Drainage pipe network intelligent operation and maintenance method and system based on big data analysis
The drainage network system uses big data analysis and One-Class SVM models to improve anomaly detection accuracy, addressing inefficiencies in manual and sensor-based methods by identifying and preventing network failures.
Patent Information
- Application Number
- CN202411206591.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-12-20
AI Technical Summary
Existing drainage network anomaly detection methods, relying on manual inspections and sensor monitoring, are inefficient, costly, and lack accuracy in harsh environments, failing to timely detect issues such as siltation, excessive pollution, and pipe gaps.
A drainage network intelligent operation and maintenance system using big data analysis, employing One-Class SVM models trained with historical data to analyze water quality indicators and external water inflow ratios, combined with chemical mass balance equations, to identify anomalies like siltation, pollution, and pipe gaps.
Enhances anomaly detection accuracy, enabling timely identification and prevention of drainage network failures, reducing maintenance costs and risks from waterlogging incidents.
Smart Images
Figure CN119170129A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of drainage pipe operation and maintenance, and specifically relates to a drainage pipe network intelligent operation and maintenance method and system based on big data analysis. Background Art
[0002] Urban municipal drainage networks are an important part of urban infrastructure, and their safe and stable operation is directly related to the quality of life of urban residents and the safety of the urban environment. In recent years, with the continuous expansion of urban scale and the continuous increase in population density, drainage networks are facing increasingly severe challenges, such as aging of networks, increased sewage discharge, urban waterlogging and other problems, resulting in frequent abnormalities in drainage networks, posing a major threat to urban safety and the environment. Existing drainage network anomaly detection methods mainly rely on manual and sensor monitoring systems, which have the following shortcomings:
[0003] 1. Manual inspections are inefficient, have limited coverage, and are difficult to detect in a timely manner. Manual inspections and regular testing require a lot of manpower and material resources, and are costly.
[0004] 2. Flow measurement equipment has poor measurement accuracy and is not suitable for the harsh environment of drainage pipes. In the closed, highly corrosive, and low-visibility drainage pipe environment, the current electronic sensor equipment cannot accurately and reliably carry out dynamic monitoring of flow rate and flow and collect data for real-time upload. Summary of the invention
[0005] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a drainage network intelligent operation and maintenance method and system based on big data analysis.
[0006] The purpose of the present invention can be achieved through the following technical solutions:
[0007] A drainage network intelligent operation and maintenance method and system based on big data analysis, comprising the following steps:
[0008] S1) setting the external water inflow position of each section of the drainage network, marked as #1, #2, #3...#n; the drainage network outflow position is marked as #0; and recording the external water type corresponding to each position;
[0009] Preferably, the external water includes surface water, domestic sewage, and industrial sewage;
[0010] S2) Collect the water quality index data of each section of the drainage network under normal operation under different data factor conditions in history to form a historical index data set [C'];
[0011] Preferably, the water quality indicators include one or more combinations of conductivity, phosphate, ammonia nitrogen, total phosphorus or total nitrogen.
[0012] Preferably, the data factors include season, rainfall and maintenance cycle;
[0013] S3) detecting the water quality index data of each section of the current drainage network to form a current index data set [C];
[0014] S4) comparing and analyzing the historical indicator data set [C'] and the current indicator data set [C], and then determining whether the current indicator data set is abnormal;
[0015] Preferably, the specific process of the comparative analysis is as follows:
[0016] Data preprocessing: Collect historical indicator data sets [C'] and current indicator data sets [C]; use StandardScaler to standardize the data so that the mean of each feature is 0 and the variance is 1;
[0017] Model training: Use the historical indicator dataset [C'] to train the One-Class SVM model; select the RBF kernel function and define the gamma and nu parameters;
[0018] Anomaly detection: Use the trained model to detect the current indicator data set [C]. The output of One-Class SVM is 1 for normal and -1 for abnormal;
[0019] S5) Select and detect various water quality indicators of each section of the drainage network; according to the chemical mass balance model (CMBM), that is, by detecting the pollutant concentration of the water quality indicators before and after the drainage network is inflowed by external water, establish the mass conservation equation of pollutants flowing into and out of the drainage network, which is used to quantitatively calculate the proportion of external water inflow into the drainage network; the relationship is:
[0020]
[0021] Where: i is the water quality index i concentration at the outflow of the drainage network; C ij is the concentration of water quality index i of external water j in the drainage network; X j is the proportion of external water j inflow in the drainage network; n is the number of external water types in the drainage network;
[0022] S6) Collect the external water inflow ratio of each section of the drainage network under normal operation under different data factor conditions in history to form a historical ratio data set [X'];
[0023] S7) detecting water quality index data of each current section of the drainage network, and then quantitatively calculating the proportion of external water inflow into each current section of the drainage network; forming a current proportion data set [X];
[0024] S8) comparing and analyzing the historical proportion data set [X'] with the current proportion data set [X], and then determining whether the current proportion data set is abnormal;
[0025] The specific process of the comparative analysis is as follows:
[0026] Data preprocessing: Collect historical indicator data sets [X'] and current indicator data sets [X]; use StandardScaler to standardize the data so that the mean of each feature is 0 and the variance is 1;
[0027] Model training: Use the historical normal data set [X'] to train the One-Class SVM model; select the RBF kernel function (kernel = 'rbf'), and define the gamma parameter and nu parameter;
[0028] Anomaly detection: Use the trained model to detect the current test data set [X]. The output of One-Class SVM is 1 for normal and -1 for abnormal;
[0029] S9) on the basis of judging that the current indicator data set [C] is abnormal in step S4), comparing the value of the current indicator data set [C] with that of the historical indicator data set to see whether it is increased, if the value is increased, it is determined that siltation occurs at the position mark or drainage pollution exceeds the standard;
[0030] At the same time, referring to steps S5-S8), the current proportion data set upstream of the position marker is calculated, and the One-Class SVM model is used for detection. If the current proportion data set upstream of the position marker is detected as abnormal, it is determined that blockage occurs at the position marker;
[0031] If normal, it is determined that drainage pollution exceeds the standard at the location marked.
[0032] S10) on the basis of determining in step S4) that the current indicator data set [C] is abnormal, comparing the value of the current indicator data set [C] with that of the historical indicator data set to see whether the value has decreased;
[0033] If the value decreases, at the same time, refer to steps S5-S8), calculate the current proportional data set downstream of the position marker, and use the One-Class SVM model for detection. If the current proportional data set downstream of the position marker is detected as abnormal, it is determined that a pipeline gap-type drainage leakage occurs at the position marker.
[0034] Beneficial effects of the present invention:
[0035] 1. By monitoring multiple water quality indicators, such as conductivity, phosphate, ammonia nitrogen, total phosphorus and total nitrogen, and calculating the proportion of external water inflow in combination with the chemical mass balance equation, the operating status of the drainage network can be more comprehensively reflected and the accuracy of anomaly detection can be improved.
[0036] 2. Use historical data to train the One-Class SVM model, learn the water quality indicators and external water inflow ratio data sets under normal operating conditions, and establish a model to identify abnormal data. According to the specific changes in water quality indicators and external water inflow ratios, it can timely identify abnormal conditions in the drainage network, such as siltation, excessive drainage pollution, and network gaps, and take timely measures to avoid municipal waterlogging.
[0037] 3. Use big data analysis and algorithms to identify abnormal situations early and arrange maintenance work in a timely manner, avoid greater losses caused by sudden disasters such as heavy rains, reduce pipeline maintenance costs, and provide guarantees for the safe operation of drainage pipelines. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.
[0039] Figure 1 This is a structural schematic diagram of an intelligent operation and maintenance method for a drainage network based on big data analysis and a marker for the inflow position of external water in the system of the present invention.
[0040] Figure 2 This is a flow chart of a method for intelligent operation and maintenance of a drainage network based on big data analysis according to the present invention.
[0041] Figure 3 This is a flow chart of the construction of a One-Class SVM model in a smart operation and maintenance method for a drainage network based on big data analysis in the present invention. DETAILED DESCRIPTION
[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments 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.
[0043] See also Figure 1-Figure 3 As shown, a smart operation and maintenance method for a drainage network based on big data analysis includes the following steps:
[0044] S1) Set the external water inflow position of each section of the drainage network, marked as #1, #2, #3...#n; the outflow position of the drainage network is recorded as #0; and record the external water type corresponding to each position; provide basic information for the chemical mass balance model.
[0045] Optionally, the external water includes surface water, domestic sewage, and industrial sewage.
[0046] S2) Collect the water quality index data of each section of the drainage network under normal operation under different data factor conditions in the past to form a historical index data set [C'];
[0047] S3) detecting the water quality index data of each section of the current drainage network to form a current index data set [C];
[0048] S4) comparing and analyzing the historical indicator data set [C'] and the current indicator data set [C], and then determining whether the current indicator data set is abnormal;
[0049] Optionally, the water quality index includes one or more combinations of conductivity, phosphate, ammonia nitrogen, total phosphorus or total nitrogen. The basis for this is that many researchers at home and abroad have clearly pointed out in their reports that there are significant differences in water quality indicators between different water sources. In addition, some domestic scholars have also used the conductivity index to conduct diagnostic research on the problem of external water infiltration in the drainage system. The present invention selects different combinations of water quality indicators, and the calculation results are relatively accurate and within a reasonable range, which can be used as reference data for qualitative diagnosis of drainage networks.
[0050] Specifically, sensors or sampling and analysis methods can be used for detection. Sensors can monitor the conductivity of water quality in real time; while the concentration data of phosphate and ammonia nitrogen require regular collection of water samples for laboratory testing. Among them, TN is determined by potassium persulfate oxidation-ultraviolet spectrophotometry; NH3-N is determined by Nessler's reagent spectrophotometry; TP and PO4 3- -P was determined by molybdenum antimony counterspectrophotometry.
[0051] Optionally, the data factors include season, rainfall and maintenance cycle; the season can be divided into four seasons: spring, summer, autumn and winter; or divided into evenly spaced time in the 24 solar terms; and converted into numerical values using one-hot coding. As follows:
[0052] Spring:
[1000]
[0053] Summer:
[0100]
[0054] Autumn:
[0010]
[0055] Winter:
[0001] .
[0056] The rainfall is determined based on the information released by the local weather station and is usually measured in millimeters.
[0057] The maintenance cycle is defined as the time interval from the last maintenance time point to the current time point at the drainage network location, and is measured in days.
[0058] In this embodiment, the specific process of the comparative analysis is as follows:
[0059] Data preprocessing: Collect historical indicator data sets [C'] and current indicator data sets [C]; use StandardScaler to standardize the data so that the mean of each feature is 0 and the variance is 1;
[0060] Model training: Use the historical indicator dataset [C'] to train the One-Class SVM model; select the RBF kernel function and define the gamma and nu parameters;
[0061] Anomaly detection: Use the trained model to detect the current indicator data set [C]. The output of One-Class SVM is 1 for normal and -1 for abnormal;
[0062] As a preferred solution of this embodiment, the conductivity is selected as the water quality indicator in the historical indicator data set [C'] and the current indicator data set [C]; the conductivity of each section of the drainage network under normal operation of different data factors is compared and analyzed,
[0063] For example:
[0064] Historical indicator dataset [C']:
[0065] At #1: [C1(season = spring, rainfall = 13, maintenance cycle = 30, conductivity = 750), C2, ...];
[0066] Among them, all samples of the historical indicator dataset should be normal samples, that is, samples without failure. In addition, for categorical features (such as season), one-hot encoding can be used to convert them into numerical values.
[0067] Current indicator dataset [C]:
[0068] At #1: [C1(Beginning of Spring, rainfall, maintenance cycle, conductivity), C2], where the current indicator data set [C] contains at least two samples.
[0069] It should be noted that One-ClassSVM (Single Class Support Vector Machine) is a supervised learning algorithm for anomaly detection. Its main goal is to find anomalies or rare samples in a data set without requiring a large number of normal samples for training. This makes One-ClassSVM effective in dealing with high-dimensional data and very sparse anomaly detection problems.
[0070] The basic principle of One-Class SVM is to find a hyperplane that separates normal samples from abnormal samples. This hyperplane is called the "separation hyperplane" and is supported by support vectors, that is, the normal sample points closest to the separation hyperplane. The specific model construction steps can refer to the existing technology, usually implemented using Python or the machine learning library Scikit-Learn.
[0071] In model training, select the RBF kernel function; set the gamma parameter and nu parameter at the same time; for example, gamma = 'auto'; nu = 0.1-0.5 (control the error rate, that is, how many normal data points are allowed to be identified as abnormal by the model. The smaller the value, the stricter the model).
[0072] Specifically, during the implementation of the vehicle, the model training should be performed on the different marked positions of the drainage network (#1, #2#3...#n) to improve the adaptability of the model to data at different positions. For example, the One-Class SVM model at position #1, position #2, etc. can be trained separately to more accurately identify anomalies at different positions.
[0073] S5) Select and detect various water quality indicators of each section of the drainage network; according to the chemical mass balance model (CMBM), that is, by detecting the pollutant concentration of the water quality indicators before and after the drainage network is inflowed by external water, establish the mass conservation equation of pollutants flowing into and out of the drainage network, which is used to quantitatively calculate the proportion of external water inflow into the drainage network; the relationship is:
[0074]
[0075] Where: i is the water quality index i concentration at the outflow of the drainage network; C ij is the concentration of water quality index i of external water j in the drainage network; X j is the proportion of external water j inflow in the drainage network; n is the number of external water types in the drainage network;
[0076] like Figure 1 As shown, PO4 3- Taking the combination of -P, NH3-N, and TP water quality indicators as an example, the following mass balance equations are established based on the water quality and water quantity at the location marks (#1, #2, #3, and #0) of the drainage network:
[0077]
[0078] S6) Collect and organize historical Chinese formulas: PO , NH , TP They are PO4 at the outflow of the pipeline 3- -P, NH3-N and TP concentrations, mg / L;
[0079] C #1，PO The PO4 values at the drainage network inflow locations (#1, #2, #3) are 3- -P, NH3-N and TP concentrations, mg / L;
[0080] X #1 etc. are the proportion of inflow water at the drainage network location mark, %;
[0081] The concentration values of PO43--P, NH3-N, and TP at the inflow position marks (#1, #2, and #3) of the drainage network and the corresponding concentration value at the outflow position mark (#0) of the drainage network are measured by water quality monitoring instruments. Substituting them into the formula, we can solve X #1 , X #2 , X #3 The proportion of inflow water.
[0082] S6) Collect the external water inflow ratio of each section of the drainage network under normal operation under different data factor conditions in history to form a historical ratio data set [X'];
[0083] S7) detecting water quality index data of each current section of the drainage network, and then quantitatively calculating the proportion of external water inflow into each current section of the drainage network; forming a current proportion data set [X];
[0084] S8) comparing and analyzing the historical proportion data set [X'] with the current proportion data set [X], and then determining whether the current proportion data set is abnormal;
[0085] The specific process of the comparative analysis is as follows:
[0086] Data preprocessing: Collect historical scale data sets [X'] and current scale data sets [X]; use StandardScaler to standardize the data so that the mean of each feature is 0 and the variance is 1;
[0087] Model training: Use the historical proportion data set [X'] to train the One-Class SVM model; select the RBF kernel function (kernel = 'rbf'), and define the gamma parameter and nu parameter;
[0088] Anomaly detection: Use the trained model to detect the current scale data set [X]. The output of One-Class SVM is 1 for normal and -1 for abnormal;
[0089] S9) on the basis of judging that the current indicator data set [C] is abnormal in step S4), comparing the value of the current indicator data set [C] with that of the historical indicator data set to see whether it is increased, if the value is increased, it is determined that siltation occurs at the position mark or drainage pollution exceeds the standard;
[0090] At the same time, referring to steps S5-S8), the current proportion data set upstream of the position marker is calculated, and the One-Class SVM model is used for detection. If the current proportion data set upstream of the position marker is detected as abnormal, it is determined that blockage occurs at the position marker;
[0091] If normal, it is determined that drainage pollution exceeds the standard at the location marked.
[0092] S10) on the basis of determining in step S4) that the current indicator data set [C] is abnormal, comparing the value of the current indicator data set [C] with that of the historical indicator data set to see whether the value has decreased;
[0093] If the value decreases, at the same time, refer to steps S5-S8), calculate the current proportional data set downstream of the position marker, and use the One-Class SVM model for detection. If the current proportional data set downstream of the position marker is detected as abnormal, it is determined that a pipeline gap-type drainage leakage occurs at the position marker.
[0094] The above steps S9)-S10) are the judgment logic of the present invention. It is worth noting that the abnormal situation that occurs is comprehensively judged by using the changes in water quality indicators at different position marks of the drainage network and the changes in the proportion of external water inflow;
[0095] Wherein, the process of blockage determination in step S9) is as follows:
[0096] Changes in water quality indicators: When siltation occurs at a certain location mark in the drainage network, the concentration of water quality indicators upstream will gradually increase, reflecting the accumulation of pollutants and deterioration of water quality.
[0097] Changes in the proportion of external water inflow: Due to poor drainage caused by siltation, the proportion of external water inflow from upstream will be abnormal. This may be because the external water cannot flow into the downstream normally, resulting in an increase in the amount of water accumulation in the upstream area.
[0098] Therefore, when the concentration of the water quality index in the drainage network increases and the proportion of external water inflow upstream of the position mark becomes abnormal, it can be determined that blockage occurs at the position mark.
[0099] Step S10) The process of determining the pipe network gap type drainage leakage is as follows:
[0100] Changes in water quality indicators: Due to leakage from the pipe network gap, the water flow at the location mark will decrease, resulting in a decrease in the water quality indicator value.
[0101] Changes in the proportion of external water inflow: Due to leakage in the pipe network, water flows to the inlet at the surrounding position mark, which will cause an abnormality in the proportion of external water inflow at the nearest position mark downstream.
[0102] Therefore, by observing the comprehensive changes in the water quality index values and the proportion of external water inflow, it can be determined whether a gap leakage occurs at the position mark.
[0103] On the basis of the intelligent operation and maintenance method of the drainage network based on big data analysis, the present application also provides a system for the intelligent operation and maintenance method of the drainage network based on big data analysis, which system includes: a memory, a processor and a communication module.
[0104] The memory may be configured to store various data to support operations on the system, and to store one or more program instructions. Including any application program operating on the system and the instructions of steps 1-9) of the above-mentioned intelligent operation and maintenance method of the drainage network based on big data analysis. The memory may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, disk or optical disk.
[0105] The processor is coupled to the memory and is used to execute one or more program instructions in the memory. Specifically, it is used to execute steps S1-S8 of the above-mentioned intelligent operation and maintenance method of the drainage network based on big data analysis).
[0106] Further, the processor is used in the data analysis process for building a One-Class SVM model based on Python or the machine learning library Scikit-Learn, including:
[0107] Data preprocessing: Collect historical scale data sets [X'] and current scale data sets [X]; use StandardScaler to standardize the data so that the mean of each feature is 0 and the variance is 1;
[0108] Model training: Use the historical proportion data set [X'] to train the One-Class SVM model; select the RBF kernel function (kernel = 'rbf'), and define the gamma parameter and nu parameter;
[0109] Anomaly detection: Use the trained model to detect the current scale data set [X]. The output of One-Class SVM is 1 for normal and -1 for abnormal;
[0110] Furthermore, after determining that the data at the current position mark is abnormal, the processor is also used to: implement a comprehensive determination of the abnormal type of the drainage network according to the model output result and the above steps S9-S10).
[0111] The communication module is configured to be used for wired or wireless communication connection between the device in the system and other devices. The device in which the communication module is located can access a wireless network based on a communication standard, such as WiFi, 2G, 3G, 4G or 5G, or a combination thereof. In an exemplary embodiment, the communication module can be implemented based on near field communication technology (NFC), radio frequency identification technology (RFID), infrared technology (IrDA), ultra-wideband technology (UWB), etc.
[0112] In summary, the present invention provides a smart operation and maintenance method for drainage network based on big data analysis, which realizes early identification and early warning of abnormal conditions in the drainage network by monitoring multiple water quality indicators, combining the chemical mass balance equation to calculate the proportion of exogenous water inflow, and using the One-Class SVM model to learn normal operating status data.
[0113] The method and system of the present invention can more comprehensively reflect the operating status of the drainage network, improve the accuracy of abnormality detection, timely identify abnormal situations such as siltation, excessive drainage pollution, and network gaps, and take corresponding measures in a timely manner according to the abnormal situations, effectively avoiding greater losses caused by municipal waterlogging incidents and sudden rainstorms, reducing network maintenance costs, and providing guarantees for the safe operation of the drainage network.
[0114] In the specific embodiments provided in the present application, it should be understood that the terms "include", "comprising" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of further restrictions, the elements defined by the sentence "comprising ... one or more combinations" do not exclude the presence of other identical elements in the process, method, commodity or device including the elements.
[0115] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.
[0116] In the description of the specification, the description with reference to the terms "preferred embodiment", "optionally", "particularly", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.
[0117] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.
Claims
1. A smart operation and maintenance method for a drainage network based on big data analysis, characterized in that: The following steps are involved: S1) setting the external water inflow position of each section of the drainage network, marked as #1, #2, #3...#n; the drainage network outflow position is marked as #0; and recording the external water type corresponding to each position; S2) Collect the water quality index data of each section of the drainage network under normal operation under different data factor conditions in history to form a historical index data set [C']; S3) detecting the water quality index data of each section of the current drainage network to form a current index data set [C]; S4) comparing and analyzing the historical indicator data set [C'] and the current indicator data set [C], and then determining whether the current indicator data set is abnormal; S5) Select and detect various water quality indicators of each section of the drainage network; according to the chemical mass balance model, that is, by detecting the pollutant concentration of the water quality indicators before and after the drainage network is inflowed by external water, establish the mass conservation equation of pollutants flowing into and out of the drainage network, which is used to quantitatively calculate the proportion of external water inflow into the drainage network; the relationship is: Where: i is the water quality index i concentration at the outflow of the drainage network; C ij is the concentration of water quality index i of external water j in the drainage network; X j is the proportion of external water j inflow in the drainage network; n is the number of external water types in the drainage network; S6) Collect the external water inflow ratio of each section of the drainage network under normal operation under different data factor conditions in history to form a historical ratio data set [X']; S7) detecting water quality index data of each current section of the drainage network, and then quantitatively calculating the proportion of external water inflow into each current section of the drainage network; Form the current scale data set [X]; S8) Compare and analyze the historical proportion data set [X'] with the current proportion data set [X] to determine whether the current proportion data set is abnormal.
2. According to claim 1, a method for intelligent operation and maintenance of drainage pipe network based on big data analysis is characterized in that: The external water includes surface water, domestic sewage and industrial sewage.
3. According to the method of intelligent operation and maintenance of drainage network based on big data analysis according to claim 1, it is characterized by: The water quality indicators include one or more combinations of conductivity, phosphate, ammonia nitrogen, total phosphorus or total nitrogen.
4. The intelligent operation and maintenance method of the drainage network based on big data analysis according to claim 1 is characterized by: The data factors include season, rainfall and maintenance cycle.
5. The intelligent operation and maintenance method of the drainage network based on big data analysis according to claim 1 is characterized by: The specific process of comparative analysis in step S4) is as follows: Data preprocessing: Collect historical indicator data sets [C'] and current indicator data sets [C]; use StandardScaler to standardize the data so that the mean of each feature is 0 and the variance is 1; Model training: Use the historical indicator dataset [C'] to train the One-Class SVM model; select the RBF kernel function and define the gamma and nu parameters; Anomaly detection: Use the trained model to detect the current indicator dataset [C]. The output of One-Class SVM is 1 for normal and -1 for abnormal.
6. The intelligent operation and maintenance method of the drainage network based on big data analysis according to claim 5 is characterized by: The conductivity is selected as the water quality indicator in the historical indicator data set [C'] and the current indicator data set [C].
7. The intelligent operation and maintenance method of a drainage network based on big data analysis according to claim 1 is characterized in that: The specific process of comparative analysis in step S8) is as follows: Data preprocessing: Collect historical indicator data sets [X'] and current indicator data sets [X]; use StandardScaler to standardize the data so that the mean of each feature is 0 and the variance is 1; Model training: Use the historical normal data set [X'] to train the One-Class SVM model; select the RBF kernel function and define the gamma parameter and nu parameter; Anomaly detection: Use the trained model to detect the current test data set [X]. The output of One-Class SVM is 1 for normal and -1 for abnormal.
8. The intelligent operation and maintenance method of drainage network based on big data analysis according to claim 5 is characterized in that: Also includes step S9): In step S4), on the basis of judging that the current indicator data set [C] is abnormal, comparing the value of the current indicator data set [C] with that of the historical indicator data set to see whether it is increased. If the value is increased, it is judged that siltation occurs at the position mark or drainage pollution exceeds the standard; At the same time, referring to steps S5-S8), the current proportion data set upstream of the position marker is calculated, and the One-ClassSVM model is used for detection. If the current proportion data set upstream of the position marker is detected as abnormal, it is determined that blockage occurs at the position marker; If normal, it is determined that drainage pollution exceeds the standard at the location marked.
9. The intelligent operation and maintenance method of drainage network based on big data analysis according to claim 5 is characterized in that: Also includes step S10): In step S4), on the basis of determining that the current indicator data set [C] is abnormal, comparing the value of the current indicator data set [C] with that of the historical indicator data set to see whether it decreases; If the value decreases, at the same time, refer to steps S5-S8), calculate the current proportional data set downstream of the position marker, and use the One-Class SVM model for detection. If the current proportional data set downstream of the position marker is detected as abnormal, it is determined that a pipeline gap-type drainage leakage occurs at the position marker.
10. A drainage network intelligent operation and maintenance system based on big data analysis, used to execute the steps in the method according to any one of claims 1 to 9, characterized in that: include: Memory, processor and communication modules; The memory is used to store one or more program instructions; The processor is used to execute one or more program instructions; The communication module is used for wired or wireless communication connection between various devices in the system.