RFID-based intelligent inspection method and system for drainage pipe network

Abstract

The invention relates to the technical field of pipe network maintenance, and discloses an RFID-based intelligent inspection method and system for a drainage pipe network, and the method comprises the steps: carrying out the regional division of a target drainage pipe network through a background server, and obtaining drainage pipe network inspection regions; the mobile terminal performs path planning on the drainage pipe network inspection area to obtain a global optimal inspection path; the RFID reader-writer positions the RFID tag based on the global optimal inspection path to obtain the spatial position information of the RFID tag; and the mobile terminal patrols the target drainage pipe network by using the spatial position information of the RFID tag to obtain a drainage pipe network patrolling result. According to the invention, efficient planning of drainage pipe network inspection is realized, and the coverage and real-time performance of inspection work are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of pipe network maintenance, and in particular to an RFID-based intelligent inspection method and system for drainage pipe networks. Background Art

[0002] Urban drainage pipeline networks are an important part of infrastructure, and their maintenance is directly related to urban safety, residents' quality of life, and the sustainable development of the ecological environment. At the same time, the scientific maintenance of urban drainage pipeline networks is the lifeline for ensuring public safety, preventing urban flooding disasters, protecting the ecological environment, maintaining residents' health, and promoting sustainable urban development.

[0003] At present, the inspection of drainage pipelines is carried out through manual inspection, which faces problems such as low efficiency, many missed inspections, and high labor costs. At the same time, faced with the difficulties of a large pipeline system, many and extremely scattered inspection points, a large amount of inspection data, and a low level of intelligent information management, the relevant methods lack efficient inspection planning means and cannot guarantee the coverage and real-time performance of the inspection work. Summary of the Invention

[0004] In view of this, the present invention provides an RFID-based intelligent inspection method and system for drainage pipe networks to solve the problem that related drainage pipe network inspection methods lack efficient inspection planning means and cannot ensure the coverage and real-time performance of inspection work.

[0005] In a first aspect, the present invention provides an RFID-based intelligent inspection system for drainage pipe networks, the system comprising: a backend server, a mobile terminal, an RFID reader / writer, and an RFID tag disposed on a drainage pipe network facility; the RFID tag, the RFID reader / writer, the mobile terminal, and the backend server are connected in sequence; The backend server is used to divide the target drainage network into regions and obtain drainage network inspection areas; Mobile terminals are used to plan routes for inspection areas of the drainage network and obtain the globally optimal inspection route; RFID reader / writer, used to locate the RFID tag based on the global optimal inspection path and obtain the spatial location information of the RFID tag; The mobile terminal is used to inspect the target drainage network using the spatial location information of the RFID tag and obtain the drainage network inspection results.

[0006] An embodiment of the present invention provides an RFID-based intelligent inspection system for drainage pipe networks. Through the seamless connection between RFID readers and mobile terminals, it realizes the rapid collection and synchronous update of inspection data, ensuring the integrity and traceability of the inspection data. With RFID tags as the core data carrier, combined with RFID readers, mobile terminals, background servers, and multiple path planning algorithms, a closed-loop system is formed from inspection area division, path planning and navigation, tag identification and data collection, and inspection result review to maintenance and archiving, thereby improving the intelligence of drainage pipe network inspection.

[0007] In a second aspect, the present invention provides an RFID-based drainage network intelligent inspection method, which is applied to an RFID-based drainage network intelligent inspection system. The method includes: The backend server divides the target drainage network into regions and obtains the drainage network inspection areas; The mobile terminal plans the route for the inspection area of the drainage network and obtains the global optimal inspection route; The RFID reader locates the RFID tag based on the global optimal inspection path and obtains the spatial location information of the RFID tag; The mobile terminal uses the spatial location information of the RFID tag to inspect the target drainage network and obtain the drainage network inspection results.

[0008] The RFID-based intelligent inspection method for drainage pipe networks provided in this embodiment realizes the full-process informatization and automated management of drainage pipe network facility inspections through the deep integration of RFID technology and the intelligent inspection system for drainage pipe networks. The RFID tag is used as the core data carrier, which not only stores the static information (number, type and coordinates) and dynamic data (operating status and detection parameters) of the facility, but also realizes intelligent path planning by combining with the background server, greatly improving the inspection efficiency. The seamless connection between the RFID reader and the mobile terminal realizes the rapid collection and synchronous update of the inspection data, ensuring the integrity and traceability of the data.

[0009] In an optional implementation, the backend server divides the target drainage network into regions to obtain drainage network inspection areas, including: The backend server uses the improved hierarchical clustering algorithm to divide the target drainage network into regions and obtain the initial inspection area; The backend server optimizes and adjusts the boundaries of the initial inspection area to obtain the drainage network inspection area.

[0010] The RFID-based intelligent inspection method for drainage network provided in this embodiment utilizes an improved hierarchical clustering algorithm for spatial partitioning, which can adaptively handle the distribution of facilities with different densities, reflecting the actual accessibility between facilities. It also improves the rationality and accuracy of the drainage network inspection area division by optimizing and adjusting the boundaries of the initial inspection area.

[0011] In an optional implementation, the backend server uses an improved hierarchical clustering algorithm to divide the target drainage network into regions to obtain initial inspection areas, including: The backend server obtains the actual road network distance in the target drainage network, the priority of each facility point, and the historical operation and maintenance data of the facility point; The backend server calculates the composite distance between each facility point based on the actual road network distance, the priority of each facility point, and the historical operation and maintenance data of the facility point; The backend server clusters multiple facility points based on the composite distance between each facility point to obtain multiple clusters; The background server calculates the silhouette coefficients and workload balance corresponding to multiple clusters respectively, and compares the silhouette coefficients and workload balance with the iteration termination condition; If both the silhouette coefficient and the workload balance meet the iteration termination conditions, the background server divides the target drainage network based on multiple clusters to obtain the initial inspection area.

[0012] In the RFID-based intelligent inspection method for drainage pipe networks provided in this embodiment, the backend server calculates the composite distance between each facility point based on the actual road network distance, the priority of each facility point and the historical operation and maintenance data of the facility point, and clusters multiple facility points based on the composite distance between each facility point. In the clustering process, the influence of the actual road network distance, the priority of each facility point and the historical operation and maintenance data of the facility point on the area division is comprehensively considered, thereby improving the adaptability of the initial inspection area; by comparing the silhouette coefficient and the workload balance with the iteration termination condition, the influence of the workload on the area division is considered in the process of initial inspection area division, thereby improving the rationality of the initial inspection area setting.

[0013] In an optional embodiment, the mobile terminal performs path planning for the drainage network inspection area to obtain a global optimal inspection path, including: The mobile terminal performs global path planning for the drainage network inspection area and obtains a global access sequence; The mobile terminal uses a heuristic search algorithm to perform local path planning based on the global access sequence and obtains the optimal local path; Based on the global access sequence, the mobile terminal splices the optimal local paths to obtain the global optimal inspection path.

[0014] The RFID-based intelligent inspection method for drainage pipe networks provided in this embodiment performs global path planning for the inspection area of the drainage pipe network through a mobile terminal, obtains a global access sequence, ensures the rationality of the global optimal inspection path, utilizes a heuristic search algorithm for local path planning, improves the efficiency of local path planning, realizes the rational planning of the optimal local path, and splices the optimal local paths based on the global access sequence, which can not only obtain the approximate globally optimal inspection sequence, but also quickly adapt to dynamic road condition changes, flexibly plan local paths, and realize efficient multi-task point inspection.

[0015] In an optional implementation, the mobile terminal performs global path planning on the drainage network inspection area to obtain a global access sequence, including: The mobile terminal obtains the historical pheromone concentration and heuristic information of the path between the starting facility point and other facility points in the drainage network inspection area; The mobile terminal selects the next facility point based on the historical pheromone concentration and heuristic information of the path between the starting facility point and other facilities, and constructs the initial inspection path based on the starting facility point and the next facility point; The mobile terminal obtains performance indicator data of the initial patrol path, and incrementally updates the pheromone concentration of the initial patrol path based on the performance indicator data of the initial patrol path; The mobile terminal updates the initial patrol path based on the pheromone concentration of the incrementally updated initial patrol path until the pheromone concentration corresponding to the updated patrol path meets a preset condition, and then determines a global access sequence based on the updated patrol path.

[0016] The RFID-based intelligent inspection method for drainage pipe networks provided in this embodiment realizes the rational planning of the initial inspection path by selecting the next facility point based on the historical pheromone concentration and heuristic information of the path between the starting facility point and other facility points through a mobile terminal; by incrementally updating the pheromone concentration of the initial inspection path and determining the global access sequence based on the pheromone concentration of the initial inspection path after the incremental update, the problem of local optimality is avoided and the rationality of the global access sequence is ensured.

[0017] In an optional implementation, the mobile terminal performs local path planning based on the global access sequence using a heuristic search algorithm to obtain an optimal local path, including: The mobile terminal obtains the coordinates of adjacent facility points, the traffic adjustment coefficient and the road section congestion index in the global access sequence, and establishes a heuristic function based on the coordinates of the adjacent facility points, the traffic adjustment coefficient and the road section congestion index; The mobile terminal obtains the multi-dimensional factors of the actual path between adjacent facility points in the global access sequence, and determines the actual cost based on the multi-dimensional factors of the actual path between adjacent facility points; The mobile terminal calculates a cost function based on the heuristic function and the actual cost, and determines an optimal local path based on the cost function.

[0018] The RFID-based intelligent inspection method for drainage pipe networks provided in this embodiment calculates a cost function based on a heuristic function and actual cost, and determines the optimal local path based on the cost function, so that the optimal local path can quickly adapt to dynamic road condition changes, thereby improving the rationality of optimal local path planning.

[0019] In an optional embodiment, the RFID reader locates the RFID tag based on the global optimal inspection path to obtain the spatial location information of the RFID tag, including: The RFID reader receives the RFID tag navigation coordinates sent by the mobile terminal, transmits a reference signal to the RFID tag based on the RFID tag navigation coordinates, and receives multiple reception signals returned by the RFID tag; The RFID reader determines a plurality of phase differences based on the reference signal and the received signal; The RFID reader obtains an operating wavelength of the reader antenna and calculates a communication propagation distance between the reader antenna and the RFID tag based on the operating wavelength of the reader antenna and a phase difference between a reference signal and a received signal; wherein the RFID reader includes a reader antenna; The RFID reader obtains the coordinates of the reader antenna and determines the spatial location information of the RFID tag based on the coordinates of the reader antenna and the communication propagation distance between the reader antenna and the RFID tag.

[0020] The RFID-based intelligent inspection method for drainage pipe networks provided in this embodiment is based on the RFID tag navigation coordinates sent by the mobile terminal and utilizes the built-in phase detection technology of the RFID reader to accurately obtain the spatial position information of the RFID tag, achieving sub-meter positioning accuracy, and further guiding inspection personnel to accurately find the location of the facilities. The dual positioning method combining the RFID tag navigation coordinates and the RFID phase detection technology greatly improves the efficiency of finding hidden facilities such as underground manhole covers.

[0021] In an optional implementation, the backend server reviews the drainage network inspection results to obtain the drainage network inspection review results.

[0022] The RFID-based intelligent inspection method for drainage network provided in this embodiment reviews the inspection results of the drainage network through the background server to obtain the inspection and review results of the drainage network, thereby avoiding the problem of possible deviations in manual inspections and greatly improving the efficiency of subsequent verification.

[0023] In an optional implementation, the backend server performs data analysis on the inspection results of the drainage network and generates a drainage network maintenance plan based on the data analysis results.

[0024] The RFID-based intelligent inspection method for drainage pipe networks provided in this embodiment performs data analysis on the inspection results of the drainage pipe networks through a background server, and generates a drainage pipe network maintenance plan based on the data analysis results, providing scientific decision-making support for the maintenance of drainage pipe network facilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0026] Figure 1 1 is a structural block diagram of an RFID-based intelligent inspection system for drainage pipe networks according to an embodiment of the present invention; Figure 2 1 is a schematic diagram of the workflow of the RFID-based intelligent inspection system for drainage pipe networks according to an embodiment of the present invention; Figure 3 1 is a flow chart of an RFID-based intelligent inspection method for a drainage network according to an embodiment of the present invention (I); Figure 4 1 is a flow chart of an RFID-based intelligent inspection of a drainage network according to an embodiment of the present invention; Figure 5 2 is a flow chart of an RFID-based intelligent inspection method for a drainage network according to an embodiment of the present invention; Figure 6 1 is a flow chart of an RFID-based intelligent inspection method for a drainage network according to an embodiment of the present invention (III); Figure 7 1 is a flow chart of an RFID-based intelligent inspection method for a drainage network according to an embodiment of the present invention (IV); Figure 8 1 is a flow chart of an RFID-based intelligent inspection method for a drainage network according to an embodiment of the present invention (V). DETAILED DESCRIPTION

[0027] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are 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 those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0028] The embodiment of the present invention provides an RFID-based intelligent inspection system for drainage pipe networks. To address the problem that inspection route planning relies on subjective experience, is inefficient, and lacks scientificity, the system stores geographic information based on RFID tags and combines multiple intelligent algorithms to achieve scientific and intelligent planning of inspection routes, significantly improving the efficiency and standardization of inspection work. Secondly, to address the problem that manual inspections are "difficult to find points" and prone to "jump points" and "missed inspections", the system uses RFID tag phase recognition and mobile terminal coupling precise positioning technology to achieve precise positioning and navigation of pipe network inspection points. Furthermore, to address the problem that inspection record information has a large amount of data and is difficult to input and manage, the system achieves standardized, digital recording, and real-time updating and maintenance of pipe network inspection information through information attribute classification management and intelligent reading and recognition of RFID tags, and establishes a complete intelligent management system for pipe network facility inspection archives. Finally, to address the problem that manual inspections may have deviations, the system implements intelligent verification of manual inspection information based on multiple deep learning algorithms, greatly improving the efficiency of later verification. The above steps fundamentally change the limitations of the inspection model and establish an intelligent and information-based inspection management system.

[0029] This embodiment provides an RFID-based intelligent inspection system for drainage pipe networks. Figure 1 As shown in the figure, the system includes an RFID-based intelligent inspection device for drainage pipe network, such as Figure 1 As shown, the RFID-based intelligent inspection device for drainage network includes: a background server 101, a mobile terminal 102, an RFID reader 103 and an RFID tag 104 set on the drainage network facility; the RFID tag 104, the RFID reader 103, the mobile terminal 102 and the background server 101 are connected in sequence.

[0030] The backend server 101 is used to divide the target drainage network into regions to obtain drainage network inspection areas.

[0031] Specifically, the backend server 101 is responsible for regional division of the target drainage network, which requires comprehensive consideration of multiple factors, the use of scientific and intelligent methods for division, and the establishment of a dynamic optimization mechanism; the backend server builds a comprehensive data analysis model, integrating key information in multiple dimensions, including: basic geographic data, including: static information such as facility spatial distribution, road network structure and traffic flow; business priority data, including: key network nodes, fault-frequent areas and complaint hotspots, etc. areas that require special attention; historical operation and maintenance data, including: fault records, maintenance statistics, inspection feedback and other historical experiences; through data standardization processing, these heterogeneous information are converted into quantifiable evaluation indicators, providing a data basis for scientific division.

[0032] Furthermore, the backend server 101 plays a key role in data processing and distribution. It obtains basic data in the drainage network (including static data such as facility number, type and location coordinates, dynamic monitoring data such as strain and temperature, and multimedia data such as inspection records, photos and voice) from RFID (Radio Frequency Identification) tags 104 and mobile terminals 102, standardizes, quality-checks and classifies and stores these multiple types of basic data to generate a structured data set.

[0033] Furthermore, the processed basic data is sent to the path planning module for optimizing inspection routes, to the AI (Artificial Intelligence) review module for intelligent review, and to the maintenance decision module for generating maintenance plans.

[0034] Furthermore, in terms of algorithm models, the intelligent processing module of the background server integrates a variety of advanced algorithms: in terms of inspection path planning, it combines the ant colony algorithm with The algorithm (a heuristic search algorithm) can dynamically optimize inspection routes based on real-time traffic information and provide the best navigation solution. The AI review module uses a deep learning model and, through image recognition and natural language processing technology, achieves intelligent matching and anomaly detection between inspection images and text descriptions, greatly improving review efficiency. In terms of maintenance strategy analysis, based on machine learning algorithms, it comprehensively analyzes inspection data and historical maintenance records to provide scientific decision-making recommendations for facility maintenance.

[0035] Furthermore, in terms of interaction and security control, the backend server 101 adopts a multi-layer security protection architecture, ensuring the security of data transmission through HTTPS (Hypertext Transfer Protocol Secure) and JWT (JSON WebToken, an open standard) token mechanism; the backend server 101 supports the reliable transmission of multiple types of data such as pictures, text, audio and coordinate trajectories, and implements role-based access control, providing differentiated permission management for different users such as inspectors, auditors and project leaders; at the same time, the integrated instant messaging module supports real-time communication with the dispatch center, ensuring timely information transmission and rapid response to problems during the inspection process.

[0036] Furthermore, the backend server 101 achieves a virtuous cycle of development by establishing a unified data archiving system and continuous optimization mechanism, and establishes a three-tier data architecture to ensure data integrity and traceability; the first tier is the business data layer, which stores basic data such as drainage network inspection results and maintenance implementation records according to business types; the second tier is the model data layer, which stores training data sets, verification results and model version information of various algorithm models; the third tier is the knowledge graph layer, which constructs a facility-fault-solution associated knowledge network to support intelligent decision-making.

[0037] Furthermore, the backend server 101 has built a multi-dimensional evaluation system to evaluate the quality of inspection work based on indicators such as GPS (Global Positioning System) trajectory, operation timing and data integrity; track the changes in facility status after maintenance and evaluate the actual effect of maintenance plans; monitor the accuracy, response time and other technical indicators of various algorithm models to evaluate performance.

[0038] Furthermore, the backend server 101 conducts iterative optimization based on the evaluation results, and dynamically adjusts the inspection cycle and inspection point density according to data such as the actual facility failure frequency and regional characteristics; based on the actual application effect and combined with the newly added annotation data, it continuously optimizes the recognition accuracy and processing efficiency of various AI models; and promptly adds newly discovered typical fault cases and effective handling methods to the expert knowledge base. The backend server 101 pushes the optimized parameters and model updates to the mobile terminal 102 and the RFID reader 103.

[0039] Furthermore, to address the problems of large amounts of inspection record information and difficulty in data entry and management, the data backend in the backend server adopts a distributed database architecture, and realizes efficient data management and access through a three-tier data storage structure: 1) Business data layer, which stores basic data such as original inspection records and maintenance implementation records by business type to ensure the security and integrity of core data; 2) Model data layer, which stores training data sets, verification results and model version information of various algorithm models, and supports high-concurrency access and real-time data updates; 3) Knowledge graph layer, which builds a facility-fault-solution association knowledge network for algorithm analysis and result visualization, providing a data foundation for decision support. The backend server also integrates a local SQLite (a database) database as an offline storage solution for mobile terminals, and cooperates with breakpoint resumption and incremental update mechanisms to ensure that data synchronization can be completed efficiently after network recovery.

[0040] Furthermore, the key data stored in the background server 101 is the inspection content of the facilities, and the inspection content of each type of facility is different; the inspection content of pipeline and channel box facilities includes whether the road above the pipeline has collapsed, whether there is illegal occupation, whether there is unauthorized takeover, inspection of construction sites and surrounding drainage facilities, etc.

[0041] Furthermore, the inspection contents of the inspection wells include: whether the manhole cover is missing, whether sewage is overflowing, whether the manhole frame cover is deformed or damaged, whether the manhole cover is buried or illegally occupied, whether the manhole cover is displaced, the height difference between the manhole cover and the manhole frame, whether the gap exceeds the limit, whether there are protrusions or depressions between the manhole cover and the manhole frame, whether there is jumping or noise between the manhole cover and the manhole frame, whether the manhole cover markings are incorrect, whether the road around the manhole cover is under construction, etc.; the inspection contents of the rainwater outlet include whether there is water accumulation on the road surface, whether the rainwater grate is lost or damaged, whether the rainwater grate is filled or illegally occupied, whether the rainwater outlet frame is damaged, whether the height difference and gap between the cover frame exceed the limit, whether the rainwater grate holes are blocked, whether the rainwater outlet frame is protruding, depressed or jumping, whether the rainwater grate is opened or displaced and whether it emits odor, etc.

[0042] Furthermore, the inspection contents of open channels include: whether the slope protection, retaining walls and capping of the channel banks made of blocks of stone and concrete blocks have cracks, subsidence, tilting, defects, weathering and grouting detachment, whether the open channel ancillary facilities such as guardrails, mileage posts, warning signs and trails are intact, and whether there is sewage overflow or backflow in the sewage pipe network within the control range of the open channel.

[0043] Furthermore, the inspection contents of the ditch include: whether the cover plate is damaged, whether the wall structure has cracks, subsidence, tilt, defects, weathering and joint detachment, whether there are abnormalities in water level and flow, whether there is siltation, whether there is illegal occupation and planting along the channel, whether the slope is unstable and whether there is any unauthorized takeover.

[0044] The mobile terminal 102 is used to plan the route for the inspection area of the drainage network to obtain the global optimal inspection route.

[0045] Specifically, the mobile terminal 102 adopts equipment that meets industrial-grade protection standards, is equipped with a reinforced anti-fall shell and a high-capacity battery. The interface of the mobile terminal 102 provides an ergonomic interactive experience, and the integrated navigation software SDK (Software Development Kit) supports offline map use. It is also equipped with intelligent auxiliary functions such as OCR (Optical Character Recognition) text recognition and AR (Augmented Reality) facility labeling, which comprehensively improves the efficiency of inspection operations.

[0046] Furthermore, the mobile terminal 102 serves as a unified mobile operation platform for inspection personnel, providing digital support for the entire process from task receipt to completion submission. The mobile terminal 102 ensures data security through a strict identity authentication mechanism and assigns personalized task lists to inspection personnel. During the inspection process, the mobile terminal 102 not only supports standardized data collection and recording, but also provides intelligent auxiliary functions such as voice input, OCR text recognition and AR facility labeling, which significantly improve work efficiency. Taking into account the field working environment, the mobile terminal is designed with a complete offline working mode to ensure that all basic inspection operations can still be performed even when the network is unstable.

[0047] Furthermore, at the technical implementation level, the mobile terminal 102 adopts a series of advanced solutions to provide a smooth user experience: the interface design incorporates interactive logic that conforms to ergonomics, including details such as dark mode adaptation, and the navigation function is developed based on the navigation software SDK, which can not only obtain real-time traffic conditions, bus and subway information, but also combine the background JPS (Jump Point Search) algorithm and The algorithm implements intelligent path planning and supports offline maps to cope with network-free environments. The data collection module integrates multiple collection methods, including automatic reading of RFID signals, manual entry, voice notes, and on-site multimedia recording. All collected data will automatically be attached with GPS geotags and timestamps to ensure authenticity and match the corresponding RFID tag 104; in addition, the mobile terminal 102 also provides standardized inspection forms, supports dynamic generation of inspection items according to different facility types, and realizes standardized recording of the inspection process.

[0048] Furthermore, the data interaction system of the mobile terminal 102 adopts a two-way communication architecture. On the one hand, it establishes a stable connection with the RFID reader 103 through near-field communication technologies such as Bluetooth 5.0 and USB Type-C (a hardware interface form of a universal serial bus), obtains basic facility information (including static data such as facility number, facility type and installation time) and monitoring data (including dynamic data such as strain and temperature) in the drainage network, standardizes and inspects the above data, generates structured inspection records, and sends these records to the background server 101 for further analysis.

[0049] Furthermore, the mobile terminal 102 establishes a secure data transmission channel with the back-end server through the HTTPS protocol and the JWT token mechanism, realizing the reliable transmission of multiple types of data including pictures, text, audio and coordinate tracks. The mobile terminal 102 uses the local SQLite database to provide offline storage capabilities, and cooperates with the breakpoint resumption and incremental update mechanisms to ensure that data synchronization can be completed efficiently after the network is restored. At the same time, the integrated instant messaging module supports real-time communication with the dispatch center, facilitating timely handling of various emergencies encountered during the inspection process.

[0050] The RFID reader 103 is used to locate the RFID tag 104 based on the global optimal inspection path to obtain the spatial position information of the RFID tag 104.

[0051] Specifically, the RFID tag 104 adopts an industrial-grade IP67 (Ingress Protection Rating 67, a level of protection safety) waterproof and dustproof casing design, which is specially adapted to the high humidity and corrosion risks in the drainage network environment. The interior of the RFID tag 104 is treated with a special anti-corrosion coating to withstand corrosive gases such as hydrogen sulfide in underground pipes for a long time. The casing material is made of high-strength engineering plastic that can withstand external pressure and impact, and a special sealing process is used to ensure the long-term reliable operation of the internal circuit board.

[0052] Furthermore, RFID tags 104 need to be installed or pre-buried in various facilities of the drainage network, including but not limited to pipelines, inspection wells, rainwater inlets, channel boxes, open channels and ditches. RFID tags 104 serve as the core data collection and storage unit of the drainage network inspection device. They obtain real-time status data of the facilities through built-in sensors, including key parameters such as strain, temperature and humidity. The background storage information structure of RFID tags 104 of different facility types needs to be designed differently, including static data and dynamic data fields.

[0053] Furthermore, static data such as facility number, facility type, new construction or installation time and coordinate location, which are long-term unchanged or rarely changed information, and dynamic data such as monitoring data of the strain gauge strain and temperature sensor in the RFID tag 104, are stored in the memory of the RFID tag 104 in a certain data format.

[0054] Furthermore, dynamic data such as the current operating status of the facility, facility-related conditions, the most recent inspection time, abnormal records, and maintenance logs, which need to be confirmed on-site and updated in the background server, are stored in a certain data format in the background server. Each time an RFID tag is identified, the corresponding background data can be retrieved. After the above data is read and preliminarily processed by the RFID reader 103, it will be transmitted to the mobile terminal 102 for further analysis and display.

[0055] Furthermore, as an important carrier of spatial information, the RFID tag 104 stores facility coordinates in the standard WGS84 (World Geodetic System 1984) geographic coding system, which enables the RFID tag 104 data to be seamlessly connected to the GIS (Geographic Information System) platform. The RFID tag 104 uses a built-in high-precision positioning module to collect and update the precise geographic coordinates of the facility in real time, providing basic data support for the spatial analysis function. On the background server 101, the RFID tag 104 coordinates are uniformly managed through spatial databases such as PostGIS (a spatial database), which not only supports efficient spatial index queries but also realizes accurate geographic fencing functions.

[0056] Furthermore, based on the powerful spatial analysis capabilities of the GIS platform, the backend server 101 can comprehensively consider multiple factors such as facility density distribution and road network topology to scientifically divide inspection areas. At the same time, the GIS network analysis module can combine real-time road conditions information and traffic control constraints to provide more accurate distance and time cost estimates for inspection route planning. Through WebGIS (World Wide Web Geographic Information System) technology, managers can intuitively view the spatial distribution of RFID tags 104, support multi-level dynamic scaling and attribute information query, thereby fully understanding the distribution status of facilities within the jurisdiction and providing strong support for inspection management decisions.

[0057] Furthermore, the RFID reader / writer 103 integrates a high-gain antenna array, which significantly improves reading stability through circular polarization design, and can achieve concurrent reading and writing of multiple RFID tags 104 at a long distance. The core control adopts a high-performance MCU (Microcontroller Unit) processor, supports multiple ultra-high frequency RFID communication protocols including EPCglobal UHF (an ultra-high frequency RFID communication protocol), and is equipped with phase detection-based signal processing technology to achieve high-precision RFID tag 104 positioning and anti-collision.

[0058] Furthermore, the RFID reader 103 adopts a dual-channel communication architecture and integrates Bluetooth 5.0 and USB Type-C interfaces to achieve flexible connection with the mobile terminal 102. Bluetooth communication supports real-time two-way data transmission and is suitable for wireless inspection scenarios. The USB Type-C interface provides a stable wired connection for rapid transmission of large amounts of data.

[0059] Furthermore, the RFID reader 103 uses two communication interfaces, Bluetooth 5.0 and USB Type-C, to interact with the mobile terminal for data. Bluetooth communication supports real-time two-way data transmission and is suitable for wireless inspection scenarios. The USB Type-C interface provides a stable wired connection method and can be used for rapid transmission of large amounts of data. The RFID reader is designed with a communication exception handling mechanism. When the Bluetooth connection is interrupted, it will automatically switch to USB communication to ensure reliable collection and transmission of inspection data.

[0060] Furthermore, the RFID reader 103 operates based on the electromagnetic induction or electromagnetic coupling principle of radio frequency identification technology, and uses the ultra-high frequency band to achieve wireless communication with the RFID tag 104. The RFID reader 103 has an integrated high-gain antenna array, and the circular polarization design significantly improves the reading stability, which can realize concurrent reading and writing operations of multiple tags within a long range. The core control uses a high-performance MCU processor, supports multiple ultra-high frequency RFID communication protocols including EPCglobal UHF, and is equipped with an advanced anti-collision mechanism.

[0061] Furthermore, the RFID reader 103 performs a detailed scan of the RFID tag 104 in the facility within a safe distance, reading basic information including the facility number, type, and maintenance record. The backend server 101 automatically retrieves the historical inspection data and precautions of the facility, and matches the corresponding standardized inspection form according to the facility type. All collected data will be automatically associated with the RFID tag 104 to avoid data storage errors.

[0062] The mobile terminal 102 is used to inspect the target drainage network using the spatial location information of the RFID tag 104 to obtain the drainage network inspection result.

[0063] Specifically, the mobile terminal 102 adopts a local priority storage strategy for data management. All data (including basic inspection records, key parameters, high-definition pictures and videos, etc.) are stored locally first and the original backup is performed; after the inspection task is completed, the mobile terminal 102 will prompt the user to confirm whether the data needs to be uploaded. When the user confirms to upload, the mobile terminal 102 will use an intelligent compression algorithm to compress large-capacity files and upload all data to the cloud in batches. The above-mentioned local priority storage strategy not only saves mobile data traffic, but also avoids data transmission problems caused by network instability. When abnormal data is found, the mobile terminal 102 will promptly remind the inspection personnel to review or supplement the collection.

[0064] The RFID-based intelligent inspection system for drainage pipe networks provided in this embodiment realizes the rapid collection and synchronous update of inspection data through the seamless connection between RFID readers and mobile terminals, ensuring the integrity and traceability of inspection data; with RFID tags as the core data carrier, combined with RFID readers, mobile terminals, background servers, and multiple path planning algorithms, a closed-loop system is formed from inspection area division, path planning and navigation, tag identification and data collection, and inspection result review to maintenance and archiving, thereby improving the intelligence of drainage pipe network inspection.

[0065] The following is a specific example to illustrate the workflow of the RFID-based intelligent inspection system for drainage pipe networks.

[0066] Example 1: With RFID tags as the core data carrier, combined with RFID readers, mobile terminals, backend servers and databases, as well as multiple path planning algorithms and artificial intelligence audit models, a closed-loop system is formed from inspection area division, path planning and navigation, tag identification and data collection, inspection result review to maintenance and archiving.

[0067] like Figure 2 As shown in the figure, through scientific area division, intelligent route planning, standardized execution process, strict data review and continuous optimization and improvement, the drainage network inspection work can be carried out efficiently. The workflow of the RFID-based drainage network intelligent inspection system is mainly divided into six parts: inspection area division, multi-task point inspection route planning, inspection task execution, data review and analysis, maintenance plan generation, data archiving and continuous optimization, including: RFID tags store static data such as facility number and facility coordinate location through memory chips, and collect dynamic monitoring data such as strain and temperature through built-in sensors to form the original tag data set.

[0068] The RFID reader uses phase detection technology to obtain the tag's spatial coordinate data, and adopts a multi-tag anti-collision algorithm (to prevent multiple tag data from interfering with each other) to parse and clean the original data. At the same time, it retrieves related data such as facility maintenance logs from the background database to generate an enhanced facility data packet with spatial attributes; the mobile terminal APP receives the enhanced data packet through the Bluetooth 5.0 / USB Type-C interface, combines GPS positioning data with the multimedia acquisition module, and generates a structured inspection record containing time and space stamps, on-site photos, and voice notes, which is encrypted and transmitted to the background server via the HTTPS protocol.

[0069] The backend server integrates the navigation path data generated by the path planning algorithm, the real-time inspection data uploaded by the mobile terminal, and the historical maintenance records through the PostGIS spatial database, and uses multi-source data fusion technology to construct a three-dimensional decision-making data set.

[0070] The AI audit model uses an improved YOLOv8 (You Only Look Once version 8, a target detection network) target detection algorithm to identify defects in on-site photos, combines natural language processing technology to analyze voice comments, and outputs a graded audit report. Ultimately, the approved inspection data and maintenance plans are written to the business data layer of the distributed database, while updating the facility status relationship network in the knowledge graph layer to form a traceable maintenance archive record.

[0071] Through data connection and collaborative processing between the above modules, combined with intelligent IoT devices, multi-source data fusion, AI algorithm optimization and process automation, the inspection efficiency of the drainage network and the quality of maintenance decisions have been significantly improved.

[0072] According to an embodiment of the present invention, an embodiment of an RFID-based intelligent inspection method for a drainage network is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0073] In this embodiment, a drainage network intelligent inspection method based on RFID is provided, which can be used in the above-mentioned drainage network intelligent inspection system based on RFID. Figure 3 FIG. 1 is a flow chart of an RFID-based intelligent inspection method for a drainage network according to an embodiment of the present invention. Figure 3 As shown, the process includes the following steps: In step S301, the backend server divides the target drainage network into regions to obtain drainage network inspection areas.

[0074] Specifically, the method for dividing the drainage network inspection areas adopts the strategy of "double-layer optimization + manual verification".

[0075] In step S302, the mobile terminal performs path planning for the inspection area of the drainage network to obtain a global optimal inspection path.

[0076] Specifically, after the drainage network inspection area is divided, the mobile terminal receives the inspection area data. To address the problems of strong subjectivity and low efficiency in inspection route planning, an intelligent path planning method based on a multi-level algorithm is proposed. First, the input inspection area data is comprehensively processed, and all RFID tag points in the area are spatially indexed. Then, road condition information provided by the navigation software API (Application Programming Interface) is obtained in real time, including road congestion, construction closures, and other status. At the same time, a complete path planning constraint model is constructed by combining business constraints such as point priority and timeliness requirements.

[0077] In step S303, the RFID reader locates the RFID tag based on the global optimal inspection path to obtain spatial location information of the RFID tag.

[0078] Specifically, the inspection personnel first go to the approximate area of the target facility through the global optimal inspection route sent by the mobile terminal. After arriving near the facility, they use the RFID reader to scan the surrounding RFID tags. The RFID reader uses built-in phase detection technology to perform triangulation positioning calculations based on the phase difference between the received signal and the reference signal, and can accurately obtain the spatial location information of the RFID tag.

[0079] Furthermore, after reaching the navigation coordinates of the RFID tag, the RFID reader scans the surrounding area. The RFID reader uses built-in phase detection technology to perform triangulation positioning calculations based on the phase difference between the received signal and the reference signal, and can accurately obtain the spatial position information of the RFID tag.

[0080] In step S304, the mobile terminal inspects the target drainage network using the spatial location information of the RFID tag to obtain the drainage network inspection result.

[0081] Specifically, when the RFID tag signal of the target facility is read, the RFID reader will automatically synchronize the calculated precise three-dimensional coordinates to the mobile terminal, achieving sub-meter positioning, and further guiding inspection personnel to accurately find the location of the RFID tag (i.e. the location of the facility).

[0082] Furthermore, if Figure 4As shown, after the inspection personnel carry the RFID reader into the target location, they read the static and dynamic data in the RFID tag; if problems are found or manual filling is required (such as the damage level of the manhole cover, whether it is illegally occupied, etc.), the new information can be written back to the tag or cached to the mobile terminal, and synchronized to the background server when the network environment permits; the facility location is found based on the spatial location of the RFID tag. After arriving at the facility, the RFID reader is used to perform a detailed scan of the RFID tag corresponding to the facility within a safe distance, reading basic information including the facility number, type, maintenance record, etc. The background server automatically retrieves the historical inspection data and precautions of the facility, and matches the corresponding standardized inspection form according to the facility type. All collected data will be automatically associated with the RFID tag to avoid data storage errors.

[0083] Furthermore, after the facilities to be inspected are found, the inspection work is recorded in a standardized multimedia format through the mobile terminal. The mobile terminal has a built-in professional inspection and photography module that supports both panoramic and close-up modes. It automatically verifies the lighting, clarity, and composition integrity of the photos, and adds geographic location and time watermarks. The photos taken above will automatically be associated with the currently scanned RFID tag to ensure that each photo can accurately correspond to a specific facility. After the shooting is completed, the mobile terminal interface automatically switches to a structured form interface. The form is divided into four areas: basic information area (automatically associated with facility information read from RFID tags), status check area (dynamically loads 20-30 standard inspection items according to facility type), abnormal record area (supports handwriting input or voice recognition), and disposal suggestion area.

[0084] Furthermore, the mobile terminal also provides a convenient voice note function, allowing inspectors to record 15-30 seconds of on-site audio instructions with one click. The audio recording will also be automatically matched with the RFID tag to form a complete multimedia inspection record. The mobile terminal will send the collected inspection data to the background server via the HTTPS protocol for review.

[0085] Furthermore, for different types of facilities, the mobile terminal will automatically match the corresponding inspection checklist: pipeline and channel box categories include 10 indicators such as pipeline damage, cracks, misalignment, leakage, pipeline siltation depth, and illegal buildings around the pipeline; inspection wells focus on checking 12 indicators such as manhole cover integrity, anti-fall net status, well shaft structure, ladder firmness, and well bottom siltation; rainwater inlets check 8 indicators such as water inlet grates, water collection wells, connection pipe integrity, drainage smoothness, and surrounding road surface waterlogging; open channels and side ditches include 15 indicators such as channel wall structure, slope protection status, cross-sectional dimensions, water flow smoothness, and illegal discharge.

[0086] Furthermore, the mobile terminal uses a four-level assessment system (A-normal, B-minor problems, C-obvious defects and D-serious failures) to automatically calculate the status level of each facility. When the assessment result is C or D, the mobile terminal will force the supplementation of detailed photos and text descriptions; all collected data will undergo real-time quality control, including photo clarity detection, recording quality assessment, and required item completeness verification. For unqualified records, there will be an immediate prompt to re-collect.

[0087] Furthermore, the mobile terminal adopts a local priority storage strategy for data management. All data (including basic inspection records, key parameters, high-definition pictures, videos, etc.) are saved locally first and the original backup is performed. After the inspection task is completed, the mobile terminal will prompt the user to confirm whether the data needs to be uploaded. When the user confirms to upload, the mobile terminal will use an intelligent compression algorithm to compress large-capacity files and upload all data to the cloud in batches. The local priority storage strategy not only saves mobile data traffic, but also avoids data transmission problems caused by network instability. When abnormal data is found, the mobile terminal will promptly remind the inspection personnel to review or supplement the collection.

[0088] The RFID-based intelligent inspection method for drainage pipe networks provided in this embodiment realizes the full-process informatization and automated management of drainage pipe network facility inspections through the deep integration of RFID technology and the RFID-based intelligent inspection system for drainage pipe networks. The RFID tag is used as the core data carrier, which not only stores the static information (number, type and coordinates) and dynamic data (operating status and detection parameters) of the facility, but also realizes intelligent path planning by combining with the background server, greatly improving the inspection efficiency. The seamless connection between the RFID reader and the mobile terminal realizes the rapid collection and synchronous update of the inspection data, ensuring the integrity and traceability of the data.

[0089] In this embodiment, a drainage network intelligent inspection method based on RFID is provided, which can be used in the above-mentioned drainage network intelligent inspection system based on RFID. Figure 5 FIG. 1 is a flow chart of an RFID-based intelligent inspection method for a drainage network according to an embodiment of the present invention. Figure 5 As shown, the process includes the following steps: In step S501, the backend server divides the target drainage network into regions to obtain drainage network inspection areas.

[0090] Specifically, the above step S501 includes: In step S5011, the backend server divides the target drainage network into regions using the improved hierarchical clustering algorithm to obtain initial inspection areas.

[0091] Specifically, an improved hierarchical clustering algorithm is used for spatial partitioning. The improved hierarchical clustering algorithm can adaptively handle the distribution of facilities with different densities and reflect the actual accessibility between facilities through the distance matrix.

[0092] In some optional implementations, the above step S5011 includes: In step a1, the backend server obtains the actual road network distance, the priority of each facility point, and the historical operation and maintenance data of the facility point in the target drainage network.

[0093] In step a2, the backend server calculates the composite distance between each facility point based on the actual road network distance, the priority of each facility point, and the historical operation and maintenance data of the facility point.

[0094] Specifically, the composite distance between facilities The calculation formula is as follows: (1) in, is the actual road network distance, and Represents facilities and facilities Priority, Represents facilities and facilities Historical operation and maintenance data of the distance between them.

[0095] In step a3, the backend server clusters the multiple facility points based on the composite distances between the facility points to obtain multiple clusters.

[0096] Specifically, during the iterative merging process, the background server selects two facility points with the smallest composite distance to merge, thereby obtaining multiple clusters.

[0097] In step a4, the background server calculates the silhouette coefficients and workload balance corresponding to the multiple clusters respectively, and compares the silhouette coefficients and workload balance with the iteration termination condition.

[0098] Specifically, the silhouette coefficient is used to measure the density within a cluster. The calculation formula is as follows: (2) in, For facilities The average distance to other facilities in the same cluster, For facilities The minimum average distance to the nearest facility in another cluster, Indicates the number of facilities within the cluster.

[0099] Furthermore, workload balance The calculation formula is as follows: (3) in, For the The workload of the area, It is obtained by calculating the sum of the priorities of all facilities in the cluster.

[0100] In step a5, if the silhouette coefficient and the workload balance both meet the iteration termination conditions, the backend server divides the target drainage network based on multiple clusters to obtain initial inspection areas.

[0101] Specifically, when the silhouette coefficient >0.7 and workload balance When <0.15, the iterative merging is terminated and the initial inspection area is obtained.

[0102] In step S5012, the backend server optimizes and adjusts the boundaries of the initial inspection area to obtain the drainage network inspection area.

[0103] Specifically, by setting multiple constraints such as workload balance, route rationality and facility integrity, the boundaries of the initial inspection area are optimized and adjusted.

[0104] Furthermore, experienced managers conduct on-site assessments and fine-tuning to ensure that the inspection workload of each area can be completed scientifically within the specified time (usually an 8-hour work system). The backend server will send the divided drainage network inspection areas to the mobile terminal.

[0105] Furthermore, a data-driven continuous optimization system is established, and the backend server regularly collects the actual operation data of each drainage network inspection area, including key indicators such as inspection completion rate, time consumption, distance statistics and problem discovery rate; through time series analysis, problems such as uneven workload distribution and efficiency anomalies are identified; when it is found that certain drainage network inspection areas are under obvious operational pressure, the local optimization process is initiated. At the same time, the inspection area division is flexibly adjusted according to temporary factors such as seasonal changes (such as key areas during flood season), project information and traffic control to ensure the continuous and efficient implementation of the inspection work.

[0106] Step S502: The mobile terminal performs route planning for the inspection area of the drainage network to obtain the global optimal inspection route. Figure 3 Step S302 of the illustrated embodiment will not be described in detail here.

[0107] In step S503, the RFID reader locates the RFID tag based on the global optimal inspection path to obtain the spatial location information of the RFID tag. Figure 3 Step S303 of the illustrated embodiment will not be described in detail here.

[0108] Step S504: The mobile terminal inspects the target drainage network using the spatial location information of the RFID tag to obtain the inspection results of the drainage network. Figure 3 Step S304 of the illustrated embodiment will not be described in detail here.

[0109] The RFID-based intelligent inspection method for drainage network provided in this embodiment utilizes an improved hierarchical clustering algorithm for spatial partitioning, which can adaptively handle the distribution of facilities with different densities, reflecting the actual accessibility between facilities. It also improves the rationality and accuracy of the drainage network inspection area division by optimizing and adjusting the boundaries of the initial inspection area.

[0110] In this embodiment, a drainage network intelligent inspection method based on RFID is provided, which can be used in the above-mentioned drainage network intelligent inspection system based on RFID. Figure 6 FIG. 1 is a flow chart of an RFID-based intelligent inspection method for a drainage network according to an embodiment of the present invention. Figure 6 As shown, the process includes the following steps: Step S601: The backend server divides the target drainage network into regions to obtain drainage network inspection areas. Figure 5 Step S501 of the illustrated embodiment will not be described in detail here.

[0111] In step S602, the mobile terminal performs path planning on the inspection area of the drainage network to obtain a global optimal inspection path.

[0112] Specifically, the path planning adopts a two-layer algorithm strategy of "global optimization + local optimization". The first layer uses the improved ant colony algorithm to generate a global access sequence that takes priority into account, and avoids local optimality through the pheromone update mechanism to ensure the rationality of the overall path; in the second layer, when calculating the optimal local path, in order to ensure the path finding efficiency, the mobile terminal calls The algorithm calculates the actual path and comprehensively considers multiple factors such as distance, time, road conditions (one-way streets, construction closures, time limits, etc.), real-time traffic information (congestion level, accident control, etc.) and time window constraints (avoidance during peak hours, facility inspection time requirements, etc.) to ensure the feasibility of the actual path.

[0113] The above step S602 includes: In step S6021, the mobile terminal performs global path planning on the inspection area of the drainage network to obtain a global access sequence.

[0114] In some optional implementations, step S6021 includes: In step b1, the mobile terminal obtains historical pheromone concentrations and heuristic information of the path between the starting facility point and other facility points in the inspection area of the drainage network.

[0115] Specifically, the heuristic information includes path distance and actual cost.

[0116] In step b2, the mobile terminal selects the next facility point based on the historical pheromone concentration and heuristic information of the path between the starting facility point and other facility points, and constructs an initial inspection path based on the starting facility point and the next facility point.

[0117] Specifically, the path planning starts from the starting facility point, selects the next facility point according to a certain probability based on the historical pheromone concentration and heuristic information of the path between the starting facility point and other facilities, and then constructs the initial inspection path based on the starting facility point and the next facility point.

[0118] In step b3, the mobile terminal obtains performance indicator data of the initial patrol path, and incrementally updates the pheromone concentration of the initial patrol path based on the performance indicator data of the initial patrol path.

[0119] Specifically, the performance indicator data of the initial inspection path includes the total length of the path, inspection efficiency or other performance indicators.

[0120] Furthermore, in order to further improve the accuracy and efficiency of inspection path planning, the pheromone update formula is introduced into the ant colony algorithm. The pheromone update formula is as follows: (4) in, For the iteration Ants are always at the facility point With facilities The pheromone concentration on the path between the facilities indicates that the facility With facilities How recommended is this path? is the pheromone volatility coefficient, which is used to gradually dilute historical experience and avoid the solidification of path selection. Indicates the Only ants can track the path in one iteration The pheromone contribution indicates the recommendation degree of the current path. is the number of ants.

[0121] Furthermore, the above pheromone update formula can effectively balance the accumulation and volatilization of pheromones, reducing the decline in inspection efficiency caused by falling into local optimality.

[0122] In step b4, the mobile terminal updates the initial patrol path based on the pheromone concentration of the incrementally updated initial patrol path until the pheromone concentration corresponding to the updated patrol path meets a preset condition, and then determines a global access sequence based on the updated patrol path.

[0123] Specifically, based on the above-mentioned pheromone update formula, the path planning algorithm generates a global path through repeated iterations. In each iteration, the path planning starts from the starting facility point and selects the next facility point with a certain probability based on the historical pheromone concentration of the path and the preset heuristic information. The selection of the next facility point is based on two factors, including: the recommendation degree of the path (i.e., pheromone concentration). The recommendation degree of the path reflects the excellence of the path in historical iterations. The path with a higher pheromone concentration is more likely to be selected; the distance of the path or other cost factors will also affect the selection. By introducing the pheromone volatility coefficient, the excessive solidification of the path selection is avoided, ensuring that the path planning will not fall into the local optimal solution.

[0124] Furthermore, after each iteration, all initial inspection paths incrementally update the pheromones on the initial inspection paths according to their performance index data. The better the performance index data, the larger the pheromone increment on the path, which makes the initial inspection path more likely to be preferred in subsequent iterations. Through the gradual update of pheromones, the path planning can be guided to gradually move towards the global optimal path. At the same time, the pheromone volatilization mechanism ensures that historically inefficient paths gradually disappear, avoiding the solidification of path selection.

[0125] Furthermore, after multiple iterations, the pheromone concentration gradually reflects the globally optimal or near-optimal inspection path, and the mobile terminal can effectively converge to an optimal solution. By balancing local search with global pheromone accumulation and volatilization, path planning avoids the dilemma of local optimal solutions and ultimately generates an efficient and accurate global access sequence.

[0126] Step S6022: The mobile terminal performs local path planning based on the global access sequence using a heuristic search algorithm to obtain an optimal local path.

[0127] Specifically, in order to ensure the efficiency of path finding, the local path calculation is combined with Adaptive selection heuristic function.

[0128] In some optional implementations, step S6022 includes: In step c1, the mobile terminal obtains the coordinates of adjacent facility points, traffic adjustment coefficients, and road congestion indexes in the global access sequence, and establishes a heuristic function based on the coordinates of the adjacent facility points, traffic adjustment coefficients, and road congestion indexes.

[0129] Specifically, the heuristic function uses an improved city block distance calculation method. The calculation formula is as follows: (5) in, Indicates the coordinates of the current facility point. represents the coordinates of the target facility point, is the road condition adjustment coefficient, is the congestion index of the current road section.

[0130] In step c2, the mobile terminal obtains multi-dimensional factors of the actual path between adjacent facility points in the global access sequence, and determines the actual cost based on the multi-dimensional factors of the actual path between adjacent facility points.

[0131] Specifically, the multi-dimensional factors of the actual path between adjacent facility points include: path distance, expected travel time, road condition influencing factors and other factors.

[0132] In step c3, the mobile terminal calculates a cost function based on the heuristic function and the actual cost, and determines an optimal local path based on the cost function.

[0133] Specifically, for the Euclidean distance between points, the cost function The calculation formula is as follows: (6) in, Represents the distance from the starting facility to the current facility the actual cost.

[0134] In step S6023, the mobile terminal splices the optimal local paths based on the global access sequence to obtain a global optimal inspection path.

[0135] Specifically, the global access sequence is used to determine the access order of the facility points. Based on the access order of the facility points, the optimal local paths between the facility points are spliced to obtain the global optimal inspection path.

[0136] In step S603, the RFID reader locates the RFID tag based on the global optimal inspection path to obtain the spatial location information of the RFID tag. Figure 5 Step S503 of the illustrated embodiment will not be described in detail here.

[0137] Step S604: The mobile terminal inspects the target drainage network using the spatial location information of the RFID tag to obtain the inspection results of the drainage network. Figure 5 Step S504 of the illustrated embodiment will not be described in detail here.

[0138] The RFID-based intelligent inspection method for drainage pipe networks provided in this embodiment performs global path planning for the inspection area of the drainage pipe network through a mobile terminal, obtains a global access sequence, ensures the rationality of the global optimal inspection path, utilizes a heuristic search algorithm for local path planning, improves the efficiency of local path planning, realizes the rational planning of the optimal local path, and splices the optimal local paths based on the global access sequence, which can not only obtain the approximate globally optimal inspection sequence, but also quickly adapt to dynamic road condition changes, flexibly plan local paths, and realize efficient multi-task point inspection.

[0139] In this embodiment, a drainage network intelligent inspection method based on RFID is provided, which can be used in the above-mentioned drainage network intelligent inspection system based on RFID. Figure 7 FIG. 1 is a flow chart of an RFID-based intelligent inspection method for a drainage network according to an embodiment of the present invention. Figure 7 As shown, the process includes the following steps: Step S701: The backend server divides the target drainage network into regions to obtain drainage network inspection areas. Figure 6 Step S601 of the illustrated embodiment will not be described in detail here.

[0140] Step S702: The mobile terminal performs route planning for the drainage network inspection area to obtain the global optimal inspection route. Figure 6 Step S602 of the illustrated embodiment will not be described in detail here.

[0141] In step S703, the RFID reader locates the RFID tag based on the global optimal inspection path to obtain spatial location information of the RFID tag.

[0142] Among them, after completing the inspection route planning, the inspection personnel carry RFID readers and writers and start to perform inspections according to the path navigation of the mobile terminal; before starting the inspection task, the inspection personnel need to scan their work permits through the mobile terminal for identity authentication and confirm the inspection authority. After the authentication is passed, the inspection task list for the day is automatically loaded; in order to solve the problems of "difficulty in finding points", "jumping points" and "missed inspections" during inspections, the innovative dual positioning technology of RFID tag phase recognition and GPS coupling is adopted.

[0143] Specifically, the above step S703 includes: In step S7031, the RFID reader receives the navigation coordinates of the RFID tag sent by the mobile terminal, transmits a reference signal to the RFID tag based on the navigation coordinates of the RFID tag, and receives multiple reception signals returned by the RFID tag.

[0144] Specifically, the mobile terminal obtains the RFID tag navigation coordinates from the navigation software SDK, and sends the RFID tag navigation coordinates to the RFID reader.

[0145] In step S7032, the RFID reader determines a plurality of phase differences based on the reference signal and the received signal.

[0146] In step S7033, the RFID reader obtains the operating wavelength of the reader antenna and calculates the communication propagation distance between the reader antenna and the RFID tag based on the operating wavelength of the reader antenna and the phase difference between the reference signal and the received signal; wherein the RFID reader includes the reader antenna.

[0147] Specifically, in order to improve positioning accuracy, the phase detection principle is adopted in the RFID reader to achieve high-precision tag positioning by analyzing the phase difference between the received signal and the reference signal. The calculation formula is as follows: (7) in, Indicates the communication propagation distance between the reader antenna and the RFID tag, Indicates the operating wavelength of the reader antenna.

[0148] In step S7034, the RFID reader obtains the coordinates of the reader antenna and determines the spatial location information of the RFID tag based on the coordinates of the reader antenna and the communication propagation distance between the reader antenna and the RFID tag.

[0149] Specifically, when the RFID reader uses a multi-antenna array, multiple baselines can be deployed to obtain multiple The spatial position information of the RFID tag is determined by combining the triangulation positioning equation. The calculation formula of the spatial position information of the RFID tag is as follows: (8) in, For the The position of the reader antenna array element, is the spatial location information of the RFID tag to be located, is the communication propagation distance between the reader antenna and the RFID tag.

[0150] Furthermore, when the distance between each reader antenna and the RFID tag can be measured, the accuracy of the RFID tag's spatial position information can be improved by minimizing the measurement error. The expression for minimizing the measurement error is: (9) in, is the measured distance between each reader antenna and the RFID tag.

[0151] Step S704: The mobile terminal inspects the target drainage network using the spatial location information of the RFID tag to obtain the inspection results of the drainage network. Figure 6 Step S604 of the illustrated embodiment will not be described in detail here.

[0152] The RFID-based intelligent inspection method for drainage pipe networks provided in this embodiment is based on the RFID tag navigation coordinates sent by the mobile terminal and utilizes the built-in phase detection technology of the RFID reader to accurately obtain the spatial position information of the RFID tag, achieving sub-meter positioning accuracy, and further guiding inspection personnel to accurately find the location of the facilities. The dual positioning method combining the RFID tag navigation coordinates and the RFID phase detection technology greatly improves the efficiency of finding hidden facilities such as underground manhole covers.

[0153] In this embodiment, a drainage network intelligent inspection method based on RFID is provided, which can be used in the above-mentioned drainage network intelligent inspection system based on RFID. Figure 8 FIG. 1 is a flow chart of an RFID-based intelligent inspection method for a drainage network according to an embodiment of the present invention. Figure 8 As shown, the process includes the following steps: Step S801: The backend server divides the target drainage network into regions to obtain drainage network inspection areas. Figure 7 Step S701 of the illustrated embodiment will not be described in detail here.

[0154] Step S802: The mobile terminal performs route planning for the inspection area of the drainage network to obtain the global optimal inspection route. Figure 7 Step S702 of the illustrated embodiment will not be described in detail here.

[0155] In step S803, the RFID reader locates the RFID tag based on the global optimal inspection path and obtains the spatial location information of the RFID tag. Figure 7 Step S703 of the illustrated embodiment will not be described in detail here.

[0156] Step S804: The mobile terminal inspects the target drainage network using the spatial location information of the RFID tag to obtain the inspection results of the drainage network. Figure 7 Step S704 of the illustrated embodiment will not be described in detail here.

[0157] Step S805: The backend server reviews the drainage network inspection result to obtain the drainage network inspection review result.

[0158] Specifically, after the inspection is completed, the drainage network inspection results need to be efficiently and accurately reviewed and comprehensively analyzed to ensure the closed-loop implementation of data quality and problem rectification; in response to the problems of low efficiency and difficulty in ensuring accuracy of manual inspection data verification, a multi-level intelligent audit system of "real-time verification-AI automatic audit-manual review" has been constructed. The background server integrates multiple AI algorithms such as computer vision, natural language processing, and spatiotemporal data analysis, realizing intelligent and efficient audit of drainage network inspection results.

[0159] Furthermore, the backend server integrates multiple AI algorithms for collaborative operation, building a comprehensive intelligent audit system. The specific steps for the backend server to audit the drainage network inspection results include: 1) Use YOLOv8 for image analysis. The YOLOv8 multi-task loss function is: (10) in, 、 and They are classification loss weight, bounding box regression loss weight, and distribution focus loss weight, respectively. 、 and They are classification loss, bounding box regression loss, and distribution focus loss, represents the total loss of multiple tasks, is the total number of facility defect categories (such as cracks, leakage, etc., 12 categories), is the true category label, is the predicted category probability, is the difficult sample focusing factor, It is an improved bounding box evaluation metric that comprehensively considers the degree of box overlap, center point distance, and aspect ratio consistency. and are the coordinates of the predicted box and the real box respectively, is the distribution score after discretization of the bounding box.

[0160] 2) Intelligent Image-Text Matching Verification: By building an image-text matching evaluation model, we ensure that on-site photos are consistent with the text description. We first perform semantic understanding of the inspection text description and use the BERT-base (Bidirectional Encoder Representation from Transformers, a pre-trained language model) model to extract key information. We then calculate the semantic matching between the image recognition results and the text description. Finally, we use the attention mechanism to compare key features to determine whether the image and text describe the same issue.

[0161] 3) Spatiotemporal Data Analysis: A spatiotemporal sequence analysis model is established using an LSTM (Long Short-Term Memory) recurrent neural network. The input features of this spatiotemporal sequence analysis model include multidimensional data such as longitude and latitude coordinates, time intervals between adjacent inspection points (to detect inspection speed anomalies), duration of stay at a single point (to identify perfunctory inspections), time difference between taking photos and recording (to detect supplementary recordings), operation sequence coding (to verify process standardization), completion rate of required items, and number of multimedia records. By analyzing the rationality of inspection routes, abnormal working hours, and standardization of operation sequences, this model can effectively identify possible data falsification or inspection omissions.

[0162] For example, when the Chinese description of a drainage network mentions "damaged manhole covers with cracks on the edges," the backend server uses YOLOv8 to confirm the location and damage of the manhole covers in the inspection image, uses Transformer (an attention mechanism model) for semantic matching of images and text, and combines the spatiotemporal data analyzed by LSTM to verify the rationality of the inspection process. This multi-model collaborative intelligent audit method can significantly improve the efficiency and accuracy of drainage network verification.

[0163] 4) The backend server has established a complete manual review process for issues identified as serious or suspicious by AI audits: Auditors will conduct an in-depth investigation of the issues, carefully checking the degree of match between images and text descriptions and verifying the on-site conditions. Auditors will give a clear "pass" or "correction required" judgment result and automatically push the conclusion to the relevant inspection or maintenance department through the backend server for follow-up work.

[0164] Furthermore, to ensure the continued improvement in the accuracy of drainage network inspection and audit results, the audit analysis adopted a key optimization strategy. By regularly retraining the AI model with newly added labeled data to improve model performance, typical problems discovered during the audit process are continuously incorporated into the knowledge base for rule maintenance and improvement. At the same time, through statistical analysis of the time and accuracy of each audit, the bottlenecks in the operation of the backend server are identified and targeted improvements are made to continuously improve the overall audit efficiency.

[0165] In step S806, the backend server performs data analysis on the inspection results of the drainage network and generates a drainage network maintenance plan based on the data analysis results.

[0166] Specifically, after the drainage network inspection results are reviewed and approved, the backend server is responsible for conducting in-depth analysis and historical data mining of the drainage network inspection results, and combining multi-objective optimization algorithms to intelligently generate maintenance plans to provide scientific decision-making support for facility maintenance. It is mainly divided into three core links: problem diagnosis, trend prediction, and plan generation. The backend server will send the generated maintenance plan to the mobile terminal for relevant maintenance personnel to execute.

[0167] Furthermore, the inspection results of the drainage network are used for problem diagnosis and classification. The backend server systematically analyzes the multi-dimensional data collected during the inspection. First, the cause of the problem facilities is diagnosed based on the expert rule library of case-based reasoning (CBR). Differentiated processing strategies are formulated for the four types of assessment results: Class A (normal), Class B (minor problems), Class C (obvious defects), and Class D (serious failures). At the same time, the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) density clustering algorithm is used to analyze the spatial distribution characteristics of the problem facilities, identify potential regional hidden dangers, and provide a scientific basis for subsequent regional remediation plans.

[0168] Furthermore, trend prediction and risk assessment are carried out: the backend server uses a bidirectional LSTM network to build a time series prediction model, comprehensively analyzing the historical status indicators of the facilities (such as pipeline siltation, crack width, etc.), environmental factors (rainfall and surface subsidence, etc.), and human factors (surrounding construction and traffic load, etc.). Through the attention mechanism, key time features are automatically identified, and the seasonal decomposition algorithm is combined to process periodic changes to achieve accurate prediction of the facility status. Based on the prediction results, an improved PageRank (a technology that calculates hyperlinks between web pages) algorithm is used to rank the risk levels of the facilities and scientifically determine the priority of preventive maintenance.

[0169] Furthermore, maintenance plans are formulated based on the classification of multi-objective optimization algorithms. For immediate rectification projects (C / D level problems), standard maintenance processes are matched through knowledge graphs, genetic algorithms are used to optimize resource allocation, and heuristic algorithms are used to determine the optimal construction time window. For preventive maintenance projects, the background server analyzes historical maintenance effects through a reinforcement learning model and uses Q-learning (a value-based reinforcement learning algorithm) to continuously optimize maintenance strategies, seeking the optimal balance between cost and effect. Based on the optimized maintenance plan, natural language generation (NLG) technology is used to automatically generate professional maintenance reports, covering key information such as problem statistics, diagnostic instructions, rectification suggestions, and resource budgets, thus establishing a complete closed-loop maintenance management system.

[0170] The RFID-based intelligent inspection method for drainage pipe networks provided in this embodiment reviews the inspection results of the drainage pipe networks through a background server to obtain the inspection and review results of the drainage pipe networks, thereby avoiding the problem of possible deviations in manual inspections and greatly improving the efficiency of subsequent verification. The inspection results of the drainage pipe networks are then analyzed, and a maintenance plan for the drainage pipe networks is generated based on the data analysis results, providing scientific decision-making support for the maintenance of drainage pipe network facilities.

[0171] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the embodiments of this application.

[0172] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0173] In the several embodiments provided in the embodiments of the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0174] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0175] 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.

[0176] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0177] Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention. Such modifications and variations are all within the scope defined by the appended claims.

Claims

1. An RFID-based intelligent inspection system for drainage pipe networks, characterized in that: The system includes: a backend server, a mobile terminal, an RFID reader and an RFID tag installed on a drainage network facility; the RFID tag, the RFID reader, the mobile terminal and the backend server are connected in sequence; The backend server is used to divide the target drainage network into regions to obtain drainage network inspection areas; The mobile terminal is used to plan a path for the inspection area of the drainage network to obtain a global optimal inspection path; The RFID reader is configured to locate the RFID tag based on the global optimal inspection path to obtain spatial location information of the RFID tag; The mobile terminal is used to inspect the target drainage network using the spatial location information of the RFID tag to obtain the drainage network inspection result.

2. An RFID-based intelligent inspection method for drainage pipe networks, characterized in that: Applied to the RFID-based intelligent inspection system for drainage pipe networks according to claim 1, the method comprises: The backend server divides the target drainage network into regions and obtains the drainage network inspection areas; The mobile terminal performs path planning on the inspection area of the drainage network to obtain a global optimal inspection path; The RFID reader locates the RFID tag based on the global optimal inspection path to obtain spatial location information of the RFID tag; The mobile terminal inspects the target drainage network using the spatial location information of the RFID tag to obtain the drainage network inspection result.

3. The method according to claim 2, characterized in that The backend server divides the target drainage network into regions to obtain drainage network inspection areas, including: The backend server divides the target drainage network into regions using an improved hierarchical clustering algorithm to obtain initial inspection areas; The backend server optimizes and adjusts the boundaries of the initial inspection area to obtain the drainage network inspection area.

4. The method according to claim 3, characterized in that The backend server uses an improved hierarchical clustering algorithm to divide the target drainage network into regions to obtain initial inspection areas, including: The backend server obtains the actual road network distance, the priority of each facility point and the historical operation and maintenance data of the facility point in the target drainage network; The backend server calculates the composite distance between each facility point based on the actual road network distance, the priority of each facility point and the historical operation and maintenance data of the facility point; The backend server clusters the plurality of facility points based on the composite distance between the facility points to obtain a plurality of clusters; The backend server calculates the silhouette coefficients and workload balance corresponding to the plurality of clusters respectively, and compares the silhouette coefficients and the workload balance with an iteration termination condition; If both the silhouette coefficient and the workload balance meet the iteration termination condition, the backend server divides the target drainage network based on the multiple clusters to obtain the initial inspection area.

5. The method according to claim 2, characterized in that The mobile terminal performs path planning on the drainage network inspection area to obtain a global optimal inspection path, including: The mobile terminal performs global path planning on the drainage network inspection area to obtain a global access sequence; The mobile terminal performs local path planning based on the global access sequence using a heuristic search algorithm to obtain an optimal local path; The mobile terminal splices the optimal local paths based on the global access sequence to obtain the global optimal inspection path.

6. The method according to claim 5, characterized in that The mobile terminal performs global path planning on the drainage network inspection area to obtain a global access sequence, including: The mobile terminal obtains historical pheromone concentrations and heuristic information of the path between the starting facility point and other facility points in the drainage network inspection area; The mobile terminal selects a next facility point based on historical pheromone concentrations and heuristic information of a path between the starting facility point and the other facility points, and constructs an initial inspection path based on the starting facility point and the next facility point; The mobile terminal obtains performance indicator data of the initial patrol path, and incrementally updates the pheromone concentration of the initial patrol path based on the performance indicator data of the initial patrol path; The mobile terminal updates the initial patrol path based on the pheromone concentration of the incrementally updated initial patrol path until the pheromone concentration corresponding to the updated patrol path meets a preset condition, and then determines the global access sequence based on the updated patrol path.

7. The method according to claim 5, characterized in that The mobile terminal performs local path planning based on the global access sequence using a heuristic search algorithm to obtain an optimal local path, including: The mobile terminal obtains the coordinates of adjacent facility points, the traffic adjustment coefficient, and the road section congestion index in the global access sequence, and establishes a heuristic function based on the coordinates of the adjacent facility points, the traffic adjustment coefficient, and the road section congestion index; The mobile terminal obtains multi-dimensional factors of an actual path between adjacent facility points in a global access sequence, and determines an actual cost based on the multi-dimensional factors of the actual path between adjacent facility points; The mobile terminal calculates a cost function based on the heuristic function and the actual cost, and determines the optimal local path based on the cost function.

8. The method according to claim 2, characterized in that The RFID reader locates the RFID tag based on the global optimal inspection path to obtain spatial location information of the RFID tag, including: The RFID reader receives the RFID tag navigation coordinates sent by the mobile terminal, transmits a reference signal to the RFID tag based on the RFID tag navigation coordinates, and receives multiple reception signals returned by the RFID tag; The RFID reader determines a plurality of phase differences based on the reference signal and the received signal; The RFID reader / writer obtains an operating wavelength of a reader / writer antenna, and calculates a communication propagation distance between the reader / writer antenna and the RFID tag based on the operating wavelength of the reader / writer antenna and a phase difference between the reference signal and the received signal; wherein the RFID reader / writer includes the reader / writer antenna; The RFID reader / writer obtains the coordinates of the reader / writer antenna and determines the spatial position information of the RFID tag based on the coordinates of the reader / writer antenna and the communication propagation distance between the reader / writer antenna and the RFID tag.

9. The method according to claim 2, characterized in that Also includes: The backend server reviews the drainage network inspection result to obtain the drainage network inspection review result.

10. The method according to claim 9, characterized in that Also includes: The backend server performs data analysis on the drainage network inspection results and generates a drainage network maintenance plan based on the data analysis results.

Images