Intelligent fire hydrant cap and water identification and energy collaborative management method thereof

Abstract

The application provides a smart fire hydrant cap and a water use identification and energy collaborative management method thereof. The fire hydrant cap comprises a data acquisition unit for acquiring water use state data; an energy collection unit comprising a hydroelectric generator and a piezoelectric vibration energy collector; an energy storage unit comprising a parallelly connected farad capacitor and a solid-state thin film battery; a communication unit comprising a dual-mode wireless communication module for transmitting data to a supervision platform; a lock control unit for locking or unlocking a fire hydrant valve; an edge computing unit comprising a controller and a lightweight neural network accelerator, the controller being capable of identifying water use behavior according to the acquired data; the energy collection unit and the energy storage unit form a power supply module, and the two cooperate to dynamically supply power to the fire hydrant; the edge computing unit is electrically connected with the data acquisition unit, the energy collection unit, the energy storage unit, the communication unit and the lock control unit to control the units.

Description

Technical Field

[0001] This invention relates to the field of smart fire protection technology, and in particular to a smart fire hydrant cap and its water usage identification and energy collaborative management method, as well as a smart fire hydrant system. Background Technology

[0002] Fire hydrants are critical nodes in urban fire protection and water supply networks. Traditional fire hydrants lack intelligent monitoring, leading to serious problems such as unauthorized opening, water theft, and damage, resulting in significant water loss and impacting fire emergency response. Intelligent fire hydrant caps are integrated smart terminals that replace traditional caps, enabling non-destructive installation and real-time remote status transmission. However, existing cap structures have the following technical shortcomings: Energy management is disconnected from the working mode. Existing self-generating power solutions (such as hydroelectric power generation) simply charge the battery and adopt a crude working mode of "generating when there is water and not managing when there is no water". They cannot dynamically adjust the sampling frequency, inference depth and communication strategy of the system according to whether there is water flow, resulting in ineffective energy consumption during non-water flow periods.

[0003] The lack of remote monitoring of fire hydrant water usage status makes it difficult to identify water usage behavior and abnormal status in real time, resulting in delayed response, low management efficiency, and complete loss of monitoring and intervention capabilities when the network is down, making it difficult to meet the real-time and accuracy requirements of smart fire protection.

[0004] The metal body of the fire hydrant is shielded from signals, and the cap is directly installed on the top of the metal fire hydrant. The built-in antenna is affected by the strong coupling of the hydrant body, which greatly reduces the radiation efficiency, resulting in a NB-IoT / LoRa communication success rate of less than 70%.

[0005] Therefore, there is an urgent need for an intelligent fire hydrant cap that can maintain reliable communication in a metallic environment and identify water usage behavior and provide dynamic energy supply in real time, as well as a method for water usage identification and energy collaborative management. Summary of the Invention

[0006] The present invention aims to provide an intelligent fire hydrant cap to overcome the shortcomings of the prior art.

[0007] To solve the above-mentioned technical problems, the technical solution of the present invention is: an intelligent fire hydrant cap, comprising: The data acquisition unit includes at least a water pressure sensor, a flow sensor, a triaxial accelerometer, and a conductivity sensor, used to collect water usage status data; An energy harvesting unit includes a hydroelectric generator and a piezoelectric vibration energy harvester, wherein the hydroelectric generator generates hydroelectric power and the piezoelectric vibration energy harvester is used to harvest vibration energy; The energy storage unit, comprising a parallel-connected supercapacitor and a solid-state thin-film battery, is used for power supply; The communication unit includes a dual-mode wireless communication module for transmitting data to the regulatory platform; The locking unit, including a miniature geared motor and a locking pin, is used to lock or unlock fire hydrant valves; The edge computing unit includes a controller and a lightweight neural network accelerator. The controller is equipped with a water use behavior recognition model, which can identify water use behavior based on the collected data. The lightweight neural network accelerator is used to improve the data processing speed of the controller. The energy harvesting unit and the energy storage unit together form an energy supply module, which works together to provide dynamic energy supply for the fire hydrant; the edge computing unit is electrically connected to the data acquisition unit, the energy harvesting unit, the energy storage unit, the communication unit, and the lock control unit, and controls each unit.

[0008] Furthermore, the aforementioned intelligent fire hydrant cap also includes a metal outer shell. The top of the metal outer shell has an annular groove filled with a high-frequency dielectric material with a dielectric constant of 2.0-2.5. The annular groove and the high-frequency dielectric material form an annular slot antenna located on the top of the metal outer shell. The feed point of the annular slot antenna is located on the axis of symmetry of the annular groove and is connected to a communication unit located inside the metal outer shell via a microstrip line. Preferably, the high-frequency dielectric material is PTFE with a dielectric constant of 2.2.

[0009] Furthermore, in the aforementioned intelligent fire hydrant cap, the width of the annular groove is 1-2mm, and the circumference is 100-150mm.

[0010] This invention provides a method for water usage identification and energy collaborative management of intelligent fire hydrant caps, comprising the following steps: S1. Working condition adaptation and wake-up: The control system is in deep sleep mode by default, and only maintains wake-up monitoring; when the acceleration amplitude of the cap exceeds the threshold or the water pressure change rate exceeds the threshold, the control system is woken up. The acceleration amplitude threshold is 0.1g, where g is the acceleration due to gravity, and the water pressure change rate threshold is 0.01MPa / s. Data collection begins. S2. Multi-source data acquisition: After waking up, multi-source water usage status data is acquired at a sampling frequency of 100-1000Hz. The water usage status data includes water pressure, flow rate, acceleration, and conductivity. S3. Edge feature extraction: Extract time-domain and frequency-domain features from the controller's database. The extracted features include at least the water pressure drop slope, flow duration and cumulative water consumption, acceleration spectrum main frequency, and conductivity abrupt change index. S4. Water use behavior recognition: Input the above features into the lightweight model to identify water use behavior categories; S5. Differentiated response: The system makes different interventions based on different water use behavior categories; S6. Energy Coordination Management: Monitors the output power of the hydroelectric generator installed in the water inlet channel of the cap, and controls the system to automatically switch to different energy management modes based on the output power to dynamically supply energy.

[0011] Furthermore, in the above-mentioned method for water usage identification and energy collaborative management of intelligent fire hydrant caps, the water usage behavior categories in step S4 include at least normal fire-fighting water use, municipal sanitation water intake, hose theft, tampering to steal water, and pipeline leakage; among which, normal fire-fighting water use and municipal sanitation water intake are considered normal water use.

[0012] Furthermore, the aforementioned method for water usage identification and energy management of the intelligent fire hydrant cap is applicable when the water pressure drop slope is 0.05-0.15 MPa / s, the flow duration is 30-180 seconds, and the cumulative water consumption is 5-50 m³. 3 When the conductivity mutation index is <0.05 and the acceleration spectrum has no sustained dominant frequency, the current water use behavior is identified as normal fire-fighting water use.

[0013] Furthermore, the aforementioned method for water usage identification and energy management of the intelligent fire hydrant cap is applicable when the water pressure drop slope is 0.02-0.08 MPa / s, the flow duration is 5-15 minutes, and the cumulative water consumption is 1-10 m³. 3 When the conductivity mutation index is <0.05 and the acceleration spectrum has a wide frequency range without sharp peaks, the current water use behavior is identified as municipal sanitation water intake.

[0014] Furthermore, the aforementioned intelligent fire hydrant cap water usage identification and energy collaborative management method identifies the current water usage behavior as hose theft when the water pressure drop slope is >0.12 MPa / s, the flow duration exceeds 5 minutes, the conductivity mutation index is >0.08, and the acceleration spectrum main frequency maintains the main peak characteristics of 150-500Hz with an amplitude >0.5g.

[0015] Furthermore, the aforementioned method for water usage identification and energy management of the intelligent fire hydrant cap is applicable when the water pressure drop slope is >0.25 MPa / s, the flow duration is 30-180 seconds, and the cumulative water consumption is 5-50 m³. 3 When the conductivity mutation index is greater than 0.12 and the acceleration spectrum main frequency is greater than 1kHz, and the amplitude is greater than 1g, the current water use behavior is identified as water theft through demolition.

[0016] Furthermore, the aforementioned method for water usage identification and energy management of the intelligent fire hydrant cap is applicable when the water pressure drop slope is 0.001-0.01 MPa / s, the flow duration is 30-180 seconds, and the cumulative water consumption is 5-50 m³. 3When the conductivity mutation index is <0.05 and the acceleration has no sustained dominant frequency and the amplitude is weak, the current water use behavior is identified as water theft through demolition.

[0017] Furthermore, in the aforementioned method for water usage identification and energy collaborative management of intelligent fire hydrant caps, the differentiated response in step S5 includes: if water theft through a hose or water theft through tampering is identified, the controller controls the locking unit to close the valve and reports an alarm and sensor snapshots taken 5 seconds before and after to the monitoring platform; if pipeline leakage is identified, it is reported periodically; if normal fire-fighting water use or municipal sanitation water intake is identified, only the water usage is accumulated and the water usage statistics are uploaded to the monitoring platform.

[0018] Furthermore, in the above-mentioned intelligent fire hydrant cap water identification and energy collaborative management method, the energy management mode in step S6 includes deep sleep mode, balance mode, and full speed mode; the energy collaborative management method includes: setting output power thresholds P1 and P2, where P1=5µW and P2=50Mw; When P h ≤P1, P h When the current output power of the hydroelectric generator is at its maximum and there is no wake-up event, the control system enters deep sleep mode and ultra-low power standby. When P1 < P h When ≤P2, the control system enters the balance mode, the control system is awakened, and enters the normal working state. It collects multi-source data at a medium sampling frequency, performs model quantization inference, and communication is intermittently enabled. When P h When the signal reaches P2, the control system enters full-speed mode and high-performance operation. It collects multi-source data at a high-speed sampling frequency, performs full-precision model inference, and maintains communication either always open or reports at high frequency.

[0019] Furthermore, in the above-mentioned method for water identification and energy collaborative management of intelligent fire hydrant caps, when the energy management mode is deep sleep mode, the solid-state thin-film battery is the only continuous power source, providing the sustaining current to maintain wake-up monitoring, the farad capacitor is in a floating charge standby state, does not discharge externally, and waits for energy replenishment, the energy collected by the piezoelectric vibration collector is preferentially charged into the farad capacitor, and the hydroelectric generator does not work. When the energy management mode is in balance mode, the hydroelectric generator is the main power source, directly providing the system with the electricity it currently generates, reducing the burden on the battery; the solid-state thin-film battery provides supplementary power, filling the power gap when the hydroelectric power generation is insufficient; the supercapacitor dynamically stabilizes the voltage; and the energy collected by the piezoelectric vibration collector continuously and weakly replenishes the supercapacitor. When the energy management mode is in full-speed mode, the hydroelectric generator is the only power source, and the excess power is charged into the supercapacitor and solid-state thin-film battery in order of priority; the energy collected by the piezoelectric vibration collector continues to weakly supplement the supercapacitor. The energy management mode also includes a burst mode. When water usage behavior detects water theft through hoses or water theft through tampering, the control system switches to burst mode. The supercapacitor instantly releases a large current to power the locking unit and communication unit. The locking unit locks the valve, and the communication unit transmits data. After the burst mode ends, the control system returns to the original energy management mode before the burst and restarts monitoring the output power of the hydroelectric generator.

[0020] The present invention also provides a smart fire hydrant system, including multiple smart fire hydrant caps as described above, and a cloud platform for wireless communication with them; the cloud platform receives data transmitted by the smart fire hydrant caps, including reported alarm events, water usage status data, etc., and generates a pipeline network operation status diagram.

[0021] Compared with the prior art, the beneficial effects of the present invention are: This invention features a ring-shaped slot antenna on the top of the metal casing of the cap, using the entire cap body as a reference ground, thereby increasing the communication success rate to over 99% and ensuring communication reliability.

[0022] This invention employs multi-source data acquisition, integrating four dimensions: water pressure, flow rate, vibration, and water quality, significantly improving the accuracy of water theft detection and reducing false alarms. Furthermore, it integrates a complete closed-loop edge intelligence—from wake-up, data acquisition, feature extraction, model inference to lock control—on a specific carrier called a fire hydrant cap. This is fundamentally different from gateway-level or cloud-level edge computing, offering faster response and greater security, and avoiding the complete loss of monitoring and intervention capabilities during network outages.

[0023] This invention employs a capacity-coordinated power management approach, which can dynamically switch energy management modes based on the output power of the hydroelectric generator, dynamically supply power, and can operate in ultra-low power standby mode during periods of no wake-up (when there is no water flow). This solves the problem of the coarse working mode in existing technologies, ensures effective energy management, and saves energy. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a schematic diagram of the structure of the present invention; Figure 2 This is a schematic diagram of the annular slot antenna structure of the present invention; Figure 3This is a schematic diagram of the VSWR test for the annular slot antenna of the present invention; Figure 4 This is a schematic diagram of the water usage identification method of the present invention; Figure 5 This is a schematic diagram of the capability collaborative management method of the present invention; In the diagram: 1. Metal casing; 2. Annular groove; 3. Power supply point. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] Example like Figure 1-2 As shown, a smart fire hydrant cap includes: The data acquisition unit includes at least a water pressure sensor, a flow sensor, a triaxial accelerometer, and a conductivity sensor, used to collect water usage status data; The water pressure sensor, flow sensor, and conductivity sensor are all located in the water inlet channel of the cap. The water pressure sensor can determine whether water is coming out, distinguish the water intensity, and identify leakage. The conductivity sensor detects conductivity and can be used to distinguish between fire water and tap water, and assist in identifying abnormal water use. The triaxial accelerometer is located inside the cap and is generally installed on the main control board inside the cap. It can be used to assist in identifying water theft methods (hose / breakage) and detect external impacts. The energy harvesting unit includes a hydroelectric generator and a piezoelectric vibration energy harvester. The hydroelectric generator generates electricity and is installed inside the water inlet channel of the cap, with an output power of 50-200mW. The piezoelectric vibration energy harvester is attached to the inner wall of the cap's metal outer shell and is used to collect vibration energy from road traffic, etc., with an output power of 5-30µW. The energy storage unit, including parallel-connected supercapacitors and solid-state thin-film batteries, is used for power supply and can also be used to smooth power generation spikes and maintain a real-time clock during deep sleep. The communication unit includes a dual-mode wireless communication module (NB-IoT and LoRa) for transmitting data to the regulatory platform; The locking unit, including a miniature geared motor and a locking pin, is used to lock or unlock fire hydrant valves; The edge computing unit includes a controller and a lightweight neural network accelerator. The controller uses a low-power ARM Cortex M4 MCU (64MHz clock speed, built-in floating-point unit). The controller has a water use behavior recognition model, which can identify water use behavior based on the collected data. The lightweight neural network accelerator is used to improve the data processing speed of the controller. The energy harvesting unit and the energy storage unit together form an energy supply module, which works together to provide dynamic energy to the fire hydrant; the edge computing unit is electrically connected to the data acquisition unit, energy harvesting unit, energy storage unit, communication unit, and lock control unit to control each unit.

[0028] This invention provides power through the joint operation of an energy harvesting unit and an energy storage unit. The energy harvesting unit can generate its own electricity without the need for external power supply maintenance, thus reducing dependence on external power supply. Furthermore, the energy harvesting unit itself has a long lifespan, thereby improving the overall lifespan of the power supply module, reducing power supply maintenance and lowering maintenance costs.

[0029] This invention identifies water usage behavior in real time and on-site through edge computing units and intervenes quickly without uploading to the cloud for computing. It offers a faster and safer response and avoids the complete loss of monitoring and intervention capabilities when the network is down.

[0030] like Figure 2 As shown, the aforementioned intelligent fire hydrant cap also includes a metal outer shell 1. The top of the metal outer shell 1 has an annular groove 2. The top of the metal outer shell 1 can be flat or curved. The annular groove 2 is filled with a high-frequency dielectric material with a dielectric constant of 2.0-2.5. The annular groove 2 and the high-frequency dielectric material form an annular slot antenna located on the top of the metal outer shell 1. Specifically, the annular groove region is a non-conductive region. When a high-frequency signal is fed in through the feed point, this slot will radiate electromagnetic waves outwards like an antenna. The feed point 3 of the annular slot antenna is located on the axis of symmetry of the annular groove 2 and is connected to a communication unit located inside the metal outer shell via a 50Ω microstrip line. In this embodiment, the high-frequency dielectric material is PTFE with a dielectric constant of 2.2. Figure 3 As shown, by using the entire metal casing and the fire hydrant body as a grounding reference surface, a standing wave ratio of <1.5 and a gain of ≥2dBi can be achieved in the 800-1000MHz frequency band.

[0031] In the above structure, the width of the annular groove 2 is 1-2 mm, and the circumference is 100-150 mm. This size of the annular groove ensures the communication success rate of the annular slot antenna. In addition, in this embodiment, the wall thickness of the metal shell 1 is 2-3 mm, and the depth of the annular groove 2 is 1 / 2 the wall thickness of the metal shell 1.

[0032] The dielectric constant of PTFE (εr≈2.2) can shorten the wavelength of electromagnetic waves, resulting in a lower resonant frequency for the antenna within the same physical perimeter. This helps to achieve target frequency band matching within a limited space. PTFE also prevents dust and moisture from entering the annular groove and maintains a constant groove width, providing mechanical protection and support.

[0033] This invention addresses the severe electromagnetic shielding of fire hydrant metal bodies by directly creating an annular gap (annular groove) on the cap's metal outer shell and using the entire hydrant body as a reference ground, thereby increasing the communication success rate to over 99% and ensuring communication reliability.

[0034] like Figure 4-5 As shown, based on the above-mentioned intelligent fire hydrant cap, the present invention provides a method for water usage identification and energy collaborative management of an intelligent fire hydrant cap, comprising the following steps: S1. Working condition adaptation and wake-up: The control system is in deep sleep mode by default (power consumption ≤3µA), and only maintains wake-up monitoring. This control system includes an edge computing unit. When the acceleration amplitude of the cap exceeds the threshold or the water pressure change rate exceeds the threshold, the control system is woken up. The MCU wakes up within 2ms. The acceleration amplitude threshold is 0.1g, where g is the acceleration due to gravity. The water pressure change rate threshold is 0.01MPa / s. Data collection begins and the collection duration is at least 5s. S2. Multi-source data acquisition: After waking up, multi-source water usage status data is acquired at a sampling frequency of 100-1000Hz. The water usage status data includes water pressure, flow rate, acceleration, and conductivity. S3. Edge feature extraction: Extract time-domain and frequency-domain features from the controller's database. The extracted features include at least the water pressure drop slope, flow duration and cumulative water consumption, acceleration spectrum main frequency, and conductivity abrupt change index. S4. Water use behavior recognition: Input the above features into the lightweight model to identify water use behavior categories; Specifically, a one-dimensional convolutional neural network (1D-CNN, structure: input layer → convolutional layer (3×8) → pooling layer → fully connected layer (32) → softmax output layer) or a random forest classifier (maximum depth 6, number of trees 50) is pre-trained and quantized into 8-bit integers to identify behavior categories within 200ms, which is fast and accurate.

[0035] S5. Differentiated response: The system makes different interventions based on different water use behavior categories; S6. Energy Coordination Management: Monitors the output power of the hydroelectric generator installed in the water inlet channel of the cap, and controls the system to automatically switch to different energy management modes based on the output power to dynamically supply energy.

[0036] It should be noted that step S6 is performed synchronously with water usage behavior recognition, and the initial system is powered on and initialized.

[0037] In the feature extraction of step S3, the slope of the water pressure drop k is included. P = dP / dt, where P is the current water pressure in the fire hydrant cavity and t is time; Flow duration T Q The time difference between the flow sensor detecting positive flow and the flow returning to zero is given; the cumulative water consumption V is the water consumption during the duration. The acceleration spectrum is analyzed to extract the main frequency and its amplitude, which is used to distinguish between situations such as hose friction and mechanical damage. The conductivity mutation index Δσ = |σ(t)-σ(t-10s)| / σ0 is the ratio of the conductivity change to the reference value, where σ(t) is the current conductivity, σ(t-10s) is the conductivity 10 seconds ago, and σ0 is the reference conductivity, which is the stable conductivity value under normal conditions and is automatically calibrated during system initialization.

[0038] This invention employs multi-source data acquisition, integrating four dimensions: water pressure, flow rate, vibration, and water quality, significantly improving the accuracy of water theft detection and reducing false alarms. Furthermore, it integrates a complete closed-loop edge intelligence—from wake-up, data acquisition, feature extraction, model inference to lock control—on a specific carrier called a fire hydrant cap. This is fundamentally different from gateway-level or cloud-level edge computing, offering faster response and greater security, and avoiding the complete loss of monitoring and intervention capabilities during network outages.

[0039] The water use behavior categories in step S4 include at least normal fire-fighting water use, municipal sanitation water intake, hose theft, tampering to steal water, and pipeline leakage; among them, normal fire-fighting water use and municipal sanitation water intake are considered normal water use.

[0040] When the water pressure drop slope is 0.05-0.15 MPa / s, the flow duration is 30-180 seconds, and the cumulative water consumption is 5-50 m³ / s... 3 If the conductivity mutation index is <0.05, and the acceleration spectrum has no sustained dominant frequency and the amplitude is small, the current water use behavior is identified as normal fire-fighting water use.

[0041] When the water pressure drop slope is 0.02-0.08 MPa / s, the flow duration is 5-15 minutes, and the cumulative water consumption is 1-10 m³ 3 When the conductivity mutation index is <0.05, and the acceleration spectrum has a wide frequency, no sharp main peak, and medium amplitude, the current water use behavior is identified as municipal sanitation water intake.

[0042] When the water pressure drop slope is greater than 0.12 MPa / s, the flow duration exceeds 5 minutes, the conductivity mutation index is greater than 0.08, and the acceleration spectrum frequency maintains a main peak characteristic of 150-500Hz with an amplitude greater than 0.5g, the current water use behavior is identified as hose water theft. Because the flow rate fluctuates during hose water theft, the cumulative water consumption feature is not extracted.

[0043] When the water pressure drop slope is >0.25 MPa / s, the flow duration is 30-180 seconds, and the cumulative water consumption is 5-50 m³ / s. 3 When the conductivity mutation index is >0.12, and the acceleration spectrum main frequency is >1kHz, with a broadband impact amplitude >1g, the current water use behavior is identified as water theft through breaching. Because the flow rate fluctuates drastically during water theft through breaching, the duration of the flow rate and the cumulative water consumption are not extracted.

[0044] When the water pressure drop slope is 0.001-0.01 MPa / s, the flow duration is 30-180 seconds, and the cumulative water consumption is 5-50 m³ / s... 3 When the conductivity mutation index is <0.05 and the acceleration has no sustained dominant frequency and the amplitude is weak, the current water use behavior is identified as water theft through demolition.

[0045] The differentiated response in step S5 includes: if it is identified as water theft through a hose or water theft by breaking and tampering, the controller controls the locking unit to close the valve and reports an alarm and sensor snapshots for 5 seconds before and after to the monitoring platform through the communication unit; if it is identified as pipeline leakage, it reports periodically and does not perform locking; if it is identified as normal fire-fighting water or municipal sanitation water intake, it only accumulates the water consumption and uploads the water consumption statistics to the monitoring platform.

[0046] The energy management modes in step S6 include deep sleep mode, balance mode, and full speed mode.

[0047] In step S6, the capability collaborative management method includes: setting output power thresholds P1 and P2, where P1 = 5µW and P2 = 50Mw; When P h ≤P1, essentially no water flow, P h When the current output power of the hydroelectric generator is at its maximum and there is no wake-up event, the control system enters a deep sleep mode, with ultra-low power standby and total power consumption ≤ 3 µA. When P1 < P h When ≤P2, the control system enters the balanced mode, the control system is woken up and enters the normal working state, and collects multi-source data at a medium speed of 100 Hz, performs model quantization inference, and communication is intermittently enabled; specifically, it adopts 8-bit quantization model low-precision inference, with low power consumption.

[0048] When P hWhen the signal reaches P2, the control system enters full-speed mode and high-performance operation. It collects multi-source data at a high-speed sampling frequency of 1000Hz, performs full-precision model inference, and has the highest recognition accuracy. Communication is always on or high-frequency reporting is also enabled.

[0049] When the energy management mode is in deep sleep mode, the solid-state thin-film battery is the only continuous power source, providing the sustaining current to maintain wake-up monitoring. The farad capacitor is in a float charge standby state, not discharging externally, waiting for energy replenishment. The energy collected by the piezoelectric vibration collector is preferentially charged into the farad capacitor, and the hydroelectric generator does not work.

[0050] In the current mode, the solid-state thin-film battery has a very small discharge current (3 µA), which is beneficial to extending battery life. Although the weak energy collected by the piezoelectric vibration harvester (5-30 µW) cannot power the system, it can continuously trickle charge the farad capacitor, accumulating little by little to store energy.

[0051] When the energy management mode is in balance mode, the hydroelectric generator is the main power source, directly providing the system with the electricity it currently generates, reducing the burden on the battery; the solid-state thin-film battery provides supplementary power, filling the power gap when the hydroelectric power generation is insufficient; the supercapacitor dynamically regulates the voltage, specifically, when the MCU or sensor generates an instantaneous current spike, the supercapacitor instantly releases its charge to eliminate the current spike and prevent voltage drop; the energy collected by the piezoelectric vibration collector continuously and weakly replenishes the supercapacitor.

[0052] In the current mode, hydropower generation bypasses the battery and supplies power directly along with it. This "direct supply + differential compensation" mode keeps the battery in a shallow discharge state (low current, short time), significantly extending battery cycle life. Furthermore, when there is water flow but the flow rate is unstable, the hydropower generation fluctuates, and the supercapacitor acts as a filter, ensuring the MCU receives a stable voltage.

[0053] When the energy management mode is in full-speed mode, the hydroelectric generator is the only power source, and the excess power is charged into the supercapacitor and solid-state thin-film battery in order of priority; the energy collected by the piezoelectric vibration collector continues to weakly supplement the supercapacitor.

[0054] In the current mode, the water flow is large, the hydropower generation energy is sufficient, and the entire system is in a state of energy surplus, prioritizing the charging of the supercapacitor to prepare for the next interlocking and communication. Furthermore, the system operates at maximum performance, consuming no battery power, allowing the battery to "rest and recuperate."

[0055] The energy management mode also includes a burst mode. When water usage behavior detects water theft through hoses or water theft through tampering, the control system switches to burst mode. The supercapacitor instantly releases a large current to power the locking unit and communication unit. The locking unit locks the valve, and the communication unit transmits data. After the burst mode ends, the control system returns to the original energy management mode before the burst and restarts monitoring the output power of the hydroelectric generator.

[0056] In the current mode, the hydroelectric generator and solid-state thin-film battery do not participate. After the burst ends, the hydroelectric generator or solid-state thin-film battery charges the supercapacitor to prepare for the next burst.

[0057] Furthermore, it should be noted that when the burst mode is triggered and the energy stored in the farad capacitor is insufficient to complete a burst action, the farad capacitor and the solid-state thin-film battery can work together to power the burst mode. Specifically, the solid-state thin-film battery provides the baseline current, and the farad capacitor provides the transient peak current. When the energy stored in the farad capacitor and the solid-state thin-film battery is insufficient to complete a burst action, only wake-up monitoring is maintained, all non-essential functions are suspended, and the system waits for replenishment from hydroelectric power generation or vibration energy.

[0058] This invention employs a capacity-coordinated power management approach, which can dynamically switch energy management modes based on the output power of the hydroelectric generator, dynamically supply power, and can operate in ultra-low power standby mode during periods of no wake-up (when there is no water flow). This solves the problem of the coarse working mode in existing technologies, ensures effective energy management, and saves energy.

[0059] The present invention also provides a smart fire hydrant system, including multiple smart fire hydrant caps as described above, and a cloud platform that communicates wirelessly with them. The cloud platform is also a monitoring platform. The cloud platform receives alarm events, water usage status and other data reported by the smart fire hydrant caps, and generates a pipeline network operation status diagram for easy remote viewing by personnel.

[0060] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0061] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A smart fire hydrant cap, characterized in that: include: The data acquisition unit includes at least a water pressure sensor, a flow sensor, a triaxial accelerometer, and a conductivity sensor, used to collect water usage status data; An energy harvesting unit includes a hydroelectric generator and a piezoelectric vibration energy harvester, wherein the hydroelectric generator generates hydroelectric power and the piezoelectric vibration energy harvester is used to harvest vibration energy; The energy storage unit, comprising a parallel-connected supercapacitor and a solid-state thin-film battery, is used for power supply; The communication unit includes a dual-mode wireless communication module for transmitting data to the regulatory platform; The interlocking unit is used to lock or unlock fire hydrant valves; The edge computing unit includes a controller and a lightweight neural network accelerator. The controller is equipped with a water use behavior recognition model, which can identify water use behavior based on the collected data. The lightweight neural network accelerator is used to improve the data processing speed of the controller. The energy harvesting unit and the energy storage unit together form an energy supply module, which works together to provide dynamic energy supply for the fire hydrant; the edge computing unit is electrically connected to the data acquisition unit, the energy harvesting unit, the energy storage unit, the communication unit, and the lock control unit, and controls each unit.

2. The intelligent fire hydrant cap according to claim 1, characterized in that: It also includes a metal casing with an annular groove on the top. The annular groove is filled with a high-frequency dielectric material with a dielectric constant of 2.0-2.

5. The annular groove and the high-frequency dielectric material form an annular slot antenna located on the top of the metal casing. The feed point of the annular slot antenna is located on the axis of symmetry of the annular groove and is connected to a communication unit located inside the metal casing via a microstrip line.

3. The intelligent fire hydrant cap according to claim 2, characterized in that: The width of the annular groove is 1-2 mm, and the circumference is 100-150 mm.

4. A method for water usage identification and energy collaborative management based on the intelligent fire hydrant cap as described in any one of claims 1-3, characterized in that: Includes the following steps: S1. Working condition adaptation and wake-up: The control system is in deep sleep mode by default, and only maintains wake-up monitoring; when the acceleration amplitude of the cap exceeds the threshold or the water pressure change rate exceeds the threshold, the control system is woken up. The acceleration amplitude threshold is 0.1g, where g is the acceleration due to gravity, and the water pressure change rate threshold is 0.01MPa / s. Data collection begins. S2. Multi-source data acquisition: After waking up, multi-source water usage status data is acquired at a sampling frequency of 100-1000Hz. The water usage status data includes water pressure, flow rate, acceleration, and conductivity. S3. Edge feature extraction: Extract time-domain and frequency-domain features from the controller's database. The extracted features include at least the water pressure drop slope, flow duration and cumulative water consumption, acceleration spectrum main frequency, and conductivity abrupt change index. S4. Water use behavior recognition: Input the above features into the lightweight model to identify water use behavior categories; S5. Differentiated response: The system makes different interventions based on different water use behavior categories; S6. Energy Coordination Management: Monitors the output power of the hydroelectric generator installed in the water inlet channel of the cap, and controls the system to automatically switch to different energy management modes based on the output power to dynamically supply energy.

5. The method for water usage identification and energy collaborative management of intelligent fire hydrant caps according to claim 4, characterized in that: The water use behavior categories in step S4 include at least normal fire-fighting water use, municipal sanitation water intake, hose theft, tampering to steal water, and pipeline leakage; among them, normal fire-fighting water use and municipal sanitation water intake are considered normal water use.

6. The method for water usage identification and energy collaborative management of intelligent fire hydrant caps according to claim 5, characterized in that: When the water pressure drop slope is 0.05-0.15 MPa / s, the flow duration is 30-180 seconds, and the cumulative water consumption is 5-50 m³ / s... 3 When the conductivity mutation index is <0.05 and the acceleration spectrum has no sustained dominant frequency, the current water use behavior is identified as normal fire-fighting water use.

7. The method for water usage identification and energy collaborative management of intelligent fire hydrant caps according to claim 5, characterized in that: When the water pressure drop slope is 0.02-0.08 MPa / s, the flow duration is 5-15 minutes, and the cumulative water consumption is 1-10 m³ 3 When the conductivity mutation index is <0.05 and the acceleration spectrum has a wide frequency range without sharp peaks, the current water use behavior is identified as municipal sanitation water intake.

8. The method for water usage identification and energy collaborative management of intelligent fire hydrant caps according to claim 5, characterized in that: When the water pressure drop slope is greater than 0.12 MPa / s, the flow duration exceeds 5 minutes, the conductivity mutation index is greater than 0.08, and the acceleration spectrum main frequency maintains the main peak characteristics of 150-500Hz with an amplitude greater than 0.5g, the current water use behavior is identified as hose water theft.

9. The method for water usage identification and energy collaborative management of intelligent fire hydrant caps according to claim 5, characterized in that: When the water pressure drop slope is >0.25 MPa / s, the flow duration is 30-180 seconds, and the cumulative water consumption is 5-50 m³ / s. 3 When the conductivity mutation index is greater than 0.12 and the acceleration spectrum main frequency is greater than 1kHz, and the amplitude is greater than 1g, the current water use behavior is identified as water theft through demolition.

10. The method for water usage identification and energy collaborative management of intelligent fire hydrant caps according to claim 5, characterized in that: When the water pressure drop slope is 0.001-0.01 MPa / s, the flow duration is 30-180 seconds, and the cumulative water consumption is 5-50 m³ / s... 3 When the conductivity mutation index is <0.05 and the acceleration has no sustained dominant frequency and the amplitude is weak, the current water use behavior is identified as water theft through demolition.

11. The method for water usage identification and energy collaborative management of intelligent fire hydrant caps according to claim 4, characterized in that: The differentiated response in step S5 includes: if it is identified as water theft through a hose or water theft through tampering, the controller controls the locking unit to close the valve and reports an alarm and sensor snapshots for 5 seconds before and after to the monitoring platform; if it is identified as pipeline leakage, it reports periodically; if it is identified as normal fire-fighting water or municipal sanitation water intake, it only accumulates the water consumption and uploads the water consumption statistics to the monitoring platform.

12. The method for water usage identification and energy collaborative management of intelligent fire hydrant caps according to claim 4, characterized in that: The energy management modes in step S6 include deep sleep mode, balance mode, and full speed mode; the energy coordination management method includes setting output power thresholds P1 and P2, where P1=5µW and P2=50Mw. When P h ≤P1, P h When the current output power of the hydroelectric generator is at its maximum and there is no wake-up event, the control system enters deep sleep mode and ultra-low power standby. When P1 < P h When ≤P2, the control system enters the balance mode, the control system is awakened, and enters the normal working state. It collects multi-source data at a medium sampling frequency, performs model quantization inference, and communication is intermittently enabled. When P h When the signal reaches P2, the control system enters full-speed mode and high-performance operation. It collects multi-source data at a high-speed sampling frequency, performs full-precision model inference, and maintains communication either always open or reports at high frequency.

13. The method for water usage identification and energy collaborative management of intelligent fire hydrant caps according to claim 12, characterized in that: When the energy management mode is deep sleep mode, the solid-state thin-film battery is the only continuous power source, providing the sustaining current to maintain wake-up monitoring. The farad capacitor is in a floating charge standby state, does not discharge externally, and waits for energy replenishment. The energy collected by the piezoelectric vibration collector is preferentially charged into the farad capacitor, and the hydroelectric generator does not work. When the energy management mode is in balance mode, the hydroelectric generator is the main power source, directly providing the system with the electricity it currently generates, reducing the burden on the battery; the solid-state thin-film battery provides supplementary power, filling the power gap when the hydroelectric power generation is insufficient; the supercapacitor dynamically stabilizes the voltage; and the energy collected by the piezoelectric vibration collector continuously and weakly replenishes the supercapacitor. When the energy management mode is in full-speed mode, the hydroelectric generator is the only power source, and the excess power is charged into the supercapacitor and solid-state thin-film battery in order of priority; the energy collected by the piezoelectric vibration collector continues to weakly supplement the supercapacitor. The energy management mode also includes a burst mode. When water usage behavior detects water theft through hoses or water theft through tampering, the control system switches to burst mode. The supercapacitor instantly releases a large current to power the locking unit and communication unit. The locking unit locks the valve, and the communication unit transmits data. After the burst mode ends, the control system returns to the original energy management mode before the burst and restarts monitoring the output power of the hydroelectric generator.

14. A smart fire hydrant system, characterized in that: It includes multiple smart fire hydrant caps as described in any one of claims 1-3, and a cloud platform that communicates wirelessly with them; the cloud platform receives data transmitted by the smart fire hydrant caps and generates a pipeline network operation status diagram from it.

Images