Wireless water leakage detection and alarm method and system for hidden space in building

Abstract

The invention relates to the technical field of a wireless water leakage detection and alarm method used in a building, and discloses a wireless water leakage detection and alarm method and system used in a hidden space in the building. The technical problems that in the prior art, in a complex electromagnetic environment, wireless signal transmission stability is poor, durability of a water leakage sensor is insufficient, and the accurate positioning capacity of a water leakage position is poor are solved. According to the method and system, a distributed water leakage detection unit, a wireless data gateway and a central management platform are included, and accurate positioning of water leakage and intelligent water source identification are achieved. By adopting the technical scheme, the problems of signal robustness and sensor durability can be effectively solved, accurate positioning of water leakage and water source inference are realized, response time is shortened, loss is reduced, and building safety management is improved.

Description

Technical Field

[0001] This invention belongs to the field of Internet of Things (IoT) technology, and specifically relates to a wireless water leakage detection and alarm method and system for concealed spaces inside buildings. Background Technology

[0002] As a vital carrier of human social activities, the maintenance and management of the internal environment of buildings has always been a core issue in property operation and asset preservation and appreciation. Especially in concealed spaces within buildings, such as inside walls, above ceilings, under floors, and in areas like pipe shafts and equipment compartments, the complex structures and lack of daily access make them highly susceptible to hidden leaks, such as burst pipes, equipment leaks, or external wall seepage. Failure to detect and effectively address these leaks in a timely manner can lead to serious structural damage, mold growth, and damage to building materials, and even substantial property losses and safety hazards, profoundly impacting the building's lifespan and quality of life. Given the destructive nature and potential risks of such leaks, establishing an efficient, reliable, and highly adaptable leak detection and early warning mechanism has become a critical technical challenge urgently needing to be addressed in modern building management.

[0003] In traditional practice, leak detection inside buildings primarily relies on manual inspections and wired sensor systems. While manual inspections offer some flexibility, limitations in manpower, inspection frequency, and the physical accessibility of concealed spaces often hinder early, real-time detection of leaks. Typically, leaks are only detected when they have progressed to the point of leaving visible water stains, mold, or structural damage. Repairs at this stage often require higher costs, longer construction periods, and may even involve disruptive alterations to the building's aesthetics and functionality. On the other hand, wired sensor systems, as an automated detection solution, connect sensors to a central control unit via physical cables, theoretically providing relatively stable data transmission and power supply. However, wired systems face significant challenges in practical deployment: their wiring is complex, time-consuming, and labor-intensive, especially in renovation projects of existing buildings, requiring extensive wall excavation, ceiling dismantling, or floor demolition. This not only increases construction difficulty and costs but also severely impacts the building's normal use. In addition, the cables themselves are also at risk of corrosion and damage. Once the line fails, it will directly affect the reliability of the entire detection network. Summary of the Invention

[0004] To achieve the aforementioned objectives, this invention provides a wireless leak detection and alarm method and system for concealed spaces within buildings, aiming to solve the technical problems of poor wireless signal transmission stability in complex electromagnetic environments, insufficient durability of leak sensors, and lack of precise leak location capabilities in existing technologies. This invention introduces a multi-modal wireless communication strategy, adaptive sensor health management technology, and a multi-source data fusion spatial positioning system to construct a highly integrated, robust, reliable leak monitoring and early warning solution with precise location capabilities.

[0005] In a first aspect, the present invention provides a wireless water leakage detection and alarm system for concealed spaces inside buildings.

[0006] It includes: multiple leak detection units deployed in a distributed manner, used to monitor the moisture status of their respective areas in real time and transmit the monitoring data through a wireless network; at least one wireless data gateway unit, coupled to the multiple leak detection units, used to receive and aggregate wireless data from the leak detection units and upload the data; The system also includes a central management platform coupled to the wireless data gateway unit for receiving, storing, analyzing, and displaying data from the wireless data gateway unit, and for early warning, location, and management of water leakage events. The water leakage detection unit comprises: a sensor body employing non-contact capacitive or photoelectric sensing principles for moisture detection, with its encapsulation material being polytetrafluoroethylene (PTFE) or polyvinylidene fluoride (PVDF); a low-power microprocessor coupled to the sensor body for digitizing, preprocessing, and performing local analysis on the raw data collected by the sensor, and operating in an ultra-low-power standby mode during non-working periods; and a multi-mode wireless communication module coupled to the low-power microprocessor. The system includes at least two wireless communication submodules configured with different frequency bands or different communication protocols to enable data transmission between the leak detection unit and the wireless data gateway unit. These submodules adaptively select the communication mode based on real-time environmental channel status information or a preset communication strategy using an intelligent switching algorithm. An energy harvesting and management module, coupled to the low-power microprocessor and the multi-mode wireless communication module, provides a sustainable energy supply to the leak detection unit. An integrated diagnostic and calibration module, coupled to the low-power microprocessor and the sensor body, monitors the internal operating status of the leak detection unit in real time and performs fault warnings and performance calibration.

[0007] Optionally, the sensor body specifically includes: a sensing element composed of staggered microelectrodes, wherein the electrode array substrate material is selected from FR4 or a polyimide flexible substrate; The microelectrode is coated with a hydrophilic polymer film. After absorbing water, the dielectric constant of the film changes, causing a change in the capacitance of the sensor body. This change is positively correlated with the moisture content. The low-power microprocessor integrates a high-precision analog-to-digital converter and a programmable logic controller to periodically wake up the sensor body, execute data acquisition instructions, filter data, and perform preliminary moisture threshold judgment. The low-power microprocessor can be woken up from standby mode by an external interrupt or timer.

[0008] Optionally, the multimodal wireless communication module is configured with at least two wireless communication sub-modules of different frequency bands or different communication protocols, including: a sub-module supporting Sub-GHz low-power wide-area network communication protocols, including LoRa or NB-IoT, for realizing long-distance, high-penetration data transmission; The system also includes another submodule supporting ultra-wideband communication protocols, conforming to the IEEE 802.15.4a standard, for achieving high-precision ranging and short-range high-bandwidth data transmission. The multi-mode wireless communication module uses an intelligent switching algorithm to dynamically select Sub-GHz or UWB communication mode based on channel state information estimated from parameters such as real-time received signal strength indication, signal-to-noise ratio, bit error rate, or packet error rate. The transmit power, modulation order, and coding rate of the multi-mode wireless communication module can be dynamically adjusted by the low-power microprocessor based on real-time signal-to-noise ratio and bit error rate to achieve link self-adaptation. The multi-mode wireless communication module is also equipped with multiple sets of orthogonally polarized antennas or spatially separated antenna arrays for achieving spatial diversity or multiple-input multiple-output communication.

[0009] Optionally, the energy harvesting and management module includes at least one photovoltaic panel and / or one thermoelectric generator. The photovoltaic panel is used to harvest solar energy in a shaded space with illumination, and the thermoelectric generator is used to harvest thermal energy from the surface of the pipe or from the temperature difference in the environment. The energy harvesting and management module also includes an energy storage unit, which uses a solid-state lithium-ion battery or a supercapacitor, to store the harvested energy and provide power when energy harvesting is insufficient. The energy management circuit integrates a maximum power point tracking controller and has overcharge, over-discharge, and short-circuit protection functions. The integrated diagnostic and calibration module includes an internal temperature sensor, a humidity sensor, and an internal power supply voltage monitor. The diagnostic function determines whether the device performance has degraded by monitoring parameters such as the drift of the reference capacitance value of the sensor body, the transmission success rate of the communication module, and the number of charge-discharge cycles of the battery, combined with preset health thresholds or an anomaly detection model based on machine learning. The calibration function supports remote or periodic self-calibration by recording the sensor reference value and correcting sensor drift under waterless conditions.

[0010] Optionally, the wireless data gateway unit is configured with a multimode wireless receiving module that matches the multimode wireless communication module of the leak detection unit. This receiving module can simultaneously or alternately receive Sub-GHz and UWB signals. The wireless data gateway unit has advanced signal processing capabilities, including multipath signal combining, such as maximum ratio combining or selective combining algorithms, interference suppression, forward error correction decoding, and data integrity verification. The wireless data gateway unit has a local data caching function for temporarily storing data when the connection with the central management platform is interrupted, and synchronizing the data after the connection is restored. The wireless data gateway unit has a built-in high-performance microcontroller for performing preliminary timestamping and encryption processing on the received leak data, and executing the data synchronization protocol with the central management platform. The wireless data gateway unit also supports some edge computing functions, including performing localized preliminary location calculations on the received leak events, or executing simple linkage control logic.

[0011] Optionally, the central management platform includes: a data storage and management module for persistently storing all leak detection data, sensor diagnostic data, network topology information, and historical alarm records, wherein the module adopts a distributed database architecture; a data analysis and location module for accurately locating the leak and intelligently identifying the leak source; a visualization and alarm module for intuitively displaying the location and related information of the leak event through a graphical user interface and supporting multiple alarm methods; and a user interaction interface providing functions such as administrator permission management, system configuration, historical data query, report generation, and remote device diagnostic and calibration command issuance.

[0012] Optionally, the data analysis and positioning module includes: an indoor positioning engine based on multi-source fusion, which combines UWB ranging data, RSSI fingerprint data, and hop count information from the sensor network, and uses weighted least squares, extended Kalman filtering, or particle filtering algorithms to estimate the precise location of the alarm leak detection unit in real time; the positioning engine is pre-loaded with building information model data, including two-dimensional / three-dimensional structural diagrams of the building, pipe routing diagrams, wall material information, and equipment layout diagrams; and a leak source inference module based on a fluid dynamics model, which, when one or more leak detection units alarm, uses the... The BIM data contains information on building structure and piping layout. Combined with the initial location of the leak point, a simplified water flow path simulation algorithm is used to infer the direction of water diffusion and possible leak source areas. This module can consider the influence of gravity, surface tension, capillary action, and material absorption characteristics on the water flow path, and can refer to a preset water flow diffusion model or use historical leak event data for machine learning training. An abnormal event correlation analysis module correlates leak alarm data with related building equipment operation data, including pump start / stop status, valve opening, and pipeline pressure sensor data, to further narrow down the range of leak sources.

[0013] Optionally, the visualization and alarm module can accurately mark the location of the detection unit for water leakage alarm on the 3D view or 2D plan of the BIM model in the form of highlights, animations, or icons, and simultaneously display the inferred water leakage source area; the alarm methods supported by the visualization and alarm module include SMS service notifications, email notifications, mobile application push notifications, and linkage with building automation systems or fire alarm systems; the alarm information includes key information such as water leakage location, inferred source, alarm time, and sensor health status.

[0014] Secondly, the present invention provides a wireless water leakage detection and alarm method for concealed spaces inside buildings.

[0015] The method includes the following steps: Step 1: Deployment and initialization of leakage detection units, including the distributed deployment of multiple leakage detection units in concealed spaces within the building, recording the initial geographical location information of each leakage detection unit and spatially mapping and binding it with the building information model of the central management platform, and performing self-checks and initialization; Step 2: Multimodal wireless communication and data transmission, including the leakage detection units periodically or based on event triggers collecting moisture status data, and transmitting the data to a wireless data gateway unit through a multimodal wireless communication module. The wireless data gateway unit receives the wireless signals, performs signal fusion processing, and then uploads the data to the central management platform; Step 3: Intelligent identification and adaptive early warning of leakage events, including the central management platform receiving and analyzing the leakage detection data, and when the moisture level... When the value continuously exceeds the preset leakage threshold, a leakage event is determined to have occurred, and a multi-level adaptive early warning mechanism is triggered based on the leakage level and the importance of the affected area; Step 4: Precise location of the leakage and inference of the water source, including the data analysis and positioning module using the multi-source fusion indoor positioning engine to accurately calculate the spatial coordinates of the alarm leakage detection unit, combining the BIM model and fluid dynamics model leakage source inference module to infer the diffusion direction of water flow and the leakage source area, and performing correlation analysis with the building automation system data through the abnormal event correlation analysis module; Step 5: Fault diagnosis and preventive maintenance, including the data analysis and positioning module of the central management platform continuously monitoring the health status data of each leakage detection unit, generating a fault warning when the parameters deviate from the health threshold, and supporting remote parameter configuration and firmware upgrades.

[0016] Optionally, step four specifically includes: based on the indoor positioning engine, using the distance measurement data between the UWB signal emitted by the alarm leak detection unit and the preset UWB anchor point, combined with a polygonal measurement algorithm or a nonlinear optimization algorithm, to accurately calculate the two-dimensional or three-dimensional spatial coordinates of the alarm leak detection unit, and using RSSI fingerprint spectrum as auxiliary correction; the positioning engine matches the calculated precise location with the BIM model, and highlights the alarm point in the digital twin environment; the fluid dynamics model leak source inference module, based on the precise location of the alarm point, combined with the detailed information about the building structure and piping system in the BIM model, uses a preset fluid diffusion model to trace back to the most likely leak source area, and refers to the pattern recognition results of historical leak data; the abnormal event correlation analysis module performs real-time correlation analysis between the leak alarm information and data such as water pressure, water flow sensors, and valve status from the building automation system, so as to improve the inference accuracy of the leak source to the specific pipe or equipment.

[0017] Through the above-mentioned technical solution, this invention not only effectively solves the robustness problem of wireless signal transmission in complex building interior environments and significantly improves the durability and maintenance-free cycle of water leakage sensors, but more importantly, it realizes the accurate location of water leakage events in hidden spaces and the intelligent inference of the source of leakage, thereby greatly shortening the fault response time, reducing the economic losses caused by water leakage, and improving the safety management and operational efficiency of buildings. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the wireless water leakage detection and alarm system for concealed spaces inside buildings according to the present invention. Figure 2 This is a schematic diagram of the structure of the leakage detection unit of the present invention; Figure 3 This is a flowchart illustrating the wireless water leakage detection and alarm method for concealed spaces inside buildings according to the present invention.

[0019] The attached diagram is labeled as follows: 1. Leakage detection unit; 2. Wireless data gateway unit; 3. Central management platform; 4. Sensor body; 5. Low-power microprocessor; 6. Multimodal wireless communication module; 7. Energy harvesting and management module; 8. Integrated diagnostic and calibration module. Detailed Implementation

[0020] In one specific embodiment, the wireless water leakage detection and alarm system for concealed spaces inside buildings according to the present invention, such as... Figure 1 As shown, it mainly consists of multiple distributed leak detection units 1, at least one wireless data gateway unit 2, and a central management platform 3. These components interact and collaborate via wireless or wired networks to achieve comprehensive monitoring, intelligent identification, precise location, and timely early warning of leaks in concealed spaces within the building.

[0021] Leakage detection unit 1, as the sensing terminal of the system, is designed to operate stably in a hidden space where no one is on duty for a long time, and to monitor the moisture status of the area in real time. Figure 2 The internal structure of the leak detection unit 1 is shown in detail. Its core components include: a sensor body 4, a low-power microprocessor 5, a multi-mode wireless communication module 6, an energy harvesting and management module 7, and an integrated diagnostic and calibration module 8.

[0022] The sensor body 4 is a key component of the water leakage detection unit 1, sensing moisture. Its working principle is based on non-contact capacitive sensing. Specifically, the sensor body 4 employs an interlaced array of microelectrodes made from high-purity copper or nickel alloy through a precision etching process. Its geometry is optimized, with electrode widths adjustable from 0.2 mm to 0.5 mm and electrode spacing from 0.1 mm to 0.3 mm to ensure a balance between high sensitivity and low parasitic capacitance. The electrode array substrate is made of moisture-resistant, low-dielectric-loss FR4 or polyimide flexible substrate. A hydrophilic polymer film, such as polyvinyl alcohol (PVA) or a polyacrylamide derivative, with a thickness of approximately 5 to 15 micrometers, is uniformly coated on the electrode surface. This hydrophilic film exhibits a specific dielectric constant in its dry state. When it adsorbs moisture, the penetration of water molecules (with a high dielectric constant of approximately 80) significantly alters the overall dielectric constant of the film, thereby increasing the capacitance of the sensor body 4. The change in capacitance value exhibits an approximately linear or nonlinear positive correlation with the moisture content absorbed by the thin film, which can be quantified using a pre-calibrated curve. To enhance the sensor's durability and stability in extreme environments, the sensor body 4 is encapsulated using high-performance fluoropolymer materials such as polytetrafluoroethylene (PTFE) or polyvinylidene fluoride (PVDF). These materials, with their excellent chemical resistance, UV resistance, weather resistance, hydrophobicity, extremely low coefficient of friction, and superior electrical insulation properties, ensure that the sensor body 4 maintains its structural integrity even under long-term high humidity or even short-term immersion conditions, preventing electrode oxidation, corrosion, and leakage, thereby significantly extending the lifespan of the sensor body 4 and reducing maintenance frequency. This encapsulation typically employs injection molding or thermoforming processes to ensure a good seal with the internal circuitry.

[0023] The low-power microprocessor 5, which can be an STM32L4 series (such as STM32L476RG) or Nordic nRF52 series (such as nRF52840), is coupled to the sensor body 4. Its main function is to perform high-precision digitization, preprocessing, and local analysis on the raw analog data acquired by the sensor body 4. The low-power microprocessor 5 integrates a 12-bit to 16-bit high-precision analog-to-digital converter (ADC) with a sampling frequency configurable from 100Hz to 1kHz to accurately capture minute changes in capacitance. The low-power microprocessor 5 also has a built-in programmable logic controller (PLC) function, or similar logic can be implemented via firmware, for precisely controlling the periodic wake-up of the sensor body 4, the generation of capacitance measurement excitation signals (e.g., generating a square wave or sine wave through a high-frequency oscillator or relaxor oscillator and measuring its response frequency or amplitude), executing data acquisition instructions, and eliminating environmental noise interference through digital filtering algorithms (e.g., moving average filtering, Kalman filtering, or exponential smoothing filtering). The filtered capacitance data is compared in real time with a preset leakage threshold for preliminary moisture threshold determination. During non-working cycles, such as between data acquisition intervals (configurable intervals, e.g., every 5 or 10 minutes), the low-power microprocessor 5 enters an ultra-low-power standby or sleep mode, where its power consumption can be reduced to the microampere level (e.g., less than 2 μA) to maximize energy savings. The low-power microprocessor 5 can be woken from standby mode through several mechanisms: one is periodic wake-up via an internal timer to perform routine data acquisition; another is event-triggered wake-up, such as when the capacitance value detected by the sensor body 4 undergoes a sudden change (exceeding a specific rapid change rate threshold, indicating a possible leakage event), the low-power microprocessor 5 is quickly woken up via an external interrupt pin, thus achieving near real-time response.

[0024] The multimodal wireless communication module 6, coupled to the low-power microprocessor 5, is the core of reliable data transmission between the leak detection unit 1 and the external network. This multimodal wireless communication module 6 is configured with at least two independent wireless communication sub-modules operating on different frequency bands or using different communication protocols. For example, the first sub-module can support Sub-GHz Low-Power Wide-Area Network (LPWAN) communication protocols, such as LoRaWAN or NB-IoT. If LoRaWAN is selected, its operating frequency band can be set to the 868MHz ISM band in Europe or the 915MHz ISM band in North America, and it supports various spreading factors (SF7 to SF12) and coding rates (4 / 5 to 4 / 8) to achieve line-of-sight transmission distances of up to 15 kilometers and indoor penetration distances of tens of meters under different signal-to-noise ratio conditions. Typical data rates range from 0.3kbps to 50kbps, and transmit power can be adjusted up to a maximum of 20dBm. If NB-IoT is selected, it operates on operator-licensed frequency bands, offering wider coverage and stronger penetration, making it particularly suitable for areas with poor signal coverage, such as basements, but with relatively higher power consumption. The second submodule supports the Ultra-Wideband (UWB) communication protocol, conforming to the IEEE 802.15.4a standard, and typically operates in the frequency band between 3.1 GHz and 10.6 GHz. UWB, with its extremely narrow pulses (typically nanoseconds) and wide bandwidth (typically greater than 500 MHz), achieves centimeter-level precise positioning capabilities (e.g., ranging accuracy within 10 cm) and high-bandwidth data transmission over short distances (up to tens of Mbps), with typical transmission distances ranging from 30 meters to 100 meters. The multimodal wireless communication module 6, through its built-in intelligent switching algorithm, adaptively selects the optimal communication mode based on real-time environmental channel state information (CSI) or a preset communication strategy. Channel state information can be estimated from parameters such as Received Signal Strength Indication (RSSI), Signal-to-Noise Ratio (SNR), Bit Error Rate (BER), or Packet Error Rate (PER). For example, when the RSSI is detected to be below -110 dBm or the BER is above 10^-3, the system can preferentially select the Sub-GHz band for transmission, utilizing its longer wavelength and stronger diffraction and penetration capabilities to ensure reliable delivery of data packets. When precise ranging and positioning are required (e.g., when a water leak alarm is triggered and precise positioning is needed), the system quickly switches to UWB mode for high-precision ranging data transmission. The transmit power, modulation order (e.g., QPSK, 16QAM), and coding rate of the multi-mode wireless communication module 6 can all be dynamically adjusted by the low-power microprocessor 5 based on real-time signal-to-noise ratio and bit error rate, realizing link adaptation. For example, when channel quality is good, high-order modulation and a high coding rate can be used to improve data throughput; when channel quality deteriorates, it degrades to low-order modulation and a low coding rate, sacrificing throughput for transmission robustness, thereby ensuring data transmission reliability and energy efficiency.To further improve signal reception quality and transmission reliability in complex multipath fading environments, the multimode wireless communication module 6 is also equipped with multiple sets of orthogonally polarized antennas (e.g., one vertically polarized antenna and one horizontally polarized antenna) or spatially separated antenna arrays (e.g., two or more antennas with a spacing greater than half a wavelength) to implement spatial diversity or MIMO (multiple-input multiple-output) communication technologies. Spatial diversity can effectively combat fading by receiving and combining signal copies from different paths to improve the received signal-to-noise ratio; while MIMO can further improve spectral efficiency and link robustness.

[0025] The energy harvesting and management module 7, coupled to the low-power microprocessor 5 and the multimode wireless communication module 6, is designed to provide a sustainable energy supply for the leak detection unit 1 and significantly extend its maintenance-free operating cycle. This energy harvesting and management module 7 includes at least one or more energy harvesting devices. For example, for partially concealed spaces with weak lighting (such as weak light transmitted through light wells, glass bricks, or vents), one or more micro-amorphous silicon or CIGS (copper indium gallium selenide) photovoltaic panels with dimensions of approximately 20mm x 20mm to 50mm x 50mm can be integrated, which can still provide microwatt to milliwatt-level power output in low-illuminance environments (e.g., below 500 lux). For areas near pipes, HVAC systems, or other locations with temperature differences, one or more thermoelectric generators (TEGs) can be integrated to convert ambient temperature differences (e.g., a 5°C to 10°C temperature difference between the pipe surface and ambient air) into electrical energy using the Seebeck effect. A single TEG can generate tens to hundreds of microwatts of power under typical temperature differences. The energy harvesting and management module 7 also includes an energy storage unit that uses a long-life, wide-temperature-range (-20°C to 60°C) solid-state lithium-ion battery (e.g., a thin-film lithium-ion battery with a capacity between 10mAh and 100mAh) or a supercapacitor (with a capacity between 0.1F and 1F) to store the harvested energy and power the leak detection unit 1 when energy harvesting is insufficient or when there are instantaneous high power demands (such as peak current for UWB transmission). The energy management circuit integrates a high-efficiency maximum power point tracking (MPPT) controller, such as a DC-DC converter based on perturb and observe or incremental conductance, to optimize energy harvesting efficiency in real time. Furthermore, the circuit features comprehensive battery protection functions, including overcharge protection, over-discharge protection, overcurrent protection, and short-circuit protection, ensuring the safe and long-term operation of the energy storage unit. Through precise power budget management and a sleep / wake-up mechanism, the average power consumption of the entire leakage detection unit 1 can be controlled to between a few microwatts and tens of microwatts, enabling maintenance-free operation for several years or even more than ten years without external power supply.

[0026] An integrated diagnostic and calibration module 8, coupled to a low-power microprocessor 5 and a sensor body 4, is used to monitor the internal operating status of the leakage detection unit 1 in real time and to provide fault warnings and performance calibration. This integrated diagnostic and calibration module 8 includes multiple embedded sensors, such as a high-precision NTC thermistor or digital temperature sensor (e.g., DS18B20) for detecting the internal ambient temperature; a capacitive or resistive humidity sensor for monitoring internal humidity; and a sophisticated power supply voltage monitor (e.g., reading the voltage on the voltage divider network via an ADC) for real-time tracking of battery voltage and the health of the system power rails. Diagnostic functions are achieved through continuous monitoring of a series of key parameters. These include, but are not limited to: the long-term drift trend of the reference capacitance value of the sensor body 4 (i.e., the initial capacitance value under waterless conditions), which may indicate sensor aging, contamination, or packaging failure; the transmission success rate of the multimode wireless communication module 6 (by statistically analyzing the ratio of sent / received acknowledgments); the charge / discharge cycle count, internal resistance change rate, and predicted remaining lifespan of the energy storage unit; and whether the internal temperature and humidity of the module exceed normal operating ranges. These parameters are collected in real time and compared with preset health thresholds, or input into anomaly detection algorithms based on lightweight machine learning models (e.g., a small support vector machine (SVM) or decision tree model trained on baseline data collected during normal device operation) to determine whether the device is experiencing performance degradation or potential malfunctions. Once any anomaly is detected, such as a sensor baseline drift exceeding 10% or a communication success rate below 95%, the integrated diagnostic and calibration module 8 immediately sends a detailed diagnostic report or warning to the central management platform 3 via the multimodal wireless communication module 6. This report includes the device ID, anomaly type, specific parameter values, occurrence time, and suggested handling measures. The calibration function supports remote triggering or periodic self-calibration. For example, the system can perform a self-calibration process during a preset waterless period (e.g., 3 AM to 4 AM) or upon receiving a remote command: first, ensuring the sensor body 4 is dry, and then recording the current capacitance value of the sensor body 4 as the new baseline value. The new baseline value is then compared with historical data, and a compensation algorithm (e.g., a polynomial fitting compensation model based on temperature and humidity) is used to correct for sensor drift caused by environmental changes or aging, ensuring long-term accuracy of moisture measurement.

[0027] The wireless data gateway unit 2, acting as a bridge between the leak detection unit 1 and the central management platform 3, receives wireless data from multiple leak detection units 1, aggregates and pre-processes it, and then securely and efficiently uploads it to the central management platform 3. This wireless data gateway unit 2 is equipped with a multi-mode wireless receiving module that matches the multi-mode wireless communication module 6 of the leak detection unit 1. This module can simultaneously or alternately receive Sub-GHz (LoRa / NB-IoT) and UWB signals. Specifically, it includes independent Sub-GHz and UWB receiving links and is equipped with multiple RF switches and antenna arrays to optimize signal reception performance. The wireless data gateway unit 2 possesses advanced signal processing capabilities, including but not limited to: multipath signal combining techniques for Sub-GHz signals, such as Maximum Ratio Combining (MRC) or Selection Combining (SC) algorithms. When receiving the same data copy from different spatial paths, the MRC algorithm weights and superimposes the data based on the signal-to-noise ratio (SNR) of each copy to maximize the SNR of the synthesized signal, thereby significantly improving the data reception success rate in strong multipath fading environments. For UWB signals, the wireless data gateway unit 2 can perform high-time-resolution pulse acquisition and time-of-arrival (ToA) or time-of-difference (TDoA) measurement, which is crucial for accurate positioning. Furthermore, the wireless data gateway unit 2 performs interference suppression (e.g., avoiding other interference within the ISM band through adaptive filters or spectrum analysis), forward error correction (FEC) decoding (e.g., decoding of convolutional codes or LDPC codes), and data integrity verification (e.g., CRC check) to ensure the accuracy and reliability of received data. The wireless data gateway unit 2 also features a local data caching function, integrating a 1GB to 4GB eMMC or SPI NAND Flash memory. This cache is used to temporarily store received leak data, diagnostic information, and alarm events when the network connection with the central management platform 3 is interrupted (e.g., Ethernet cable disconnection or loss of cellular network signal). Once the network connection is restored, the wireless data gateway unit 2 automatically initiates a data synchronization mechanism, uploading the cached data to the central management platform 3 in batches according to time sequence, ensuring no data loss.The wireless data gateway unit 2 integrates a high-performance microcontroller, such as an embedded processor based on the ARM Cortex-A series (e.g., NXPi.MX6UL), with a main frequency of 600MHz to 1GHz. This microcontroller manages the gateway's own operating status, performs initial timestamping and encryption of received leakage data using AES-128 or AES-256 algorithms, and executes data synchronization protocols with the central management platform 3, such as MQTT (Message Queuing Telemetry Transport) or HTTPS (Hypertext Transfer Protocol Secure). The wireless data gateway unit 2 connects to the central management platform 3 in various ways, including through a standard Ethernet interface (RJ45), Wi-Fi (IEEE 802.11n / ac standard), or a cellular network module (such as a communication module supporting 4G LTE Cat M1 or 5G NR). The wireless data gateway unit 2 also supports some edge computing functions, such as performing preliminary location calculations on received water leakage events locally (using a simple location algorithm based on a small number of UWB anchor points or RSSI fingerprints), or executing simple linkage control logic (such as controlling nearby local valves to close in advance by outputting a relay signal when a local area water leakage is detected), thereby reducing the computational burden on the central management platform 3 and reducing the system's response delay to water leakage events.

[0028] The central management platform 3 is the core intelligent hub of the entire system, used to receive, store, analyze, and display data from all wireless data gateway units 2, and to provide early warning, location, and management of water leakage events. This platform 3 is typically deployed on cloud servers (such as AWS, Azure, or Alibaba Cloud) or local server clusters. Its core functional modules include: a data storage and management module, a data analysis and location module, a visualization and alarm module, and a user interaction interface.

[0029] The data storage and management module is designed to persistently store all leak detection data, sensor diagnostic data, network topology information, and historical alarm records. This module employs a distributed database architecture, such as a NoSQL database based on Apache Cassandra or MongoDB, to support the storage and efficient querying of petabyte-scale data, while maintaining high availability and scalability. All data is encrypted before storage and is regularly backed up to prevent data loss. The data model is meticulously designed, including fields such as sensor ID, timestamp, moisture level, battery voltage, signal strength, fault code, alarm level, and location coordinates.

[0030] The data analysis and location module is one of the core innovations of this invention. Its design goal is to achieve precise location of the leak and intelligent identification of the leak source. This module 10 integrates three key sub-modules: First, an indoor positioning engine based on multi-source fusion. This engine combines UWB ranging data, RSSI (Received Signal Strength Indicator) fingerprint data, and hop count information from the sensor network. UWB ranging data is acquired through Two-Way Ranging (TWR) or Time Difference of Arrival (TDoA) techniques, providing high-precision distance information. The RSSI fingerprint is constructed by offline measurement of the target area before deployment, forming a database recording RSSI characteristics at different locations. During online positioning, location estimation is performed using K-Nearest Neighbor (K-NN) or probabilistic matching algorithms. The hop count information provides a rough distance indication of the network topology between the leak detection unit 1 and the wireless data gateway unit 2. The positioning engine 13 uses Weighted Least Squares (WLS), Extended Kalman Filter (EKF), or Particle Filter (PF) algorithms to fuse these multi-source heterogeneous data to estimate the precise location of the alarm leak detection unit 1 in real time. The WLS algorithm minimizes positioning error by assigning weights to different data sources; the EKF algorithm estimates the state iteratively by predicting and updating; and the PF algorithm estimates the state in a nonlinear system by randomly sampling particles. The positioning engine 13 preloads detailed Building Information Modeling (BIM) data, typically imported in IFC (Industry Foundation Classes) format, including 2D / 3D structural diagrams of the building (such as floor plans and elevations), pipe routing diagrams (precise 3D paths of water supply, drainage, and HVAC pipes), wall material information (such as concrete, gypsum board, and brick walls, and their dielectric constants and attenuation characteristics), and equipment layout diagrams (such as the locations of pumps, valves, and fire hydrants). BIM data not only provides static geometric and semantic information but is also used to construct UWB signal propagation models (e.g., considering wall attenuation and multipath reflection effects) and assist the positioning engine in identifying and correcting non-line-of-sight (NLOS) errors, thereby significantly improving positioning accuracy in complex indoor environments.

[0031] Second, a leak source inference module based on a fluid dynamics model. When one or more leak detection units 1 alarm, the inference module 14 utilizes the building structure information (such as the location and height of partition walls, floor height differences, ground drainage slope, and permeability coefficients of wall and floor materials) and pipe layout information contained in the aforementioned BIM data, combined with the preliminary location coordinates of the leak point, to intelligently infer the diffusion direction of water flow and possible leak source areas through a simplified water flow path simulation algorithm. This simulation algorithm can be based on the grid method or finite element method, dividing the building space into a fine grid to simulate the flow and diffusion of water droplets under the influence of gravity, surface tension, capillary action (for porous materials such as concrete and gypsum board), and material absorption characteristics (such as water absorption rate and saturation). For example, the water flow simulation can consider gravity-driven diffusion along slopes, as well as obstruction when encountering walls or capillary climbing along wall corners and pipe outer walls. During the inference process, module 14 can refer to preset water flow diffusion model parameters, which can be laboratory-calibrated based on different building materials and water flow rates. In addition, the system can also perform machine learning training using historical leak event data. For example, by using support vector machines or simple neural network models, it can learn to associate known leak points and diffusion patterns with actual leak sources to optimize the accuracy of future inferences.

[0032] Third, an abnormal event correlation analysis module. This module 15 performs in-depth correlation analysis between water leakage alarm data and related building equipment operation data from the Building Automation System (BAS), further narrowing down the scope of the leakage source. BAS data is typically integrated via protocols such as BACnet / IP, Modbus / TCP, or OPC UA, acquiring information such as pump start / stop status, water flow valve opening, pipeline pressure sensor data, liquid level sensor data, and HVAC system operating status. For example, if a water leakage alarm is accompanied by an abnormal drop in water supply pipeline pressure in a certain area (e.g., a sudden drop from the normal value of 0.4 MPa to 0.2 MPa), the system will significantly increase the probability score of a specific section of the water supply pipeline being a ruptured leak source. Through this fusion and cross-validation of multi-source data, the system can improve the accuracy of leak source inference from a vague regional level to a specific part of a specific pipe or equipment, and even identify whether the damage is to a pipe joint, valve gasket, or the pipe itself.

[0033] The visualization and alarm module is used to visually display the location and related information of a leak through a graphical user interface (GUI). This module 11 can render a 3D view of the BIM model in real time on a web browser using technologies such as WebGL or Three.js, or on a 2D plan. When a leak alarm occurs, the system can accurately mark the location of the alarming leak detection unit 1 through highlighting, animation (such as a spreading water wave animation), or custom icons, and simultaneously display the inferred leak source area (e.g., using area blocks of different colors or transparency to indicate probability). Module 11 supports multiple alarm methods to ensure timely information delivery: including but not limited to SMS notifications, email notifications, mobile application (App) push notifications, and linkage with building automation systems (BAS) or fire alarm systems via API interfaces or OPC UA / BACnet protocols to trigger a higher level of response. The alarm information includes the precise coordinates of the leak location, the inferred leak source area, the alarm time, the ID of the affected sensor, and the real-time health status of the sensor (such as battery level and signal strength), providing managers with comprehensive decision-making information.

[0034] The user interface provides a comprehensive set of system management functions, including administrator permission management (multi-level user permission control), system configuration (remote configuration of sensor thresholds, alarm rules, and communication parameters), historical data query (supporting multi-dimensional filtering by time period, region, device ID, etc.), report generation (generating PDF or CSV reports such as daily, weekly, and monthly leak event statistics and device health reports), and remote device diagnostic and calibration command issuance. This interface enables maintenance personnel to manage and maintain the entire system efficiently and conveniently.

[0035] This invention also provides a wireless water leakage detection and alarm method for concealed spaces inside buildings, such as... Figure 3 As shown, the method specifically includes the following steps: Step 1: Deployment and Initialization of Leak Detection Units. This step is carried out in concealed spaces within the building. These concealed spaces include, but are not limited to, the interior of walls, above ceilings, under floors (such as raised floors and underfloor heating layers), equipment compartments, pipe shafts, cable trays, ventilation ducts, and damp-risk areas such as under sink cabinets and behind toilets in kitchens, bathrooms, and laundry rooms. During deployment, maintenance personnel distribute multiple leak detection units 1 according to a preset topology (e.g., densely deployed in a grid pattern in critical areas, or linearly deployed along pipes and equipment areas) or based on empirical rules (denser deployment in high-risk areas and sparse deployment in low-risk areas). To ensure accurate subsequent location, the initial geographical location information of each leak detection unit 1 is accurately recorded. This can be accomplished using various high-precision measurement tools, such as using a laser rangefinder for 3D coordinate measurement, or combining it with an RTK (Real-Time Kinematic) GPS system for sub-centimeter-level accurate coordinate acquisition. These coordinate information are then input into the central management platform 3 and spatially mapped and bound to the pre-imported BIM model to ensure that each sensor has a precise corresponding location in the digital twin environment. After deployment, each leak detection unit 1 immediately performs self-testing and initialization, which includes baseline calibration of the sensor body 4 (measuring and recording the initial capacitance value under waterless conditions), channel scanning and power adaptive adjustment of the multimodal wireless communication module 6 (e.g., scanning LoRa or NB-IoT channels to select the channel with the least interference and estimating the required transmit power based on the link budget), and establishing a preliminary, secure communication link with the nearest wireless data gateway unit 2 (e.g., through key exchange and authentication).

[0036] Step Two: Multimodal Wireless Communication and Data Transmission. During normal system operation, each leak detection unit 1 collects information such as moisture status data, its own health status data, and battery level according to a preset period (e.g., every 5 minutes, 10 minutes, or 30 minutes) or based on event triggers (e.g., a sudden change in sensor capacitance exceeding a preset threshold, indicating a potential leak). This data is then transmitted to the wireless data gateway unit 2 via the multimodal wireless communication module 6. Specifically, the leak detection unit 1 dynamically selects Sub-GHz (e.g., LoRa) or UWB communication mode based on real-time channel status information, such as received signal strength index (RSSI), signal-to-noise ratio (SNR), link quality index (LQI), and packet error rate (PER). When channel quality deteriorates (e.g., RSSI is below -110dBm) or stronger penetration and coverage are required, Sub-GHz mode is preferentially selected for transmission, utilizing its stronger diffraction capability and wider coverage. When the system requires high-precision ranging and positioning (typically triggered by a command issued by the central management platform 3 after a water leak alarm is detected), it switches to UWB mode, utilizing its high time resolution and short pulse transmission characteristics to achieve centimeter-level distance measurement accuracy. The multi-mode wireless communication module 6 employs adaptive modulation and coding (AMC) technology when transmitting data, dynamically adjusting the modulation order (e.g., from BPSK, QPSK to 16QAM, 64QAM) and coding rate (e.g., from 1 / 2 to 3 / 4) according to current channel conditions. This maximizes the transmission rate or minimizes the transmit power while maintaining a low bit error rate, achieving energy efficiency optimization. Simultaneously, to effectively address multipath fading, spatial diversity (e.g., MRC combining at the receiver) or transmit diversity (e.g., encoding at the transmitter using Alamouti codes) is achieved through multi-antenna arrays. Data packets are decomposed into multiple copies and transmitted through different antennas or different frequency / time / coding channels, significantly improving the robustness of data transmission. Wireless data gateway unit 2 receives wireless signals from multiple leak detection units 1. For each received multipath signal copy, wireless data gateway unit 2 performs advanced signal fusion processing, such as using the Maximum Ratio Combining (MRC) algorithm to weighted superimpose all received valid signal copies to maximize the signal-to-noise ratio of the synthesized signal, thereby recovering the original leak detection data. The data undergoes preliminary decryption at the gateway (using a pre-shared key or digital certificate for TLS / DTLS encrypted connection), integrity verification (CRC or SHA256 hash), and high-precision timestamp marking (synchronized via an NTP server). Subsequently, the processed data is securely uploaded to the central management platform 3 via wired (Ethernet) or cellular network (4G / 5G).

[0037] Step 3: Intelligent Identification and Adaptive Early Warning of Leakage Events. After receiving leakage detection data from the wireless data gateway unit 2, the central management platform 3 first performs data parsing and verification to ensure that the data format is correct and has not been tampered with. Subsequently, the data analysis and positioning module performs real-time and continuous analysis of the received moisture status data. When the moisture value (e.g., the change in capacitance value) measured by the sensor body 4 of any one or more leakage detection units 1 continuously exceeds a preset leakage threshold, the central management platform 3 determines that a leakage event has occurred. This threshold is not a fixed value, but can be dynamically adjusted or automatically learned through machine learning algorithms based on the actual building environment, the water absorption characteristics of the installation materials (e.g., the different responses of wooden floors and concrete structures to moisture), and historical data. For example, the threshold can be set as a slight leakage if the sensor capacitance value increases by 0.1pF relative to the reference value, a moderate leakage if it increases by 0.5pF, and a severe leakage if it increases by 2.0pF. Once a leakage event is identified, the system triggers a multi-level adaptive early warning mechanism. Based on the leakage level (minor leak, moderate leak, severe leak) and the importance of the affected area (e.g., high priority for important equipment rooms such as data centers, archives, and power distribution rooms; medium priority for public corridors and general office areas), the system can automatically select different alarm levels and notification methods. For example, for minor leaks, it may only notify designated management personnel through internal system messages and emails; for moderate leaks, it will trigger SMS notifications and mobile app push notifications; for severe leaks or high-priority areas, the system will immediately trigger multi-channel alarms (SMS, telephone calls) and, through linkage with the building automation system (BAS), simultaneously initiate emergency drainage procedures, automatically shut off water supply valves in the relevant areas, or cut off power to specific equipment to minimize losses. The warning information includes a preliminary leak alarm unit identification code and an accurate timestamp, as well as preliminary area location information.

[0038] Step Four: Precise Location and Source Inference of Leakage. After a leak is identified, the data analysis and location module immediately initiates a high-precision location and source inference process. First, based on the indoor positioning engine 13, the system uses the UWB signal emitted by the alarm leak detection unit 1 (or triggers it to enter UWB ranging mode via a specific command) to measure the distance between the signal and the pre-set UWB anchor points (e.g., an anchor point network composed of Decawave DW1000 series chipsets with a positioning accuracy of ±10cm) located inside the building and precisely corresponding to the spatial coordinates of the BIM model. Using a trilateration algorithm or a more complex nonlinear optimization algorithm (such as the Levenberg-Marquardt algorithm), the two-dimensional or three-dimensional spatial coordinates of the alarm leak detection unit 1 are precisely calculated. Simultaneously, the advantages of UWB ranging accuracy being less affected by the environment (high multipath resistance) and a pre-constructed RSSI fingerprint map (containing RSSI measurements from hundreds of reference points) are used as auxiliary correction data sources. For example, UWB provides absolute distance, while RSSI fingerprinting provides relative location and environmental feature information. The fusion of these two technologies can further improve positioning accuracy and robustness in complex indoor environments. Secondly, the positioning engine 13 matches the calculated precise location with the BIM model, accurately displaying the alarm point in a highlighted or animated form within the digital twin environment (e.g., on a 3D rendered building model). Thirdly, the fluid dynamics model leakage source inference module 14, based on the precise location of the alarm point and combined with detailed information from the BIM model regarding building structure (such as the precise geometric location, thickness, and material properties of walls, and the structure and drainage slope of floors) and piping system (detailed 3D geometric information, pipe diameter, and material of water supply, drainage, fire protection, and HVAC pipelines), utilizes a preset fluid diffusion model. This model can simulate water flow using a simplified grid method or finite element method, considering the influence of gravity, surface tension, capillary action, and material absorption characteristics on the water flow path, tracing back to the most likely origin point or area of ​​the water flow. For example, the simulated water flow will diffuse along the lowest slope, bypassing or seeping through obstacles, and the time and path to each sensor are calculated. During this process, the system also references pattern recognition results from historical leakage data, using machine learning algorithms (such as clustering-based algorithms to identify common leakage patterns, or rule-based reasoning systems) to identify the correlation between specific leakage patterns and water sources, further optimizing the accuracy of inferences. Finally, the abnormal event correlation analysis module performs in-depth correlation analysis between leakage alarm information and real-time data from the building automation system (BAS), including water pressure sensor data, water flow sensor data, valve status, and pump start / stop status. For example, if a leak occurs near a water supply pipe in a certain area, and the BAS simultaneously detects a sudden drop in water pressure in that area (e.g., a drop of 0.01 MPa per second), the system will significantly increase the probability score of that water supply pipe being the source of the leak.Through the fusion and cross-validation of multi-source data, the system can improve the accuracy of inferring the source of leakage from the regional level to a specific part of a particular pipe or equipment, such as locating a valve, pipe weld, or specific connection point.

[0039] Step 5: Fault Diagnosis and Preventive Maintenance. The data analysis and location module of the central management platform 3 continuously monitors the health status data of each leak detection unit 1. This data is reported periodically through the integrated diagnostic and calibration module 8, including but not limited to key operating parameters such as the reference drift of the sensor body 4, the remaining battery power, energy harvesting efficiency, the signal transmission success rate (i.e., the data packet reception confirmation rate) of the wireless communication multimode wireless communication module 6, and the internal temperature and humidity of the equipment. When any parameter deviates from the preset health threshold (e.g., battery voltage below 3.0V or sensor reference drift exceeding 15%) or is reported as abnormal through the integrated diagnostic and calibration module 8 (e.g., "sensor aging", "battery life nearing its end", "communication link unstable"), the central management platform 3 will automatically generate a fault warning and send a detailed diagnostic report to maintenance personnel. This diagnostic report typically includes: device ID, precise location, current value and historical trend of abnormal parameters, diagnostic results (e.g., "high risk of sensor corrosion", "low battery power, replacement recommended", "intermittent communication link interruption"), and recommended maintenance measures (e.g., "check if the sensor body is contaminated", "arrange battery replacement", "check if there are new sources of interference in the communication environment"). Furthermore, through long-term analysis of historical health data and past leakage events, the system can utilize predictive analytics models (e.g., battery life prediction based on time series analysis or regression models, or failure mode identification based on equipment operating characteristics) to identify potential failure modes and degradation trends of the equipment, thereby supporting the implementation of preventative maintenance strategies. For example, based on the average battery life prediction of a sensor group in a certain area, the system can schedule batch battery checks or replacements several months in advance, thus transforming reactive maintenance into proactive predictive maintenance, significantly reducing unplanned downtime, and minimizing the risk of leakage and repair costs due to equipment failure. The system also supports remote parameter configuration (such as data acquisition frequency and alarm thresholds) and firmware over-the-air (FOTA) upgrades for the leakage detection unit 1, ensuring that the equipment always operates in optimal condition and receives timely functional updates and security patches. Example

[0040] This embodiment aims to further illustrate the actual performance and advantages of the present invention, "A Wireless Method and System for Detecting and Alarming Leakage in Concealed Spaces Inside Buildings," through specific scenarios and quantitative data.

[0041] In a typical commercial office building, 200 leakage detection units 1 of this invention are deployed in key areas, including the underground data center (below the raised floor), multimedia server room (above the ceiling), archives (inside the walls), and the pipe shafts and valve areas of the core electromechanical equipment layer. These units are deployed according to a strategy combining a pre-defined grid and pipeline paths, with the initial position of each unit precisely mapped using a laser rangefinder and a BIM model. Ten wireless data gateway units 2 are deployed in the system, evenly distributed across different floors and areas, and connected to a central management platform 3 located in the cloud via the building's existing Ethernet backbone network. The BIM model includes detailed structural information of the building, the routing of HVAC ducts, fire protection system ducts, the materials and layout of water supply and drainage pipes, and the importance level of each functional area.

[0042] After deployment, all leak detection units 1 underwent self-testing and initialization, with the average reference capacitance value of their sensor bodies 4 stabilizing at 50pF ± 0.5pF. The multi-mode wireless communication module 6 primarily uses Sub-GHz LoRa mode in daily monitoring, achieving an average transmission success rate of over 98%. Battery life prediction indicates that, with data collection and transmission frequency of 12 times per day (once every 2 hours), it can support over 7 years of maintenance-free operation on average.

[0043] Scenario 1: Data center air conditioning condensate leak At 2:37 AM, a water leakage detection unit 1, numbered DC-L2-S038 (deployed inside the raised floor beneath an air conditioning unit), located in the data center area on the second basement level of the building, detected abnormal moisture. The capacitance of the sensor body 4 rapidly increased from 50.2 pF to 53.8 pF within 30 seconds, triggering the moderate water leakage alarm threshold. The low-power microprocessor 5 immediately transmitted this event information to the nearest wireless data gateway unit 2 via LoRa mode.

[0044] After receiving the alarm information, the wireless data gateway unit 2 performs preliminary processing and uploads the information to the central management platform 3. The data analysis and positioning module of the central management platform 3 receives the alarm, identifies it as a moderate water leakage event, and immediately triggers a high-priority alarm mechanism. Simultaneously, the positioning engine 13 sends a UWB ranging command to the DC-L2-S038 unit. The DC-L2-S038 unit switches to UWB mode and performs bidirectional ranging with three nearby UWB anchor points (Point A: X = 10.5m, Y = 5.2m, Z = 0.8m; Point B: X = 15.0m, Y = 6.0m, Z = 0.8m; Point C: X = 12.0m, Y = 10.0m, Z = 0.8m), obtaining distances of 2.5m, 3.8m, and 4.2m respectively. Combining these UWB ranging data with the pre-imported BIM model, the positioning engine 13 accurately calculates the precise three-dimensional coordinates of the DC-L2-S038 leakage detection unit 1 using the weighted least squares method as X = 11.5m, Y = 7.0m, Z = 0.8m (positioning error less than 10cm).

[0045] Subsequently, the leakage source inference module of the fluid dynamics model, based on this precise location and combined with information from the BIM model regarding the routing of air conditioning condensate pipes and drainage pipes under the data center's raised floor, the floor slope, and the characteristics of the pipe material (PVC) and floor material (anti-static flooring with a certain degree of hydrophobicity), simulated the water flow diffusion path. The simulation results showed that the water flow diffused from this point in a north-northeast direction and traced upstream. Simultaneously, the abnormal event correlation analysis module correlated with the data center building automation system (BAS). The system detected that approximately one minute before the DC-L2-S038 alarm, the condensate pump operation status of the adjacent air conditioning unit (CRAC-05) suddenly changed from "normal" to "alarm," and its condensate pan level sensor reading exceeded the preset upper limit; the pump failed to start after the alarm.

[0046] Based on data from multiple sources, the central management platform 3 ultimately determined that the source of the leak was a "blockage or rupture in the condensate drain pipe of air conditioning unit CRAC-05," which was precisely displayed on the BIM 3D model. The leaking area was highlighted, and the possible direction of water flow was indicated. The system immediately notified the operations supervisor and on-duty engineer via SMS, email, and app, including detailed leak location coordinates, the deduced source, alarm time, and relevant BAS data. The on-duty engineer arrived at the site within 10 minutes of receiving the alarm and, using the indicators in the BIM model, directly located the CRAC-05 unit, confirmed the overflow of its condensate pan, and found a slight blockage in the drain pipe. The entire process, from alarm to accurate location, source deduction, and manual response, was extremely short, effectively preventing further damage to the data center equipment caused by the leak.

[0047] Scenario 2: Early warning and preventive maintenance of aging pipelines and leaks During the six-month operation period, the fault diagnosis and preventive maintenance module of the central management platform 3 continuously monitored the health status of all leak detection units 1. The system found that the leak detection unit 1, numbered BA-L5-S120 (deployed in the wall behind the toilet), located in a bathroom (non-critical area) on the fifth floor of the building, had its sensor body 4's reference capacitance value gradually drift from an initial 50.1pF to 51.5pF. Although this did not reach the leak alarm threshold, it exceeded the preset "sensor reference drift warning" threshold (1.0pF). At the same time, the transmission success rate of the unit's wireless communication multimode wireless communication module 6 dropped from 99% to 96%, and although the battery voltage was still within the safe range, its rate of decline was slightly higher than the average.

[0048] The diagnostic module 10 analyzes historical data and, based on subtle trends in these parameters, uses a built-in anomaly detection model to identify potential "aging and leakage risks" and "minor battery performance degradation" in the unit. The system generates a diagnostic report, recommending that maintenance personnel pay close attention to this area during the next routine inspection.

[0049] During a subsequent routine inspection, maintenance personnel, based on the precise location reported by the system, inspected the wall containing unit BA-L5-S120 and discovered slight, barely noticeable dampness on the wall surface. Further inspection confirmed a very minute, chronic leak at the connection point of the embedded water supply pipe, virtually undetectable to the naked eye. Thanks to this early warning, maintenance personnel repaired the leak before it worsened, replacing the aging pipe joint and preventing a larger leak and subsequent repair costs. Simultaneously, the unit's battery was also replaced, ensuring its long-term stable operation.

[0050] This embodiment fully demonstrates the invention's outstanding ability to accurately detect, precisely locate, intelligently infer the source of water leakage, and provide proactive fault warnings in practical applications.

[0051] Comparative Example To further highlight the technical advantages of this invention, a comparative example is provided below with a common leak detection system in the prior art. This comparative example uses a traditional resistive leak sensor and single-mode (such as Wi-Fi) wireless communication, and lacks high-precision positioning and intelligent analysis capabilities.

[0052] Comparative Example 1: Traditional resistive sensors vs. a single Wi-Fi communication system In a commercial office building similar to the example described above, 200 traditional resistive water leakage sensors are deployed. These sensors typically consist of two exposed metal electrodes; when water comes into contact with the electrodes, the resistance decreases, triggering an alarm. Wireless communication uses a standard Wi-Fi module (IEEE 802.11n) connected to the building's existing Wi-Fi network. The system does not include UWB positioning capabilities; positioning primarily relies on Wi-Fi RSSI fingerprinting or simple AP location association. The central management platform has basic functionality, providing only alarm display and historical data query.

[0053] Scenario 1 (Simulation): Data Center Air Conditioning Condensate Leakage Suppose that in the same data center area, a traditional resistive sensor, model DC-L2-S038, detects a water leak. Because resistive sensors require direct water flow between the two electrodes to trigger, if the leak is a slow seepage or a small amount of water droplets, it may not trigger an alarm in time. Even if it does trigger, the sensor electrodes are highly susceptible to electrochemical corrosion and oxidation in long-term humid environments, leading to decreased sensitivity or false alarms.

[0054] Upon detecting a leak, the unit sends an alarm via its Wi-Fi module. In the complex electromagnetic environment of a data center, filled with numerous metal structures and server racks, the multipath effect and attenuation of Wi-Fi signals are severe. Testing revealed numerous dead zones and weak signal areas in the deployed environment, with a packet loss rate as high as 15%-25%. This could prevent alarm information from being transmitted to the central management platform in a timely and reliable manner.

[0055] Even if the alarm information reaches the platform, due to the lack of a UWB positioning module, the platform can only determine the approximate coverage area of ​​which Wi-Fi AP the leaking unit is located in based on Wi-Fi RSSI or a rough AP association, such as "the data center area on the second basement level". The positioning accuracy is low, usually within a range of 5-10 meters or even larger, and cannot pinpoint the exact location of the cabinet or equipment underneath.

[0056] More importantly, the system lacks fluid dynamics modeling or abnormal event correlation analysis capabilities. The platform can only display "DC-L2-S038 alarm" and cannot intelligently deduce the source of the leak. After receiving the alarm, maintenance personnel must manually check all air conditioning units, pipes, and equipment in the area one by one to find the source of the leak. This is not only time-consuming and labor-intensive but may also delay fault handling, leading to greater economic losses. For example, the location and source identification that can be completed within 10 minutes in the above embodiment may take 1-2 hours or even longer in this comparative example, causing condensate overflow to further damage critical servers.

[0057] Scenario 2 (Simulation): Pipeline Aging and Leakage Early Warning For chronic pipe leaks like those in the BA-L5-S120 example, traditional resistive sensors offer virtually no early warning. Because the leak is slow, water may only wet the absorbent material around the sensor, failing to form a continuous water film sufficient to conduct electricity between the two electrodes. The sensor only triggers when the leak accumulates enough to directly contact and conduct electricity to the electrodes, by which time the leak has often been ongoing for a considerable time, even causing localized damage. The sensor itself lacks internal diagnostic and calibration modules, making it unable to monitor performance drift or battery health. Maintenance personnel can only detect problems through regular inspections or after visible leaks or even serious malfunctions, completely lacking preventative maintenance capabilities. Regarding battery life, traditional Wi-Fi modules consume significantly more power than LPWAN modules, typically lasting only a few months to a year, requiring frequent battery replacements and increasing maintenance costs.

[0058] By comparing this embodiment with the comparative example, the present invention demonstrates significant technical superiority in terms of sensor durability, wireless communication robustness, accurate leak location, intelligent source inference, and preventive maintenance capabilities.

[0059] The table below summarizes the key performance indicators of the present invention and the comparative example: Through the detailed description and data comparison of the above embodiments and comparative examples, it is fully demonstrated that the present invention, "a wireless water leakage detection and alarm method and system for concealed spaces inside buildings," has significant technical advancement and practicality. It can effectively solve key problems in the prior art and provide a more reliable, efficient, and intelligent solution for building safety management.

Claims

1. A wireless water leakage detection and alarm system for concealed spaces inside buildings, characterized in that, include: Multiple leak detection units (1) are deployed in a distributed manner to monitor the moisture status of their respective areas in real time and transmit the monitoring data through a wireless network; At least one wireless data gateway unit (2) is coupled to the plurality of water leakage detection units (1) for receiving and aggregating wireless data from the water leakage detection units (1) and uploading the data; And a central management platform (3), coupled to the wireless data gateway unit (2), for receiving, storing, analyzing and displaying data from the wireless data gateway unit (2), and for early warning, location and management of water leakage events; wherein, the water leakage detection unit (1) includes: a sensor body (4), which uses non-contact capacitive or photoelectric sensing principle for moisture detection, and its encapsulation material is polytetrafluoroethylene or polyvinylidene fluoride; a low-power microprocessor (5), coupled to the sensor body (4), for performing digital conversion, preprocessing and local analysis of the raw data collected by the sensor, and in ultra-low power standby mode during non-working cycles; a multi-mode wireless communication module (6), coupled to the low-power microprocessor (5), equipped with The system is equipped with at least two wireless communication sub-modules with different frequency bands or different communication protocols to realize data transmission between the water leakage detection unit (1) and the wireless data gateway unit (2), and to adaptively select the communication mode according to real-time environmental channel status information or preset communication strategy through an intelligent switching algorithm; an energy harvesting and management module (7) coupled to the low-power microprocessor (5) and the multi-mode wireless communication module (6) to provide sustainable energy supply for the water leakage detection unit (1); and an integrated diagnostic and calibration module (8) coupled to the low-power microprocessor (5) and the sensor body (4) to monitor the internal operating status of the water leakage detection unit (1) in real time and to perform fault warning and performance calibration.

2. The wireless water leakage detection and alarm system for concealed spaces inside buildings according to claim 1, characterized in that, The sensor body (4) specifically includes: a sensing element composed of staggered microelectrodes, wherein the electrode array substrate material is selected from FR4 or polyimide flexible substrate; And a hydrophilic polymer film coated on the surface of the microelectrode. After the film absorbs water, its dielectric constant changes, thereby causing a change in the capacitance value of the sensor body (4). The amount of change is positively correlated with the moisture content. The low-power microprocessor (5) integrates a high-precision analog-to-digital converter and a programmable logic controller for periodically waking up the sensor body (4), executing data acquisition instructions, filtering, and performing preliminary moisture threshold judgment. The low-power microprocessor (5) can be woken up from standby mode by external interrupt or timer.

3. The wireless water leakage detection and alarm system for concealed spaces inside buildings according to claim 1, characterized in that, The multimodal wireless communication module (6) is configured with at least two wireless communication sub-modules of different frequency bands or different communication protocols, including: a sub-module that supports Sub-GHz low-power wide-area network communication protocols, including LoRa or NB-IoT, for long-distance data transmission; And another sub-module supporting ultra-wideband communication protocol, following the IEEE 802.15.4a standard, is used to realize high-precision ranging and short-range high-bandwidth data transmission; the multi-mode wireless communication module (6) dynamically selects Sub-GHz or UWB communication mode based on channel state information estimated by parameters such as real-time received signal strength indication, signal-to-noise ratio, bit error rate or packet error rate through intelligent switching algorithm; the transmit power, modulation order and coding rate of the multi-mode wireless communication module (6) can be dynamically adjusted by the low-power microprocessor (5) according to the real-time signal-to-noise ratio and bit error rate to realize link self-adaptation; the multi-mode wireless communication module (6) is also equipped with multiple sets of orthogonal polarized antennas or spatially separated antenna arrays to realize spatial diversity or multiple-input multiple-output communication.

4. The wireless water leakage detection and alarm system for concealed spaces inside buildings according to claim 1, characterized in that, The energy harvesting and management module (7) includes at least one photovoltaic panel and / or one thermoelectric generator. The photovoltaic panel is used to harvest light energy in a concealed space with light, and the thermoelectric generator is used to harvest heat energy from the surface of the pipe or the temperature difference of the environment. The energy harvesting and management module (7) also includes an energy storage unit, which uses a solid-state lithium-ion battery or a supercapacitor to store the harvested energy and provide power when the energy harvesting is insufficient. The energy management circuit integrates a maximum power point tracking controller and has overcharge, over-discharge and short-circuit protection functions. The integrated diagnostic and calibration module (8) includes an internal temperature sensor, a humidity sensor and an internal power supply voltage monitor. The diagnostic function judges whether the equipment performance has degraded by monitoring parameters such as the drift of the reference capacitance value of the sensor body (4), the transmission success rate of the communication module and the number of charge and discharge cycles of the battery, combined with a preset health threshold or an anomaly detection model based on machine learning. The calibration function supports remote or periodic self-calibration by recording the sensor reference value and correcting the sensor drift under waterless conditions.

5. The wireless water leakage detection and alarm system for concealed spaces inside buildings according to claim 1, characterized in that, The wireless data gateway unit (2) is equipped with a multimode wireless receiving module that matches the multimode wireless communication module (6) of the water leakage detection unit (1). This receiving module can simultaneously or alternately receive Sub-GHz and UWB signals. The wireless data gateway unit (2) has advanced signal processing capabilities, including multipath signal combining, interference suppression, forward error correction decoding, and data integrity verification. The wireless data gateway unit (2) also has a local data caching function, which is used to temporarily store data when the connection with the central management platform (3) is interrupted, and to synchronize the data after the connection is restored. The wireless data gateway unit (2) has a built-in high-performance microcontroller, which is used to perform preliminary timestamp marking and encryption processing on the received water leakage data, and to execute the data synchronization protocol with the central management platform (3). The wireless data gateway unit (2) also supports some edge computing functions, including performing localized preliminary location calculations on the received water leakage events, or executing simple linkage control logic.

6. The wireless water leakage detection and alarm system for concealed spaces inside buildings according to claim 1, characterized in that, The central management platform (3) includes: a data storage and management module for persistently storing all leak detection data, sensor diagnostic data, network topology information and historical alarm records, the module adopting a distributed database architecture; a data analysis and positioning module for accurately locating the leak location and intelligently identifying the leak source; and a visualization and alarm module for intuitively displaying the location and related information of the leak event through a graphical user interface, and supporting multiple alarm methods. It also includes a user interface that provides administrator permission management, system configuration, historical data query, report generation, and remote device diagnostic and calibration command issuance functions.

7. The wireless water leakage detection and alarm system for concealed spaces inside buildings according to claim 6, characterized in that, The data analysis and positioning module includes: an indoor positioning engine based on multi-source fusion, which combines UWB ranging data, RSSI fingerprint spectrum data, and hop count information of sensor networks, and uses weighted least squares, extended Kalman filtering, or particle filtering algorithms to estimate the precise location of the alarm leak detection unit (1) in real time; the positioning engine is preloaded with building information model data, including two-dimensional / three-dimensional structural diagrams of the building, pipe routing diagrams, wall material information, and equipment layout diagrams; and a leak source inference module based on a fluid dynamics model, which, when one or more leak detection units (1) alarm, uses the data to determine the source of the leak. The BIM data contains information on building structure and pipe layout. Combined with the initial location of the leak point, a simplified water flow path simulation algorithm is used to infer the direction of water flow diffusion and possible leak source areas. The module can consider the influence of gravity, surface tension, capillary action, and material absorption characteristics on the water flow path, and can refer to a preset water flow diffusion model or use historical leak event data for machine learning training. An abnormal event correlation analysis module correlates leak alarm data with related building equipment operation data, including pump start / stop status, valve opening, and pipe pressure sensor data, to further narrow down the range of leak sources.

8. The wireless water leakage detection and alarm system for concealed spaces inside buildings according to claim 7, characterized in that, The visualization and alarm module can accurately mark the location of the detection unit for water leakage alarm on the 3D view or 2D plan of the BIM model in the form of highlights, animations, or icons, and simultaneously display the inferred water leakage source area. The visualization and alarm module supports alarm methods including SMS service notifications, email notifications, mobile application push notifications, and linkage with building automation systems or fire alarm systems. The alarm information includes key information such as water leakage location, inferred source, alarm time, and sensor health status.

9. A wireless water leakage detection and alarm method for concealed spaces inside buildings, based on the wireless water leakage detection and alarm system for concealed spaces inside buildings as described in any one of claims 1-8, characterized in that, Includes the following steps: Step 1: Deployment and initialization of the leak detection unit (1), including the distributed deployment of multiple leak detection units (1) in concealed spaces inside the building, recording the initial geographical location information of each leak detection unit (1) and spatially mapping and binding it with the building information model of the central management platform (3), and performing self-checking and initialization; Step 2: Multimodal wireless communication and data transmission, including the leak detection unit (1) periodically or based on event triggering to collect moisture status data, and transmitting the data to the wireless data gateway unit (2) through the multimodal wireless communication module (6), the wireless data gateway unit (2) receiving the wireless signal and performing signal fusion processing, and then uploading it to the central management platform (3); Step 3: Intelligent identification and adaptive early warning of leak events, including the central management platform (3) receiving and analyzing the leak event data. The process involves five steps: Step 1: Water detection data. When the water level consistently exceeds a preset leakage threshold, a leakage event is identified, and a multi-layered adaptive early warning mechanism is triggered based on the leakage level and the importance of the affected area. Step 2: Precise location and source inference of the leak. This includes the data analysis and positioning module using the multi-source fusion indoor positioning engine to accurately calculate the spatial coordinates of the alarm leak detection unit, combining the BIM model and fluid dynamics model, and the leak source inference module to infer the direction of water flow diffusion and the leak source area. This is further analyzed in conjunction with data from the building automation system through the abnormal event correlation analysis module. Step 3: Fault diagnosis and preventative maintenance. This includes the central management platform's data analysis and positioning module continuously monitoring the health status data of each leak detection unit. When parameters deviate from the health threshold, a fault warning is generated, and remote parameter configuration and firmware upgrades are supported.

10. The wireless water leakage detection and alarm method for concealed spaces inside buildings according to claim 9, characterized in that, Step four specifically includes: based on the indoor positioning engine, using the distance measurement data between the UWB signal emitted by the alarm leak detection unit (1) and the preset UWB anchor point, combined with the polygonal measurement algorithm or nonlinear optimization algorithm, accurately calculating the two-dimensional or three-dimensional spatial coordinates of the alarm leak detection unit (1), and combining it with the RSSI fingerprint spectrum as auxiliary correction; the positioning engine matches the calculated accurate position with the BIM model, and highlights the alarm point in the digital twin environment; the fluid dynamics model leak source inference module, based on the accurate position of the alarm point, combined with the detailed information about the building structure and piping system in the BIM model, uses the preset fluid diffusion model to trace back the most likely leak source area, and refers to the pattern recognition results of historical leak data; the abnormal event correlation analysis module performs real-time correlation analysis between the leak alarm information and the data such as water pressure, water flow sensors, and valve status from the building automation system, so as to improve the inference accuracy of the leak source to the specific pipe or equipment.

Images