A water quality and quantity joint scheduling method based on a distributed hydrological model

Abstract

The present application relates to the technical field of water environment treatment, and in particular to a water quality and quantity joint scheduling method based on a distributed hydrological model, which comprises the following steps: deploying high-precision monitoring equipment to collect multi-source data, constructing a basic data set, analyzing the temporal and spatial distribution characteristics and dynamic change law of pollutants, designing a multi-layer neural network model for training and optimization, and generating water quality prediction results and scheduling strategies. The present application solves or at least alleviates the problem of the limitation of the ability of the non-point source pollution model in describing the migration path and mechanism of pollutants, and provides a water quality and quantity joint scheduling method based on a distributed hydrological model. Through the fusion of multi-source data and dynamic optimization of the control strategy, the present application realizes accurate prediction and efficient treatment of pollutants under complex hydrological conditions.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of water resource management and water environment governance, and particularly relates to a water quality and quantity joint scheduling method based on a distributed hydrological model. BACKGROUND

[0002] Water quality and quantity joint scheduling is a core link in the management of river basin water environment, especially under the background of increasing non-point source pollution, how to accurately predict and effectively respond to the migration and transformation process of pollutants under complex hydrological conditions has become a key problem to be solved. In recent years, non-point source pollution models, as an important tool for quantitatively characterizing pollution load characteristics and screening the best prevention and control measures, have developed rapidly, mainly including two categories of empirical models and physical models. Empirical models are established through statistical analysis, such as output coefficient method, list analysis method and pollution coefficient method, which are widely used due to their advantages of few parameters, simple structure and acceptable precision. However, such models are difficult to describe the specific path and mechanism of pollutant migration, limiting their further application under complex hydrological conditions. Although physical models can more accurately simulate the migration and transformation process of pollutants, they still face many challenges in practical application due to their high requirements for data quality and computing resources.

[0003] At present, distributed hydrological models are increasingly applied at the basin scale, providing a new technical means for water quality and quantity joint scheduling. However, existing methods still have the following shortcomings: first, traditional models do not accurately depict the temporal and spatial distribution characteristics of rainfall, runoff and pollutant concentration, making it difficult to meet the demand for high-precision prediction; second, the model training process lacks self-adaptive ability and cannot dynamically adjust parameters according to real-time monitoring data, resulting in a large deviation between predicted results and actual situations; third, existing scheduling strategies are mostly based on static rules and are difficult to respond to dynamic changes such as sudden changes in rainfall intensity, abnormal fluctuations in runoff and accelerated diffusion of pollutants, lacking emergency response capability. In addition, the application of multi-source data fusion technology in water quality prediction is not mature, and the integration efficiency of remote sensing images, meteorological data and ground monitoring data is low, affecting the comprehensiveness and accuracy of the prediction results.

[0004] Therefore, it is of great practical significance to develop a water quality and quantity joint scheduling method based on a distributed hydrological model, which can accurately predict the temporal and spatial distribution of pollutant concentration and develop targeted dynamic scheduling strategies. The present application aims to improve the accuracy and timeliness of water quality prediction by constructing a high-precision monitoring system, combining a multi-layer neural network model and a self-adaptive weight adjustment mechanism; at the same time, through multi-source data fusion technology and spectral analysis methods, high-risk areas and periods are identified, and a hierarchical and classified scheduling strategy is developed, thereby comprehensively improving the scientificity and effectiveness of river basin water environment management. SUMMARY

[0005] The present application aims to overcome the deficiencies in the prior art, solve or at least alleviate the problem of the limitation of the ability of the non-point source pollution model in describing the migration path and mechanism of pollutants, and provide a water quality and quantity joint scheduling method based on a distributed hydrological model. The method realizes accurate prediction and efficient control of pollutants under complex hydrological conditions by fusing multi-source data and dynamically optimizing the control strategy.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical scheme: a water quality and quantity joint scheduling method based on a distributed hydrological model, characterized in that it comprises the following steps:

[0007] S1, high-precision monitoring equipment is arranged in the target river basin to synchronously collect rainfall, runoff, pollutant concentration and meteorological data;

[0008] S2, hydrological boundary conditions, water quality parameters and spatial distribution data in different time periods are recorded in the initial stage, and a basic data set is constructed;

[0009] S3, the spatio-temporal distribution characteristics of rainfall intensity on runoff process are obtained through experimental data in the basic data set;

[0010] S4, the dynamic variation law of pollutant concentration in time series is extracted based on the experimental data;

[0011] S5, the spatio-temporal distribution characteristics and the dynamic variation law are comprehensively analyzed, a multi-layer neural network model is designed, and hydrological, meteorological and pollutant concentration data are fused for training;

[0012] S6, the multi-layer neural network model is iteratively optimized according to actual monitoring data to form a prediction model and algorithm;

[0013] S7, the optimized model is applied to water quality prediction to generate a spatio-temporal distribution map of typical reservoir inlet tributaries, and a corresponding scheduling strategy is developed.

[0014] To further implement the present application, the following technical solutions can be preferred:

[0015] Preferably, the step S2 comprises the following steps:

[0016] S201, a set of hydrodynamic equations is constructed, the Navier-Stokes equation under the Cartesian coordinate system is adopted as the basis, and the material transport equation is used to describe the migration and transformation process of pollutants;

[0017] S202, the calculation area is meshed, quadrilateral meshes are used to simulate river bank edges, and triangular meshes are used to simulate reservoir bays, and mesh sizes are set to adapt to the topographic features;

[0018] S203, setting boundary conditions and initial conditions, wherein the boundary conditions include inflow and outflow and pollutant concentration, and the initial conditions include initial water level, flow rate and pollutant distribution.

[0019] Preferably, in the meshing process, the quadrilateral mesh size is set to 200 m x 200 m, which is suitable for the main channel of the river and the regular coastline area; the triangular mesh size is set to 5000 m 2 , which is suitable for the reservoir bay and irregular coastline area; the two kinds of meshes are seamlessly connected by sharing nodes to ensure the overall continuity of the calculation domain;

[0020] The setting of the boundary conditions includes hydrodynamic boundary conditions and water quality boundary conditions, the hydrodynamic boundary conditions are controlled by upstream flow and downstream water level, and the water quality boundary conditions are controlled by upstream pollutant concentration and downstream pollutant flux; the initial conditions are generated by interpolation of historical data, including initial water level, flow field and pollutant concentration field.

[0021] Preferably, the dynamic change rule is obtained by wavelet transform to decompose the time series signal into multi-scale components, and the main frequency amplitude is the amplitude of the low-frequency component of the decomposed signal, which reflects the main trend of the change of the pollutant concentration with time.

[0022] Preferably, in the model training process, an adaptive weight adjustment mechanism is adopted, including calculating the long-term mean and short-term fluctuation of the input data, extracting abnormal features, and judging whether the model weight needs to be updated based on Markov chain; when the abnormal feature coefficient is greater than a set threshold, the model weight adjustment is triggered;

[0023] The adaptive weight adjustment mechanism also includes monitoring the change trend of the pollutant concentration in the time window, and adjusting the model parameters to adapt to the new hydrological conditions when the change trend deviates from the expected range.

[0024] Preferably, the step S7 is specifically:

[0025] In the water quality prediction process, a multi-source data fusion technology is used to integrate remote sensing images, meteorological data and ground monitoring data into the model input layer; the prediction result is output in the form of a space-time distribution map to show the change trend of the pollutant concentration with time and space;

[0026] In the scheduling strategy development process, high-risk areas and periods are identified according to the prediction results, and targeted management measures are proposed; vegetation buffer zones are added in areas with high pollution load, and the operation of pollution interception facilities is strengthened in high flow periods;

[0027] The deviation between the corrected pollutant concentration value and the actual monitoring value is analyzed to determine the abnormal type, and the scheduling strategy is selected according to the abnormal type; the abnormal type includes sudden change of rainfall intensity, abnormal fluctuation of runoff, accelerated diffusion of pollutants and noise interference.

[0028] Among them, the rainfall intensity mutation is that the pollutant concentration shows a rapid rising trend with the change of rainfall intensity, the runoff abnormal fluctuation is that the pollutant concentration appears nonlinear fluctuation with the change of runoff, the pollutant diffusion acceleration is that the pollutant concentration changes significantly in a short time, and the noise interference is that there is a random high-frequency component in the pollutant concentration.

[0029] Preferably, the step S7 comprises the following steps:

[0030] S71, identifying high-risk areas and periods

[0031] Based on the space-time distribution map, the continuous area with pollutant concentration exceeding the warning value by 20% is extracted, the high-risk period with duration exceeding 3 monitoring periods is marked, and the composite risk area in the watershed which simultaneously meets the runoff mutation rate > 15% and the pollutant concentration gradient > 0.3 mg / (L·km) is identified;

[0032] S72, making targeted scheduling strategy

[0033] For high-risk areas: start key monitoring mode, and increase data sampling frequency to 200% of the regular value;

[0034] For the duration period: deploy mobile emergency treatment equipment to the predicted pollution diffusion front 1 hour in advance;

[0035] For the composite risk area: activate the triple defense and control system of pump station, gate and dosing device synchronously;

[0036] S73, execute basic scheduling

[0037] Start level 1 control within 5km upstream of high-risk area: close leisure water intake, and limit shipping passage;

[0038] Implement pre-scheduling 2 hours before high-risk period: reduce reservoir water level to 90% of design value;

[0039] Set up emergency buffer zone 10km downstream of composite risk area: deploy adsorption barrier and online monitoring buoy.

[0040] Preferably, the step S7 further comprises the following steps:

[0041] S74, judge abnormal type

[0042] Analyze the spectral characteristics of the monitoring data:

[0043] Rainfall mutation type: 0.5-2Hz low frequency band energy ratio > 40%;

[0044] Runoff abnormal type: 2-5Hz mid-frequency band appears continuous oscillation waveform;

[0045] Pollution diffusion type: pulse cluster characteristics exist in the high frequency band of 5-10 Hz;

[0046] S75, dynamically adjust the strategy parameters

[0047] When the rainfall suddenly changes: for every increase of 10 mm / h of rainfall, the pump station power is increased by 15%;

[0048] When the runoff is abnormal: for every 5% deviation of the flow rate from the baseline value, the gate opening is adjusted by 3%;

[0049] When pollution diffuses: for every 0.1 mg / (L·h) increase in concentration gradient, the dosing interval is shortened by 20%.

[0050] Preferably, when the parameter adjustment of step S75 is completed, if at least one device parameter change rate > 15%, the following step is immediately executed:

[0051] S76, strategy optimization control

[0052] When rainfall and runoff anomalies occur simultaneously, the gate adjustment is started 5 minutes after the pump station power is increased to 120%;

[0053] When runoff and pollution anomalies occur simultaneously, the automatic dosing device is activated immediately when the flow rate stabilizes within ±5%; when three anomalies occur simultaneously, the pump station drainage is started after 60% of the first round of drug dosing is completed;

[0054] S77, transition control start

[0055] When the rainfall mutation is removed and the pump station power falls below 110%, the power gradually decreases program is triggered;

[0056] When the runoff fluctuation stabilizes within ±5% for 30 minutes, the gate reset program is triggered;

[0057] When the pollution concentration is below the warning value for 2 consecutive hours, the monitoring frequency reduction program is triggered.

[0058] Preferably, when the transition control of step S77 is completed, if the device parameter fluctuation rate < 5% for 1 hour, the following step is executed:

[0059] S78, scene-based aftercare:

[0060] After the treatment at the reservoir inlet is completed, the diversion channel is kept open until the water turbidity decreases by 50%;

[0061] During the treatment at the tributary junction, the dissolved oxygen content downstream is detected after each purification cycle is completed;

[0062] After the treatment at the dam front area is completed, the ultrafiltration device is maintained at 50% power for 12 hours;

[0063] S79, effect verification modeling:

[0064] Rainfall treatment verification: calculate the energy consumption ratio of unit rainfall pumping, optimize the combination scheme of pump stations;

[0065] Runoff treatment verification: analyze the correlation between gate adjustment frequency and flow stability;

[0066] Pollution treatment verification: establish a dose-effect model of reagent dosage and concentration reduction.

[0067] The beneficial effects of the present application are:

[0068] The present application realizes the accurate description of the migration path of pollutants under complex hydrological conditions by constructing a distributed hydrological model to fuse multi-source data, solves the problem that the empirical model is difficult to describe the migration mechanism of pollutants, and improves the applicability and prediction accuracy of the model.

[0069] At the same time, the present application introduces an adaptive weight adjustment mechanism and a dynamic optimization scheduling strategy, which significantly enhances the response ability of the model to abnormal conditions, can quickly adjust the treatment measures under complex scenes such as rainfall mutation and runoff anomaly, and guarantees the safety of the water environment of the watershed.

[0070] In addition, the multi-layer neural network model combined with wavelet transform technology proposed by the present application can effectively extract the dynamic change rule of pollutant concentration, and provides reliable technical support for formulating a scientific and reasonable water quality and water quantity joint scheduling strategy. BRIEF DESCRIPTION OF DRAWINGS

[0071] Figure 1 is the overall flowchart of the present application.

[0072] Figure 2 is the detailed flowchart of step S2 of the present application.

[0073] Figure 3 is the step flowchart of step S7 of the present application.

[0074] Figure 4 is the grid division schematic diagram of the present application.

[0075] Figure 5 is the adaptive weight adjustment mechanism flowchart of the present application.

[0076] Figure 6 is the abnormal type judgment flowchart of the present application.

[0077] Figure 7 is the strategy optimization control flowchart of the present application.

[0078] Figure 8 is the transition control start flowchart of the present application.

[0079] Figure 9 Scenario-based post-disaster handling flowchart of the present application. DETAILED DESCRIPTION

[0080] In the description of the present application, it should be further pointed out that, unless otherwise explicitly specified and limited, the terms "arrangement", "installation", "connection", "connection" should be broadly understood, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be connected inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0081] The technical solutions in the embodiments of the present application will be described clearly and completely in the following with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0082] Embodiment one

[0083] Currently, distributed hydrological models are increasingly applied at the basin scale, providing new technical means for water quality and quantity joint scheduling. However, the existing methods still have the following shortcomings: First, the traditional model does not describe the temporal and spatial distribution characteristics of rainfall, runoff and pollutant concentration in detail, which is difficult to meet the demand of high-precision prediction; second, the model lacks self-adaptive ability during the training process, and cannot dynamically adjust the parameters according to real-time monitoring data, resulting in a large deviation between the prediction results and the actual situation; third, the existing scheduling strategy is mostly based on static rules, which is difficult to cope with dynamic change scenarios such as sudden change of rainfall intensity, abnormal fluctuation of runoff and accelerated diffusion of pollutants, and lacks emergency response capability. In addition, the application of multi-source data fusion technology in water quality prediction is not mature, and the integration efficiency of remote sensing images, meteorological data and ground monitoring data is low, which affects the comprehensiveness and accuracy of the prediction results.

[0084] Reference Figure 1 The embodiment discloses a water quality and quantity joint scheduling method based on a distributed hydrological model, comprising the following steps:

[0085] S1, high-precision monitoring equipment is arranged in the target basin for synchronous acquisition of rainfall, runoff, pollutant concentration and meteorological data;

[0086] S2, record the hydrological boundary conditions, water quality parameters and spatial distribution data in different time periods in the initial stage, and construct a basic data set;

[0087] S3, acquire the time and space distribution characteristics of rainfall intensity on runoff process through the experimental data in the basic data set;

[0088] S4, extract the dynamic change rule of pollutant concentration in time series based on the experimental data;

[0089] S5, comprehensively analyze the time and space distribution characteristics and the dynamic change rule, design a multi-layer neural network model, and train the model by fusing hydrological, meteorological and pollutant concentration data;

[0090] S6, iteratively optimize the multi-layer neural network model according to actual monitoring data, form a prediction model and algorithm;

[0091] S7, apply the optimized model to water quality prediction, generate a time and space distribution map of a typical reservoir inlet tributary, and develop a corresponding scheduling strategy.

[0092] Referring to Figure 2 , the step S2 comprises the following steps:

[0093] S201, construct a set of hydrodynamic equations, adopt Navier-Stokes equations in Cartesian coordinates as the basis, and combine a material transport equation to describe the migration and transformation process of pollutants;

[0094] S202, divide the calculation region into grids, adopt quadrilateral grids to simulate the river bank, adopt triangular grids to simulate the reservoir bay, and set the grid size to adapt to the topographic features;

[0095] S203, set boundary conditions and initial conditions, wherein the boundary conditions include inflow and outflow and pollutant concentration, and the initial conditions include initial water level, flow rate and pollutant distribution.

[0096] In step S201, the two-dimensional hydrodynamic control equation is a set of Navier-Stokes equations in Cartesian coordinates, which consists of a water flow continuity equation, a momentum equation along the water flow direction (x direction) and a momentum equation perpendicular to the water flow direction (y direction).

[0097]

[0098] In the above formulas, η is the water surface elevation, h is the total water depth, g is the acceleration of gravity, ρ is the density of water, ρ0 is the reference density of fresh water, f = 2Ωsinφ is the Coriolis force coefficient (Ω is the rotation angular velocity, φ is the geographic latitude), P a is the atmospheric pressure, s ij is the radiation stress tensor. S and (u s , v sQ and V, respectively, are the discharge and velocity of the point source. and is the average value of the flow velocity over the depth, defined as:

[0099]

[0100] (τ sx ,τ sy ) and (τ bx ,τ by ) are the wind stress tensor at the water surface and the bed stress tensor at the river bed, respectively. The bed stress can be determined by the quadratic law of resistance (frictional resistance is proportional to the square of the flow velocity):

[0101]

[0102] c f is the drag coefficient or the bed friction, is the average flow velocity over the depth of the water at the river bed, and the frictional velocity at the river bed is The bed friction can be estimated by the Chezy number C or the Manning number M:

[0103]

[0104] The unit of the Chezy number is m 1 / 2 / s, and the unit of the Manning number is m 1 / 3 / s. The Manning number and the bed roughness height (roughness) k s are related as follows:

[0105]

[0106] The value of the Manning number is generally between 20 and 40 m 1 / 3 / s.

[0107] (τ sx ,τ sy ) is the wind stress tensor at the water surface, and the wind stress can be obtained by the following empirical formula

[0108]

[0109] where p a is the air density, c d is the air drag coefficient, is the wind speed 10 m above the water surface. The friction velocity due to the wind stress can be expressed as

[0110]

[0111] T ijFor lateral stresses, including viscous friction, turbulent friction and differential convection, they can be estimated by the eddy viscosity coefficient formula based on the average flow velocity gradient of water depth:

[0112]

[0113] where A is the horizontal eddy viscosity coefficient.

[0114] According to the Kolmogrov and Prandtle theory, the turbulent eddy viscosity coefficient v τ is proportional to the square root of the turbulent kinetic energy k and the characteristic eddy viscosity scale l. If the dissipation scale is taken as l (the dissipation rate ε = κ 3 / 2 / l), the eddy viscosity coefficient expression represented by κ and ε can be obtained:

[0115]

[0116] where c μ is an empirical constant.

[0117] The logarithmic rate eddy viscosity coefficient can be calculated by

[0118]

[0119] where U τ = max(U τs , U τb ), c1 and c2 are constants, and when c1 = 0.41 and c2 = -0.41, the expression is a standard parabola.

[0120] Smagorinsky proposed a formula in 1996 that relates the effective eddy viscosity coefficient at the sub-grid scale to the characteristic length:

[0121]

[0122] c s is called the "Smagorinsky constant", l represents the characteristic length, and the deformation rate is defined as:

[0123]

[0124] According to the law of conservation of mass, considering the factors such as convection, diffusion and degradation in the process of pollutant transport, the transport equation of pollutants can be written as

[0125]

[0126] where c is the concentration of pollutants; (D x , D y ) is the diffusion coefficient in x and y directions. K dThe linear degradation coefficient of the pollutant, i.e., the degradation of the pollutant, conforms to a first-order reaction formula:

[0127]

[0128] With reference to Figure 4 , in the meshing process, the size of the quadrilateral mesh is set to 200m*200m, which is suitable for the main stream of the river and the regular coastline area; the size of the triangular mesh is set to 5000m 2 , which is suitable for the reservoir bay and irregular coastline area; the two kinds of meshes are seamlessly connected by sharing nodes to ensure the overall continuity of the calculation domain;

[0129] The setting of the boundary conditions includes the hydrodynamic boundary condition and the water quality boundary condition; the hydrodynamic boundary condition is controlled by the upstream flow and the downstream water level, and the water quality boundary condition is controlled by the upstream pollutant concentration and the downstream pollutant flux; the initial conditions are generated by interpolation of historical data, including the initial water level, flow velocity field and pollutant concentration field.

[0130] The dynamic change law decomposes the time series signal into multi-scale components through wavelet transform, and the main frequency amplitude is the amplitude of the low-frequency component in the decomposed signal, which reflects the main trend of the change of the pollutant concentration with time.

[0131] With reference to Figure 5 , in the model training process, an adaptive weight adjustment mechanism is adopted, including calculating the long-term mean and short-term fluctuation of the input data, extracting abnormal features, judging whether the model weight needs to be updated based on Markov chain, and triggering the model weight adjustment when the abnormal feature coefficient is greater than the set threshold;

[0132] The adaptive weight adjustment mechanism further includes monitoring the change trend of the pollutant concentration in the time window, and adjusting the model parameters to adapt to the new hydrological conditions when the change trend deviates from the expected range.

[0133] The step S7 specifically includes:

[0134] In the water quality prediction process, a multi-source data fusion technology is adopted to integrate remote sensing image meteorological data and ground monitoring data into the model input layer; the prediction result is output in the form of a space-time distribution map to show the change trend of the pollutant concentration with time and space;

[0135] In the scheduling strategy formulation process, the high-risk areas and periods are identified according to the prediction results, and targeted control measures are proposed; vegetation buffer belts are added in areas with high pollution load, and the operation of pollution interception facilities is strengthened during periods with high flow;

[0136] With reference to Figure 6, analyze the deviation between the corrected pollutant concentration value and the actual monitoring value, determine the abnormal type, and select the scheduling strategy according to the abnormal type, the abnormal type including rainfall intensity mutation, runoff abnormal fluctuation, pollutant diffusion acceleration and noise interference;

[0137] Wherein, the rainfall intensity mutation is that the pollutant concentration shows a rapid rising trend with the change of rainfall intensity, the runoff abnormal fluctuation is that the pollutant concentration appears nonlinear fluctuation with the change of runoff, the pollutant diffusion acceleration is that the pollutant concentration changes significantly in a short time, and the noise interference is that there is a random high-frequency component in the pollutant concentration.

[0138] With reference to Figure 3 , Figure 7 , Figure 8 And Figure 9 , the step S7 comprises the following steps:

[0139] S71, identifying high-risk areas and periods

[0140] Based on the space-time distribution map, the continuous area where the pollutant concentration exceeds the warning value by 20% is extracted, the high-risk period with a duration of more than 3 monitoring periods is marked, and the composite risk area in the watershed which simultaneously satisfies the runoff mutation rate > 15% and the pollutant concentration gradient > 0.3 mg / (L·km) is identified;

[0141] S72, making targeted scheduling strategy

[0142] For high-risk areas: start the key monitoring mode, and increase the data sampling frequency to 200% of the regular value;

[0143] For the duration: deploy mobile emergency treatment equipment to the predicted pollution diffusion front 1 hour in advance;

[0144] For the composite risk area: activate the three-in-one prevention and control system of pump station, gate and dosing device synchronously;

[0145] S73, execute basic scheduling

[0146] Start level one prevention and control within 5km upstream of the high-risk area: close the leisure water intake, and limit the navigation passage;

[0147] Implement pre-scheduling 2 hours before the high-risk period: reduce the reservoir water level to 90% of the design value;

[0148] Set up an emergency buffer zone 10km downstream of the composite risk area: deploy adsorption barriers and online monitoring buoys.

[0149] The step S7 further comprises the following steps:

[0150] S74, judging abnormal type

[0151] Analysis of monitoring data spectrum characteristics:

[0152] Rainfall mutation type: 0.5-2Hz low frequency band energy ratio >40%;

[0153] Abnormal runoff type: 2-5Hz mid-frequency band appears continuous oscillation waveform;

[0154] Pollution diffusion type: 5-10Hz high frequency band exists pulse cluster characteristics;

[0155] S75, dynamic adjustment of strategy parameters

[0156] When the rainfall mutates: for every increase of 10mm / h of rainfall, the pump station power is increased by 15%;

[0157] When the runoff is abnormal: for every 5% deviation of the flow rate from the baseline value, the gate opening is adjusted by 3%;

[0158] When the pollution diffuses: for every 0.1mg / (L·h) increase in concentration gradient, the dosing interval is shortened by 20%.

[0159] After completing the parameter adjustment of step S75, if at least one device parameter change rate >15%, immediately execute the following steps:

[0160] S76, strategy optimization control

[0161] When rainfall and runoff anomalies occur simultaneously, after the pump station power is increased to 120%, the gate adjustment is started after a delay of 5 minutes;

[0162] When runoff and pollution anomalies occur simultaneously, activate the automatic dosing device immediately when the flow rate stabilizes within ±5%; when three anomalies occur simultaneously, start the pump station drainage after completing 60% of the first round of drug dosing;

[0163] S77, transition control start

[0164] When the rainfall mutation is removed and the pump station power falls below 110%, trigger the power gradual reduction program;

[0165] When the runoff fluctuation stabilizes within ±5% for 30 minutes, trigger the gate reset program;

[0166] When the pollution concentration is continuously below the warning value for 2 hours, trigger the monitoring frequency reduction program.

[0167] After completing the transition control of step S77, if the device parameter fluctuation rate <5% for 1 hour, execute the following steps:

[0168] S78, scene-based aftercare:

[0169] After the disposal at the reservoir entrance is completed, keep the diversion channel open until the water turbidity decreases by 50%.

[0170] During the branch confluence disposal, the downstream dissolved oxygen content is detected after each purification cycle is completed;

[0171] After the disposal in the dam front area is completed, the ultrafiltration device is maintained at 50% power for 12 hours;

[0172] S79, effect verification modeling:

[0173] Rainfall disposal verification: calculate the energy consumption ratio of unit rainfall pumping and drainage, and optimize the pump station combination scheme;

[0174] Runoff disposal verification: analyze the correlation between gate adjustment times and flow velocity stability;

[0175] Pollution disposal verification: establish a dose-effect model of reagent dosage and concentration reduction.

[0176] The present application realizes the accurate description of the migration path of pollutants under complex hydrological conditions by constructing a distributed hydrological model to fuse multi-source data, solves the problem that the empirical model is difficult to describe the migration mechanism of pollutants, and improves the applicability and prediction accuracy of the model.

[0177] At the same time, the present application significantly enhances the response ability of the model to abnormal conditions by introducing an adaptive weight adjustment mechanism and a dynamic optimization scheduling strategy, can quickly adjust the treatment measures under complex scenes such as rainfall mutation and runoff anomaly, and guarantees the safety of the water environment of the watershed.

[0178] Embodiment two

[0179] In order to better enable relevant persons in the art to fully understand and implement the present application, the specific implementation principles of the present application are supplemented below in conjunction with a specific application scenario.

[0180] The present application provides a water quality and quantity joint scheduling method based on a distributed hydrological model, which is characterized by realizing accurate prediction and efficient treatment of the migration path and concentration change of pollutants under complex hydrological conditions through multi-source data fusion and dynamic optimization control strategy.

[0181] Laying out high-precision monitoring equipment within the target watershed is the first step of this method. These devices include rain gauges, flow meters, water quality sensors, and weather stations, etc., for synchronous collection of rainfall, runoff, pollutant concentration, and meteorological data. The distribution of these devices should cover key nodes throughout the watershed, such as the main river, tributary junctions, reservoir bays, and key areas susceptible to pollution. For example, in a typical watershed application, a set of rain gauges and flow meters are placed every 5 kilometers along the main river, while water quality sensors are added near the tributary inlet to the reservoir to capture changes in pollutant concentration. All monitoring equipment transmits data in real-time to the central processing system through a wireless communication network, which is responsible for preliminary cleaning and arrangement of the data to ensure the accuracy of subsequent analysis.

[0182] In the initial stage, hydrological boundary conditions, water quality parameters, and spatial distribution data over different time periods are recorded, and a basic data set is constructed. This process is based on historical data combined with experimental data obtained from field sampling. For example, rainfall, runoff, and pollutant concentration data from the past three years are selected as the baseline, and an initial water level field, flow field, and pollutant concentration field are generated through interpolation. Subsequently, a hydrodynamic model is established based on the Navier-Stokes equation in the Cartesian coordinate system, and a material transport equation is used to describe the migration and transformation process of pollutants. The calculation region uses grid division technology, with a quadrilateral grid size of 200 meters x 200 meters for the main river and regular shoreline areas, and a triangular grid size of 5000 square meters for reservoir bays and irregular shoreline areas. The two types of grids are seamlessly connected by sharing nodes to ensure the overall continuity of the calculation domain. The boundary conditions include hydrodynamic boundary conditions and water quality boundary conditions, with the former controlled by upstream flow and downstream water level, and the latter controlled by upstream pollutant concentration and downstream pollutant flux.

[0183] Extracting the spatial and temporal distribution characteristics of rainfall intensity on runoff processes through experimental data in the basic data set is an important part of this method. Specifically, statistical analysis tools are used to analyze the correlation between collected rainfall and runoff data, and the relationship between rainfall intensity and runoff generation is extracted. For example, by plotting the scatter plot of rainfall intensity and runoff coefficient, it is found that they show a nonlinear growth trend, and further fitting gives the empirical formula Q = aP^b + c, where Q is the runoff, P is the rainfall intensity, a, b, and c are fitting parameters. This formula can effectively reflect the influence of rainfall intensity on the runoff process, providing an important basis for subsequent model training.

[0184] Based on experimental data, the dynamic change rule of pollutant concentration in time series is extracted, which is one of the core technologies of this method. In order to improve the prediction accuracy, wavelet transform technology is used to decompose the time series signal of pollutant concentration into multi-scale components, and the main frequency amplitude of low frequency component is extracted as the main trend index. For example, the monitoring data of a certain basin shows that the pollutant concentration presents periodic fluctuation in rainy season, and after wavelet transform decomposition, it is found that the amplitude of low frequency component is significantly higher than that of high frequency component, indicating that the main change trend of pollutant concentration is driven by long-term rainfall pattern. Based on this, the dynamic change model of pollutant concentration C(t) = A sin(ωt + φ) + B can be constructed, where C(t) is the change of pollutant concentration with time, A is the main frequency amplitude, ω is the angular frequency, φ is the phase angle, and B is the baseline concentration. This model can accurately describe the dynamic change rule of pollutant concentration, laying a foundation for subsequent model training.

[0185] The spatio-temporal distribution characteristics and dynamic change rule are analyzed comprehensively, and a multi-layer neural network model is designed to integrate hydrological, meteorological and pollutant concentration data for training. Specifically, the multi-layer neural network model includes input layer, hidden layer and output layer, among which the input layer receives rainfall, runoff, pollutant concentration and meteorological data, the hidden layer adopts multi-layer perceptron structure to extract features, and the output layer predicts the spatio-temporal distribution of pollutant concentration. During model training, an adaptive weight adjustment mechanism is introduced to extract abnormal features by calculating the long-term mean and short-term fluctuation of input data, and to judge whether the model weight needs to be updated based on Markov chain. For example, when the abnormal feature coefficient is greater than the set threshold, the model weight adjustment is triggered, and at the same time the change trend of pollutant concentration in the monitoring time window is monitored, and when the change trend deviates from the expected range, the model parameters are adjusted to adapt to the new hydrological conditions. In addition, cross-validation technique is used to evaluate the prediction performance during model training to ensure good generalization ability of the model.

[0186] According to the actual monitoring data, the multi-layer neural network model is iteratively optimized, which is one of the key steps of this method. Specifically, the predicted results of the model are compared with the actual monitoring values, the errors are calculated and the model parameters are adjusted. For example, the prediction result shows that the pollutant concentration in a certain period is significantly lower than the actual monitoring value, and through analysis it is found that the rainfall intensity in this period suddenly changes, causing abnormal fluctuation of runoff, and then affecting the distribution of pollutant concentration. For such cases, Bayesian optimization algorithm is used to adjust the model hyperparameters, and the number of training data samples is increased to improve the response ability of the model to abnormal conditions. After several iterations of optimization, the prediction model and algorithm are finally formed, with prediction accuracy of more than 90%.

[0187] The optimized model is applied to water quality prediction, generating a spatiotemporal distribution map of typical tributaries entering the reservoir, and formulating corresponding scheduling strategies. Specifically, multi-source data fusion technology is used to integrate remote sensing images, meteorological data, and ground monitoring data into the model input layer. The prediction results are output in the form of a spatiotemporal distribution map, showing the trend of pollutant concentration changes over time and space. For example, a prediction result shows that the pollutant concentration near the mouth of a tributary entering the reservoir exceeds the warning value by 20% and lasts for more than 3 monitoring periods, identifying it as a high-risk area. For such high-risk areas, the key monitoring mode is started, the data sampling frequency is increased to 200% of the regular value, and mobile emergency treatment equipment is deployed to the predicted pollution diffusion front 1 hour in advance. For composite risk areas, i.e., areas that simultaneously meet the conditions of runoff mutation rate greater than 15% and pollutant concentration gradient greater than 0.3 milligrams per liter per kilometer, the three-in-one prevention and control system of pump stations, gates, and dosing devices is activated simultaneously.

[0188] In the scheduling strategy formulation process, the spectral characteristics of the monitoring data are further analyzed to determine the type of anomaly and dynamically adjust the strategy parameters. For example, the rainfall mutation type is characterized by a low-frequency energy proportion greater than 40% in the 0.5 to 2 Hz band, the runoff anomaly type is characterized by a sustained oscillation waveform in the 2 to 5 Hz band, and the pollution diffusion type is characterized by pulse cluster characteristics in the 5 to 10 Hz band. Strategy parameters are adjusted according to the type of anomaly. For every 10 millimeters per hour of rainfall, the pump station power is increased by 15%, for every 5% deviation of flow rate from the baseline value, the gate opening is adjusted by 3%, and for every 0.1 milligrams per liter per hour increase in concentration gradient, the dosing interval is shortened by 20%. For example, a rainfall mutation caused the pump station power to increase to 120%, and after a 5-minute delay, the gate adjustment was started. The automatic dosing device was activated immediately when the flow rate stabilized within ±5%, and the pump station drainage was started after 60% of the first round of drug dosing was completed. When the rainfall mutation is removed and the pump station power falls below 110%, the power gradually decreases; when the runoff fluctuation stabilizes within ±5% for 30 minutes, the gate reset program is triggered; and when the pollution concentration is below the warning value for 2 consecutive hours, the monitoring frequency reduction program is triggered.

[0189] After the transition control is completed, if the device parameter fluctuation rate is less than 5% for 1 hour, subsequent operations are performed. For example, after the treatment at the entrance of the reservoir area is completed, the diversion channel is kept open until the water flow turbidity decreases by 50%; during the treatment at the confluence of the tributaries, the downstream dissolved oxygen content is detected after each purification cycle is completed; after the treatment at the dam area is completed, the ultrafiltration device is maintained at 50% power for 12 hours. In addition, by calculating the energy consumption ratio of unit rainfall pumping, the correlation between the gate adjustment frequency and the flow velocity stability is analyzed, and the dose-effect model of the reagent dosage and the concentration reduction amplitude is established. For example, the experimental results show that the energy consumption ratio of unit rainfall pumping is 0.8 kilowatt-hour per millimeter when it is optimal, the gate adjustment frequency and the flow velocity stability have a negative correlation, and the dose-effect model of the reagent dosage and the concentration reduction amplitude is ΔC=k·log(M), where ΔC is the concentration reduction amplitude, M is the reagent dosage, and k is the proportional coefficient.

[0190] To sum up, the present application realizes the accurate description of the migration path of pollutants under complex hydrological conditions by constructing a distributed hydrological model to fuse multi-source data, solves the problem that the empirical model is difficult to describe the migration mechanism of pollutants, and improves the applicability and prediction accuracy of the model. By introducing the self-adaptive weight adjustment mechanism and the dynamic optimization scheduling strategy, the response ability of the model to abnormal conditions is significantly enhanced, and the treatment measures can be quickly adjusted under complex scenes such as rainfall mutation and runoff anomaly, thereby ensuring the safety of the water environment of the watershed. The multi-layer neural network model combined with the wavelet transform technology can effectively extract the dynamic change law of the pollutant concentration, and provides reliable technical support for formulating a scientific and reasonable water quality and water quantity joint scheduling strategy.

[0191] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any skilled person in the art can make equivalent replacement or change according to the technical solution and the inventive concept of the present application within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A water quality and quantity joint scheduling method based on a distributed hydrological model, characterized in that, Comprise the following steps: S1, high-precision monitoring equipment is arranged in the target basin, for synchronous acquisition of rainfall, runoff, pollutant concentration and meteorological data; S2, record the hydrological boundary conditions, water quality parameters and spatial distribution data in different time periods in the initial stage, and build a basic data set; S3, obtain the time and space distribution characteristics of rainfall intensity on runoff process through experimental data in the basic data set; S4, based on the experimental data, the time series signal of pollutant concentration is decomposed into multi-scale components by wavelet transform, and the dynamic change law reflecting the main trend of the change of pollutant concentration with time is extracted; S5, the time and space distribution characteristics and the dynamic change law are comprehensively analyzed, a multi-layer neural network model is designed, and hydrological, meteorological and pollutant concentration data are fused for training; S6, according to the actual monitoring data, the adaptive weight adjustment mechanism is used to iteratively optimize the multi-layer neural network model to form a prediction model, wherein the adaptive weight adjustment mechanism comprises: calculating the long-term mean and short-term fluctuation of the input data, extracting the abnormal features, and judging whether the model weight needs to be updated based on Markov chain, and triggering the model weight adjustment when the abnormal feature coefficient is greater than the set threshold; S7, the optimized model is applied to water quality prediction, the time and space distribution graph of the typical reservoir inlet branch is generated, and the corresponding scheduling strategy is made; in the water quality prediction process, multi-source data fusion technology is used to integrate remote sensing image meteorological data and ground monitoring data to the model input layer; the prediction result is output in the form of time and space distribution graph, which shows the change trend of pollutant concentration with time and space; in the scheduling strategy making process, the high-risk area and period are identified according to the prediction result, and the targeted control measures are put forward. 2.The water quality and quantity joint scheduling method based on a distributed hydrological model of claim 1, wherein, The step S2 comprises the following steps: S201, construct the hydrodynamic equation set, use the Navier-Stokes equation in Cartesian coordinate system as the basis, and combine the material transport equation to describe the migration and transformation process of pollutants; S202, grid division is carried out on the calculation region, quadrilateral grid is used to simulate the river bank, triangular grid is used to simulate the reservoir bay, and grid size is set respectively to adapt to the terrain characteristics; S203, set boundary conditions and initial conditions, wherein the boundary conditions include inflow and outflow and pollutant concentration, and the initial conditions include initial water level, flow rate and pollutant distribution.

3. The water quality and quantity joint scheduling method based on a distributed hydrological model according to claim 2, characterized in that, In the grid division process, the size of quadrilateral grid is set to 200 m*200 m, which is suitable for the main stream of river and regular shoreline area; the size of triangular grid is set to 5000 m², which is suitable for reservoir bay and irregular shoreline area; the two kinds of grids are seamlessly connected through shared nodes to ensure the overall continuity of the calculation domain; The setting of boundary conditions includes hydrodynamic boundary conditions and water quality boundary conditions, the hydrodynamic boundary conditions are controlled by upstream flow and downstream water level, and the water quality boundary conditions are controlled by upstream pollutant concentration and downstream pollutant flux; the initial conditions are generated by interpolation of historical data, including initial water level, flow field and pollutant concentration field.

4. The water quality and quantity joint scheduling method based on a distributed hydrological model according to claim 1, characterized in that, The step S7 is specifically: In the water quality prediction process, multi-source data fusion technology is used to integrate remote sensing image meteorological data and ground monitoring data into the model input layer; the prediction result is output in the form of a space-time distribution map, showing the trend of pollutant concentration change over time and space; In the scheduling strategy formulation process, high-risk areas and time periods are identified based on the prediction results, and targeted control measures are proposed; Increase vegetation buffer zones in areas with high pollution loads and strengthen the operation of sewage interception facilities during high flow periods; Analyze the deviation between the corrected pollutant concentration value and the actual monitoring value, determine the type of anomaly, and select a scheduling strategy based on the type of anomaly. The types of anomalies include sudden changes in rainfall intensity, abnormal fluctuations in runoff, accelerated pollutant diffusion, and noise interference. Wherein, sudden change of rainfall intensity is manifested as rapid rise of pollutant concentration with rainfall intensity, abnormal fluctuation of runoff is manifested as nonlinear fluctuation of pollutant concentration with runoff, accelerated diffusion of pollutants is manifested as significant change of pollutant concentration in a short time, and noise interference is manifested as random high-frequency component in pollutant concentration.

5. The water quality and quantity joint scheduling method based on a distributed hydrological model according to claim 4, characterized in that, The step S7 includes the following steps: S71, identify high-risk areas and time periods Extract continuous areas where pollutant concentration exceeds 20% of the warning value based on the space-time distribution map, mark high-risk time periods with a duration of more than 3 monitoring periods, and identify composite risk areas that meet both runoff mutation rate > 15% and pollutant concentration gradient > 0.3 mg / (L·km) in the watershed; S72, formulate targeted scheduling strategy For high-risk areas: start the key monitoring mode, and increase the data sampling frequency to 200% of the regular value; For continuous time periods: deploy mobile emergency treatment equipment to the predicted pollution diffusion front 1 hour in advance; For composite risk areas: activate the three-in-one prevention and control system of pump station, gate, and dosing device simultaneously; S73, execute basic scheduling Start level 1 control within 5 km upstream of the high-risk area: close leisure water intake and restrict shipping passage; Implement pre-scheduling 2 hours before the high-risk period: lower the reservoir water level to 90% of the design value; Set up an emergency buffer zone 10 km downstream of the composite risk area: deploy adsorption barriers and online monitoring buoys.

6. The water quality and quantity joint scheduling method based on a distributed hydrological model according to claim 5, characterized in that, The step S7 also includes the following steps: S74, determine the type of anomaly Analyze the spectral characteristics of the monitoring data: Rainfall mutation type: 0.5-2Hz low frequency band energy ratio > 40%; Runoff anomaly type: 2-5Hz mid-frequency band appears continuous oscillation waveform; Pollution diffusion type: 5-10Hz high frequency band has pulse cluster characteristics; S75, dynamically adjust strategy parameters When rainfall mutates: for every 10mm / h increase in rainfall, increase pump station power by 15%; When runoff is abnormal: for every 5% deviation of flow rate from the baseline value, adjust the gate opening by 3%; When pollution diffuses: for every 0.1 mg / (L·h) increase in concentration gradient, shorten the dosing interval by 20%.

7. The water quality and quantity joint scheduling method based on a distributed hydrological model according to claim 6, characterized in that, After completing the parameter adjustment in step S75, if at least one device parameter changes by > 15%, execute the following step immediately: S76, strategy optimization control When rainfall and runoff anomalies occur simultaneously, delay 5 minutes to start gate adjustment after pump station power is increased to 120%. When the runoff is concurrent with pollution anomaly, activate the automatic dosing device immediately when the flow rate is stable within ±5% range; When the triple anomaly is concurrent, start the pump station drainage after 60% of the first round of agent is put into the water; S77, transition control start When the rainfall mutation is removed and the pump station power falls below 110%, trigger the power ramp-down program; When the runoff fluctuation is stable within ±5% range for 30 minutes, trigger the gate reset program; When the pollution concentration is below the warning value for 2 hours, trigger the monitoring frequency reduction program.

8. The water quality and quantity joint scheduling method based on a distributed hydrological model according to claim 7, characterized in that, When the transition control in step S77 is completed, if the device parameter fluctuation rate is <5% for 1 hour, execute the following steps: S78, scenario-based post-disposal treatment: After the disposal at the reservoir inlet is completed, keep the diversion channel open until the water turbidity drops by 50%; During the disposal at the tributary junction, detect the downstream dissolved oxygen content after each purification cycle is completed; After the disposal at the dam front area is completed, maintain the ultrafiltration device at 50% power for 12 hours; S79, effect verification modeling: Rainfall disposal verification: calculate the energy consumption ratio per unit of rainfall for pumping and drainage, and optimize the pump station combination scheme; Runoff disposal verification: analyze the correlation between the number of gate adjustments and the flow rate stability; Pollution disposal verification: establish a dose-effect model of the agent dosage and the concentration reduction.

Images