Incoming water control and tracing method and system, medium and program product

Abstract

The invention discloses an incoming water control and tracing method and system, a medium and a program product, and relates to the technical field of sewage management, the method comprises the following steps: a uniform water sample detection step: sampling a drainage pipeline of each incoming water enterprise at regular time according to a preset period to form a uniform water sample; screening the water quality data of the uniform water sample through a pre-detection module; if the pre-detection index exceeds a contract preset management and control threshold value, a precise detection module is started to carry out multi-parameter precise detection on the water quality; performing grading treatment according to a fine detection result; an instantaneous water sample detection step: deploying dynamic monitoring nodes in a total water inlet pipeline of a sewage treatment plant, and monitoring a plurality of early warning indexes in real time; when any early warning index exceeds a preset multiple of a dynamic threshold value, whole-network instantaneous sampling is triggered; collecting instantaneous water samples of the drainage branch pipes of all the incoming water enterprises, and determining the drainage branch pipes with overproof incoming water; when the water quality detection exceeds the standard, triggering a standard-exceeding traceability instruction; according to the invention, a high-efficiency and high-precision sewage monitoring system is established.

Description

Technical Field

[0001] This application relates to the field of wastewater management technology, and in particular to a method, system, medium, and program product for controlling and tracing incoming water. Background Technology

[0002] The existing sewage treatment system collects sewage from multiple sewage treatment plants through a centralized pipe network for unified treatment. Although sewage discharge standards are agreed upon with partner companies, due to the characteristics of the centralized end-of-pipe treatment process, some sewage treatment plants may illegally discharge sewage or discharge sewage into the sewage treatment plant that exceeds the sewage discharge standards agreed upon in the contract. This leads to an increase in the overall sewage treatment cost and a reduction in profit margin.

[0003] However, existing wastewater treatment systems cannot trace the source of pollutants exceeding standards after the wastewater is mixed, making it difficult to accurately locate illegal discharges. The lack of real-time monitoring means that evidence cannot be collected promptly after violations occur, which can encourage companies to take chances, exacerbate illegal discharges, and worsen the situation of uncollected wastewater treatment fees. Furthermore, some companies may intermittently discharge high-concentration wastewater (such as electroplating wastewater), causing disruption to the biological treatment system. Therefore, establishing a highly efficient and accurate wastewater monitoring system to address the technical deficiencies of existing wastewater treatment systems has practical application value and economic benefits. Summary of the Invention

[0004] In order to establish a highly efficient and accurate wastewater monitoring system and address the technical deficiencies of existing wastewater treatment systems, this application provides a method, system, medium, and program product for incoming water control and source tracing.

[0005] Firstly, the objective of this invention is achieved through the following technical solution: A method for controlling and tracing the source of incoming water includes: Homogeneous water sample testing steps: Sampling is performed on the drainage pipes of each water-receiving enterprise at preset intervals to form uniform water samples. The water quality data of the uniform water samples is screened by the pre-detection module. If the pre-detection indicators exceed the preset control thresholds in the contract, the fine detection module is activated to perform multi-parameter precise detection of the water quality. According to the fine detection results, graded handling is carried out: if the standard is exceeded, the additional fee procedure is initiated, and the disputed sample is sealed at the same time. Instantaneous water sample testing steps: Dynamic monitoring nodes are deployed in the main inlet pipeline of the sewage treatment plant to monitor multiple early warning indicators in real time. When any early warning indicator exceeds a preset multiple of the dynamic threshold, instantaneous sampling of the entire network is triggered. Instantaneous water samples are collected from the drainage branch pipes of each water-receiving enterprise. The concentration gradient difference between the instantaneous water sample and the corresponding uniform water sample is compared to determine the drainage branch pipe of the water exceeding the standard. When water quality exceeds the standard, a source tracing command is triggered.

[0006] By adopting the above-mentioned technical solution, which combines dual-track detection of uniform water sample detection and instantaneous water sample detection with intelligent source tracing and automatic execution, this application adopts a proactive wastewater management approach with full-process supervision, unlike the end-of-pipe treatment of traditional technical solutions. The uniform water sample detection reduces the frequency of redundant detection through a pre-inspection and fine-inspection grading mechanism, and can promptly identify enterprises that exceed the emission standards. Enterprises that exceed the emission standards will be charged for exceeding the standards. In wastewater treatment using uniform water sample detection, exceeding the standards mainly refers to the concentration of discharged wastewater exceeding the standards agreed upon in the contract. Instantaneous water sample testing utilizes a dynamic real-time control mechanism combining real-time monitoring and early warning with precise testing to detect illegal discharges by water-receiving enterprises. In actual production, wastewater treatment is limited to wastewater from cleaning electroplating tanks. Electroplating waste liquid needs to be treated by specialized treatment facilities that handle electroplating waste liquid. However, due to the extremely high cost of treating electroplating waste liquid, some enterprises illegally discharge some of it into the wastewater, resulting in extremely high concentrations. For example, if the normal discharge concentration is 50, illegal discharge may lead to concentrations as high as 5,000 or even 50,000, which will affect subsequent wastewater treatment production. Therefore, in addition to incurring additional costs, it is also necessary to shut down the influent to prevent illegally discharged wastewater from entering the equalization tank of the wastewater treatment plant and affecting wastewater treatment production. The real-time instantaneous water sampling detection in this application is a further high-precision detection measure taken when anomalies are detected in the sampling detection of the main inlet pipe. Through the real-time monitoring of the instantaneous water sampling detection steps, problematic drainage branch pipes can be identified and shut down to prevent the impact of illegally discharged wastewater on the biological treatment system (such as the death of bacteria caused by a sudden increase in heavy metal ion concentration). The treatment efficiency and the accuracy of wastewater inflow control are high. It maintains the stable compliance of the effluent quality of the wastewater treatment plant with standards (such as GB 18918). Then, combined with the source tracing process for exceeding standards, the retention of disputed samples, and the third-party testing arbitration mechanism, the fairness and compliance of water sampling detection are guaranteed. Thus, this application establishes a high-efficiency and high-precision wastewater monitoring system to solve the technical defects of existing wastewater treatment systems.

[0007] In a preferred embodiment of this application, the homogeneous water sample detection step specifically includes: A fixed volume of water sample is collected by a timed sampler, and the samples are continuously collected for a fixed period of time to form a mixed homogeneous water sample. The pre-detection module detects the pH value, oxidation-reduction potential and conductivity of the homogeneous water sample. If the values ​​exceed the contract-preset control threshold, the values ​​are within a preset multiple range of the control threshold agreed in the contract. The tiered handling based on the precision detection results includes charging additional fees at a tiered rate if the concentration exceeding the standard is within a first preset multiple of the contract value, and sending a production stoppage order to the enterprise responsible for the exceeding standard if the concentration exceeds a second preset multiple of the contract value.

[0008] By adopting the above technical solutions, the uniform water sample detection step ensures the representativeness of the samples by collecting mixed water samples with fixed volume and duration; the pre-detection module checks basic parameters such as pH value, oxidation-reduction potential and conductivity, providing a basis for preliminary screening; the graded treatment mechanism takes different measures according to the degree of exceedance, which helps to refine the management of water resource usage costs.

[0009] In a preferred embodiment of this application, the method further includes: Collect water quality characteristic data and wastewater pollutant index data of historical drainage samples from enterprises, construct drainage sample dataset and corresponding feature matrix; perform feature dimensionality reduction on the feature matrix and extract the feature vectors corresponding to the top K largest eigenvalues; The eigenvalues ​​of K eigenvectors are calculated using a neural network backpropagation algorithm and linear weighting. Set a combination of thresholds to be adjusted that includes multiple thresholds to be adjusted, and calculate the false alarm rate and false negative rate under each threshold to be adjusted; The backpropagation algorithm of the neural network is used, and a loss function based on the false positive rate and the false negative rate is used to numerically adjust multiple thresholds to be adjusted, so as to determine the optimal feature thresholds corresponding to K feature vectors. In actual water sample testing, based on the K feature vectors and the corresponding optimal feature thresholds, a preliminary water quality exceedance judgment is made on the uniform water sample and the instantaneous water sample to obtain the water quality exceedance prediction result.

[0010] By employing the above technical solutions and extracting key feature vectors, the dimensionality of data analysis can be effectively reduced, and computational efficiency improved. Combining this with neural network algorithms to optimize feature thresholds enables more accurate identification of water quality exceeding standards, reducing false alarms and missed alarms.

[0011] In a preferred embodiment of this application: the K eigenvectors are three eigenvectors α, β, and γ; the eigenvalues ​​of the eigenvectors are calculated using the following formula: N(α,β,γ)=ω1×α+ω2×β+ω3×γ Where ω1, ω2, and ω3 are the weight factors optimized by the backpropagation algorithm of the neural network.

[0012] By adopting the above technical solution, three feature vectors are used as the basis for water quality assessment to simplify the model complexity, thereby improving the processing speed of the system while ensuring accuracy.

[0013] In a preferred example of this application, the loss function constructed based on the false positive rate and the false negative rate is: Loss=λ1η1+λ2η2 Where η1 is the false alarm rate; η2 is the false negative rate; λ1 and λ2 are balance coefficients, λ1+λ2=1 and their values ​​can be adjusted; The loss function is solved using a neural network backpropagation algorithm, and the weight factors ω1, ω2 and ω3 are updated to obtain the minimum value of the loss function corresponding to the plurality of thresholds to be adjusted, so as to determine the optimal feature threshold.

[0014] By adopting the above technical solution, a loss function based on false alarm rate and false negative rate is introduced. By adjusting the weight factor to minimize the loss function, it is ensured that the selection of the optimal feature threshold takes into account both reducing the possibility of false alarms and avoiding the omission of actual exceeding events.

[0015] In a preferred embodiment of this application, the optimal feature threshold satisfies the condition that both the false positive rate and the false negative rate are lower than a preset ratio value; The water quality exceedance prediction results include water quality exceeding the standard and water quality not exceeding the standard. The water quality exceeding the standard is defined as the feature value of any feature vector of the K feature vectors exceeding the corresponding optimal feature threshold; the water quality not exceeding the standard is defined as the feature value of all feature vectors of the K feature vectors not exceeding the corresponding optimal feature threshold.

[0016] By adopting the above technical solution, the criteria for determining water quality exceeding standards have been clarified, which facilitates operation and implementation.

[0017] In a preferred example of this application, the formula for calculating the false alarm rate η1 is as follows: Wherein, C represents the number of false alarm water samples in the drainage sample dataset; A represents water samples in the drainage sample dataset that may not exceed the standard; and B represents water samples in the drainage sample dataset that may exceed the standard. The formula for calculating the false negative rate η2 is as follows: Where D represents the number of water samples that were missed during screening, i.e., the number of water samples that met the standards after testing B potentially exceeding the standards; A and B are obtained by comparing the threshold to be adjusted with the drainage sample dataset.

[0018] By adopting the above technical solution, the accuracy and effectiveness of the system can be objectively evaluated by comparing water samples that may exceed or not exceed the standards in the drainage sample dataset.

[0019] Secondly, the objective of this invention is achieved through the following technical solution: A water inflow control and traceability system, the system comprising: The uniform water sample collection module is used to periodically sample the drainage pipes of each water-consuming enterprise according to a preset cycle to form a uniform water sample. The pre-detection module is used to perform preliminary screening of water quality data from homogeneous water samples; The precision detection module is activated when the pre-detection module detects that water quality indicators exceed the pre-set control thresholds in the contract. It is used to perform precise multi-parameter detection of water quality. The graded disposal module is used to perform graded disposal operations based on the results of the precision test. If it is confirmed that the standard is exceeded, the additional fee procedure will be initiated, and the sample will be sealed simultaneously if there is a disputed sample. Dynamic monitoring nodes are deployed in the main inlet pipeline of the wastewater treatment plant to monitor multiple early warning indicators in real time; The instantaneous sampling trigger module is used to trigger instantaneous sampling across the entire network when any early warning indicator exceeds a preset multiple of the dynamic threshold; the instantaneous water sample acquisition module is used to collect instantaneous water samples from the drainage branch pipes of each water-producing enterprise. The concentration gradient comparison analysis module is used to analyze the concentration gradient difference between instantaneous water samples and corresponding homogeneous water samples to determine the drainage branch pipes of water exceeding the standard. The excessive water quality traceability instruction module is triggered when water quality tests show excessive levels of pollutants, and is used to initiate the traceability process. Thirdly, the objective of this invention is achieved through the following technical solution: A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the aforementioned water inflow control and traceability method.

[0020] Fourthly, the objective of this invention is achieved through the following technical solution: A computer program product includes a computer program / instructions that, when executed by a processor, implement the steps of a water source control and traceability method as described above.

[0021] In summary, this application includes at least one of the following beneficial technical effects: 1. Incoming water control and source tracing methods, including uniform water sample testing and instantaneous water sample testing. By periodically sampling the drainage pipes of each enterprise to form uniform water samples for pre-testing, water quality problems can be detected early. A precision testing module is used to further analyze samples exceeding standards, providing more detailed water quality information and supporting tiered treatment decisions. The instantaneous water sample testing step deploys monitoring nodes in the main influent pipeline of the wastewater treatment plant to monitor and warn indicators in real time. When an anomaly occurs, it triggers full-network sampling. By comparing the concentration gradient difference between instantaneous and uniform water samples, the source of pollution can be quickly located. 2. The method of building a feature matrix using historical data can effectively reduce the dimensions of data analysis and improve computational efficiency by performing feature dimensionality reduction on these data and extracting key feature vectors. The preliminary screening of all samples based on the water quality exceedance prediction results is beneficial to saving monitoring costs. Attached Figure Description

[0022] Figure 1This is a flowchart of a water inflow control and source tracing method according to an embodiment of this application; Figure 2 This is another flowchart of a water inflow control and tracing method in one embodiment of this application. Detailed Implementation

[0023] The present application will be further described in detail below with reference to the accompanying drawings.

[0024] In one embodiment, such as Figure 1 As shown, this application discloses a method for controlling and tracing the source of incoming water, which specifically includes the following steps: S1: Homogeneous water sample testing procedure: Sampling is performed on the drainage pipes of each water-receiving enterprise at preset intervals to form uniform water samples. The water quality data of the uniform water samples is screened by the pre-detection module. If the pre-detection indicators exceed the preset control thresholds in the contract, the fine detection module is activated to perform multi-parameter precise detection of the water quality. According to the fine detection results, graded handling is carried out: if the standard is exceeded, the additional fee procedure is initiated, and the disputed sample is sealed at the same time.

[0025] In this embodiment, a fixed volume (e.g., 50 ml) of water sample is collected using a timed sampler, and the samples are continuously collected for a fixed duration (e.g., 24 hours) to form a homogeneous mixed water sample. Samples are collected at a preset cycle, such as every ten minutes. A pre-detection module detects pH, oxidation-reduction potential (ORP), and conductivity. Samples exceeding the contractually preset control threshold are considered to be within a preset multiple of the contractually agreed control threshold, such as ±10% of the contractually agreed control threshold. If the indicators are within ±10% of the contractual threshold, the sample is marked as compliant, and the basic fee rate applies. Disputed samples are sealed and simultaneously sent to a third-party institution for review.

[0026] Specifically, based on the results of the precision testing, a tiered approach is taken: if the concentration exceeding the standard is within the first preset multiple of the contract value, a tiered fee is charged; if the concentration exceeding the standard exceeds the second preset multiple of the contract value, a shutdown order is sent to the branch pipe enterprise. The first preset multiple is, for example, 120%, and the second preset multiple is, for example, 200%.

[0027] S2: Instantaneous water sample testing procedure: Dynamic monitoring nodes are deployed in the main inlet pipeline of the sewage treatment plant to monitor multiple early warning indicators in real time. When any early warning indicator exceeds a preset multiple of the dynamic threshold, instantaneous sampling of the entire network is triggered. Instantaneous water samples are collected from the drainage branch pipes of each water-receiving enterprise. The concentration gradient difference between the instantaneous water sample and the corresponding uniform water sample is compared to determine the drainage branch pipe of the water exceeding the standard.

[0028] In this embodiment, sensors are deployed in the main water inlet pipe to monitor and warn of conductivity, turbidity, and temperature in real time; the preset multiple of the dynamic threshold is 150% of the historical average. Instantaneous water samples are collected in 100ml samples per branch pipe.

[0029] S3: Trigger the source tracing command when water quality exceeds the standard.

[0030] In this embodiment, each water supply company has two branch pipes installed on its drainage branch pipes. Each branch pipe is equipped with a switch valve to take uniform water samples and instantaneous water samples respectively. During normal operation, the average water sample valve is opened at fixed intervals for sampling, while the instantaneous water sample valve is closed. The instantaneous water sample valve is only opened for sampling when there is an abnormality in the indicators of the main water inlet pipe.

[0031] Specifically, upon receiving the source tracing instruction for exceeding the standard, the concentration gradient difference between the instantaneous water sample and the homogeneous water sample is calculated: ΔE=|ORP 瞬时 -ORP 均匀 | Where ΔE is the redox potential difference; ORP 瞬时 ORP is the instantaneous redox potential of a water sample, in mV. 均匀 The redox potential of a homogeneous water sample is ΔE. If ΔE is greater than 50mV, it indicates that the redox state of the instantaneous water sample is significantly different from that of the homogeneous water sample (e.g., illegal discharge of high-concentration reducing or oxidizing wastewater). Based on historical data, ΔE of 50mV can effectively distinguish between normal fluctuations (ΔE≤50mV) and abnormal discharges (ΔE>50mV).

[0032] Where R is the COD concentration gradient ratio, COD 均匀 Chemical oxygen demand (COD) of a homogeneous water sample. 瞬时 R represents the chemical oxygen demand (COD) of the instantaneous water sample. If R > 3, it indicates that the COD concentration of the instantaneous water sample is significantly higher than that of the homogeneous water sample (i.e., the pollution load of the instantaneous water sample increases sharply).

[0033] If ΔE>50mV and R>3, then directly lock the branch pipe of the company illegally discharging water, shut off the inlet valve of the facility company, backflush the pipe and charge an additional fee for exceeding the standard.

[0034] Furthermore, the API interface is linked to smart contracts to automatically settle fees according to a tiered formula: In one embodiment, such as Figure 2 As shown, a method for controlling and tracing incoming water also includes: S10: Collect water quality characteristic data and wastewater pollutant index data of historical drainage samples from enterprises, construct drainage sample dataset and corresponding feature matrix; perform feature dimensionality reduction on feature matrix and extract feature vectors corresponding to the top K largest eigenvalues.

[0035] In practical applications, the sheer volume of samples required for quantitative analysis after collection is enormous. For an industrial park with 30 companies, each with 7 water samples and 6 indicators per sample, a single monitoring session would require storing 1260 indicators. If each indicator is measured in 10 minutes, this would take approximately 210 hours to complete. Assuming each person works 8 hours a day, 26 people would be needed daily to perform these measurements. Assuming a lab worker's monthly salary is 5000 yuan, the annual cost would be 1.56 million yuan. Even assuming a cost of 1 yuan per indicator, the cost would be 1260 yuan, resulting in an annual cost of 460,000 yuan. The total cost reaches 2.02 million yuan, and this is only for one monitoring session per day. For controlling the water quality of incoming water from companies, multiple monitoring sessions daily are necessary, making the costs extremely high. This is the main reason why it is currently impossible to achieve real-time monitoring of every single water sample from every company.

[0036] Given the large volume of wastewater treatment plant testing workload and costs associated with high-volume data collection, this paper proposes a preliminary screening method to predict sample exceedance rates in order to reduce the overall cost of water sample monitoring. This method utilizes a logical algorithm with low-cost and time-efficient indicators to predict exceedance rates, allowing for manual retesting of the identified water samples. Prediction can reduce the number of samples tested to 30% of the original number, significantly lowering the overall monitoring cost.

[0037] In this embodiment, the wastewater sample dataset is a combination of historical wastewater sample data from a large number of enterprises receiving wastewater, including wastewater treatment plants. For example, it may collect historical data (≥1 year) from no fewer than 100 electroplating enterprises covering 5 industrial parks. The wastewater sample dataset associates sample identification information with all samples, including enterprise, water sample, and water quality characteristics. The water quality characteristic data are those with lower detection costs, including pH, ORP, conductivity, temperature, turbidity, and color. The wastewater pollutant index data are those with higher detection costs, including COD, ammonia nitrogen, and heavy metals (nickel, chromium, etc.).

[0038] Specifically, in the constructed feature matrix, each row represents a water sample, and each column corresponds to an original feature index, forming an m×n matrix, where m is the number of samples and n is the number of indicators; Example matrix structure is

[0039] S20: The backpropagation algorithm of the neural network and linear weighting are used to calculate the eigenvalues ​​of the K eigenvectors.

[0040] In this embodiment, Principal Component Analysis (PCA) is used to extract the eigenvectors corresponding to the top K largest eigenvalues. In this embodiment, K = 3, so the K eigenvectors are three eigenvectors: α, β, and γ. The formula for calculating the eigenvalues ​​of the eigenvectors is: N(α, β, γ) = ω1×α + ω2×β + ω3×γ Wherein, ω1, ω2 and ω3 are weight factors optimized by the backpropagation algorithm of the neural network. The initial values ​​of ω1, ω2 and ω3 are randomly generated, such as ω1 = 0.3, ω2 = 0.5 and ω3 = 0.2.

[0041] For example, the first three largest features are pH, ORP, and conductivity. The new feature vector after dimensionality reduction is: α = 0.62 × pH + 0.15 × ORP - 0.31 × conductivity; β = -0.21 × pH + 0.78 × ORP + 0.42 × conductivity; γ = 0.53 × pH - 0.32 × ORP + 0.65 × conductivity.

[0042] The original multiple detection indicators were reduced to three to make predictions of water sample exceedances and reduce the workload of testing.

[0043] S30: Set a combination of thresholds to be adjusted that includes multiple thresholds to be adjusted, and calculate the false alarm rate and false negative rate under each threshold to be adjusted.

[0044] S40: The backpropagation algorithm of the neural network is used, combined with the loss function constructed based on the false alarm rate and the false negative rate, to numerically adjust multiple thresholds to be adjusted, and determine the optimal feature thresholds for the corresponding K feature vectors.

[0045] In this embodiment, the loss function constructed based on the false positive rate and the false negative rate is: Loss=λ1η1+λ2η2 Where η1 is the false alarm rate; η2 is the false negative rate; λ1 and λ2 are balancing coefficients used to adjust the weights of the false alarm and false negative rates; λ1 + λ2 = 1 and its value can be adjusted. The default values ​​are λ1 = λ2 = 0.5.

[0046] Furthermore, the loss function also incorporates fuzzy logic control, dynamically adjusting λ1 and λ2 based on the company's historical violation frequency. For example, λ2 is increased to 0.7 for companies that are habitual offenders of illegally discharging wastewater.

[0047] Specifically, the formula for calculating the false alarm rate η1 is as follows: Where C represents the number of false alarm water samples in the drainage sample dataset; A represents water samples in the drainage sample dataset that may not exceed the standard; and B represents water samples in the drainage sample dataset that may exceed the standard.

[0048] The formula for calculating the false negative rate η2 is as follows: Where D represents the number of water samples that were missed during screening, i.e., the number of water samples that met the standards after testing B potentially exceeding the standards; A and B are obtained by comparing the threshold to be adjusted with the drainage sample dataset.

[0049] Specifically, a backpropagation algorithm using a neural network is employed to solve the loss function and update the weight factors ω1, ω2, and ω3. This yields the minimum value of the loss function corresponding to multiple thresholds to be adjusted, thus determining the optimal feature threshold. The optimal feature threshold satisfies the condition that both the false positive rate and the false negative rate are below a preset proportion, such as 20%.

[0050] Furthermore, weight optimization employs a neural network backpropagation algorithm, updating the weights with the objective of minimizing the loss function; the backpropagation formula is: ω is updated iteratively using gradient descent. i ω i Corresponding to ω1, ω2 and ω3.

[0051] For example, the threshold iteration process is as follows: Set the combination of thresholds to be adjusted (also known as the threshold candidate set) N 预 ∈[0.5, 2.0] (step size 0.1); For each candidate threshold N 预 Calculate the prediction result, where N is greater than N 预 False alarms (C) and false negatives (D) are recorded. Choose N with the smallest loss function value. 预 The optimal feature threshold N is used.

[0052] For example, the process of determining the optimal feature threshold N according to steps S20 to S40 is as follows: (1) Based on more than one year and no less than 30,000 water samples from a single electroplating park, the characteristic values ​​N (α, β, γ) of each water sample were obtained; (2) N(α, β, γ) is compared with the preset threshold Npre. Water samples with a value less than the preset threshold Npre are defined as water samples that may not exceed the standard, and the number of such samples is A. Water samples with a value greater than the preset threshold Npre are defined as water samples that may exceed the standard, and the number of such samples is B. (3) A water samples that are unlikely to exceed the standard are manually screened in the laboratory. The judgment is made by the indicators of wastewater pollutants. Water samples that meet the standard are defined as accurately screened water samples; water samples that do not meet the standard are defined as false alarm water samples, and the number of such samples is C. The false alarm rate is the proportion of false alarm water samples, which is calculated using... Formula calculation.

[0053] (4) The B potentially exceeding water standards were manually screened in the laboratory using wastewater pollutant indicators. Water samples that did not meet the standards were defined as accurately screened samples; those that met the standards were defined as underreported samples, with a count of D samples. The underreporting rate, i.e., the proportion of underreported samples, was determined using... Formula calculation.

[0054] (5) Determine whether both the false alarm rate and the missed alarm rate are below 20%. If not, adjust the preset threshold N. 预 The A+B water samples were compared again to obtain new false alarm and missed alarm rates. This process was repeated until both the false alarm and missed alarm rates were below 20%, at which point the preset threshold N was set. 预 The required optimal feature threshold NS50 is used to determine whether water quality exceeds the standard in actual water sample testing. Based on K feature vectors and the corresponding optimal feature threshold, a preliminary judgment of water quality exceeding the standard is made for uniform water samples and instantaneous water samples, and the water quality exceeding the standard prediction result is obtained.

[0055] In this embodiment, the water quality exceedance prediction result includes water quality exceeding the standard and water quality not exceeding the standard. Water quality exceeding the standard means that the feature value of any feature vector of the K feature vectors exceeds the corresponding optimal feature threshold; water quality not exceeding the standard means that the feature values ​​of all feature vectors of the K feature vectors do not exceed the corresponding optimal feature threshold.

[0056] Specifically, in actual operation, only characteristic values ​​are measured on water samples to obtain characteristic values ​​N(α, β, γ). N(α, β, γ) are compared with the optimal characteristic threshold N. Water samples with values ​​greater than the optimal characteristic threshold N are initially defined as exceeding the water quality standard, and water samples with values ​​lower than the optimal characteristic threshold N are initially defined as not exceeding the water quality standard.

[0057] The technical solution proposed in this application can reduce the amount of testing by 70% through predictive screening, and reduce the annual cost from RMB 2.02 million to RMB 606,000. This data is obtained by comparing the monitoring of 30 companies in an industrial park, each company with 7 water samples, each water sample with 6 indicators, and a monitoring of 1260 indicators required for one monitoring.

[0058] In one embodiment, a method for controlling and tracing incoming water further includes: S100: Establish a database of illegal discharge characteristics to record the ΔE (oxidation-reduction potential difference), R (COD concentration gradient ratio), and time-series waveforms of historical illegal discharge events of various water-receiving enterprises.

[0059] S200: Instantaneous or uniform water samples are randomly selected from enterprises that habitually discharge illegally using sampling methods. When the instantaneous or uniform water sample meets the conditions of ΔE > 65mV and R > 4.2, it is marked as "intentional illegal discharge" in the matching feature library and an alarm message is triggered.

[0060] Specifically, the frequency of monitoring and sampling will be increased for water supply companies that are suspected of "deliberately discharging wastewater".

[0061] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0062] In one embodiment, a water inflow control and tracing system is provided, which corresponds to a water inflow control and tracing method described in the above embodiments.

[0063] A water inflow control and source tracing system includes a uniform water sample collection module, a pre-detection module, a precision detection module, a graded treatment module, a dynamic monitoring node, an instantaneous sampling trigger module, an instantaneous water sample collection module, a concentration gradient comparison analysis module, and an exceedance source tracing instruction module. Detailed descriptions of each functional module are as follows: The uniform water sample collection module is used to periodically sample the drainage pipes of each water-consuming enterprise according to a preset cycle to form a uniform water sample. The pre-detection module is used to perform preliminary screening of water quality data from homogeneous water samples; The precision detection module is activated when the pre-detection module detects that water quality indicators exceed the pre-set control thresholds in the contract. It is used to perform precise multi-parameter detection of water quality. The graded disposal module is used to perform graded disposal operations based on the results of the precision test. If it is confirmed that the standard is exceeded, the additional fee procedure will be initiated, and the sample will be sealed simultaneously if there is a disputed sample. Dynamic monitoring nodes are deployed in the main inlet pipeline of the wastewater treatment plant to monitor multiple early warning indicators in real time; The instantaneous sampling trigger module is used to trigger instantaneous sampling across the entire network when any early warning indicator exceeds a preset multiple of the dynamic threshold; the instantaneous water sample acquisition module is used to collect instantaneous water samples from the drainage branch pipes of each water-producing enterprise. The concentration gradient comparison analysis module is used to analyze the concentration gradient difference between instantaneous water samples and corresponding homogeneous water samples to determine the drainage branch pipes of water exceeding the standard. The source tracing instruction module is triggered when water quality exceeds the standard, and is used to start the source tracing process.

[0064] Optionally, a water inflow control and traceability system may also include: The data acquisition module is used to collect water quality characteristic data and wastewater pollutant index data of historical drainage samples from enterprises, construct drainage sample datasets, and generate corresponding feature matrices. The feature processing module is used to perform feature dimensionality reduction on the feature matrix and extract the feature vectors corresponding to the top K largest eigenvalues; The weight calculation module is used to calculate the feature weights of the K feature vectors using a neural network backpropagation algorithm and a linear weighting method. The threshold adjustment module is used to set a threshold combination containing multiple thresholds to be adjusted, and to calculate the false alarm rate and false negative rate under each threshold; it is further used to use a neural network backpropagation algorithm, combined with a loss function constructed based on the false alarm rate and false negative rate, to numerically optimize the multiple thresholds to be adjusted, and to determine the optimal feature threshold corresponding to the K feature vectors. The water quality assessment module is used to make a preliminary judgment on water quality exceeding the standard for uniform water samples and instantaneous water samples based on the K feature vectors and their corresponding optimal feature thresholds during the actual water sample testing process, and output the water quality exceeding the standard prediction result.

[0065] For specific limitations regarding a water inflow control and traceability system, please refer to the limitations of a water inflow control and traceability method mentioned above, which will not be repeated here. Each module in the aforementioned water inflow control and traceability system can be implemented entirely or partially through software, hardware, or a combination thereof. Each module can be embedded in or independent of the processor in a computer device in hardware form, or it can be stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0066] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: S1: Homogeneous water sample testing procedure: Sampling is carried out on the drainage pipes of each water-receiving enterprise at preset intervals to form uniform water samples; the water quality data of the uniform water samples is screened by the pre-detection module; if the pre-detection indicators exceed the preset control thresholds in the contract, the fine detection module is activated to conduct multi-parameter accurate detection of water quality; and graded handling is carried out according to the fine detection results: if the standard is exceeded, the additional fee procedure is initiated, and the disputed sample is sealed at the same time. S2: Instantaneous water sample testing procedure: Dynamic monitoring nodes are deployed in the main inlet pipeline of the sewage treatment plant to monitor multiple early warning indicators in real time. When any early warning indicator exceeds a preset multiple of the dynamic threshold, instantaneous sampling of the entire network is triggered. Instantaneous water samples are collected from the drainage branch pipes of each water-receiving enterprise. The concentration gradient difference between the instantaneous water sample and the corresponding uniform water sample is compared to determine the drainage branch pipe of the water exceeding the standard. S3: Trigger the source tracing command when water quality exceeds the standard.

[0067] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0068] In one embodiment, particularly according to an embodiment of the invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, embodiments of the invention include a computer program product comprising a computer program / instructions that, when executed by a processor, implement the steps of the water inflow control and traceability method described above. In such embodiments, the computer program can be downloaded and installed from a network via a communication module, and / or installed from a removable medium. When the computer program is executed by a central processing unit (CPU), it performs the various functions defined in this invention.

[0069] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0070] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for controlling and tracing incoming water, characterized in that, include: Homogeneous water sample testing steps: Sampling is performed on the drainage pipes of each water-receiving enterprise at preset intervals to form uniform water samples. The water quality data of the uniform water samples is screened by the pre-detection module. If the pre-detection indicators exceed the preset control thresholds in the contract, the fine detection module is activated to perform multi-parameter precise detection of the water quality. According to the fine detection results, graded handling is carried out: if the standard is exceeded, the additional fee procedure is initiated, and the disputed sample is sealed at the same time. Instantaneous water sample testing steps: Dynamic monitoring nodes are deployed in the main inlet pipeline of the sewage treatment plant to monitor multiple early warning indicators in real time. When any early warning indicator exceeds a preset multiple of the dynamic threshold, instantaneous sampling of the entire network is triggered. Instantaneous water samples are collected from the drainage branch pipes of each water-receiving enterprise. The concentration gradient difference between the instantaneous water sample and the corresponding uniform water sample is compared to determine the drainage branch pipe of the water exceeding the standard. When water quality exceeds the standard, a source tracing command is triggered.

2. The method for controlling and tracing incoming water according to claim 1, characterized in that, The homogeneous water sample detection step specifically includes: A fixed volume of water sample is collected by a timed sampler, and the samples are continuously collected for a fixed period of time to form a mixed homogeneous water sample. The pre-detection module detects the pH value, oxidation-reduction potential and conductivity of the homogeneous water sample. If the values ​​exceed the contract-preset control threshold, the values ​​are within a preset multiple range of the control threshold agreed in the contract. The tiered handling based on the precision detection results includes charging additional fees at a tiered rate if the concentration exceeding the standard is within a first preset multiple of the contract value, and sending a production stoppage order to the enterprise responsible for the exceeding standard if the concentration exceeds a second preset multiple of the contract value.

3. The method for controlling and tracing incoming water according to claim 1, characterized in that, The method further includes: Collect water quality characteristic data and wastewater pollutant index data of historical drainage samples from enterprises, construct drainage sample dataset and corresponding feature matrix; perform feature dimensionality reduction on the feature matrix and extract the feature vectors corresponding to the top K largest eigenvalues; The eigenvalues ​​of K eigenvectors are calculated using a neural network backpropagation algorithm and linear weighting. Set a combination of thresholds to be adjusted that includes multiple thresholds to be adjusted, and calculate the false alarm rate and false negative rate under each threshold to be adjusted; The backpropagation algorithm of the neural network is used, and a loss function based on the false positive rate and the false negative rate is used to numerically adjust multiple thresholds to be adjusted, so as to determine the optimal feature thresholds corresponding to K feature vectors. In actual water sample testing, based on the K feature vectors and the corresponding optimal feature thresholds, a preliminary water quality exceedance judgment is made on the uniform water sample and the instantaneous water sample to obtain the water quality exceedance prediction result.

4. The method for controlling and tracing incoming water according to claim 3, characterized in that, The K eigenvectors are three eigenvectors α, β, and γ; the eigenvalues ​​of the eigenvectors are calculated using the following formula: N(α,β,γ)=ω1×α+ω2×β+ω3×γ Where ω1, ω2, and ω3 are the weight factors optimized by the backpropagation algorithm of the neural network.

5. The method for controlling and tracing incoming water according to claim 4, characterized in that, The loss function constructed based on the false positive rate and the false negative rate is as follows: Loss=λ1η1+λ2η2 Where η1 is the false alarm rate; η2 is the false negative rate; λ1 and λ2 are balance coefficients, λ1+λ2=1 and their values ​​can be adjusted; The loss function is solved using a neural network backpropagation algorithm, and the weight factors ω1, ω2 and ω3 are updated to obtain the minimum value of the loss function corresponding to the plurality of thresholds to be adjusted, so as to determine the optimal feature threshold.

6. The method for controlling and tracing incoming water according to claim 5, characterized in that, The optimal feature threshold satisfies the condition that both the false alarm rate and the false negative rate are lower than a preset ratio value; The water quality exceedance prediction results include water quality exceeding the standard and water quality not exceeding the standard. The water quality exceeding the standard is defined as the feature value of any feature vector of the K feature vectors exceeding the corresponding optimal feature threshold; the water quality not exceeding the standard is defined as the feature value of all feature vectors of the K feature vectors not exceeding the corresponding optimal feature threshold.

7. A method for controlling and tracing incoming water according to claim 3 or 5, characterized in that, The formula for calculating the false alarm rate η1 is as follows: Wherein, C represents the number of false alarm water samples in the drainage sample dataset; A represents water samples in the drainage sample dataset that may not exceed the standard; and B represents water samples in the drainage sample dataset that may exceed the standard. The formula for calculating the false negative rate η2 is as follows: Where D represents the number of water samples that were missed during screening, i.e., the number of water samples that met the standards after testing B potentially exceeding the standards; A and B are obtained by comparing the threshold to be adjusted with the drainage sample dataset.

8. A water inflow control and traceability system, characterized in that, The system includes: The uniform water sample collection module is used to periodically sample the drainage pipes of each water-consuming enterprise according to a preset cycle to form a uniform water sample. The pre-detection module is used to perform preliminary screening of the water quality data of the homogeneous water sample; The precision detection module is activated when the pre-detection module detects that the water quality indicators exceed the contract-preset control threshold, and is used to perform precise multi-parameter detection of water quality. The graded disposal module is used to perform graded disposal operations based on the results of the precision test. If it is confirmed that the standard is exceeded, the additional fee procedure will be initiated, and the sample will be sealed simultaneously if there is a disputed sample. Dynamic monitoring nodes are deployed in the main inlet pipeline of the wastewater treatment plant to monitor multiple early warning indicators in real time; The instantaneous sampling trigger module is used to trigger instantaneous sampling across the entire network when any early warning indicator exceeds a preset multiple of the dynamic threshold; the instantaneous water sample acquisition module is used to collect instantaneous water samples from the drainage branch pipes of each water-producing enterprise. The concentration gradient comparison analysis module is used to perform concentration gradient difference analysis between the instantaneous water sample and the corresponding uniform water sample to determine the drainage branch of the water exceeding the standard. The source tracing instruction module is triggered when water quality exceeds the standard, and is used to start the source tracing process.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the water inflow control and tracing method as described in any one of claims 1 to 7.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the steps of the water inflow control and tracing method as described in any one of claims 1 to 7.

Images