Urban drainage pipe network damage detection system based on intelligent analysis

Abstract

The invention relates to the technical field of urban infrastructure intelligent detection, and discloses an urban drainage pipe network damage detection system based on intelligent analysis. A multi-source data acquisition module acquires pipe network static structure parameters and real-time operation monitoring data through a distributed sensor network; generating a topological characteristic value, a dynamic operation state characteristic value and a dynamic detection threshold set of each damage type; the data preprocessing module filters noise of the monitoring data, eliminates abnormal values and extracts time domain and frequency domain feature vectors; the feature fusion module fuses and generates a multi-dimensional fusion feature matrix based on the topology and operation feature values; the intelligent analysis engine calculates and outputs a damage type diagnosis result through mode matching; the spatial positioning module generates three-dimensional coordinate positioning data of the damaged area in combination with the topological characteristic value; and the dynamic learning module optimizes the detection rule according to the maintenance data. The system realizes intelligent damage detection and accurate positioning, improves the detection accuracy and efficiency, adapts to a complex environment, and is high in intelligent level.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent detection of urban infrastructure, and specifically to a detection system for damaged urban drainage pipe networks based on intelligent analysis. Background Technique

[0002] As an important part of urban infrastructure, the operation status of urban drainage pipe networks is directly related to urban flood control and drainage, water resource utilization, and ecological environment protection. With the acceleration of the urbanization process, the scale of drainage pipe networks is constantly expanding, and problems such as pipe network aging and damage are becoming increasingly prominent. Traditional manual inspection and offline detection methods have disadvantages such as low efficiency, high cost, and poor accuracy, and are difficult to meet the needs of modern urban refined management.

[0003] Currently, urban drainage pipe network detection faces many challenges. On the one hand, the pipe network structure is complex, including a large number of pipe segments with different diameters, pipe materials, and laying years. Traditional detection methods are difficult to comprehensively and accurately obtain the static structure parameters and dynamic operation data of the pipe network. On the other hand, the pipe network operation environment is complex and changeable, and environmental factors such as groundwater level, soil moisture content, and temperature will have a significant impact on the detection results. Traditional fixed-threshold detection methods cannot adapt to the detection needs under different working conditions. In addition, existing detection technologies often lack the ability to fuse and analyze multi-source data, and it is difficult to extract effective damage features from a large amount of monitoring data, resulting in low accuracy and reliability of the detection results.

[0004] With the rapid development of sensor technology, Internet of Things technology, and intelligent analysis technology, new ideas and methods have been provided for the detection of damaged urban drainage pipe networks. By deploying a distributed sensor network in the drainage pipe network, multi-source data such as pressure, flow rate, and acoustic fingerprint of the pipe network can be collected in real time; using intelligent analysis algorithms to fuse and process multi-source data and perform pattern matching can achieve accurate identification and spatial positioning of the damaged types of the pipe network. However, existing technologies still have deficiencies in aspects such as multi-source data collection and preprocessing, feature fusion and intelligent analysis, dynamic threshold adjustment and spatial positioning, and need to be further optimized and improved. Summary of the Invention

[0005] The purpose of the present invention is to provide a detection system for damaged urban drainage pipe networks based on intelligent analysis to solve the problems raised in the above background technique.

[0006] To achieve the above purpose, the present invention provides the following technical solution: A detection system for damaged urban drainage pipe networks based on intelligent analysis, the system includes: The multi-source data acquisition module is used to collect the static structure parameters and real-time operation monitoring data of the drainage pipe network through a distributed sensor network, analyze and process them to generate the topological characteristic values of the pipe network and the dynamic operation state characteristic values, and analyze and process them based on the standard parameter sets of various damage types in the historical maintenance database to generate the dynamic detection threshold sets of various damage types; The data preprocessing module is used to receive the real-time operation monitoring data, perform noise filtering and outlier removal, screen out the effective monitoring data, and extract the time-domain feature vectors and frequency-domain feature vectors of the effective monitoring data; The feature fusion module is based on the topological characteristic values of the pipe network and the dynamic operation state characteristic values, and performs fusion processing on the time-domain feature vectors and frequency-domain feature vectors to generate a multi-dimensional fusion feature matrix; The intelligent analysis engine is used to perform pattern matching calculations on the multi-dimensional fusion feature matrix and the dynamic detection threshold sets of various damage types, output the matching degree values of various damage types, and screen out the damage type corresponding to the maximum matching degree value as the current pipe network state diagnosis result; The space positioning module is used to index the spatial distribution patterns of the corresponding damage types in the historical maintenance database according to the diagnosis results, and calculate and generate the three-dimensional coordinate positioning data of the damaged area in combination with the topological characteristic values of the pipe network.

[0007] Preferably, the process of collecting the static structure parameters and real-time operation monitoring data of the drainage pipe network is as follows: When the detection equipment of the pipe network system is initially deployed, collect the structure parameters such as pipe diameter size, pipe material type, and laying age, establish a pipe network topological relationship map based on the structure parameters, and start the pressure sensor, flow meter, and acoustic detector for baseline data collection; When the pipe network system has historical detection records, extract the operation pressure fluctuation data, flow change data, and acoustic fingerprint feature data of the last three maintenance cycles from the historical maintenance database, and record them as the current real-time monitoring reference data set.

[0008] Preferably, the generation of the dynamic detection threshold sets of various damage types specifically includes: Extract the standard parameter sets of various damage types from the historical maintenance database, including the standard pressure gradient value of the leakage state, the standard acoustic frequency spectrum features of the pipe body cracks, and the standard amplitude of the flow rate mutation under the blockage condition; Obtain the current environmental variable data, including the groundwater level height, soil moisture content, and temperature change parameters, and calculate and generate an environmental correction coefficient; Multiply each parameter item in the standard parameter set by the environmental correction coefficient to generate a dynamic detection threshold set suitable for the current working condition; The calculation of the environmental correction coefficient includes: Establish a linear regression model between the groundwater level height and the pipe wall stress value, and obtain the correction factor of the temperature change on the elastic modulus of the pipe material; Use the grey relational analysis method to calculate the correlation coefficient between the soil moisture content and the pressure sensor reading fluctuation, and generate multi-dimensional environmental compensation parameters.

[0009] Preferably, the generation of the multi-dimensional fusion feature matrix specifically includes: Perform feature concatenation on the pressure mean value, pressure fluctuation variance, and flow rate change rate included in the time-domain feature vector and the main frequency component of the acoustic fingerprint and harmonic energy distribution included in the frequency-domain feature vector; Based on the pipe section connection relationship in the pipe network topology eigenvalue, perform spatial correlation analysis on the concatenated features, and generate a multi-dimensional fusion feature matrix containing 16-dimensional feature indicators.

[0010] Preferably, the noise filtering and outlier removal specifically include: Perform moving average filtering on the original data sequence collected by the pressure sensor to remove sudden interference pulses; Use the box plot analysis method to identify the outliers in the flow monitoring data, and perform linear interpolation correction on the values exceeding 1.5 times the interquartile range.

[0011] Preferably, the pattern matching calculation specifically includes: Calculate the Euclidean distance between each feature dimension in the multi-dimensional fusion feature matrix and the dynamic detection threshold of each damage type; After normalizing the Euclidean distances of each dimension, use the weighted summation algorithm to generate the comprehensive matching degree values of each damage type.

[0012] Preferably, the screening of the damage type corresponding to the maximum matching degree value as the current pipe network status diagnosis result specifically includes: Establish a descending order queue of the matching degree values, and select the damage type at the head of the queue as the priority diagnosis result; When the difference between the first two matching degree values is less than the set tolerance threshold, trigger manual review marking and record the pending confirmation status.

[0013] Preferably, it further includes a dynamic learning module, specifically including: After each maintenance operation is completed, obtain the actual damage type and location information confirmed by on-site investigation; Compare and analyze the actual data with the system diagnosis result, and calculate the feature recognition error amount; According to the error amount, adjust the parameter weight coefficients in the dynamic detection threshold set, and update the pattern matching rules in the historical maintenance database.

[0014] Preferably, the calculation and generation of three-dimensional coordinate positioning data of the damaged area in combination with the pipe network topological characteristic values ​​specifically includes: Based on the pipe segment coding rules in the pipe network topology diagram, the sensor node locations where abnormal data fluctuations occur are determined; According to the pipe section connection direction and spatial coordinate data, the triangulation positioning method is used to calculate the spatial position of the damaged point relative to the nearest three monitoring nodes.

[0015] Preferably, the parameter weight coefficients in the dynamic detection threshold set are adjusted according to the error amount, and the specific processing process is: Collect statistics on the recognition accuracy of each feature dimension in the last N maintenance operations; Weight attenuation is performed on feature dimensions whose accuracy is lower than the set threshold, and the attenuation amplitude is proportional to the accuracy deviation value; Implement weight enhancement adjustments for feature dimensions whose accuracy is consistently higher than the benchmark value.

[0016] Compared with the prior art, the present invention has the following beneficial effects: The system collects static structural parameters and real-time operation monitoring data of the drainage network through a distributed sensor network, and can obtain the network status information in a comprehensive and real-time manner. During the initial deployment, the system collects structural parameters such as pipe diameter, pipe type, and laying year, and establishes a network topology relationship map. At the same time, it combines pressure sensors, flow meters, and acoustic detectors to collect baseline data; when there are historical inspection records, the operating data of the last three maintenance cycles are extracted as the benchmark reference data set, ensuring the comprehensiveness of data collection and the scientific nature of the benchmark, providing a solid data foundation for subsequent inspections.

[0017] The data preprocessing module performs noise filtering and outlier removal on the real-time operation monitoring data, uses sliding average filtering to process the raw data of the pressure sensor to remove sudden interference pulses, and uses the box plot analysis method to identify outliers in the flow monitoring data and perform linear interpolation correction, which effectively improves the data quality and ensures the accuracy of subsequent analysis.

[0018] The feature fusion module fuses the time domain feature vector and the frequency domain feature vector based on the network topology feature value and the dynamic operation state feature value to generate a multi-dimensional fusion feature matrix containing 16-dimensional feature indicators. By cascading the time domain features such as the pressure mean and the pressure fluctuation variance with the frequency domain features such as the main frequency component of the soundprint and the harmonic energy distribution, and combining the connection relationship of the pipe sections for spatial correlation analysis, the organic fusion of multi-dimensional features is achieved, which can more comprehensively and accurately reflect the operation state of the pipeline network and improve the ability to extract damage features.

[0019] The intelligent analysis engine performs pattern matching calculations on the multi-dimensional fusion feature matrix and the dynamic detection threshold set of each damage type. By calculating the Euclidean distance, normalizing, and performing weighted summation to generate a comprehensive matching degree value, it can accurately identify the damage type of the pipe network. At the same time, a queue is established for arranging the matching degree values in descending order, and the first one is selected as the priority diagnosis result. When the difference between the first two is less than the set tolerance threshold, manual review is triggered, which not only ensures the accuracy of detection but also improves the reliability of detection.

[0020] The generation of the dynamic detection threshold set fully considers the influence of environmental factors. The standard parameter set of each damage type is extracted from the historical maintenance database, environmental variable data such as the height of the groundwater level, soil moisture content, and temperature change are obtained, and the environmental correction coefficient is calculated. The standard parameters are multiplied by the correction coefficient to generate dynamic thresholds suitable for the current working conditions. By establishing a linear regression model between the groundwater level and the pipe wall stress, obtaining the correction factor of temperature on the elastic modulus of the pipe material, and using the grey relational analysis method to calculate the correlation coefficient between the soil moisture content and the fluctuation of the pressure sensor readings to generate multi-dimensional environmental compensation parameters, the detection threshold can adapt to environmental changes in real time, improving the adaptability and accuracy of detection.

[0021] The spatial positioning module indexes the spatial distribution pattern of the corresponding damage type in the historical maintenance database according to the diagnosis result, and calculates and generates the three-dimensional coordinate positioning data of the damaged area in combination with the topological characteristic values of the pipe network. Based on the pipe segment coding rules of the pipe network topological relationship map, the position of the sensor node with abnormal data fluctuation is determined, and then the position of the damage point is deduced by using the triangulation method according to the pipe segment connection direction and spatial coordinate data, realizing the accurate positioning of the damaged area, providing a clear target for the pipe network maintenance, and improving the maintenance efficiency.

[0022] The system also includes a dynamic learning module. After each maintenance operation is completed, the actual damage type and location information confirmed by on-site investigation are obtained, compared and analyzed with the system diagnosis result to calculate the feature recognition error amount, and the parameter weight coefficient in the dynamic detection threshold set is adjusted according to the error amount, and the pattern matching rule in the historical maintenance database is updated. Through continuous learning and optimization, the system can gradually improve the accuracy and reliability of detection, realize the continuous improvement of detection ability, and adapt to the changes in the operation status of the pipe network and the development of detection requirements. Description of the Drawings

[0023] Figure 1 It is the working principle diagram of the urban drainage pipe network damage detection system based on intelligent analysis described in the present invention; Figure 2 It is the working principle diagram of the spatial positioning module; Figure 3 It is the module diagram of dynamic learning; Figure 4 It is the adjustment diagram of the parameter weight coefficient. Detailed implementation manners

[0024] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0025] Please refer to Figures 1-4 , the urban drainage network damage detection system based on intelligent analysis involved in the present invention includes a multi-source data acquisition module, a data preprocessing module, a feature fusion module, an intelligent analysis engine, and a spatial positioning module. The specific implementation steps are as follows: The multi-source data acquisition module collects static structure parameters and real-time operation monitoring data of the drainage network through a distributed sensor network, analyzes and processes them to generate network topology characteristic values and dynamic operation state characteristic values, and analyzes and processes them based on the standard parameter sets of various damage types in the historical maintenance database to generate dynamic detection threshold sets for various damage types.

[0026] The data preprocessing module receives real-time operation monitoring data and performs noise filtering and outlier removal, screens out valid monitoring data, and extracts time-domain feature vectors and frequency-domain feature vectors of the valid monitoring data.

[0027] The feature fusion module performs fusion processing on the time-domain feature vectors and frequency-domain feature vectors based on the network topology characteristic values and dynamic operation state characteristic values to generate a multi-dimensional fusion feature matrix.

[0028] The intelligent analysis engine performs pattern matching calculations on the multi-dimensional fusion feature matrix and the dynamic detection threshold sets of various damage types, outputs the matching degree values of various damage types, and screens out the damage type corresponding to the maximum matching degree value as the current network state diagnosis result.

[0029] The spatial positioning module indexes the spatial distribution patterns of the corresponding damage types in the historical maintenance database according to the diagnosis results, and calculates and generates three-dimensional coordinate positioning data of the damaged area in combination with the network topology characteristic values.

[0030] Next, the present invention will be further described in conjunction with Embodiments 1 to 5: Embodiment 1

[0031] This embodiment relates to the specific process of collecting static structure parameters and real-time operation monitoring data of the drainage network in the multi-source data acquisition module. This process is divided into two different data acquisition modes according to whether the detection equipment is first deployed in the network system, aiming to provide comprehensive, accurate and comparable basic data for the entire damage detection system, and ensure the reliability and effectiveness of subsequent detection and analysis.

[0032] When the detection equipment is initially deployed in the pipe network system, the multi-source data acquisition module first collects static structure parameters. Specifically, key structure parameters such as pipe diameter size, pipe material type, and laying age need to be obtained. The pipe diameter size directly reflects the flow capacity of the pipe network. There are significant differences in the water flow characteristics and pressure distribution of pipe networks with different pipe diameters; the pipe material type determines the physical properties of the pipe network, such as strength, corrosion resistance, etc. Different pipe materials may have different types and patterns of damage during long-term use; the laying age is related to the aging degree of the pipe network. Pipe networks with a long laying age are more likely to have problems such as cracks and leaks. By collecting these structure parameters, a pipe network topology relationship map can be established. This map visually displays information such as the connection relationship and spatial layout of each pipe segment in the pipe network in a graphical manner, providing a basic framework for subsequent analysis of the topological characteristic values of the pipe network.

[0033] After establishing the pipe network topology relationship map, the module activates pressure sensors, flow meters, and acoustic detectors to collect baseline data. The pressure sensor is used to monitor the pressure changes in the pipe network. By collecting pressure data in real time, it can reflect the operating load of the pipe network and whether there are abnormal pressure fluctuations; the flow meter is used to measure the flow rate in the pipe network. The change in flow rate can indirectly reflect the flow state of the pipe network, such as whether there is a blockage; the acoustic detector can obtain voiceprint feature data by capturing the sound signals in the pipe network. Different types of damage (such as leaks, cracks, etc.) will produce different voiceprint features. These three types of sensors work together to comprehensively collect real-time operation monitoring data of the pipe network in the initial state from three dimensions: pressure, flow rate, and acoustics, forming baseline data. These baseline data represent the various indicators of the pipe network in the normal operating state and are an important benchmark for subsequent judgment of whether the pipe network is damaged.

[0034] When there are historical detection records in the pipe network system, the focus of the multi-source data acquisition module's work shifts to extracting relevant data from the historical maintenance database. The specific operation is to extract the operating pressure fluctuation data, flow rate change data, and voiceprint feature data of the last three maintenance cycles and record them as the reference dataset for the current real-time monitoring. The data of the last three maintenance cycles are selected because they can be closer to the current operating condition of the pipe network and reduce data deviation caused by too long a time interval. The operating pressure fluctuation data can reflect the change trend and abnormal fluctuation of the pipe network pressure during the past maintenance cycle; the flow rate change data can reflect the fluctuation law of the flow rate at different time periods and whether there is a sudden change in the flow rate; the voiceprint feature data records the abnormal sound characteristics that may occur in the pipe network during historical operation. Using these data as the reference dataset for comparison can provide a more timely basis for the current real-time monitoring data.

[0035] During the data acquisition process, certain principles need to be followed for the deployment of sensors. The node layout of the distributed sensor network should be optimized according to the topological structure of the pipe network and actual requirements to ensure comprehensive coverage of key parts of the pipe network, such as pipe section joints, elbows, and areas near valves, where damage is prone to occur. The sampling frequency of the sensors should be set according to the operating characteristics of the pipe network and the requirements of detection accuracy, ensuring that subtle abnormal changes can be captured while avoiding data redundancy caused by too high sampling frequency, which increases the burden of subsequent data processing.

[0036] For the collected static structure parameters and real-time operation monitoring data, the multi-source data acquisition module needs to perform preliminary analysis and processing. For static structure parameters, by integrating and calculating data such as pipe diameter size, pipe material type, and laying age, topological characteristic values of the pipe network are generated, which can quantify the topological structure information of the pipe network, such as pipe section length, connection angle, node density, etc. For real-time operation monitoring data, the module generates dynamic operation state characteristic values by analyzing the change rules and statistical characteristics of data such as pressure, flow rate, and acoustic fingerprint, such as the average value, maximum value, and minimum value of pressure, the change rate of flow rate, and the main frequency distribution of acoustic fingerprint. The generation of these characteristic values provides a more representative and analytically valuable data form for subsequent data preprocessing, feature fusion, and other links.

[0037] In terms of data storage, all the collected data and the generated characteristic values need to be stored in the historical maintenance database in a timely and accurate manner. The database design should have good scalability and query performance, enabling convenient retrieval and invocation of data from different time periods and of different types. At the same time, a data backup and recovery mechanism needs to be established to ensure the security and integrity of the data and prevent the normal operation of the system from being affected by data loss.

[0038] In addition, the multi-source data acquisition module also needs to have the ability to interact with other modules. After completing data acquisition and preliminary processing, it timely transmits the topological characteristic values of the pipe network, dynamic operation state characteristic values, and real-time operation monitoring data to the data preprocessing module for subsequent operations such as noise filtering and outlier removal. At the same time, it can receive feedback information from other modules, such as abnormal data prompts found by the intelligent analysis engine during the pattern matching process, so as to adjust the focus and frequency of data acquisition and further improve the pertinence and effectiveness of data acquisition.

[0039] In practical applications, whether it is baseline data collection at the time of initial deployment or extraction of benchmark reference data sets based on historical records, the authenticity and accuracy of the data must be ensured. Before data collection, staff must calibrate and debug sensor equipment to ensure that its measurement accuracy meets the requirements; during data collection, sensor equipment should be regularly inspected and maintained to promptly detect and handle equipment failures to avoid data anomalies caused by equipment problems. At the same time, data in the historical maintenance database needs to be strictly reviewed and verified to ensure data integrity and reliability, and to prevent historical data errors from affecting the accuracy of current test results.

[0040] Through the above two data collection modes, the multi-source data collection module can provide comprehensive, accurate and timely basic data for the urban drainage pipe network damage detection system, laying a solid foundation for subsequent damage detection and diagnosis. Whether it is a newly deployed pipe network system or a pipe network system with historical detection records, this module can ensure that the system can timely and accurately detect damage problems in the pipe network through reasonable data collection and processing methods, and provide strong technical support for the maintenance and management of urban drainage pipe networks.

[0041] Example 2

[0042] This embodiment is centered around generating a dynamic detection threshold set for each damage type. This process combines historical data with real-time environmental variables to dynamically modify standard detection parameters to meet the needs of pipeline damage detection under different working conditions. Specifically, it includes four core links: standard parameter extraction, environmental variable acquisition, environmental correction coefficient calculation, and dynamic threshold generation. Each link is closely connected to ensure that the generated threshold set is scientific and practical.

[0043] The standard parameter sets for each damage type are extracted from the historical maintenance database. The historical maintenance database stores a large amount of inspection and maintenance data accumulated during the long-term operation of the pipeline network, covering the typical characteristic parameters of different damage types. For example, the standard parameters of the leakage state include the standard value of the pressure gradient, which reflects the typical amplitude of the pressure drop in the unit length of the pipe section when the pipeline network leaks; the standard parameters of the pipe body cracks include the standard characteristics of the soundprint spectrum, which can form a recognizable spectrum pattern by analyzing the frequency distribution characteristics of the sound waves generated by the cracks; the standard parameters of the blockage condition involve the standard amplitude of the flow mutation, which is used to characterize the degree of sudden decrease in flow when blockage occurs. These standard parameter sets are based on the statistical analysis of a large number of previous damage cases, are widely representative, and provide a basic reference for current detection.

[0044] Obtain the current environmental variable data, including the groundwater level height, soil moisture content, and temperature change parameters. The change in the groundwater level height will generate different hydrostatic pressures on the pipe wall, thereby affecting the stress state of the pipe network; the level of soil moisture content may change the supporting force and frictional force of the soil on the pipe network, resulting in uneven stress on the pipe material; temperature changes will cause the pipe material to expand and contract thermally, changing its elastic modulus and internal stress distribution. To accurately obtain these environmental variables, corresponding monitoring devices can be deployed around the pipe network: the groundwater level height can be monitored in real time through a liquid level sensor, the soil moisture content can be measured using a soil moisture sensor, and the environmental temperature change can be recorded with a temperature sensor. These monitoring devices need to have high precision and stability to ensure that the collected data can truly reflect the current environmental conditions.

[0045] After obtaining the environmental variable data, enter the calculation stage of the environmental correction factor. First, establish a linear regression model between the groundwater level height and the pipe wall stress value. Through the measured values of the groundwater level height and the pipe wall stress in the historical data, use the linear regression analysis method to determine the functional relationship between the two. For example, assume that the groundwater level height is , and the pipe wall stress value is . Through regression analysis, a linear equation of is obtained, where and are regression coefficients. This model can quantify the influence degree of the change in the groundwater level height on the pipe wall stress, providing a basis for subsequent correction of pressure-related parameters.

[0046] At the same time, obtain the correction factor of the temperature change on the elastic modulus of the pipe material. The elastic modulus is an important indicator to measure the ability of a material to resist elastic deformation. Temperature changes will cause the elastic modulus of the pipe material to change, thereby affecting the mechanical properties of the pipe network. By referring to the physical property parameter table of the pipe material or conducting a material mechanics experiment, determine the change ratio of the elastic modulus of the pipe material in different temperature ranges to form a temperature correction factor. For example, when the temperature of a certain pipe material rises by each time, the elastic modulus decreases by , then this ratio can be used as the temperature correction factor.

[0047] Grey correlation analysis is used to calculate the correlation coefficient between soil moisture content and pressure sensor reading fluctuation. Grey correlation analysis is a systematic analysis method used to study small amounts of data and uncertain factors, and is suitable for quantitative analysis of the complex relationship between soil moisture content and pressure fluctuation. After dimensionless processing of the soil moisture content data sequence and the pressure sensor reading fluctuation data sequence, the correlation coefficient and correlation degree between the two are calculated to obtain the coefficient value reflecting the correlation between the two. The larger the coefficient value, the more significant the impact of soil moisture changes on pressure sensor reading fluctuations, and vice versa. Through the above steps, environmental compensation parameters containing multi-dimensional influencing factors such as groundwater level, temperature, and soil moisture content are generated, which comprehensively reflect the impact of environmental variables on the operation status of the pipeline network.

[0048] Finally, each parameter item in the standard parameter set is multiplied by the environmental correction coefficient to generate a dynamic detection threshold set that adapts to the current working conditions. For example, for the standard value of the pressure gradient in the leakage state, assuming that its original value is , the pressure gradient threshold under the current working condition is calculated by the environmental correction coefficient: ,in It is a comprehensive environmental correction coefficient, including the influence of groundwater level, temperature, soil moisture content, etc. Similarly, the standard characteristics of the soundprint spectrum of pipe cracks and the standard amplitude of flow mutation in blocking conditions are corrected accordingly to form a complete set of dynamic detection thresholds.

[0049] In practical applications, attention should be paid to the real-time and dynamic nature of environmental variables. Factors such as groundwater level, soil moisture content, and temperature may change significantly over time, season, weather, and other conditions. Therefore, the environmental correction coefficient needs to be updated and calculated regularly to ensure that the dynamic detection threshold set can reflect the current environmental conditions in a timely manner. For example, when the groundwater level rises during the rainy season, the environmental variable data should be recollected and the environmental correction coefficient should be updated to avoid deviations in the detection results due to threshold lag.

[0050] In addition, the integrity and accuracy of the historical maintenance database are crucial to the extraction of the standard parameter set. A strict data entry and review mechanism needs to be established to ensure that the parameter information of each damage record is authentic and reliable. At the same time, as new damage cases continue to accumulate, the standard parameter set should be regularly updated and expanded to include characteristic parameters of new types of damage, so as to enhance the system's ability to detect complex damage situations.

[0051] When calculating the environmental correction coefficient, it is necessary to reasonably select the analysis method and model according to the actual layout of the pipe network and the characteristics of the pipe material. For example, for pipes of different materials (such as cast iron pipes and plastic pipes), the influence degree of temperature on their elastic modulus may be different, and correction models need to be established separately; for areas with complex geological conditions, the correlation between soil moisture content and pressure fluctuation may show non-linear characteristics, and non-linear regression or other data analysis methods can be considered for processing.

[0052] The generated set of dynamic detection thresholds needs to be adapted to the pattern matching algorithm of the intelligent analysis engine. When designing the threshold parameters, it is necessary to consider the dimension and range of the feature dimensions to ensure comparability with the feature indicators in the multi-dimensional fusion feature matrix. For example, the pressure gradient threshold is consistent with the dimensions of time-domain features such as the pressure mean and the pressure fluctuation variance, and the voiceprint spectrum threshold matches the dimensions of frequency-domain features such as the main frequency component of the voiceprint and the harmonic energy distribution, so as to ensure the accuracy of the pattern matching calculation.

[0053] The set of dynamic detection thresholds generated through the above steps not only retains the reference value of historical experience data but also fully considers the influence of current environmental factors, enabling the detection system to maintain high sensitivity and accuracy under different working conditions. This mechanism avoids the limitations of using fixed thresholds for detection, improves the adaptability of the system to complex environments, and provides a scientific and reliable basis for judgment for the real-time damage detection of urban drainage pipe networks.

[0054] Embodiment 3

[0055] This embodiment focuses on the process of noise filtering and outlier removal in the data preprocessing module and the generation of the multi-dimensional fusion feature matrix in the feature fusion module. In the data preprocessing stage, for the original data sequence collected by the pressure sensor, the moving average filtering method is a key step. This method smooths the signal by calculating the local average of the data points and effectively removes sudden interference pulses. Specifically, for a given original pressure data sequence, a sliding window with a fixed length is defined, and the window moves point by point on the data sequence. After each move, the average value of the data points within the window is calculated as the filtered value of this point. This method can retain the main trend of the data while reducing the influence of random noise, making the pressure data more stable and reliable.

[0056] For the flow monitoring data, the box plot analysis method is used to identify the outliers. The box plot divides the data into four parts through quartiles, where the interquartile range (IQR) is defined as the difference between the upper quartile (Q3) and the lower quartile (Q1). The outliers are defined as the values that exceed 1.5 times the interquartile range. For the identified outliers, linear interpolation correction is performed. Linear interpolation is a method of estimating the value of an unknown data point between two known data points. By constructing a linear function using the normal data points before and after the outlier, a reasonable replacement value is calculated to ensure the accuracy and effectiveness of the flow data.

[0057] After noise filtering and outlier removal, the effective monitoring data are screened out, and the time-domain feature vectors and frequency-domain feature vectors of the effective monitoring data are extracted. The time-domain feature vectors mainly focus on the statistical characteristics of the signal in the time domain, including the pressure mean, pressure fluctuation variance, flow rate change rate, etc. The pressure mean reflects the average level of the pipe network pressure, the pressure fluctuation variance measures the degree of dispersion of the pressure data relative to the mean, and the flow rate change rate describes how fast the flow rate changes over time. The frequency-domain feature vectors convert the time-domain signal to the frequency domain through Fourier transform to analyze the frequency components of the signal, including the main frequency component of the acoustic signature, harmonic energy distribution, etc. The main frequency component of the acoustic signature is the frequency component with the strongest energy in the acoustic signature signal, and different types of damage often have different main frequency characteristics; the harmonic energy distribution describes the proportion of the energy of each harmonic, providing richer frequency-domain information.

[0058] In the feature fusion module, the time-domain feature vectors and frequency-domain feature vectors are cascaded. Feature cascading is to connect the feature vectors from different sources in sequence into a longer vector, so that different types of feature data are combined with each other to form a more comprehensive feature representation. Specifically, the features such as the pressure mean, pressure fluctuation variance, and flow rate change rate included in the time-domain feature vectors are arranged in sequence with the features such as the main frequency component of the acoustic signature and harmonic energy distribution included in the frequency-domain feature vectors to form a high-dimensional feature vector.

[0059] Then, based on the pipe segment connection relationship in the pipe network topological feature values, spatial correlation analysis is performed on the cascaded features. The pipe network topological feature values describe information such as the connection methods and spatial position relationships of each pipe segment in the pipe network. By analyzing these relationships, the correlation between sensor data at different positions can be understood. For example, the pressure and flow rate changes of adjacent pipe segments often have a certain correlation. When a certain pipe segment is damaged, the operating status of its adjacent pipe segments will also be affected. Using this spatial correlation, the cascaded features are weighted, with higher weights assigned to the sensor data closer to the damage point and lower weights assigned to the sensor data farther away, so as to highlight the key information related to damage detection.

[0060] Based on spatial correlation analysis, a multi-dimensional fusion feature matrix containing 16-dimensional feature indicators is generated. These 16-dimensional feature indicators comprehensively consider the topological structure information of the pipe network, as well as the time-domain and frequency-domain characteristics of the operating state. For example, they may include features such as the average pressure of sensors at different locations, the variance of pressure fluctuations, the flow rate change rate, the main frequency component of the acoustic fingerprint, the harmonic energy distribution, etc., and the spatial correlation features calculated based on the topological relationship of the pipe network. These feature indicators complement each other and describe the operating state of the pipe network from multiple perspectives, providing rich and comprehensive feature data for the intelligent analysis engine to perform pattern matching calculations.

[0061] The generation process of the multi-dimensional fusion feature matrix needs to strictly follow the data processing flow. First, standardize the concatenated feature vectors to map the values of each feature dimension to the same scale range, eliminate the influence of the dimensionality of different feature dimensions, and ensure the accuracy of subsequent calculations. Then, determine the spatial weights of each feature dimension according to the topological feature values of the pipe network and apply the weights to the standardized feature vectors. Finally, organize the processed feature vectors into a matrix form according to certain rules to form the multi-dimensional fusion feature matrix.

[0062] In practical applications, the effect of feature fusion is affected by various factors. The layout density and location of sensors will affect the quality of the acquired raw data and the accuracy of spatial correlation analysis. If the sensor layout is too sparse, it may not be able to capture the subtle changes caused by local damage; if the sensor layout location is unreasonable, it may lead to weak data correlation and affect the effect of feature fusion. The complexity of the topological structure of the pipe network also poses challenges to spatial correlation analysis. For complex pipe network systems, the connection relationships between pipe segments are intricate, and more advanced graph theory algorithms and spatial analysis methods are required to accurately describe and analyze such relationships.

[0063] In addition, the selection and combination of feature dimensions are also crucial. Too many feature dimensions may lead to data redundancy, increase the computational complexity, and may also introduce noise; while too few feature dimensions may not be able to comprehensively describe the operating state of the pipe network and affect the accuracy of damage detection. Therefore, it is necessary to reasonably select and combine feature dimensions according to the actual characteristics of the pipe network and detection requirements to ensure that the multi-dimensional fusion feature matrix can fully reflect the operating state of the pipe network and has high computational efficiency.

[0064] The collaborative work between the data preprocessing module and the feature fusion module is also very critical. The quality of data preprocessing directly affects the subsequent feature extraction and fusion effects. If the noise filtering is not thorough or the outlier removal is inaccurate, it may cause the extracted feature vectors to contain incorrect information, thereby affecting the quality of the multi-dimensional fusion feature matrix. Therefore, the two modules need to cooperate closely and continuously optimize the processing flow to ensure the accuracy and reliability of the entire data processing process.

[0065] After the multi-dimensional fusion feature matrix is generated, it will be used as the input data for the intelligent analysis engine for subsequent pattern matching calculations. Each feature dimension in the matrix represents an aspect of the pipeline network operation status. By comprehensively analyzing the values of these feature dimensions, it is possible to more accurately determine whether there are damages in the pipeline network and the types of damages. For example, when abnormal changes occur simultaneously in the main frequency component of the acoustic fingerprint and the variance of pressure fluctuations in the matrix, it may indicate damages such as cracks or leaks in the pipeline network.

[0066] Through noise filtering, outlier removal, and feature extraction in the data preprocessing module, as well as feature concatenation, spatial correlation analysis, and matrix generation in the feature fusion module, a comprehensive and accurate description of the pipeline network operation status is achieved. This multi-dimensional fusion feature representation method fully utilizes the information in the time domain, frequency domain, and spatial domain, improves the sensitivity and specificity of damage detection, and provides strong technical support for the health monitoring of urban drainage pipeline networks.

[0067] Example 4

[0068] The process of the intelligent analysis engine performing pattern matching calculations and screening diagnostic results is the core link of the urban drainage pipeline network damage detection system. This process realizes the intelligent identification and diagnosis of the pipeline network damage types through the precise comparison of the multi-dimensional fusion feature matrix with the dynamic detection threshold set.

[0069] When performing pattern matching calculations, the intelligent analysis engine first calculates the Euclidean distance between each feature dimension in the multi-dimensional fusion feature matrix and the dynamic detection thresholds of each damage type. The Euclidean distance is a classic method for measuring the distance between two points in a multi-dimensional space. By calculating the Euclidean distance between the feature vector and the threshold vector, the difference degree between the two can be quantified. Specifically, for each feature vector in the multi-dimensional fusion feature matrix, subtract the value of each dimension from the corresponding dimension value of the dynamic detection threshold vector of the corresponding damage type, square the results and sum them, and then take the square root to obtain the Euclidean distance between the feature vector and the threshold of this damage type. The smaller the Euclidean distance, the higher the matching degree between the feature vector and this damage type; conversely, the lower the matching degree.

[0070] To eliminate the influence of the dimensions of different feature dimensions and make the distances of each dimension comparable, it is necessary to normalize the calculated Euclidean distance. The normalization process maps the Euclidean distance to the interval [0,1]. Common normalization methods include min-max normalization, z-score normalization, etc. In this system, the min-max normalization method is adopted. Subtract the minimum value of each dimension from the Euclidean distance of each feature dimension, and then divide by the difference between the maximum value and the minimum value of this dimension to obtain the normalized distance value. After such processing, the distance values of different feature dimensions are in the same scale range, which is convenient for subsequent comprehensive calculations.

[0071] After obtaining the normalized Euclidean distance, the weighted summation algorithm is used to generate the comprehensive matching degree values for each type of damage. Since the importance of different feature dimensions may vary in damage detection, different weights need to be assigned to each feature dimension. The determination of these weights is based on the statistical analysis of historical data and the empirical knowledge of domain experts. For example, in leak detection, the pressure gradient feature may be more important than the flow rate change rate feature, so a higher weight is assigned to the pressure gradient feature. By multiplying the normalized Euclidean distance by the weight of the corresponding feature dimension and then summing, the comprehensive matching degree values for each type of damage are obtained. The smaller the comprehensive matching degree value, the higher the matching degree between the feature vector and the type of damage.

[0072] When screening the diagnostic results, first establish a queue arranged in descending order of the matching degree values. Arrange the comprehensive matching degree values of each type of damage calculated in descending order to form a queue. The first item in the queue is the type of damage with the largest matching degree value, that is, the most likely type of damage in the current pipe network state.

[0073] Normally, select the type of damage at the head of the queue as the priority diagnostic result. However, when the difference between the first two matching degree values is less than the set tolerance threshold, it means that the matching degrees of these two types of damage are relatively close and it is difficult to directly determine the accurate type of damage. At this time, the system will trigger an artificial review mark and record the pending confirmation status. The set tolerance threshold is to avoid misjudgment caused by calculation errors or data fluctuations and ensure the reliability of the diagnostic results. The triggering of the artificial review mark reminds the staff that they need to further check and confirm the detection results, and determine the final type of damage by combining on-site investigation or other detection means.

[0074] In practical applications, the accuracy of pattern matching calculation is affected by various factors. The rationality of the dynamic detection threshold set is directly related to the reliability of the matching results. If the threshold is set too loose, it may lead to misjudgment, misjudging the normal state as a damaged state; if the threshold is set too strict, it may lead to missed detection and fail to detect real damage in time. Therefore, it is necessary to continuously optimize and adjust the dynamic detection threshold set according to the actual pipe network operation data and historical damage cases.

[0075] The selection of feature dimensions and weight assignment also have important impacts on the pattern matching results. If the selected feature dimensions cannot effectively distinguish different types of damage, or the weight assignment is unreasonable, it may lead to inaccurate calculation of the matching degree. During the system design and optimization process, a large number of experiments and analyses are required to select the most representative and discriminative feature dimensions and determine a reasonable weight assignment scheme.

[0076] To improve the intelligence level and adaptive ability of the system, the intelligent analysis engine can also introduce machine learning algorithms for pattern recognition and classification. For example, algorithms such as Support Vector Machine (SVM) and Random Forest can be used to train historical data to establish a damage type classification model. Then, the multi-dimensional fusion feature matrix is used as the input, and through the trained model, classification prediction is carried out to obtain the diagnostic result of the damage type. This method can automatically learn the complex relationship between features and damage types, improving the accuracy and efficiency of diagnosis.

[0077] When processing large-scale pipe network data, computational efficiency is also an important factor to consider. Pattern matching calculation involves a large number of comparison operations between feature vectors and threshold vectors. If the computational efficiency is low, it may lead to an overly long system response time, affecting the effect of real-time detection. To improve computational efficiency, parallel computing technology can be adopted to distribute computational tasks to multiple processors or computing nodes for simultaneous processing, greatly shortening the computational time.

[0078] The intelligent analysis engine can also cooperate closely with other modules of the system to form a complete closed-loop detection system. For example, in cooperation with the spatial positioning module, the location of the damaged area can be quickly determined according to the diagnostic result; in cooperation with the dynamic learning module, the diagnostic result and the actual damage situation are fed back to the system to continuously update and optimize the detection model and threshold set, improving the detection ability and adaptability of the system.

[0079] The pattern matching calculation and diagnostic result screening process of the intelligent analysis engine are the key links of the urban drainage pipe network damage detection system. Through accurate Euclidean distance calculation, reasonable normalization processing, and weighted summation algorithm, combined with a scientific diagnostic result screening mechanism, the damage type of the pipe network can be accurately identified, providing strong technical support for the maintenance and management of the pipe network. In practical applications, it is also necessary to continuously optimize algorithms and models to improve the accuracy, reliability, and intelligence level of the system to adapt to the complex and changeable operating environment of the urban drainage pipe network.

[0080] Example 5

[0081] This example involves the process of generating three-dimensional coordinate positioning data for the damaged area in the dynamic learning module and the spatial positioning module. These two modules respectively improve the adaptability and maintenance efficiency of the detection system from the perspectives of system self-optimization and precise positioning, forming a complete detection closed-loop.

[0082] The core function of the dynamic learning module is to continuously optimize the system's detection ability by maintaining the actual data feedback after maintenance operations. After each maintenance operation is completed, the system first obtains the actual damage type and location information confirmed through on-site inspection. The on-site inspection data is obtained through manual on-site detection, image recording, etc., to ensure the authenticity and accuracy of the data. For example, staff use equipment such as pipeline endoscopes and ultrasonic detectors to conduct on-site verification of the pipe sections diagnosed as leaking, and record information such as the pipe section numbers of the actual damage locations and the damage forms (such as crack lengths, leakage hole diameters).

[0083] Next, the actual data is compared and analyzed with the system diagnosis results to calculate the feature recognition error amount. During the comparison process, for each feature dimension (such as the mean pressure, the main frequency component of the acoustic fingerprint, etc.), the deviation between the system prediction value and the actual value is calculated to form an error vector. For example, if the actual damage type of a certain pipe section is a crack but the system diagnoses it as a leak, then it is necessary to analyze the error distribution of feature dimensions such as the pressure gradient and the acoustic frequency spectrum to determine the key features leading to misjudgment.

[0084] Based on the error amount analysis, the system enters the parameter weight adjustment phase. First, the recognition accuracy data of each feature dimension in the recent N maintenance operations is statistically analyzed. The value of N can be set according to the pipe network scale and maintenance frequency, usually 50 - 100 times, to ensure the reliability of the statistical results. The accuracy calculation method is: the ratio of the number of times a certain feature dimension correctly identifies the damage type in all maintenance operations to the total number of times. For example, if the pressure gradient feature correctly assisted in identifying the damage type 60 times in 80 maintenance operations, then its accuracy is 75%.

[0085] For feature dimensions with an accuracy lower than the set threshold (such as 70%), weight attenuation processing is performed, and the attenuation amplitude is proportional to the accuracy deviation value. Suppose the accuracy of a certain feature dimension is 60% and the deviation value is 10%, then its weight attenuation amplitude can be set to 10% of the initial weight. The attenuation process is realized through a linear function, that is, the weight adjustment coefficient = 1 - k×(threshold - accuracy), where k is the attenuation coefficient (0 < k < 1). Conversely, for feature dimensions with an accuracy continuously higher than the benchmark value (such as 85%), weight enhancement adjustment is implemented, and the enhancement amplitude is related to the proportion exceeding the benchmark value. For example, if it exceeds 10%, the weight increases by 5%. Through this dynamic adjustment mechanism, the system can automatically weaken the influence of redundant or inefficient features, strengthen the role of key features, and gradually optimize the parameter weight coefficients in the dynamic detection threshold set.

[0086] After the weight adjustment is completed, the system will update the actual damage data, diagnostic error information and adjusted weight coefficients to the historical maintenance database to form a new pattern matching rule. The updated database provides a more accurate reference for subsequent detection. For example, when the weight of a feature dimension is enhanced, the intelligent analysis engine will pay more attention to the difference between the feature and the threshold during pattern matching, thereby improving the recognition accuracy of the corresponding damage type.

[0087] In the spatial positioning module, the process of generating three-dimensional coordinate positioning data of the damaged area is based on the pipe network topology relationship map and the sensor node layout. First, according to the pipe segment coding rules in the pipe network topology relationship map, the sensor node location where the abnormal data fluctuation occurs is determined. The pipe segment coding rules use a hierarchical coding method, such as "area code-pipe segment serial number-node number", to ensure that each sensor node corresponds to a unique physical location. After the intelligent analysis engine outputs the damage diagnosis result, the system locates the specific node position by querying the sensor ID corresponding to the abnormal feature in the multi-dimensional fusion feature matrix.

[0088] Then, according to the pipe connection direction and spatial coordinate data, the triangulation method is used to calculate the spatial position of the damaged point relative to the nearest three monitoring nodes. The principle of the triangulation method is to use three monitoring nodes (A, B, C) with known coordinates to calculate the coordinates of the damaged point P by measuring the time difference or signal strength difference of the signal sent by the damaged point to each node. The specific steps are as follows: Determine the three-dimensional coordinates of monitoring nodes A, B, and C respectively , , ; Get the time difference or distance difference between the damaged point signal and each node, for example (in is the signal propagation speed); Establish a hyperbolic equation system and determine the coordinates of the damage point P by solving the equation system .

[0089] In practical applications, considering the spatial distribution characteristics of underground pipe networks, the spatial coordinates of monitoring nodes are obtained through the geographic information system (GIS) combined with pipeline laying drawings, and the accuracy must reach the centimeter level. The signal propagation speed is calibrated according to the pipe network medium (such as water, air) and sensor type (such as pressure sensor, acoustic detector). For example, the propagation speed of sound waves in water is about 1500 meters per second.

[0090] After generating the three-dimensional coordinate positioning data, the system presents the results in a visual manner in the pipeline network topology map, marking the specific location of the damaged point, the predicted damage type, and the confidence level. Staff can view the positioning information in real time through a mobile terminal or a monitoring platform, and combine data such as the pipeline network burial depth and the surrounding environment to formulate the optimal maintenance plan. For example, if the damaged point is located under a major traffic artery, the system can automatically prompt the adoption of trenchless repair technology to reduce the impact on traffic.

[0091] The coordinated operation of the dynamic learning module and the spatial positioning module is reflected in: the dynamic learning module improves the diagnostic accuracy by optimizing the feature weights, providing a more reliable prerequisite for spatial positioning; while the accurate results of the spatial positioning module provide real feedback data for the dynamic learning module, forming a closed loop of "detection - positioning - optimization". For example, when positioning deviations occur in a certain area multiple times, the dynamic learning module can identify the weight problems of relevant feature dimensions and improve the detection and positioning accuracy of that area after targeted adjustment.

[0092] In terms of data storage and transmission, the error data, weight adjustment records, and positioning results generated during the dynamic learning process need to be encrypted and stored in the historical maintenance database, and the blockchain technology is used to ensure the data cannot be tampered with. At the same time, the positioning data is transmitted to the maintenance personnel's terminal in real time through a dedicated communication network, supporting the offline map loading and navigation functions to ensure that the damaged area can be accurately reached even in a network-free environment.

[0093] In addition, the system needs to have the ability to fuse multi-source data. For example, by combining geological exploration data (such as soil type, groundwater level) and pipeline network design parameters (such as pipe material strength, slope), the triangulation positioning results can be corrected. For complex pipeline network structures (such as multi-branch intersections, deep underground pipeline networks), 3D modeling technology can be introduced to construct a pipeline network spatial model, and model simulation can be used to assist positioning calculations to improve the positioning accuracy in complex scenarios.

[0094] Through the adaptive optimization of the dynamic learning module and the accurate calculation of the spatial positioning module, the urban drainage pipeline network damage detection system has achieved an upgrade from "passive detection" to "active optimization", which can not only quickly identify the damage type, but also continuously improve the detection ability, providing key technical support for the refined management and efficient maintenance of the pipeline network.

[0095] It should be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0096] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An urban drainage pipe network damage detection system based on intelligent analysis, characterized in that, Including: A multi-source data acquisition module, which is used to collect static structure parameters and real-time operation monitoring data of the drainage pipe network through a distributed sensor network, analyze and process them to generate pipe network topology eigenvalue and dynamic operation state eigenvalue, and analyze and process based on the standard parameter set of each damage type in the historical maintenance database to generate a dynamic detection threshold set for each damage type; A data preprocessing module, which is used to receive real-time operation monitoring data, perform noise filtering and outlier removal, screen out effective monitoring data, and extract the time-domain feature vector and frequency-domain feature vector of the effective monitoring data; A feature fusion module, which is based on the pipe network topology eigenvalue and dynamic operation state eigenvalue, and performs fusion processing on the time-domain feature vector and frequency-domain feature vector to generate a multi-dimensional fusion feature matrix; An intelligent analysis engine, which is used to perform pattern matching calculations on the multi-dimensional fusion feature matrix and the dynamic detection threshold set of each damage type, output the matching degree values of each damage type, and screen out the damage type corresponding to the maximum matching degree value as the current pipe network state diagnosis result; A spatial positioning module, which is used to index the spatial distribution pattern of the corresponding damage type in the historical maintenance database according to the diagnosis result, and calculate and generate the three-dimensional coordinate positioning data of the damaged area in combination with the pipe network topology eigenvalue.

2. The urban drainage pipe network damage detection system based on intelligent analysis according to claim 1, wherein: The specific process of collecting the static structure parameters and real-time operation monitoring data of the drainage pipe network is as follows: When the detection equipment is initially deployed in the pipe network system, collect the structure parameters of the pipe diameter size, pipe material type, and laying age, establish a pipe network topology relationship map based on the structure parameters, and start the pressure sensor, flowmeter, and acoustic detector to collect baseline data; When there are historical detection records in the pipe network system, extract the operation pressure fluctuation data, flow change data, and acoustic fingerprint feature data of the last three maintenance cycles from the historical maintenance database, and record them as the reference data set for the current real-time monitoring.

3. The urban drainage network damage detection system based on intelligent analysis according to claim 1, characterized in that: The specific generation of the dynamic detection threshold set for each damage type includes: Extract the standard parameter set of each damage type from the historical maintenance database, including the standard value of the pressure gradient in the leakage state, the standard acoustic fingerprint spectrum feature of the pipe body crack, and the standard amplitude of the flow rate mutation in the blockage condition; Obtain the current environmental variable data, including the groundwater level height, soil moisture content, and temperature change parameters, and calculate and generate an environmental correction coefficient; Multiply each parameter item in the standard parameter set by the environmental correction coefficient to generate a dynamic detection threshold set suitable for the current working condition; The calculation of the environmental correction coefficient includes: Establish a linear regression model between the groundwater level height and the pipe wall stress value, and obtain the correction factor of the temperature change on the pipe material elastic modulus; Use the grey relational analysis method to calculate the correlation coefficient between the soil moisture content and the pressure sensor reading fluctuation, and generate a multi-dimensional environmental compensation parameter.

4. The urban drainage network damage detection system based on intelligent analysis according to claim 3, characterized in that: The specific generation of the multi-dimensional fusion feature matrix includes: Perform feature concatenation on the pressure mean value, pressure fluctuation variance, and flow rate change rate included in the time-domain feature vector and the main acoustic fingerprint frequency component and harmonic energy distribution included in the frequency-domain feature vector; Based on the pipe section connection relationship in the pipe network topology eigenvalue, perform spatial correlation analysis on the concatenated features to generate a multi-dimensional fusion feature matrix containing 16-dimensional feature indicators.

5. The urban drainage pipe network damage detection system based on intelligent analysis according to claim 4, characterized in that: The specific implementation of noise filtering and outlier removal includes: Perform sliding average filtering on the raw data sequence collected by the pressure sensor to remove sudden interference pulses; The box plot analysis method was used to identify outliers in the flow monitoring data, and linear interpolation correction was performed on the values ​​that exceeded 1.5 times the interquartile range.

6. The urban drainage pipe network damage detection system based on intelligent analysis according to claim 4, characterized in that: The pattern matching calculation specifically includes: Calculate the Euclidean distance between each feature dimension in the multidimensional fusion feature matrix and the dynamic detection threshold of each damage type; After normalizing the Euclidean distances of each dimension, a weighted sum algorithm is used to generate the comprehensive matching value of each damage type.

7. The urban drainage pipe network damage detection system based on intelligent analysis according to claim 6, characterized in that: The method of selecting the damage type corresponding to the maximum matching value as the current pipe network status diagnosis result specifically includes: Establish a queue with the matching degree values ​​arranged in descending order, and select the damage type at the top of the queue as the priority diagnosis result; When the difference between the first two matching values ​​is less than the set tolerance threshold, the manual review mark is triggered and the pending confirmation status is recorded.

8. The urban drainage pipe network damage detection system based on intelligent analysis according to claim 1, characterized in that: Also included are dynamic learning modules, including: After each maintenance operation is completed, obtain the actual damage type and location information confirmed by on-site investigation; Compare and analyze the actual data with the system diagnosis results to calculate the feature recognition error; The parameter weight coefficients in the dynamic detection threshold set are adjusted according to the error amount, and the pattern matching rules in the history maintenance database are updated.

9. The urban drainage pipe network damage detection system based on intelligent analysis according to claim 3, wherein: The method of calculating and generating three-dimensional coordinate positioning data of the damaged area in combination with the pipe network topological characteristic values ​​specifically includes: Based on the pipe segment coding rules in the pipe network topology relationship map, the sensor node location where abnormal data fluctuation occurs is determined; According to the pipe section connection direction and spatial coordinate data, the triangulation positioning method is used to calculate the spatial position of the damaged point relative to the three nearest monitoring nodes.

10. The urban drainage pipe network damage detection system based on intelligent analysis according to claim 8, characterized in that: The parameter weight coefficients in the dynamic detection threshold set are adjusted according to the error amount, and the specific processing process is as follows: Collect statistics on the recognition accuracy of each feature dimension in the last N maintenance operations; Weight attenuation is performed on feature dimensions whose accuracy is lower than the set threshold, and the attenuation amplitude is proportional to the accuracy deviation value; Implement weight enhancement adjustments for feature dimensions whose accuracy is consistently higher than the benchmark value.

Images