Rainwater and sewage pipe network rainfall inflow and overflow load estimation method based on high-frequency inflow data of sewage treatment plant, storage medium and equipment

By using high-frequency inflow data and STL and HMM algorithms to identify overflow periods in sewage treatment plants, an overflow water volume and load estimation model is constructed, which solves the problem of identification and estimation of rainwater mixed into sewage pipelines, and reduces monitoring costs and increases water Effects of environmental management level.

CN118966045APending Publication Date: 2024-11-15BEIJING YINGTELIWEI ENVIRONMENTAL TECH CO LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202410922350.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-10
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The existing technology is difficult to effectively identify and estimate the situation of rainwater mixing into the sewage pipeline network, resulting in an increase in the risk of overflow overflow in the confluence system and the lack of real-time monitoring methods, which affects the efficiency of sewage treatment plants and downstream water quality, and puts pressure on urban water environment management.

Method used

The rainfall inflow and overflow load estimation method based on high-frequency inflow data of sewage treatment plants is used to identify the overflow period through time series decomposition STL algorithm and hidden Markov model HMM algorithm, dynamically identify the water volume and water quality of the benchmark inflow plant to build a city Overflow water volume and load estimation model calculates rainwater inflow, overflow water volume and overflow load, providing scientific data support.

Benefits of technology

It realizes rapid and accurate identification of the scale of rainfall inflow and overflow without relying on a huge water quality monitoring network, reduces monitoring costs, improves the scientificity and intelligence level of urban water environment management, and ensures the stable operation of sewage treatment plants and downstream water quality. .

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure FT_1
    Figure FT_1
  • Figure FT_2
    Figure FT_2
  • Figure FT_3
    Figure FT_3
Patent Text Reader

Abstract

The invention relates to a rainfall inflow and overflow load estimation method for a rain sewage pipe network based on high-frequency inflow data of a sewage treatment plant, a storage medium and equipment, and belongs to the technical field of urban water environment management. The method comprises the following steps: acquiring high-frequency water inlet monitoring data and regional rainfall high-frequency monitoring data of a sewage treatment plant; based on a time sequence decomposition STL algorithm and a hidden Markov model HMM algorithm, identifying an overflow occurrence time period of the rain and sewage pipe network; dynamically identifying the reference water inlet quantity and water quality concentration of the sewage treatment plant in the non-overflow period; constructing an urban overflow water quantity and load estimation model; and calculating to obtain the rainwater inflow, the overflow water quantity and the overflow load of each overflow event. By utilizing the technology provided by the invention, an urban management department can be helped to quickly identify and accurately grasp the overall overflow condition of the rain sewage pipe network, so that data support is provided for scientifically formulating a coping scheme, further checking and finally eliminating overflow points and improving urban water environment quality.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to a method, storage medium and equipment for estimating rainfall inflow and overflow load of a rainwater and sewage pipe network based on high-frequency inflow data of a sewage treatment plant, and belongs to the technical field of urban water environment management. Background Art

[0002] As the urbanization process continues to accelerate, the underground rainwater and sewage pipe network, as a key urban infrastructure, plays an increasingly prominent role in ensuring water safety and improving the water environment. Through the rapid discharge of the rainwater pipe network, surface runoff can be quickly transferred to the downstream, reducing the risk of upstream waterlogging. The sewage pipe network is responsible for transferring domestic sewage to the sewage treatment plant, and after purification to meet the standards, it forms a water resource that can be safely discharged and recycled. In most old urban areas, from the perspective of operating costs, rainwater and sewage pipe networks usually adopt a combined system. During large rainfall events, rainwater will mix into the sewage pipe network, causing a combined system overflow (CSO). After rainwater mixes with sewage, it will occupy the transfer space and capacity of the sewage pipe network, dilute domestic sewage, and significantly reduce the water quality concentration of the sewage treatment plant, resulting in the inability to meet the living environment of microorganisms in the sewage treatment link, thereby greatly reducing the treatment efficiency of the sewage treatment plant and increasing operating costs. At the same time, the combined system overflow of the sewage pipe network is often directly discharged into the downstream water body, causing downstream water pollution. Even in areas where the rainwater-sewage separation has been relatively well-developed, the complex connection of the underground pipe network and the slope of the terrain may still cause rainwater to mix into the sewage pipe network during rainfall, causing combined system overflow pollution. At the same time, the trend of climate change around the world has also led to a significant increase in the frequency and intensity of rainfall, causing the surface runoff flow brought by rainfall to far exceed the design value of the pipe network, further increasing the risk of rainwater mixing into the sewage pipe network, and bringing great pressure to the reliable operation of the rainwater and sewage pipe network and urban water resources management. In order to reduce or even eliminate the risk of combined sewer overflows, the first thing to do is to be able to promptly identify the occurrence of overflows and scientifically estimate the scale of overflows, so as to provide timely data support for subsequent on-site inspections and disposal. However, since the rainwater and sewage pipes are buried underground, and the water volume and water quality monitoring equipment are expensive and difficult to install, it is impossible to carry out large-scale real-time monitoring. Therefore, it is not realistic to rely solely on the monitoring network to identify and estimate overflows. In addition, the sudden occurrence and time-varying characteristics of overflows also bring more complex challenges to carrying out effective responses and formulating scientific management strategies. To this end, it is necessary to develop an overflow identification and estimation method that does not rely on a large water quantity and quality monitoring network, to help management departments quickly and accurately identify the mixing of rainwater into the sewage network, and to estimate the scale of overflow caused by the mixing of rainfall inflow, thereby providing reliable data support for the management department's response decisions. Summary of the invention

[0003] The purpose of the present invention is to provide a method, storage medium and device for estimating rainfall inflow and overflow load based on high-frequency inflow data of sewage treatment plants. The present invention will be able to provide more detailed and comprehensive data basis for management decisions related to combined sewer overflow, help management departments accurately identify rainfall inflow and estimate the overflow scale caused by inflow, thereby improving the scientificity and intelligence level of urban water environment management. The technical solution is as follows: Step 1: Obtain high-frequency water inflow monitoring data and rainfall monitoring data of the sewage treatment plant; Step 2: Identify the overflow period based on the time series decomposition STL algorithm and the hidden Markov model HMM algorithm; Step 3: Dynamically identify the baseline inflow and water quality of the sewage treatment plant during the non-overflow period; Step 4: Construct an urban overflow water volume and load estimation model; Step 5: Calculate the rainwater inflow, overflow volume and overflow load of each overflow event; Through the above technical solution, the mixing of rainfall inflow in the sewage pipe network in the study area can be quickly identified, and the corresponding rainwater mixing amount, overflow water volume and overflow load of different indicators within the water collection range of each sewage treatment plant can be estimated, thereby providing data support for subsequent on-site inspections and response measures. Preferably, the acquisition of high-frequency inflow monitoring data and rainfall monitoring data of sewage treatment plants specifically includes collecting inflow water volume, ammonia nitrogen concentration, total nitrogen concentration, total phosphorus concentration and COD of a single sewage treatment plant. Cr The hourly scale monitoring data of concentration is used to identify the water collection range of the sewage treatment plant based on the upstream and downstream relationships and outlet nodes in the pipe network topology, so as to obtain the hourly scale monitoring data of rainfall at at least one rainfall monitoring station within the water collection range. Preferably, the overflow occurrence period is identified based on the time series decomposition STL algorithm and the hidden Markov model HMM algorithm, specifically including data cleaning of the high-frequency inlet monitoring data of the sewage treatment plant, and identifying the inlet ammonia nitrogen concentration, total nitrogen concentration, total phosphorus concentration and COD based on the time series decomposition STL algorithm. Cr The concentration and inflow water volume trends are based on the Hidden Markov Model (HMM) algorithm to identify the low-value periods of ammonia nitrogen concentration and the high-value periods of inflow water, and the rainfall monitoring data to identify the rainfall events, so as to take the intersection of the low-value periods of ammonia nitrogen concentration, the high-value periods of inflow water and the rainfall periods to obtain the overflow period. Among them: 1) Cleaning of high-frequency inbound monitoring data of sewage treatment plants: In order to ensure the accuracy and reliability of the data, the original monitoring data is cleaned. The data cleaning process mainly includes two steps: extreme high value removal and zero negative value removal. First, extreme high value removal is to exclude extreme outliers that may be caused by equipment failure or abnormal conditions. When performing this step, a threshold is set, and data points exceeding the threshold will be regarded as outliers and eliminated. Secondly, zero negative value elimination is to exclude zero or negative values ​​in the inbound monitoring data. In actual operation, the inbound monitoring data of sewage treatment plants sometimes have zero or negative values, which may be due to data anomalies caused by sensor failure, equipment shutdown or other reasons. In order to ensure the accuracy of the data, zero and negative values ​​need to be eliminated to avoid affecting subsequent analysis and calculations. 2) Identify the trend of inlet ammonia nitrogen concentration and inlet water volume based on the time series decomposition STL algorithm: The time series decomposition STL algorithm is a powerful time series decomposition method. By performing local regression smoothing on the time series, it can be decomposed into three parts: trend, seasonality, and residual. It can help understand and analyze time series at different time scales and has a wide range of applications in the fields of denoising and time series data analysis. Through the STL algorithm, trend components can be extracted from the time series data of inlet ammonia nitrogen concentration and inlet water volume. The trend part can reveal the long-term change trend in the data, such as the gradual increase or decrease in ammonia nitrogen concentration or water volume. This decomposition method can help identify potential change patterns and promptly identify possible situations where rainfall inflow is mixed into the sewage network. The calculation formula is: Y(t)=T(t)+S(t)+R(t) Among them, Y(t) represents the original monitoring data time series of the ammonia nitrogen concentration or water volume entering the plant; T(t) represents the trend part, which reflects the long-term trend of the data; S(t) represents the seasonal part, which shows the periodic changes of the data; R(t) represents the residual part, which represents the random noise that is not explained by the trend and seasonality. Estimation of seasonal component: S(t)=D(t)+I(t) Among them, D(t) represents the low-frequency component of the seasonal part, representing the long-term changes of seasonality; I(t) represents the high-frequency component of the seasonal part, representing the short-term fluctuations of seasonality. Smoothed estimate of the seasonal component: D(t)=SMOOTH(X(t),span) Among them, SMOOTH represents the smoothing function, X(t) represents the seasonal part of the original time series, and span represents the size of the smoothing window. Estimation of the trend component: T(t)=L(t)+H(t) Among them, L(t) represents the low-frequency component of the trend, representing the smooth change of the trend; H(t) represents the high-frequency component of the trend, representing the rapid change of the trend. Smoothed estimate of the trend component: L(t)=SMOOTH(Y(t),span) Wherein, Y(t) represents the time series monitoring data of the original ammonia nitrogen concentration or water volume entering the plant, SMOOTH represents the smoothing function, and span represents the size of the smoothing window. 3) Identify periods of low ammonia nitrogen concentration and high inflow water volume based on the Hidden Markov Model (HMM) algorithm: The HMM algorithm uses a probability matrix to establish the relationship between hidden states and observation sequences. In the application of identifying periods of low ammonia nitrogen concentration and high inflow water volume, the hidden state can represent a specific state of the water body, such as low ammonia nitrogen concentration or high inflow water volume, while the observation sequence corresponds to the actual observed ammonia nitrogen concentration and inflow water volume data. By training the HMM model, the known ammonia nitrogen concentration and inflow water volume data can be used to estimate the state transition probability matrix A, the observation probability matrix B, and the initial state probability vector π to describe the dynamic evolution process of the system; then, by calculating the forward probability and backward probability, the probability of the system being in a specific state under a given observation sequence can be obtained. Based on these probabilities, the state can be predicted to identify periods of low ammonia nitrogen concentration and high inflow water volume. The calculation formula is: State transition probability (Transition Probability): A={a_ij} Among them, A represents the state transition probability matrix, and a_ij represents the probability of transitioning from state i to state j. Emission Probability: B={b_j(k)} Where B represents the observation probability matrix, and b_j(k) represents the probability of observing observation value k in state j. Initial State Probability: π={π_i} Among them, π represents the initial state probability vector, and π_i represents the probability of being in state i at the initial moment. Forward Probability: α(t,j)=P(O1,O2,...,Ot,q(t)=j|λ) Among them, α(t,j) represents the probability that the system is in state j at time t and observes O1, O2, ..., Ot, and λ represents the parameters of the HMM model. Backward Probability: β(t,j)=P(O(t+1),O(t+2),...,OT|q(t)=j,λ) Among them, β(t,j) represents the probability of the system being in state j at time t and the observed value from time t+1 to the last time T, and λ represents the parameters of the HMM model. 4) Identify rainfall events: The criteria for identifying rainfall events are defined as total rainfall greater than 0.2 mm and no rainfall for more than 6 hours. The start and end time of a single rainfall event can be obtained through rainfall monitoring data, thereby identifying the rainfall period and non-rainfall period within a specific start and end time. 5) Identify overflow periods based on periods of low ammonia nitrogen concentration, high inflow water volume and rainfall periods: First, identify the start and end times of each rainfall event through rainfall monitoring data; second, identify periods of low inflow ammonia nitrogen concentration and high inflow water volume through the HMM algorithm; finally, take the intersection of the rainfall event time period, the low ammonia nitrogen concentration time period and the high inflow water volume time period to determine the period when combined sewer overflow occurs. Preferably, the baseline inflow water volume and water quality of the sewage treatment plant in the non-overflow period are dynamically identified, specifically, firstly based on the identification result of the combined system overflow event period in step 2, determine the time period between the two overflow events, that is, the time period between the end of overflow event A and the beginning of overflow event B, which is the time period when no rainwater mixing and sewage pipe network overflow occurs; then, for each time point in each non-overflow period, extract the corresponding monitoring value of inflow water volume and ammonia nitrogen concentration; finally, perform statistical analysis on the monitoring data in each non-overflow period, calculate the mean of inflow water volume and ammonia nitrogen concentration, as the inflow water volume and ammonia nitrogen concentration of the sewage treatment plant in this period when there is no rainwater mixing and sewage pipe network overflow, that is, the baseline inflow water volume and ammonia nitrogen concentration of the next overflow event after the non-overflow period. . Preferably, a model for estimating urban overflow water volume and load is constructed, specifically including identifying the water collection range of the sewage treatment plant based on the topological relationship of the pipe network, obtaining the characteristic pipe length and characteristic pipe diameter data of the sewage pipe network within the water collection range of the sewage treatment plant, dynamically calculating the amount of rainwater mixed in each simulation time step and the overflow water volume of the sewage pipe network based on the principles of water balance and mass balance, and dynamically calculating the overflow load of different water quality indicators within each simulation time step. Among them: 1) Based on the topological relationship of the pipe network, identify the water collection range of the sewage treatment plant: First, determine the location of the sewage treatment plant through the geographic information system (GIS) data of the study area, and then determine the water inlet node of the sewage treatment plant as the starting point for upstream tracing; determine the upstream of the water inlet node of the sewage treatment plant according to the elevation and connection relationship of the sewage pipeline node; analyze upstream along the sewage pipeline node according to the node elevation and connection relationship until all upstream nodes are traversed. All upstream areas where domestic sewage is discharged into these nodes constitute the water collection range of the sewage treatment plant. 2) Obtain the characteristic pipe length and characteristic pipe diameter data of the sewage pipe network within the water collection range of the sewage treatment plant: First, collect the sewage pipe network data within the water collection range of the sewage treatment plant, which includes the location, length, diameter and other information of the sewage pipe; determine the upstream sewage pipeline of each sewage treatment plant with the water collection range as the boundary, and extract its data; calculate the characteristic pipe length of each sewage pipe based on the extracted pipeline data. The characteristic pipe length refers to the effective length of the pipe length and the water flow path, that is, the effective length of the pipe after considering factors such as flow direction, elbows, and branches; determine the characteristic pipe diameter data of each sewage pipe. The characteristic pipe diameter refers to the effective flow area in the pipe, which is usually expressed by the inner diameter or equivalent diameter of the pipe. The characteristic pipe diameter information can be obtained through pipeline design documents and on-site measurements. 3) Dynamically calculate the amount of rainwater mixed into and overflow water volume of the sewage pipe network at each time step during the overflow period: Based on the principles of water balance and mass balance, calculate the amount of rainwater mixed into and overflow water volume at each time step respectively. The calculation formula is: V=π*r 2 *L Q3=Q1+Q2-Q4 Among them, V is the network volume, L is the characteristic pipe length of the sewage network, r is the characteristic pipe diameter of the sewage network, Q1 is the benchmark inflow water volume during the non-overflow period, C1 is the benchmark inflow ammonia nitrogen concentration during the non-overflow period, Q2 is the amount of rainwater mixed in, C2 is the inflow ammonia nitrogen concentration after mixing with rainwater, Q3 is the overflow water volume, ΔC3 is the change in the inflow ammonia nitrogen concentration of the sewage treatment plant, and Q4 is the inflow water volume. 4) Dynamically calculate the overflow load of different water quality indicators of the sewage network at each time step during the overflow period: Dynamically calculate ammonia nitrogen, total nitrogen, total phosphorus, COD cr The overflow load is calculated as follows: in, is the overflow load of the water quality index numbered n at time i, is the concentration of water quality index numbered n entering the factory at time i. Preferably, the rainwater inflow, overflow volume and overflow load of each overflow event are calculated, specifically: the rainwater mixing volume, overflow volume, ammonia nitrogen overflow load, total nitrogen overflow load, total phosphorus overflow load and COD in each overflow period calculated according to the urban overflow volume and load estimation model cr Overflow load: sum all overflow events within the target period to obtain the total rainwater mixing amount, overflow water volume, ammonia nitrogen overflow load, total nitrogen overflow load, total phosphorus overflow load and COD within the target period. cr Overflow load. In summary, the present application has the following beneficial technical effects: 1. The rainfall inflow and overflow load estimation technology for the rainwater and sewage pipe network based on the high-frequency inflow data of the sewage treatment plant proposed in this application can effectively reduce the high dependence of conventional rainfall inflow and overflow analysis methods on monitoring data. There is no need to carry out comprehensive and high-frequency monitoring of the hydrology and water quality of the sewage pipe network. Only relying on conventional rainfall monitoring and sewage treatment plant inlet monitoring data can more accurately estimate the rainfall inflow and overflow load of different water quality indicators when combined system overflow occurs, which can significantly save monitoring costs and workload. 2. The rainfall inflow and overflow load estimation technology for the rainwater and sewage pipe network based on the high-frequency inflow data of the sewage treatment plant proposed in this application can be deployed online and run in a rolling manner, read the rainfall monitoring data and the sewage treatment plant inlet monitoring data, and calculate in real time to identify whether rainfall inflow mixing has occurred in the sewage pipe network system, and the corresponding overflow load when mixing occurs, so as to enable the management department to quickly deploy response measures and conduct on-site inspections to gain valuable response time, ensure the stable operation of the sewage treatment plant, and reduce the risk of water environment pollution downstream. 3. The rainfall inflow and overflow load estimation technology for the rainwater and sewage pipe network based on the high-frequency inflow data of the sewage treatment plant proposed in this application establishes a relatively accurate and reliable combined sewer overflow identification and analysis framework through intelligent data analysis and modeling methods. This technology is not affected by the randomness and spatiotemporal variability of combined sewer overflow events, and is universal. It can be quickly deployed and applied to the operation status analysis of any urban rainwater and sewage pipe network, significantly improving the efficiency of urban water environment management, and has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0004] Figure 1 It is a flow chart of a method for estimating rainfall inflow and overflow loads of a rainwater and sewage pipe network based on high-frequency inflow data of a sewage treatment plant in an embodiment of the present application. Figure 2 It is a schematic diagram of the S002 process in the embodiment of the present application. Figure 3 It is a schematic diagram of the S004 process in the embodiment of the present application. Figure 4This is a graph showing the trend decomposition result of the time series of ammonia nitrogen concentration entering the plant based on the STL algorithm in the embodiment of the present application. Figure 5 This is a graph showing the trend decomposition result of the time series of water volume entering the plant based on the STL algorithm in an embodiment of the present application. Figure 6 This is a diagram showing the result of identifying the time series state of ammonia nitrogen concentration entering the plant based on the HMM algorithm in an embodiment of the present application. Figure 7 This is a diagram showing the result of identifying the time series state of the water volume entering the plant based on the HMM algorithm in an embodiment of the present application. DETAILED DESCRIPTION

[0005] The following will be combined with the accompanying drawings in the examples of the present invention to clearly and completely describe the technical solutions in the examples of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not 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. Example refer to Figure 1 , a method for estimating rainfall inflow and overflow load of a rainwater and sewage pipe network based on high-frequency inflow data of a sewage treatment plant mainly includes the following steps: S001: Obtain high-frequency monitoring data of water inflow and rainfall at sewage treatment plants. Select the sewage treatment plants to be studied and collect high-frequency online monitoring data of each sewage treatment plant. Common indicators include inflow water volume, ammonia nitrogen concentration, total nitrogen concentration, total phosphorus concentration and COD Cr The hourly scale monitoring data of concentration is used to identify the water collection range of the sewage treatment plant based on the upstream and downstream relationships and outlet nodes in the pipe network topology, so as to obtain the hourly scale monitoring data of rainfall at at least one rainfall monitoring station within the water collection range. S002: Based on the STL and HMM algorithms, the overflow period is identified, which specifically includes data cleaning of the high-frequency inlet monitoring data of the sewage treatment plant, where the abnormal identification algorithm can use the 3sigma algorithm; based on the STL algorithm, the inlet ammonia nitrogen concentration (such as Figure 4 ), total nitrogen concentration, total phosphorus concentration and COD Cr Concentration and inlet water volume (such as Figure 5 ) trend, and analyzed it using the statsmodels module of Python software; based on the HMM algorithm, the low-value periods of ammonia nitrogen concentration (such as Figure 6 ) and periods of high water inflow (such as Figure 7), and analyzed using the hmmlearn module of Python software; the rainfall events were identified based on the rainfall monitoring data, and the intersection of the low-value period of ammonia nitrogen concentration, the high-value period of water inflow and the rainfall period was taken to obtain the overflow occurrence period. S003: Dynamic identification of baseline inflow and water quality of sewage treatment plants. Specifically, the identified overflow events are numbered, and the same number is used for the non-overflow period before each overflow event, that is, each overflow event can be matched with a non-overflow event; the average inflow and average ammonia nitrogen concentration of each non-overflow event are calculated, and the results are used as the baseline inflow and ammonia nitrogen concentration corresponding to the overflow events with the same number. S004: Construct an urban overflow water volume and load estimation model. The input data of the model are the characteristic pipe length and characteristic pipe diameter of the sewage pipe network within the water collection range of the sewage treatment plant, the high-frequency water volume and water quality monitoring data of the sewage treatment plant, and at least one rainfall monitoring time series within the water collection range; the calculation principle of the model is water volume balance and mass balance; the output results of the model are the rainwater mixing volume, overflow water volume and overflow load of different water quality indicators in each simulation time step of all overflow events during the study period. S005: Calculate the rainwater inflow, overflow volume and overflow load of each overflow event. Specifically, the simulation results of the urban overflow volume and load estimation model are summed according to the overflow period to obtain the total rainwater mixing volume, overflow volume, ammonia nitrogen overflow load, total nitrogen overflow load, total phosphorus overflow load and COD of each overflow event. Cr Overflow load. Among them, Figure 2 As shown, the identification of overflow occurrence period based on STL and HMM algorithm mainly includes the following steps: S201: Cleaning of high-frequency inflow monitoring data of sewage treatment plants: remove extremely high values ​​and zero negative values ​​from the inflow water volume time series and the concentration time series of different inflow water quality indicators. S202: Identify the changing trend of the ammonia nitrogen concentration and water volume entering the plant based on the STL algorithm: Decompose and extract the time series of the changing trend of the ammonia nitrogen concentration and water volume entering the plant through the STL algorithm. The decomposition results of the time series trend of the ammonia nitrogen concentration entering the plant based on the STL algorithm are as follows: Figure 4 As shown in the figure, the results of the decomposition of the time series trend of the water inflow based on the STL algorithm are as follows: Figure 5 shown. S203: Identify the low-value period of ammonia nitrogen concentration and the high-value period of water inflow based on the HMM algorithm: Use the HMM algorithm to identify the low-value period of ammonia nitrogen concentration and the high-value period of water inflow. The results of the time series state recognition of the ammonia nitrogen concentration inflow based on the HMM algorithm are as follows: Figure 6 As shown in the figure, the results of the time series state recognition of the water inflow based on the HMM algorithm are as follows: Figure 7shown. S204: Identify rainfall events: Obtain the start time and end time of rainfall events within the study period through rainfall event identification criteria, thereby identifying rainfall events. S205: Identify overflow occurrence periods based on periods of low ammonia nitrogen concentration, high inflow water volume and rainfall periods: Take the intersection of the identification results in S203 and S204 to find periods during the rainfall event when both low ammonia nitrogen concentration and high inflow water volume are present, which are overflow occurrence periods. Among them, Figure 3 As shown in Figure 1, the construction of the urban overflow water volume and load estimation model mainly includes the following steps: S401: Based on the pipe network topology, identify the water collection range of the sewage treatment plant: Take the sewage treatment plant water inlet node as the tracing starting point, find the upstream node connected to the sewage treatment plant water inlet according to the upstream and downstream topological relationship of the pipeline, until all upstream nodes are traversed, thereby identifying the water collection range of the sewage treatment plant. S402: Obtain the characteristic pipe length and characteristic pipe diameter data of the sewage pipe network within the water collection range of the sewage treatment plant: collect the information such as the location, length, diameter, etc. of the sewage pipe network within the water collection range of the sewage treatment plant. The characteristic pipe length is the total length of all pipes within the water collection range, and the characteristic pipe diameter is the weighted average pipe diameter according to the length. S403: Dynamically calculate the amount of rainwater mixed into and overflow water volume for each simulation time step: Based on the principles of water balance and mass balance, the amount of rainwater mixed into and overflow water volume of the sewage network is calculated for each overflow event. S404: Dynamically calculate the overflow load of different water quality indicators for each simulation time step: Based on the overflow water volume and inlet water quality concentration at each moment during the overflow period, calculate the ammonia nitrogen, total nitrogen, total phosphorus, COD at each moment. Cr Overflow load.

Claims

1. A method, storage medium and device for estimating rainfall inflow and overflow loads based on high-frequency inflow data of a sewage treatment plant, characterized in that: The method comprises: Step 1: Obtain high-frequency water inflow monitoring data and rainfall monitoring data of the sewage treatment plant; Step 2: Identify overflow periods in the rainwater and sewage pipe network based on the time series decomposition STL algorithm and the hidden Markov model HMM algorithm; Step 3: Dynamically identify the baseline inflow volume and quality of the sewage treatment plant during non-overflow periods; Step 4: Construct an overflow water volume and load estimation model for the urban rainwater and sewage pipe network; Step 5: Calculate the rainwater inflow, overflow volume and overflow load for each overflow event.

2. The method according to claim 1, characterized in that The first step is to obtain high-frequency monitoring data of sewage treatment plant water inflow and rainfall, including collecting sewage treatment plant water inflow, ammonia nitrogen concentration, total nitrogen concentration, total phosphorus concentration and COD Cr Hourly monitoring data on concentration and hourly rainfall monitoring data at at least one rainfall monitoring station within the water intake range of the sewage treatment plant.

3. The method according to claim 1, characterized in that The step 2 of identifying the overflow period of the rainwater and sewage pipe network based on the time series decomposition algorithm STL and the hidden Markov model algorithm HMM refers to cleaning the high-frequency inlet monitoring data of the sewage treatment plant, and identifying the inlet ammonia nitrogen concentration, total nitrogen concentration, total phosphorus concentration, COD concentration based on the time series decomposition algorithm STL. Cr Based on the changing trends of ammonia nitrogen concentration and inflow water volume, the hidden Markov model algorithm HMM is used to identify the periods of low ammonia nitrogen concentration and high inflow water volume, and the rainfall events are identified based on the rainfall monitoring data. The intersection of the periods of low ammonia nitrogen concentration, high inflow water volume and rainfall periods is then taken to obtain the periods when overflow occurs in the rainwater and sewage pipe network.

4. The method according to claim 1, wherein The dynamic identification of the baseline inflow volume and water quality of the sewage treatment plant during the non-overflow period in step 3 refers to the dynamic identification of the baseline inflow volume and water quality based on the concentration of ammonia nitrogen, total nitrogen, total phosphorus and COD. Cr The trend analysis results of concentration and influent volume are used to calculate the influent volume and water quality of the sewage treatment plant if no rainwater mixing occurs and no overflow occurs during each overflow period. The main method is to find the mean of influent volume and water quality concentration in the time period between two overflow events.

5. The method according to claim 1, wherein The overflow water volume and load estimation model of the urban rainwater and sewage pipe network is constructed in the fourth step, which refers to identifying the water collection range of the sewage treatment plant based on the topological relationship of the pipe network, obtaining the characteristic pipe length and characteristic pipe diameter data within the collection range of the sewage treatment plant, and dynamically calculating the rainwater mixing volume and overflow water volume for each simulation time step based on the principles of water balance and mass balance, and dynamically calculating the overflow load of different water quality indicators for each simulation time step.

6. The method according to claim 1, wherein The calculation of the rainwater inflow, overflow volume and overflow load of each overflow event in step 5 refers to running the overflow volume and load estimation model of the urban rainwater and sewage pipe network, summarizing the model calculation results, and obtaining the rainwater mixing volume, overflow volume, ammonia nitrogen overflow load, total nitrogen overflow load, total phosphorus overflow load and COD corresponding to each overflow event. Cr Overflow load.

Citation Information

Cited By

  • Multi-segmentation-point sewage distribution method and equipment considering spatial heterogeneity of sewage pipe network

    CN120450239A

  • Sewage pipe network rainfall-inflow event identification method based on semi-supervised iterative training

    CN120597183A