CN115713448A - Catchment unit-based method for quickly tracing conventional factors of basin water pollution - Google Patents
Catchment unit-based method for quickly tracing conventional factors of basin water pollution Download PDFInfo
- Publication number
- CN115713448A CN115713448A CN202211426681.9A CN202211426681A CN115713448A CN 115713448 A CN115713448 A CN 115713448A CN 202211426681 A CN202211426681 A CN 202211426681A CN 115713448 A CN115713448 A CN 115713448A
- Authority
- CN
- China
- Prior art keywords
- pollution
- river
- section
- daily
- source
- 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
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a method for quickly tracing the source of conventional factors of basin water pollution based on a catchment unit aiming at non-sudden water pollution factors such as chemical oxygen demand, total phosphorus, ammonia nitrogen and the like in daily water quality monitoring.
Description
Technical Field
The invention relates to a computer system of a specific calculation model and a water area treatment scientific technology, in particular to a drainage basin water pollution conventional factor fast tracing method based on a catchment unit. The invention requires the same applicant with the application number of 202111617892.6 and the invention name of the method is 'a method for accurately identifying the type of the cross section scale river-entering pollution source'; the benefit of the chinese patent application No. 202111628531.1 entitled "a method for estimating emissions from sources of pollution into rivers" is incorporated herein by reference in its entirety.
Background
The scientific estimation of the pollution source emission amount is required for accurate source tracing, however, the establishment of the current pollution source list mainly depends on the field investigation work with large workload and long period, besides the hourly emission data of a few industrial enterprises, the estimation of the pollution emission amount of most point sources and area sources is mostly calculated according to the year, the middle process of pollutant emission is neglected, but the water quality problem in each month is different, the dynamic estimation according to the pollution source cannot be achieved, and the scientific management requirement cannot be met.
According to the research of the prior art, the drainage basin scales COD and BOD can be known 5 、NH 3 The discharge amount of H, TN and TP is in certain relation with the pollution occurring before, i.e. the pollution at the current moment can affect a period of time in the future. In order to better simulate the process, the input of the model should be a group of time series and corresponding pollution situation data, and common machine learning regression algorithms (such as xgboost, random forest, support vector machine, SVR, nearest neighbor algorithm) are difficult to simulate the process, and the influence period of pollution has certain uncertainty.
In the prior art, aiming at the problems that the tracing method for the sudden pollution event is more, and the tracing method for the water environment problem of the accumulation type monitoring factor has the problems of pollution tracing delay, no dynamic tracing time period, no precision of tracing area and the like, in view of the above, a rapid tracing method with timely tracing, dynamic tracing time period and precision of tracing area is needed to be provided.
Disclosure of Invention
Aiming at the problems and aiming at non-sudden water pollution factors such as chemical oxygen demand, total phosphorus, ammonia nitrogen and the like of daily water quality monitoring, the invention provides a method for quickly tracing the conventional factors of the river basin water pollution based on a catchment unit. Specifically, the method for quickly tracing the conventional factors of the basin water pollution based on the catchment unit comprises the following steps:
step 1: establishing a tracing service logic, establishing a basin catchment relation and laying a foundation for accurate tracing; the tracing service logic is as follows: tracing the pollution source step by step from a downstream river reach to an upstream according to pollution identification, pollution section determination, pollution ballast determination, pollution class determination and pollution positioning; and 2, step: estimating the discharge amount of land pollution sources and the daily river inflow; and step 3: calculating pollution contribution; and 4, step 4: and carrying out accurate tracing of pollution sources.
Preferably, the establishing of the basin catchment relation further comprises: the method comprises the steps of defining a river basin range based on a DEM, wherein the DEM is a digital high-rise model; dividing a river hierarchical relationship, namely dividing the river into a plurality of levels such as a first-level river, a second-level river, a third-level river and the like, and establishing a river main branch flow relationship; establishing a corresponding relation between the section and the river reach, determining the upstream and downstream relation according to the upstream and downstream positions of the section, and numbering the section from bottom to top; the water collection unit is defined, and the water collection unit is used for dividing each section based on the DEM in the flow field; determining river inlets of the catchment units, and determining one or more main river inlets for each catchment unit; determining a list of river inlets and determining the positive discharge relation of pollution sources.
Preferably, the estimating of the land pollution source emission amount and the daily river inflow amount includes:
identifying the type of a pollution source, and dividing the pollution source into an industrial pollution point source, a living pollution point source, an agricultural pollution point source and a non-point source; and wherein the industrial pollution point source comprises an industrial park and an industrial enterprise; the domestic pollution point source comprises a sewage treatment plant, and the agricultural pollution point source comprises livestock and poultry breeding and the like; the surface source comprises: aquaculture, soil erosion, organic fertilizers, chemical fertilizers, straws, rural domestic sewage, rural domestic garbage, urban non-point sources and the like;
determining the daily river inflow of the pollution source, wherein the method comprises the following two calculation modes:
mode 1: when the annual daily rainfall can be obtained, the following method is adopted for calculation:
industrial pollution point source: the method comprises the following steps of (1) having an industrial pollution source monitored on line, wherein daily river entering quantity = online monitoring daily discharge quantity and river entering coefficient; converting the annual emission data of other industrial pollution point sources into daily emission according to the electricity consumption, wherein the daily river entering amount = daily emission amount and river entering coefficient; a life pollution point source: in a domestic sewage treatment plant with online monitoring, the daily river entering amount = online monitoring daily discharge amount and river entering coefficient; the rest annual discharge data of the used living pollution point sources are converted into daily discharge according to the seasonal production relationship, and the daily river entering amount = daily discharge amount and river entering coefficient;
mode 2: when only rainfall in a certain day can be obtained and no rainfall in the whole year exists, a depth-adaptive multiple selectable cyclic neural network model is established, and drainage and river inflow of a river basin are simulated and calculated.
Preferably, the depth-adaptive multiple selectable cyclic neural network model adopts an LSTM network, which comprises a point source-livestock algorithm, a point source-industrial wastewater algorithm, a point source-urban domestic sewage algorithm and an 8-plane source algorithm model.
Preferably, the process of training and optimizing the deep adaptive multiple selectable recurrent neural network model is as follows: processing data according to the historical data to form a complete database; according to the call request of the user, the system automatically completes the training of data; after the network model is trained successfully, automatically updating the model database so as to call the request of the data; when a system detects a request of data calculation, firstly, which algorithm model should be adopted is judged according to tag, and the forward calculation of a neural network is completed by adopting model deployment trained by an LSTM network according to a depth self-adaptive multiple selectable circular neural network model; and verifying the calculated result, and returning to call after the verification is successful.
Preferably, the step of calculating the pollution contribution comprises:
calculating upstream incoming water pollution, wherein the upstream incoming water pollution is pollution contribution brought by the upstream of the section, and the concentration value of a certain pollutant concentration value of the upstream section after being naturally reduced by the section is taken as a calculated value, namely:
C on the upper part : concentration value of upstream section, mg/l; c Attenuation (a) : the concentration value of the upstream section reaches the concentration value of the downstream section after the upstream section is attenuated, and the concentration value is mg/l;
Calculating the polluted river entering amount of the catchment unit; and calculating a pollution contribution, wherein,
upstream incoming water contribution = C Attenuation (a) /C 0 X 100%; the self-pollution contribution of the fracture surface = C 0 -C Attenuation(s) /C 0 ×100％；C 0 : concentration value of the section is mg/l; c Attenuation (a) : the concentration value of the upstream section reaches the concentration value of the downstream section in mg/l after the upstream section is attenuated;
the contribution of a certain pollution type of the section = the pollution river inflow of the pollution type of the section/the pollution river inflow in the catchment unit of the section;
the pollution contribution of a certain point source or land block of the section = the pollution river inflow amount of the point source or the land block of the section/the pollution river inflow amount in the water collection unit of the section.
Preferably, the calculating the polluted river inflow amount of the catchment unit comprises:
the daily river inflow of pollution in the water collection unit of the section = daily river inflow of the water collection unit industrial park + daily river inflow of the water collection unit industrial enterprise + daily river inflow of the water collection unit sewage treatment plant + daily river inflow of the water collection unit livestock and poultry breeding day + daily river inflow of the water collection unit aquaculture + daily river inflow of the water collection unit soil erosion + daily river inflow of the water collection unit organic fertilizer + daily river inflow of the water collection unit chemical fertilizer + daily river inflow of the water collection unit straw + daily river inflow of the water collection unit rural domestic sewage + daily river inflow of the water collection unit rural domestic garbage + daily river inflow of the water collection unit urban surface source;
the river inflow of a certain pollution type of the catchment unit = the sum of the river inflow of all pollution sources of the type in the catchment unit.
Preferably, the developing the accurate tracing of the pollution source comprises: identifying a problem river reach and starting tracing; calculating the contribution degree of the upstream incoming water and the contribution degree of the local river reach by taking the problem river reach as a starting point, thereby determining the source tracing river reach and realizing the pollution section determination; determining a tracing key area and a pollution area; determining a river reach leading pollution class; and realizing accurate tracing according to the ratio of the accumulated river entering quantity.
Preferably, the method for calculating the upstream incoming water contribution and the local river reach contribution is as follows:
if the concentration value of the downstream section corresponding to the river reach-the concentration value after the upstream section is attenuated is greater than 0, then:
the upstream incoming water contribution = the concentration value after the upstream cross section is attenuated/the concentration value of the present cross section × 100%;
the river reach contribution degree = (concentration value after attenuation of the section at the upstream side)/section concentration value 100%;
if the concentration value of the downstream section is less than 0 after the attenuation of the upstream section, which indicates that the upstream pollution is large, the upstream section tracing analysis is carried out:
the upstream incoming water contribution degree = (concentration value after attenuation of upstream cross section-concentration value of the current cross section)/concentration value of the current cross section is 100%;
the river reach contribution = 1-the upstream incoming water contribution.
Preferentially, the catchment unit with contribution degree accounting for 85% is a key area for tracing the source of a problematic river reach, the pollution source type with accumulated discharge amount accounting for 85% is a leading pollution type of the current river reach, the source with accumulated river inflow accounting for 85% is a leading suspected pollution source list of the current river reach, and points can be traced accurately according to the longitude and latitude of the list.
Drawings
Various embodiments or examples ("examples") of the disclosure are disclosed in the following detailed description and accompanying drawings. The drawings are not necessarily to scale. In general, the operations of the disclosed methods may be performed in any order, unless otherwise specified in the claims. In the drawings:
FIG. 1 is a flow chart of a rapid tracing method for a conventional factor of basin water pollution based on a catchment unit according to the invention;
FIG. 2 is a flow chart of a depth adaptive multiple selectable recurrent neural network algorithm employed in the present invention;
FIG. 3 is an LSTM network employed by the recurrent neural network;
FIG. 4 is a flow chart of the overall training and optimization operation of the deep adaptive multiple selectable recurrent neural network model system;
FIG. 5 is a schematic diagram of the accurate tracing of the pollution source development according to the present invention;
FIG. 6 is a schematic diagram of determining the dominant river reach pollution type according to the present invention;
fig. 7 is a schematic diagram of the present invention for realizing accurate tracing according to the ratio of the accumulated river inflow.
Detailed Description
Before explaining one or more embodiments of the present disclosure in detail, it is to be understood that the embodiments are not limited in their application to the details of construction and to the steps or methods set forth in the following description or illustrated in the drawings.
Fig. 1 is a flow chart of a basin water pollution conventional factor fast tracing method based on a water collection unit according to the invention.
As shown in fig. 1, the flow of the basin water pollution conventional factor fast tracing method based on the catchment unit is as follows:
1. establishing tracing service logic, establishing basin catchment relation and laying foundation for accurate tracing
The tracing service logic is as follows: based on the catchment relation of the drainage basin, the method traces the source of the pollution according to the pollution identification, pollution section determination, pollution ballast determination, pollution class determination and pollution positioning, finally forms the idea of tracing the source list, traces the source from the downstream river reach to the upstream step by step, and finds the problem accurately and timely. The basin catchment relation is established as follows:
1. demarcating river basin range based on DEM
And (3) extracting the water flow direction, confluence cumulant, water flow length and the like of a watershed and surface water flow runoff model by using a hydrological analysis tool based on a DEM (digital high-rise model), and delimiting a river basin range.
2. Dividing river hierarchy relationship
The river is divided into a plurality of levels such as a first-level river, a second-level river, a third-level river and the like, and a river main branch flow relation is established.
3. Establishing the corresponding relation between the section and the river reach
And determining the upstream and downstream relation according to the upstream and downstream positions of the sections, and numbering the sections from bottom to top. In a space superposition mode, the whole river is divided into a plurality of river sections by sections, and the corresponding relation between the sections and each river section is established.
4. Demarcating water collecting unit
Dividing water collection units (also called section water collection areas) of each section based on a DEM (digital elevation model) in a flow field, setting a water outlet for each water collection unit, and establishing input and output relations among the water collection units; meanwhile, establishing corresponding relations between the water collecting unit and the river reach and between the water collecting unit and the water outlet; and finding out pollution sources on land in each water collection unit through the spatial range of the water collection units, and establishing an association relationship between the land pollution sources and the water collection units so as to establish an association relationship between the land pollution sources and the water collection units and between the river reach and the section.
5. Determining the river mouth of the catchment unit
One or more main river inlets are determined for each catchment unit, and the main river inlets comprise a main natural discharge opening and a heavy pollution source straight discharge opening. When no major point source straight discharge port exists, a main natural discharge port is selected as a catchment unit river inlet; when a heavy pollution source straight discharge port is arranged in the water collection unit, the heavy pollution source straight discharge port is required to be included besides 1 natural discharge port.
6. Determining a logic estuary list and determining a positive emission relation of pollution sources
And collecting the river inlets of the catchment units to form a catchment unit river inlet list, and finally establishing a forward discharge logic relation of the pollution sources discharged from the sections of land pollution sources- > catchment units- > river inlets- > river reach to rivers.
2. Estimating land area pollution source discharge amount and daily river inflow amount
1. Identifying pollution source types
The pollution sources are divided into four categories, namely industrial pollution point sources, living pollution point sources, agricultural pollution point sources and surface sources. The industrial pollution point source comprises an industrial park, an industrial enterprise and the like. The point source of the domestic pollution comprises a sewage treatment plant. The agricultural pollution point source comprises livestock and poultry breeding and the like. The surface source comprises: aquaculture, soil erosion, organic fertilizer, chemical fertilizer, straw, rural domestic sewage, rural domestic garbage, urban non-point source and the like.
2. Determining daily river entering amount of pollution source
(1) When the annual daily rainfall can be obtained, the following method is adopted for calculation:
industrial pollution point source: the method comprises the following steps of (1) having an industrial pollution source monitored on line, wherein daily river entering quantity = online monitoring daily discharge quantity and river entering coefficient; and the annual emission data of the use of other industrial pollution point sources is converted into daily emission according to the power consumption, and the daily river entering amount = daily emission.
Point source of life pollution: in a domestic sewage treatment plant with online monitoring, daily river entering quantity = online monitoring daily discharge quantity and river entering coefficient; and converting other annual emission data of the point source of the life pollution into daily emission according to the seasonal production relationship, wherein daily river inflow = daily emission and river inflow coefficient.
Agricultural pollution point source: the livestock and poultry farm is monitored on line, and the daily river entering quantity = online monitoring daily discharge quantity and river entering coefficient; and (3) breeding other livestock and poultry: daily river inflow = annual discharge data-river inflow coefficient-percentage daily runoff.
A non-point source: aquaculture, soil erosion, organic fertilizers, chemical fertilizers, straws, rural domestic sewage, rural domestic garbage and urban non-point sources are calculated by adopting annual discharge, and the daily river inflow = annual discharge data and river inflow coefficient and daily runoff percentage.
(2) When only rainfall in a certain day can be obtained and no rainfall in the whole year exists, a depth-adaptive multiple selectable cyclic neural network model is established, and drainage basin discharge and river inflow are simulated and calculated
The depth-adaptive multiple selectable recurrent neural network model of the present invention is shown in fig. 2. According to figure 2, in the point source-livestock algorithm analysis, year, month and livestock stock-keeping quantity, daily pollution discharge coefficient processing rate and direct access to river are input, model training is carried out through a deep self-adaptive multiple selectable circulating neural network, and TP, TN, COD and BOD are completed 5 、NH 3 And (4) calculating the discharge amount of indexes such as _Nand the like and the river entering amount, and further calculating the river entering and exiting occupation ratio.
In the point source-industrial wastewater algorithm analysis, input data comprise monthly, yearly and monthly power consumption, annual discharge of wastewater of affiliated industries and enterprises, discharge mode, treatment rate and the like, model parameter training is completed through a deep self-adaptive multiple selectable circulating neural network, and TP, TN, COD and BOD are calculated 5 、NH 3 N, and finally calculating the river entering and exiting proportion; in the point source-urban domestic sewage algorithm analysis, input indexes comprise year, month, service population, treatment rate, day-by-day contribution degree and the like, model training is carried out according to related data, and TP, TN, COD and BOD are calculated finally 5 、NH 3 N and river inflow.
In the aspect of the surface source, 8 surface sources are included, specifically: chemical fertilizer, straw, organic fertilizer, aquaculture, soil erosion, domestic sewage, domestic garbage and urban non-point source. Calculating the non-point source of the fertilizer, namely training each model parameter of the model according to indexes of pure TN and TP, rural farmland area, farmland runoff daily percentage-surface runoff, interflow and runoff total amount, pollutant emission coefficient-TP and TN applied to the rural fertilizer according to year, month and day, rural fertilizer application, and circulating neural network multiple selectable to complete the model parameters of the model, and completing TN and TP of rural pollutant emission amount and TN and TP of river entering amount; calculating straws, namely completing the training of model parameters according to the multiple self-adaptive selectable recurrent neural networks of year, month, day, grain yield of villages and towns, farmland area of villages and towns, farmland runoff rate daily percentage-surface runoff rate, interflow, runoff total amount, percentage, pollutant emission coefficient-TP, TN index, straw output coefficient, straw utilization rate, pollutant content and the like according to the depth, and finally outputting the pollutant emission amount _ TN and TP of the villages and the towns and TN and TP of river inflow; in the calculation of the organic fertilizer, according to data such as year, month, day, application amount of the organic fertilizer in the villages and towns, area of farmland in the villages and towns, daily percentage of runoff of the farmland, surface runoff, interflow, total runoff, percentage, pollutant content of the organic fertilizer and the like, a multiple selectable recurrent neural network is self-adapted according to depth to complete the training of model parameters, and discharge amount _ TN and TP of pollutants in the villages and towns and TN and TP of river entering amount are output; in the aspects of aquaculture, soil erosion, domestic sewage, domestic garbage and urban non-point sources, the training of the parameters of the finished model is completed by combining a depth self-adaptive multiple selectable circulating neural network according to corresponding input indexes and calculating the final result.
With further reference to fig. 2, the network inputs the pollution factor sequence at the previous time t (i.e. the pollution factor data at the previous time t is included), wherein the recurrent neural network structure calculates the information at the current time by sensing the information at the previous time in the current state, and here, the sensing complexity can be increased by using the multiple recurrent neural network structure, so that the point source and area source output index data at the current time can be calculated more accurately. Wherein the recurrent neural network element structure employs an LSTM network as shown in fig. 3. In contrast to conventional recurrent neural networks, the LSTM is still based on the current time instant input x t And the output h of the previous moment t-1 To calculate the output h of the current time t Only the internal structure is designed more elaborately, and an input gate i is added t Forgetting door f t And an output gate o t Three gates and an internal memory unit c t . The input gate controls how much the currently calculated new state is updated into the memory unit, the forgetting gate controls how much the information in the previous memory unit is forgotten, the output gate controls how much the current output depends on the current memory unit, and the information transfer formula in the LSTM is as follows:
the input gate formula: i.e. i t ＝σ(W i x t +U i h t-1 +b i )
Forget gate formula: f. of t ＝σ(W f x t +U f h t-1 +b f )
Output gate formula: o. o t ＝σ(W o x t +U o h t-1 +b o )
Candidate layer formula:
the memory cell update formula:
each layer outputs the formula: h is a total of t ＝o t ■Tanh(c t )
Wherein 9632is an exclusive OR operation, i.e. the same value is 1, and different values are 0
σ is the Sigmoid activation function and Tanh is the hyperbolic tangent activation function.
Each cycle of LSTM may pass two values into the next cycle: c. C t And h t Here, long-term memory and short-term memory are to be understood. The network can effectively solve the problem that the gradient of the recurrent neural network disappears, has a memory function, can better utilize the prior pollution monitoring factors, can recognize that important information is transferred, unimportant information is forgotten, and automatically performs node adjustment and dimension adjustment according to a loss function of training, and has the defect of huge training cost.
For the deep adaptive multiple selectable recurrent neural network model shown in fig. 2 and 3, data processing is performed according to historical data to form a complete database, and data in the database can be flexibly called and operated. And then, the system automatically finishes the training of the data according to the call request of the user. And a training stage, which adopts a deep self-adaptive multiple selectable cyclic neural network, wherein the network specifies input data and output data of each model. After the network model is trained successfully, the model database is automatically updated so as to be called by the request of the data. When the system detects a request of data calculation, firstly, which calculation model should be adopted is judged according to tag, and the forward calculation of the neural network is completed by data according to model deployment of a tensoflow serving platform. And the calculation result is verified, and after the verification is successful, the call is returned. The overall training and optimizing operation flow chart of the deep self-adaptive multiple selectable circular neural network model system is shown in FIG. 4.
3. Calculating pollution contribution
The section pollution is divided into upstream water pollution and section catchment unit pollution.
1. Calculating upstream water pollution
The upstream water pollution refers to the pollution contribution brought by the upstream of the section, and the concentration value of a certain pollutant on the upstream section is naturally reduced by the section of the river and then is used as a calculated value. Namely:
C on the upper part : concentration value of upstream section, mg/l;
C attenuation(s) : the concentration value of the upstream section reaches the concentration value of the downstream section in mg/l after the upstream section is attenuated;
2. Calculating the amount of polluted water entering the river of the catchment unit
The daily pollution contribution is taken as an example to illustrate the calculation of the pollution contribution.
The sum of daily river inflow of certain pollutants of various types of pollution sources in the inland area of the section catchment unit is as follows:
the daily river inflow of pollution in the water collection unit of the section = daily river inflow of the water collection unit industrial park + daily river inflow of the water collection unit industrial enterprise + daily river inflow of the water collection unit sewage treatment plant + daily river inflow of the water collection unit livestock and poultry breeding day + daily river inflow of the water collection unit aquaculture + daily river inflow of the water collection unit soil erosion day + river inflow of the water collection unit organic fertilizer + daily river inflow of the water collection unit chemical fertilizer + daily river inflow of the water collection unit straw + daily river inflow of the water collection unit rural domestic sewage + daily river inflow of the water collection unit rural domestic garbage + daily river inflow of the water collection unit urban surface source
The river inflow of a certain pollution type of the catchment unit = the sum of the river inflow of all pollution sources of the type in the catchment unit.
3. Calculating the contribution degree of pollution
Upstream incoming water contribution = C Attenuation (a) /C 0 ×100％；
Self-contamination contribution of the fracture surface = C 0 -C Attenuation (a) /C 0 ×100％；
C 0 : the concentration value of the section is mg/l;
C attenuation (a) : the concentration value of the upstream section reaches the concentration value of the downstream section in mg/l after the upstream section is attenuated;
the contribution of a certain pollution type of the section = the pollution river inflow of the pollution type of the section/the pollution river inflow in the catchment unit of the section.
The pollution contribution of a certain point source or land block of the section = the pollution river inflow amount of the point source or the land block of the section/the pollution river inflow amount in the water collection unit of the section.
4. Develop accurate tracing to source of pollution source
1. Identifying problematic river reach, and starting tracing
And performing section standard exceeding analysis by automatically monitoring the daily monitoring value and the monthly monitoring value of the section, and calculating to obtain a problem section list with standard exceeding according to the section-section corresponding relation. And selecting any problematic river reach and starting tracing.
2. Determining the tracing river reach range by taking the problem river reach as a starting point to realize pollution section determination
Taking the problem river reach as a starting point, and tracing the river reach of the pollution source upwards towards the incoming water direction of the river. The tracing range of a certain river reach should sequentially trace the source to the upstream of the river according to the river reach with positive contribution as the starting point of tracing the source until the concentration value of the upstream cross section is lower than that of the downstream cross section or until the source is traced, and the judgment relationship is as shown in fig. 5. The found river reach is the river reach with greater pollution contribution.
Wherein each river reach contribution needs to reduce the influence of upstream incoming water.
【1】 If the concentration value of the downstream section corresponding to the river reach-the concentration value after the upstream section is attenuated is greater than 0, then:
the upstream incoming water contribution = the concentration value after the upstream cross section is attenuated/the concentration value of the present cross section × 100%;
the river reach contribution degree = (concentration value of the cross section-concentration value after attenuation of the upstream cross section)/concentration value of the cross section = 100%.
【2】 If the concentration value of the downstream section is less than 0 after the attenuation of the upstream section, which indicates that the upstream pollution is large, the upstream section tracing analysis is carried out:
the upstream incoming water contribution degree = (concentration value after attenuation of upstream cross section-concentration value of the current cross section)/concentration value of the current cross section is 100%;
the local river reach contribution = 1-upstream incoming water contribution.
3. Determining tracing key area and pollution area
And the catchment units corresponding to the river reach are all tracing areas. In the tracing area, the water catchment units are ranked according to the contribution degree of the discharge of the inflow amount of the water catchment units to the river reach, and the accumulated contribution degree accounts for 85 percent, so that the tracing area is the important area of the problem river reach.
4. Determining a river reach leading pollution type
According to a list of key catchment units for accounting traceability, selecting pollution factors, sequencing the point source and non-point source discharge amount in a space area, finding out the pollution source type with the accumulated discharge amount accounting for 85%, and determining the dominant pollution type of the river reach, as shown in fig. 6. Once the pollution source type is located, the emission amount and the proportion of the analyzed pollution source type can be displayed in a list mode.
5. Realizing accurate tracing according to the ratio of the accumulated river entering quantity
On the premise of determining the leading pollution type of the current river reach, according to a main pollution source list under the type, a source with the accumulated river volume accounting for 85% is a leading suspected pollution source list of the current river reach, and according to the longitude and latitude of the list, the source can be accurately traced to a point. The point sources are sequenced and calculated according to lists of single industrial enterprises, sewage plants and livestock and poultry with the discharge amount accounting for 85% of the discharge amount of the point sources; the area source is calculated by taking the village and the town as the minimum unit, and the specific pollution source quantity and the proportion are obtained through analysis, as shown in fig. 7.
Claims (10)
1. A catchment unit-based method for quickly tracing conventional factors of basin water pollution includes: COD and BOD 5 、NH 3 H, TN and TP, the rapid and accurate tracing method comprises the following steps:
step 1: establishing a tracing service logic, establishing a basin catchment relation and laying a foundation for accurate tracing; the tracing service logic is as follows: tracing the pollution source step by step from a downstream river reach to an upstream according to pollution identification, pollution section determination, pollution ballast determination, pollution class determination and pollution positioning;
step 2: estimating land pollution source discharge amount and daily river entering amount;
and step 3: calculating the pollution contribution;
and 4, step 4: and carrying out accurate tracing of the pollution source.
2. The method as claimed in claim 1, wherein the establishing of the basin water catchment relation further comprises:
the method comprises the steps of (1) defining a river basin range based on a DEM (digital elevation model), wherein the DEM is a digital high-rise model;
dividing a river hierarchical relationship, namely dividing the river into a plurality of levels such as a first-level river, a second-level river, a third-level river and the like, and establishing a river main branch flow relationship;
establishing a corresponding relation between the section and the river reach, determining the upstream and downstream relation according to the upstream and downstream positions of the section, and numbering the section from bottom to top; dividing the whole river into a plurality of river sections by sections in a space superposition mode, and establishing a corresponding relation between the sections and each river section;
the method comprises the following steps of defining water collection units, wherein the water collection units are divided into sections based on a DEM (digital elevation model) in a flow field, each water collection unit is provided with a water outlet, and the input and output relations among the water collection units are established; meanwhile, establishing corresponding relations between the catchment units and the river reach as well as between the catchment units and the water outlet; finding out a pollution source on land in each water collection unit through the spatial range of the water collection unit, and establishing an incidence relation between the land pollution source and the water collection unit so as to establish an incidence relation between the land pollution source and the water collection unit and between the river reach and the cross section;
determining river inlets of the catchment units, and determining one or more main river inlets including a main natural discharge opening and a direct discharge opening of a heavy pollution source for each catchment unit;
and determining a logic estuary list and determining a positive emission relation of the pollution source.
3. The method for fast and accurately tracing the watershed water pollution conventional factor according to claim 1, wherein the estimating of the discharge amount of the land pollution source and the daily river inflow comprises:
identifying the type of a pollution source, and dividing the pollution source into an industrial pollution point source, a domestic pollution point source, an agricultural pollution point source and a surface source; and wherein the industrial pollution point source comprises an industrial park and an industrial enterprise; the domestic pollution point source comprises a sewage treatment plant, and the agricultural pollution point source comprises livestock and poultry breeding and the like; the surface source comprises: aquaculture, soil erosion, organic fertilizers, chemical fertilizers, straws, rural domestic sewage, rural domestic garbage, urban non-point sources and the like;
determining the daily river inflow of the pollution source, wherein the method comprises the following two calculation modes:
mode 1: when the annual daily rainfall can be obtained, the following method is adopted for calculation:
industrial pollution point source: the method comprises the following steps of (1) having an industrial pollution source monitored on line, wherein daily river entering quantity = online monitoring daily discharge quantity and river entering coefficient; the emission data of other industrial pollution point sources in the use years is converted into daily emission according to the electricity consumption, and the daily river entering amount = daily emission amount and river entering coefficient; a life pollution point source: in a domestic sewage treatment plant with online monitoring, daily river entering quantity = online monitoring daily discharge quantity and river entering coefficient; converting other annual emission data of the life pollution point sources into daily emission according to the seasonal production relationship, wherein daily river entering amount = daily emission amount and river entering coefficient;
mode 2: when only rainfall in a certain day can be obtained and no rainfall in the whole year exists, a depth-adaptive multiple selectable cyclic neural network model is established, and drainage and river inflow of a river basin are simulated and calculated.
4. The method for fast and accurately tracing the conventional factors of the watershed water pollution according to claim 3, wherein the depth-adaptive multiple selectable cyclic neural network model adopts an LSTM network, which comprises a point source-livestock algorithm, a point source-industrial wastewater algorithm, a point source-municipal domestic sewage algorithm and an 8-surface source algorithm model.
5. The method for fast and accurately tracing the watershed water pollution conventional factor as claimed in claim 4, wherein the process of training and optimizing the deep adaptive multiple selectable recurrent neural network model comprises:
processing data according to the historical data to form a complete database; according to the calling request of the user, the system automatically completes the training of data; after the network model is trained successfully, automatically updating the model database so as to be convenient for the request call of the data; when a system detects a request of data calculation, firstly, judging which algorithm model should be adopted according to tag, and finishing the forward calculation of the neural network by adopting model deployment trained by an LSTM network according to a depth self-adaptive multiple selectable circular neural network model; and verifying the calculated result, and returning to call after the verification is successful.
6. The method for fast and accurately tracing the watershed water pollution conventional factor according to claim 1, wherein the step of calculating the pollution contribution comprises the following steps:
calculating upstream incoming water pollution, wherein the upstream incoming water pollution is a pollution contribution brought by the upstream of the section, and the upstream incoming water pollution takes a concentration value of a certain pollutant of the upstream section after being naturally reduced by the section as a calculated value, namely:
C on the upper part : concentration value of upstream section, mg/l; c Attenuation (a) : the concentration value of the upstream section reaches the concentration value of the downstream section in mg/l after the upstream section is attenuated;
Calculating the polluted river entering amount of the catchment unit; and calculating a pollution contribution, wherein,
upstream incoming water contribution = C Attenuation (a) /C 0 X is 100%; self-contamination contribution of the fracture surface = C 0 -C Attenuation(s) /C 0 ×100％；C 0 : the concentration value of the section is mg/l; c Attenuation (a) : the concentration value of the upstream section reaches the concentration value of the downstream section after the upstream section is attenuated, and the concentration value is mg/l;
the contribution of a certain pollution type of the section = the pollution river inflow of the pollution type of the section/the pollution river inflow in the catchment unit of the section;
and the pollution contribution of a certain point source or land block of the section = the pollution river inflow of the point source or the land block of the section/the pollution river inflow of the water collection unit of the section.
7. The method as claimed in claim 6, wherein the calculating of the polluted river inflow amount of the catchment unit comprises:
the daily river inflow of pollution in the water collection unit of the section = daily river inflow of the water collection unit industrial park + daily river inflow of the water collection unit industrial enterprise + daily river inflow of the water collection unit sewage treatment plant + daily river inflow of the water collection unit livestock and poultry breeding day + daily river inflow of the water collection unit aquaculture + daily river inflow of the water collection unit soil erosion + daily river inflow of the water collection unit organic fertilizer + daily river inflow of the water collection unit chemical fertilizer + daily river inflow of the water collection unit straw + daily river inflow of the water collection unit rural domestic sewage + daily river inflow of the water collection unit rural domestic garbage + daily river inflow of the water collection unit urban surface source;
the river inflow of a certain pollution type of the catchment unit = the sum of the river inflow of all pollution sources of the type in the catchment unit.
8. The method for fast and accurately tracing the conventional factors of the watershed water pollution as claimed in claim 1, wherein the developing the accurate tracing of the pollution sources comprises:
identifying a problem river reach and starting tracing; calculating the contribution degree of the upstream incoming water and the contribution degree of the local river reach by taking the problem river reach as a starting point, thereby determining the source tracing river reach and realizing the pollution section determination; determining a tracing key area and a pollution area; determining a river reach leading pollution class; and realizing accurate tracing according to the ratio of the accumulated river entering quantity.
9. The method as claimed in claim 8, wherein the method for calculating the contribution of the upstream incoming water and the contribution of the local river reach comprises:
if the concentration value of the downstream section corresponding to the river reach-the concentration value after the attenuation of the upstream section is greater than 0, then:
the upstream incoming water contribution = upstream cross section attenuated concentration value/this cross section concentration value 100%;
the river reach contribution degree = (concentration value after attenuation of the section at the upstream side)/section concentration value 100%;
if the concentration value of the downstream section-the concentration value of the upstream section after attenuation is less than 0, which indicates that the upstream pollution is large, the upstream section tracing analysis should be carried out:
the upstream incoming water contribution degree = (concentration value after attenuation of upstream cross section-concentration value of the current cross section)/concentration value of the current cross section is 100%;
the local river reach contribution = 1-upstream incoming water contribution.
10. The method for fast and accurately tracing the source of the conventional factors of the watershed water pollution according to claim 9, wherein the catchment unit with the contribution degree of 85 percent is a key area for tracing the source of the problematic river reach, the pollution source type with the accumulated discharge amount of 85 percent is a dominant pollution type of the river reach, the source with the river inflow amount of 85 percent is accumulated, the list of the dominant suspected pollution sources of the river reach is obtained, and the source can be accurately traced to the point according to the longitude and latitude of the list.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211426681.9A CN115713448A (en) | 2022-11-15 | 2022-11-15 | Catchment unit-based method for quickly tracing conventional factors of basin water pollution |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211426681.9A CN115713448A (en) | 2022-11-15 | 2022-11-15 | Catchment unit-based method for quickly tracing conventional factors of basin water pollution |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115713448A true CN115713448A (en) | 2023-02-24 |
Family
ID=85233104
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211426681.9A Pending CN115713448A (en) | 2022-11-15 | 2022-11-15 | Catchment unit-based method for quickly tracing conventional factors of basin water pollution |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115713448A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116930445A (en) * | 2023-09-06 | 2023-10-24 | 水电水利规划设计总院 | Tracing method for water pollution of boundary river in administrative area |
| CN117195135A (en) * | 2023-11-01 | 2023-12-08 | 潍坊德瑞生物科技有限公司 | Water pollution anomaly traceability detection method and system |
-
2022
- 2022-11-15 CN CN202211426681.9A patent/CN115713448A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116930445A (en) * | 2023-09-06 | 2023-10-24 | 水电水利规划设计总院 | Tracing method for water pollution of boundary river in administrative area |
| CN116930445B (en) * | 2023-09-06 | 2023-12-15 | 水电水利规划设计总院 | Tracing method for water pollution of boundary river in administrative area |
| CN117195135A (en) * | 2023-11-01 | 2023-12-08 | 潍坊德瑞生物科技有限公司 | Water pollution anomaly traceability detection method and system |
| CN117195135B (en) * | 2023-11-01 | 2024-02-27 | 潍坊德瑞生物科技有限公司 | Water pollution anomaly traceability detection method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108664647B (en) | Basin fine management system of integrated water environment model | |
| CN104318325B (en) | Many basin real-time intelligent water quality prediction methods and system | |
| CN103810537B (en) | A kind of Regional environment risk appraisal procedure based on water quality model | |
| Zhao et al. | Water resources risk assessment model based on the subjective and objective combination weighting methods | |
| CN115713448A (en) | Catchment unit-based method for quickly tracing conventional factors of basin water pollution | |
| CN102854296B (en) | Sewage-disposal soft measurement method on basis of integrated neural network | |
| CN110728035A (en) | Pollutant total amount control method based on control of section water quality reaching standard | |
| CN108573302A (en) | A kind of simulation of basin non-point source pollution loading and Best Management Practices optimization method | |
| CN110175948A (en) | A kind of ecological environment water demand threshold value quantization method based on river holistic health | |
| CN110909484A (en) | Watershed grey water footprint evaluation method and water environment treatment strategy making method | |
| CN108717453A (en) | A kind of Plain River Network Areas pollutional load computational methods based on GIS platform | |
| CN109598428B (en) | Pollutant reduction and distribution method based on administrative units and water system | |
| CN117236199B (en) | Method and system for improving water quality and guaranteeing water safety of river and lake in urban water network area | |
| CN104182794A (en) | Method for soft measurement of effluent total phosphorus in sewage disposal process based on neural network | |
| Wang et al. | A full-view management method based on artificial neural networks for energy and material-savings in wastewater treatment plants | |
| CN110765213A (en) | Method for compiling emission list (dynamic list) of pollution sources in surface water basin | |
| CN109613197A (en) | A kind of water quality monitoring early warning feedback response method based on the river network of rivers | |
| CN114858207A (en) | Soft measurement-based gridding source tracing investigation method for drain outlet of river channel | |
| CN114418446A (en) | Quantitative assessment method for water resource shortage | |
| CN114022008A (en) | Estuary suitable ecological flow assessment method based on water ecological zoning theory | |
| CN114595631A (en) | Water quality prediction method based on EFDC model and machine learning algorithm | |
| CN116484647B (en) | Distributed water resource allocation method and system for overall coordination | |
| Kim et al. | Operator decision support system for integrated wastewater management including wastewater treatment plants and receiving water bodies | |
| CN117610208A (en) | River water quality and quantity prediction method based on urban water system model interaction | |
| CN112287613A (en) | Pollutant reduction method for watershed water environment control section |
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 |