Phosphorus pollution load monitoring and evaluating method applied to watershed scale

Abstract

The invention belongs to the field of environmental protection, and particularly relates to a phosphorus pollution load monitoring and evaluating method applied to a watershed scale. The objective of the invention is to solve the problem of low accuracy of phosphorus pollution load measurement in a drainage basin. The phosphorus pollution load monitoring and evaluating method applied to the watershed scale comprises the following steps: step 1, constructing a phosphorus pollution monitoring system; the phosphorus pollution monitoring system in the step 1 comprises a data acquisition module; a phosphorus pollution load accounting module; the phosphorus pollution load accounting module comprises a trained pollution emission coefficient prediction network, a trained emission generation coefficient prediction network and a phosphorus pollution load accounting model; 2, acquiring environmental data of the monitored drainage basin in real time, and inputting the acquired environmental data of the monitored drainage basin into a phosphorus pollution monitoring system to obtain a phosphorus pollution load value of the monitored drainage basin; the problem that the accuracy of the measured phosphorus pollution load in the drainage basin is low is solved.

Description

Technical Field

[0001] The invention belongs to the field of environmental protection, and in particular relates to a phosphorus pollution load monitoring and evaluation method applied at a watershed scale. Background Art

[0002] Phosphorus pollution is one of the important factors of water eutrophication. With the advancement of industrialization, the "three phosphorus enterprises" such as the exploitation of phosphorus resources, the production of phosphorus chemical products and the construction of phosphogypsum storage enterprises have had a significant impact on the environment, especially on water bodies in phosphorus-rich areas such as the Yangtze River, causing phosphorus pollution and other problems.

[0003] Traditional phosphorus pollution monitoring methods mostly focus on single emission sources or water body monitoring, which makes it difficult to accurately reflect the pollution distribution of phosphorus load in the basin, and lacks timeliness and spatial resolution. Due to the limitations of traditional methods, it is impossible to effectively monitor the temporal and spatial distribution of phosphorus pollution load in the basin driven by the three phosphorus enterprises. There is a problem of low accuracy in measuring phosphorus pollution load in the basin. Summary of the invention

[0004] The purpose of the present invention is to solve the problem of low accuracy of the measured phosphorus pollution load in a watershed. A method for monitoring and evaluating phosphorus pollution load at a watershed scale is provided, comprising:

[0005] S1. Obtain the census data of phosphorus pollution sources of three phosphorus enterprises in the monitoring basin; determine N monitoring points in the monitoring basin according to the census data of phosphorus pollution sources of three phosphorus enterprises in the monitoring basin; deploy a sensor at each monitoring point;

[0006] S2: Obtain a first training set, and train a pollution emission coefficient prediction network according to the first training set to obtain a trained pollution emission coefficient prediction network;

[0007] S3: Obtain a second training set, and train the emission occurrence coefficient prediction network according to the second training set to obtain a trained emission occurrence coefficient prediction network;

[0008] S4: construct a phosphorus pollution load accounting model; construct a phosphorus pollution monitoring system based on the data acquisition module, the trained pollution emission coefficient prediction network, the trained emission occurrence coefficient prediction network, and the phosphorus pollution load accounting model;

[0009] S5: Acquire the environmental data of the monitored river basin in real time through sensors, input the acquired environmental data of the monitored river basin into the phosphorus pollution monitoring system, and obtain the phosphorus pollution load value of the monitored river basin.

[0010] The beneficial effects of the present invention are:

[0011] The present invention deploys an automated monitoring system at a monitoring point, including sensors, data acquisition terminals, communication modules, etc. The sensors can monitor relevant data such as phosphorus concentration and flow in water bodies in real time, and the data acquisition terminals transmit the collected data to a central database through a wireless network for processing and analysis, and use hydrological analysis tools and geographic information systems to divide the phosphorus pollution basins of three phosphorus enterprises, and use national control section data to calculate the later allowable emission limits. According to the analysis and calculation results, a phosphorus pollution load report is automatically generated, and the control measures are dynamically adjusted. The method has efficient monitoring and convenient and fast prediction, and can evaluate the phosphorus environmental risks of three phosphorus enterprises in a basin and their impact on the water environment of the basin in real time and quickly, providing a scientific basis for the phosphorus pollution control of the basin. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a schematic flow chart of a phosphorus pollution load monitoring and evaluation method applied at a watershed scale according to the present invention. DETAILED DESCRIPTION

[0013] The present invention discloses a phosphorus pollution load monitoring and evaluation method applied at a watershed scale. The general process is as follows: Step 1: constructing a phosphorus pollution monitoring system;

[0014] The phosphorus pollution monitoring system in step 1 includes a data acquisition module; a phosphorus pollution load calculation module;

[0015] The data acquisition module includes: N sensors; N is a positive integer;

[0016] The phosphorus pollution load accounting module includes: a trained pollution emission coefficient prediction network, a trained emission occurrence coefficient prediction network and a phosphorus pollution load accounting model;

[0017] Step 2: Acquire the environmental data of the monitoring basin in real time, input the acquired environmental data of the monitoring basin into the phosphorus pollution monitoring system, and obtain the phosphorus pollution load value of the monitoring basin;

[0018] Specific implementation method 1: Combination Figure 1 Describe the specific implementation mode of the present invention:

[0019] S1. Obtain the census data of phosphorus pollution sources of three phosphorus enterprises in the monitoring basin; determine N monitoring points in the monitoring basin according to the census data of phosphorus pollution sources of three phosphorus enterprises in the monitoring basin; deploy a sensor at each monitoring point;

[0020] S2: Obtain a first training set, and train a pollution emission coefficient prediction network according to the first training set to obtain a trained pollution emission coefficient prediction network;

[0021] S3: Obtain a second training set, and train the emission occurrence coefficient prediction network according to the second training set to obtain a trained emission occurrence coefficient prediction network;

[0022] S4: construct a phosphorus pollution load accounting model; construct a phosphorus pollution monitoring system based on the data acquisition module, the trained pollution emission coefficient prediction network, the trained emission occurrence coefficient prediction network, and the phosphorus pollution load accounting model;

[0023] S5: Obtain the environmental data of the monitored basin in real time through sensors, input the acquired environmental data of the monitored basin into the phosphorus pollution monitoring system, and obtain the phosphorus pollution load value of the monitored basin

[0024] Specific implementation method 2: The difference between this implementation method and specific implementation method 1 is that:

[0025] In S1, the census data of phosphorus pollution sources of three phosphorus enterprises in the monitoring basin are obtained; N monitoring points are determined in the monitoring basin according to the census data of phosphorus pollution sources of three phosphorus enterprises in the monitoring basin; the specific process is:

[0026] S1.1: Collect the phosphorus pollution source emission data of all three phosphorus enterprises in the monitoring basin and the geographic information data in the monitoring basin;

[0027] The three phosphorus enterprises include: phosphate mine enterprises, phosphate fertilizer enterprises, phosphorus-containing pesticide enterprises, yellow phosphorus enterprises, and phosphogypsum storage enterprises;

[0028] The phosphorus pollution source emission data of all three phosphorus enterprises in the monitoring basin include: phosphorus emission amount, emission point, emission method (such as gas, liquid or solid), emission time, pollutant concentration, etc. of the three phosphorus enterprises in the monitoring basin.

[0029] The data source can be the environmental protection department, the enterprise's emission report or the report provided by a third-party monitoring agency.

[0030] The geographic information data within the monitored river basin include the basin scope, river direction, important water bodies, sewage outlet locations, enterprise locations, and nearby ecologically sensitive areas.

[0031] S1.2: According to S1.1, collect the phosphorus pollution source emission data of all three phosphorus enterprises in the monitoring basin and the geographic information data in the monitoring basin to analyze the phosphorus pollution sources and obtain N monitoring points;

[0032] For example, the pollution source monitoring points can be selected according to the emission volume and concentration of pollutants emitted by the pollution source.

[0033] The selection basis is: determine the monitoring points according to the spatial distribution of pollution sources, water characteristics in the basin, and protection requirements of ecologically sensitive areas. The following factors can be considered: The monitoring points include: areas where rivers, lakes, groundwater and other water bodies meet and areas that may be affected by pollution.

[0034] The other steps and parameters are the same as those in the first specific implementation.

[0035] Specific implementation method three: This implementation method is different from the specific implementation method one in that:

[0036] The pollution emission coefficient prediction network in S2 includes: a first pollution emission coefficient prediction subnetwork, a second pollution emission coefficient prediction subnetwork, a third pollution emission coefficient prediction subnetwork, a fourth pollution emission coefficient prediction subnetwork, a fifth pollution emission coefficient prediction subnetwork,

[0037] The first pollution emission coefficient prediction subnetwork, the second pollution emission coefficient prediction subnetwork, the third pollution emission coefficient prediction subnetwork, the fourth pollution emission coefficient prediction subnetwork, and the fifth pollution emission coefficient prediction subnetwork are all LSTM networks;

[0038] The other steps and parameters are the same as those in the first to second embodiments.

[0039] Specific implementation method 4: This implementation method is different from specific implementation methods 1 to 4 in that:

[0040] In S2, a first training set is obtained, and a pollution emission coefficient prediction network is trained according to the first training set to obtain a trained pollution emission coefficient prediction network; the specific process is:

[0041] S2.1: Obtain historical phosphorus emission data; preprocess the historical phosphorus emission data to obtain preprocessed historical phosphorus emission data, and obtain a first training set based on the preprocessed historical phosphorus emission data;

[0042] The historical phosphorus emission data include: historical phosphorus emission data of phosphate mines, historical phosphorus emission data of phosphate fertilizers, historical phosphorus emission data of phosphorus-containing pesticides, historical phosphorus emission data of yellow phosphorus enterprises, and historical phosphorus emission data of phosphogypsum storage enterprises;

[0043] The specific process is:

[0044] The data preprocessing process is: normalizing, standardizing, and cleaning the historical phosphorus emission data in turn.

[0045] The historical phosphorus emission data can be obtained through annual reports of enterprises, historical monitoring or third-party monitoring agencies; the normalization, standardization and data cleaning processes are well known to those skilled in the art;

[0046] The preprocessed historical phosphorus emission data are divided into an LSTM training set, an LSTM validation set, and an LSTM test set in chronological order; the first 70% of the historical phosphorus emission data arranged in time are used as the LSTM training set, the next 15% of the historical phosphorus emission data are used as the LSTM validation set, and the last 15% of the historical phosphorus emission data are used as the LSTM test set;

[0047] S2.2: The historical phosphorus emission data of the phosphate mining enterprises in the first training set is used as the input of the first pollution emission coefficient prediction subnetwork, and the first pollution emission coefficient prediction subnetwork outputs the predicted pollution emission coefficient of the phosphate mining enterprises;

[0048] The historical phosphorus emission data of the phosphate fertilizer enterprises in the first training set are used as the input of the second pollution emission coefficient prediction subnetwork, and the second pollution emission coefficient prediction subnetwork outputs the predicted pollution emission coefficient of the phosphate fertilizer enterprises;

[0049] The historical phosphorus emission data of the phosphorus-containing pesticide enterprises in the first training set are used as the input of the third pollution emission coefficient prediction subnetwork, and the third pollution emission coefficient prediction subnetwork outputs the predicted pollution emission coefficient of the phosphorus-containing pesticide enterprises;

[0050] The historical phosphorus emission data of yellow phosphorus enterprises in the first training set are used as the input of the fourth pollution emission coefficient prediction subnetwork, and the fourth pollution emission coefficient prediction subnetwork outputs the predicted pollution emission coefficient of the yellow phosphorus enterprises;

[0051] The historical phosphorus emission data of the phosphogypsum storage enterprises in the first training set are used as the input of the fifth pollution emission coefficient prediction subnetwork, and the fifth pollution emission coefficient prediction subnetwork outputs the predicted pollution emission coefficient of the phosphogypsum storage enterprises;

[0052] S2.3: Calculate the LSTM first loss according to the input and output of the first pollution emission coefficient prediction subnetwork, train the first pollution emission coefficient prediction subnetwork according to the LSTM first loss, and stop training when the LSTM first loss is the smallest or the number of iterations is the largest, to obtain the trained first pollution emission coefficient prediction subnetwork;

[0053] Calculate the LSTM second loss according to the input and output of the second pollution emission coefficient prediction subnetwork, train the second pollution emission coefficient prediction subnetwork according to the LSTM second loss, and stop training when the LSTM second loss is the smallest or the number of iterations is the largest, to obtain a trained second pollution emission coefficient prediction subnetwork;

[0054] Calculate the LSTM third loss according to the input and output of the third pollution emission coefficient prediction subnetwork, train the third pollution emission coefficient prediction subnetwork according to the LSTM third loss, and stop training when the LSTM third loss is the smallest or the number of iterations is the largest, to obtain a trained third pollution emission coefficient prediction subnetwork;

[0055] Calculate the LSTM fourth loss according to the input and output of the fourth pollution emission coefficient prediction subnetwork, train the fourth pollution emission coefficient prediction subnetwork according to the LSTM fourth loss, and stop training when the LSTM fourth loss is the smallest or the number of iterations is the largest, to obtain a trained fourth pollution emission coefficient prediction subnetwork;

[0056] Calculate the LSTM fifth loss according to the input and output of the fifth pollution emission coefficient prediction subnetwork, train the fifth pollution emission coefficient prediction subnetwork according to the LSTM fifth loss, and stop training when the LSTM fifth loss is the smallest or the number of iterations is the largest, to obtain a trained fifth pollution emission coefficient prediction subnetwork;

[0057] S2.4: The trained first pollution emission coefficient prediction subnetwork, the trained second pollution emission coefficient prediction subnetwork, the trained third pollution emission coefficient prediction subnetwork, the trained fourth pollution emission coefficient prediction subnetwork and the trained fifth pollution emission coefficient prediction subnetwork constitute a trained pollution emission coefficient prediction network;

[0058] The following is an introduction to the LSTM layer processing flow. The LSTM workflow is carried out through three gates and cell states. The three gates in LSTM are the forget gate, input gate, and output gate.

[0059] The forget gate is used to determine which information should be forgotten from the cell state, and the forget gate output is controlled by the sigmoid function. The output value of the forget gate output is a value between 0 and 1.

[0060] The input gate is used to determine which new information will be stored in the cell state. The input gate consists of two parts: a sigmoid layer, which is used to decide which values ​​need to be updated, and a tanh layer, which is used to generate new candidate values ​​and add them to the cell state. The cell state is then updated based on the information from the forget gate and the input gate.

[0061] The output gate is used to determine which part of the cell state will be output to the hidden state at the next time step.

[0062] The gate functions and state transfer process in the LSTM layer are expressed as follows:

[0063] f t =σ(W f ·[h t-1 , x t ]+b f )#(2)

[0064] i t =σ(W i ·[h t-1 , x t]+b i )#(3)

[0065]

[0066] O t =σ(W o ·[h t-1 , x t ]+b o )#(6)

[0067] h t =O t *tanh(C t )#(7)

[0068] Among them, f t represents the output vector of the forget gate at time t, σ is the Sigmoid activation function, W f Represents the forget gate weight matrix H t-1 represents the hidden state at time t, x t represents the input of the LSTM layer at time t; b f It is the offset top of the forget gate;

[0069] i t represents the output vector of the input gate at time t, b i is the bias term of the input gate; W i is the weight matrix of the input gate; b i The input to the gate is the bias term;

[0070] is the candidate cell state at time t, W c is the weight matrix of the candidate cells, tanh represents the hyperbolic tangent activation function, b c is the bias term of the candidate cell;

[0071] C t represents the cell state after the update at time t, C t-1 represents the cell state at time t-1; * represents the element product,

[0072] O t represents the output vector of the output gate at time t, W o represents the weight matrix of the output gate, b o represents the bias term of the output gate;

[0073] The LSTM network is trained according to the input and output of the LSTM network to obtain a trained pollution emission coefficient prediction network; the specific process is:

[0074] The first training set is divided into: LSTM training set: the first 70% of the data, LSTM validation set: the next 15%, LSTM test set: the last 15%. Pre-training is performed through the LSTM training set, and the loss function is calculated through the LSTM validation set to adjust the hyperparameters. The LSTM test set is used to evaluate the performance of the model. The trained LSTM model can predict the pollution emission coefficient at future moments, which can be further used to calculate the emission legacy of total phosphorus.

[0075] The other steps and parameters are the same as those in the first to third embodiments.

[0076] Specific implementation method 5: This implementation method is different from specific implementation methods 1 to 4 in that:

[0077] The loss function of the LSTM loss is RMSE,

[0078] RMSE is the root mean square error, and the calculation formula is well known to those skilled in the art.

[0079] In addition, the training indicator for the accuracy of the model in this scheme can also be the coefficient of determination, or other losses such as mean absolute error;

[0080] The other steps and parameters are the same as those in Specific Embodiments 1 to 4.

[0081] Specific implementation method 6: This implementation method is different from specific implementation methods 1 to 5 in that:

[0082] The emission occurrence coefficient prediction network in S3 includes a first emission occurrence coefficient prediction subnetwork, a second emission occurrence coefficient prediction subnetwork, a third emission occurrence coefficient prediction subnetwork, a fourth emission occurrence coefficient prediction subnetwork, a fifth emission occurrence coefficient prediction subnetwork,

[0083] The first emission occurrence coefficient prediction subnetwork, the second emission occurrence coefficient prediction subnetwork, the third emission occurrence coefficient prediction subnetwork, the fourth emission occurrence coefficient prediction subnetwork, and the fifth emission occurrence coefficient prediction subnetwork are all FNN networks;

[0084] Feedforward Neural Network (FNN) is a basic artificial neural network structure, which consists of multiple layers, including an input layer, one or more hidden layers, and an output layer. Information flows in only one direction in the network, from the input layer to the output layer. The neurons in each layer are connected to the neurons in the next layer, but there is no connection between neurons in the same layer. It is a network well known to those in the field;

[0085] ; Other steps and parameters are the same as those in Specific Implementation Methods 1 to 5.

[0086] Specific implementation method 7: This implementation method is different from specific implementation methods 1 to 6 in that:

[0087] In S3, a second training set is obtained, and the emission occurrence coefficient prediction network is trained according to the second training set to obtain a trained emission occurrence coefficient prediction network; the specific process is:

[0088] S3.1: Acquire emission data of three phosphorus enterprises, pre-process the emission data of three phosphorus enterprises, and obtain the pre-processed emission data of three phosphorus enterprises as the second training set;

[0089] The three phosphorus enterprise emission data include: historical emission data of phosphate mine enterprises, historical emission data of phosphate fertilizer enterprises, historical emission data of phosphorus-containing pesticide enterprises, historical emission data of yellow phosphorus enterprises, and historical emission data of phosphogypsum storage enterprises;

[0090] The data preprocessing process is: normalizing, standardizing, and cleaning the historical phosphorus emission data in turn.

[0091] The historical phosphorus emission data can be obtained through annual reports of enterprises, historical monitoring or third-party monitoring agencies; the normalization, standardization and data cleaning processes are well known to those skilled in the art;

[0092] S3.2: The historical emission data of the phosphate mining enterprises in the second training set is used as the input of the first emission occurrence coefficient prediction subnetwork, and the first emission occurrence coefficient prediction subnetwork outputs the predicted emission occurrence coefficient of the phosphate mining enterprises;

[0093] The historical emission data of phosphate fertilizer enterprises in the second training set are used as the input of the second emission occurrence coefficient prediction subnetwork, and the second emission occurrence coefficient prediction subnetwork outputs the predicted emission occurrence coefficient of the phosphate fertilizer enterprises;

[0094] The historical emission data of phosphorus-containing pesticide enterprises in the second training set are used as the input of the third emission occurrence coefficient prediction subnetwork, and the third emission occurrence coefficient prediction subnetwork outputs the predicted emission occurrence coefficient of phosphorus-containing pesticide enterprises;

[0095] The historical emission data of yellow phosphorus enterprises in the second training set are used as the input of the fourth emission occurrence coefficient prediction subnetwork, and the fourth emission occurrence coefficient prediction subnetwork outputs the predicted emission occurrence coefficient of yellow phosphorus enterprises;

[0096] The historical emission data of the phosphogypsum storage enterprises in the second training set are used as the input of the fifth emission occurrence coefficient prediction subnetwork, and the fifth emission occurrence coefficient prediction subnetwork outputs the predicted emission occurrence coefficient of the phosphogypsum storage enterprises;

[0097] S3.3: Calculate the first loss of FNN according to the input and output of the first emission occurrence coefficient prediction subnetwork, train the first emission occurrence coefficient prediction subnetwork according to the first loss of FNN, and stop training when the first loss of FNN is the smallest or the number of iterations is the largest, to obtain the trained first emission occurrence coefficient prediction subnetwork;

[0098] Calculate the FNN second loss according to the input and output of the second emission occurrence coefficient prediction subnetwork, train the second emission occurrence coefficient prediction subnetwork according to the FNN second loss, stop training when the FNN second loss is minimum or the number of iterations is maximum, and obtain the trained second emission occurrence coefficient prediction subnetwork;

[0099] Calculate the FNN third loss according to the input and output of the third emission occurrence coefficient prediction subnetwork, train the third emission occurrence coefficient prediction subnetwork according to the FNN third loss, stop training when the FNN third loss is minimum or the number of iterations is maximum, and obtain the trained third emission occurrence coefficient prediction subnetwork;

[0100] Calculate the fourth loss of FNN according to the input and output of the fourth emission coefficient prediction subnetwork, train the fourth emission coefficient prediction subnetwork according to the fourth loss of FNN, and stop training when the fourth loss of FNN is the smallest or the number of iterations is the largest, to obtain a trained fourth emission coefficient prediction subnetwork;

[0101] Calculate the fifth loss of FNN according to the input and output of the fifth emission coefficient prediction subnetwork, train the fifth emission coefficient prediction subnetwork according to the fifth loss of FNN, stop training when the fifth loss of FNN is the smallest or the number of iterations is the largest, and obtain the trained fifth emission coefficient prediction subnetwork;

[0102] S3.4: The first emission occurrence coefficient prediction subnetwork, the second emission occurrence coefficient prediction subnetwork, the third emission occurrence coefficient prediction subnetwork, the fourth emission occurrence coefficient prediction subnetwork, and the fifth emission occurrence coefficient prediction subnetwork constitute a trained emission occurrence coefficient prediction network

[0103] The first to fifth FNN losses are RMSE

[0104] The other steps and parameters are the same as those in Specific Embodiments 1 to 6.

[0105] Specific implementation eight: This implementation differs from specific implementations one to seven in that:

[0106] The historical emission data of phosphate mining enterprises in S3.1 include: the total phosphorus emission TP1 of phosphate mining enterprises, the phosphorus content in dust emitted by phosphate mining enterprises C1, the dust volume Q1 generated by daily ore mining of phosphate mining enterprises; the phosphorus concentration in mine water C2; the flow rate of mine water Q2; the phosphorus concentration in phosphate tailings C3; the emission volume Q3 of phosphate tailings; the phosphorus concentration in phosphorus-containing sludge C4; the emission volume Q4 of phosphorus-containing sludge; the phosphorus concentration C5 in flushing wastewater; the emission volume Q5 of flushing wastewater;

[0107] The historical emission data of phosphate fertilizer enterprises include: the total phosphorus emission TP2 of phosphorus-containing pesticide enterprises; the concentration of soluble phosphate in phosphogypsum produced by phosphate fertilizer enterprises C6; the total mass of phosphogypsum produced by phosphate fertilizer enterprises M6; the phosphorus concentration of wastewater after treatment produced by phosphate fertilizer enterprises C7; the flow rate of treated wastewater produced by phosphate fertilizer enterprises Q7; the total phosphorus concentration of phosphorus-containing dust produced by phosphate fertilizer enterprises C8; the total mass of phosphorus-containing dust discharged into the atmosphere by phosphate fertilizer enterprises M8;

[0108] The historical emission data of phosphorus-containing pesticide enterprises include: the total phosphorus emission of phosphorus pesticide enterprises TP3; the phosphorus concentration in phosphorus-containing wastewater in the synthesis process of phosphorus-containing pesticides phosphate esterification C9; the emission of phosphorus-containing wastewater in the synthesis process of phosphorus-containing pesticides phosphate esterification Q9; the phosphorus concentration in phosphorus-containing wastewater in the synthesis process of phosphite phosphorus-containing pesticides C 10 ; Discharge of phosphorus-containing wastewater in the synthesis process of phosphite-containing pesticides Q 10 ; Phosphorus concentration in phosphorus-containing wastewater during the synthesis of phosphine-based pesticides C 11 ; Discharge of phosphorus-containing wastewater in the synthesis process of phosphine-based pesticides Q 11 ; Phosphorus concentration in phosphorus-containing wastewater during the synthesis of organophosphorus pesticides C 12 ; Discharge of phosphorus-containing wastewater in the synthesis process of organophosphorus pesticides Q 12 ;

[0109] The historical emission data of yellow phosphorus enterprises include: the total phosphorus emission TP4 of yellow phosphorus enterprises; the phosphorus concentration C 13 ; Emission flow rate Q of phosphorus-containing gas in yellow phosphorus electric furnace process 13 ;

[0110] The historical emission data of the phosphogypsum storage enterprises include: the total phosphorus emission TP5 of the phosphogypsum storage enterprises; the phosphorus concentration C of the phosphorus-containing wastewater in the drainage and slope protection process of the phosphogypsum storage yard; 14 ; The amount of phosphorus-containing wastewater discharged during drainage and slope protection of phosphogypsum storage yard Q 14 ; Phosphorus concentration in leachate from phosphogypsum storage enterprise yard C 15 ; Flow rate Q of leachate from the phosphogypsum storage enterprise yard 15 ;

[0111] The other steps and parameters are the same as those in Specific Embodiments 1 to 7.

[0112] Specific implementation method 9: This implementation method is different from specific implementation methods 1 to 8 in that:

[0113] The emission occurrence coefficients of phosphate mining enterprises in S3.2 include: dust settling rate R1; proportion of dust entering water bodies F1; proportion of mine water discharge meeting standards after treatment R2; proportion of phosphate tailings entering water bodies through runoff or seepage R3; proportion of phosphorus-containing sludge entering water bodies through runoff or seepage R4; proportion of flushing wastewater entering water bodies through surface runoff R5;

[0114] The emission occurrence coefficient of the phosphate fertilizer enterprise includes: the leaching rate of soluble phosphate R6; the proportion of dust washed out and entering the water body during rainwater scouring or atmospheric deposition R8;

[0115] The emission occurrence coefficient of the phosphorus-containing pesticide enterprise includes: the removal efficiency of residual phosphorus after wastewater treatment in the synthesis process of phosphate-containing phosphorus pesticides R9; the emission ratio of residual phosphorus in the wastewater after treatment in the synthesis process of phosphite-containing phosphorus pesticides R 10 ;

[0116] The discharge ratio R of residual phosphorus in the wastewater after treatment in the synthesis process of phosphine-based pesticides 11 ; The discharge ratio of residual phosphorus in the wastewater after treatment in the synthesis process of organophosphorus pesticides R 12 ;

[0117] The predicted emission coefficient of yellow phosphorus enterprises includes: the precipitation ratio R of phosphorus gas entering the water body during rainwater scouring or atmospheric precipitation in the yellow phosphorus electric furnace method 13 ;

[0118] The emission occurrence coefficient of the phosphogypsum storage enterprise includes: the proportion of wastewater entering the water body through runoff or leakage during the drainage and slope protection of the phosphogypsum storage yard R 14 The ratio of leachate from the phosphogypsum storage yard entering the water body R 15 ; The range of the emission occurrence coefficient obtained by the present invention is shown in Table 1.

[0119] Table 1 Emission occurrence coefficients of three phosphorus enterprises

[0120]

[0121] The other steps and parameters are the same as those in Specific Embodiments 1 to 8.

[0122] Specific implementation method 10: This implementation method is different from specific implementation methods 1 to 9 in that:

[0123] The phosphorus pollution load calculation model in S4 is expressed by the formula:

[0124]

[0125] In the formula, C t represents the phosphorus pollution load value of the monitored watershed; t represents the time variable; ΔTP i TP represents the total phosphorus legacy discharged historically by the i-th triphosphorus enterprise; i represents the total phosphorus emission of the i-th three-phosphorus enterprise; V represents the volume of water in the monitored basin; k1 represents the sedimentation rate constant (s-1), and the general value range is: 10 -7 -10 -4 ; k2 represents the dilution rate constant (s-1), generally ranging from 10 -7 -10 -5 ; k3 represents the bioabsorption rate constant (s-1), generally ranging from 10 -8 -10 -5

[0126] i=1 indicates a phosphate mining enterprise; i=2 indicates a phosphate fertilizer enterprise; i=3 indicates a phosphorus-containing pesticide enterprise; i=4 indicates a yellow phosphorus enterprise; and i=5 indicates a phosphogypsum storage enterprise;

[0127]

[0128] Where ΔTP1(t) represents the total phosphorus legacy discharged by the phosphate mining enterprises in history; j represents the period number, n represents the total number of periods, represents the pollution emission coefficient of the phosphate mining enterprise in period j; the periods in the present invention are divided according to years, and j=1 represents the first year of production;

[0129]

[0130] In the formula, ΔTP2(t) represents the total phosphorus legacy discharged by phosphate fertilizer enterprises in history; represents the pollution emission coefficient of the phosphate fertilizer enterprise in period j;

[0131]

[0132] In the formula, ΔTP3(t) represents the total phosphorus legacy discharged historically by phosphorus-containing pesticide enterprises; represents the pollution emission coefficient of phosphorus pesticide enterprises in period j;

[0133]

[0134] In the formula, ΔTP4(t) represents the total phosphorus legacy discharged historically by yellow phosphorus enterprises; represents the pollution emission coefficient of the yellow phosphorus enterprise in period j

[0135]

[0136] In the formula, ΔTP5(t) represents the total phosphorus legacy discharged historically by the phosphogypsum storage enterprise; represents the pollution emission coefficient of the phosphogypsum storage enterprise in period j;

[0137]

[0138] The environmental data of the monitoring basin obtained in the step 2 include: C1 phosphorus content in dust (unit: g / kg); Q1 dust amount generated by daily ore mining (unit: kg / d); C2 phosphorus concentration in mine water; Q2 mine water flow rate; C3 phosphorus concentration in phosphate tailings (unit: g / t); Q3 phosphorus tailings discharge (unit: t / d); C4 phosphorus concentration in phosphorus-containing sludge (unit: g / m 3 ); Q4 Phosphorus-containing sludge discharge (unit: m 3 / d); Phosphorus concentration in C5 flushing wastewater (unit: g / m 3 ); Q5 flushing wastewater discharge (unit: m 3 / day); the concentration of soluble phosphate in phosphogypsum produced by C6 phosphate fertilizer enterprises (unit: g / t); the total mass of phosphogypsum produced by M6 phosphate fertilizer enterprises (unit: t / d); the phosphorus concentration of treated wastewater produced by C7 phosphate fertilizer enterprises (unit: g / m3); the flow rate of treated wastewater produced by Q7 phosphate fertilizer enterprises (unit: m 3 / day); C8 The total phosphorus concentration in phosphorus-containing dust produced by phosphate fertilizer enterprises (unit: g / t); M8 The total mass of phosphorus-containing dust discharged into the atmosphere by phosphate fertilizer enterprises (unit: t / day); C9 The phosphorus concentration in phosphorus-containing wastewater in the synthesis process of phosphorus-containing pesticides by phosphate esterification (unit: g / m 3 ); Q9 Phosphorus-containing wastewater discharge in the synthesis process of phosphorus-containing pesticides (unit: m 3 / day); C 10 Phosphorus concentration in phosphorus-containing wastewater during the synthesis of phosphite-containing pesticides (unit: g / m 3 );Q 10 Discharge of phosphorus-containing wastewater in the synthesis process of phosphite-containing pesticides (unit: m 3 / day); C 11 Phosphorus concentration in phosphorus-containing wastewater during the synthesis of phosphine-based pesticides (unit: g / m 3 );Q 11 Discharge of phosphorus-containing wastewater in the synthesis process of phosphine-based pesticides (unit: m 3 / day); C 12 Phosphorus concentration in phosphorus-containing wastewater during the synthesis of organophosphorus pesticides (unit: g / m 3 );Q 12Discharge of phosphorus-containing wastewater from the synthesis process of organophosphorus pesticides (unit: m 3 / day); C 13 Phosphorus concentration in phosphorus-containing gas in yellow phosphorus electric furnace method (unit: g / m 3 );Q 13 Emission flow of phosphorus-containing gas in the yellow phosphorus electric furnace process (unit: m 3 / day); C 14 Phosphorus concentration of phosphorus-containing wastewater during drainage and slope protection of phosphogypsum storage yard (unit: mg / L); Q 14 The amount of phosphorus-containing wastewater discharged during drainage and slope protection of phosphogypsum storage yard (unit: m 3 / day); C 15 Phosphorus concentration in the leachate of gypsum storage yard (unit: g / m 3 );Q 15 Flow rate of leachate from phosphogypsum storage yard (unit: m 3 / day);

[0139] The other steps and parameters are the same as those in Specific Implementations 1 to 9.

[0140] The above only describes the preferred embodiments of the present invention. It should be understood that the present invention is not limited to the above specific implementation methods. Although the present invention has been disclosed as above in the preferred embodiments, it is not used to limit the present invention. Any technician familiar with the profession can make some changes or modify the technical contents disclosed above into equivalent embodiments with equivalent changes without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent replacement and improvement made to the above embodiments without departing from the content of the technical solution of the present invention, based on the technical essence of the present invention, within the spirit and principles of the present invention, still fall within the protection scope of the technical solution of the present invention.

Claims

1. A method for monitoring and evaluating phosphorus pollution load at a watershed scale, characterized in that: include: S1. Obtain the census data of phosphorus pollution sources of three phosphorus enterprises in the monitoring basin; determine N monitoring points in the monitoring basin according to the census data of phosphorus pollution sources of three phosphorus enterprises in the monitoring basin; deploy a sensor at each monitoring point; S2: Obtain a first training set, and train a pollution emission coefficient prediction network according to the first training set to obtain a trained pollution emission coefficient prediction network; S3: Obtain a second training set, and train the emission occurrence coefficient prediction network according to the second training set to obtain a trained emission occurrence coefficient prediction network; S4: construct a phosphorus pollution load accounting model; construct a phosphorus pollution monitoring system based on the data acquisition module, the trained pollution emission coefficient prediction network, the trained emission occurrence coefficient prediction network, and the phosphorus pollution load accounting model; S5: Acquire the environmental data of the monitored river basin in real time through sensors, input the acquired environmental data of the monitored river basin into the phosphorus pollution monitoring system, and obtain the phosphorus pollution load value of the monitored river basin.

2. The method for monitoring and evaluating phosphorus pollution load at a watershed scale according to claim 1, characterized in that: In S1, the census data of phosphorus pollution sources of three phosphorus enterprises in the monitoring basin are obtained; N monitoring points are determined in the monitoring basin according to the census data of phosphorus pollution sources of three phosphorus enterprises in the monitoring basin; the specific process is: S1.1: Collect the phosphorus pollution source emission data of all three phosphorus enterprises in the monitoring basin and the geographic information data in the monitoring basin; The three phosphorus enterprises include: phosphate mine enterprises, phosphate fertilizer enterprises, phosphorus-containing pesticide enterprises, yellow phosphorus enterprises, and phosphogypsum storage enterprises; S1.2: According to S1.1, the phosphorus pollution source emission data of all three phosphorus enterprises in the monitoring basin and the geographic information data in the monitoring basin are collected to analyze the phosphorus pollution sources and obtain N monitoring points.

3. The method for monitoring and evaluating phosphorus pollution load at a watershed scale according to claim 2, characterized in that: The pollution emission coefficient prediction network in S2 includes: a first pollution emission coefficient prediction subnetwork, a second pollution emission coefficient prediction subnetwork, a third pollution emission coefficient prediction subnetwork, a fourth pollution emission coefficient prediction subnetwork, a fifth pollution emission coefficient prediction subnetwork, The first pollution emission coefficient prediction subnetwork, the second pollution emission coefficient prediction subnetwork, the third pollution emission coefficient prediction subnetwork, the fourth pollution emission coefficient prediction subnetwork, and the fifth pollution emission coefficient prediction subnetwork are all LSTM networks.

4. The method for monitoring and evaluating phosphorus pollution load at a watershed scale according to claim 3, characterized in that: In S2, a first training set is obtained, and a pollution emission coefficient prediction network is trained according to the first training set to obtain a trained pollution emission coefficient prediction network; the specific process is: S2.1: Obtain historical phosphorus emission data; preprocess the historical phosphorus emission data to obtain preprocessed historical phosphorus emission data, and obtain a first training set based on the preprocessed historical phosphorus emission data; The historical phosphorus emission data include: historical phosphorus emission data of phosphate mines, historical phosphorus emission data of phosphate fertilizers, historical phosphorus emission data of phosphorus-containing pesticides, historical phosphorus emission data of yellow phosphorus enterprises, and historical phosphorus emission data of phosphogypsum storage enterprises; S2.2: The historical phosphorus emission data of the phosphate mining enterprises in the first training set is used as the input of the first pollution emission coefficient prediction subnetwork, and the first pollution emission coefficient prediction subnetwork outputs the predicted pollution emission coefficient of the phosphate mining enterprises; The historical phosphorus emission data of the phosphate fertilizer enterprises in the first training set are used as the input of the second pollution emission coefficient prediction subnetwork, and the second pollution emission coefficient prediction subnetwork outputs the predicted pollution emission coefficient of the phosphate fertilizer enterprises; The historical phosphorus emission data of the phosphorus-containing pesticide enterprises in the first training set are used as the input of the third pollution emission coefficient prediction subnetwork, and the third pollution emission coefficient prediction subnetwork outputs the predicted pollution emission coefficient of the phosphorus-containing pesticide enterprises; The historical phosphorus emission data of yellow phosphorus enterprises in the first training set are used as the input of the fourth pollution emission coefficient prediction subnetwork, and the fourth pollution emission coefficient prediction subnetwork outputs the predicted pollution emission coefficient of the yellow phosphorus enterprises; The historical phosphorus emission data of the phosphogypsum storage enterprises in the first training set are used as the input of the fifth pollution emission coefficient prediction subnetwork, and the fifth pollution emission coefficient prediction subnetwork outputs the predicted pollution emission coefficient of the phosphogypsum storage enterprises; S2.3: Calculate the LSTM first loss according to the input and output of the first pollution emission coefficient prediction subnetwork, train the first pollution emission coefficient prediction subnetwork according to the LSTM first loss, and stop training when the LSTM first loss is the smallest or the number of iterations is the largest, to obtain the trained first pollution emission coefficient prediction subnetwork; Calculate the LSTM second loss according to the input and output of the second pollution emission coefficient prediction subnetwork, train the second pollution emission coefficient prediction subnetwork according to the LSTM second loss, and stop training when the LSTM second loss is the smallest or the number of iterations is the largest, to obtain a trained second pollution emission coefficient prediction subnetwork; Calculate the LSTM third loss according to the input and output of the third pollution emission coefficient prediction subnetwork, train the third pollution emission coefficient prediction subnetwork according to the LSTM third loss, and stop training when the LSTM third loss is the smallest or the number of iterations is the largest, to obtain a trained third pollution emission coefficient prediction subnetwork; Calculate the LSTM fourth loss according to the input and output of the fourth pollution emission coefficient prediction subnetwork, train the fourth pollution emission coefficient prediction subnetwork according to the LSTM fourth loss, and stop training when the LSTM fourth loss is the smallest or the number of iterations is the largest, to obtain a trained fourth pollution emission coefficient prediction subnetwork; Calculate the LSTM fifth loss according to the input and output of the fifth pollution emission coefficient prediction subnetwork, train the fifth pollution emission coefficient prediction subnetwork according to the LSTM fifth loss, and stop training when the LSTM fifth loss is the smallest or the number of iterations is the largest, to obtain a trained fifth pollution emission coefficient prediction subnetwork; S2.4: The trained first pollution emission coefficient prediction subnetwork, the trained second pollution emission coefficient prediction subnetwork, the trained third pollution emission coefficient prediction subnetwork, the trained fourth pollution emission coefficient prediction subnetwork and the trained fifth pollution emission coefficient prediction subnetwork constitute a trained pollution emission coefficient prediction network.

5. The method for monitoring and evaluating phosphorus pollution load at a watershed scale according to claim 4, characterized in that: The loss functions of the LSTM first loss, LSTM second loss, LSTM third loss, LSTM fourth loss, and LSTM fifth loss are mean square error, cross entropy, or logarithmic loss.

6. The method for monitoring and evaluating phosphorus pollution load at a watershed scale according to claim 5, characterized in that: The emission occurrence coefficient prediction network in S3 includes a first emission occurrence coefficient prediction subnetwork, a second emission occurrence coefficient prediction subnetwork, a third emission occurrence coefficient prediction subnetwork, a fourth emission occurrence coefficient prediction subnetwork, a fifth emission occurrence coefficient prediction subnetwork, The first emission occurrence coefficient prediction subnetwork, the second emission occurrence coefficient prediction subnetwork, the third emission occurrence coefficient prediction subnetwork, the fourth emission occurrence coefficient prediction subnetwork, and the fifth emission occurrence coefficient prediction subnetwork are all FNN networks.

7. The method for monitoring and evaluating phosphorus pollution load at a watershed scale according to claim 6, characterized in that: In S3, a second training set is obtained, and the emission occurrence coefficient prediction network is trained according to the second training set to obtain a trained emission occurrence coefficient prediction network; the specific process is: S3.1: Acquire emission data of three phosphorus enterprises, pre-process the emission data of three phosphorus enterprises, and obtain the pre-processed emission data of three phosphorus enterprises as the second training set; The three phosphorus enterprise emission data include: historical emission data of phosphate mine enterprises, historical emission data of phosphate fertilizer enterprises, historical emission data of phosphorus-containing pesticide enterprises, historical emission data of yellow phosphorus enterprises, and historical emission data of phosphogypsum storage enterprises; S3.2: The historical emission data of the phosphate mining enterprises in the second training set is used as the input of the first emission occurrence coefficient prediction subnetwork, and the first emission occurrence coefficient prediction subnetwork outputs the predicted emission occurrence coefficient of the phosphate mining enterprises; The historical emission data of phosphate fertilizer enterprises in the second training set are used as the input of the second emission occurrence coefficient prediction subnetwork, and the second emission occurrence coefficient prediction subnetwork outputs the predicted emission occurrence coefficient of the phosphate fertilizer enterprises; The historical emission data of phosphorus-containing pesticide enterprises in the second training set are used as the input of the third emission occurrence coefficient prediction subnetwork, and the third emission occurrence coefficient prediction subnetwork outputs the predicted emission occurrence coefficient of phosphorus-containing pesticide enterprises; The historical emission data of yellow phosphorus enterprises in the second training set are used as the input of the fourth emission occurrence coefficient prediction subnetwork, and the fourth emission occurrence coefficient prediction subnetwork outputs the predicted emission occurrence coefficient of yellow phosphorus enterprises; The historical emission data of the phosphogypsum storage enterprises in the second training set are used as the input of the fifth emission occurrence coefficient prediction subnetwork, and the fifth emission occurrence coefficient prediction subnetwork outputs the predicted emission occurrence coefficient of the phosphogypsum storage enterprises; S3.3: Calculate the first loss of FNN according to the input and output of the first emission occurrence coefficient prediction subnetwork, train the first emission occurrence coefficient prediction subnetwork according to the first loss of FNN, and stop training when the first loss of FNN is the smallest or the number of iterations is the largest, to obtain the trained first emission occurrence coefficient prediction subnetwork; Calculate the FNN second loss according to the input and output of the second emission occurrence coefficient prediction subnetwork, train the second emission occurrence coefficient prediction subnetwork according to the FNN second loss, stop training when the FNN second loss is minimum or the number of iterations is maximum, and obtain the trained second emission occurrence coefficient prediction subnetwork; Calculate the FNN third loss according to the input and output of the third emission occurrence coefficient prediction subnetwork, train the third emission occurrence coefficient prediction subnetwork according to the FNN third loss, stop training when the FNN third loss is minimum or the number of iterations is maximum, and obtain the trained third emission occurrence coefficient prediction subnetwork; Calculate the fourth loss of FNN according to the input and output of the fourth emission coefficient prediction subnetwork, train the fourth emission coefficient prediction subnetwork according to the fourth loss of FNN, and stop training when the fourth loss of FNN is the smallest or the number of iterations is the largest, to obtain a trained fourth emission coefficient prediction subnetwork; Calculate the fifth loss of FNN according to the input and output of the fifth emission coefficient prediction subnetwork, train the fifth emission coefficient prediction subnetwork according to the fifth loss of FNN, stop training when the fifth loss of FNN is the smallest or the number of iterations is the largest, and obtain the trained fifth emission coefficient prediction subnetwork; S3.4: The first emission occurrence coefficient prediction subnetwork, the second emission occurrence coefficient prediction subnetwork, the third emission occurrence coefficient prediction subnetwork, the fourth emission occurrence coefficient prediction subnetwork, and the fifth emission occurrence coefficient prediction subnetwork constitute a trained emission occurrence coefficient prediction network; The loss functions of the FNN first loss, the FNN second loss, the FNN third loss, the FNN fourth loss, and the FNN fifth loss are mean square error, cross entropy, or logarithmic loss.

8. The method for monitoring and evaluating phosphorus pollution load at a watershed scale according to claim 7, characterized in that: The historical emission data of phosphate mining enterprises in S3.1 include: the total phosphorus emission TP1 of phosphate mining enterprises, the phosphorus content in dust emitted by phosphate mining enterprises C1, the dust volume Q1 generated by daily ore mining of phosphate mining enterprises; the phosphorus concentration in mine water C2; the flow rate of mine water Q2; the phosphorus concentration in phosphate tailings C3; the emission volume Q3 of phosphate tailings; the phosphorus concentration in phosphorus-containing sludge C4; the emission volume Q4 of phosphorus-containing sludge; the phosphorus concentration C5 in flushing wastewater; the emission volume Q5 of flushing wastewater; The historical emission data of phosphate fertilizer enterprises include: the total phosphorus emission TP2 of phosphorus-containing pesticide enterprises; the concentration of soluble phosphate in phosphogypsum produced by phosphate fertilizer enterprises C6; the total mass of phosphogypsum produced by phosphate fertilizer enterprises M6; the phosphorus concentration of wastewater after treatment produced by phosphate fertilizer enterprises C7; the flow rate of treated wastewater produced by phosphate fertilizer enterprises Q7; the total phosphorus concentration of phosphorus-containing dust produced by phosphate fertilizer enterprises C8; the total mass of phosphorus-containing dust discharged into the atmosphere by phosphate fertilizer enterprises M8; The historical emission data of phosphorus-containing pesticide enterprises include: the total phosphorus emission of phosphorus pesticide enterprises TP3; the phosphorus concentration in phosphorus-containing wastewater in the synthesis process of phosphorus-containing pesticides phosphate esterification C9; the emission of phosphorus-containing wastewater in the synthesis process of phosphorus-containing pesticides phosphate esterification Q9; the phosphorus concentration in phosphorus-containing wastewater in the synthesis process of phosphite phosphorus-containing pesticides C 10 ; Discharge of phosphorus-containing wastewater in the synthesis process of phosphite-containing pesticides Q 10 ; Phosphorus concentration in phosphorus-containing wastewater during the synthesis of phosphine-based pesticides C 11 ; Discharge of phosphorus-containing wastewater in the synthesis process of phosphine-based pesticides Q 11 ; Phosphorus concentration in phosphorus-containing wastewater during the synthesis of organophosphorus pesticides C 12 ; Discharge of phosphorus-containing wastewater in the synthesis process of organophosphorus pesticides Q 12 ; The historical emission data of yellow phosphorus enterprises include: the total phosphorus emission TP4 of yellow phosphorus enterprises; the phosphorus concentration C 13 ; Emission flow rate Q of phosphorus-containing gas in yellow phosphorus electric furnace process 13 ; The historical emission data of the phosphogypsum storage enterprises include: the total phosphorus emission TP5 of the phosphogypsum storage enterprises; the phosphorus concentration C of the phosphorus-containing wastewater in the drainage and slope protection process of the phosphogypsum storage yard; 14 ; The amount of phosphorus-containing wastewater discharged during drainage and slope protection of phosphogypsum storage yard Q 14 ; Phosphorus concentration in leachate from phosphogypsum storage enterprise yard C 15 ; Flow rate Q of leachate from the phosphogypsum storage enterprise yard 15 .

9. The method for monitoring and evaluating phosphorus pollution load at a watershed scale according to claim 8, characterized in that: The emission occurrence coefficients of phosphate mining enterprises in S3.2 include: dust settling rate R1; proportion of dust entering water bodies F1; proportion of mine water discharge meeting standards after treatment R2; proportion of phosphate tailings entering water bodies through runoff or seepage R3; proportion of phosphorus-containing sludge entering water bodies through runoff or seepage R4; proportion of flushing wastewater entering water bodies through surface runoff R5; The emission occurrence coefficient of the phosphate fertilizer enterprise includes: the leaching rate of soluble phosphate R6; the proportion of dust washed out and entering the water body during rainwater scouring or atmospheric deposition R8; The emission occurrence coefficient of the phosphorus-containing pesticide enterprise includes: the removal efficiency of residual phosphorus after wastewater treatment in the synthesis process of phosphate-containing phosphorus pesticides R9; the emission ratio of residual phosphorus in the wastewater after treatment in the synthesis process of phosphite-containing phosphorus pesticides R 10 ; The discharge ratio R of residual phosphorus in the wastewater after treatment in the synthesis process of phosphine-based pesticides 11 ; The discharge ratio of residual phosphorus in the wastewater after treatment in the synthesis process of organophosphorus pesticides R 12 ; The predicted emission coefficient of yellow phosphorus enterprises includes: the precipitation ratio R of phosphorus gas entering the water body during rainwater scouring or atmospheric precipitation in the yellow phosphorus electric furnace method 13 ; The emission occurrence coefficient of the phosphogypsum storage enterprise includes: the proportion of wastewater entering the water body through runoff or leakage during the drainage and slope protection of the phosphogypsum storage yard R 14 The ratio of leachate from the phosphogypsum storage yard entering the water body R 15 .

10. The method for monitoring and evaluating phosphorus pollution load at a watershed scale according to claim 9, characterized in that: The phosphorus pollution load calculation model in S4 is expressed by the formula: In the formula, C t represents the phosphorus pollution load value of the monitored watershed; t represents the time variable; ΔTP i TP represents the total phosphorus legacy discharged historically by the i-th triphosphorus enterprise; i represents the total phosphorus emission of the i-th triphosphate enterprise; V represents the volume of the monitored watershed; k1 represents the deposition rate constant (s-1); k2 represents the dilution rate constant (s-1); k3 represents the biological absorption rate constant (s-1), i=1 indicates a phosphate mining enterprise; i=2 indicates a phosphate fertilizer enterprise; i=3 indicates a phosphorus-containing pesticide enterprise; i=4 indicates a yellow phosphorus enterprise; and i=5 indicates a phosphogypsum storage enterprise; Where ΔTP1(t) represents the total phosphorus legacy discharged by the phosphate mining enterprises in history; j represents the period number, n represents the total number of periods, represents the pollution emission coefficient of the phosphate mining enterprise in period j; In the formula, ΔTP2(t) represents the total phosphorus legacy discharged by phosphate fertilizer enterprises in history; represents the pollution emission coefficient of the phosphate fertilizer enterprise in period j; In the formula, ΔTP3(t) represents the total phosphorus legacy discharged historically by phosphorus-containing pesticide enterprises; represents the pollution emission coefficient of phosphorus-containing pesticide enterprises in period j; In the formula, ΔTP4(t) represents the total phosphorus legacy discharged historically by yellow phosphorus enterprises; represents the pollution emission coefficient of the yellow phosphorus enterprise in period j In the formula, ΔTP5(t) represents the total phosphorus legacy discharged historically by the phosphogypsum storage enterprise; represents the pollution emission coefficient of the phosphogypsum storage enterprise in period j;

Images