Full pipe operation early warning analysis method for sewage pipe network

Abstract

The invention relates to the technical field of municipal pipeline detection, and discloses a sewage pipe network full pipe operation early warning analysis method comprising the following steps: S1, arranging sensors at key nodes of a sewage pipe network, and collecting real-time operation data of the sewage pipe network; s2, acquiring real-time meteorological data including rainfall capacity, rainfall intensity and other related meteorological factors; s3, performing data cleaning and standardization processing on the collected sewage pipe network data and meteorological data; s4, on the basis of multivariable regression analysis, establishing a relation model among the flow, the liquid level and the pipe fullness risk; and S5, performing time sequence prediction by using a long short-term memory network, and predicting future flow and liquid level trend. Time sequence prediction is carried out through the long-short-term memory network, the early warning threshold value is dynamically adjusted in combination with the self-adaptive algorithm, the running state change of the sewage pipe network is responded in real time, the future flow and liquid level trend is accurately predicted, and the early warning accuracy and the response speed to the sudden pipe fullness risk are improved.

Description

Technical Field

[0001] The invention relates to the technical field of municipal pipeline detection, and in particular to a full-pipe operation early warning analysis method for a sewage pipe network. Background Art

[0002] With the acceleration of urbanization and the increase in extreme weather events, the operating pressure faced by urban sewage pipe networks continues to increase, and full pipe operation and overflow problems occur from time to time, posing a severe challenge to the stable operation and environmental protection of urban drainage systems. Existing sewage pipe network monitoring and early warning systems mostly use simple rule methods based on static thresholds. This method usually relies on fixed empirical values ​​to set early warning thresholds, and fails to fully consider the complexity and dynamic changes of the actual operating environment. As a result, when encountering sudden events or changes in operating conditions, false alarms and missed alarms are prone to occur, and full pipe risks cannot be warned in a timely and effective manner. In addition, traditional early warning methods do not take meteorological factors into consideration sufficiently, especially under extreme weather conditions (such as heavy rains and typhoons). It is difficult for the system to accurately assess the impact of meteorological changes on pipe network operation, thus affecting the accuracy and reliability of early warnings.

[0003] Abnormal detection in existing technologies mostly uses simple threshold judgment or rule matching methods. For complex and changeable sewage pipe network operation data, these methods are often unable to accurately identify potential abnormal operation states, especially for atypical events such as illegal discharge and pipe blockage. The detection effect is poor, which seriously affects the intelligence level and practicality of the early warning system. When dealing with multi-source data fusion, traditional methods lack effective algorithm support, and the correlation and complex interactive relationships between data cannot be fully utilized, resulting in large deviations in early warning analysis results. In addition, existing technologies are prone to overfitting problems in the multivariate modeling process, which makes the model perform poorly when facing new data, reducing the robustness and adaptability of the early warning system.

[0004] To this end, those skilled in the art provide a full pipe operation early warning analysis method for a sewage pipe network to solve the problems raised in the above background technology. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a full-pipe operation early warning analysis method for a sewage pipe network, which solves the problem of poor accuracy and adaptability of the existing early warning analysis method during use.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A full pipe operation early warning analysis method for a sewage pipe network comprises the following steps: S1. Install sensors at key nodes of the sewage pipe network to collect real-time operation data of the sewage pipe network, including flow, liquid level and pressure data; S2. Obtain real-time meteorological data, including rainfall, rainfall intensity and other relevant meteorological factors; S3. Clean and standardize the collected sewage pipe network data and meteorological data; S4. Based on multivariate regression analysis, a relationship model between flow rate, liquid level and full pipe risk is established; S5. Use long short-term memory network for time series prediction to predict future flow and level trends; S6. Dynamically adjust the full pipe warning threshold according to the time series prediction result and the adaptive algorithm; S7. Integrate real-time meteorological data and use deep learning models to establish a nonlinear model of the impact of meteorology on the sewage network; S8. Calculate the final warning parameters by combining the dynamically adjusted thresholds with the meteorological impact forecast results; S9. Fusion analysis of multi-source data is performed, and unsupervised learning algorithms are used to identify abnormal operating conditions in the sewage network.

[0007] Preferably, the relationship model established by the multivariate regression analysis in step S4 is:

[0008] in, is the full-pipeline risk index, is the flow rate at different monitoring points, is the liquid level at different monitoring points, is the pressure at different monitoring points, , , are model parameters, which are optimized by fitting the historical data under full pipe operation conditions.

[0009] Preferably, the specific model of the long short-term memory network for time series prediction in step S5 is:

[0010] in, It is the predicted flow rate or liquid level at the next moment, used for full pipe state prediction; It is the current hidden state, reflecting the current operation status of the pipeline network; is the current input, representing real-time monitoring data, , is the weight matrix, is the bias term, is the activation function.

[0011] Preferably, in step S6, the dynamic adjustment of the full pipe warning threshold adopts an adaptive algorithm, and the adjustment formula is:

[0012] in, It is the warning threshold after dynamic adjustment, reflecting the real-time response to the full pipe warning; is the basic threshold; It is the adjustment factor, and the amplitude is adjusted to match the changes in the operation status of the pipeline network; The difference between the predicted and current values ​​directly affects the full pipe risk.

[0013] Preferably, the deep learning model in step S7 adopts a hybrid model of a convolutional neural network and a long short-term memory network, which is specifically used to calculate the correction of the meteorological effect on the full pipe risk, and the expression is:

[0014]

[0015] in, is the meteorological influence correction factor, which is used to correct the full pipe warning threshold; , is the model weight, which reflects the impact of meteorological data on the load capacity of the sewage network.

[0016] Preferably, the calculation formula of the final warning parameter in step S8 is:

[0017] in, It is the final warning threshold, which is directly used to trigger the full pipe warning. To adjust the weights and control the impact of meteorological corrections on the thresholds; For the amplification factor, adjust the sensitivity of the threshold response.

[0018] Preferably, the model parameters of the multivariate regression analysis are By fitting the historical data of full-pipe operation, the least square method and regularization optimization are used to obtain the objective function:

[0019] in, is the number of samples, is the full pipe risk predicted by the model, It is a regularization parameter to control the complexity of the model and prevent overfitting.

[0020] Preferably, the identification of abnormal operation state in step S9 adopts the unsupervised learning algorithm Isolation Forest, and its abnormal detection scoring function is:

[0021] in, It is an abnormal score, and a lower score value indicates a possible full tube risk; is the average path length of the data points, reflecting the degree of abnormal operation of the sewage network.

[0022] Preferably, the loss function of the LSTM model is weighted mean square error, which is calculated as follows:

[0023] in, is the sample weight, which is reversely adjusted by the anomaly score S related to the full pipe risk to enhance the model's sensitivity to full pipe events.

[0024] Preferably, the warning threshold When the actual monitoring data is exceeded, the system automatically triggers a full pipe warning signal and initiates the corresponding emergency response plan to prevent sewage overflow and environmental pollution.

[0025] The present invention provides a method for early warning analysis of full pipe operation of a sewage pipe network, which has the following beneficial effects: 1. The present invention uses a long short-term memory network to perform time series prediction and combines an adaptive algorithm to dynamically adjust the warning threshold, so that the system can respond to changes in the operating status of the sewage pipe network in real time and accurately predict future flow and liquid level trends, thereby significantly improving the accuracy of the warning and the response speed to sudden full pipe risks.

[0026] 2. The present invention adopts a hybrid model of convolutional neural network and LSTM to incorporate the nonlinear impact of real-time meteorological data into the full pipe risk assessment, and realizes multi-source data fusion analysis and intelligent anomaly detection through the unsupervised learning algorithm isolation forest, which effectively identifies abnormal conditions in pipe network operation, and significantly improves the accuracy and reliability of the early warning system, especially under extreme weather conditions.

[0027] 3. The present invention introduces regularization technology to optimize the multivariate regression analysis model to prevent overfitting. Combined with the comprehensive analysis and real-time adjustment of multi-dimensional data, the system can automatically adapt to the complex and changeable operating environment to ensure the robustness and consistency of the early warning analysis results, thereby providing reliable technical support for the efficient management and safe operation of the urban drainage system. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a schematic diagram of the method flow of the present invention. DETAILED DESCRIPTION

[0029] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings 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.

[0030] Embodiment 1: Please see attached Figure 1 The embodiment of the present invention provides a method for early warning analysis of full pipe operation of a sewage pipe network, comprising the following steps: S1. Install sensors at key nodes of the sewage pipe network to collect real-time operation data of the sewage pipe network, including flow, liquid level and pressure data; Specifically, the sensor types include high-precision ultrasonic flowmeters, electromagnetic level gauges and pressure sensors, which are used to monitor the flow rate, water level and internal pressure of sewage in the pipe network. The location of each sensor should be determined through detailed system analysis and modeling to ensure the comprehensiveness of data coverage and the focus on key nodes. The sensor network is connected to the central control system through a wireless communication module to achieve remote transmission of real-time data. In order to improve the timeliness of data, the system adopts a high-frequency data acquisition mode, combined with advanced data compression and transmission technology to reduce the risk of transmission delay and data loss. During the data acquisition process, the system will eliminate invalid data caused by sensor failure or external interference through automated data cleaning and outlier filtering mechanisms to ensure data reliability. Real-time operation data will be automatically stored by the central control system and used as the basic data source for subsequent hydraulic model calculations, early warning analysis and emergency response. Through this data acquisition process, not only can the real-time operation status of the sewage pipe network be accurately grasped, but also scientific data support can be provided for the maintenance, optimization and emergency management of the pipe network, effectively improving the operation efficiency and safety level of the urban sewage treatment system. The implementation of this data collection will help to achieve comprehensive monitoring of the complex operating status of the sewage network, quickly respond to potential operating anomalies, and ultimately improve the level of refinement of urban water environment management.

[0031] S2. Obtain real-time meteorological data, including rainfall, rainfall intensity and other relevant meteorological factors; Specifically, these meteorological data include not only rainfall and rainfall intensity, but also key meteorological parameters such as wind speed, temperature, humidity, and atmospheric pressure that may affect the operation of the sewage pipe network. By accessing real-time data from local meteorological stations, remote sensing satellites, and high-precision radars, the system can obtain full coverage of meteorological information. In addition, the system adopts a real-time update mechanism for meteorological data based on an API interface to synchronize the acquired meteorological data with the monitoring data of the sewage pipe network to ensure the timeliness and consistency of the data. In terms of data processing, the system uses data fusion and correction technology to integrate and verify multi-source meteorological data, remove outliers, and perform necessary smoothing to reduce the impact of sensor errors and data noise on the analysis results. The acquisition and processing of real-time meteorological data provides the necessary boundary conditions for the hydraulic model of the sewage pipe network, so that the full pipe operation warning analysis can more accurately reflect the actual situation. Especially when extreme weather events such as heavy rain and typhoons occur, the system can adjust the warning parameters in time through the keen capture and rapid response of meteorological data, significantly improve the prediction ability of the full pipe risk of the sewage pipe network and the accuracy of the warning response, and provide a scientific basis for the emergency dispatch and operation and maintenance of the urban drainage system. Through the collection and application of this real-time meteorological data, the system can respond quickly to weather changes, thereby further improving the overall intelligence and reliability of sewage network early warning analysis.

[0032] S3. Clean and standardize the collected sewage pipe network data and meteorological data; Specifically, during the data cleaning process, the system first conducts a preliminary integrity check on the collected sewage network data (including flow, liquid level and pressure) and meteorological data (including rainfall, rainfall intensity, wind speed, etc.), eliminating missing, abnormal or obviously unreasonable data points caused by sensor failure, network transmission problems or other unexpected reasons. For detected outliers and noise data, the system uses statistical analysis methods, such as mean filtering, time series analysis and other technologies, to smooth or interpolate the data, thereby eliminating the impact of fluctuations in a short period of time, while ensuring the continuity and rationality of the data.

[0033] Standardization is to eliminate the dimensional differences between different data types and monitoring points, so that all input data can be compared and analyzed on a unified scale. The system uses standardization methods such as normalization or standard normal distribution transformation to convert various types of data into a relatively stable range (for example, between 0 and 1 or a distribution with a mean of 0 and a standard deviation of 1) to avoid a certain type of data from causing excessive bias in the analysis results. Through standardization, data such as flow, liquid level, pressure and meteorology can be processed under the same weight in subsequent model calculations and machine learning processes, thereby improving the robustness of the model and the efficiency of calculations.

[0034] After data cleaning and standardization, the system will verify the processed data to ensure that the cleaning and standardization steps have not introduced new anomalies or distortions. After these processing steps, the data set will be of high quality, uniformity and consistency, providing a solid data foundation for subsequent analysis modules such as multivariate regression analysis, time series prediction, dynamic threshold adjustment and anomaly detection. Ultimately, through rigorous cleaning and standardization, the system can significantly improve the accuracy and response speed of sewage pipe network full pipe operation warning, reduce the probability of false alarms and missed alarms, and provide reliable data support for the stable operation of urban sewage treatment and drainage systems.

[0035] S4. Based on multivariate regression analysis, a relationship model between flow rate, liquid level and full pipe risk is established; the relationship model established by multivariate regression analysis is:

[0036] in, is the full-pipeline risk index, is the flow rate at different monitoring points, is the liquid level at different monitoring points, is the pressure at different monitoring points, , , are the model parameters, which are optimized by fitting the historical data under full pipe operation conditions; Model parameters for multivariate regression analysis By fitting the historical data of full-pipe operation, the least square method and regularization optimization are used to obtain the objective function:

[0037] in, is the number of samples, is the full pipe risk predicted by the model, It is a regularization parameter to control the complexity of the model and prevent overfitting.

[0038] Specifically, first, the method uses the collected multi-dimensional data, including flow, liquid level and pressure at different monitoring points, as input variables, and sets the full pipe risk index as the output target. Through regression analysis, the method can determine the influence and relative importance of each input variable on the full pipe risk, thereby establishing a statistical relationship model reflecting the operation status of the pipeline network and the full pipe risk.

[0039] To ensure the accuracy and adaptability of the model, this method uses historical operating data for fitting when establishing the model, and applies the least squares method to minimize the error between the model prediction value and the actual full-pipe risk index. In addition, to prevent the model from overfitting when processing complex data, this method introduces regularization technology, which reduces the complexity of the model by constraining the model parameters, and improves the model's adaptability and stability to new data.

[0040] Through this multivariate regression analysis and parameter optimization method, an accurate full pipe risk prediction model can be established. This model can not only evaluate the current operating status of the sewage pipe network in real time, but also predict the possible full pipe risk in the future based on real-time data, providing a scientific basis for early warning.

[0041] S5. Use the long short-term memory network to perform time series prediction to predict future flow and liquid level trends. The specific model for long short-term memory network to perform time series prediction is:

[0042] in, It is the predicted flow rate or liquid level at the next moment, used for full pipe state prediction; It is the current hidden state, reflecting the current operation status of the pipeline network; is the current input, representing real-time monitoring data, , is the weight matrix, is the bias term, is the activation function; The loss function of the LSTM model is the weighted mean square error, which is calculated as:

[0043] in, is the sample weight, which is reversely adjusted by the anomaly score S related to the full pipe risk to enhance the model's sensitivity to full pipe events.

[0044] Specifically, this method constructs an LSTM model and makes full use of its advantages in processing time series data to capture the time dependence and complex nonlinear relationship of flow and liquid level changes in the sewage network. The LSTM model adopts a gating mechanism, which can memorize long-term historical information and effectively suppress the influence of irrelevant information, making it more accurate and stable when predicting flow and liquid level trends in the future.

[0045] In this method, the input of the LSTM model includes the flow rate, liquid level and other relevant real-time monitoring data at the current moment. These input data are processed by the embedding layer and the hidden layer of the network to output the predicted flow rate or liquid level value at the next moment. The hidden state used in the model can reflect the current operation status of the pipeline network, while the activation function (such as Sigmoid or Tanh function) is used to introduce nonlinear features, so that the model can capture the complex dynamic features in the data.

[0046] The LSTM model learns the relationship between input and output by continuously updating the weight matrix and bias terms. In order to improve the model's sensitivity to full pipe events, this method uses weighted mean square error (WMSE) as the loss function. In the design of the loss function, the error of each sample is assigned different weights, and these weights are inversely adjusted with the anomaly score associated with full pipe risk. By giving higher weights to high-risk samples, the model can focus more on data that may cause full pipe events during training, thereby enhancing the ability to predict full pipe status.

[0047] By using LSTM for time series prediction, this method can not only accurately predict the future flow and liquid level trends of the sewage network, provide reliable basic data for early warning of full pipe status, but also respond to dynamic changes in the sewage network in a timely manner.

[0048] S6. Dynamically adjust the full pipe warning threshold according to the time series prediction result and the adaptive algorithm. The dynamic adjustment of the full pipe warning threshold adopts the adaptive algorithm, and the adjustment formula is:

[0049] in, It is the warning threshold after dynamic adjustment, reflecting the real-time response to the full pipe warning; is the basic threshold; It is the adjustment factor, and the amplitude is adjusted to match the changes in the operation status of the pipeline network; The difference between the predicted and current values ​​directly affects the full pipe risk.

[0050] Specifically, the adaptive algorithm evaluates the change in the operating status of the pipeline network by comparing the difference between the predicted value and the current actual value, and adjusts the warning threshold according to the degree of change. Specifically, the adaptive algorithm will first set a basic warning threshold, which is determined based on historical data and experience to reflect the full pipe risk of the pipeline network under standard conditions. Then, based on the difference between the real-time predicted flow or liquid level and the current value, the algorithm calculates an adjustment factor that reflects the magnitude of the change in the operating status of the pipeline network. By introducing the adjustment factor into the threshold calculation, the adaptive algorithm can dynamically enlarge or reduce the warning threshold to make it closer to the current operation reality.

[0051] The dynamically adjusted warning threshold can not only respond to changes in the status of the pipeline network in real time, but also warn of potential full pipe risks. Through this adaptive dynamic adjustment method, the warning system can significantly reduce the incidence of false alarms and missed alarms, and improve its ability to respond to emergencies.

[0052] S7. Integrate real-time meteorological data and use a deep learning model to establish a nonlinear model of the impact of meteorology on the sewage pipe network. The deep learning model uses a hybrid model of a convolutional neural network and a long short-term memory network, which is specifically used to calculate the correction of meteorology to the full pipe risk. The expression is:

[0053]

[0054] in, is the meteorological influence correction factor, which is used to correct the full pipe warning threshold; , is the model weight, which reflects the impact of meteorological data on the load capacity of the sewage network.

[0055] Specifically, S8. Calculate the final warning parameters by combining the dynamically adjusted threshold and the meteorological impact forecast results. The calculation formula for the final warning parameters is:

[0056] in, It is the final warning threshold, which is directly used to trigger the full pipe warning. To adjust the weights and control the impact of meteorological corrections on the thresholds; is the amplification factor, which adjusts the sensitivity of the threshold response; the warning threshold When the actual monitoring data is exceeded, the system automatically triggers a full pipe warning signal and initiates the corresponding emergency response plan to prevent sewage overflow and environmental pollution.

[0057] Specifically, this method uses real-time meteorological data, such as rainfall, rainfall intensity, wind speed, temperature, etc., and combines these meteorological factors with the operating conditions of the sewage pipe network through a deep learning model to establish a nonlinear impact model of meteorology on full pipe risk. In order to more accurately reflect the dynamic impact of meteorological changes on the load capacity of the pipe network, the method adopts a hybrid model of convolutional neural network (CNN) and long short-term memory network (LSTM).

[0058] Convolutional neural networks (CNNs) are mainly used to extract spatial features of meteorological data and can identify complex patterns of sewage network loads under different meteorological conditions. Through multi-level convolution and pooling operations of the convolutional layer, CNN can effectively capture the local correlation and spatial distribution characteristics of meteorological data. These features are further input into LSTM, which is good at processing time series data and can learn the temporal dynamic changes of meteorological factors and their long-term impact on the status of the network. Therefore, the hybrid model of CNN and LSTM can give full play to their respective advantages and establish a comprehensive meteorological impact assessment model by jointly learning the spatiotemporal features of meteorological data.

[0059] In the model, the feature map extracted by CNN is processed by the activation function and passed as input to the LSTM network, which calculates the meteorological impact correction coefficient. This correction coefficient is used to dynamically adjust the full pipe warning threshold so that the threshold can reflect the actual impact of current meteorological conditions on the operation of the pipe network. The corrected warning threshold can respond more flexibly to extreme weather events, such as heavy rain or high temperature, and significantly improve the warning system's ability to identify these sudden risks.

[0060] The application of this deep learning model enables the early warning analysis method of the sewage pipe network to not only rely on static operation data, but also dynamically incorporate the influence of meteorological factors to form a comprehensive assessment of the risk of full pipes. Through nonlinear modeling and correction of meteorological data, the system can more accurately predict the risk of full pipes under different weather conditions, thereby providing more reliable decision support for urban sewage treatment and drainage management.

[0061] S9. Fusion analysis of multi-source data is performed, and unsupervised learning algorithms are used to identify abnormal operating conditions in the sewage pipe network. The identification of abnormal operating conditions uses the unsupervised learning algorithm Isolation Forest, and its anomaly detection scoring function is:

[0062] in, It is an abnormal score, and a lower score value indicates a possible full tube risk; is the average path length of the data points, reflecting the degree of abnormal operation of the sewage network.

[0063] Specifically, in order to identify abnormal operating conditions in the pipeline network, the method adopts the unsupervised learning algorithm Isolation Forest, which is specifically used to detect anomalies in data. It separates data points by constructing multiple random binary trees, and uses the difficulty of separating data points in the tree to measure anomalies.

[0064] The core of the Isolation Forest algorithm is to construct isolated trees by randomly selecting features and split points, and evaluate their abnormality by the path length required for data points to be separated in these trees. Specifically, the algorithm calculates the average path length of each data point in all isolated trees. A shorter path means that the data point is easier to be separated, that is, it may be an abnormal point. Conversely, if the average path length of a data point is longer, it means that it is more difficult to be isolated in the data set and may be a normal point.

[0065] The anomaly detection scoring function is used to quantify the degree of anomaly of each data point, where the lower the score, the more abnormal the data point is, indicating that there may be a risk of full pipe. The advantage of the isolation forest algorithm lies in its sensitivity to multidimensional data anomalies. It can effectively process complex and nonlinear sewage pipe network operation data and quickly identify abnormal operation modes. Through this method, potential problems in the sewage pipe network, such as pipe blockage, illegal discharge or other abnormal conditions, can be discovered early, so that preventive and response measures can be taken in time.

[0066] Embodiment 2: A specific implementation of early warning analysis according to Example 1: background: In the urban sewage pipe network system, flow meters, level meters and pressure sensors are installed at multiple key nodes to monitor the operation status of the pipe network. This paper will take the actual collected data as an example to demonstrate how the method of the present invention can be applied to the full pipe operation early warning analysis of the sewage pipe network.

[0067] Data collection Node 1 (backbone pipeline): flow h Liquid level m pressure bar Node 2 (branch pipeline): flow h Liquid level m pressure bar Weather data: Current rainfall mm Rainfall intensity mm / h Implementation steps 1. Multivariate regression analysis to establish a relationship model Based on historical data and the above real-time data, a full-pipeline risk model is established using multivariate regression analysis: 2. Timing prediction and dynamic adjustment The LSTM model is used to predict the flow rate and liquid level at the next moment, and the output is: Forecast traffic ; Based on the prediction results, an adaptive algorithm is applied to adjust the full pipe warning threshold: , .

[0068] 3. Integration and correction of meteorological influences A hybrid model of convolutional neural network and LSTM is used to calculate the meteorological impact correction coefficient , the output of the meteorological model is: .

[0069] 4. Final warning parameter calculation The final warning threshold is calculated by combining the dynamically adjusted threshold and the meteorological impact correction coefficient:

[0070]

[0071] 5. Anomaly detection and early warning The isolation forest algorithm is used for anomaly detection, and the anomaly score S\approx0.15S≈0.15 is obtained, which is much lower than the warning threshold (for example, set to 0.5). Therefore, the system triggers a full-pipe warning signal and notifies the operation and maintenance personnel to take emergency response through the warning platform.

[0072] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for early warning analysis of full pipe operation of a sewage pipe network, characterized in that: The following steps are involved: S1. Install sensors at key nodes of the sewage pipe network to collect real-time operation data of the sewage pipe network, including flow, liquid level and pressure data; S2. Obtain real-time meteorological data, including rainfall, rainfall intensity and other relevant meteorological factors; S3. Clean and standardize the collected sewage pipe network data and meteorological data; S4. Based on multivariate regression analysis, a relationship model between flow rate, liquid level and full pipe risk is established; S5, use long short-term memory network for time series prediction to predict future flow and level trends; S6. Dynamically adjust the full pipe warning threshold according to the time series prediction result and the adaptive algorithm; S7. Integrate real-time meteorological data and use deep learning models to establish a nonlinear model of the impact of meteorology on the sewage network; S8. Calculate the final warning parameters by combining the dynamically adjusted thresholds with the meteorological impact forecast results; S9. Fusion analysis of multi-source data is performed, and unsupervised learning algorithms are used to identify abnormal operating conditions in the sewage network.

2. A sewage pipe network full pipe operation early warning analysis method according to claim 1, characterized in that: The relationship model established by the multivariate regression analysis in step S4 is: in, is the full-pipeline risk index, is the flow rate at different monitoring points, is the liquid level at different monitoring points, is the pressure at different monitoring points, , , are model parameters, which are optimized by fitting the historical data under full pipe operation conditions.

3. The method for early warning analysis of full pipe operation of a sewage pipe network according to claim 1 is characterized in that: The specific model of the long short-term memory network for time series prediction in step S5 is: in, It is the predicted flow rate or liquid level at the next moment, used for full pipe state prediction; It is the current hidden state, reflecting the current operation status of the pipeline network; is the current input, representing real-time monitoring data, , is the weight matrix, is the bias term, is the activation function.

4. The method for early warning analysis of full pipe operation of a sewage pipe network according to claim 1, characterized in that: In step S6, the dynamic adjustment of the full pipe warning threshold adopts an adaptive algorithm, and the adjustment formula is: in, It is the warning threshold after dynamic adjustment, reflecting the real-time response to the full pipe warning; is the basic threshold; It is the adjustment factor, and the amplitude is adjusted to match the changes in the operation status of the pipeline network; The difference between the predicted and current values ​​directly affects the full pipe risk.

5. The method for early warning analysis of full pipe operation of a sewage pipe network according to claim 1, characterized in that: The deep learning model in step S7 adopts a hybrid model of convolutional neural network and long short-term memory network, which is specifically used to calculate the correction of meteorological conditions to full pipe risk, and the expression is: in, is the meteorological influence correction factor, which is used to correct the full pipe warning threshold; , is the model weight, which reflects the impact of meteorological data on the load capacity of the sewage network.

6. The method for early warning analysis of full pipe operation of a sewage pipe network according to claim 1, characterized in that: The calculation formula of the final warning parameter in step S8 is: in, It is the final warning threshold, which is directly used to trigger the full pipe warning. To adjust the weights and control the impact of meteorological corrections on the thresholds; For the amplification factor, adjust the sensitivity of the threshold response.

7. A sewage pipe network full pipe operation early warning analysis method according to claim 2, characterized in that: The model parameters of multivariate regression analysis By fitting the historical data of full-pipe operation, the least square method and regularization optimization are used to obtain the objective function: in, is the number of samples, is the full pipe risk predicted by the model, It is a regularization parameter to control the complexity of the model and prevent overfitting.

8. The method for early warning analysis of full pipe operation of a sewage pipe network according to claim 1, characterized in that: The identification of abnormal operation status in step S9 adopts the unsupervised learning algorithm Isolation Forest, and its abnormal detection scoring function is: in, It is an abnormal score, and a lower score value indicates a possible full tube risk; is the average path length of the data points, reflecting the degree of abnormal operation of the sewage network.

9. A sewage pipe network full pipe operation early warning analysis method according to claim 3, characterized in that: The loss function of the LSTM model is the weighted mean square error, which is calculated as follows: in, is the sample weight, which is reversely adjusted by the anomaly score S related to the full pipe risk to enhance the model's sensitivity to full pipe events.

10. A sewage pipe network full pipe operation early warning analysis method according to claim 6, characterized in that: The warning threshold When the actual monitoring data is exceeded, the system automatically triggers a full pipe warning signal and initiates the corresponding emergency response plan to prevent sewage overflow and environmental pollution.

Images