CN116090611A - Remote intelligent optimal control method for primary rain interception facility - Google Patents
Remote intelligent optimal control method for primary rain interception facility Download PDFInfo
- Publication number
- CN116090611A CN116090611A CN202211644380.3A CN202211644380A CN116090611A CN 116090611 A CN116090611 A CN 116090611A CN 202211644380 A CN202211644380 A CN 202211644380A CN 116090611 A CN116090611 A CN 116090611A
- Authority
- CN
- China
- Prior art keywords
- rainfall
- water quality
- model
- data
- standard
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 62
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 122
- 238000004088 simulation Methods 0.000 claims abstract description 36
- 230000008569 process Effects 0.000 claims abstract description 31
- 238000013135 deep learning Methods 0.000 claims abstract description 28
- 239000003344 environmental pollutant Substances 0.000 claims abstract description 20
- 231100000719 pollutant Toxicity 0.000 claims abstract description 20
- 230000000694 effects Effects 0.000 claims description 19
- 238000012549 training Methods 0.000 claims description 18
- 238000009825 accumulation Methods 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 14
- 238000012795 verification Methods 0.000 claims description 9
- 238000009991 scouring Methods 0.000 claims description 8
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 claims description 6
- 238000011010 flushing procedure Methods 0.000 claims description 6
- 230000007246 mechanism Effects 0.000 claims description 6
- 238000013136 deep learning model Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims description 4
- 238000011156 evaluation Methods 0.000 claims description 4
- 230000008595 infiltration Effects 0.000 claims description 4
- 238000001764 infiltration Methods 0.000 claims description 4
- 229910052760 oxygen Inorganic materials 0.000 claims description 4
- 239000001301 oxygen Substances 0.000 claims description 4
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 claims description 3
- XKMRRTOUMJRJIA-UHFFFAOYSA-N ammonia nh3 Chemical compound N.N XKMRRTOUMJRJIA-UHFFFAOYSA-N 0.000 claims description 3
- 229910052757 nitrogen Inorganic materials 0.000 claims description 3
- 229910052698 phosphorus Inorganic materials 0.000 claims description 3
- 239000011574 phosphorus Substances 0.000 claims description 3
- 230000000452 restraining effect Effects 0.000 claims description 3
- 239000007787 solid Substances 0.000 claims description 3
- 238000003860 storage Methods 0.000 claims description 3
- 239000000126 substance Substances 0.000 claims description 3
- 230000001970 hydrokinetic effect Effects 0.000 claims description 2
- 238000013508 migration Methods 0.000 claims description 2
- 230000005012 migration Effects 0.000 claims description 2
- 230000009466 transformation Effects 0.000 claims 1
- 239000003403 water pollutant Substances 0.000 claims 1
- 238000004364 calculation method Methods 0.000 description 5
- 238000005457 optimization Methods 0.000 description 3
- 239000010865 sewage Substances 0.000 description 3
- 239000000356 contaminant Substances 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000005293 physical law Methods 0.000 description 2
- 238000010200 validation analysis Methods 0.000 description 2
- KHOITXIGCFIULA-UHFFFAOYSA-N Alophen Chemical compound C1=CC(OC(=O)C)=CC=C1C(C=1N=CC=CC=1)C1=CC=C(OC(C)=O)C=C1 KHOITXIGCFIULA-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 239000002352 surface water Substances 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 238000002759 z-score normalization Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A20/00—Water conservation; Efficient water supply; Efficient water use
- Y02A20/152—Water filtration
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Databases & Information Systems (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Educational Administration (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Entrepreneurship & Innovation (AREA)
- Remote Sensing (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Feedback Control In General (AREA)
Abstract
The invention relates to a remote intelligent optimal control method for a primary rain interception facility. The rainfall prediction is carried out based on deep learning; acquiring accumulated rainfall corresponding to the opening interception facilities, namely, an initial rain interception standard through the initial rain pollution control standard of the area and analysis of future rainfall; simulating pollutant concentration by using a hydrodynamic water quality model, and evaluating whether a water quality exceeding time period exists in a future rainfall process; if so, determining the accumulated rainfall in the exceeding time period, adjusting the initial rain interception standard, re-developing water quality simulation until no exceeding water quality, determining the time range from the rainfall starting time to the time corresponding to the determined accumulated rainfall as the interception facility starting time, and generating a remote control scheme; based on the scheme, the remote control of the interception facility is realized by using configuration software. The invention can realize advanced early warning and remote control of the interception of the initial rain, can more effectively play the role of the collection and treatment of the initial rain of the interception facility, and ensures that the water environment quality of the area reaches the standard.
Description
Technical Field
The invention relates to the field of primary rain interception control, in particular to a remote intelligent optimal control method for a primary rain interception facility.
Background
The primary rain interception is an effective means for treating the pollution of the primary rain runoff and improving the water quality of river water. The initial rainwater pollution is a main source of urban non-point source pollution, and at present, the treatment of the initial rainwater pollution mainly adopts facilities such as a intercepting well, a regulating reservoir, sewage treatment equipment and the like to intercept, regulate and purify sewage, treat the sewage before entering a river and the like. However, at present, the shutoff measures cannot realize remote regulation and control, the shutoff standard is determined only by the related data, the starting time of the shutoff facility is manually judged, and the shutoff facility is manually started. This method is not only excessively time-consuming and laborious, but also may not achieve the desired shut-off effect. Therefore, predicting rainfall condition in advance, determining the initial rain interception standard and performing water quantity and water quality simulation are important tasks for realizing remote regulation and control of initial rain interception.
At present, the water environment time sequence prediction method mainly comprises two types of traditional model prediction and deep learning prediction. However, the conventional prediction method generally needs to input too many boundary conditions, and the uncertain parameters are more but difficult to obtain, and often the simulation result has errors with the actual result. Compared with the traditional model, the deep learning prediction method can calculate and process various data more quickly, acquire optimal parameters of the model, and has simpler modeling process. Therefore, deep learning prediction is gradually evolving into a mainstream method in water quality prediction. However, the deep learning model often ignores physical laws, such as energy conservation law, substance conservation law, and the like, so that accuracy and reliability of a prediction result are poor, and applicability and generalization capability of the model are poor.
At present, a plurality of areas do not analyze the accumulation condition of regional rainfall and the water quality change characteristics in the rainfall process, but directly set a primary rain interception standard by taking regional primary rain pollution control standard data as a reference, so that the standard has great uncertainty, and the primary rain interception effect is possibly inconspicuous.
Therefore, the water environment time sequence prediction method based on the deep learning model ignores the uncertainty of the physical law and the initial rain interception standard, and is a technical problem to be solved.
Disclosure of Invention
In order to solve the technical problems, the invention provides a deep learning prediction method based on physical guidance and a rainfall runoff pipe network hydrodynamic water quality model, which realize advanced early warning of initial rainwater pollution, generate a remote interception control scheme, realize remote regulation and control of interception facilities through configuration software, more effectively play the initial rainwater collection processing capacity of the initial rainwater interception facilities, and ensure that the regional water environment quality reaches the standard.
In order to achieve the above purpose, the invention provides a remote intelligent optimization control method for a primary rain interception facility, which is characterized by comprising the following steps:
s1, a deep learning prediction model based on physical guidance is used for predicting rainfall in a future time period by utilizing historical rainfall data and meteorological data, and a rainfall time sequence is generated;
s2, analyzing accumulated rainfall according to a predicted rainfall time sequence, determining the accumulated rainfall corresponding to the opening of a closure facility according to an area initial rain pollution control standard, namely an initial rain closure standard, and simulating future water quality concentration based on a rainfall runoff pipe network hydrodynamic water quality model according to the initial rain closure standard;
s3, judging whether the water quality concentration exceeds the standard, if so, determining the accumulated rainfall in the exceeding period, adjusting the initial rain interception standard, and re-carrying out water quantity and water quality simulation until the exceeding water quality does not occur in the rainfall process; step S4 is entered; if not, directly entering step S4;
s4, determining the time range from the rainfall start time to the time corresponding to the determined accumulated rainfall as the closure facility opening time, and the rest time as the closure facility closing time, generating a remote control scheme, and based on the remote control scheme, utilizing configuration software to realize remote control of the closure facility.
Further, S1 specifically includes:
s11, adding a punishment item of a hydrologic physical process into a loss function of a deep learning prediction model, and restraining a predicted value violating a physical mechanism;
s12, carrying out standardized processing on historical rainfall data and meteorological data to serve as model input, and selecting a time sequence of the first 80% as a training set and a time sequence of the last 20% as a verification set;
s13, performing super-parameter debugging and training on the deep learning prediction model by using a training set, and determining the optimal super-parameters of the model according to the performance of the trained deep learning prediction model on a verification set to obtain the trained deep learning prediction model based on physical guidance;
s14, inputting the currently measured rainfall time series data into a trained physical guidance-based deep learning prediction model, and outputting the rainfall time series of a future time period.
Optionally, the physical mechanism includes one or more of law of conservation of energy, law of conservation of matter, physical processes in hydrodynamics of water, and physical equations involved in migration and conversion of water contaminants.
Further, the simulation of the future water quality concentration based on the rainfall runoff pipe network hydrodynamic water quality model in the step S2 comprises the following steps:
s21, dividing a sub-catchment area by adopting a Thiessen polygon method according to the underlying surface data and the DEM data, carrying out pipe network generalization according to drainage pipe network data, removing pipelines with the pipe length smaller than a first threshold value and the pipe diameter smaller than a second threshold value, and removing unnecessary nodes;
s22, analyzing accumulated rainfall according to the rainfall time sequence predicted in the S1, determining the accumulated rainfall corresponding to the opening of the intercepting facility, namely, the initial rain intercepting standard by combining with the locally determined initial rain pollution control standard of the area, and taking the rainfall time sequence as input data of a rainfall runoff pipe network hydrodynamic water quality model, wherein the time sequence length is t1;
s23, determining the size of relevant parameters of each sub-catchment area, each node and each pipeline in the model according to the existing pipe network data and the literature data, determining the types of pollutants, and selecting a proper function equation to simulate the accumulation and flushing process of the pollutants;
s24, determining the water quantity and water quality parameters in a rainfall runoff pipe network hydrodynamic water quality model by selecting rainfall data and water quality data in a rainfall event, and repeatedly debugging related parameters by referring to a model user manual and literature data; selecting another rainfall event, setting the rated water quantity and water quality parameters in the model, and verifying the simulation effect of the model;
s25, selecting Nash efficiency coefficient NSE as an evaluation index of a model simulation effect, if NSE is more than 0.8, the model simulation effect is better, and directly outputting the simulation concentration of the water quality index in the future time period; if NSE <0.6, the model simulation effect is poor, and the steps S23-S24 are repeated until the simulation effect meets NSE >0.6, and the simulation concentration of the water quality index in the future time period is output.
Further, the step S3 specifically includes:
s31, according to the future water quality concentration simulated by the rainfall runoff pipe network hydrodynamic water quality model, comparing the water quality concentration of the check section in the water body with the standard concentration, and evaluating whether the water quality exceeds the standard time period in the future rainfall process;
s32, if yes, determining accumulated rainfall in an exceeding period, adjusting a primary rain interception standard, re-carrying out water quantity and water quality simulation until no exceeding water quality exists in the rainfall process, and entering a step S33; if not, directly proceeding to step S33;
s33, outputting the time range from the rainfall start time to the time corresponding to the determined accumulated rainfall, wherein the time range is the closure facility opening time;
s34, generating a remote control scheme according to the opening time of the interception facility, and based on the remote control scheme, utilizing configuration software to realize remote control of the interception equipment.
Optionally, the deep-learning predictive model described in step S1 includes one or more of RNN, CNN, LSTM, GRU, LSTM-seq2seq and variants or combinations of these deep-learning models.
Further, the penalty term added to the hydrokinetic physical process in step S11 includes:
and adding the difference value before and after the iteration of the physical equation in hydrodynamics to the loss function of the model, so that the model is participated in model training and is continuously optimized.
Optionally, the rainfall runoff network hydrodynamic water quality model described in step S2 includes one or more of SWMM, MIKE, HSPF, STORM and Infoworks CS models.
Further, the specific process of determining the model-related parameters in step S23 is as follows:
s231, dividing land utilization types according to the land utilization data, calculating the occupation ratio of each land utilization type of each sub-catchment area, counting the water impermeability of each sub-catchment area, and calculating the area, width and gradient of each sub-catchment area by adopting ArcGIS;
s232, determining Manning coefficients of the impermeable area and the permeable area, water storage depth of the depression, maximum and minimum infiltration rate, attenuation constant, pipeline roughness coefficient, maximum accumulation amount of pollutants, accumulation rate constant, pollutant scouring coefficient and scouring index according to a model user manual and related literature data.
Optionally, the water quality indicator described in step S25 includes one or more of ammonia nitrogen, total phosphorus, total nitrogen, chemical oxygen demand, dissolved oxygen, suspended solids, permanganate index, and pH.
The technical scheme provided by the invention has the following beneficial effects:
according to the invention, a physical mechanism is added in the deep learning prediction model, and the rainfall runoff pipe network hydrodynamic water quality model is combined to obtain the water quality prediction value of the future time period, so that the water quality prediction performance is improved; meanwhile, the invention can realize advanced early warning and remote control of the primary rain interception facility, optimize the primary rain control standard, more effectively exert the primary rain collecting and processing capacity of the primary rain interception facility, and ensure that the water environment quality of the area reaches the standard.
Drawings
The invention will be further described with reference to the accompanying drawings and examples, in which:
FIG. 1 is a technical roadmap of a remote intelligent optimization control method for a primary rain interception facility of the present invention;
FIG. 2 is a flow chart of the method for predicting the deep learning water quality based on physical guidance of the present invention;
FIG. 3 is a flow chart of the water quantity and quality simulation based on the rainfall runoff network hydrodynamic water quality model.
Detailed Description
For a clearer understanding of technical features, objects and effects of the present invention, a detailed description of embodiments of the present invention will be made with reference to the accompanying drawings.
Referring to fig. l, fig. 1 is a technical scheme of a remote intelligent optimization control method of a first rain interception facility according to the present invention, which mainly includes the following steps:
s1, a deep learning water quality prediction model based on physical guidance is used for predicting rainfall in a future time period by utilizing historical rainfall data and meteorological data, and a rainfall time sequence is generated;
referring to fig. 2, S1 in this embodiment specifically includes:
s11, adding a punishment item of a hydrologic physical process into a loss function of a deep learning prediction model, and restraining a predicted value violating a physical mechanism; wherein, the loss function expression without adding the punishment items of the hydrodynamics physical process is as follows: loss (y, y) p ) The penalty function expression generalizes to:
P(y，y p ，**args)＝|m(y，**args)-m(y p ，**args)|-β (1)
therefore, the loss function expression of the added hydrodynamics physical process penalty term is:
Loss＝|P(y，y p ，**args)|×loss(y，y p ) (2)
wherein y is an actual measurement value, y p Is the predictive value, args refers to other variable parameters, m (y, args) and m (y) p Args) are functions of calculating physical states corresponding to the measured value and the predicted value, respectively, and beta is setThe physical deviation threshold.
S12, carrying out standardized processing on historical rainfall data and meteorological data at a time t before the current time as model input, and selecting a time sequence of the first 80% as a training set and a time sequence of the last 20% as a verification set;
in this embodiment, the rainfall is selected as the prediction index, and the input rainfall data is calculated according to 8:2 into a training set and a validation set, and performing Z-score normalization treatment on the training set and the validation set, wherein a normalization formula is as follows:
wherein x is i As the original rainfall data, the rainfall data,
std is the standard deviation of the rainfall data of the training set, which is the mean value of the rainfall data of the training set.
S13, performing super-parameter debugging and training on the deep learning prediction model by using a training set, and determining the optimal super-parameters of the model according to the performance of the trained deep learning prediction model on a verification set to obtain the trained deep learning prediction model based on physical guidance;
selecting random search as a parameter regulator, instantiating the regulator, creating and executing a callback to stop training after the verification loss reaches a specific value, and then running a super-parameter search to obtain different super-parameters; repeatedly training the obtained different super-parameter models by using a training set, and verifying the prediction performance of the different models by using a verification set to obtain the optimal super-parameters; finally, the hyper-parametric model is re-instantiated and re-trained using the optimal hyper-parameters.
S14, inputting the currently measured rainfall time series data into a trained physical guidance-based deep learning prediction model, and outputting the rainfall time series from the current moment to the future moment t 1.
In this embodiment, it is preferable to construct an LSTM-seq2seq model using tensorflow, and perform super-parametric tuning using Keras, and select a nash efficiency coefficient (NSE) as a model evaluation index, where the calculation formula is as follows:
wherein n is the number of predicted values, Y is the actual measured value of rainfall for a certain predicted time, Y p As a predicted value of the amount of rainfall,
is the mean value of the original rainfall data.
The better the predictive performance of the model when NSE is closer to 1; when NSE >0.8, the model predictive performance can be considered to perform well; model predictive performance can be considered to be within acceptable limits when NSE > 0.6; when NSE <0.6, the model prediction performance is considered to be poor and unacceptable.
It should be noted that in other embodiments, the deep learning predictive model may be one or more of RNN, CNN, LSTM, GRU and a variant or combination of these deep learning models.
S2, analyzing accumulated rainfall according to a predicted rainfall time sequence, determining the accumulated rainfall corresponding to the opening of a closure facility according to an area initial rain pollution control standard, namely an initial rain closure standard, and simulating future water quality concentration based on a rainfall runoff pipe network hydrodynamic water quality model according to the initial rain closure standard;
in this embodiment, the rainfall runoff pipe network hydrodynamic water quality model is preferably a SWMM model, and it should be noted that in other embodiments, the rainfall runoff pipe network hydrodynamic water quality model may also be one or more of MIKE, HSPF, STORM and info oks CS models.
Referring to fig. 3, S2 in this embodiment specifically includes:
s21, dividing each sub-catchment area by utilizing the underlying surface data and the DEM data through a Thiessen polygon tool in the ArcGIS, carrying out pipe network generalization according to drainage pipe network data, removing the pipe with the pipe length smaller than a first threshold value (preferably 30 m) and the pipe diameter smaller than a second threshold value (preferably 200 mm), and removing unnecessary nodes;
s22, analyzing accumulated rainfall according to the rainfall time sequence obtained in the S1, determining the accumulated rainfall corresponding to the opening of the intercepting facility, namely, the initial rain intercepting standard by combining with the initial rain pollution control standard of the locally determined area, taking the rainfall time sequence obtained in the S1 as input data of an SWMM model, wherein the time sequence length of the input data is t1, the time sequence length of the output data is t2, and the model output data is water quality concentration data of all water quality indexes from the current moment to the later t 2;
s23, determining the size of relevant parameters of each sub-catchment area, each node and each pipeline in the model according to the existing pipe network data and the literature data, determining the types of pollutants, and selecting a proper function equation to simulate the accumulation and flushing process of the pollutants;
calculating the area, width, gradient and other attributes of each sub-catchment area by adopting an ArcGIS, importing the sub-catchment area file and the pipe network file into an SWMM model, dividing land utilization types according to land utilization data, calculating the occupation ratio of each land utilization type of each sub-catchment area, and inputting the waterproof rate of each sub-catchment area; and acquiring elevation of each node according to the related official network data.
In addition, three water quality indexes of NH3-N, TP and COD are set in the SWMM model to serve as pollutant indexes, the rainfall time sequence obtained in the above is used as an input time sequence of the SWMM model, and uncertainty parameters such as a Manning coefficient, a depression water storage depth, a maximum minimum infiltration rate, an attenuation constant, a pipeline roughness coefficient, a maximum accumulation amount of pollutants, an accumulation rate constant, a pollutant scouring coefficient, a scouring index and the like of a water impermeable zone and a water permeable zone are determined according to a model user manual and related literature data.
And (3) carrying out surface production flow calculation by adopting a nonlinear reservoir method, and obtaining the surface production flow by solving an equation set combining a continuity equation and a Manning equation. And selecting a motion wave calculation method, and carrying out one-dimensional pipe network and river channel confluence calculation by solving a complete Saint View south equation. Selecting a Horton equation to simulate and calculate the infiltration amount of rainfall runoff, designating the pollutant types as ammonia nitrogen, total phosphorus, total nitrogen and suspended solids, selecting an exponential accumulation function as an accumulation calculation equation (formula 5) of the pollutant and an exponential flushing function as a pollutant flushing equation (formula 6), and calculating the accumulation amount and flushing amount of each pollutant through the equations.
Wherein B is the accumulation amount of pollutants, C 1 For maximum accumulation of contaminants, C 2 T is the accumulation time, which is the accumulation rate constant; w is the scouring amount of pollutants, S 1 For the scouring coefficient S 2 Q is the runoff per unit area for the flush index.
S24, determining the water quantity and water quality parameters in a rainfall runoff pipe network hydrodynamic water quality model by selecting rainfall data and water quality data in a rainfall event, and repeatedly debugging related parameters by referring to a model user manual and literature data; selecting another rainfall event, setting the rated water quantity and water quality parameters in the model, and verifying the simulation effect of the model;
s25, selecting Nash efficiency coefficient NSE as an evaluation index of a model simulation effect, if NSE is more than 0.8, the model simulation effect is better, and directly outputting the simulation concentration of the water quality index in the future time period; if NSE <0.6, the model simulation effect is poor, and the steps S23-S24 are repeated until the simulation effect is within an acceptable range (namely NSE > 0.6), and then the simulation concentration of the water quality index from the current moment to the future moment t2 is output.
And selecting rainfall data and water quality data of a rainfall event as reference data for calibrating the SWMM model, taking the rainfall data as an input time sequence of the SWMM model, waiting for the rainfall data to be put into the SWMM model with well-defined parameters, simulating three water quality indexes of NH3-N, TP and COD, comparing and analyzing the simulated water quality concentration with actual water quality data in the reference data, and evaluating the performance of the model by utilizing Nash efficiency coefficients. If NSE <0.6, the model simulation effect is poor, the relevant parameters are repeatedly debugged by referring to a model user manual and literature data until NSE >0.6, the model performance is within an acceptable range, the rated model parameters are substituted into another rainfall event, water quality simulation is conducted again, whether the rated parameters are applicable in the rainfall event is verified, namely whether NSE in the verification process is larger than 0.6, if so, the rated parameters are applicable to other rainfall events, and if so, the rated parameters are conducted again.
S3, the example combines the related requirements of Chinese surface water environment quality standard (GB 3838-2002) and the requirements of a local area on water quality standard (each water quality index reaches IV type standard, namely water quality standard), judges whether the water quality concentration of each water quality index simulated by the SWMM model reaches the standard or not, when one water quality index exceeds the standard, the accumulated rainfall in the exceeding time period is required to be determined, the initial rain interception standard is readjusted, the SWMM model is used again for water quality simulation until the simulation concentration of the three water quality indexes is in the standard range, and then the step S4 is carried out; and when all three water quality indexes reach the standard, directly entering the step S4.
S4, determining the time range from the rainfall start time to the time corresponding to the determined accumulated rainfall as the closure facility opening time, and the rest time as the closure facility closing time, generating a remote control scheme, and based on the remote control scheme, utilizing configuration software to realize remote control of the closure facility.
In this embodiment, S3 and S4 specifically include:
s31, comparing the water quality concentration of the examination section in the water body with the standard concentration according to the future water quality concentration simulated by the SWMM model, and evaluating whether the water quality exceeds the standard time period in the future rainfall process;
s32, if yes, determining accumulated rainfall in an exceeding period, adjusting a primary rain interception standard, re-carrying out water quantity and water quality simulation until no exceeding water quality exists in the rainfall process, and entering a step S33; if not, directly proceeding to step S33;
s33, outputting the time range from the rainfall start time to the time corresponding to the determined accumulated rainfall, wherein the time range is the closure facility opening time;
s34, generating a remote control scheme according to the opening time of the interception facility, and based on the remote control scheme, utilizing configuration software to realize remote control of the interception equipment.
In the embodiment, a control scheme of the primary rain interception facility is generated according to the obtained interception facility opening time, an intelligent scheduling functional module of the primary rain interception facility is developed and designed through configuration software, a remote control scheme of the primary rain interception facility is generated, and visual display is performed through a system platform.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The foregoing embodiment numbers of the present invention are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments. In the unit claims enumerating several means, several of these means may be embodied by one and the same item of hardware. The use of the terms first, second, third, etc. do not denote any order, but rather the terms first, second, third, etc. are used to interpret the terms as labels.
The foregoing description is only of the preferred embodiments of the present invention, and is not intended to limit the scope of the invention, but rather is intended to cover any equivalents of the structures or equivalent processes disclosed herein or in the alternative, which may be employed directly or indirectly in other related arts.
Claims (10)
1. The remote intelligent optimal control method for the primary rain interception facility is characterized by comprising the following steps of:
s1, a deep learning prediction model based on physical guidance is used for predicting rainfall in a future time period by utilizing historical rainfall data and meteorological data, and a rainfall time sequence is generated;
s2, analyzing accumulated rainfall according to a predicted rainfall time sequence, determining the accumulated rainfall corresponding to the opening of a closure facility according to an area initial rain pollution control standard, namely an initial rain closure standard, and simulating future water quality concentration based on a rainfall runoff pipe network hydrodynamic water quality model according to the initial rain closure standard;
s3, judging whether the water quality concentration exceeds the standard, if so, determining the accumulated rainfall in the exceeding period, adjusting the initial rain interception standard, and re-carrying out water quantity and water quality simulation until the exceeding water quality does not occur in the rainfall process; step S4 is entered; if not, directly entering step S4;
s4, determining the time range from the rainfall start time to the time corresponding to the determined accumulated rainfall as the closure facility opening time, and the rest time as the closure facility closing time, generating a remote control scheme, and based on the remote control scheme, utilizing configuration software to realize remote control of the closure facility.
2. The method for remotely and intelligently optimizing control of a primary rain interception facility according to claim 1, wherein S1 specifically comprises:
s11, adding a punishment item of a hydrologic physical process into a loss function of a deep learning prediction model, and restraining a predicted value violating a physical mechanism;
s12, carrying out standardized processing on historical rainfall data and meteorological data to serve as model input, and selecting a time sequence of the first 80% as a training set and a time sequence of the last 20% as a verification set;
s13, performing super-parameter debugging and training on the deep learning prediction model by using a training set, and determining the optimal super-parameters of the model according to the performance of the trained deep learning prediction model on a verification set to obtain the trained deep learning prediction model based on physical guidance;
s14, inputting the currently measured rainfall time series data into a trained physical guidance-based deep learning prediction model, and outputting the rainfall time series of a future time period.
3. The method of claim 2, wherein the physical mechanism comprises one or more of the laws of conservation of energy, laws of conservation of matter, physical processes in hydrodynamics, and physical equations involved in migration and transformation of water pollutants.
4. The method for remotely and intelligently optimizing and controlling the primary rain interception facility according to claim 1, wherein the rainfall runoff pipe network hydrodynamic water quality model-based future water quality concentration simulation in the step S2 comprises the following steps:
s21, dividing a sub-catchment area by adopting a Thiessen polygon method according to the underlying surface data and the DEM data, carrying out pipe network generalization according to drainage pipe network data, removing pipelines with the pipe length smaller than a first threshold value and the pipe diameter smaller than a second threshold value, and removing unnecessary nodes;
s22, analyzing accumulated rainfall according to the rainfall time sequence predicted in the S1, determining the accumulated rainfall corresponding to the opening of the intercepting facility, namely, the initial rain intercepting standard by combining with the locally determined initial rain pollution control standard of the area, and taking the rainfall time sequence as input data of a rainfall runoff pipe network hydrodynamic water quality model, wherein the time sequence length is t1;
s23, determining the size of relevant parameters of each sub-catchment area, each node and each pipeline in the model according to the existing pipe network data and the literature data, determining the types of pollutants, and selecting a proper function equation to simulate the accumulation and flushing process of the pollutants;
s24, determining the water quantity and water quality parameters in a rainfall runoff pipe network hydrodynamic water quality model by selecting rainfall data and water quality data in a rainfall event, and repeatedly debugging related parameters by referring to a model user manual and literature data; selecting another rainfall event, setting the rated water quantity and water quality parameters in the model, and verifying the simulation effect of the model;
s25, selecting Nash efficiency coefficient NSE as an evaluation index of a model simulation effect, if NSE is more than 0.8, the model simulation effect is better, and directly outputting the simulation concentration of the water quality index in the future time period; if NSE <0.6, the model simulation effect is poor, and the steps S23-S24 are repeated until the simulation effect meets NSE >0.6, and the simulation concentration of the water quality index in the future time period is output.
5. The method for remotely and intelligently optimizing control of a primary rain interception facility according to claim 1, wherein steps S3 and S4 specifically comprise:
s31, according to the future water quality concentration simulated by the rainfall runoff pipe network hydrodynamic water quality model, comparing the water quality concentration of the check section in the water body with the standard concentration, and evaluating whether the water quality exceeds the standard time period in the future rainfall process;
s32, if yes, determining accumulated rainfall in an exceeding period, adjusting a primary rain interception standard, re-carrying out water quantity and water quality simulation until no exceeding water quality exists in the rainfall process, and entering a step S33; if not, directly proceeding to step S33;
s33, outputting the time range from the rainfall start time to the time corresponding to the determined accumulated rainfall, wherein the time range is the closure facility opening time;
s34, generating a remote control scheme according to the opening time of the interception facility, and based on the remote control scheme, utilizing configuration software to realize remote control of the interception equipment.
6. The method according to claim 2, wherein the deep learning predictive model in step S1 includes one or more of RNN, CNN, LSTM, GRU, LSTM-seq2seq and variants or combinations of these deep learning models.
7. The method for remotely and intelligently optimizing control of a primary rain interception facility according to claim 2, wherein the penalty term added to the hydrokinetic physical process in step S11 comprises:
and adding the difference value before and after the iteration of the physical equation in hydrodynamics to the loss function of the model, so that the model is participated in model training and is continuously optimized.
8. The method for remotely and intelligently optimizing control of a primary rain interception facility according to claim 1, wherein the rainfall runoff network hydrodynamic water quality model in step S2 comprises one or more of SWMM, MIKE, HSPF, STORM and info rks CS models.
9. The method for remotely and intelligently optimizing control of a primary rain interception facility according to claim 4, wherein the specific process of determining the model-related parameters in step S23 is as follows:
s231, dividing land utilization types according to the land utilization data, calculating the occupation ratio of each land utilization type of each sub-catchment area, counting the water impermeability of each sub-catchment area, and calculating the area, width and gradient of each sub-catchment area by adopting ArcGIS;
s232, determining Manning coefficients of the impermeable area and the permeable area, water storage depth of the depression, maximum and minimum infiltration rate, attenuation constant, pipeline roughness coefficient, maximum accumulation amount of pollutants, accumulation rate constant, pollutant scouring coefficient and scouring index according to a model user manual and related literature data.
10. The method according to claim 4, wherein the water quality index in step S25 includes one or more of ammonia nitrogen, total phosphorus, total nitrogen, chemical oxygen demand, dissolved oxygen, suspended solids, permanganate index, and pH.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211644380.3A CN116090611A (en) | 2022-12-20 | 2022-12-20 | Remote intelligent optimal control method for primary rain interception facility |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211644380.3A CN116090611A (en) | 2022-12-20 | 2022-12-20 | Remote intelligent optimal control method for primary rain interception facility |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116090611A true CN116090611A (en) | 2023-05-09 |
Family
ID=86201676
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211644380.3A Pending CN116090611A (en) | 2022-12-20 | 2022-12-20 | Remote intelligent optimal control method for primary rain interception facility |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116090611A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116842851A (en) * | 2023-08-03 | 2023-10-03 | 北京市市政工程设计研究总院有限公司广东分院 | Model system for water service data perception and mechanism analysis based on drainage basin subsystem |
-
2022
- 2022-12-20 CN CN202211644380.3A patent/CN116090611A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116842851A (en) * | 2023-08-03 | 2023-10-03 | 北京市市政工程设计研究总院有限公司广东分院 | Model system for water service data perception and mechanism analysis based on drainage basin subsystem |
| CN116842851B (en) * | 2023-08-03 | 2024-04-19 | 北京市市政工程设计研究总院有限公司 | Model system for water service data perception and mechanism analysis based on drainage basin subsystem |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Freni et al. | Uncertainty in urban stormwater quality modelling: The effect of acceptability threshold in the GLUE methodology | |
| Chen et al. | An edge intelligence empowered flooding process prediction using Internet of things in smart city | |
| Henriksen et al. | Assessment of exploitable groundwater resources of Denmark by use of ensemble resource indicators and a numerical groundwater–surface water model | |
| Jiang et al. | Risk assessment of water quality using Monte Carlo simulation and artificial neural network method | |
| CN103903452B (en) | Forecasting Approach for Short-term Traffic Flow | |
| Radwan et al. | Sensitivity and uncertainty analysis for river quality modelling | |
| Alamdari et al. | A multiobjective simulation-optimization tool for assisting in urban watershed restoration planning | |
| Zhang et al. | Applying the first-order error analysis in determining the margin of safety for total maximum daily load computations | |
| Yang et al. | An online water quality monitoring and management system developed for the Liming River basin in Daqing, China | |
| Mannina et al. | Integrated urban water modelling with uncertainty analysis | |
| Khan et al. | Comparing A Bayesian and Fuzzy Number Approach to Uncertainty Quantification in Short-Term Dissolved Oxygen Prediction. | |
| Oxley et al. | Optimization–simulation model for detention basin system design | |
| Mannina et al. | Receiving water quality assessment: comparison between simplified and detailed integrated urban modelling approaches | |
| CN111737853A (en) | Low-impact development multi-target interval optimization configuration method based on SWMM model | |
| Freni et al. | Assessment of data availability influence on integrated urban drainage modelling uncertainty | |
| CN116090611A (en) | Remote intelligent optimal control method for primary rain interception facility | |
| Shabri | A hybrid wavelet analysis and adaptive neuro-fuzzy inference system for drought forecasting | |
| Okkan | Assessing the effects of climate change on monthly precipitation: proposing of a downscaling strategy through a case study in Turkey | |
| Yin et al. | Identification of pollution sources in rivers using a hydrodynamic diffusion wave model and improved Bayesian-Markov chain Monte Carlo algorithm | |
| Gao et al. | A framework for automatic calibration of SWMM considering input uncertainty | |
| Mannina et al. | A hydrodynamic water quality model for propagation of pollutants in rivers | |
| Keupers et al. | Impact analysis of CSOs on the receiving river water quality using an integrated conceptual model | |
| Fu et al. | Simulation of urban wastewater systems using artificial neural networks: embedding urban areas in integrated catchment modelling | |
| Zhang et al. | DeepCSO: forecasting of combined sewer overflow at a citywide level using multi-task deep learning | |
| CN114595631B (en) | Water quality prediction method based on EFDC model and machine learning algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |