Municipal sewage treatment plant environmental benefit and resource load evaluation method

Abstract

The invention discloses an environmental benefit and resource load evaluation method for an urban sewage treatment plant, and belongs to the field of sustainable evaluation. The implementation method comprises the following steps: training a random forest model by taking pollutant effluent concentration, energy consumption intensity and medicament consumption intensity of a national town sewage treatment plant as output variables and taking other historical operation information as input variables; on the basis, quantile regression is used for calculating quantiles of actual values of three operation indexes of each factory under the original processing condition; constructing an environmental benefit evaluation index system represented by pollutant removal efficiency and a resource load evaluation system represented by consumption efficiency of resources such as energy and chemicals; through obtaining environmental benefit and resource load grading benchmark values of the town sewage treatment plant, core functions and necessary input of each plant are accurately quantified, and real environmental contribution of the town sewage treatment industry is clarified, so that evaluation of sustainability of the town sewage treatment plant is more objective and fair, and sustainable development of urban environment infrastructures is facilitated.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of sustainable evaluation method, and particularly relates to a method for evaluating environmental benefits and resource load of a municipal wastewater treatment plant. BACKGROUND

[0002] The fundamental purpose of developing a municipal wastewater treatment plant is to reduce the degree of urban water pollution. For a long time, it has been generally believed in the society that the municipal wastewater treatment plant has made positive contributions to the protection of urban water environment and the stable growth of economy. However, in fact, in order to ensure that the effluent quality meets the discharge standard, the municipal wastewater treatment industry has invested a large amount of energy and reagents and other resources in the operation process, resulting in many negative effects such as greenhouse gas emission and operating cost increase, which leads to the risk of future development of the industry. How to comprehensively evaluate the real benefits of the municipal wastewater treatment plant has a positive significance for promoting its reduction of dependence on external resource input and then realizing sustainable development (Wang et al., 2012). The core function of the municipal wastewater treatment plant is to remove pollutants such as organic matter and TN, to promote the virtuous cycle of clean water resources by reducing the total amount and flux of pollutants discharged to the urban water body, and to generate environmental benefits. At the same time, the consumption of electric energy, external carbon source, phosphorus removal reagent and dewatering reagent in the wastewater treatment process also produces resource load. The sustainable evaluation of the municipal wastewater treatment plant should be systematically carried out around the above two key aspects, that is, to consider not only the positive role of wastewater treatment but also the negative effects of the process itself (Wang et al., 2015). At present, researches focusing on the evaluation of various indexes of the municipal wastewater treatment plant have emerged, but they all have defects such as small sample size, rough assumption of wastewater treatment process and not deep data rule mining. SUMMARY

[0003] In order to solve the problem that the negative effects caused by the consumption of energy and reagents in the municipal wastewater treatment plant have been ignored for a long time, and the problem that the evaluation is not objective due to the dependence on artificial preference for sorting different indexes, the purpose of the present application is to provide a method for evaluating environmental benefits and resource load of a municipal wastewater treatment plant. By establishing an evaluation framework of quantile random forest, the benchmark values of the main pollutant removal efficiency and the energy and reagent consumption efficiency of each plant are accurately determined, so that the environmental benefits, resource load and final sustainability evaluation of the municipal wastewater treatment plant are more accurate, and then the plants can reduce the operating cost while improving the operating efficiency, realizing efficient and sustainable operation.

[0004] The purpose of the present application is realized by the following technical scheme:

[0005] The method for evaluating environmental benefits and resource load of a municipal wastewater treatment plant disclosed by the present application comprises the following steps:

[0006] Step one: Take the effluent concentration of various pollutants in the urban sewage treatment plant as the output object, train the random forest model F1, and use the quantile regression method to calculate the quantile of the effluent concentration of various pollutants under the original treatment condition on the basis of F1, and obtain the removal efficiency of various pollutants in each plant. According to the removal efficiency, the environmental benefits of each plant are obtained.

[0007] Step two: Take the energy consumption intensity of various energy sources in the urban sewage treatment plant as the output object, train the random forest model F2, and use the quantile regression method to calculate the quantile of the energy consumption intensity of various energy sources under the original treatment condition on the basis of F2, and obtain the consumption efficiency of various energy sources in each plant.

[0008] Take the consumption intensity of various reagents in the urban sewage treatment plant as the output object to train the random forest model F3, and use the quantile regression method to calculate the quantile of the consumption intensity of various reagents under the original treatment condition on the basis of F3, and obtain the consumption efficiency of various reagents in each plant.

[0009] According to the consumption efficiency of various energy sources and various reagents in each plant, the resource load of each plant is calculated.

[0010] Step three: According to the environmental benefits of each plant obtained in step one and the resource load of each plant obtained in step two, the sustainability score of each plant is calculated, that is, the environmental benefits and resource load evaluation of the urban sewage treatment plant is realized.

[0011] Further, the random forest model F1 in step one is constructed by taking part of the information in the monthly operation data of all urban sewage treatment plants as input variables, including the treatment water volume of each plant, the influent concentration of various pollutants, the consumption intensity of various energy sources and reagents, the treatment process type, the season and the city, and the output variable is the effluent concentration of various pollutants. Then, on the basis of F1, the quantile of the effluent concentration of various pollutants under the original treatment condition is calculated by using the quantile regression method, and the specific process is shown in formulas (1), (2) and (3):

[0012]

[0013] In formula (1), F1(y|X=x k ) represents the probability that the output variable Y is less than or equal to the specified value y when the input variable X is the kth monthly operation data x k , I{Y j ≤y} is an indicator function, which is equal to 1 when the jth output variable Y j is less than or equal to the specified value y, and is equal to 0 otherwise, and w tj (x k) represents the weight coefficients assigned to each decision tree by the quantile regression method based on the overall distribution of the input variables, T is the number of decision trees in the random forest algorithm, and n represents the total number of monthly data entries.

[0014] Formula (2) means that the values ​​of the output variable Y are sorted in ascending order, and denoted as Y1, Y2, ... Y i-1 Y i Y i+1 For a given value y, find the values ​​Y of two adjacent output variables Y. i-1 and Y i If y is exactly between the two values, then the quantile τ corresponding to the specified value y can be estimated by formula (3), where G is the ECDF (Empirical Cumulative Distribution Function), and G(Y) is the quantile τ of the specified value y. i-1 ) and G(Y i ) respectively represent less than or equal to Y i-1 and Y i The proportion of Y in Y, that is, the proportion of the output variable Y in Y i-1 To Y i It changes approximately linearly within a small interval.

[0015] Furthermore, the six types of pollutants mentioned in step one include COD, BOD, TN, NH3-N, TP, and SS. The removal efficiency of any pollutant i is calculated using formula (4):

[0016] RP i =1-τ i (4)

[0017] In formula (4) RP i For the removal efficiency of pollutant i, τ i Let be the quantile of the effluent concentration of pollutant i under the original treatment conditions.

[0018] Furthermore, the environmental benefits of each plant mentioned in step one are obtained using formula (5):

[0019]

[0020] In formula (5), S Environmental Benefits For the environmental benefits of each factory, W 1i The weighting coefficient is assigned to the removal efficiency of pollutant i.

[0021] Furthermore, the random forest model F2 described in step two is constructed using partial information from the monthly operation data of all urban wastewater treatment plants as input variables, including the treated water volume of each plant, the removal amount of various pollutants, the consumption intensity of various reagents, the treatment process type, the season, and the city. The output variable is the consumption intensity of various energy sources. Then, based on F2, the quantile regression method is used to calculate the quantiles of the consumption intensity of various energy sources under the original treatment conditions. The specific process is shown in formulas (6), (7), and (8):

[0022]

[0023] In formula (6), F2(y|X=x) k This indicates that when the input variable X is the k-th monthly operational data point x... k When the output variable Y is less than or equal to a specified value y, I{Y j ≤y} is the indicator function, when the j-th output variable Y j The indicator function is 1 when the value is less than or equal to the specified value y, and 0 otherwise. tj (x k ) represents the weight coefficients assigned to each decision tree by the quantile regression method based on the overall distribution of the input variables, T is the number of decision trees in the random forest algorithm, and n represents the total number of monthly data entries.

[0024] Formula (7) means that the values ​​of the output variable Y are sorted in ascending order, and denoted as Y1, Y2, ... Y i-1 Y i Y i+1 For a given value y, find the values ​​Y of two adjacent output variables Y. i-1 and Y i If y lies exactly between the two values, then the quantile τ corresponding to the specified value y can be estimated using formula (8), where G is the ECDF (Empirical Cumulative Distribution Function), and G(Y) is the quantile τ of the specified value y. i-1 ) and G(Y i ) respectively represent less than or equal to Y i-1 and Y i The proportion of Y in Y, that is, the proportion of the output variable Y in Y i-1 To Y i It changes approximately linearly within a small interval.

[0025] Furthermore, the energy sources mentioned in step two total two types, including electrical energy and thermal energy, and the consumption efficiency of any energy source j is calculated using formula (9):

[0026] CP j =τj (9)

[0027] In formula (9) CP j For the energy consumption efficiency of j, τ j Let be the quantile of energy consumption intensity of energy j under the original treatment conditions.

[0028] Furthermore, the random forest model F3 described in step two is constructed using partial information from the monthly operation data of all urban wastewater treatment plants as input variables, including the treated water volume of each plant, the removal amount of various pollutants, the consumption intensity of various energy sources, the type of treatment process, the season, and the city. The output variable is the consumption intensity of various chemicals. Then, based on F3, the quantile regression method is used to calculate the quantiles of the consumption intensity of various chemicals under the original treatment conditions. The specific process is shown in formulas (10), (11), and (12):

[0029]

[0030] In formula (10), F3(y|X=x) k This indicates that when the input variable X is the k-th monthly operational data point x... k When the output variable Y is less than or equal to a specified value y, I{Y j ≤y} is the indicator function, when the j-th output variable Y j The indicator function is 1 when the value is less than or equal to the specified value y, and 0 otherwise. tj (x k ) represents the weight coefficients assigned to each decision tree by the quantile regression method based on the overall distribution of the input variables, T is the number of decision trees in the random forest algorithm, and n represents the total number of monthly data entries.

[0031] Formula (11) means that the values ​​of the output variable Y are sorted in ascending order, and denoted as Y1, Y2, ... Y i-1 Y i Y i+1 For a given value y, find the values ​​Y of two adjacent output variables Y. i-1 and Y i If y lies exactly between the two values, then the quantile τ corresponding to the specified value y can be estimated using formula (12), where G is the ECDF (Empirical Cumulative Distribution Function), and G(Y) is the quantile τ of the specified value y. i-1 ) and G(Y i ) respectively represent less than or equal to Y i-1 and Y i The proportion of Y in Y, that is, the proportion of the output variable Y in Y i-1 To Yi It changes approximately linearly within a small interval.

[0032] Furthermore, there are a total of three types of agents mentioned in step two, including an external carbon source, a phosphorus removal agent, and a dehydration agent. The consumption efficiency of any agent k is calculated using formula (13):

[0033] CP k =τ k (13)

[0034] In formula (13) CP k τ represents the consumption efficiency of drug k. k Let be the quantile of the consumption intensity of agent k under the original treatment conditions.

[0035] Furthermore, the resource load of each plant mentioned in step two is obtained through formula (14):

[0036]

[0037] In formula (14), S Resource Loads To manage the resource load of each plant, W 2j and W 3k These are the weighting coefficients for the consumption efficiency allocated to energy j and medicine k, respectively.

[0038] Furthermore, the sustainability score of each factory mentioned in step three is obtained using formula (15):

[0039] S Sustainability =S Environmental Benefits -S Resource Loads (15)

[0040] In formula (15) S Sustainability This assigns a sustainability score to each plant. The score represents the environmental benefits (S) generated by each plant through the removal of pollutants including COD, BOD, TN, NH3-N, TP, and SS. Environmental Benefits At the same time, resource load S will also be generated due to the consumption of electricity and heat, as well as external carbon sources, phosphorus removal agents and dehydration agents. Resource Loads The difference between the two represents the true sustainability of each plant. When a town's wastewater treatment plant removes more pollutants or consumes less energy and chemicals, S... Sustainability The larger the value, the greater the positive contribution of the factory to the ecology of the city.

[0041] Beneficial effects:

[0042] 1. This invention discloses a method for evaluating the environmental benefits and resource load of urban wastewater treatment plants. The evaluation object is accurate to each urban wastewater treatment plant. The evaluation content includes key operational indicators such as pollutant effluent concentration, energy consumption intensity, and reagent dosage intensity of each plant. The aim is to obtain accurate classification and benchmark values ​​of each key operational indicator under corresponding treatment conditions, and then rationally formulate targeted adjustment strategies. The evaluation results are based entirely on real historical operational data, eliminating the subjective interference that is usually unavoidable in previous evaluation studies. In actual production, it can provide accurate reference for improving pollutant removal efficiency and controlling energy and reagent consumption costs of urban wastewater treatment plants in various provinces and cities under different influent conditions.

[0043] 2. This invention discloses a method for evaluating the environmental benefits and resource load of urban wastewater treatment plants. By incorporating the resource load caused by energy and chemical consumption of urban wastewater treatment plants into the scope of consideration, it improves the sustainability evaluation system of urban wastewater treatment plants, making the evaluation results more scientific. At the same time, it introduces cutting-edge data science methods to quantify and make the environmental benefits and resource load of urban wastewater treatment plants comparable, which helps to promote the efficient and sustainable development of China's urban wastewater treatment industry. The evaluation system proposed in this invention has the characteristics and advantages of convenience and speed, and can clearly know the relative size and composition of the sustainability of each plant through a highly automated calculation process.

[0044] 3. This invention discloses a method for evaluating the environmental benefits and resource load of urban wastewater treatment plants. Using the effluent concentrations of various pollutants from the urban wastewater treatment plant as the output, a random forest model F1 is trained. Based on F1, quantile regression is used to calculate the quantiles of the effluent concentrations of various pollutants under the original treatment conditions, obtaining the removal efficiency of various pollutants in each plant. Based on the removal efficiency, the environmental benefits of each plant are obtained. Using the energy consumption intensity of various types of energy from the urban wastewater treatment plant as the output, a random forest model F2 is trained. Based on F2, quantile regression is used to calculate the quantiles of the energy consumption intensity of various types of energy under the original treatment conditions, obtaining the energy consumption efficiency of each plant. Using the consumption intensity of various chemicals from the urban wastewater treatment plant as the output, a random forest model F3 is trained. Based on F3, quantile regression is used to calculate the quantiles of the chemical consumption intensity of various types of chemicals under the original treatment conditions, obtaining the chemical consumption efficiency of each plant. Based on the energy and chemical consumption efficiency of each plant, the resource load of each plant is calculated. Based on the training of random forest models F1, F2, and F3, and combined with the constructed formula models for key operational indicators such as pollutant effluent concentration, energy consumption intensity, and reagent dosage intensity, quantile regression is used to calculate the quantiles of the actual values ​​of the above three operational indicators for each plant under the original treatment conditions. The results are used as the basis for characterizing the environmental benefits and resource load of urban wastewater treatment plants, and an evaluation index system for the environmental benefits and resource load of urban wastewater treatment plants is constructed. This system enables a systematic and quantitative evaluation of the environmental benefits and resource load of urban wastewater treatment plants, accurately quantifies the core functions and necessary inputs of each plant, clarifies the environmental contribution of the urban wastewater treatment industry, and makes the evaluation of the sustainability of urban wastewater treatment plants more objective and fair, which is conducive to the sustainable development of urban environmental infrastructure. Attached Figure Description

[0045] Figure 1 This invention discloses a plant-level environmental benefit scheme for urban wastewater treatment. Environmental Benefits Resource load S Resource Loads and sustainability score S Sustainability A schematic diagram of the evaluation process. Detailed Implementation

[0046] To better illustrate the purpose and advantages of the present invention, the invention will be further described below in conjunction with the accompanying drawings and examples.

[0047] Example 1:

[0048] This embodiment discloses a method for evaluating the environmental benefits and resource load of urban wastewater treatment plants. It can be used to quantitatively assess the environmental benefits, resource load, and sustainability scores of typical urban wastewater treatment plants with the largest treatment capacity in China's most developed cities, such as the Nanshan Wastewater Treatment Plant in Shenzhen, Guangdong Province, during a specified period, such as January 2020. By revealing the multiple reasons behind the heterogeneity between plants, it assists China's urban wastewater treatment industry in achieving sustainable development efficiently and accurately.

[0049] like Figure 1 As shown in the figure, this embodiment discloses a method for evaluating the environmental benefits and resource load of urban wastewater treatment plants. The specific implementation steps are as follows:

[0050] Step 1: Using the monthly operational data of urban wastewater treatment plants in various provinces and cities from 2007 to 2020 provided by the Ministry of Housing and Urban-Rural Development, including treated water volume, influent concentrations of COD, BOD, TN, NH3-N, TP, and SS, electricity consumption, external carbon source, consumption intensity of phosphorus removal and dehydration agents, treatment process type, season, and city, as input variables, and the effluent concentrations of COD, BOD, TN, NH3-N, TP, and SS, as output variables, train a random forest model F1. Based on the F1 model, quantile regression is then applied. The method calculates the quantiles of COD, BOD, TN, NH3-N, TP, and SS concentrations in the effluent of Shenzhen Nanshan Wastewater Treatment Plant in January 2020 under the original treatment conditions, which are 1.00, 0.00, 0.01, 0.02, 0.00, and 0.01, respectively. This yields the removal efficiencies of COD, BOD, TN, NH3-N, TP, and SS at Shenzhen Nanshan Wastewater Treatment Plant in January 2020, which are 0.00, 1.00, 0.99, 0.98, 1.00, and 0.99, respectively.

[0051] Based on the removal efficiency, and considering the rationality of the integration among evaluation indicators, the removal of various pollutants and the consumption of various energy and reagents are all key operating indicators of urban wastewater treatment plants. Their actual significance in production does not show a clear relationship of importance. Therefore, each pollutant removal efficiency and each resource consumption efficiency are assigned an equal weight coefficient, thereby achieving the quantification of the final environmental benefits, resource load, and sustainability scores. The environmental benefits of Shenzhen Nanshan Wastewater Treatment Plant in January 2020 were 4.96.

[0052] Step 2: Using the monthly operation data of urban wastewater treatment plants in various provinces and cities from 2007 to 2020 provided by the Ministry of Housing and Urban-Rural Development, including the treated water volume, COD, BOD, TN, NH3-N, TP and SS influent and effluent concentrations, external carbon source, consumption intensity of phosphorus removal and dehydration agents, treatment process type, season and city, as input variables and the energy consumption intensity as output variable, a random forest model F2 is trained. Based on F2, the quantile regression method is used to calculate the quantile of the energy consumption intensity of the Nanshan Wastewater Treatment Plant in Shenzhen in January 2020 under the original treatment conditions, which is 0.01. Thus, the energy consumption efficiency of the Nanshan Wastewater Treatment Plant in Shenzhen in January 2020 is obtained, which is 0.01.

[0053] Using the monthly operation data of urban wastewater treatment plants in various provinces and cities from 2007 to 2020 provided by the Ministry of Housing and Urban-Rural Development, including treated water volume, COD, BOD, TN, NH3-N, TP and SS influent and effluent concentrations, power consumption intensity, treatment process type, season and city, as input variables, and the consumption intensity of added carbon source, phosphorus removal agent and dehydration agent as output variables, a random forest model F3 was trained. Based on F3, the quantile regression method was used to calculate the quantiles of the consumption intensity of added carbon source, phosphorus removal agent and dehydration agent of Nanshan Wastewater Treatment Plant in Shenzhen in January 2020 under the original treatment conditions, which were 0.02, 0.00 and 0.03, respectively. Thus, the consumption efficiency of added carbon source, phosphorus removal agent and dehydration agent of Nanshan Wastewater Treatment Plant in Shenzhen in January 2020 was obtained, which were 0.02, 0.00 and 0.03, respectively.

[0054] Based on the consumption efficiency, each pollutant removal efficiency and each resource consumption efficiency are assigned an equal weight coefficient, thereby quantifying the final environmental benefits, resource load, and sustainability score. The resource load of Shenzhen Nanshan Wastewater Treatment Plant in January 2020 was 0.06.

[0055] Step 3: Based on the environmental benefit score of 4.96 obtained in Step 1 for Shenzhen Nanshan Wastewater Treatment Plant in January 2020 and the resource load score of 0.06 obtained in Step 2, the sustainability score of Shenzhen Nanshan Wastewater Treatment Plant in January 2020 is calculated to be 4.90.

[0056] Similarly, the environmental benefits, resource load, and sustainability scores of the Shenzhen Nanshan Wastewater Treatment Plant for other designated months between 2007 and 2020 can be obtained. Comparing these results shows that the environmental benefits of the Shenzhen Nanshan Wastewater Treatment Plant significantly exceed its resource load, but the environmental benefits exhibit a downward trend. Except for TN, the plant's removal efficiency for other pollutants exceeds the median level under the same treatment conditions. Even the lowest-valued COD removal efficiency reaches 0.69. However, the gap between environmental benefits and resource load is narrowing because the effluent concentrations of some pollutants are increasing. For example, the effluent concentration of NH3-N increased by about two-thirds, while the influent concentrations of COD, SS, and TP decreased by at least 24.3%, directly leading to a lower pollutant removal efficiency. Therefore, the Shenzhen Nanshan Wastewater Treatment Plant should promptly investigate the root causes of these changes in pollutant concentrations to prevent the erosion of its existing environmental benefits. As the largest urban wastewater treatment facility in a developed region, this plant's sustainable transformation will play a significant leading role in the industry. By controlling the removal efficiency of each pollutant and the consumption efficiency of each energy and reagent to above and below the median values ​​corresponding to the same treatment conditions, respectively, it can ensure the efficient operation of the urban wastewater treatment process. Each plant can formulate improvement measures for various pollutant removal and energy and reagent consumption that are suitable for its own actual situation based on the evaluation results of environmental benefits and resource load, so as to accurately improve operational efficiency and reduce economic costs, thereby promoting the improvement of the overall sustainability score of the industry and helping to achieve efficient and sustainable operation.

[0057] The above detailed description further illustrates the purpose, technical solution, and beneficial effects of the invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for evaluating the environmental benefits and resource load of urban wastewater treatment plants, characterized in that: Includes the following steps: Step 1: Using the effluent concentrations of various pollutants from urban wastewater treatment plants as the output, train a random forest model F1. Based on the random forest model F1, use quantile regression to calculate the quantiles of the effluent concentrations of various pollutants under the original treatment conditions, and obtain the removal efficiency of various pollutants in each plant; based on the removal efficiency, obtain the environmental benefits of each plant. Step 2: Using the energy consumption intensity of various types of energy in urban wastewater treatment plants as the output, train a random forest model F2. Based on the random forest model F2, use the quantile regression method to calculate the quantiles of the energy consumption intensity of various types of energy under the original treatment conditions, and obtain the energy consumption efficiency of various types of energy in each plant. A random forest model F3 was trained using the consumption intensity of various chemicals in urban wastewater treatment plants as the output. Based on the random forest model F3, the quantile regression method was used to calculate the quantiles of the consumption intensity of various chemicals under the original treatment conditions, so as to obtain the consumption efficiency of various chemicals in each plant. Calculate the resource load of each plant based on the consumption efficiency of various types of energy and various types of pharmaceuticals. Step 3: Based on the environmental benefits of each plant obtained in Step 1 and the resource load of each plant obtained in Step 2, calculate the sustainability score of each plant, thus realizing the environmental benefits and resource load evaluation of urban wastewater treatment plants.

2. The method as described in claim 1, characterized in that: The random forest model F1 described in step one is constructed using partial information from the monthly operation data of all urban wastewater treatment plants as input variables. This partial information includes the treatment volume of each plant, the influent concentration of various pollutants, the consumption intensity of various energy and reagents, the treatment process type, the season, and the city. The output variable is the effluent concentration of various pollutants. Then, based on F1, the quantile regression method is used to calculate the quantiles of the effluent concentration of various pollutants under the original treatment conditions. The specific process is shown in formulas (1), (2), and (3): AND i-1 ≤y≤Y i (2) In formula (1), F1(y|X=x) k This indicates that when the input variable X is the k-th monthly operational data point x... k When the output variable Y is less than or equal to a specified value y, I{Y j ≤y} is the indicator function, when the j-th output variable Y j The indicator function is 1 when the value is less than or equal to the specified value y, and 0 otherwise. tj (x k ) represents the weight coefficients assigned to each decision tree by the quantile regression method based on the overall distribution of the input variables, T represents the number of decision trees in the random forest algorithm, and n represents the total number of monthly data entries; Formula (2) means that the values ​​of the output variable Y are sorted in ascending order, and denoted as Y1, Y2, ... Y i-1 Y i Y i+1 For a given value y, find the values ​​Y of two adjacent output variables Y. i-1 and Y i If y is exactly between the two values, then the quantile τ corresponding to the specified value y is estimated by formula (3), where G is the empirical cumulative distribution function ECDF, G(Y i-1 ) and G(Y i ) respectively represent less than or equal to Y i-1 and Y i The proportion of Y in Y, that is, the proportion of the output variable Y in Y i-1 To Y i It changes approximately linearly within a small interval.

3. The method as described in claim 2, characterized in that: The pollutants mentioned in step one total six types, including chemical oxygen demand (COD), biochemical oxygen demand (BOD), total nitrogen (TN), ammonia nitrogen (NH3-N), total phosphorus (TP), and suspended solids (SS). The removal efficiency of any pollutant i is calculated using formula (4): RP i =1-τ i (4) In formula (4) RP i For the removal efficiency of pollutant i, τ i Let be the quantile of the effluent concentration of pollutant i under the original treatment conditions.

4. The method as described in claim 3, characterized in that: The environmental benefits of each plant mentioned in step one are obtained through formula (5): In formula (5), S Environmental Benefits For the environmental benefits of each factory, W 1i The weighting coefficient is assigned to the removal efficiency of pollutant i.

5. The method as described in claim 4, characterized in that: The random forest model F2 described in step two is constructed using partial information from the monthly operation data of all urban wastewater treatment plants as input variables. This partial information includes the treated water volume of each plant, the removal amount of various pollutants, the consumption intensity of various reagents, the treatment process type, the season, and the city. The output variable is the consumption intensity of various energy sources. Then, based on F2, the quantile regression method is used to calculate the quantiles of the consumption intensity of various energy sources under the original treatment conditions. The specific process is shown in formulas (6), (7), and (8): AND i-1 ≤y≤Y i (7) In formula (6), F2(y|X=x) k This indicates that when the input variable X is the k-th monthly operational data point x... k When the output variable Y is less than or equal to a specified value y, I{Y j ≤y} is the indicator function, when the j-th output variable Y j The indicator function is 1 when the value is less than or equal to the specified value y, and 0 otherwise. tj (x k ) represents the weight coefficients assigned to each decision tree by the quantile regression method based on the overall distribution of the input variables, T represents the number of decision trees in the random forest algorithm, and n represents the total number of monthly data entries; Formula (7) means that the values ​​of the output variable Y are sorted in ascending order, and denoted as Y1, Y2, ... Y i-1 Y i Y i+1 For a given value y, find the values ​​Y of two adjacent output variables Y. i-1 and Y i If y lies exactly between the two values, then the quantile τ corresponding to the specified value y can be estimated using formula (8), where G is the ECDF (Empirical Cumulative Distribution Function), and G(Y) is the quantile τ of the specified value y. i-1 ) and G(Y i ) respectively represent less than or equal to Y i-1 and Y i The proportion of Y in Y, that is, the proportion of the output variable Y in Y i-1 To Y i It changes approximately linearly within a small interval.

6. The method as described in claim 5, characterized in that: Step 2 describes two types of energy sources, including electrical energy and thermal energy. The energy consumption efficiency of any energy source j is calculated using formula (9): CP j =τ j (9) In formula (9) CP j For the energy consumption efficiency of j, τ j Let be the quantile of energy consumption intensity of energy j under the original treatment conditions.

7. The method as described in claim 6, characterized in that: The random forest model F3 described in step two is constructed using partial information from the monthly operation data of all urban wastewater treatment plants as input variables. The partial information includes the treated water volume of each plant, the removal amount of various pollutants, the consumption intensity of various energy sources, the type of treatment process, the season, and the city. The output variable is the consumption intensity of various chemicals. Then, based on F3, the quantile regression method is used to calculate the quantiles of the consumption intensity of various chemicals under the original treatment conditions. The specific process is shown in formulas (10), (11), and (12): AND i-1 ≤y≤Y i (11) In formula (10), F3(y|X=x) k This indicates that when the input variable X is the k-th monthly operational data point x... k When the output variable Y is less than or equal to a specified value y, I{Y j ≤y} is the indicator function, when the j-th output variable Y j The indicator function is 1 when the value is less than or equal to the specified value y, and 0 otherwise. tj (x k ) represents the weight coefficients assigned to each decision tree by the quantile regression method based on the overall distribution of the input variables, T represents the number of decision trees in the random forest algorithm, and n represents the total number of monthly data entries; Formula (11) means that the values ​​of the output variable Y are sorted in ascending order, and denoted as Y1, Y2, ... Y i-1 Y i Y i+1 For a given value y, find the values ​​Y of two adjacent output variables Y. i-1 and Y i If y lies exactly between the two values, then the quantile τ corresponding to the specified value y can be estimated using formula (12), where G is the ECDF (Empirical Cumulative Distribution Function), and G(Y) is the quantile τ of the specified value y. i-1 ) and G(Y i ) respectively represent less than or equal to Y i-1 and Y i The proportion of Y in Y, that is, the proportion of the output variable Y in Y i-1 To Y i It changes approximately linearly within a small interval.

8. The method as described in claim 7, characterized in that: Step 2 describes a total of three types of agents, including an external carbon source, a phosphorus removal agent, and a dehydration agent. The consumption efficiency of any agent k is calculated using formula (13): CP k =τ k (13) In formula (13) CP k τ represents the consumption efficiency of drug k. k Let be the quantile of the consumption intensity of agent k under the original treatment conditions.

9. The method as described in claim 8, characterized in that: The resource load of each plant mentioned in step two is obtained through formula (14): In formula (14), S Resource Loads To manage the resource load of each plant, W 2j and W 3k These are the weighting coefficients for the consumption efficiency allocated to energy j and medicine k, respectively.

10. The method as described in claim 9, characterized in that: The sustainability scores for each factory mentioned in step three are obtained using formula (15): S Sustainability =S Environmental Benefits -S Resource Loads (15) In formula (15) S Sustainability This assigns a sustainability score to each plant. The score represents the environmental benefits (S) generated by each plant through the removal of pollutants including COD, BOD, TN, NH3-N, TP, and SS. Environmental Benefits At the same time, resource load S will also be generated due to the consumption of electricity and heat, as well as external carbon sources, phosphorus removal agents and dehydration agents. Resource Loads The difference between the two represents the true sustainability of each plant. When a town's wastewater treatment plant removes more pollutants or consumes less energy and chemicals, S... Sustainability The larger the value, the greater the positive contribution of the factory to the ecology of the city.

Images