the reverse backwater direction to the tail of the reservoir or the downstream river section to obtain a set of hydrodynamic evolution paths.

[0012] In one embodiment of the present invention, S3 further includes: S31, For the water segment of the p-th path in the set of hydrodynamic evolution paths, number each network unit according to the path order; S32, based on the velocity vector of each grid cell along the mainstream direction, along the hydrodynamic evolution path, accumulate the reverse flow intensity of each grid cell in the mainstream direction. Based on the safe velocity reference value, the average velocity of the m-th grid cell in the p-th path, and the total number of grid cells in the p-th path, calculate and obtain the reverse hydrodynamic chain index. When the reverse hydrodynamic chain index is higher than the first hydrodynamic chain index threshold but lower than the first hydrodynamic chain index threshold, the grid cell has the risk of causing local water inrush in the reservoir area, triggering the first reverse flow risk warning instruction. When the reverse hydrodynamic chain index is higher than the second hydrodynamic chain index threshold, the grid cell has the risk of causing hydrodynamic instability or structural risk, triggering the second reverse flow risk warning instruction. S33: Obtain the effective reservoir capacity correction coefficient for the current water level, normal storage level and grid unit volume of each grid unit on the hydrodynamic evolution path. When the effective reservoir capacity correction coefficient is higher than the first correction coefficient threshold and lower than the second correction coefficient threshold, the grid unit has the risk of increased reservoir pressure and triggers the first reservoir pressure risk warning instruction. When the effective reservoir capacity correction coefficient is higher than the second correction coefficient threshold, the grid unit has the risk of reservoir safety or scheduling and triggers the second reservoir pressure risk warning instruction. S34, record the water level time series of each grid unit on each hydrodynamic evolution path, calculate the standard deviation of water level changes, and obtain the water level resonance index. When the water level resonance index is greater than the first water level resonance index threshold and less than the second water level resonance index threshold, the grid unit has a local resonance trend risk, triggering the first resonance risk alarm command. When the water level resonance index is greater than the second water level resonance index threshold, there is a hydrodynamic resonance that causes facility safety hazards, triggering the second resonance risk alarm command. S35 marks the water body segments corresponding to the path endpoints of the first countercurrent risk warning instruction, the second countercurrent risk warning instruction, the first reservoir pressure risk warning instruction, the second reservoir pressure risk warning instruction, the first resonance risk alarm instruction, and the second resonance risk alarm instruction as alarm risk water body segments, forming a set of key risk units.

[0013] In one embodiment of the present invention, the method for determining whether a water level pulse event has occurred in step S4 is as follows: For each grid cell in the set of key risk units, extract its water level time series. ,in, f =1,2,…,F represents the sampling sequence number, and the time interval is... ; And based on the water level time series Calculate and obtain the peak value of water level change ;

[0014] Find the sampling point where the water level begins to rise. 1. Locate the water level that has reached its peak value. sampling points And find sampling points where the water level has dropped to a stable level. Calculate the duration of water level fluctuations :

[0015] And calculate the slope of the water level rise. and the slope of water level drop :

[0016]

[0017] When the water level changes peak And the duration of water level fluctuations And the slope of the water level rise or the slope of the water level drop If this occurs, it is determined to be a water level pulse event.

[0018] In one embodiment of the present invention, the method for determining whether a reverse flow event has occurred in step S4 is as follows: For each grid cell in the set of key risk units, extract its velocity time series. Where f=1,2,…,F are the sampling sequence numbers, and the time interval is… ; And based on the flow velocity time series Calculate and obtain the reverse peak flow velocity :

[0019] Among them, going with the current is positive, and going against the current is negative; Find the flow rate start Reverse sampling points Sampling points that reach the reverse peak And find sampling points where the flow rate recovers downstream or decays to a stable level. Calculate the duration of reverse flow :

[0020] And calculate the reverse acceleration. and attenuation slope :

[0021]

[0022] When the flow velocity reverses its peak value And the duration of reverse flow And reverse acceleration or attenuation slope If so, it is determined to be a reverse flow event.

[0023] In one embodiment of the present invention, the method for generating the protection strategy corresponding to the triggering warning command in step S4 is as follows: Upon receiving the first backflow risk warning instruction, a first backflow protection strategy is generated, including: increasing valve opening by 5%–10% and simultaneously reducing unit intake flow by 5%–10%; maintaining flood discharge flow at 80%–90% of the design value; In response to the second backflow risk warning, the first backflow protection strategy is generated, including: increasing valve opening by 15%–25%; rapidly reducing the unit intake flow by 15%–20%; reinforcing the reservoir bank and activating the baffles; and controlling the flood discharge flow at 70%–80%. Upon receiving the first reservoir capacity pressure risk warning instruction, a first reservoir capacity pressure protection strategy is generated, including: increasing valve opening by 5%–10%, increasing unit outflow by 5%–10%, and reducing reservoir capacity by 5%–8%; Upon receiving the warning instruction regarding the pressure risk of the second reservoir, a pressure protection strategy for the second reservoir is generated, including: increasing valve opening by 20%–30%; adjusting the unit's outflow rate by 15%–25% and shutting down 30% of the units; and reducing the reservoir capacity by 10%–15%. Upon receiving the first resonance risk warning command, a first resonance protection strategy is generated, including: slowly adjusting the valve opening by ±5%–10%; steadily regulating the unit flow rate by ±5%–10%; and maintaining the flood discharge flow rate at 85%–90%. In response to the warning instruction for the second resonance risk, a second resonance protection strategy is generated, including: quickly adjusting the valve opening by 15%–20%; significantly adjusting the unit flow rate by 20%–30%; and controlling the flood discharge flow rate at 60%–70%.

[0024] To achieve the above objectives, a second aspect of the present invention provides an alarm system for a hydropower plant, characterized in that it includes: The data acquisition module is used to acquire digital elevation models, water depth measurement data, and dam structure monitoring data of the target reservoir area, and to receive hydrodynamic data collected by flow velocity sensors, water level gauges, and ADCP monitoring equipment. The grid generation module is used to establish a regular grid based on the digital elevation model, divide the target reservoir area into several hydrological-hydrodynamic grid units, and construct the average flow velocity vector, slope stability coefficient, water level change rate, and water level change rate of the i-th grid unit. An anomaly identification module is used to identify potential abnormal hydrodynamic units within the grid cell based on the average flow velocity vector, slope stability coefficient, water level change rate, and water level change rate of the i-th grid cell, and to form a set of abnormal units. The path tracing module is used to trace the set of abnormal units along the runoff direction or the reverse backwater direction to the tail end of the reservoir or the downstream river section, and construct a set of hydrodynamic evolution paths. The risk assessment module is used to calculate and assess the reverse hydrodynamic chain index, effective reservoir capacity correction coefficient, and water level resonance index in the path set, and generate corresponding first countercurrent risk warning instructions, second countercurrent risk warning instructions, first reservoir capacity pressure risk warning instructions, second reservoir capacity pressure risk warning instructions, first resonance risk alarm instructions, and second resonance risk alarm instructions based on the assessment results. The risk marking module is used to mark the water segment at the end of the path that triggers a warning or alarm command as an alarm risk water segment, forming a set of key risk units; The event determination module is used to extract the peak value of water level change, the duration of water level fluctuation, the slope of water level rise and the slope of water level fall, the peak value of reverse flow velocity, the duration of reverse flow, the reverse acceleration and the attenuation slope for each grid cell in the set of key risk units, to determine whether a water level pulse event and a reverse flow event have occurred, and to identify the actual reverse hydrodynamic event. The protection strategy generation module is used to generate corresponding reverse flow protection strategies, reservoir pressure protection strategies, or resonance protection strategies based on the triggered risk warning instructions for the actual reverse hydrodynamic events, and output them to the hydropower plant dispatching system.

[0025] To achieve the above objectives, a third aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect.

[0026] The methods, systems, and storage media of this invention enable full-chain tracing of hydrodynamic risks in hydropower plant reservoir areas and accurate identification of complex events, thereby improving the accuracy of risk warnings and the efficiency of dispatch response.

[0027] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0028] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart of a hydropower plant alarm method according to an embodiment of the present invention; Figure 2 This is a structural diagram of a hydropower plant alarm system according to an embodiment of the present invention. Detailed Implementation

[0029] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0030] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0031] The following description, with reference to the accompanying drawings, describes an alarm method and system for a hydropower plant according to an embodiment of the present invention.

[0032] Example 1 Please see Figure 1 This invention provides an alarm method for a hydropower plant, comprising the following steps: Using a regular network partitioning method, several target reservoir areas are divided into several hydrological-hydrodynamic grid units. Within each grid unit... Within, construct the average velocity vector of the i-th grid cell. Slope stability coefficient Flow velocity reverse ratio and water level change rate Identify and mark potential anomalous hydrodynamic units. For each potential anomalous hydrodynamic unit, trace it along the runoff direction or the reverse backwater direction to the tail of the reservoir or the downstream river section to obtain a set of hydrodynamic evolution paths. For each water segment in the set of hydrodynamic evolution paths, construct an inverse hydrodynamic chain index. Effective storage capacity correction coefficient Resonance index with water level The corresponding water body segments corresponding to the path endpoints of the first countercurrent risk warning instruction, the second countercurrent risk warning instruction, the first reservoir pressure risk warning instruction, the second reservoir pressure risk warning instruction, the first resonance risk alarm instruction, and the second resonance risk alarm instruction are marked as alarm risk water body segments, forming a set of key risk units; For each grid cell in the set of key risk units, extract the peak value of water level change. Duration of water level fluctuations Water level rise slope and the slope of water level drop To determine whether a water level pulse event has occurred; and to extract the reverse peak flow velocity. Duration of reverse flow Reverse acceleration and attenuation slope To determine whether a reverse flow event has occurred; When the conditions for a water level pulse event and a reverse flow event are met simultaneously at the same time along the same path, it is determined to be an actual reverse hydrodynamic event. The same path of the grid cells is extracted, and the corresponding risk warning instruction is generated to produce the corresponding protection strategy.

[0033] In this embodiment, existing technologies only focus on single parameters such as water level and flow rate, failing to capture the concealment and suddenness of nonlinear hydrodynamic evolution in the reservoir area (such as local reverse flow and water level resonance). This method divides the "entire reservoir area" into precisely traceable "micro-units" by dividing it into hydrological-hydrodynamic grid units using a regular network, and simultaneously monitors multi-dimensional parameters such as "average flow velocity vector, slope stability coefficient, and water level change rate." This not only locates the initial position of "potentially abnormal hydrodynamic units" but also fully reconstructs the hydrodynamic evolution path (such as the reverse propagation process of tributary recharge) by "tracing along the runoff / backflow direction to the reservoir tail / downstream." This completely solves the pain point of "knowing only the anomaly but not the source and direction of diffusion," achieving an upgrade from "single-point monitoring" to "full-chain tracing," and significantly improving the accuracy and timeliness of risk identification.

[0034] Current technologies lack a systematic classification and assessment of hydrodynamic risks, making it difficult to distinguish the severity of different risks such as "reverse flow," "sudden changes in reservoir pressure," and "water level resonance." This method constructs three core assessment indices (reverse hydrodynamic chain index, effective reservoir capacity correction coefficient, and water level resonance index) to specifically address three key risk categories and generate "first / second" graded early warning instructions. In response to the "backflow risk", the intensity and impact range of backflow are quantified by the reverse hydrodynamic chain index, and the early warning instructions can be directly linked to the risk of the unit's water intake flow pattern (such as vortex, cavitation); To address the "risk of reservoir pressure," an effective reservoir capacity correction coefficient is used to reflect the degree to which the backflow weakens the "effective reservoir capacity and storage capacity," thus avoiding misjudgment of available reservoir capacity during scheduling. To address the "water level resonance risk," a water level resonance index is used to capture sudden changes in reservoir bank water levels, providing early warnings of bank erosion and revetment damage. This precise "risk-index-early warning" correspondence allows dispatchers to quickly determine the type and level of risk, avoiding resource waste or insufficient warnings caused by a "one-size-fits-all" approach. A single "water level pulse" or "reverse flow" in the reservoir area may be triggered by normal dispatching (such as valve regulation), and misjudging it as a risk event could lead to over-dispatching; conversely, when both are superimposed (such as sudden heavy rainfall + tributary backflow), they can easily form a serious reverse hydrodynamic event, which may be missed due to "single parameter meeting the standard." This method uses dual event judgment criteria (water level pulse events: peak water level change, fluctuation duration, etc.; reverse flow events: reverse flow velocity peak, reverse acceleration, etc.) to clarify that only when both are met at the same time can it be judged as an "actual reverse hydrodynamic event"—which not only eliminates the interference of misjudgment due to "single parameter anomaly" but also accurately identifies "composite risk events," avoiding "dispatch lag" or "overprotection" caused by ambiguous judgment, and ensuring the accuracy of risk response.

[0035] Existing technologies mostly remain at the "early warning" level, lacking a direct link to "protection strategies," resulting in scheduling strategies lagging behind risk evolution. This method, after "determining the actual reverse hydrodynamic event," directly "extracts the grid cells corresponding to the risk early warning command and generates targeted protection strategies"—for example: If the warning instruction points to "downstream backflow caused by tributary backflow", the protection strategy can accurately locate the grid unit corresponding to the backflowing tributary and guide the adjustment of the tributary inflow flow or the main discharge rhythm. If the warning command is associated with "water level pulse near the unit + reverse flow", the strategy can be directly directed to adjust the unit's inlet valve to avoid the impact of vortices on the unit's efficiency.

[0036] Reservoirs need to fulfill multiple tasks, including flood control, power generation, and water supply. Existing technologies, due to insufficient risk identification, often lead to contradictions such as "ignoring reverse flow to ensure power generation, resulting in reduced reservoir capacity" and "causing downstream flooding safety issues due to flood discharge for flood control purposes." This method accurately identifies "actual reverse hydrodynamic events" and their impact range: On the power generation side, the impact of countercurrent on the unit can be avoided in advance, ensuring stable operation and power generation efficiency of the unit; On the flood control side, the water level resonance index is used to warn of reservoir bank pressure and prevent reservoir bank damage from affecting flood control capacity; On the water supply side, available reservoir capacity is accurately calculated using an effective reservoir capacity correction coefficient to avoid water supply imbalances caused by backflow. Ultimately, this achieves coordinated optimization of flood control, power generation, and water supply tasks under the premise of "controllable risk," reducing the overall benefit losses caused by prioritizing a single task.

[0037] Example 2 This embodiment is an explanation of Embodiment 1. Specifically, it involves acquiring a high-resolution digital elevation model (DEM), water depth measurement data, and dam area structure monitoring data for the target reservoir area. Based on DEM data, a regular grid partitioning method was used to divide the target reservoir area into several hydrological-hydrodynamic grid units, denoted as: ; In one implementation method, the side length of the grid cell is set to 20-50 meters, and n is the total number of grid cells. In each grid cell Within, construct the average velocity vector of the i-th grid cell. and slope stability coefficient ; The average velocity vector of the i-th grid cell The method of obtaining it is: Discretized grid cells in the reservoir area Within, satisfying the two-dimensional shallow water flow continuity equation, the source and sink terms are obtained:

[0038] in, Indicates the average water depth. and These represent the average velocity components in the x and y directions, respectively. In fluid dynamics equations, the left side describes the flow of water itself (horizontal migration) and changes in volume (water level rise and fall), while the right side... It is specifically used to describe the increase or decrease in water volume caused by non-horizontal flow. This indicates that water has entered this grid cell out of thin air (relative to the two-dimensional horizontal plane). This indicates that water has disappeared from that grid cell; source / sink item This term characterizes vertical water exchange within a grid cell, excluding horizontal flow, including rainfall recharge, surface evaporation, groundwater seepage, and artificial water intake / extraction fluxes. By introducing this term into the continuity equation, errors in water level changes caused by vertical water exchange can be corrected, thus accurately retrieving the true horizontal velocity vector caused solely by hydrodynamic transport. ; Flow monitoring equipment, including an acoustic Doppler current profiler (ADCP) and an underwater current meter, was deployed at the boundary section of the reservoir area to obtain boundary flow data. :

[0039] in, This represents the total cross-sectional area of ​​the boundary section. dA represents the instantaneous velocity distribution on the cross section; dA represents the cross section being divided by countless small area elements, each with an instantaneous velocity. ; There is an exchange flow between each grid cell i and its neighboring grid cell j:

[0040] in, It is the average water depth between units. and It is the average flow velocity between grid cell i and its neighboring grid cell j. It is the boundary normal of grid cell i and its neighboring grid cell j. It is the common side length of grid cell i and its neighboring grid cell j; The reservoir boundary, including the inlet or outlet, has a boundary flow rate. The monitored boundary flow As a boundary condition participating in constraints:

[0041] in, This represents the collection of all small water pipes at the boundary of the reservoir area; Water level increment within the water grid cell Substituting into the continuity equation, we construct a system of flow conservation equations and inversely obtain the average velocity vector of the i-th grid cell. :

[0042] in, and These are the exchange flows between adjacent units in the x and y directions, respectively; the conservation condition ensures that the inflow and outflow of each unit are consistent with the total monitored flow; dx is the average flow distribution obtained by integrating over the entire side length after subdividing the unit boundary into many small segments in the x direction; dy: Similarly, in the y direction, the unit boundary is subdivided into many small segments, and the length of each small segment is dy, which is integrated over the entire side length. Based on the average velocity vector of the i-th grid cell Projecting onto the mainstream direction, the flow direction determined by the DEM is used to obtain the velocity vector along the mainstream direction. :

[0043] Wherein, vector dot product / vector magnitude; This represents the mainstream direction vector of the unit. This represents the length (magnitude) of the mainstream direction vector. Vector projection yields the velocity vector along the main flow direction; In the time series, the negative magnitude of the flow velocity along the mainstream direction, i.e. the reverse flow portion, is statistically analyzed to calculate the flow velocity reverse ratio. :

[0044] in, , indicating downstream, , indicating the opposite current; The negative sign is used to focus only on the reverse flow portion. The absolute value represents the total magnitude of the current flow rate.

[0045] Slope stability coefficient of the i-th grid cell The method of obtaining it is: Based on the Mohr-Coulomb strength criterion, shear stress was calculated using shear strength, pore water pressure, and slope at monitoring points along the reservoir bank. and effective normal stress The ratio of shear strength to actual shear stress is calculated to obtain the slope stability coefficient of the i-th grid element. :

[0046] in, Cohesion, derived from monitored soil samples. For effective normal stress, It is the internal friction angle. Shear stress; shear stress Where H represents the slope height, Indicates the unit weight of the soil. Indicates the angle of inclination; ,in, This indicates pore water pressure.

[0047] In each grid cell Inside, water level time series were collected. And calculate the rate of change of water level. ; Water level change rate The formula for calculating the moving average is:

[0048] in, This represents the water level value of the i-th grid cell monitored at time i (sampling time). Indicates the first The water level value of the i-th grid cell monitored at time t. This represents the time interval between two adjacent samples; N represents the length of the sliding window. Mesh cells will be used when one of the following conditions is met. Marked as a potentially anomalous hydrodynamic unit: Water level change rate Exceeding the change threshold Hthr; As one implementation method, in the prior art, the reference values ​​for the variation threshold Hthr are: small reservoirs / slow-moving reservoirs: 0.05–0.1 m / h; medium-sized reservoirs: 0.1–0.3 m / h; large reservoirs / flood season reservoirs: 0.3–0.5 m / h. Flow rate reverse ratio Exceeding the ratio threshold Rthr; slope stability coefficient Below the stability threshold Sthr; Stable: >1.5; Potential landslide risk: 1.2–1.5; High risk / unstable: <1.2; All potentially anomalous hydrodynamic units are denoted as the first anomalous set; and for each hydrological-hydrodynamic grid unit within the first anomalous set, the hydrodynamic evolution path is traced along the runoff direction or the reverse backwater direction to the reservoir tail or downstream river segment, thus obtaining a set of hydrodynamic evolution paths: .

[0049] In this embodiment, traditional methods for calculating reservoir flow often employ a "generalized" approach, making it difficult to capture microscopic hydrodynamic phenomena such as local backflow and backwater. This method achieves refined calculation of water flow by using regular grid division and inversion of the flow conservation equations: the reservoir area is decomposed into appropriately sized micro-units with regular grid divisions of 20-50 meters—avoiding the loss of local flow details due to overly coarse grids (such as small-scale backflow in tributary recharge) while avoiding computational redundancy due to overly fine grids, thus balancing "computational accuracy" and "efficiency"; through "flow monitoring at boundary cross-sections using ADCP / underwater velocity meters," "flow calculation through exchange between grid units," and "flow constraints at reservoir boundaries," a flow conservation equation set is constructed to invert the average velocity vector of each grid unit. This calculation method not only satisfies the physical laws of the two-dimensional shallow water flow continuity equation, but also ensures that the velocity calculation results are consistent with the actual water flow through the dual constraints of "boundary measured flow rate + inter-unit exchange flow rate", solving the problems of "susceptibility to interference and large fluctuations in accuracy" in traditional single equation inversion. By determining the mainstream direction through DEM, the average velocity vector is projected onto the mainstream direction, and then the "velocity reverse ratio" is statistically analyzed to accurately quantify the intensity and proportion of countercurrent, avoiding the misjudgment of the degree of countercurrent (such as the distinction between weak countercurrent and strong countercurrent) caused by the traditional method of "only judging the positive and negative of the velocity".

[0050] Traditional anomaly identification often relies on a single water level change rate, which can easily misjudge water level fluctuations caused by normal operations as anomalies, or miss the potential risk of "normal flow velocity but slope instability". By setting differentiated thresholds (0.05-0.5 m / h) based on reservoir size (small / medium / large) and operational phase (normal / flood season), misjudgments caused by a "one-size-fits-all" threshold are avoided. For example, a low threshold (0.05-0.1 m / h) can be used for small, gently flowing reservoirs to promptly capture subtle abnormal water level fluctuations; a high threshold (0.3-0.5 m / h) can be used for large reservoirs during flood season to avoid misjudging rapid water level changes caused by normal flood discharge. Furthermore, by setting judgment criteria directly targeting "backflow risk", the hidden anomalies of "normal water level but localized backflow" can be accurately identified, compensating for the shortcoming of traditional water level monitoring in failing to capture backflow. Based on the Mohr-Coulomb criterion, the coefficients are divided into three levels: "stable (>1.5), potential risk (1.2-1.5), and high risk (<1.2)". This not only quantifies the safety status of the reservoir bank structure but also identifies in advance the risk of "no abnormal water flow but the slope is on the verge of instability", thus avoiding reservoir bank collapse accidents. The three types of parameters are identified in synergy, forming a three-dimensional anomaly screening system of "water level-water flow-structure", which greatly reduces the probability of missed or false alarms and ensures that no potential anomaly units are missed.

[0051] For each unit in the "first anomaly set", tracing along the "runoff direction (downstream)" or "reverse backflow direction (countercurrent)" to the reservoir tail or downstream can clearly show the propagation path of the anomaly. For example, if the anomaly unit is located at the tributary inlet, tracing can reveal that the countercurrent is formed from the tributary supported by the main stream and is spreading upstream along the tributary; if the anomaly unit is located near the generating unit, tracing can determine whether the countercurrent comes from the reservoir tail backflow. The generated "hydrodynamic evolution path set" provides spatial dimension basis for subsequent risk assessment (such as whether the backflow will affect the effective reservoir capacity or whether it will impact the generating units), allowing dispatchers to not only know "where the anomaly is", but also to know "where the anomaly comes from and where it will go", providing precise spatial guidance for subsequent development of protection strategies to "block the spread of the anomaly" (such as adjusting tributary gates and optimizing the flood discharge rhythm).

[0052] Example 3 This embodiment is an explanation of Embodiment 1. Specifically, for the water segment of the p-th path in the set of hydrodynamic evolution paths, its grid cells are numbered in the order of the path as m=1,2,…,Np, where Np is the total number of grid cells for that path. Each grid cell has its own average velocity vector. After projecting onto the mainstream direction, we get Going with the current is positive, going against the current is negative; Along the hydrodynamic evolution path, the reverse flow intensity in the mainstream direction of each grid cell is accumulated, and the reverse hydrodynamic chain index is calculated. :

[0053] in, This is a safe flow rate reference value; This represents the average flow velocity of the m-th grid cell in the p-th path; When 0.3≤ <0.5; indicates that there is a slight backflow in the local area of ​​the path, which slightly disturbs the hydrodynamics and does not trigger an alarm command; When 0.5≤ <0.7; indicates that the backflow along this path is significant, posing a risk of localized water inrush in the reservoir area, triggering the first backflow risk warning instruction; when ≥0.7 indicates that there is a strong countercurrent in the path, which may lead to hydrodynamic instability or structural risk, triggering the second countercurrent risk warning instruction.

[0054] In this embodiment, the mainstream flow velocity of the grid cells is sequentially taken along the path. Using the safe flow velocity as a benchmark, the abstract countercurrent is transformed into a quantifiable value by accumulating the reverse flow intensity reverse hydrodynamic chain index. This avoids the limitations of traditional methods that only qualitatively determine the presence or absence of countercurrent, accurately reflecting the cumulative impact of countercurrent along the path. Three levels of risk are classified based on the index value (0.3 ≤...). <0.5 indicates a slight perturbation, 0.5≤ <0.7 indicates significant backflow. (≥0.7 indicates strong backflow), corresponding to different early warning instructions. Minor disturbances do not trigger alarms to reduce interference, while significant and strong backflows trigger different early warnings, enabling dispatchers to take precise measures according to the risk level, avoiding excessive or insufficient warnings, and improving the efficiency of risk response.

[0055] Example 4 This embodiment is an explanation of Embodiment 1. Specifically, for the water segment of the p-th path in the set of hydrodynamic evolution paths, its grid cells are numbered in the order of the path as m=1,2,…,Np, where Np is the total number of grid cells for that path. The current water level is available for each grid cell in the path element. Normal water level and grid cell volume Obtain the effective storage capacity correction coefficient :

[0056] When 0.05≤ <0.1; indicates that the water level along the path deviates from the normal storage level, but is within a controllable range, and no alarm command is triggered; When 0.1≤ <0.2; indicates that the water level deviation along the path is abnormal, posing a risk of increased reservoir pressure, triggering the first reservoir pressure risk warning instruction; when ≥0.2 indicates an abnormal water level deviation along the path, posing a risk to reservoir safety or scheduling, triggering a warning instruction for pressure risk in the second reservoir.

[0057] In this embodiment, based on the current water level, normal storage level, and volume calculation coefficient of each grid unit along the path, "water level deviation" is directly correlated with "effective storage capacity change," transforming the traditionally vague "water level anomaly" into a quantifiable storage capacity impact indicator. This clearly reflects the degree to which the water body section along the path weakens the reservoir's effective storage capacity, avoiding the one-sidedness of judging storage capacity risk solely based on water level values. Three levels of risk are classified according to coefficient values ​​(0.05≤...). <0.1 indicates a controllable offset; 0.1≤ <0.2 indicates storage capacity pressure risk. (≥0.2 indicates a safety scheduling risk), corresponding to differentiated early warning strategies. Controllable deviations do not trigger alarms, reducing invalid interference. Risk levels trigger corresponding early warnings, allowing dispatchers to accurately grasp the storage capacity pressure level, avoiding "excessive early warnings disrupting scheduling" or "insufficient early warnings ignoring risks," and improving the accuracy of storage capacity safety management.

[0058] Example 5 This embodiment is an explanation of Embodiment 1. Specifically, for the water segment of the p-th path in the set of hydrodynamic evolution paths, its grid cells are numbered in order of path as m=1,2,…,Np, and each grid cell records the water level time series. And calculate the standard deviation of water level changes. Calculate the water level resonance index :

[0059] in, This is for reference to the range of water level fluctuations within a safe range; When 0.3≤ <0.5; indicates that the water level fluctuation along the path is within the acceptable range and no alarm command will be triggered; When 0.5≤ <0.7; indicates that the water level fluctuation along the path is abnormal, and there is a risk of local resonance, triggering the first resonance risk alarm command; when ≥0.7 indicates that the water level fluctuation along the path is abnormal, and there is a potential safety hazard due to hydrodynamic resonance, triggering the second resonance risk alarm command; The water segments corresponding to the endpoints of the paths that trigger the first backflow risk warning, the second backflow risk warning, the first reservoir pressure risk warning, the second reservoir pressure risk warning, the first resonance risk alarm, and the second resonance risk alarm are marked as alarm risk water segments, forming a set of key risk units.

[0060] In this embodiment, the standard deviation of water level changes is calculated using a grid cell water level time series, and a water level resonance index is constructed by combining it with a reference safe fluctuation amplitude. This transforms the traditionally difficult-to-capture "water level resonance" phenomenon into a quantifiable numerical indicator. This index clearly reflects the degree to which water level fluctuations deviate from the safe range in the path water segment, avoiding the subjectivity and lag of judging resonance risk solely by visual observation of water level changes, and achieving scientific identification of water level resonance risk. Three levels of risk are classified based on the index value (0.3 ≤ <0.5 is considered a safe fluctuation; 0.5 ≤ <0.7 indicates a local resonance trend. (≥0.7 indicates a potential safety hazard), corresponding to differentiated alarm strategies. Safety fluctuations do not trigger alarms, reducing dispatching interference; local resonance trends and potential safety hazards trigger different levels of alarms, allowing dispatchers to accurately grasp the severity of resonance risks and take timely targeted measures (such as adjusting the discharge rhythm and reinforcing bank facilities) to prevent the resonance risk from escalating and causing reservoir bank damage, unit instability, and other problems. The endpoint water sections corresponding to various risk warning instructions (backflow, reservoir pressure, water level resonance) are marked as alarm risk water sections, forming a set of key risk units. This avoids the waste of resources from indiscriminate control of the entire reservoir area, enabling dispatchers to quickly identify high-risk areas and concentrate on monitoring and protecting key units, significantly improving the targeting and efficiency of reservoir area risk management and laying the foundation for subsequent development of precise protection strategies.

[0061] Example 6 This embodiment is an explanation of Embodiment 1. Specifically, it extracts the water level time series for each grid cell in the set of key risk units. Where f=1,2,…,F are the sampling sequence numbers, and the time interval is… ; And based on the water level time series Calculate and obtain the peak value of water level change :

[0062] Find the sampling point where the water level begins to rise. 1. Locate the water level that has reached its peak value. sampling points And find sampling points where the water level has dropped to a stable level. Calculate the duration of water level fluctuations :

[0063] And calculate the slope of the water level rise. and the slope of water level drop :

[0064]

[0065] When the water level changes peak And the duration of water level fluctuations And the slope of the water level rise or the slope of the water level drop If this occurs, it is determined to be a water level pulse event; Peak water level change ≥ 0.5m: water level fluctuation range used to define risk level The 0.5m peak value setting is based on hydropower plant operation experience: water level fluctuations during normal reservoir scheduling (such as daily flood discharge and unit load adjustment) are mostly within 0.2m. If the peak value exceeds 0.5m, it may be a violent fluctuation caused by sudden rainstorms, tributary floods, or abnormal valve operation, which can easily lead to a sudden increase in reservoir pressure and sudden changes in the stress on the reservoir bank soil. It has risk attributes, so it is used as the basic threshold for "whether it is likely to be a pulse event". Fluctuation duration ≥ 20 minutes: Eliminate transient disturbances and identify persistent risks. If only the peak value meets the standard but the duration is short (e.g., <5 min), it may be a temporary fluctuation caused by instantaneous error of the monitoring equipment or water flow impact, and will not have a substantial impact on the stability of the reservoir area. The 20-minute setting can filter out fluctuations that "continuously affect the hydrodynamic balance of the reservoir area"—this duration is sufficient for the water level anomaly to be transmitted to the surrounding grid units, which may trigger a chain reaction (such as backwater backing, unit inlet water disturbance), so it needs to be included in the risk assessment. An upward / downward slope ≥ 0.05 m / s: quantifies the severity of fluctuations, excluding slow changes. A slope of 0.05 m / s corresponds to a water level change of 3 m within 1 minute, far exceeding the gradual rate of normal operation (normal slopes are mostly <0.01 m / s). A slope meeting this standard indicates a rapid water level change, which the reservoir's hydrodynamic system cannot adapt to in time, easily leading to turbulent flow, vortices, and other problems, exacerbating the impact on the generating units and dam. This condition can further distinguish between "slow rises / falls" and "violent pulses," avoiding the misjudgment of risk-free, slow water level changes as events.

[0066] For each grid cell in the set of key risk units, extract its velocity time series. Where f=1,2,…,F are the sampling sequence numbers, and the time interval is… ; And based on the flow velocity time series Calculate and obtain the reverse peak flow velocity :

[0067] Among them, going with the current is positive, and going against the current is negative; Find the sampling point where the flow rate begins to reverse. Sampling points that reach the reverse peak And find sampling points where the flow rate recovers downstream or decays to a stable level. Calculate the duration of reverse flow :

[0068] And calculate the reverse acceleration. and attenuation slope :

[0069]

[0070] When the flow velocity reverses its peak value And the duration of reverse flow And reverse acceleration or attenuation slope If so, it is determined to be a reverse flow event.

[0071] Peak reverse flow velocity ≥ 0.2 m / s: Defining the risk level of reverse flow intensity The 0.2 m / s reverse peak value setting is based on the actual operation of hydropower plants: the normal downstream flow velocity in the reservoir area is mostly stable between 0.3-1.5 m / s. If the peak reverse flow velocity reaches 0.2 m / s, it can significantly disturb the flow pattern at the unit intake (such as inducing small-scale vortices) and may lead to tributary backflow and increased sedimentation. This threshold can exclude weak reverse flows of <0.1 m / s (such as temporary reverse flows caused by local water flow turbulence) and accurately identify reverse flows that pose a real risk and may affect the hydrodynamic balance of the reservoir area. Reverse flow duration ≥ 5 min: Eliminate transient disturbances and pinpoint substantial impacts. If only the reverse peak value meets the standard but the duration is short (e.g., <1 min), it is mostly a temporary reverse flow caused by water flow impact or instantaneous equipment error, and will not have a substantial impact on unit operation or reservoir capacity regulation. The 5-minute setting can filter out reverse flows that "continuously affect the reservoir area"—this duration is sufficient for the reverse flow effect to spread to adjacent grid units, which may cause problems such as deterioration of the unit's water intake flow pattern and local scouring of the reservoir bank, thus avoiding misjudging risk-free instantaneous reverse flows as events requiring intervention. Reverse acceleration or decay slope ≥ 0.05 m / s²: Quantify the trend of reverse flow changes and identify risk aggravation / slow-release characteristics. An acceleration / attenuation slope threshold of 0.05 m / s² corresponds to a "reverse flow velocity change of 0.05 m / s within 1 second," far exceeding the gradual change rate of normal water flow (normal flow velocity change rate is mostly <0.01 m / s²). If the reverse acceleration meets the threshold, it indicates that the reverse flow intensity is rapidly increasing, and the risk will continue to escalate (e.g., it may quickly develop into a strong reverse flow impacting the turbine). If the attenuation slope meets the threshold, it indicates that although the reverse flow is weakening, its initial intensity is high and the attenuation is slow, and it will still continue to have an impact. This condition can further distinguish between "stable weak reverse flow" and "dynamic high-risk reverse flow," ensuring that only reverse flows with a risk diffusion trend are included in the judgment.

[0072] In this embodiment, by extracting the water level time series of grid cells, the peak value of water level changes, the duration of fluctuations, and the rise / fall slope are calculated to construct a multi-dimensional judgment standard (peak value, duration, and slope all meet the standard). This method overcomes the limitations of traditional methods that rely solely on water level values ​​to judge anomalies. It can accurately capture the pulse phenomenon of "severe water level fluctuations in a short period of time," avoiding misjudgment of risks due to the compliance of a single parameter (such as only a high peak value but a short duration) or the omission of potential risks due to parameter omissions (such as ignoring the degree of fluctuation reflected by the slope), thus achieving scientific identification of water level pulse events. Based on the flow velocity time series, the reverse flow peak value, reverse flow duration, and reverse acceleration / attenuation slope are calculated, similarly using multi-parameter collaborative judgment to determine reverse flow events. Compared to traditional methods that only determine the sign of flow velocity, this mechanism can further quantify the intensity (reverse peak), duration (duration), and trend (acceleration / attenuation slope) of the backflow, clearly distinguishing between "brief, weak backflow" and "continuous, strong backflow." This avoids misjudging risk-free, minor backflows as events requiring intervention, or ignoring high-risk, strong backflows, providing a precise basis for subsequent risk classification and response. The refined determination of water level pulses and backflow events is a core prerequisite for the subsequent principle that "meeting both simultaneously determines an actual reverse hydrodynamic event." By clarifying the criteria for these two types of events, truly risky composite events can be accurately screened out, eliminating interference from single events (water level pulses alone or backflow alone). This ensures that subsequent protection strategies are specifically targeted at high-risk scenarios, avoiding wasted resources on events with no actual risk, and significantly improving the accuracy and efficiency of hydropower plants in responding to complex hydrodynamic risks.

[0073] Example 7 This embodiment is an explanation of Embodiment 1. Specifically, when the conditions for a water level pulse event and a reverse flow event are simultaneously met at the same time along the same path, it is determined to be an actual reverse hydrodynamic event. The corresponding risk warning instructions for the grid cells are extracted, and corresponding protection strategies are generated, including: Upon receiving the first backflow risk warning instruction, a first backflow protection strategy is generated, including: increasing valve opening by 5%–10% to improve flow capacity; simultaneously reducing unit intake flow by 5%–10% to alleviate backflow pressure; and maintaining flood discharge flow at 80%–90% of the design value to maintain hydraulic balance. In response to the second backflow risk warning, the first backflow protection strategy is generated, including: increasing valve opening by 15%–25% to quickly release reverse hydrodynamic force; reducing unit intake flow by 15%–20% to prevent backflow; reinforcing the reservoir bank and activating the baffle plate; and controlling the flood discharge flow at 70%–80% to prevent excessive downstream impact.

[0074] Upon receiving the first reservoir pressure risk warning instruction, generate the first reservoir pressure protection strategy, including: increasing valve opening by 5%–10% to improve discharge efficiency; increasing unit outflow by 5%–10% to disperse reservoir pressure; and reducing reservoir capacity by 5%–8% to keep water level within a controllable range. In response to the warning instruction on the pressure risk of the second reservoir, a pressure protection strategy for the second reservoir is generated, including: increasing the valve opening by 20%–30% to quickly release the high reservoir pressure; adjusting the unit's outflow rate by 15%–25% and shutting down 30% of the units to avoid overload; and reducing the reservoir capacity by 10%–15% to prevent bank slippage and structural instability. In response to the first resonance risk warning command, a first resonance protection strategy is generated, including: slowly adjusting the valve opening by ±5%–10% to avoid rapid hydraulic disturbance; smoothly adjusting the unit flow by ±5%–10% to reduce the oscillation amplitude; and maintaining the flood discharge flow at 85%–90% to maintain stable system operation. In response to the warning instruction for the second resonance risk, a second resonance protection strategy is generated, including: quickly adjusting the valve opening by 15%–20% to promptly disrupt the resonance conditions; significantly adjusting the unit flow rate by 20%–30% to suppress system resonance coupling; and controlling the flood discharge flow rate at 60%–70% to avoid drastic fluctuations in the downstream water level.

[0075] In this embodiment, the actual reverse hydrodynamic event is determined by the dual conditions of "same path, same grid unit, same time" (water level pulse + reverse flow), completely eliminating the interference of single events (such as only water level fluctuations or only brief reverse flow). Compared with the traditional "respond to single parameter anomaly" mode, this judgment logic can accurately screen out the compound risks that truly threaten the safety of the reservoir area. It avoids scheduling disorder caused by misjudging and initiating unnecessary protective measures (such as frequent start-up and shutdown of units and repeated valve adjustments), and prevents serious consequences such as unit damage and reservoir bank instability caused by missing compound risks, thus improving the accuracy of risk identification. Differentiated and quantitative protection strategies are formulated for different warning instructions (first / second reverse flow, reservoir pressure, resonance) - for example, the first reverse flow risk only requires "5%-10% valve opening adjustment + 5%-10% unit flow reduction", while the second reverse flow risk requires "15%-25% opening increase + reservoir bank reinforcement + 70%-80% flood discharge control", and the strategy strength is strictly matched with the risk level. This "tiered and quantitative" design avoids the drawbacks of traditional "one-size-fits-all" protection (such as drastically adjusting equipment regardless of risk level). It can mitigate low-level risks with minimal intervention costs while rapidly containing high-level risks through intensive measures, balancing "protection effectiveness" with "equipment wear / dispatch costs." From "determining the actual reverse hydrodynamic event" to "extracting corresponding early warning instructions," and then to "generating a quantitative protection strategy," a complete response loop is formed. The strategy clearly defines specific operational parameters such as "valve opening adjustment range, unit flow change range, and flood discharge flow control interval," eliminating the need for dispatchers to perform additional calculations and directly guiding on-site operations, significantly shortening the time lag between "risk detection and protection execution." Especially in emergency scenarios such as strong backflow and high reservoir pressure, rapid and accurate strategy execution can effectively prevent risk spread, ensuring stable unit operation, reservoir bank structural safety, and downstream flood safety, strengthening the hydropower plant's dynamic control capabilities against complex hydrodynamic risks.

[0076] Existing technologies focus only on single parameters such as water level and flow rate, failing to capture the concealed and sudden nature of nonlinear hydrodynamic evolution in reservoir areas (such as local reverse flow and water level resonance). This method divides the "entire reservoir area" into precisely traceable "micro-units" by using a regular network to divide the hydrological-hydrodynamic grid units. It simultaneously monitors multi-dimensional parameters such as "average flow velocity vector, slope stability coefficient, and water level change rate." This not only locates the initial position of "potentially abnormal hydrodynamic units" but also traces the hydrodynamic evolution path "along the runoff / backflow direction to the reservoir tail / downstream," completely reconstructing the hydrodynamic evolution path (such as the reverse propagation process of tributary backflow). This completely solves the pain point of "knowing the anomaly but not its source and direction of spread," achieving an upgrade from "single-point monitoring" to "full-chain tracing," and significantly improving the accuracy and timeliness of risk identification.

[0077] Current technologies lack a systematic classification and assessment of hydrodynamic risks, making it difficult to distinguish the severity of different risks such as "reverse flow," "sudden changes in reservoir pressure," and "water level resonance." This method constructs three core assessment indices (reverse hydrodynamic chain index, effective reservoir capacity correction coefficient, and water level resonance index) to specifically address three key risk categories and generate "first / second" graded early warning instructions. In response to the "backflow risk", the intensity and impact range of backflow are quantified by the reverse hydrodynamic chain index, and the early warning instructions can be directly linked to the risk of the unit's water intake flow pattern (such as vortex, cavitation); To address the "risk of reservoir pressure," an effective reservoir capacity correction coefficient is used to reflect the degree to which the backflow weakens the "effective reservoir capacity and storage capacity," thus avoiding misjudgment of available reservoir capacity during scheduling. To address the "water level resonance risk," a water level resonance index is used to capture sudden changes in reservoir bank water levels, providing early warnings of bank erosion and revetment damage. This precise "risk-index-early warning" correspondence allows dispatchers to quickly determine the type and level of risk, avoiding resource waste or insufficient warnings caused by a "one-size-fits-all" approach. A single "water level pulse" or "reverse flow" in the reservoir area may be triggered by normal dispatching (such as valve regulation), and misjudging it as a risk event could lead to over-dispatching; conversely, when both are superimposed (such as sudden heavy rainfall + tributary backflow), they can easily form a serious reverse hydrodynamic event, which may be missed due to "single parameter meeting the standard." This method uses dual event judgment criteria (water level pulse events: peak water level change, fluctuation duration, etc.; reverse flow events: reverse flow velocity peak, reverse acceleration, etc.) to clarify that only when both are met at the same time can it be judged as an "actual reverse hydrodynamic event"—which not only eliminates the interference of misjudgment due to "single parameter anomaly" but also accurately identifies "composite risk events," avoiding "dispatch lag" or "overprotection" caused by ambiguous judgment, and ensuring the accuracy of risk response.

[0078] Example 8 Figure 2 This is a structural diagram of a hydropower plant hydrodynamic risk full-chain early warning system according to an embodiment of the present invention.

[0079] like Figure 2 As shown, a full-chain early warning system for hydrodynamic risks in hydropower plants includes: The data acquisition module is used to acquire the digital elevation model (DEM) of the target reservoir area, water depth measurement data, dam area structure monitoring data, and hydrodynamic data collected by flow velocity sensors, water level gauges, and ADCP monitoring equipment. The grid generation module is used to create a regular grid based on the DEM, dividing the target reservoir area into several hydrological-hydrodynamic grid cells, and constructing the average velocity vector of the i-th grid cell. Slope stability coefficient Water level change rate and water level change rate ; An anomaly identification module is used to identify anomalies within the grid cell based on the average flow velocity vector of the i-th grid cell. Slope stability coefficient Water level change rate and water level change rate Identify potential anomalous hydrodynamic units and form a set of anomalous units; The path tracing module is used to trace the set of abnormal units along the runoff direction or the reverse backwater direction to the tail end of the reservoir or the downstream river section, and construct a set of hydrodynamic evolution paths. The risk assessment module is used to calculate the reverse hydrodynamic chain index within the set of paths. Effective storage capacity correction coefficient Resonance index with water level And assess, and generate corresponding first backflow risk warning instructions, second backflow risk warning instructions, first reservoir capacity pressure risk warning instructions, second reservoir capacity pressure risk warning instructions, first resonance risk alarm instructions and second resonance risk alarm instructions based on the assessment results; The risk marking module is used to mark the water segment at the end of the path that triggers a warning or alarm command as an alarm risk water segment, forming a set of key risk units; The event determination module is used to extract the peak value of water level changes within the key risk unit. Duration of water level fluctuations Water level rise slope and the slope of water level drop Flow velocity reverse peak Duration of reverse flow Reverse acceleration and attenuation slope It determines whether a water level pulse event and a reverse flow event have occurred, and identifies the actual reverse hydrodynamic event; The protection strategy generation module is used to generate corresponding reverse flow protection strategies, reservoir pressure protection strategies, or resonance protection strategies based on the triggered risk warning instructions for the actual reverse hydrodynamic events, and output them to the hydropower plant dispatching system.

[0080] In this embodiment, the system forms a complete closed loop of "data acquisition - grid modeling - anomaly location - path tracing - risk assessment - event determination - strategy generation," breaking the limitations of scattered data and isolated modules in traditional monitoring. It achieves full-chain control from basic reservoir data to protection strategies, comprehensively capturing the evolution of hydrodynamic risks. The grid division module refines monitoring units, the anomaly identification module uses multiple parameters to determine anomalies, and the event determination module uses dual conditions to identify actual risks. Each module progresses layer by layer, eliminating interference from single parameters while accurately identifying complex risks, significantly reducing misjudgments and omissions in traditional monitoring and improving risk identification accuracy. The protection strategy generation module directly connects to risk warning commands, outputting directly executable quantitative operational parameters without additional calculations, shortening the time lag between "risk discovery" and "strategy execution," helping the dispatch system respond quickly to risks, ensuring the safety of generating units, reservoir banks, and downstream areas, and improving the operational stability of the hydropower plant.

[0081] The path tracing module, targeting sets of anomalous units, traces and constructs a set of hydrodynamic evolution paths along the runoff or reverse backflow direction, clearly identifying the path to which each anomalous unit belongs. This principle breaks through the limitations of traditional "isolated point monitoring," allowing for the tracing of risk sources (e.g., which unit the backflow begins from) and the understanding of diffusion direction (spreading along the path to the reservoir tail or downstream) through association along the same path. This enables dispatchers to clearly understand the risk evolution trajectory, providing spatial dimensional basis for subsequent precise control. The event determination module, targeting key risk units, determines whether a water level pulse and a reverse flow event simultaneously occur only within grid units along the "same path." This principle avoids logical confusion caused by cross-path determination (e.g., anomalous events from different paths are mistakenly identified as related risks), ensuring that only two types of events with a coupling relationship within the same path are identified as actual reverse hydrodynamic events, further reducing the misjudgment rate and making risk determination more closely aligned with the actual hydrodynamic evolution patterns.

[0082] The threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value, it is acceptable.

[0083] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned method for early warning of hydrodynamic risks across the entire hydropower plant chain.

[0084] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0085] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

Claims

1. A method for alarming a hydropower plant, characterized by the following steps: include: S1 divides the target reservoir area into hydrological-hydrodynamic grid units, and constructs the average flow velocity vector, slope stability coefficient, flow velocity reverse ratio and water level change rate of each grid unit as multidimensional parameters of the grid unit; S2, identify potential anomalous units based on the multidimensional parameters of the grid unit, and track the anomalous units along the runoff direction or the reverse backflow direction to construct a set of hydrodynamic evolution paths; S3, calculate the reverse hydrodynamic chain index, effective reservoir capacity correction coefficient and water level resonance index for each path in the set of hydrodynamic evolution paths, generate graded early warning instructions based on the calculation results and mark the alarm risk water body segments at the corresponding path endpoints to form a set of key risk units; S4. For each grid cell in the set of key risk units, determine whether a water level pulse event and a reverse flow event have occurred. When the conditions of the two types of events are met simultaneously at the same time in the same path, it is determined to be an actual reverse hydrodynamic event, and a protection strategy corresponding to the triggering warning command is generated.

2. The method as described in claim 1, characterized in that, S1 further includes: S11: Obtain the digital elevation model, water depth measurement data and dam structure monitoring data of the target reservoir area, and based on the digital elevation model, divide the target reservoir area into hydrological-hydrodynamic grid units using the planning grid division method. S12, construct a coordinate system based on the divided grid cells, calculate the boundary flow of the grid cells based on the instantaneous velocity distribution of the grid cells in the dam area structure monitoring data, use the boundary flow as a constraint, and obtain the average velocity vector of the grid cells by constructing a set of flow conservation equations; S13, based on the Mohr-Coulomb strength criterion, calculates the shear stress and effective normal stress using the shear strength, pore water pressure, and slope at monitoring points on the reservoir bank. Then, calculates the ratio of shear strength to actual shear stress to obtain the slope stability coefficient for the i-th grid element. S14. Determine the flow direction of the grid cells based on the digital elevation model, calculate the average velocity vector of the grid cells in that flow direction as the velocity vector along the mainstream direction, and statistically analyze the reverse flow portion of the grid cells in the time series to obtain the reverse flow ratio. S15: Collect water level time series data in grid cells and use the moving average calculation formula to obtain the water level change rate.

3. The method as described in claim 2, characterized in that, S12 further includes: Source and sink terms are constructed using the average water depth, x-direction average velocity component, and y-direction average velocity component of hydrological-hydrodynamic grid cells from water exploration measurement data. Error correction is performed using these source and sink terms. The instantaneous velocity distribution of grid cells in dam area structure monitoring data is integrated to obtain the boundary flow of the grid cell. The exchange flow between each grid cell and its adjacent grid cells is calculated using the average water depth, average velocity, and common side length between the grid cells. A continuous equation for the exchange flow is constructed using the boundary flow as a constraint. The water level increment in the grid cell is substituted into the continuous equation to construct a set of flow conservation equations, and the average velocity vector of the grid cell is obtained by inversion.

4. The method as described in claim 1, characterized in that, S2 further includes: S21. When the water level change rate exceeds the change threshold, when the flow velocity reverse ratio exceeds the ratio threshold, or when the slope stability coefficient is lower than the stability threshold, the grid cell is marked as an abnormal cell. S22, for each anomalous unit, trace it along the runoff direction or the reverse backwater direction to the tail of the reservoir or the downstream river section to obtain a set of hydrodynamic evolution paths.

5. The method as described in claim 1, characterized in that, S3 further includes: S31, For the water segment of the p-th path in the set of hydrodynamic evolution paths, number each network unit according to the path order; S32, based on the velocity vector of each grid cell along the mainstream direction, along the hydrodynamic evolution path, accumulate the reverse flow intensity of each grid cell in the mainstream direction. Based on the safe velocity reference value, the average velocity of the m-th grid cell in the p-th path, and the total number of grid cells in the p-th path, calculate and obtain the reverse hydrodynamic chain index. When the reverse hydrodynamic chain index is higher than the first hydrodynamic chain index threshold but lower than the first hydrodynamic chain index threshold, the grid cell has the risk of causing local water inrush in the reservoir area, triggering the first reverse flow risk warning instruction. When the reverse hydrodynamic chain index is higher than the second hydrodynamic chain index threshold, the grid cell has the risk of causing hydrodynamic instability or structural risk, triggering the second reverse flow risk warning instruction. S33: Obtain the effective reservoir capacity correction coefficient for the current water level, normal storage level and grid unit volume of each grid unit on the hydrodynamic evolution path. When the effective reservoir capacity correction coefficient is higher than the first correction coefficient threshold and lower than the second correction coefficient threshold, the grid unit has the risk of increased reservoir pressure and triggers the first reservoir pressure risk warning instruction. When the effective reservoir capacity correction coefficient is higher than the second correction coefficient threshold, the grid unit has the risk of reservoir safety or scheduling and triggers the second reservoir pressure risk warning instruction. S34, record the water level time series of each grid unit on each hydrodynamic evolution path, calculate the standard deviation of water level changes, and obtain the water level resonance index. When the water level resonance index is greater than the first water level resonance index threshold and less than the second water level resonance index threshold, the grid unit has a local resonance trend risk, triggering the first resonance risk alarm command. When the water level resonance index is greater than the second water level resonance index threshold, there is a hydrodynamic resonance that causes facility safety hazards, triggering the second resonance risk alarm command. S35 marks the water body segments corresponding to the path endpoints of the first countercurrent risk warning instruction, the second countercurrent risk warning instruction, the first reservoir pressure risk warning instruction, the second reservoir pressure risk warning instruction, the first resonance risk alarm instruction, and the second resonance risk alarm instruction as alarm risk water body segments, forming a set of key risk units.

6. The method as described in claim 1, characterized in that, The method for determining whether a water level pulse event has occurred in step S4 is as follows: For each grid cell in the set of key risk units, extract its water level time series. ,in, f =1,2,…,F represents the sampling sequence number, and the time interval is... ; And based on the water level time series Calculate and obtain the peak value of water level change ; Obtain sampling points where the water level begins to rise.

1. Locate the water level that has reached its peak value. sampling points And find sampling points where the water level has dropped to a stable level. Calculate the duration of water level fluctuations : And calculate the slope of the water level rise. and the slope of water level drop : When the water level changes peak And the duration of water level fluctuations And the slope of the water level rise or the slope of the water level drop If this occurs, it is determined to be a water level pulse event.

7. The method as described in claim 1, characterized in that, The method for determining whether a reverse flow event has occurred in step S4 is as follows: For each grid cell in the set of key risk units, extract its velocity time series. Where f=1,2,…,F are the sampling sequence numbers, and the time interval is… ; And based on the flow velocity time series Calculate and obtain the reverse peak flow velocity : Obtain the sampling point where the flow rate begins to reverse. Sampling points that reach the reverse peak And find sampling points where the flow rate recovers downstream or decays to a stable level. Calculate the duration of reverse flow : And calculate the reverse acceleration. and attenuation slope : When the flow velocity reverses its peak value And the duration of reverse flow And reverse acceleration or attenuation slope If so, it is determined to be a reverse flow event.

8. The method as described in claim 5, characterized in that, The method for generating the protection strategy corresponding to the triggering warning command in step S4 is as follows: Upon receiving the first backflow risk warning instruction, a first backflow protection strategy is generated, including: increasing valve opening by 5%–10% and simultaneously reducing unit intake flow by 5%–10%; maintaining flood discharge flow at 80%–90% of the design value; In response to the second backflow risk warning, the first backflow protection strategy is generated, including: increasing valve opening by 15%–25%; rapidly reducing the unit intake flow by 15%–20%; reinforcing the reservoir bank and activating the baffles; and controlling the flood discharge flow at 70%–80%. Upon receiving the first reservoir capacity pressure risk warning instruction, a first reservoir capacity pressure protection strategy is generated, including: increasing valve opening by 5%–10%, increasing unit outflow by 5%–10%, and reducing reservoir capacity by 5%–8%; Upon receiving the warning instruction regarding the pressure risk of the second reservoir, a pressure protection strategy for the second reservoir is generated, including: increasing valve opening by 20%–30%; adjusting the unit's outflow rate by 15%–25% and shutting down 30% of the units; and reducing the reservoir capacity by 10%–15%. Upon receiving the first resonance risk warning command, a first resonance protection strategy is generated, including: slowly adjusting the valve opening by ±5%–10%; steadily regulating the unit flow rate by ±5%–10%; and maintaining the flood discharge flow rate at 85%–90%. In response to the warning instruction for the second resonance risk, a second resonance protection strategy is generated, including: quickly adjusting the valve opening by 15%–20%; significantly adjusting the unit flow rate by 20%–30%; and controlling the flood discharge flow rate at 60%–70%.

9. A full-chain early warning system for hydrodynamic risks in hydropower plants, characterized in that, include: The data acquisition module is used to acquire digital elevation models, water depth measurement data, and dam structure monitoring data of the target reservoir area, and to receive hydrodynamic data collected by flow velocity sensors, water level gauges, and ADCP monitoring equipment. The grid generation module is used to establish a regular grid based on the digital elevation model, divide the target reservoir area into several hydrological-hydrodynamic grid units, and construct the average flow velocity vector, slope stability coefficient, water level change rate, and water level change rate of the i-th grid unit. An anomaly identification module is used to identify potential abnormal hydrodynamic units within the grid cell based on the average flow velocity vector, slope stability coefficient, water level change rate, and water level change rate of the i-th grid cell, and to form a set of abnormal units. The path tracing module is used to trace the set of abnormal units along the runoff direction or the reverse backwater direction to the tail end of the reservoir or the downstream river section, and construct a set of hydrodynamic evolution paths. The risk assessment module is used to calculate and assess the reverse hydrodynamic chain index, effective reservoir capacity correction coefficient, and water level resonance index in the path set, and generate corresponding first countercurrent risk warning instructions, second countercurrent risk warning instructions, first reservoir capacity pressure risk warning instructions, second reservoir capacity pressure risk warning instructions, first resonance risk alarm instructions, and second resonance risk alarm instructions based on the assessment results. The risk marking module is used to mark the water segment at the end of the path that triggers a warning or alarm command as an alarm risk water segment, forming a set of key risk units; The event determination module is used to extract the peak value of water level change, the duration of water level fluctuation, the slope of water level rise and the slope of water level fall, the peak value of reverse flow velocity, the duration of reverse flow, the reverse acceleration and the attenuation slope for each grid cell in the set of key risk units, to determine whether a water level pulse event and a reverse flow event have occurred, and to identify the actual reverse hydrodynamic event. The protection strategy generation module is used to generate corresponding reverse flow protection strategies, reservoir pressure protection strategies, or resonance protection strategies based on the triggered risk warning instructions for the actual reverse hydrodynamic events, and output them to the hydropower plant dispatching system.

10. A computer-readable storage medium storing a computer program that, when executed by a processor, implements the method as claimed in any one of claims 1-8.

Images