A modular intelligent sewage treatment management system

Abstract

The present invention discloses a modular intelligent sewage treatment management system, including: a data acquisition module for collecting land-related data, real-time sewage data, and historical sewage production-related data; a sewage treatment module for constructing a multi-objective optimization model based on real-time sewage data and sewage storage tank capacity data, and starting the sewage treatment process of several service areas according to the solution results; a sewage production prediction module for inputting land-related data and historical sewage production-related data of several service areas into the sewage production prediction model to obtain the predicted sewage production and the first predicted ferrous salt consumption in the next time period; a material purchase module for calculating the predicted demand for ferrous salt according to the output of the sewage production prediction module, and purchasing according to the predicted demand for ferrous salt of several sewage treatment plants. The present invention relates to the technical field of sewage treatment, and solves the technical problems of intelligent management and optimization control in the sewage treatment process.

Description

Technical Field

[0001] The present invention belongs to the field of sewage treatment, relates to equipment control technology, and specifically is a modular intelligent sewage treatment management system. Background Art

[0002] With the acceleration of urbanization and the development of industrial production, the demand for sewage treatment is growing and facing more complex situations. On the one hand, the urban population is increasing, the land use types are diversified, and the composition and quantity of sewage generated by commercial activities, industrial production and residents' lives are becoming more complex and diverse. On the other hand, the improvement of environmental awareness has put forward higher requirements for the quality and efficiency of sewage treatment, which not only ensures that sewage is discharged in compliance with standards, but also pays attention to resource recycling and cost control.

[0003] At present, many sewage treatment systems rely mainly on flow-based control methods to decide when to start equipment, that is, setting a fixed flow threshold. This method ignores changes in sewage quality and environmental conditions and has obvious limitations. For example, in the event of an accidental leak or failure in the pretreatment link during industrial production, the concentration of pollutants may rise sharply. Even if the flow rate does not reach the threshold, the equipment should be started immediately to avoid water quality deterioration and increased treatment costs. Unreasonable flow threshold settings can also cause equipment operation problems. If the flow threshold is too low, the equipment may start and stop frequently during heavy rain or peak industrial production periods, resulting in increased mechanical wear and energy consumption. On the contrary, if the flow threshold is too high, when the sewage flow is low but the water quality has begun to deteriorate, delayed equipment startup will lead to sewage accumulation and further deterioration of water quality, increasing treatment costs and environmental risks. The time-based control method is also unable to cope with sudden changes in sewage output, resulting in the inability of sewage treatment equipment to respond intelligently, which may cause sewage overflow, reduced treatment efficiency or environmental pollution, and thus affect the stability and efficiency of the entire sewage treatment system. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a modular intelligent sewage treatment management system to solve the problems of inadaptability and inefficiency of existing sewage treatment equipment control methods when dealing with complex sewage conditions.

[0005] To achieve the above objectives, the present invention provides a modular intelligent sewage treatment management system, comprising:

[0006] Data collection module: used to divide the collected land use type data according to the service area of ​​the sewage treatment plant, and collect the real-time sewage data and historical sewage production related data of several service areas, so as to obtain the land related data, real-time sewage data and historical sewage production related data of several service areas;

[0007] Sewage treatment module: used to build a multi-objective optimization model based on real-time sewage data and sewage storage tank capacity data, and solve the multi-objective optimization model to obtain the optimal equipment start-up time, and start the sewage treatment process of several service areas according to the optimal equipment start-up time;

[0008] Sewage production prediction module: used to input land-related data and historical sewage production-related data of several service areas into the sewage production prediction model to obtain the predicted sewage production and the first predicted ferrous salt consumption of several sewage treatment plants in the next time period; wherein the sewage production prediction model is constructed based on a deep learning algorithm;

[0009] Material purchasing module: used to calculate the predicted demand for ferrous salt based on the predicted sewage output in the next time period and the first predicted ferrous salt consumption, and to purchase ferrous salt based on the predicted demand for ferrous salt of several sewage treatment plants.

[0010] Based on the above technical modules, the present invention realizes intelligent control of the sewage treatment process. Among them, the sewage treatment module constructs a multi-objective optimization model, and according to the real-time sewage data and the sewage storage tank capacity data, scientifically and rationally decides on the start of the sewage treatment process, effectively balances the treatment efficiency, cost and other factors, and improves the overall treatment efficiency; the sewage production prediction module uses a deep learning algorithm to mine the laws of sewage production-related data, accurately predicts sewage production and ferrous salt consumption, and provides guidance for subsequent material preparation; the material purchase module: combined with the sewage production and ferrous salt consumption prediction results, calculates the demand for ferrous salt, realizes on-demand procurement, avoids material waste and supply shortages, and ensures the normal operation of the sewage treatment process.

[0011] Furthermore, the land-related data and sewage production-related data of the several service areas include:

[0012] Land-related data include: commercial land proportion, industrial land proportion, agricultural land proportion and residential land proportion;

[0013] Data related to historical sewage production include: traffic density, population, precipitation, number of heavy industrial plants, number of light industrial plants;

[0014] Real-time sewage data include: sewage inventory, pollutant types, pollutant concentration, ambient temperature, energy consumption cost, and equipment operating cost.

[0015] Furthermore, the multi-objective optimization model is constructed based on the real-time sewage data and the sewage storage tank capacity data, including:

[0016] A1, define the sewage overflow risk minimization function based on the sewage inventory and the sewage storage tank capacity;

[0017] A2, define the processing cost minimization function based on energy consumption cost and equipment operation cost;

[0018] A3, define the wastewater deterioration risk minimization function based on pollutant types, pollutant concentrations and ambient temperature;

[0019] A4, construct constraint conditions and form a multi-objective optimization model with the wastewater overflow risk minimization function, treatment cost minimization function and wastewater deterioration risk minimization function;

[0020] A5, using a multi-objective evolutionary algorithm to solve the multi-objective optimization model to obtain the optimal device startup time.

[0021] The multi-objective optimization function takes into account the risk of sewage overflow, treatment costs, and the risk of sewage deterioration. When starting the sewage treatment process, it does not focus on a single flow or time factor, but weighs various risks. For example, when the sewage stock is close to the capacity of the storage tank and there is a risk of overflow, even if the flow does not reach the traditional start threshold, the equipment can be started in advance according to the multi-objective optimization model to avoid sewage overflow and environmental pollution.

[0022] For the risk of sewage deterioration, it combines factors such as pollutant type, concentration and ambient temperature. In seasons with higher temperatures or when the concentration of pollutants in sewage changes, the start-up time can be adjusted in time to prevent the sewage from seriously deteriorating, leading to treatment difficulties and increased treatment costs.

[0023] Furthermore, the process of constructing the sewage overflow risk minimization function includes:

[0024] A11, calculate the average value of sewage storage within a preset period ;

[0025] A12, according to the formula Calculate the autocorrelation cumulative intensity S of sewage stock ACF ; where t represents the time index, Q t represents the sewage stock at time t, N represents the number of sewage stock data points within the preset period, and k represents the lag order, which is used to measure the sewage stock sequence {Q t}Correlation at different time intervals, k cut represents the lag order threshold, ACF(k) represents the autocorrelation function of sewage stock;

[0026] A13, according to the formula Calculate the sewage stock volatility factor VF;

[0027] A14, based on the sewage storage tank capacity W and VF, the sewage overflow risk minimization function is defined as: ; Where T represents the preset period, S t represents the amount of sewage in the sewage storage tank, α s represents the seasonal weight function, and P represents the probability.

[0028] The sewage overflow risk minimization function combines the mean, fluctuation and time series correlation of sewage stock, etc., and can more accurately assess the risk of sewage overflow. It is calculated based on real-time data and can dynamically adjust the assessment of overflow risk according to the changes in sewage stock. If the sewage stock suddenly increases or the fluctuation increases, the function can promptly reflect the increase in overflow risk, so that the sewage treatment equipment can be started in advance to avoid environmental pollution and equipment damage caused by sewage overflow.

[0029] Furthermore, the process of constructing the processing cost minimization function includes:

[0030] A21, through statistics, the fixed cost C of sewage treatment equipment startup is obtained eq0 and the variable cost C of the sewage treatment equipment per unit time eq1 ;

[0031] A22, using the straight-line depreciation method, the depreciation cost of the sewage treatment equipment at time t is calculated as: ; Where V0 represents the initial value of the sewage treatment equipment, L represents the service life, M represents the estimated total treatment capacity, and m t represents the amount processed at time t, and , r i represents the treatment capacity of the sewage treatment equipment at the i-th moment, x i represents a binary decision variable, and when the value is 1, it indicates starting the sewage treatment equipment, and when the value is 0, it indicates shutting down the sewage treatment equipment;

[0032] A23, according to the formula Calculate the equipment operating cost C eq ; where x t represents the binary decision variable at time t, and x i The meaning is the same;

[0033] A24, according to the formula Calculate the energy cost C en ; Where Et represents the electricity price at time t, a represents the energy consumption coefficient per unit of sewage treatment, η t Indicates the energy efficiency coefficient of sewage treatment equipment;

[0034] A25, according to the equipment operation cost and the energy consumption cost, the processing cost minimization function is obtained as follows: C=min(C eq +C en ).

[0035] Further, the η t The calculation formula is: ; Among them, η0 represents the initial energy consumption efficiency, and λ represents the energy consumption efficiency attenuation coefficient.

[0036] The treatment cost minimization function comprehensively considers various cost factors such as equipment startup, operation, depreciation, and energy consumption, and can accurately calculate the actual cost of sewage treatment, providing an accurate basis for cost control. In addition, during the startup and operation of the sewage treatment process, cost optimization decisions can be made based on real-time data (such as electricity prices, equipment status, sewage volume, etc.). For example, during the period of low electricity prices, the energy consumption cost can be reduced by adjusting the equipment operation time, thereby minimizing the treatment cost and improving the economic benefits of sewage treatment.

[0037] Furthermore, the process of constructing the wastewater deterioration risk minimization function includes:

[0038] A31, use online water quality monitoring instruments to obtain the pollutant type j and pollutant concentration in the sewage storage tank at time t, and obtain y j,t ;

[0039] A32, using a temperature sensor to obtain the ambient temperature of the sewage storage tank at time t, and obtain T(t);

[0040] A33, set the deterioration influence factor to CT j , CT max ,j represents the preset concentration threshold of pollutant type j;

[0041] A34, set the temperature influence factor to ; Among them, [T min ,T max ] indicates the preset suitable temperature range, T over represents the upper temperature limit that causes severe deterioration, T under Indicates the lower limit of temperature that causes severe deterioration;

[0042] A35, according to the deterioration impact factor and the temperature impact factor, the wastewater deterioration risk minimization function is obtained as follows: ; Among them, ω j The weight coefficient of pollutant j is used to indicate the relative importance of pollutant j to the deterioration of sewage in the sewage storage tank.

[0043] Sewage deterioration may cause pollutants in it to change, forming substances that are more difficult to treat. At the same time, seriously deteriorated sewage may produce harmful gases and breed harmful microorganisms during storage. These substances may escape into the surrounding environment and cause secondary pollution. By constructing a sewage deterioration risk minimization function, the risk of sewage deterioration can be monitored in real time. When the risk reaches a certain level, the sewage treatment process is started in time to reduce the residence time of sewage in the storage tank and reduce the risk of secondary pollution.

[0044] Furthermore, the constraints include:

[0045] Set the sewage storage constraint as: ;in, Indicates the processing capacity of the sewage treatment equipment at time t;

[0046] Set the sewage treatment capacity constraint as: ;

[0047] Set the sewage treatment time constraint as: ; Among them, DR th Indicates the preset sewage deterioration risk threshold.

[0048] Furthermore, the sewage production prediction model is constructed based on a deep learning algorithm, including:

[0049] B1, collecting historical sewage production-related data of several service areas at preset time intervals to obtain historical sewage production-related data, and collecting sewage production data and ferrous salt consumption data at corresponding times, and preprocessing the historical sewage production-related data, historical sewage stock data, historical ferrous salt consumption data and land-related data to obtain preprocessed data;

[0050] B2, based on the input data dimension and the output data dimension, a time series model is constructed based on the deep learning algorithm to obtain a time series prediction model; wherein the input data dimension is the sum of the number of several variables in the historical sewage production-related data and land-related data, and the output data dimension is 2, that is, the sewage production data and the ferrous salt consumption data are label data;

[0051] B3, divide the preprocessed data into training set, validation set and test set according to the preset ratio, input the training set and validation set into the time series model for iterative training and validation optimization, save the model parameters with the highest validation accuracy, and output the optimal model;

[0052] B4, input the test set into the optimal model for testing, obtain the test accuracy, and judge whether the test accuracy meets the preset accuracy requirements; if yes, output the optimal model as the sewage production prediction model; if not, adjust the parameters of the time series model and jump to B3.

[0053] Over time, the land use in the service area of ​​the sewage treatment plant may change (such as the development of new industrial areas, commercial areas, or the expansion of residential areas, etc.), and the population and industrial structure may also change. These changes will affect the sewage production and ferrous salt consumption. The sewage production prediction model can continuously learn new data patterns and adjust the prediction results in a timely manner to adapt to these dynamic changes, providing sewage treatment plants with more realistic prediction information.

[0054] Further, the calculation of the predicted demand for ferrous salt according to the predicted sewage output in the next time period and the first predicted consumption of ferrous salt includes:

[0055] C1, using a linear function and the least squares method to construct a fitting relationship between historical sewage stock data and historical ferrous salt consumption data, the fitting function is: y=kx+b; where x represents sewage stock data, y represents ferrous salt consumption data, and k and b represent the parameters of the fitting function;

[0056] C2, inputting the predicted sewage production in the next time period into the fitting function to obtain the second predicted ferrous salt consumption;

[0057] C3, performing weighted summation of the first predicted ferrous salt consumption and the second predicted ferrous salt consumption to obtain the predicted ferrous salt demand; wherein the weight coefficient of the weighted summation is adjusted based on historical applications.

[0058] By weighted summing the first predicted ferrous salt consumption and the second predicted ferrous salt consumption, and adjusting the weight coefficient according to historical applications, the accuracy of the predicted ferrous salt demand can be further improved. Different forecasting methods may have different advantages in different situations, and weighted summing can balance the advantages and disadvantages of the two methods. For example, when the sewage output fluctuates greatly but the water quality is relatively stable, the first predicted ferrous salt consumption based on the sewage output prediction model may be more reliable; while when the sewage output is relatively stable but the water quality changes greatly and affects the ferrous salt consumption, the second predicted ferrous salt consumption based on the fitting function may better reflect the actual demand. By reasonably adjusting the weight coefficient, the forecast results can be made closer to the actual ferrous salt demand according to the actual situation, thereby providing a more accurate basis for procurement tasks and reducing the risk of purchasing too much or too little ferrous salt.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] When starting the sewage treatment process, the sewage treatment module of the present invention constructs a multi-objective optimization model to comprehensively consider factors such as sewage overflow risk, treatment cost and sewage deterioration risk. Among them, the sewage overflow risk function comprehensively considers the mean value of sewage inventory, fluctuations, time series correlation and seasonal factors, and evaluates the overflow risk in a probabilistic form; the treatment cost function covers cost elements such as equipment startup, operation, depreciation and energy consumption, and scientifically calculates the treatment cost; the sewage deterioration risk function combines the type of pollutants, concentration and ambient temperature to reasonably evaluate the deterioration risk. By comprehensively analyzing these factors, the equipment startup time can be determined more accurately, avoiding the limitations of traditional methods based on single flow or time control decisions, making sewage treatment startup decisions more scientific and reasonable;

[0061] The constructed risk function can predict risks such as sewage overflow and deterioration in advance. The sewage overflow risk function predicts the possibility of overflow based on the historical data of sewage inventory and real-time fluctuations, and starts the treatment process in advance to prevent sewage overflow from polluting the environment and damaging equipment; the sewage deterioration risk function monitors the changes in pollutant concentration and ambient temperature in real time, and starts treatment in time before the risk of deterioration increases, so as to avoid sewage deterioration increasing the difficulty and cost of treatment;

[0062] The material purchase module calculates the predicted demand for ferrous salts based on the results of the sewage production prediction model. During the calculation process, the historical sewage inventory data and the historical ferrous salt consumption data are first used to construct a fitting function, and the second predicted ferrous salt consumption is obtained in combination with the predicted sewage production. The first predicted ferrous salt consumption and the second predicted ferrous salt consumption are then weighted and summed, and the weight coefficient is adjusted based on historical applications. This method can more accurately predict the actual demand for ferrous salts, avoiding excessive ferrous salt reserves due to inaccurate estimates, resulting in waste of resources, or insufficient reserves that affect the normal operation of sewage treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0064] Figure 1 A system framework diagram of a modular intelligent sewage treatment management system provided by the present invention;

[0065] Figure 2 A schematic diagram of a modular intelligent sewage treatment management system provided by the present invention;

[0066] Figure 3 A schematic diagram of a process for calculating the predicted demand for ferrous salt provided by the present invention. DETAILED DESCRIPTION

[0067] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0068] See also Figure 1-Figure 3 The embodiment of the present invention provides a modular intelligent sewage treatment management system, including:

[0069] Data collection module: used to divide the collected land use type data according to the service area of ​​the sewage treatment plant, and collect the real-time sewage data and historical sewage production related data of several service areas, so as to obtain the land related data, real-time sewage data and historical sewage production related data of several service areas;

[0070] Sewage treatment module: used to build a multi-objective optimization model based on real-time sewage data and sewage storage tank capacity data, and start the sewage treatment process in several service areas based on the solution results;

[0071] Sewage production prediction module: used to input land-related data and historical sewage production-related data of several service areas into the sewage production prediction model to obtain the predicted sewage production and the first predicted ferrous salt consumption of several sewage treatment plants in the next time period; wherein the sewage production prediction model is constructed based on a deep learning algorithm;

[0072] Material purchasing module: used to calculate the predicted demand for ferrous salt based on the predicted sewage output in the next time period and the first predicted ferrous salt consumption, and to purchase ferrous salt based on the predicted demand for ferrous salt of several sewage treatment plants.

[0073] It should be noted that the data acquisition module of the present invention is in communication connection with the sewage treatment module and the sewage production prediction module, and the sewage production prediction module is in communication connection with the material purchase module.

[0074] In order to realize intelligent control of the sewage treatment process, the data acquisition module of the present invention needs to perform the following operations:

[0075] First, collect data on land use types within the service area of ​​the sewage treatment plant from relevant geographic information system (GIS) databases, urban planning departments, or field surveys, including the distribution and area information of different types of areas such as commercial land, industrial land, agricultural land, and residential land;

[0076] Then, according to the service area boundary of the sewage treatment plant, the collected land use type data is divided to determine the proportion of each type of land use in each service area, so as to obtain the proportion of commercial land, industrial land, agricultural land and residential land in each service area, and obtain land-related data;

[0077] Next, the sewage inventory data is collected in real time through the sensor network installed at various key locations in the sewage treatment plant. Then, the types and concentrations of pollutants in the sewage are continuously monitored using online water quality monitoring instruments. For example, spectral analysis technology, electrochemical sensors and other equipment can detect the types and contents of common pollutants such as chemical oxygen demand (COD), biochemical oxygen demand (BOD), ammonia nitrogen, heavy metal ions, etc., and obtain real-time data on pollutant types and concentrations.

[0078] The ambient temperature will affect the activity of microorganisms and the chemical reaction rate in sewage, which in turn affects the sewage treatment process and water quality changes. Therefore, the temperature sensor in the sewage treatment plant is used to obtain the ambient temperature data in real time; and the equipment energy consumption cost data and equipment operation cost data are collected in real time to obtain the real-time sewage data;

[0079] At the same time, the data collection module also needs to obtain traffic flow data from the traffic management department, and calculate the average traffic density within a preset time interval (such as daily, weekly or monthly) by conducting statistical analysis on the traffic flow monitoring point data of the road network within the service area of ​​each sewage treatment plant; and collect historical population data, precipitation data, number of heavy industrial factories and number of light industrial factories in each service area within the preset time interval; and obtain historical sewage production related data.

[0080] In the sewage treatment module, a multi-objective optimization model is built based on real-time sewage data and sewage storage tank capacity data, and the sewage treatment processes of several service areas are started according to the optimal equipment start-up time obtained to minimize the cost of the sewage treatment process;

[0081] Among them, the construction process of the sewage overflow risk minimization function in the multi-objective optimization model is as follows:

[0082] First, calculate the mean value of sewage storage within the preset period , then according to the formula Calculate the autocorrelation cumulative intensity S of sewage stock ACF ; where t represents the time index, Q t represents the sewage stock at time t, N represents the number of sewage stock data points within the preset period, and k represents the lag order, which is used to measure the sewage stock sequence {Q t}Correlation at different time intervals, k cutrepresents the lag order threshold, ACF(k) represents the autocorrelation function of sewage stock;

[0083] The autocorrelation function ACF(k) measures the correlation of the sewage stock series at different time intervals. By analyzing this correlation, we can understand the trend and law of sewage stock changes, and calculating the cumulative autocorrelation intensity is to gain a deeper understanding of the correlation of sewage stocks in time series, that is, the impact of past sewage stock changes on the present and future. If there is a strong positive correlation, it means that the sewage stock may continue to increase or decrease, which helps to predict the overflow risk in advance;

[0084] Next, the sewage stock volatility factor VF is calculated, and the formula is: ;

[0085] Finally, according to the sewage storage tank capacity W and the volatility factor VF, the sewage overflow risk minimization function is defined as: , T represents the preset period, S t represents the amount of sewage in the sewage storage tank, α s represents the seasonal weight function, with a value of 1.5 in the rainy season and 1 in other seasons, and P represents the probability. This function quantifies the risk of sewage overflow in the form of probability by considering the real-time situation, volatility, and seasonal factors of sewage inventory, aiming to minimize this risk;

[0086] It should be noted that the seasonal factor in the sewage overflow risk minimization function is determined by the statistical analysis of historical precipitation data from the local meteorological department. Specifically, the rainy season can be defined as the period when the annual precipitation is relatively concentrated and the precipitation reaches a certain threshold (for example, the average monthly precipitation exceeds [X] mm for three consecutive months, and the X value can be set according to the local climate characteristics and the actual impact of rainfall on sewage generation);

[0087] In actual operation, by real-time monitoring of sewage stock data, the sewage overflow risk minimization function can dynamically calculate the risk of sewage overflow. For example, in the rainy season, the increase in precipitation leads to an increase in sewage production, and the seasonal weight function in the function will adjust the risk assessment accordingly. When the risk value exceeds the preset threshold, it indicates that there is a high possibility of sewage overflow. At this time, the system can start the sewage treatment equipment in time to avoid environmental pollution and equipment damage caused by sewage overflow. This real-time response mechanism can flexibly adjust the equipment startup time according to actual conditions, ensuring rapid intervention and processing when the risk is about to occur, rather than starting the equipment according to fixed time or flow standards, thereby improving the rationality and timeliness of equipment startup.

[0088] The calculation of the cumulative intensity of autocorrelation in the function helps to analyze the trend of sewage stock changes. If it is found that the sewage stock shows a continuous upward trend and the correlation is strong, even if the current stock has not reached the traditional dangerous level, the equipment can be started in advance to prevent potential overflow risks. This allows the equipment to start up in advance to predict risks, rather than passively waiting for problems to occur, effectively improving the sewage treatment system's ability to cope with sewage overflow risks.

[0089] The construction process of the processing cost minimization function in the multi-objective optimization model is as follows:

[0090] First, the fixed cost C of sewage treatment equipment startup is obtained through statistical analysis. eq0 and the variable cost C of the sewage treatment equipment per unit time eq1 ;

[0091] It should be noted that the variable cost of sewage treatment equipment per unit time refers to the cost that changes with the amount of sewage treatment during the sewage treatment process, including the cost of reagents, energy consumption costs and sludge treatment costs, which are obtained by recording and analyzing the historical reagent procurement details, equipment operation energy consumption data and various expenses in the sludge disposal process. These costs are directly related to the amount of sewage treated. When the amount of sewage treated increases, the variable cost increases accordingly; when the amount of sewage treated decreases, the variable cost also decreases.

[0092] Then, calculate the various costs incurred during sewage treatment:

[0093] Use the straight-line depreciation method to calculate the depreciation cost of sewage treatment equipment at time t: , where V0 represents the initial value of the sewage treatment equipment, L represents the service life, M represents the expected total treatment volume (obtained through a comprehensive evaluation of factors such as the equipment's treatment capacity and expected service life), and m t represents the amount of processing at time t (obtained through statistics of historical data of equipment operation), and , r i represents the treatment capacity of the sewage treatment equipment at time i, x i It represents a binary decision variable, and when the value is 1, it indicates starting the sewage treatment equipment, and when the value is 0, it indicates shutting down the sewage treatment equipment. The calculation of depreciation cost takes into account factors such as the initial value of the equipment, the service life, and the amount of treatment, and reasonably allocates the value loss of the equipment during use to each treatment moment;

[0094] According to the formula Calculate the equipment operating cost C eq , where x t represents a binary decision variable, which has the same meaning as x iThe formula combines the fixed cost, variable cost and depreciation cost of equipment startup, and calculates the cost of equipment startup according to the operating status of the equipment at each moment (given by x t Decision) to calculate the total equipment operating cost;

[0095] According to the formula Calculate energy cost C en , where Et represents the electricity price at time t, a represents the energy consumption coefficient per unit sewage treatment volume, η t Indicates the energy efficiency coefficient of sewage treatment equipment;

[0096] and , η0 represents the initial energy efficiency (obtained through the performance test report of the equipment when it leaves the factory), λ represents the energy efficiency attenuation coefficient (obtained through regular comprehensive performance evaluation of the equipment);

[0097] Finally, according to the equipment operation cost and energy consumption cost, the processing cost minimization function C=min(C eq +C en ), taking into account all the costs of equipment operation, and making optimization decisions with the goal of minimizing processing costs.

[0098] In order to timely evaluate and control the risk of sewage deterioration and prevent serious sewage deterioration from increasing the difficulty of treatment and the cost of treatment, the multi-objective optimization model also includes a sewage deterioration risk minimization function, and its construction process is as follows:

[0099] Use online water quality monitoring instruments to obtain the pollutant type j and pollutant concentration y in the sewage storage tank at time t j,t , and using a temperature sensor to obtain the ambient temperature T(t) of the sewage storage tank at time t;

[0100] Set the deterioration impact factor IC j (t), according to the pollutant concentration and the preset concentration threshold CT in different stages j The relationship between and determines its value: When the pollutant concentration is lower than the lower threshold, the deterioration impact factor is 0, indicating that the pollutant has little impact on sewage deterioration; when the concentration is within the threshold range, its value is calculated by a specific formula to reflect its contribution to the deterioration risk; when the concentration exceeds the upper threshold, the deterioration impact factor is 1, indicating that the pollutant may cause serious deterioration of sewage.

[0101] Set the temperature influence factor IT(t), according to the ambient temperature and the preset suitable temperature range [T min ,T max ] and the upper temperature limit T that causes severe deterioration over and lower limit T under The relationship is determined by: That is, when the temperature is within the suitable range, the temperature impact factor is 0, indicating that the temperature has little effect on sewage deterioration; when the temperature exceeds the suitable range, its value is calculated by a specific formula to reflect the degree of influence of temperature on the risk of deterioration. The higher or lower the temperature is beyond the suitable range, the greater its impact on the risk of deterioration.

[0102] According to the deterioration influence factor and temperature influence factor, the wastewater deterioration risk minimization function is obtained: , where ω j The weight coefficient of pollutant j is used to express the relative importance of pollutant j to the deterioration of sewage in the sewage storage tank, which is obtained through professional water quality monitoring analysis and expert experience evaluation;

[0103] Finally, construct constraints and form a multi-objective optimization model:

[0104] Set the sewage storage constraint as: , in order to limit the reasonable range of sewage storage, ensure that the sewage storage is always within the range that the storage tank can accommodate and can be treated normally, and avoid problems such as sewage overflow or low storage that affects treatment efficiency. t+1 Indicates the sewage storage in the next period, S t Indicates the sewage storage in the current period, Q t represents the sewage inflow at time t, r t represents the treatment capacity of the sewage treatment equipment at time t, x t represents a binary decision variable, W represents the maximum capacity of the sewage storage tank;

[0105] Set the sewage treatment capacity constraint as: , in order to clarify the upper limit of the amount of sewage that the sewage treatment equipment can treat in a unit time, to ensure that the actual operation of sewage treatment is in line with the equipment's treatment capacity, and to prevent the situation where the treatment effect is poor due to excessive treatment volume exceeding the equipment load;

[0106] Set the sewage treatment time constraint as: , where DR th It indicates the preset threshold of sewage deterioration risk, which stipulates that when the risk of sewage deterioration reaches a certain level, the sewage treatment equipment needs to be started immediately to prevent the sewage from further decaying and deteriorating.

[0107] In this embodiment, a multi-objective evolutionary algorithm (such as NSGA-II algorithm, etc.) is used to solve the constructed multi-objective optimization model. The multi-objective evolutionary algorithm has the ability to find the optimal balance solution between multiple objectives and is suitable for processing such complex multi-objective optimization problems.

[0108] During the solution process, the algorithm will continuously generate new solutions by simulating operations such as selection, crossover and mutation in the process of biological evolution, and evaluate the performance of these solutions on various objectives to find the time when the sewage storage tank is about to overflow, or the time when the sewage is about to deteriorate seriously, or the time when the treatment cost is minimized and the amount of sewage treated is the largest; when the time reaches any one of the above three, that is, as the optimal equipment start-up time, a start-up instruction is sent to the sewage treatment equipment control system to start the sewage treatment process in several service areas.

[0109] In the actual sewage treatment process, the working conditions are complex and changeable, and are affected by many factors (such as weather changes, fluctuations in sewage quality and quantity, equipment aging, etc.). The sewage overflow risk minimization function, the treatment cost minimization function, and the sewage deterioration risk minimization function work together to give the sewage treatment system strong adaptability and dynamic adjustment capabilities. When the sewage quality changes suddenly and the risk of deterioration increases, the deterioration risk minimization function will dominate the adjustment of the equipment startup time; if there are changes in cost factors such as electricity price fluctuations, the treatment cost minimization function will play a role; and in special periods such as the rainy season, the sewage overflow risk minimization function ensures that the system can respond to overflow risks in a timely manner. This synergy enables sewage treatment equipment to adjust the startup time in real time according to actual working conditions, effectively respond to various complex situations, and ensure the stability, efficiency and sustainability of the sewage treatment process.

[0110] Ferrous salts can be used in sewage treatment to remove heavy metal ions, phosphates and other pollutants in sewage by chemical precipitation. If the supply of ferrous salts is insufficient, the chemical reaction will be insufficient, pollutants cannot be effectively removed, and the treated sewage cannot meet the discharge standards. In addition, for some processes that combine biological treatment with chemical treatment, the appropriate addition of ferrous salts helps to regulate the redox potential of the microbial growth environment, promote the growth and reproduction of beneficial microorganisms, and improve the efficiency of biological treatment. At the same time, ferrous salts are also one of the important costs of sewage treatment. If the purchase amount exceeds the actual demand, it will cause waste of funds and inventory backlog. Excessive ferrous salts not only take up storage space, but may also deteriorate and fail due to long-term storage, further increasing cost losses.

[0111] Therefore, in this system, through the sewage production prediction module, the land-related data and historical sewage production-related data of several service areas obtained by the data acquisition module are input into the constructed sewage production prediction model to predict the predicted sewage production and the first predicted ferrous salt consumption of several sewage treatment plants in the next time period; then according to the material purchase module, the predicted ferrous salt demand is calculated according to the output of the sewage production prediction module to guide the sewage treatment plants to rationally plan the ferrous salt procurement plan.

[0112] Among them, the construction process of the sewage production prediction model in the sewage measurement prediction module includes:

[0113] First, according to the preset time intervals, the data collection module is used to collect the historical sewage production-related data and land-related data of several service areas, and the sewage production data and ferrous salt consumption data of the corresponding time are collected;

[0114] Next, preprocessing operations are performed on the collected historical sewage production-related data, historical sewage inventory data, historical ferrous salt consumption data, and land-related data; including but not limited to data cleaning, data encoding, and other operations, to obtain preprocessed data that can be used for model training;

[0115] Then, according to the input data dimension (i.e., the sum of the number of several variables in the historical sewage production-related data and land-related data) and the output data dimension (sewage production data and ferrous salt consumption data, which are 2-dimensional), a time series model is constructed based on deep learning algorithms (such as long short-term memory network LSTM, gated recurrent unit GRU, etc.);

[0116] The preprocessed data is divided into three data sets according to preset proportions (such as 70% for training set, 20% for validation set, and 10% for test set). First, the training set and validation set are input into the constructed time series model for iterative training. During the training process, the model continuously adjusts the model parameters according to the input data and the corresponding label data (wastewater production and ferrous salt consumption data) to minimize the prediction error. At the same time, the validation set is used to verify the model in the training process to verify the performance of the model on the validation set. After each iteration, the model parameters with the highest verification accuracy are saved. Through continuous iterative optimization, the model with the best performance on the validation set, that is, the optimal model, is finally obtained;

[0117] The test set is input into the optimal model for testing, and the test accuracy of the model on the test set is calculated (such as taking the F1 value as the test accuracy indicator); then determine whether the test accuracy meets the preset accuracy requirements. If the preset accuracy requirements are met, it means that the model performance is good and can accurately predict the sewage output and ferrous salt consumption. At this time, the optimal model is output as the sewage output prediction model, which can be used for actual prediction work; if the preset accuracy requirements are not met, adjust the parameters of the time series model (such as adjusting the number of hidden layers, learning rate, regularization coefficient, etc.), and then re-train, verify and optimize until a model that meets the accuracy requirements is obtained.

[0118] In order to improve the accuracy and reliability of model prediction, the input data of the sewage production prediction model include land-related data and historical sewage production-related data. Among them, land-related data reflects the distribution of sewage sources and adapts to regional changes; traffic density is related to personnel flow and sewage generation, reflecting the impact of urban dynamic changes; population size directly affects the amount of domestic sewage, reflecting long-term trends and seasonal fluctuations; precipitation affects the mixing of rainwater runoff and sewage and the influence of seasonal and climatic factors; the number of heavy and light industrial factories determines the scale of industrial sewage generation and tracks the impact of industrial structure adjustment.

[0119] In the material purchase module, the forecast demand for ferrous salt is calculated based on the output of the sewage production forecast module, including:

[0120] Taking the historical sewage stock data as the independent variable x and the historical ferrous salt consumption data as the dependent variable y, a fitting relationship between the two is constructed, and the values ​​of k and b are calculated by the least squares method to obtain a fitting function. This fitting function reflects the approximate quantitative relationship between sewage stock and ferrous salt consumption in the historical data, and can be used to predict ferrous salt consumption based on sewage stock;

[0121] Then, the predicted sewage production in the next time period obtained by the sewage production prediction module is used as input and substituted into the fitting function y=kx+b constructed above to calculate the second predicted ferrous salt consumption;

[0122] Next, the first predicted ferrous salt consumption obtained by the sewage production prediction module is obtained, and then the predicted ferrous salt demand Y is calculated according to the weighted summation formula, wherein the weight coefficient of the weighted summation is adjusted based on historical applications;

[0123] Finally, purchases are made based on the predicted demand for ferrous salts at several sewage treatment plants to ensure an adequate supply of ferrous salts during sewage treatment and avoid waste of resources due to excessive purchases.

[0124] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0125] Working principle of the present invention:

[0126] The present invention collects land use type and real-time and historical sewage production-related data through a data acquisition module to provide a data basis for subsequent modules; the sewage treatment module constructs a multi-objective optimization model including functions such as sewage overflow, treatment cost and deterioration risk and set constraints based on real-time data and storage tank capacity, and uses a multi-objective evolutionary algorithm to solve and obtain the optimal equipment start-up time to start the sewage treatment process; the sewage output prediction module uses a deep learning algorithm to take land and historical sewage production-related data as input to predict sewage output and ferrous salt consumption; finally, the material purchasing module calculates the ferrous salt demand according to the prediction results, and purchases on demand to ensure the operation of the sewage treatment process.

[0127] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A modular intelligent sewage treatment management system, characterized in that: include: Data collection module: used to divide the collected land use type data according to the service area of ​​the sewage treatment plant, and collect the real-time sewage data and historical sewage production related data of several service areas, so as to obtain the land related data, real-time sewage data and historical sewage production related data of several service areas; Sewage treatment module: used to build a multi-objective optimization model based on real-time sewage data and sewage storage tank capacity data, and solve the multi-objective optimization model to obtain the optimal equipment start-up time, and start the sewage treatment process of several service areas according to the optimal equipment start-up time; Sewage production prediction module: used to input land-related data and historical sewage production-related data into the sewage production prediction model to obtain the predicted sewage production and the first predicted ferrous salt consumption of several sewage treatment plants in the next time period; wherein the sewage production prediction model is constructed based on a deep learning algorithm; Material purchasing module: used to calculate the predicted demand for ferrous salt according to the predicted sewage output in the next time period and the first predicted consumption of ferrous salt, and several sewage treatment plants purchase according to the predicted demand for ferrous salt; The multi-objective optimization model is constructed based on the real-time sewage data and the sewage storage tank capacity data, including: A1, define the sewage overflow risk minimization function based on the sewage inventory and the sewage storage tank capacity; A2, define the processing cost minimization function based on energy consumption cost and equipment operation cost; A3, define the wastewater deterioration risk minimization function based on pollutant types, pollutant concentrations and ambient temperature; A4, construct constraint conditions and form a multi-objective optimization model with the wastewater overflow risk minimization function, treatment cost minimization function and wastewater deterioration risk minimization function; A5, using a multi-objective evolutionary algorithm to solve the multi-objective optimization model to obtain the optimal device startup time; The sewage production prediction model is constructed based on a deep learning algorithm, including: B1, collecting historical sewage production-related data of several service areas at preset time intervals to obtain historical sewage production-related data, and collecting sewage production data and ferrous salt consumption data at corresponding times, and preprocessing the historical sewage production-related data, historical sewage stock data, historical ferrous salt consumption data and land-related data to obtain preprocessed data; B2, based on the input data dimension and the output data dimension, a time series model is constructed based on the deep learning algorithm to obtain a time series prediction model; wherein the input data dimension is the sum of the number of several variables in the historical sewage production-related data and land-related data, and the output data dimension is 2, that is, the sewage production data and the ferrous salt consumption data are label data; B3, divide the preprocessed data into training set, validation set and test set according to the preset ratio, input the training set and validation set into the time series model for iterative training and validation optimization, save the model parameters with the highest validation accuracy, and output the optimal model; B4, input the test set into the optimal model for testing, obtain the test accuracy, and judge whether the test accuracy meets the preset accuracy requirements; if yes, output the optimal model as the sewage production prediction model; if not, adjust the parameters of the time series model and jump to B3.

2. A modular intelligent sewage treatment management system according to claim 1, characterized in that: The land-related data include: the proportion of commercial land, the proportion of industrial land, the proportion of agricultural land and the proportion of residential land; The historical sewage production-related data include: traffic density, population, precipitation, number of heavy industrial plants, number of light industrial plants; The real-time sewage data includes: sewage inventory, pollutant types, pollutant concentration, ambient temperature, energy consumption cost, and equipment operation cost.

3. A modular intelligent sewage treatment management system according to claim 2, characterized in that: The construction process of the sewage overflow risk minimization function includes: A11, calculate the average value of sewage storage within a preset period ; A12, according to the formula Calculate the autocorrelation cumulative intensity S of sewage stock ACF ; where t represents the time index, Q t represents the sewage stock at time t, N represents the number of sewage stock data points within the preset period, and k represents the lag order, which is used to measure the sewage stock sequence {Q t }Correlation at different time intervals, k cut represents the lag order threshold, ACF(k) represents the autocorrelation function of sewage stock; A13, according to the formula Calculate the sewage stock volatility factor VF; A14, based on the sewage storage tank capacity W and VF, the sewage overflow risk minimization function is defined as: ; Where T represents the preset period, S t represents the amount of sewage in the sewage storage tank, α s represents the seasonal weight function, and P represents the probability.

4. A modular intelligent sewage treatment management system according to claim 3, characterized in that: The process of constructing the processing cost minimization function includes: A21, through statistics, the fixed cost C of sewage treatment equipment startup is obtained eq0 and the variable cost C of the sewage treatment equipment per unit time eq1 ; A22, using the straight-line depreciation method, the depreciation cost of the sewage treatment equipment at time t is calculated as: ; Where V0 represents the initial value of the sewage treatment equipment, L represents the service life, M represents the estimated total treatment capacity, and m t represents the amount processed at time t, and , r i represents the treatment capacity of the sewage treatment equipment at the i-th moment, x i represents the binary decision variable at the i-th moment; A23, according to the formula Calculate the equipment operating cost C eq ; where x t represents the binary decision variable at time t; A24, according to the formula Calculate the energy cost C en ; Among them, E t represents the electricity price at time t, a represents the energy consumption coefficient per unit sewage treatment volume, η t Indicates the energy efficiency coefficient of sewage treatment equipment; A25, according to the equipment operation cost and the energy consumption cost, the processing cost minimization function is obtained as follows: C=min(C eq +C en ).

5. A modular intelligent sewage treatment management system according to claim 4, characterized in that: The η t The calculation formula is: ; Among them, η0 represents the initial energy consumption efficiency, and λ represents the energy consumption efficiency attenuation coefficient.

6. A modular intelligent sewage treatment management system according to claim 4, characterized in that: The construction process of the sewage deterioration risk minimization function includes: A31, use online water quality monitoring instruments to obtain the pollutant type j and pollutant concentration in the sewage storage tank at time t, and obtain y j,t ; A32, using a temperature sensor to obtain the ambient temperature of the sewage storage tank at time t, and obtain T(t); A33, set the deterioration influence factor to CT j , CT max ,j represents the preset concentration threshold of pollutant type j; A34, set the temperature influence factor to ; Among them, [T min ,T max ] indicates the preset suitable temperature range, T over represents the upper temperature limit that causes severe deterioration, T under Indicates the lower limit of temperature that causes severe deterioration; A35, according to the deterioration impact factor and the temperature impact factor, the wastewater deterioration risk minimization function is obtained as follows: ; Among them, ω j represents the weight coefficient of pollutant j, J represents the total number of pollutant types, and DR(t) represents the risk of sewage deterioration at time t.

7. A modular intelligent sewage treatment management system according to claim 6, characterized in that: The constraints include: Set the sewage storage constraint as: ;in, Indicates the processing capacity of the sewage treatment equipment at time t; Set the sewage treatment capacity constraint as: ; Set the sewage treatment time constraint as: ; Among them, DR th Indicates the preset sewage deterioration risk threshold.

8. A modular intelligent sewage treatment management system according to claim 1, characterized in that: The method of calculating the predicted demand for ferrous salt according to the predicted sewage output in the next time period and the first predicted consumption of ferrous salt comprises: C1, using a linear function and the least squares method to construct a fitting relationship between historical sewage stock data and historical ferrous salt consumption data, the fitting function is: y=kx+b; where x represents sewage stock data, y represents ferrous salt consumption data, and k and b are parameter values ​​determined by the least squares method; C2, inputting the predicted sewage production in the next time period into the fitting function to obtain the second predicted ferrous salt consumption; C3, performing weighted summation on the first predicted ferrous salt consumption and the second predicted ferrous salt consumption to obtain the predicted ferrous salt demand.

Images