Reservoir discharge scheduling method and system considering flow distortion of gate opening and closing transition zone

Abstract

The application discloses a reservoir discharge scheduling method considering gate opening and closing transition zone flow distortion, and relates to the technical field of reservoir operation management. The existing discharge scheduling calculates the discharge amount according to the static flow rate setting curve, and ignores the flow distortion caused by water flow inertia in the gate opening and closing transition zone when the gate is frequently opened and closed in a small range, so that the discharged water amount has systematic cumulative deviation. The gate opening degree, opening degree movement speed, acceleration and upstream and downstream water head are collected in real time, the transient flow deviation of the actual discharged flow relative to the quasi-steady discharged flow is estimated on-line through a transition zone flow distortion estimation model when the gate is in the opening and closing transition zone, the discharged water amount correction amount is obtained by integrating along the transition duration, the corrected real discharged amount is obtained by superposition, the water balance equation is updated according to the corrected real discharged amount, the discharge scheduling optimization model is solved, and the gate opening and closing actuator is driven to act, so that the cumulative deviation of the discharged water amount under the condition of frequent small-range opening and closing is eliminated, and the water amount precision of ecological flow tracking and load following is improved.

Description

Technical Field

[0001] This invention relates to the field of reservoir discharge scheduling technology, and in particular to a reservoir discharge scheduling method, system, electronic device, and computer-readable storage medium that takes into account the flow distortion in the gate opening and closing transition zone. Background Technology

[0002] Reservoir discharge scheduling is a core aspect of reservoir operation and management, undertaking multiple tasks including flood control, water supply, power generation, and downstream ecological protection. In discharge scheduling, the system needs to calculate the discharge flow from the gates based on the pre-calibrated gate flow rate calibration curve, the gate opening degree, and the upstream and downstream water heads. Based on this, the system then schedules the gate opening and closing operations to ensure the reservoir's discharge flow meets the corresponding scheduling objectives. The aforementioned gate flow rate calibration curve is typically obtained through hydraulic model tests or field measurements after the gate has reached a certain opening degree and the flow pattern has fully stabilized. It reflects the quasi-steady-state discharge flow at various gate opening degrees and is therefore also known as the static flow rate calibration curve.

[0003] In recent years, with the increasing demands for river ecological protection and the need for power grids to support the load of hydropower units, reservoir discharge scheduling has gradually shifted from large-scale opening and closing to frequent small-scale adjustments. For example, to track the ecological flow process of downstream rivers, the scheduling system needs to make small adjustments to the gate opening in real time based on inflow conditions and ecological needs; similarly, to coordinate with the power grid's load following, hydropower stations often need to open and close gates multiple times in a short period to quickly change the downstream flow. To improve the accuracy of such scheduling, existing technologies mostly focus on two aspects: refining calibration curves and constructing scheduling optimization models. The former aims to measure the correspondence between water level and downstream flow at various opening degrees more accurately, while the latter uses multi-objective optimization and model predictive control to find better gate opening and closing combinations.

[0004] However, the opening and closing of a gate is not instantaneous, but involves a continuous process of changing opening degree. During this process, due to the inertia of the water, the water flow passing through the gate cannot accelerate or decelerate synchronously with the change in opening degree. This causes the actual discharge flow rate during this process to deviate from the quasi-steady-state discharge flow rate obtained from the static flow rate curve: when the gate is open, the water flow cannot accelerate in time, resulting in a smaller actual flow rate; when the gate is closed, the water flow cannot decelerate in time, resulting in a larger actual flow rate due to inertia. This invention refers to this continuous process of changing gate opening degree as the opening-closing transition zone, and the deviation of the actual discharge flow rate from the quasi-steady-state discharge flow rate is called the flow distortion of the opening-closing transition zone. The duration of one opening-closing transition is typically 2 to 30 seconds. The water volume deviation caused by the distortion is very small, usually below the detection limit of the flow measurement equipment, and therefore difficult to measure directly.

[0005] Existing discharge scheduling methods only consider the discharge flow when the gates are at various stable opening degrees, neglecting the transient flow distortion during the gate opening and closing process. Even worse, some methods, when calibrating discharge curves, treat abrupt changes in outflow caused by gate opening and closing as outliers and discard them, then fit the remaining stable process, thus assuming in the scheduling model that the flow state is stable at each opening degree. When the reservoir frequently performs small openings and closings for ecological flow tracking or load following, the tiny water volume deviations caused by each opening and closing transition accumulate in the same direction. Furthermore, because the inertia of water flow is not asymmetrical during gate opening acceleration and closing deceleration, the flow distortions caused by a single opening and closing cannot completely cancel each other out. The residual deviation accumulates over time with the increase in the number of adjustments, ultimately causing a systemic deviation in the reservoir's water balance, resulting in a discrepancy between the recorded discharge volume and the downstream ecological flow assessment results. This flow distortion in the opening and closing transition zone has long been almost unmodeled, let alone compensated for in scheduling.

[0006] Therefore, how to characterize the transient flow distortion in the gate opening and closing transition zone online in the discharge scheduling, and compensate it into the water balance and optimization model, so as to eliminate the cumulative deviation of the discharge volume under frequent small opening and closing, is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0007] To address the problem that existing discharge scheduling methods in the background art neglect the flow distortion in the gate opening and closing transition zone, resulting in cumulative deviations in the discharge volume under frequent small-scale opening and closing, the present invention provides a reservoir discharge scheduling method, system, electronic device, and computer-readable storage medium that takes into account the flow distortion in the gate opening and closing transition zone, so as to estimate and compensate for the transient flow distortion in the opening and closing transition zone online and eliminate the systematic cumulative deviation of the discharge volume.

[0008] The first aspect of this invention provides a reservoir discharge scheduling method that takes into account flow distortion in the gate opening and closing transition zone, comprising the following steps: Step S1: Determine the target discharge flow for each scheduling period according to the discharge scheduling target, and collect the opening degree, opening speed, acceleration, upstream head and downstream head of the reservoir discharge gate in real time.

[0009] Step S2: Determine whether the gate is in the opening-closing transition zone where the opening continuously changes based on the opening speed. When it is determined that the gate is in the opening-closing transition zone, the gate uses the opening, opening speed, acceleration, upstream head, and downstream head as inputs to estimate the transient flow deviation between the actual discharge flow of the gate in the opening-closing transition zone and the quasi-steady-state discharge flow determined by the static flow rate calibration curve.

[0010] Step S3: Integrate the transient flow deviation over the duration of the current opening and closing transition zone to obtain the correction amount of the discharge volume caused by the current opening and closing, and add the correction amount of the discharge volume to the discharge volume calculated according to the static flow rate calibration curve to obtain the actual discharge volume after the gate correction.

[0011] Step S4: Update the water balance equation with the corrected actual discharge volume, resolve the discharge scheduling optimization model to obtain the target opening degree for the next scheduling period, and generate opening and closing commands based on the target opening degree to drive the gate's opening and closing actuator to operate, so that the actual discharge flow of the gate tracks the target discharge flow.

[0012] Optionally, determining whether the gate is in the opening / closing transition zone in step S2 can be as follows: when the absolute value of the opening speed is greater than a preset speed threshold, or the absolute value of the acceleration is greater than a preset acceleration threshold, the gate is determined to be in the opening / closing transition zone; when the absolute value of the opening speed is not greater than the preset speed threshold, the absolute value of the acceleration is not greater than the preset acceleration threshold, and the gate remains stable for a preset duration, the gate is determined to be in the quasi-steady-state zone, and the transient flow deviation is set to zero.

[0013] Optionally, the input to the transition zone flow distortion estimation model may further include the direction of change of the gate opening and the discharge flow of the gate at the previous sampling time; the transition zone flow distortion estimation model may be a regression model or a neural network model trained using measured discharge flow data or non-steady flow hydrodynamic numerical simulation data as samples and transient flow deviation as labels. The neural network model may be a single hidden layer feedforward neural network.

[0014] Optionally, the opening and closing transition zone includes an opening transition zone with an increased opening degree and a closing transition zone with a decreased opening degree. The opening transition zone and the closing transition zone can respectively adopt independent transition zone flow distortion estimation models or independent model parameters to characterize the asymmetric distortion of the outflow caused by the inertia of the water flow during the acceleration of the gate opening and the deceleration of the gate closing.

[0015] Optionally, the integration in step S3 can be: summing the product of the transient flow deviation at each sampling moment in the current opening and closing transition zone with the sampling period using a preset sampling period, and using the sum as the correction amount of the discharge volume caused by this opening and closing.

[0016] Optionally, the method may further include: statistically analyzing the cumulative water volume deviation between the corrected actual discharge volume and the target discharge flow rate over multiple completed scheduling periods, and incorporating the cumulative water volume deviation as a constraint or penalty term into the discharge scheduling optimization model, so that the cumulative water volume deviation is compensated when solving for the target opening in subsequent scheduling periods.

[0017] A second aspect of this invention provides a reservoir discharge scheduling system that takes into account flow distortion in the gate opening and closing transition zone, comprising a data acquisition module, a distortion estimation module, a compensation and correction module, and a scheduling control module. The data acquisition module is used to determine the target discharge flow for each scheduling period and to acquire the opening degree, opening velocity, acceleration, upstream head, and downstream head of the reservoir discharge gate. The distortion estimation module is used to estimate the transient flow deviation of the gate relative to the quasi-steady-state discharge flow when the gate is in the opening and closing transition zone using a flow distortion estimation model in the transition zone. The compensation and correction module is used to integrate the transient flow deviation over time along the opening and closing transition zone to obtain a discharge flow correction amount, and thereby obtain the corrected actual discharge amount of the gate. The scheduling control module is used to update the water balance equation with the corrected actual discharge amount and solve the discharge scheduling optimization model to obtain the target opening degree, and then drive the gate's opening and closing actuator to operate.

[0018] A third aspect of the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, it implements the steps of the above-described reservoir discharge scheduling method.

[0019] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described reservoir discharge scheduling method.

[0020] Compared with the prior art, the present invention has the following beneficial effects: Firstly, by introducing a transition zone flow distortion estimation model, this invention can characterize the transient flow distortion in the gate opening and closing transition zone, which has long been treated as noise and excluded, online. This fills the gap in the existing discharge scheduling which assumes that the flow state is stable at each opening degree, and makes the calculation of discharge flow move from quasi-steady state to a true reflection of the opening and closing process.

[0021] Secondly, this invention integrates the transient flow deviation along the opening and closing transition zone into a discharge volume correction, and then reinjects it into the water balance equation and the discharge scheduling optimization model. This allows the discharge volume to reflect the actual discharge process, thereby eliminating the systematic deviation of the water balance caused by the accumulation of small distortions in the same direction under frequent small opening and closing conditions, and improving the water volume accuracy of ecological flow tracking and load following.

[0022] Third, this invention models the opening transition zone and the closing transition zone separately, and depicts the asymmetric distortion of water flow inertia during opening acceleration and closing deceleration. This allows the residual deviation that cannot be canceled out by one opening and one closing to be compensated, which is especially important for load following scenarios that are mainly based on small opening and closing cycles.

[0023] Fourth, the method of the present invention is relatively lightweight. It only uses kinematic quantities such as the gate opening degree, opening speed, and acceleration, as well as the upstream and downstream water heads as inputs. It uses lightweight regression or small neural networks to estimate the transient flow deviation online. The computational load is small, which makes it easy to embed into the existing reservoir scheduling system without significantly increasing its burden. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below.

[0025] Figure 1 A schematic flowchart of a reservoir discharge scheduling method that takes into account flow distortion in the gate opening and closing transition zone, provided in an embodiment of the present invention.

[0026] Figure 2 A schematic diagram of the structure of a reservoir discharge scheduling system that takes into account the flow distortion in the gate opening and closing transition zone, provided in an embodiment of the present invention.

[0027] Figure 3 This is a schematic diagram illustrating the principle of flow distortion in the gate opening and closing transition zone in an embodiment of the present invention.

[0028] Figure 4 This is a schematic diagram of the transition zone flow distortion estimation model in an embodiment of the present invention.

[0029] Figure 5 This is a schematic diagram of transient flow deviation estimation and integral compensation in a small-amplitude opening and closing transition zone according to an embodiment of the present invention.

[0030] Figure 6 This is a schematic diagram comparing the cumulative water volume deviation after multiple opening and closing operations using the method of the present invention and without using the method of the present invention. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only for explaining the invention and are not intended to limit the invention. The specific opening values, head values, flow rates, threshold values, sampling periods, and neural network structure parameters used in the following embodiments are all illustrative examples and do not constitute a limitation on the scope of protection of this invention; those skilled in the art can adjust the corresponding parameters according to actual application scenarios, and the adjusted implementation methods still fall within the scope of protection of this invention. The following specific embodiments are described in the following order: First, combined with... Figure 1 The overall flow of the method described in this invention is summarized, and in conjunction with... Figure 3 The principle of flow distortion in the opening and closing transition zone is explained; secondly, the core terms involved in this invention are explained; and then, in conjunction with... Figure 1 , Figure 4 The specific implementation of each step of the method will be explained step by step; combined with... Figure 2 The modular structure of the system described in this invention is introduced, and the electronic equipment and computer-readable storage medium are described; finally, in conjunction with... Figure 5 , Figure 6 An end-to-end implementation is given.

[0032] like Figure 1 As shown, the reservoir discharge scheduling method that takes into account the flow distortion in the gate opening and closing transition zone provided by the present invention consists of four parts: real-time acquisition of opening degree and head, online estimation of transient flow deviation in the opening and closing transition zone, integral compensation of discharge volume, and scheduling optimization solution and gate control. These four parts correspond to steps S1 to S4 and are executed cyclically at each sampling time or scheduling period.

[0033] Before proceeding with each step, we will first explain some of the core terms involved in this invention.

[0034] The opening and closing transition zone refers to the time interval during which the reservoir spillway gate moves from one opening degree to another, and its opening degree continuously changes over time. Within this interval, the gate is in an opening or closing state, and the flow state of the water passing through the gate has not yet reached a stable state corresponding to the instantaneous opening degree. In contrast, the quasi-steady-state zone refers to the time interval during which the gate remains at a certain opening degree, and the flow state of the water passing through the gate has stabilized.

[0035] The static flow rate determination curve refers to a curve or table reflecting the relationship between the gate opening, upstream and downstream heads, and discharge flow rate, obtained through hydraulic model tests or field measurements after the gate has remained at various opening degrees and the flow state of the water passing through the gate has fully stabilized. The quasi-steady-state discharge flow rate refers to the discharge flow rate obtained from the current opening degree and upstream and downstream heads according to the static flow rate determination curve. It reflects the discharge flow rate after the flow state has stabilized at that gate opening degree.

[0036] The transient flow deviation refers to the difference between the actual discharge flow of the gate and the quasi-steady-state discharge flow obtained from the opening degree and upstream and downstream heads at the same moment according to the static flow rate determination curve within the opening-closing transition zone. Due to the inertia of the water flow, the actual discharge flow lags behind the quasi-steady-state discharge flow and is smaller when the gate is open, at which point the transient flow deviation is negative; when the gate is closed, the actual discharge flow is larger due to inertia, at which point the transient flow deviation is positive. This transient flow deviation changes over time, thus constituting the flow distortion in the opening-closing transition zone.

[0037] The transition zone flow distortion estimation model refers to a mapping model that takes the gate opening degree, opening velocity, acceleration, and upstream and downstream water heads as inputs and the transient flow deviation as output. The transition zone flow distortion estimation model can be a regression model or a small neural network. Its function is to calculate the transient flow deviation at the current moment online from easily obtainable opening kinematics and upstream and downstream water heads without directly measuring the instantaneous discharge flow.

[0038] The opening velocity refers to the first-order rate of change of the gate opening with respect to time; the acceleration refers to the second-order rate of change of the gate opening with respect to time, i.e., the rate of change of the opening velocity with respect to time. Both reflect the speed and pace of the gate's opening and closing motion. The sign of the opening velocity indicates whether the gate is open or closed, and its absolute value is closely related to the degree of flow distortion caused by the inertia of the water flow.

[0039] As for terms such as discharge volume correction, corrected actual discharge volume, water balance equation, discharge scheduling optimization model, and cumulative water volume deviation, they will be explained in conjunction with their application scenarios in the specific implementation of each step below.

[0040] The following is combined with Figure 1 , Figure 3 , Figure 4 The specific implementation of each step of the method will be explained step by step.

[0041] Step S1 involves real-time acquisition of the gate opening and head, and determination of the target discharge flow. The discharge scheduling target can be the ecological flow tracking target of the downstream river channel, the grid load following target of the hydropower unit, or a combination of both. Taking ecological flow tracking as an example, the scheduling system provides the target discharge flow that should be discharged during each scheduling period according to the ecological flow process line of the downstream river channel; taking load following as an example, the scheduling system calculates the corresponding target discharge flow based on the load instructions issued by the grid. The target discharge flow refers to the flow that the gate should discharge during each scheduling period, as determined by the discharge scheduling target. The length of a scheduling period can be from 5 minutes to 30 minutes, for example, 10 minutes.

[0042] The gate opening degree *e* is measured in real time by an opening degree sensor installed on the hoist or gate, such as an absolute encoder or a wire displacement sensor, with the dimension being length and the unit being meters. The opening speed *v* and acceleration *a* can be obtained by taking the first and second order differences of the opening degree over time, or they can be given by the motion control command feedforward of the hoisting actuator. When using the differential method, assuming the sampling period is Δt, and the opening degrees at two adjacent sampling times are *e(k−1)* and *e(k)*, then the opening speed at the current moment can be taken as *v(k)* = [*e(k)−*e(k−1)*] / Δt, and the acceleration can be taken as *a(k)* = [*v(k)−*v(k−1)*] / Δt. The sampling period Δt reflects the acquisition frequency of the opening degree and water head, and its value range is generally from 0.5s to 5s, with a typical value of 2s. If the sampling period is too large, there will be too few sampling points available for estimation in the opening and closing transition zone. If it is too small, it will amplify the influence of the opening degree measurement noise on the movement speed and acceleration after differential. In practice, it can be determined by combining the resolution of the opening degree sensor and the movement speed of the gate hoist through a stability test of numerical differentiation on the measured opening degree sequence.

[0043] The upstream head Hu and downstream head Hd of the gate are measured by water level gauges located upstream and downstream of the gate, respectively, and converted to the head relative to the gate opening after deducting the corresponding reference elevation. The difference between the upstream head and the downstream head, i.e., the effective head of the gate, is one of the main factors determining the discharge flow rate. The quantities collected and determined in step S1 constitute the input for estimating the transient flow deviation in subsequent step S2, as well as the basic data for water compensation and scheduling solutions in steps S3 and S4.

[0044] Step S2, which involves determining the opening and closing transition zone and estimating the transient flow deviation online, is one of the core components of this invention.

[0045] First, determine if the gate is in the opening / closing transition zone. For example... Figure 3As shown, when the gate remains stationary at a certain opening, the flow pattern of the water passing through the gate is stable, and the actual discharge flow rate is basically consistent with the quasi-steady-state discharge flow rate obtained from the static flow rate determination curve. However, when the gate is in the process of opening and closing, the actual discharge flow rate curve deviates from the quasi-steady-state discharge flow rate curve, and a deviation region that varies with time appears between the two. This deviation region is the flow distortion in the opening and closing transition zone. Therefore, whether or not the gate is in the opening and closing transition zone can be determined by whether the gate is in motion.

[0046] Specifically, the following criteria can be used: when the absolute value of the opening velocity is greater than a preset velocity threshold v_th, or the absolute value of the acceleration is greater than a preset acceleration threshold a_th, the gate is determined to be in the opening-closing transition zone; when the absolute value of the opening velocity is not greater than v_th, the absolute value of the acceleration is not greater than a_th, and this situation continues for a preset stabilization time τ, the gate is determined to have entered the quasi-steady-state zone from the motion process. The reason for introducing the stabilization time τ is that: at the instant the gate stops, the flow state of the water passing through the gate does not immediately stabilize, and the distortion caused by the inertia of the water flow will continue for a short period of time before decaying and disappearing; if no stabilization time is set, and the gate is immediately determined to be in quasi-steady state as soon as the opening velocity drops below the threshold, the distortion that still exists after the gate stops will be missed. The preset speed threshold v_th typically ranges from 2% to 10% of the gate's rated opening and closing speed. For example, when the rated opening and closing speed is 0.05 m / s, v_th can be 0.002 m / s. The preset acceleration threshold a_th can be selected based on the acceleration and deceleration capabilities of the gate hoist, with a typical value of 0.001 m / s². The preset stabilization time τ typically ranges from 5 s to 30 s, with a typical value of 10 s. It can be obtained statistically from the time required for the flow distortion to decay to the measured noise level after several measured opening and closing transitions. When the gate is determined to be in the quasi-steady-state region, the transient flow deviation is directly set to zero, meaning that no distortion compensation is performed on the downstream flow at this time, which is consistent with the handling of the existing scheduling method under steady-state conditions.

[0047] After determining that the gate is in the opening and closing transition zone, the transient flow deviation ΔQ at the current moment is estimated online using the gate opening degree e, opening velocity v, acceleration a, upstream head Hu, and downstream head Hd as inputs, through a pre-established transition zone flow distortion estimation model. The transition zone flow distortion estimation model describes the mapping relationship between the kinematic quantities of the gate opening and the head to the transient flow deviation. The physical root of this mapping relationship is that when the gate opens or closes, the acceleration or deceleration of the water downstream of the gate needs to overcome its own inertia, thereby causing additional head loss or backflow at the flow cross section, causing the actual flow rate to deviate from the calibration value obtained under the quasi-steady-state flow assumption; the greater the gate velocity, the greater the acceleration, and the greater the applied head, the more significant this deviation becomes.

[0048] In addition to the aforementioned gate opening, opening velocity, acceleration, upstream head, and downstream head, the input to the transition zone flow distortion estimation model may further include the direction of opening change and the gate discharge flow at the previous sampling time. The discharge flow at the previous sampling time can be the estimated discharge flow output by the model at the previous sampling time as a recursive input, or it can be the measured value from a field high-frequency flow meter. Introducing the direction of opening change (positive for opening, negative for closing) allows the model to distinguish between the two opposite flow distortion directions of opening and closing. Introducing the discharge flow at the previous sampling time allows the model to utilize the continuity of the discharge flow itself to more smoothly calculate the current transient flow deviation.

[0049] The transition zone flow distortion estimation model can employ either a regression model or a neural network model. When using a regression model, the transient flow deviation ΔQ can be expressed as a linear regression of each input quantity and its combination, for example, ΔQ = k1·v + k2·a + k3·v·√(Hu−Hd), where k1, k2, and k3 are regression coefficients obtained from measured or simulated samples using least-squares fitting. This regression formula is simple, highly interpretable, and suitable for gates with small variations in opening and closing speeds and relatively linear distortion patterns. When the distortion pattern is more complex and exhibits a nonlinear relationship with multiple input quantities, a neural network model can be used. Figure 4 As shown, the neural network model can be a single-hidden-layer feedforward neural network. A single-hidden-layer feedforward neural network refers to a feedforward neural network containing only one hidden layer, where information propagates unidirectionally from the input layer through the hidden layer to the output layer. Its input layer nodes correspond one-to-one with the aforementioned input quantities. The number of hidden layer nodes is one to three times the number of input quantities; for example, when there are 7 input quantities, there are 14 hidden layer nodes. The output layer is a single node that outputs the transient flow deviation ΔQ. The activation function of the hidden layer nodes can be the hyperbolic tangent function, and the output layer uses a linear activation function. This neural network model is trained on samples to make its output transient flow deviation estimate approximate the sample label. The reason for using a small, single-hidden-layer network instead of a deeper, larger network is that this invention aims to fit a physical mapping with a low input dimension and a relatively limited number of samples. An excessively large network is prone to overfitting and increases the computational load of online inference, which contradicts the invention's goal of being lightweight and easy to embed into existing scheduling systems.

[0050] The samples for the transition zone flow distortion estimation model can come from two sources. One is actual measurement: during several gate opening and closing transitions, the actual discharge flow is directly measured using a high-frequency flowmeter such as an acoustic Doppler flowmeter. This is then subtracted from the quasi-steady-state discharge flow obtained from the static flow rate constant curve at the same time, yielding the transient flow deviation as a label. This, along with the synchronously recorded opening kinematics and head, constitutes the sample. The second source is unsteady flow hydraulic numerical simulation: a one-dimensional unsteady flow mathematical model (such as the Saint-Venant equations) is established for the gate outflow. The upstream inflow process and downstream water level process are used as boundary conditions. A four-point implicit difference scheme is used for numerical discretization. The gate opening and closing transition process is simulated under various combinations of opening and closing velocities and applied head. The difference between the actual discharge flow at each simulated moment and the quasi-steady-state discharge flow obtained from the static flow rate constant curve under the same conditions is used as the label for the transient flow deviation at the corresponding moment. The two approaches mentioned above can be used individually, or simulated samples can be used for pre-training, and then fine-tuned with a small number of real-world samples to balance sample quantity and realism.

[0051] Considering that the inertia of water flow behaves differently during gate opening acceleration and closing deceleration—when opening, the water body is accelerated from stillness or slow speed, and inertia causes the flow rate to lag and be smaller; when closing, the water body is decelerated from rapid speed, and inertia maintains a larger flow rate—these two are not symmetrical in magnitude. This invention further distinguishes the opening and closing transition zone into an opening transition zone and a closing transition zone. The opening transition zone refers to the zone where the gate opening degree increases and the gate is in the opening motion; the closing transition zone refers to the zone where the gate opening degree decreases and the gate is in the closing motion. This invention establishes independent flow distortion estimation models for the opening and closing transition zones, or sets independent model parameters for opening and closing within the same model. In this way, the negative distortion during opening and the positive distortion during closing can be accurately characterized separately, avoiding the use of the same symmetrical model to fit two inherently asymmetrical operating conditions, resulting in inaccurate fitting for both conditions. If the distinction between opening and closing is not made, the water volume underestimated during a single opening and the water volume overestimated during a single closing will be forcibly symmetrical in the model. As a result, it is impossible to accurately calculate the distortion of a single cycle or reflect the residual deviation after the switching cycle.

[0052] To prevent the flow distortion estimation model in the transition zone from giving unreasonable estimates under extreme operating conditions, upper and lower bound protection can be set for its output: when the absolute value of the estimated transient flow deviation exceeds a preset proportion (e.g., 20%) of the current quasi-steady-state discharge flow, it is truncated to the value corresponding to that proportion. This is because the flow distortion in the normal opening and closing transition zone usually does not exceed this order of magnitude of the quasi-steady-state discharge flow. Once it does, it is mostly due to improper extrapolation of the model outside the sample or abnormal input data. Through upper and lower bound protection, it is possible to avoid individual abnormal estimates from contaminating the discharge flow correction amount for this opening and closing after integration.

[0053] Furthermore, the transition zone flow distortion estimation model can be continuously updated online during operation: whenever a new batch of measured opening and closing data is accumulated, it is used to incrementally train or refit the model parameters, enabling the model to slowly change the flow characteristics caused by gate seal wear, siltation downstream of the gate, etc., thereby maintaining estimation accuracy over a long period. For reservoirs with multiple spillway gates, a separate transition zone flow distortion estimation model can be established for each gate; when the gate structures are identical, a single model can be used, with the gate number as an additional input for differentiation, balancing estimation accuracy and modeling cost.

[0054] Step S2 outputs a transient flow deviation estimate ΔQ at each sampling time. These estimates are arranged along the time of the opening and closing transition zone, which constitutes the transient flow deviation process of this opening and closing, and are integrated by step S3 to obtain the discharge water volume correction.

[0055] Step S3 is the integral compensation for transient flow deviation, which is another core part of the present invention. After step S2, the transient flow deviation at each sampling time in the current opening and closing transition zone is obtained; step S3 is to convert this series of transient flow deviations that change with time into the actual amount of water released more or less during this opening and closing, and to correct the account of the discharged water volume accordingly.

[0056] Specifically, the transient flow deviation is integrated over the duration of the current opening / closing transition zone to obtain the discharge volume correction ΔW caused by this opening / closing. The discharge volume correction ΔW refers to the difference between the actual discharge volume in the current opening / closing transition zone and the discharge volume calculated according to the static flow rate constant curve, i.e., the volume of water accumulated by the flow distortion over the entire transition duration. Since the transient flow deviation is discretely given according to the sampling period Δt, the above integration is converted into a summation: the product of the transient flow deviation ΔQ(k) at each sampling time in the current opening / closing transition zone and the sampling period Δt is summed, and the resulting sum ΔW = ΣΔQ(k)·Δt is the discharge volume correction caused by this opening / closing. A negative discharge volume correction for the opening transition indicates that less water was released than the actual rate constant; a positive discharge volume correction for the closing transition indicates that more water was released than the actual rate constant.

[0057] After obtaining the corrected discharge volume, it is superimposed on the discharge volume calculated according to the static flow rate curve to obtain the corrected actual discharge volume of the gate. The discharge volume calculated according to the static flow rate curve refers to the discharge volume originally obtained by the dispatching system through time integration (summation) of the quasi-steady-state discharge flow rate, denoted as W0. The corrected actual discharge volume is denoted as W, and W = W0 + ΔW. It compensates for the water volume distortion in the opening and closing transition zone that was ignored by existing methods, thus being closer to the actual water volume released during this gate opening and closing. To illustrate, existing methods record the amount of water that should be released according to the opening degree, while this invention records the actual amount of water released during the opening and closing process. The discharge volume correction is the difference between these two records.

[0058] It should be noted that the correction amount for a single opening and closing of the reservoir is often very small, possibly only a few percent or even less than the total discharge volume, which is why it has been neglected for a long time. However, when the reservoir is frequently opened and closed for ecological flow tracking or load following, these tiny correction amounts accumulate continuously. In particular, since the correction amount for the opening transition is negative and the correction amount for the closing transition is positive, and as can be seen from the asymmetric modeling in step S2, the two are not equal. After a complete opening and closing cycle, the two cannot completely cancel each other out, and there will always be a net correction amount in the same direction. As the number of opening and closing increases, these net correction amounts accumulate in the same direction, forming a considerable cumulative water volume deviation.

[0059] Therefore, the method can further statistically analyze the cumulative water volume deviation between the corrected actual discharge volume and the target discharge flow over multiple completed scheduling periods, and incorporate this cumulative water volume deviation as a constraint or penalty term into the discharge scheduling optimization model, so that it is compensated when solving for the target opening in subsequent scheduling periods. The cumulative water volume deviation refers to the cumulative value of the difference between the corrected actual discharge volume and the target discharge flow over each completed scheduling period since a certain statistical starting point. When the cumulative water volume deviation is negative, it indicates that the actual discharge was insufficient in the previous period, and subsequent periods should release more water to compensate for it; when the cumulative water volume deviation is positive, subsequent periods should release less water. Thus, this invention not only compensates for the flow distortion in each opening and closing operation, but also eliminates the unidirectional accumulation of distortion in a closed loop across multiple opening and closing operations. This is the key difference between this invention and those that only perform single-point flow correction.

[0060] Step S4 involves scheduling optimization and gate control. After obtaining the corrected actual discharge volume, the water balance equation is updated accordingly. The water balance equation describes the change in reservoir storage volume over time. That is, ignoring minor water losses such as evaporation and seepage, the change in reservoir storage volume over a certain period is equal to the inflow volume minus the outflow volume during that period, where the outflow volume includes the discharge volume from the gates. This invention replaces the original discharge volume calculated based on the static flow rate curve with the corrected actual discharge volume in the water balance calculation, ensuring that the estimation of reservoir storage volume and water level is based on the actual discharge volume.

[0061] Under the constraints of the updated water balance equation, the discharge scheduling optimization model is re-solved to obtain the target opening degree for the next scheduling period. The target opening degree refers to the opening degree to which the gate should move in the next scheduling period, obtained from the discharge scheduling optimization model. The discharge scheduling optimization model refers to an optimization model with the optimization objective of meeting scheduling goals such as ecological flow tracking and load following, and constraints such as the water balance equation, gate opening and closing capacity, and upstream and downstream water level limits. Its solution can employ methods known in the art, such as linear programming, dynamic programming, and model predictive control, which will not be elaborated here. When the aforementioned cumulative water volume deviation is introduced as a constraint or penalty term, the optimization model will also consider eliminating the already formed cumulative deviation when solving for the target opening degree.

[0062] Finally, based on the target opening degree for the next scheduling period, a corresponding opening and closing command is generated to drive the gate's opening and closing actuator, causing the gate to move from the current opening degree to the target opening degree. The opening and closing actuator refers to the mechanical device that drives the gate's opening and closing movement, such as a hydraulic gate hoist or a winch-type gate hoist. During this movement, the gate enters a new opening and closing transition zone, and the method returns to step S1 to collect data on the new round of opening degree, movement speed, acceleration, and head. This process is repeated continuously, ensuring that the actual discharge flow of the gate continuously tracks the target discharge flow. It is evident that this invention embeds online estimation and compensation for flow distortion in the opening and closing transition zone into the closed loop of discharge scheduling. Each opening and closing is compensated in real time, and each accumulated deviation is immediately eliminated by subsequent scheduling.

[0063] like Figure 2 As shown, this embodiment of the invention also provides a reservoir discharge scheduling system that takes into account flow distortion in the gate opening and closing transition zone. It should be understood that the following system is a division of corresponding hardware or software functional modules for implementing the above method; this division is illustrative and represents a logical functional division, and other division methods may be used in actual implementation. The system includes a data acquisition module, a distortion estimation module, a compensation and correction module, and a scheduling control module.

[0064] The data acquisition module is configured to determine the target discharge flow rate for each scheduling period and to acquire the opening degree, opening velocity, acceleration, upstream head, and downstream head of the reservoir spillway gate. In one possible implementation, the data acquisition module includes an opening degree sensor, upstream and downstream water level gauges, and a data processing unit that performs differential calculations on the opening degree to obtain the opening velocity and acceleration.

[0065] The distortion estimation module is configured to estimate the transient flow deviation of the gate relative to the quasi-steady-state discharge flow rate when the gate is in the opening / closing transition zone using a transition zone flow distortion estimation model. In one possible implementation, the distortion estimation module configures independent transition zone flow distortion estimation models for the opening and closing transition zones respectively, to characterize the asymmetric distortion under the two operating conditions.

[0066] The compensation and correction module is configured to integrate the transient flow deviation over time along the opening and closing transition zone to obtain the discharge volume correction amount, and thereby obtain the actual discharge volume of the gate after correction.

[0067] The scheduling and control module is configured to update the water balance equation with the corrected actual discharge volume and solve the discharge scheduling optimization model to obtain the target opening degree, thereby generating opening and closing commands to drive the gate's opening and closing actuator. In one possible implementation, the scheduling and control module also calculates the cumulative water volume deviation and adds it as a constraint or penalty term to the discharge scheduling optimization model.

[0068] The specific working process and beneficial effects of each of the above modules can be found in the corresponding descriptions in the foregoing method embodiments. This system embodiment is used to execute the above reservoir discharge scheduling method and can achieve the same effect as the above method, so it will not be described again here.

[0069] This invention also provides an electronic device, including a memory and a processor. The memory stores a computer program, which, when executed by the processor, implements the steps of the above-described reservoir discharge scheduling method. The processor can be a central processing unit, a digital signal processor, or a field-programmable gate array, etc.

[0070] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described reservoir discharge scheduling method. The computer-readable storage medium may be a read-only memory, a random access memory, a magnetic disk, or an optical disk, etc.

[0071] The following example uses an arc-shaped spillway gate of a reservoir, combined with... Figure 5 An end-to-end description is provided for a complete execution of the method described in this invention.

[0072] The reservoir is responsible for tracking the ecological flow of the downstream river and the load of the power grid. The spillway gate is an arc-shaped gate with an opening width of 8m, a rated opening and closing speed of 0.05m / s, a sampling period of Δt=2s, a preset speed threshold of v_th=0.002m / s, and a preset stabilization time of τ=10s. The flow distortion estimation model in the transition zone has been trained using the measured opening and closing data of the gate. A set of model parameters is provided for the open transition zone and the closed transition zone.

[0073] During a certain scheduling period, to accommodate a slight increase in grid load, the scheduling system requires the gate to open from 1.50m to 1.65m according to the target discharge flow requirement. Step S1 collects the opening sequence in real time and calculates the opening velocity by difference: the opening lasts approximately 8 seconds, the opening velocity increases from zero to approximately 0.019m / s and then decreases, with the acceleration correspondingly being positive first and then negative. Since the absolute value of the opening velocity quickly exceeds v_th = 0.002m / s, step S2 determines that the gate has entered the opening transition zone.

[0074] Step S2 outputs a transient flow deviation estimate every Δt = 2s within the open transition zone. For example... Figure 5 As shown, according to the static flow rate curve, the quasi-steady-state discharge flow corresponding to opening degrees of 1.50m and 1.65m is approximately 133.5m³ / s and 146.8m³ / s, respectively. However, the transient flow deviation ΔQ estimated by the flow distortion estimation model in the opening transition zone is approximately −2.5m³ / s, −4.0m³ / s, −3.0m³ / s, and −1.5m³ / s at the four sampling times, all of which are negative values. This indicates that the actual discharge flow during the opening process is lagging behind and smaller due to the inertia of the water flow, and the deviation is most severe in the middle of the opening, i.e., at the point where the opening speed is the maximum.

[0075] Step S3 sums the transient flow deviations over the sampling period to obtain the correction amount ΔW for the discharge volume caused by this opening: ΔW = (−2.5−4.0−3.0−1.5)×2 = −22.0 m³. The discharge volume W0 calculated from the static flow rate curve during the same period is approximately (135+139+143+146)×2≈1126 m³. Therefore, the corrected actual discharge volume W = W0 + ΔW = 1126−22 = 1104 m³. It can be seen that, according to the existing accounting method, approximately 1126 m³ of water was discharged during this opening; however, after considering the flow distortion in the opening / closing transition zone, only approximately 1104 m³ was actually discharged, a shortfall of approximately 22 m³, with a relative deviation of approximately 1.95%. If this 22 m³ is ignored, it becomes the difference between the recorded volume and the actual discharge volume.

[0076] In the following scheduling period, to accommodate a slight load adjustment, the gate was closed from an opening of 1.65m to 1.50m, also lasting approximately 8 seconds. Step S2 determines that the gate has entered the closure transition zone, and the transient flow deviations at the four sampling times are estimated by the flow distortion estimation model for the closure transition zone to be approximately +2.0m³ / s, +3.5m³ / s, +2.5m³ / s, and +1.0m³ / s, respectively. These are all positive values, indicating that the actual discharge flow remains relatively large during the closure process due to the inertia of the water flow. Step S3 sums up the discharge correction amount ΔW′ = (2.0 + 3.5 + 2.5 + 1.0) × 2 = +18.0m³.

[0077] Consider this on-off cycle as a complete switching action: opening releases 22.0 m³ less, closing releases 18.0 m³ more, and the two are not equal. After the cycle, a net correction remains: -22.0 + 18.0 = -4.0 m³. That is, for each such small-scale switching cycle, the actual discharge is about 4.0 m³ less than what is recorded according to the static calibration curve. This 4.0 m³ is precisely the portion left by the asymmetry of water flow inertia, which cannot be offset by existing methods.

[0078] like Figure 6 As shown, approximately 50 such switching cycles occur during a typical operating day when the gate cycles through small openings and closings in response to the load. If existing discharge scheduling methods are used ( Figure 6 (Upper curve), the net deviation of approximately 4.0 m³ remaining in each cycle accumulates in the same direction, resulting in a cumulative water volume deviation of approximately 50 × (−4.0) = −200 m³ at the end of the day, i.e. Figure 6 The accumulated water deficit gradually increases to approximately 200 m³, and increases daily with the number of operating days, accumulating to approximately 6000 m³ in one month, becoming a considerable systemic deficit in ecological flow assessment and water balance; the method of this invention ( Figure 6 (The lower curve), in step S3 each time, the net correction amount of −4.0m³ is included in the corrected actual discharge volume and the reinjection water balance equation. In step S4, the cumulative water volume deviation is included as a penalty term in the discharge scheduling optimization model, so that the subsequent scheduling period releases more water to make up for it, thereby keeping the cumulative water volume deviation at a level close to zero.

[0079] Comparative Example 1: For the same gate and the same operating day, the existing discharge scheduling method, which does not account for flow distortion in the opening and closing transition zone, was used for calculation. Because the existing method records data based solely on the static flow rate curve, the approximately 4.0 m³ net deviation remaining after each switching cycle is completely ignored. This results in the recorded discharge volume at the end of an operating day being approximately 200 m³ more than the actual volume. Furthermore, since this deviation accumulates daily in the same direction with the number of operating days, after one month the deviation between the recorded and actual discharge volume expands to approximately 6000 m³, leading to a corresponding water deficit in the ecological flow assessment. In contrast to this comparative example, the embodiment of this invention compensates for the distortion at each opening and closing and eliminates the accumulated deviation at multiple time scales, resulting in the recorded discharge volume being essentially consistent with the actual discharge volume, with the deviation controlled within a few cubic meters.

[0080] The specific values ​​for opening degree, flow rate, and deviation given in this embodiment are exemplary data used for ease of explanation and do not constitute a limitation on the present invention.

[0081] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the protection scope of the present invention.

Claims

1. A reservoir discharge scheduling method that takes into account flow distortion in the gate opening and closing transition zone, characterized in that, include: The target discharge flow rate for each scheduling period is determined according to the discharge scheduling objectives, and the opening degree, opening speed, acceleration, upstream head and downstream head of the reservoir discharge gate are collected in real time. Based on the opening speed, it is determined whether the gate is in the opening-closing transition zone where the opening continuously changes. When it is determined that the gate is in the opening-closing transition zone, the opening, opening speed, acceleration, upstream head and downstream head are used as inputs. The transient flow deviation of the actual discharge flow of the gate in the opening-closing transition zone is estimated online relative to the quasi-steady-state discharge flow determined by the static flow rate curve. The transient flow deviation is integrated over the duration of the current opening and closing transition zone to obtain the correction amount of the discharge volume caused by the current opening and closing. The correction amount of the discharge volume is then superimposed on the discharge volume calculated according to the static flow rate constant curve to obtain the actual discharge volume after the gate correction. The water balance equation is updated with the corrected actual discharge volume, the discharge scheduling optimization model is resolved to obtain the target opening degree for the next scheduling period, and an opening and closing command is generated according to the target opening degree to drive the gate's opening and closing actuator to operate, so that the actual discharge flow of the gate tracks the target discharge flow.

2. The method according to claim 1, characterized in that, The step of determining whether the gate is in the opening-closing transition zone where the opening degree changes continuously based on the opening degree movement speed includes: determining that the gate is in the opening-closing transition zone when the absolute value of the opening degree movement speed is greater than a preset speed threshold or the absolute value of the acceleration is greater than a preset acceleration threshold. When the absolute value of the opening speed is not greater than the preset speed threshold, the absolute value of the acceleration is not greater than the preset acceleration threshold, and this condition continues for a preset stable duration, the gate is determined to be in the quasi-steady state region, and the transient flow deviation is set to zero.

3. The method according to claim 1, characterized in that, The input to the transition zone flow distortion estimation model also includes the direction of change of the opening degree and the discharge flow of the gate at the previous sampling time. The transition zone flow distortion estimation model is a regression model or neural network model trained with measured discharge flow data or non-steady flow hydrodynamic numerical simulation data as samples and transient flow deviation as labels.

4. The method according to claim 3, characterized in that, The neural network model is a single hidden layer feedforward neural network. The input layer nodes of the single hidden layer feedforward neural network correspond one-to-one with each input quantity. The number of hidden layer nodes is one to three times the number of input quantities. The output layer is a single node that outputs the transient flow deviation. The neural network model is trained on the samples to make the output transient flow deviation estimate approximate the sample label.

5. The method according to claim 1, characterized in that, The opening and closing transition zone includes an opening transition zone with an increased opening degree and a closing transition zone with a decreased opening degree. The opening transition zone and the closing transition zone respectively adopt independent transition zone flow distortion estimation models or independent model parameters to characterize the asymmetric distortion of the outflow caused by the inertia of the water flow during the acceleration of the gate opening and the deceleration of the gate closing.

6. The method according to claim 1, characterized in that, The step of integrating the transient flow deviation along the duration of the current opening and closing transition zone to obtain the discharge volume correction caused by the current opening and closing includes: summing the product of the transient flow deviation at each sampling time in the current opening and closing transition zone with the sampling period using a preset sampling period, and using the sum as the discharge volume correction caused by the current opening and closing.

7. The method according to claim 1, characterized in that, Also includes: The cumulative water volume deviation between the corrected actual discharge volume and the target discharge flow rate over multiple completed scheduling periods is statistically analyzed, and the cumulative water volume deviation is added as a constraint or penalty to the discharge scheduling optimization model so that the cumulative water volume deviation is compensated when solving for the target opening in subsequent scheduling periods.

8. A reservoir discharge scheduling system that takes into account flow distortion in the gate opening and closing transition zone, characterized in that, include: The data acquisition module is configured to determine the target discharge flow rate for each scheduling period and to acquire the opening degree, opening speed, acceleration, upstream head and downstream head of the reservoir spillway gate. The distortion estimation module is configured to estimate the transient flow deviation between the actual discharge flow of the gate and the quasi-steady-state discharge flow online using a flow distortion estimation model in the transition zone when the gate is in the opening and closing transition zone where the opening degree changes continuously. This is based on the opening degree, the opening degree movement speed, the acceleration, and the upstream and downstream water heads of the gate. The compensation and correction module is configured to integrate the transient flow deviation along the opening and closing transition zone to obtain the discharge volume correction amount, and superimpose it on the discharge volume calculated according to the static flow rate constant curve to obtain the actual discharge volume after the gate correction. The scheduling and control module is configured to update the water balance equation with the corrected actual discharge volume and solve the discharge scheduling optimization model to obtain the target opening degree for the next scheduling period, and then generate opening and closing commands to drive the gate's opening and closing actuator to operate.

9. An electronic device, characterized in that, The system includes a memory and a processor, wherein the memory stores a computer program that, when executed by the processor, implements the steps of the reservoir discharge scheduling method according to any one of claims 1 to 7, taking into account the flow distortion in the gate opening and closing transition zone.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the reservoir discharge scheduling method according to any one of claims 1 to 7, taking into account the flow distortion in the gate opening and closing transition zone.

Images