Intelligent water level monitoring and spillway flood volume early warning system for reservoir spillway

Abstract

The present application relates to the technical field of water level monitoring and early warning, and specifically relates to an intelligent reservoir spillway water level monitoring and spillway capacity early warning system, comprising: an anti-interference water level sensing module: an anti-silt sensing array is arranged at a key section of the spillway, and the key section includes a curved section and a gate section; a dynamic hydraulic twin module: based on the water surface transverse slope, the absolute water level and the gate vibration displacement, and combined with a hydraulic BIM model, the module is processed through a lightweight fluid network algorithm; an edge cloud collaborative flow calculation module: comprising an edge node and a cloud server; a risk adaptive early warning module: when the calibrated real-time spillway capacity breaks through the spillway threshold, the early warning is executed; the present application significantly reduces the flow calculation error under high turbulence conditions, effectively improves the accuracy of spillway capacity calculation, and ensures more accurate flood control, especially in complex curved flow conditions and significantly changed gate opening conditions.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of water level monitoring and early warning, and particularly relates to an intelligent reservoir spillway water level monitoring and spillway capacity early warning system. BACKGROUND

[0002] As an important hydraulic structure for regulating reservoir water level and ensuring safe operation of reservoir, the spillway capacity and regulation mode of the reservoir spillway directly affect the flood control efficiency and the safety of the downstream area. In the traditional reservoir management mode, the spillway flow calculation relies on a static hydraulic model to calculate the spillway flow with fixed design roughness and empirical formula. However, this method has several limitations:

[0003] The traditional method does not consider the dynamic changes of the flow pattern of the spillway, especially under complex working conditions such as nonlinear flow pattern in the curved section, gate opening change and turbulence enhancement. The fixed roughness model is difficult to accurately describe the flow characteristics, resulting in large calculation error of the spillway capacity, which may cause the risk of excessive flood discharge or insufficient regulation.

[0004] The traditional reservoir flood discharge management relies on historical experience and fixed regulation rules, and cannot effectively respond to sudden heavy rainfall or rapid increase of upstream flood. Due to the lack of flow prediction mechanism, the management personnel usually adjust the gate in a post-adjustment manner, which has the problems of response lag and water level overshoot, affecting the flood control safety. SUMMARY

[0005] The present application provides an intelligent reservoir spillway water level monitoring and spillway capacity early warning system.

[0006] The intelligent reservoir spillway water level monitoring and spillway capacity early warning system comprises:

[0007] The anti-interference water level sensing module: an anti-deposition sensing array is arranged at the key section of the spillway, including the curved section and the gate section, wherein:

[0008] The curved section adopts a dual-frequency millimeter wave radar (24GHz / 77GHz) to scan the water surface transverse slope;

[0009] The gate section is provided with a pressure-ultrasonic wave fusion sensing group to synchronously obtain the absolute water level and the gate vibration displacement;

[0010] The dynamic hydraulic twin module: based on the water surface transverse slope, the absolute water level, the gate vibration displacement and combined with the hydraulic BIM model, the following is processed through the lightweight fluid network algorithm:

[0011] ① Calculate the centrifugal force correction coefficient based on the transverse slope;

[0012] ② Convert the vibration displacement into dynamic contraction rate of the flow section;

[0013] ③ Every 5 minutes, generate a digital twin including the turbulence intensity;

[0014] Edge cloud collaborative traffic solving module: including edge node and cloud server, wherein:

[0015] The edge node performs: using the improved Manning formula to calculate the benchmark spillway, wherein the roughness coefficient is dynamically updated by the digital twin;

[0016] The cloud server performs: predicting the future traffic deviation value (30 minutes) through the long short-term memory network (LSTM), and returning to the edge node for calibration, which includes superimposing the traffic deviation value on the benchmark spillway to generate the calibrated real-time spillway;

[0017] Risk adaptive early warning module:

[0018] When the calibrated real-time spillway breaks through the spillway threshold, the following is performed:

[0019] ① According to the turbulence intensity of the digital twin, the risk level is divided;

[0020] ②, based on the risk level, trigger the gate opening control instruction, and synchronously push to the reservoir dispatching terminal.

[0021] Optionally, the layout mode of the anti-deposition sensor array includes:

[0022] In the curved section, a dual-frequency millimeter wave radar group is arranged along the path, wherein the 24GHz radar array scans the water surface transverse slope at an elevation angle of 15°, and the slope angle is solved by phase difference; the 77GHz radar captures the water surface micro deformation in a vertical incidence manner, and the water surface transverse slope value α fused is generated after Kalman filtering fusion of the dual-frequency data to resist fog and haze interference.

[0023] In the gate section, a pressure-ultrasonic fusion sensor group is installed on the water surface, wherein the pressure sensor array is distributed according to the isobaric line, a MEMS piezoresistive probe is arranged every 50 cm, the absolute water pressure value is measured, the water level is solved, and the ultrasonic transmitter is embedded in the center of the pressure sensor. A 1MHz pulse wave is emitted through a 10° deflection angle, and the gate vibration displacement Δd is solved according to the echo time delay difference.

[0024] Optionally, the dynamic hydraulic twin module specifically includes:

[0025] Centrifugal force correction coefficient calculation unit: based on the water surface transverse slope angle α fused of the curved section and the hydraulic radius R in the hydraulic BIM model, a time attenuation factor is introduced, and the centrifugal force correction coefficient K c is calculated through the lightweight fluid network algorithm.

[0026] Dynamic contraction rate conversion unit: combined with the vibration frequency and the gate design opening, the gate vibration displacement Δd is mapped to the flow area contraction rate η;

[0027] Digital twin generation unit: combined with the centrifugal force correction coefficient and the flow area contraction rate, it is input into the lightweight fluid network for numerical solution to obtain the flow field distribution. The lightweight fluid network is based on the Reynolds average N-S equation, calculates the flow velocity change of water flow in the bend and gate area, and quantifies the turbulence intensity. After calculation, the key parameters are written into the attribute table of the digital twin.

[0028] Optionally, in the digital twin generation unit, the following operations are performed every 5 minutes:

[0029] (1) input K c , η into the lightweight fluid network to solve the Reynolds average N-S equation:

[0030] Where τ ij is the Reynolds stress term, δ iθ is the centrifugal force direction tensor, is the Reynolds average flow velocity component, representing the average flow velocity of water flow in the i direction, is the Reynolds average flow velocity component, representing the average flow velocity of water flow in the j direction, combined with the spillway flow field, is the local flow velocity change rate, is the convection term, ρ is the water density, is the Reynolds average pressure, representing the average pressure of the fluid, is the pressure gradient force, ν is the dynamic viscosity, representing the dynamic viscosity of water, taking 1.0×10 -6 m 2 / s, is the viscous dissipation term, τ ij is the Reynolds stress term, representing the additional stress caused by turbulence;

[0031] (2) output the global turbulence intensity I t through the embedded solver;

[0032] (3) write K c , η, I t into the digital twin attribute table, and fuse with the geometric data of the hydraulic BIM model to generate a visual flow field atlas.

[0033] Optionally, in the edge node, the reference spillway Q b is calculated using the improved Manning formula:

[0034] Where the update of the dynamic roughness coefficient n dyn is represented as:

[0035] where n0 is the reference roughness in the hydraulic BIM model (extracted from the digital twin property library), I t is the current turbulence intensity output by the digital twin, η is the contraction ratio of the flow section, A represents the area of the flow section, R is the hydraulic radius, S is the energy slope, which is approximated by the lateral slope of the water surface, R 2 / 3 represents the 2 / 3 power of the hydraulic radius, which is used in the Manning formula to describe the dependence of the water flow on the channel geometry, S 1 / 2 represents the square root of the energy slope, which is used in the Manning formula to represent the influence of the gravitational driving force of the water flow on the flow rate.

[0036] Optionally, the cloud server uses a long short-term memory network to predict the flow deviation value in the next 30 minutes, and the input features of the long short-term memory network include:

[0037] the reference spillway volume time series data in the first hour (sampling interval 5 minutes);

[0038] I t output by the digital twin, c η;

[0039] the rainfall intensity predicted by the weather forecast in the next 30 minutes;

[0040] The output of the long short-term memory network includes the flow deviation value ΔQ(t) in the next 30 minutes, with a time resolution of 5 minutes.

[0041] Optionally, the edge node performs the following operations every 5 minutes:

[0042] receives the ΔQ(t) sequence issued by the cloud;

[0043] aligns the reference spillway volume Q b (t) by timestamp;

[0044] generates the calibrated real-time spillway volume: Q cal (t) = Q b (t) + ΔQ(t)·sigmoid(t / 30), where sigmoid(t / 30) is a time weight function that gives higher confidence to the near-end prediction, with the prediction weight decreasing from 0.82 for the next 5 minutes to 0.12 for the next 30 minutes, in line with the physical law that hydrological prediction uncertainty increases with time.

[0045] Optionally, the spillway threshold in the risk adaptive warning module is represented as Q th When the calibrated real-time spillway volume Q cal ≥ Q th , a warning is triggered, where Q designβ represents the design flood discharge capacity in the hydraulic engineering BIM model, β is the operating condition adjustment factor (1.2 during the flood season and 0.9 during the non-flood season), and I... t,max This is the maximum permissible turbulence intensity in the material durability library (25% for concrete structures).

[0046] Optionally, classifying risk levels based on the turbulence intensity of the digital twin specifically includes:

[0047] I t When the percentage is less than 12%, it is classified as Level I.

[0048] 12%≤I t If the percentage is less than 18%, it is classified as Level II.

[0049] I t If the percentage is greater than 18%, it is classified as Level III.

[0050] Optionally, the trigger gate opening control command includes:

[0051] Enter the current risk level;

[0052] Output gate priority control sequence:

[0053] If the risk level is Level I: Maintain the status quo and increase the monitoring frequency;

[0054] If the risk level is Level II: prioritize opening the gate furthest from the curve (to reduce the impact of secondary flow);

[0055] If the risk level is Level III: alternately open the gates to create staggered peak discharge (interval ≤ 30 seconds).

[0056] The beneficial effects of this invention are:

[0057] This invention introduces a dynamic hydraulic digital twin to calculate key parameters such as the lateral slope of the water surface and the vibration displacement of the gate in real time. It also combines a lightweight fluid network to solve the Reynolds-averaged Navier-Stokes equations, dynamically updating the turbulence intensity, centrifugal force correction coefficient, and cross-sectional contraction rate. By adopting an improved Manning formula and calibrating it with the dynamic roughness coefficient fed back by the twin, the flow rate calculation can be adaptively adjusted according to changes in the flow state. Compared with the traditional static roughness method, this invention significantly reduces the flow rate calculation error under high turbulence conditions, effectively improving the accuracy of spillway calculation, especially in scenarios with complex flow patterns in bends and significant changes in gate opening, ensuring more precise flood discharge control.

[0058] The application adopts edge computing + cloud intelligent prediction architecture, performs benchmark flood discharge calculation through the edge node, and the cloud predicts future traffic deviation based on LSTM, so that the system can adjust the flood discharge strategy in advance. In the data fusion process, a time-weighted correction mechanism is introduced, the trust degree is adjusted according to the prediction time distance, the near-end prediction is ensured to be more reliable, compared with the traditional flood discharge scheme based on the static hydraulic model, the application can predict the abnormal trend of flood discharge in advance, dynamically adjust the gate opening, reduce the risk of sudden excess flow, and improve the intelligent level of reservoir flood discharge.

[0059] The application adopts a dynamic threshold determination model based on turbulence intensity correction, dynamically adjusts the early warning trigger threshold according to the ratio of the current turbulence intensity and the historical maximum tolerance intensity, avoids the problem of early or late warning caused by turbulence enhancement. Combined with the hierarchical early warning mechanism, the early warning level can be divided according to the amplitude of the overflow discharge exceeding the design value, and the gate opening adjustment and the reservoir scheduling terminal are linked, so as to realize the whole-process automatic control from mild flood discharge regulation to emergency flood discharge response. BRIEF DESCRIPTION OF DRAWINGS

[0060] In order to more clearly illustrate the technical solutions in the application or prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description only illustrate the application, and other drawings can also be obtained by those skilled in the art without creative effort.

[0061] Fig. 1 The system function module schematic diagram of the embodiment of the application is shown in the figure.

[0062] Fig. 2 The dynamic hydraulic twin module schematic diagram of the embodiment of the application is shown in the figure. DETAILED DESCRIPTION

[0063] The application will be described in detail below in combination with the drawings and specific embodiments. It should be noted that in order to make the embodiments more detailed, the following embodiments are the best, preferred embodiments, and other alternative ways can also be used by those skilled in the art to implement; and the drawings are only used to describe the embodiments more specifically, and are not intended to specifically limit the application.

[0064] It should be noted that in the specification, "one embodiment", "embodiment", "exemplary embodiment", "some embodiments" and the like indicate that the described embodiments can include specific features, structures or characteristics, but not necessarily every embodiment includes the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in combination with an embodiment, it should be within the knowledge of those skilled in the art to realize this feature, structure or characteristic in combination with other embodiments (whether or not explicitly described).

[0065] Generally, terms can be understood at least partly from their use in context. For example, depending at least partly on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in a singular sense, or a combination of features, structures, or characteristics in a plural sense. Additionally, the term "based on" can be understood not necessarily to convey an exclusive set of factors, but rather, alternatively, depending at least partly on the context, to allow for the presence of other factors that are not necessarily explicitly described.

[0066] like Figs. 1-2 As shown, the intelligent reservoir spillway water level monitoring and spillway volume early warning system includes:

[0067] Anti-interference water level sensing module: Anti-siltation sensor arrays are deployed at key sections of the spillway, including bend sections and gate sections, where:

[0068] The curved section uses dual-frequency millimeter-wave radar (24GHz / 77GHz) to scan the lateral slope of the water surface;

[0069] A pressure-ultrasonic fusion sensor group is deployed in the gate section to simultaneously acquire the absolute water level and the gate vibration displacement.

[0070] Dynamic hydraulic twin module: Based on the lateral slope of the water surface, absolute water level, and gate vibration displacement, combined with the hydraulic engineering BIM model, the following are processed using a lightweight fluid network algorithm:

[0071] ① Calculate the correction coefficient for centrifugal force on curves based on lateral slope;

[0072] ② Convert the vibration displacement into the dynamic contraction rate of the flow section;

[0073] ③ Generate a digital twin including turbulence intensity every 5 minutes;

[0074] Edge-cloud collaborative traffic calculation module: includes edge nodes and cloud servers, among which:

[0075] Edge node execution: The improved Manning formula is used to calculate the baseline spillway, where the roughness coefficient is dynamically updated by the digital twin;

[0076] The cloud server executes the following: predicts the flow deviation value in the future time (30 minutes) through a Long Short-Term Memory (LSTM) network, and sends it back to the edge node for calibration. The edge node calibration includes superimposing the flow deviation value onto the baseline overflow to generate the calibrated real-time overflow.

[0077] Risk adaptive early warning module:

[0078] When the calibrated real-time overflow volume exceeds the overflow threshold, the following action is taken:

[0079] ① The spillway volume only reflects the total amount of water and cannot represent the distribution of fluid kinetic energy (such as the local damage risk of high-speed water flow impacting the gate pier). The sudden change of spillway volume often lags behind the instability of flow state (the sudden increase of turbulence intensity is earlier than the significant change of flow rate). Therefore, the risk level is divided according to the turbulence intensity of the digital twin;

[0080] ② Trigger the gate opening control instruction based on the risk level and push it to the reservoir dispatching terminal simultaneously.

[0081] The layout mode of the anti-deposition sensor array includes:

[0082] In the curved section, a dual-frequency millimeter wave radar group is arranged along the path, wherein the 24GHz radar array scans the transverse slope of the water surface at an elevation angle of 15°, and the slope angle is calculated by phase difference; the 77GHz radar captures the micro-deformation of the water surface in a vertical incidence manner, and the water surface transverse slope value α fused is generated after Kalman filtering fusion of the dual-frequency data to resist the interference of fog and haze.

[0083] Calculation of transverse slope angle in curved section:

[0084] Phase difference height measurement formula: Where Δh is the height difference between adjacent water surface measurement points, c is the speed of light, is the phase difference of 24GHz radar echo, f is the center frequency of the radar (24GHz = 24x10 9 Hz);

[0085] Slope angle calculation: Where α 24 is the transverse slope angle calculated based on 24GHz, L is the distance between adjacent radar measurement points, and α 77 is calculated in the same way, and then the weight (value 0-1) of 77GHz radar data is calculated: Where C is the real-time fog and haze concentration, C0 is the concentration threshold (50mg / m 3 ), k1 is the adjustment factor (0.1); the fused water surface transverse slope value α fused is represented as:

[0086] α fused =w 77 ·α 77 +(1-w 77 )·α 24 ;

[0087] A pressure-ultrasonic fusion sensor group is installed on the water-facing surface of the gate section, wherein a pressure sensor array is distributed along an isobaric line, a MEMS piezoresistive probe is arranged every 50 cm, absolute water pressure value is measured, water level is calculated, and the static water pressure equation P = ρgh1 + P0 is used to calculate the water level, wherein P is the total pressure measured by the pressure sensor, ρ is the water density, g is the acceleration of gravity, h1 is the absolute water level height, and P0 is the atmospheric pressure, and the calculation result is: The ultrasonic transmitter is embedded in the center of the pressure sensor, emits a 1MHz pulse wave through a 10° deflection angle, and calculates the gate vibration displacement Δd according to the echo time delay difference.

[0088] Ultrasonic vibration displacement calculation:

[0089] Time delay difference ranging: Wherein, Δd is the gate vibration displacement, v is the ultrasonic wave propagation speed in the medium (1500m / s in water), and Δt is the time delay difference of adjacent pulse echoes;

[0090] Displacement direction determination: displacement direction = sign(Δt1-Δt2)·cosθ, wherein Δt1 and Δt2 are the echo time delays of two offset probes, and θ is the ultrasonic wave emission deflection angle 10°.

[0091] Example: The hydraulic radius R of a certain reservoir spillway is 85m, and the deployment scheme is as follows:

[0092] Radar arrangement in the curved section: 1 set of dual-frequency radar is arranged every 10m along the center line of the curved section (a total of 8 sets);

[0093] The 24GHz radar detects a transverse slope angle α = 0.8°, and the 77GHz corrected α' = 0.78°;

[0094] Gate section probe installation: 5 columns × 3 rows of fusion probe matrices are installed on the water-facing surface of the 4-hole gate.

[0095] The dynamic hydraulic twin module specifically includes:

[0096] Centrifugal force correction coefficient calculation unit: based on the transverse slope angle α of the water surface in the curved section fused and the hydraulic radius R in the hydraulic BIM model, a time attenuation factor is introduced, and a centrifugal force correction coefficient K is calculated through a lightweight fluid network algorithm c :

[0097] Wherein, is the exponential decay term (time decay factor) quantifying the effect of centrifugal force on the adaptation time of the curved flow (traditional methods ignore the time accumulation effect), v is the real-time flow velocity, t is the flood duration (extracted from the hydraulic BIM model for the design flood hydrograph), τ is the curved flow regime response time constant (taken as R / 2v);

[0098] Dynamic contraction ratio conversion unit: combined with the vibration dominant frequency and the gate design opening, the gate vibration displacement Δd is mapped to the flow section contraction ratio η:

[0099] where f v is the vibration dominant frequency (extracted by FFT from the vibration waveform peak frequency), D is the gate design opening (obtained from the hydraulic BIM model), t v is the time stamp within the vibration period, and T is the gate inherent vibration period (provided by the material attribute library in the hydraulic BIM model);

[0100] Digital twin generation unit: combined with the centrifugal force correction coefficient and the flow section contraction ratio, it is input into the lightweight fluid network for numerical solution to obtain the flow field distribution. The lightweight fluid network is based on the Reynolds-averaged N-S equation, which calculates the flow velocity change of the water flow in the curved channel and the gate area, and quantifies the turbulence intensity. After the calculation is completed, the key parameters are written into the attribute table of the digital twin.

[0101] In the digital twin generation unit, the following operations are performed every 5 minutes:

[0102] (1) Input K c , η into the lightweight fluid network to solve the Reynolds-averaged N-S equation:

[0103] where τ ij is the Reynolds stress term, δ iθ is the centrifugal force direction tensor, is the Reynolds-averaged flow velocity component, representing the average flow velocity of the water flow in the i direction, mainly used to describe the velocity component of the main flow direction of the spillway flow, is the Reynolds-averaged flow velocity component, representing the average flow velocity of the water flow in the j direction, combined with the flow field of the spillway, it can be used to describe the vertical or horizontal velocity component, is the local flow velocity change rate, reflecting the flow velocity gradient changing with time, used to capture the acceleration effect of the water flow in the sudden spillway event, is the convection term, representing the rate of change of the water flow with position in space, used to calculate the change of flow velocity at different positions, such as the flow velocity distribution in the curved channel area of the spillway, and ρ is the water density, is the Reynolds-averaged pressure, representing the average pressure of the fluid, used to calculate the force of the water flow on the side wall and gate of the spillway, is the pressure gradient force, representing the driving force of water flow due to pressure difference, used to analyze the water pressure distribution in the bend area and gate section, v is the dynamic viscosity, representing the dynamic viscosity of water, taking 1.0×10 -6 m 2 / s, is the viscous dissipation term, representing the viscous force action inside the fluid, used to describe the boundary layer effect of water flow near the spillway wall surface, τ ij is the Reynolds stress term, representing the additional stress caused by turbulence, used to model the turbulence intensity in the bend section, is the centrifugal force direction tensor, used to introduce the centrifugal force action, i represents the water flow direction, is the tangential direction of water flow, when represents the action force along the tangential direction of the bend, takes 1, otherwise takes 0;

[0104] (2) The global turbulence intensity I t is output by the embedded solver where k2 is the turbulent kinetic energy, U ∞ is the characteristic flow velocity, and the Reynolds stress term τ ij is obtained by solving the Reynolds-averaged N-S equation through the turbulence model closure, which is the core input parameter of turbulence intensity;

[0105] (3) Write K c ,η,I t into the digital twin attribute table, and fuse with the geometric data of the hydraulic BIM model to generate visual flow field atlas.

[0106] The hydraulic BIM model stores the geometric, physical and engineering data of the spillway and related hydraulic structures, mainly including:

[0107] Geometric structure information (spillway shape, bend curvature, cross-section size);

[0108] Hydraulic design parameters (design flood flow, flow velocity distribution, energy loss);

[0109] Structural material information (density, elastic modulus, vibration characteristics of gate, channel and other materials);

[0110] Operation control data (gate opening history, spillway scheduling rules);

[0111] Extract the bend curve from the spillway geometric modeling data to obtain the curvature hydraulic radius;

[0112] Obtain the design flood hydrograph from the flood flow-time relationship curve stored in the hydraulic BIM model.

[0113] The maximum opening of the gate is extracted directly from the hydraulic BIM model, or the opening data under different working conditions is read from the scheduling rule library.

[0114] The inherent vibration period of the gate in the material attribute library is extracted from the material database of the hydraulic BIM model

[0115] These parameters can be extracted and automatically calculated through the API of BIM.

[0116] The improved Manning formula is used in the edge node to calculate the reference spillway Q b :

[0117] Where, the update of the dynamic roughness coefficient n dyn is expressed as:

[0118] Where, n0 is the reference roughness in the hydraulic BIM model (extracted from the digital twin attribute library), I t is the current turbulence intensity output by the digital twin, η is the contraction rate of the flow section, A represents the area of the flow section, R is the hydraulic radius, S is the energy slope, which represents the energy loss rate per unit length of the flow, R 2 / 3 represents the 2 / 3 power of the hydraulic radius, which is used in the Manning formula to describe the dependence of the flow on the channel geometry, S 1 / 2 represents the square root of the energy slope, which is used in the Manning formula to represent the influence of the gravitational driving force of the flow on the flow.

[0119] The long short-term memory network in the cloud server predicts the flow deviation value in the next 30 minutes, and the input features of the long short-term memory network include:

[0120] The reference spillway time series data in the first hour (sampling interval 5 minutes);

[0121] I t ,k c ,η output by the digital twin;

[0122] The future 30-minute rainfall intensity predicted by the weather forecast;

[0123] The output of the long short-term memory network includes the flow deviation value ΔQ(t) in the next 30 minutes, with a time resolution of 5 minutes;

[0124] The long short-term memory network architecture is as follows:

[0125] 1. Input layer design:

[0126] Input features: reference spillway time series data; standardize the past 1 hour data (12 5-minute slices); dimension: 12x1;

[0127] Input features: real-time values of [I t ,K c ,η] concatenated into a feature vector 3x1;

[0128] Input features: meteorological rainfall intensity; future 30-minute forecast values (6 5-minute slices) normalized 6x1;

[0129] 2. Bidirectional LSTM layer, structure parameters:

[0130] Number of hidden units: 64 (32 for forward and 32 for backward);

[0131] Input gate activation function: Hard sigmoid (to accelerate convergence);

[0132] Forget gate bias initialization: 1.0 (to alleviate gradient vanishing);

[0133] Computational equation: h t = LSTM(x t , h t-1 ; W lstm ), where x t is the input feature at time t, and h t is the hidden state;

[0134] 3. Spatio-temporal attention layer: dynamically focusing on key time periods and features:

[0135] (1) Calculate feature energy:

[0136] (2) Generate attention weights:

[0137] where h i is the LSTM hidden state at the i-th time step, d j is the j-th digital twin parameter v a , W a , and b a are all trainable parameters;

[0138] 4. Mixture density output layer (MDN):

[0139] Output structure: G = 3: G is the total number of Gaussian mixture components, g is the Gaussian mixture component, π g is the mixture weight, μ g , and σ g are the mean and standard deviation

[0140] Physical constraints: ensure μ g ≥ 0 (non-negative flow bias) through a ReLU activation function.

[0141] 5. Training strategy:

[0142] Training data construction: historical flood event database (>100 times) and data twin high-fidelity simulation dataset (covering various flow state combinations);

[0143] Loss function adopts negative log-likelihood loss;

[0144] Model verification: MAE prediction performance index is adopted.

[0145] Edge node performs every 5 minutes:

[0146] Receive the ΔQ(t) sequence issued by the cloud;

[0147] Align the reference spillway capacity Q b (t) by timestamp;

[0148] Generate calibrated real-time spillway capacity: Q cal (t) = Q b (t) + ΔQ(t)·sigmoid(t / 30), where sigmoid(t / 30) is a time weight function, giving higher confidence to near-end prediction, making the prediction weight 0.82 for the next 5 minutes → 0.12 for 30 minutes, consistent with the physical law that hydrological prediction uncertainty increases with time.

[0149] The spillway threshold in the risk adaptive warning module is represented as Q th , when the calibrated real-time spillway capacity Q cal ≥ Q th , trigger warning, where Q design is the design flood discharge capacity in the hydraulic BIM model, β is the working condition adjustment factor (1.2 in flood season, 0.9 in non-flood season), I t,max is the maximum allowable turbulence intensity in the material durability library (25% for concrete structures).

[0150] According to the turbulence intensity of the digital twin, the risk level is divided into:

[0151] I t <12% is set to level I;

[0152] 12% ≤ I t <18% is set to level II;

[0153] I t >18% is set to level III.

[0154] Trigger gate opening control instructions include:

[0155] Input the current risk level;

[0156] Output gate priority control sequence:

[0157] If risk level = I level: maintain the status quo, improve monitoring frequency;

[0158] If risk level = II level: preferentially open the gate farthest from the bend (reduce the impact of secondary flow);

[0159] If risk level = III level: alternately open the gate to form staggered discharge (interval ≤ 30 seconds).

[0160] The instructions are encrypted and packaged into messages through an industrial Ethernet protocol and pushed to a dispatch terminal in real time.

[0161] The present application encompasses any alternative, modification, equivalent method and scheme made on the essence and scope of the present application. In order for the public to have a thorough understanding of the present application, specific details are described in the following preferred embodiments of the present application, and the present application can also be fully understood without the description of these details by those skilled in the art. In addition, in order to avoid unnecessary confusion to the essence of the present application, well-known methods, processes, procedures, elements and circuits, etc. are not described in detail.

[0162] The above is only the preferred embodiment of the present application, and it should be pointed out that for those skilled in the art, without departing from the principle of the present application, a number of improvements and refinements can be made, and these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. An intelligent reservoir spillway water level monitoring and spillage early warning system, characterized in that, Comprise: Anti-interference water level sensing module: anti-silt sensing array is arranged at key sections of spillway, the key sections include curved section and gate section, wherein: The curved section adopts double-frequency millimeter wave radar to scan the transverse slope of water surface; The gate section is provided with pressure-ultrasonic wave fusion sensing group to synchronously obtain absolute water level and gate vibration displacement amount; Dynamic hydraulic twin module: based on the transverse slope of water surface, absolute water level, gate vibration displacement amount and combined with hydraulic BIM model, the following is processed through lightweight fluid network algorithm: ①Based on the transverse slope to calculate the centrifugal force correction coefficient of the curved section; ②Convert the vibration displacement amount into dynamic contraction rate of the flow section; ③Generate digital twin including turbulence intensity every 5 minutes; Edge cloud cooperative flow calculation module: comprising edge node and cloud server, wherein: The edge node performs: the reference spillway is calculated by using the improved Manning formula, wherein the roughness coefficient is dynamically updated by the digital twin; The cloud server performs: the future time flow deviation value is predicted by using long short-term memory network, and is returned to the edge node for calibration, the edge node calibration includes superimposing the flow deviation value on the reference spillway to generate the calibrated real-time spillway; Risk adaptive early warning module: When the calibrated real-time spillway breaks through the spillway threshold, the following is performed: ①According to the turbulence intensity of the digital twin, the risk level is divided; ②Based on the risk level, the gate opening control instruction is triggered and is synchronously pushed to the reservoir dispatching terminal.

2. The intelligent reservoir spillway water level monitoring and spillage early warning system according to claim 1, characterized in that, The arrangement mode of the anti-silt sensing array comprises: In the curved section, double-frequency millimeter wave radar groups are arranged along the path, wherein the 24GHz radar array scans the water surface transverse slope at an elevation angle of 15°, and the slope angle is solved by phase difference; the 77GHz radar captures the water surface micro-deformation in the vertical incident manner, and the double-frequency data are fused by Kalman filtering to generate the water surface transverse slope value α resistant to haze interference fused ; The pressure-ultrasonic wave fusion sensing group is installed on the water-facing surface of the gate section, wherein the pressure sensor array is distributed according to the isobaric line, the MEMS piezoresistive probe is arranged every 50 cm, the absolute water pressure value is measured, the water level is calculated, the ultrasonic wave transmitter is embedded in the center of the pressure sensor, 1 MHz pulse wave is emitted through a 10° deflection angle, and the gate vibration displacement amount Δd is calculated according to the time delay difference of the echo. 3.The intelligent reservoir spillway water level monitoring and spill volume early warning system according to claim 1, characterized in that, The dynamic hydraulic twin module specifically comprises: The centrifugal force correction coefficient calculation unit: based on the transverse slope angle of the water surface of the curved section and the hydraulic radius R in the hydraulic BIM model, a time attenuation factor is introduced, and the centrifugal force correction coefficient K is calculated through a lightweight fluid network algorithm c ; Dynamic contraction rate conversion unit: the gate vibration displacement amount Δd is mapped to the contraction rate η of the flow section by combining the vibration main frequency and the designed opening degree of the gate; Digital twin generation unit: the centrifugal force correction coefficient and the contraction rate of the flow section are input into the lightweight fluid network for numerical solution to obtain the flow field distribution, the lightweight fluid network is based on the Reynolds average N-S equation to calculate the flow velocity change of the water flow in the curved section and the gate section, and the turbulence intensity is quantified, and after the calculation is completed, the key parameters are written into the attribute table of the digital twin.

4. The intelligent reservoir spillway water level monitoring and spillage early warning system according to claim 3, characterized in that, In the digital twin generation unit, the following operations are performed every 5 minutes: (1) K c ,η is input into a light-weight fluid network to solve the Reynolds-averaged N-S equations: Where, τ ij For the Reynolds stress term, δ iθ Let the direction of the centrifugal force be the tensor. Let be the Reynolds mean velocity component, representing the average velocity of the water flow in the i-direction. Let be the Reynolds-mean velocity component, representing the average velocity of the water flow in the j-direction, combined with the spillway flow field. For the local velocity change rate, For the convection term, ρ is the density of the water. The Reynolds mean pressure represents the average pressure of the fluid. Here, is the pressure gradient force, and v is the dynamic viscosity, representing the dynamic viscosity of water. For viscous dissipation, τ ij This is the Reynolds stress term, representing the additional stress caused by turbulence; (2) The global turbulence intensity I is output by the embedded solver t ; (3) K c , η, I t write into the digital twin attribute table, and the hydraulic BIM model geometry data fusion generates a visual flow field atlas.

5. The intelligent reservoir spillway water level monitoring and spillage early warning system according to claim 3, characterized in that, The improved Manning formula is used in the edge node to calculate the reference flood discharge Q b : where the dynamic roughness coefficient n dyn is updated as where n0 is the reference roughness in the hydraulic BIM model, I t is the current turbulence intensity output by the digital twin, η is the contraction ratio of the flow section, A represents the area of the flow section, R is the hydraulic radius, S is the energy slope, which is approximated by the lateral water surface slope, R 2 / 3 represents the 2 / 3 power of the hydraulic radius and is used in the Manning formula to describe the dependence of the water flow on the channel geometry, S 1 / 2 represents the square root of the energy slope and is used in the Manning formula to characterize the influence of the gravitational driving force of the water flow on the flow rate. 6.The intelligent reservoir spillway water level monitoring and spill volume early warning system according to claim 5, characterized in that, In the cloud server, the long short-term memory network predicts the flow deviation value in the next 30 minutes, the input features of the long short-term memory network include: The time series data of the reference spillway in the past 1 hour; I t ,K c ,η; The rainfall intensity predicted by the weather in the next 30 minutes; The output of the long short-term memory network includes the flow deviation value ΔQ(t) in the next 30 minutes, and the time resolution is 5 minutes.

7. The intelligent reservoir spillway water level monitoring and spill volume early warning system according to claim 6, characterized in that, The edge node performs the following operations every 5 minutes: Receive the ΔQ(t) sequence issued by the cloud; Time stamp aligned benchmark spill Q b (t); Generate calibrated real-time flood discharge: Q cal (t) = Q b (t) + ΔQ(t) · sigmoid(t / 30), where sigmoid(t / 30) is a time weight function, giving higher confidence to near-term prediction, with prediction weight 0.82 for 5 minutes in the future → 0.12 for 30 minutes, consistent with the physical law that hydrological prediction uncertainty increases with time. 8.The intelligent reservoir spillway water level monitoring and spill volume early warning system according to claim 7, characterized in that, The flood discharge threshold in the risk adaptive warning module is represented as Q th When the real-time flood discharge Q cal ≥Q th , a warning is triggered, Where Q design is the designed flood discharge capacity in the hydraulic BIM model, β is the working condition adjustment factor, and I t,max is the maximum allowable turbulence intensity in the material durability library. 9.The intelligent reservoir spillway water level monitoring and spill volume early warning system according to claim 4, characterized in that, The risk level is divided according to the turbulence intensity of the digital twin, specifically comprising: I t <12% is set as level I; 12% < I t <18% is set as level II; I t >18% is set to level III. 10.The intelligent reservoir spillway water level monitoring and spillage early warning system according to claim 1, characterized in that, The trigger gate opening control instruction comprises: Input the current risk level; Output the gate priority control sequence: If the risk level = I level: maintain the status quo, improve monitoring frequency; If the risk level = II level: priority to open the gate farthest from the bend; If the risk level = III level: alternating gate opening to form staggered discharge.

Images