Intelligent rainwater collecting and recycling system and method

Abstract

The invention relates to an intelligent rainwater collecting and recycling system and method applied to the technical field of water resource utilization. The system comprises a collection pretreatment unit, a treatment storage unit, a distribution recycling unit and a building linkage unit. According to the invention, through the modular design, especially the water storage modules which can be vertically stacked, the pain point of land shortage in urban built-up areas is solved; through the primary treatment loop and the deep treatment loop which are arranged in parallel, graded purification of rainwater according to recycling scene requirements is achieved; a central controller in the building linkage unit can predict water balance by using an AI model based on multi-source information such as weather forecast and building water data, and execute predictive scheduling, such as active air storage capacity before rainfall and energy consumption optimization by combining time-of-use electricity price. According to the system, full-process intelligent management and control of rainwater from collection, treatment to reuse are achieved, deep cooperation with a building system is achieved, and the rainwater resource utilization rate, the operation economy and the operation and maintenance convenience are remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of water resource utilization technology, specifically relating to an intelligent rainwater collection and reuse system and method suitable for urban built-up areas. Background Technology

[0002] With the deepening implementation of the concepts of sponge cities and green buildings, rainwater harvesting and reuse systems are increasingly widely used in urban built-up areas (such as residential communities and commercial complexes). However, existing systems exhibit significant incompatibility in this specific scenario. The core contradiction lies in the disconnect between the rigidity of system operation and the dynamic complexity of the urban built environment, specifically manifested as follows:

[0003] 1. The dual deficiencies of spatial adaptability and water quality adaptability: Urban built-up areas face limited land availability, but traditional systems rely on large concrete pools and fixed sites, lacking modular, scalable, and compact designs, making them inflexible in the limited spaces surrounding buildings. Simultaneously, their simplistic treatment processes fail to meet the diverse reclaimed water quality needs within buildings, ranging from green space irrigation to cooling tower makeup water. This results in either insufficient treatment limiting the scope of reuse or excessive treatment leading to energy waste, making it difficult to balance water quality safety with treatment costs.

[0004] 2. Deep disconnect between control strategies and building operational patterns: Existing systems are mostly isolated systems operating independently. Their control logic (such as start / stop based on simple water levels) fails to integrate with the building's dynamic water usage patterns (such as peak toilet flushing times in the morning and evening, and differences between weekdays and holidays), energy consumption strategies (such as time-of-use electricity pricing), and building equipment (such as air conditioning cooling towers). This results in low rainwater resource utilization, failing to play its due role in alleviating municipal water supply pressure and assisting in peak shaving and flood reduction, leading to overall low system energy efficiency.

[0005] 3. Insufficient convenience of operation and maintenance and insufficient economic efficiency throughout the entire life cycle: Traditional on-site construction mode has a long cycle and causes great interference to the normal operation of the building; moreover, the system operation and maintenance relies on manual periodic inspections, which results in delayed response and maintenance work is prone to conflict with the peak period of building operation. Long-term operation and maintenance costs are high, which puts a heavy burden on property management.

[0006] The fundamental shortcomings mentioned above regarding spatial deployment, water quality control, collaborative operation, and long-term operation and maintenance have hindered the large-scale and efficient application of rainwater harvesting technology in urban settings. Therefore, there is an urgent need in this field for a comprehensive solution that can be deeply integrated with the urban built environment, achieving efficient space utilization, precise resource allocation, and convenient and economical operation and maintenance. Summary of the Invention

[0007] In view of the above-mentioned existing technologies, the purpose of this invention is to specifically address the core pain points such as limited space in urban built-up areas, diverse reuse scenarios, and disconnection from building systems. Through the deep integration of hardware innovation and software intelligence, it achieves refined, adaptive, and efficient management of the entire process of rainwater resource collection, treatment, and reuse.

[0008] To achieve the above-mentioned objectives, the present invention adopts the following technical solution:

[0009] In a first aspect, the present invention provides an intelligent rainwater harvesting and reuse system, comprising:

[0010] The collection pretreatment unit includes a runoff collection pipe network and a cyclone diversion device connected thereto. The cyclone diversion device includes a cyclone separation chamber, a flow sensor and a turbidity sensor installed in the chamber or on the inlet pipe, and a controlled drain valve.

[0011] The treatment and storage unit includes a water storage facility connected to the outlet of the diversion device, as well as a primary treatment circuit and a deep treatment circuit connected in parallel.

[0012] The distribution and reuse unit includes a secondary water supply network connected to the outlet of the primary treatment loop, a high-quality water supply network connected to the outlet of the deep treatment loop, and electric control valves installed on each pipeline.

[0013] Building linkage units include:

[0014] The sensing module includes multiple level gauges and water quality sensors installed within the system, as well as communication interfaces for accessing external weather forecast data and building automation system data;

[0015] The central controller is connected to the sensing module, electric control valve, and drain valve. It is configured to predict the system water balance for future periods based on weather forecast data and building water consumption data, and control the electric control valve and drain valve accordingly to proactively lower the water level of the water storage facility before rainfall and distribute recycled water of different qualities to the corresponding water use points.

[0016] As a further improvement of the present invention, the processing and storage unit also includes a vertical water storage module, which is a standard cubic structure with a dovetail groove interlocking mechanism and a sealing gasket on the side, supporting vertical stacking and horizontal splicing; the individual size of the water storage module is 500-700mm×500-700mm×500-700mm, and the maximum stacking height is no more than 3 meters.

[0017] As a further improvement of the present invention, the controller of the cyclone diversion device is configured to: comprehensively determine the opening and closing of the drain valve by three parameters: flow rate, turbidity and time; when the influent flow rate is greater than 5-10 L / min, the turbidity is greater than 40-60 NTU and the duration reaches 2-5 minutes, control the drain valve to open for diversion.

[0018] As a further improvement of the present invention, the primary treatment circuit includes a quartz sand filtration assembly and an activated carbon adsorption assembly; the deep treatment circuit includes a heavy metal chelation reaction tank, an ultrafiltration membrane assembly and an ultraviolet sterilizer; and an aminophosphonic acid-type heavy metal chelating agent is added to the heavy metal chelation reaction tank.

[0019] As a further improvement of the present invention, the central controller has a built-in AI decision engine, which integrates an LSTM prediction model and a reinforcement learning algorithm. The input factors of the LSTM prediction model include weather forecasts, real-time water levels, real-time water quality, soil moisture, building water usage time characteristics, time-of-use electricity price signals, and cooling tower operating status. The LSTM prediction model has 20-40 hidden layer neurons, 300-800 iterations, and a prediction error of no more than 10%.

[0020] As a further improvement of the present invention, the system also includes an energy recovery unit, which includes a micro turbine generator installed at the bottom of the rainwater downpipe and a supercapacitor connected thereto. The supercapacitor has a capacity of 400-600F and is used to convert the potential energy of falling rainwater into electrical energy to power the sensors and controllers in the system.

[0021] As a further improvement of the present invention, the system also includes a predictive maintenance unit, which includes differential pressure sensors installed at the inlet and outlet of the filter and / or vibration sensors installed on the pump unit; the central controller is also configured to predict component failures 48-72 hours in advance and issue maintenance warnings based on sensor data trends, and the warning information includes suggestions for maintenance windows to avoid peak water usage and activity periods.

[0022] As a further improvement of the present invention, in the distribution and reuse unit, a water quality monitoring point is set at the end of the high-quality water supply network; when the detected turbidity is greater than 5 NTU or the total number of bacteria is greater than 10 CFU / mL, the central controller controls the start of the reflux valve to return the water to the collection and pretreatment unit.

[0023] Secondly, the present invention provides an intelligent rainwater harvesting and reuse method, implemented based on the aforementioned intelligent rainwater harvesting and reuse system, and comprising the following steps:

[0024] S1. Data Acquisition and Fusion: Internal operational data is collected through the sensing module, and external weather forecast data and building automation system data are integrated.

[0025] S2, AI Prediction and Scheduling: Based on the fused data, the central controller predicts the water supply and demand and water quality changes in future periods and generates scheduling strategies, including proactively emptying the water storage capacity before rainfall and preparing water in advance according to the building's water demand.

[0026] S3. Modular graded treatment: After being diverted by the cyclone diverter, rainwater is guided to the primary treatment loop or the deep treatment loop for treatment according to the water quality requirements of the final reuse scenario.

[0027] S4. Precise Allocation and Reuse: The central controller operates the electric control valves according to the scheduling strategy and control instructions, and delivers treated water of different qualities to the corresponding water use points such as greening irrigation, toilet flushing, landscape water replenishment or cooling tower water replenishment through the corresponding pipeline network.

[0028] Preferably, in step S2, the generation of the scheduling strategy also incorporates the municipal time-of-use electricity price signal, and arranges high-energy-consuming water treatment or backwashing operations to be performed during the period when the electricity price is at its lowest; and, 1-6 hours before rainfall, when the predicted rainfall is not less than 10 mm, the water level of the water storage facility is reduced to 40%-50% of its safe capacity.

[0029] In summary, this invention, through its vertically stacked modular water storage design, completely solves the pain point of limited land in urban centers, allowing for flexible deployment in confined spaces such as green belts and parking lots. Based on AI-driven predictive scheduling, it achieves a leap from passive collection to active regulation, significantly improving rainwater resource utilization and effectively reducing peak runoff. Furthermore, by linking with building automation systems, it integrates rainwater management into building energy and water systems, achieving overall energy efficiency optimization and reducing building operating costs. In addition, this invention significantly reduces energy consumption and maintenance costs through on-demand processing and predictive maintenance, and its modular design facilitates installation and expansion, further enhancing the project's economic viability. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the intelligent rainwater harvesting and reuse system of the present invention;

[0031] Figure 2 This is a flowchart illustrating the logic of AI prediction and scheduling in this invention.

[0032] Figure 3 This is a process flow diagram of the intelligent rainwater harvesting and reuse method of the present invention. Detailed Implementation

[0033] The two embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0034] First implementation method:

[0035] Figures 1-2 This invention illustrates an intelligent rainwater harvesting and reuse system, comprising a collection and pretreatment unit, a treatment and storage unit, a distribution and reuse unit, a building linkage unit, an energy recovery unit, and a predictive maintenance unit.

[0036] The collection and pretreatment unit is the first-end treatment module for rainwater entering the system, used to achieve efficient collection of rainwater and interception of initial pollution. It includes a runoff collection pipe network and a vortex diversion device, which are connected by a flange seal.

[0037] The runoff collection network uses PE thin-walled spiral pipes, with the pipe diameter selected according to the catchment area. The outer wall of the pipe is coated with an anti-corrosion coating, and the inner wall is smooth to reduce water flow resistance. The branch points of the network are equipped with guide tees to promote the uniform flow of rainwater into the vortex diversion device. The network is laid along the perimeter of buildings, both sides of roads, or the edge of green belts, without occupying a large amount of additional space, making it suitable for the narrow deployment environment of urban built-up areas.

[0038] The cyclone diversion device is the core component for the precise interception of initial rainwater. It is made of stainless steel and includes a cyclone separator (also known as a hydrocyclone / cyclone separator), a flow sensor, a turbidity sensor, and a controlled drain valve. The cyclone separator has an inverted cone-shaped structure and uses a tangential water inlet to form a cyclone field. It has a drain port at the bottom connected to the drain valve and an outlet at the top connected to the subsequent treatment and storage unit. The side wall of the chamber has reserved sensor installation interfaces for easy maintenance and replacement.

[0039] The flow sensor and turbidity sensor can be installed in the middle of the cyclone separator or at the end of the inlet pipe. The flow sensor uses an electromagnetic measurement principle to monitor the inflow of rainwater in real time. The turbidity sensor uses a laser scattering principle to detect the turbidity of rainwater to determine the degree of pollution. Both the flow sensor and the turbidity sensor establish a signal connection with the central controller of the building linkage unit through a communication interface.

[0040] The controlled drain valve is an electromagnetic valve, whose input is electrically connected to the switch output interface of the central controller. The valve is opened and closed by the electrical signal output by the central controller. The outlet of the drain valve is connected to the municipal sewage network through a pipeline to ensure that the initial polluted rainwater is discharged in time to avoid polluting the subsequent treatment units.

[0041] The working logic of the collection and pretreatment unit is as follows: after rainwater is collected by the runoff collection pipe network, it flows into the cyclone separator. Under the action of centrifugal force, the heavier particulate pollutants in the rainwater (such as road mud and roof dust) are thrown against the chamber wall and settle to the bottom of the chamber. At the same time, the flow sensor and turbidity sensor collect rainwater flow and turbidity data synchronously and transmit them to the central controller in real time. The central controller combines the rainfall duration parameter and uses preset logic to determine whether it is initial polluted rainwater. When the preset conditions are met, the drain valve is opened to discharge the polluted rainwater into the sewage pipe network. After the water quality meets the standards, the drain valve is closed to allow high-quality rainwater to enter the treatment and storage unit.

[0042] The treatment and storage unit is used to realize the graded treatment and safe storage of rainwater. It includes a water storage facility, a primary treatment loop and a deep treatment loop. The outlet of the water storage facility is connected to the inlet of the two treatment loops through a three-way pipe. The pipe is equipped with an electric control valve to switch the water flow path.

[0043] Water storage facilities can adopt various forms, including but not limited to regulating tanks, vertical water storage modules, or a combination of both, depending on the spatial conditions and water storage capacity requirements of the application scenario. Regulating tanks are constructed with reinforced concrete or integrally molded fiberglass, with a waterproof and corrosion-resistant coating on the inner wall. The top of the tank has an inspection port and vent, while the bottom has a drain outlet, suitable for large-scale water storage needs. Vertical water storage modules are standard cubic structures and serve as the core innovative component of this unit. Their individual dimensions are 500-700mm × 500-700mm × 500-700mm. The modules have dovetail groove interlocking mechanisms on the sides, with silicone sealing gaskets embedded in the grooves to ensure good sealing performance and eliminate the risk of leakage after module assembly. All bottom drain outlets of the water storage facility are equipped with electric drain valves, which are connected to the central controller to achieve automatic drain control: In the normal drain mode, the drain valve is automatically opened for 10 minutes from 2:00 to 2:10 am on the 1st of each month (during off-peak water usage) to discharge a small amount of sediment at the bottom of the pool; In the emergency drain mode, when the water quality sensor in the water storage facility detects turbidity > 8 NTU or obvious odor, the central controller triggers the emergency drain procedure, opens the drain valve until the water quality returns to the standard, and simultaneously closes the connecting valve between the water storage facility and the treatment circuit during the drain process to avoid contamination of subsequent processes.

[0044] The vertical water storage module supports vertical stacking and horizontal splicing, with a maximum stacking height of no more than 3 meters. The overall compressive strength after stacking is ≥2.5MPa, which can withstand urban ground loads (such as being run over by small vehicles). It is suitable for deployment in confined spaces such as green belts in residential areas and parking lots in commercial complexes. The water storage cavity formed by splicing multiple modules is equipped with a water circulation system. It is connected to a circulating water pump to periodically disturb the water to prevent water quality deterioration. The water storage facility is also equipped with a level gauge, which is connected to the central controller to provide real-time feedback of water level data.

[0045] The primary treatment loop and the deep treatment loop are set up in parallel and both adopt a modular design, supporting plug-and-play installation and expansion. Both loops are equipped with electric control valves at the inlet and outlet. The electric control valves are electromagnetic proportional valves that are electrically connected to the central controller to achieve precise control of water flow and path switching.

[0046] The primary treatment circuit is used to treat rainwater that meets low water quality requirements. It includes a quartz sand filter module and an activated carbon adsorption module connected in series. The quartz sand filter module is used to intercept suspended particulate matter, colloids and other impurities in the rainwater. The activated carbon adsorption module is used to adsorb organic matter, odors and some heavy metal ions in the rainwater, which can meet the needs of reuse scenarios such as greening irrigation and road washing.

[0047] The advanced treatment loop is used to treat rainwater that meets high water quality requirements. It includes a heavy metal chelation reaction tank, an ultrafiltration membrane module, and an ultraviolet sterilizer connected in series. The heavy metal chelation reaction tank is equipped with a stirring device and a reagent dosing port. By adding an aminophosphonic acid-type heavy metal chelating agent, stable chelate precipitates are formed with complexed heavy metals (such as Pb²⁺, Cu²⁺, etc.) in the rainwater. The ultrafiltration membrane module uses a hollow fiber membrane with a pore size of 0.01μm, a molecular weight cutoff of 5000Da, and an operating pressure of 0.1-0.3MPa, which can effectively remove bacteria, viruses, and residual suspended solids in the water. The ultraviolet sterilizer is used to kill residual microorganisms in the water. The treated rainwater can meet the reuse requirements of toilet flushing, landscape water replenishment, and cooling tower water replenishment.

[0048] The distribution and reuse unit is used to achieve precise delivery and reuse of water treated to different qualities. It includes a secondary water supply network, a high-quality water supply network, electric control valves, water quality monitoring points, and a return valve. All components are connected by pipelines to form a closed-loop water supply system.

[0049] The secondary water supply network is connected to the outlet of the primary treatment circuit and is used to transport rainwater that has undergone primary treatment; the high-quality water supply network is connected to the outlet of the deep treatment circuit and is used to transport rainwater that has undergone deep treatment; both networks use PE water supply pipes, and the pipe diameter is adapted according to the water flow rate. The outer wall of the network is clearly marked to facilitate construction and maintenance and avoid misconnection and misuse.

[0050] Electric control valves are installed at the nodes of the main and branch pipes of the two pipe networks and are connected to the central controller of the building linkage unit to control the opening and closing of the corresponding pipes and the distribution of water flow. The electric control valves on the main pipes are electric gate valves with a nominal diameter adapted to the pipe network, while the electric control valves on the branch pipes are electric ball valves. All valves have manual operation functions to ensure manual emergency control in case of power failure or malfunction.

[0051] Water quality monitoring points are set up at the end of the high-quality water supply network. Each monitoring point integrates a turbidity sensor and an online bacterial count monitor. The turbidity sensor has a range of 0-20 NTU and an accuracy of ±0.1 NTU. The online bacterial count monitor uses the fluorescence counting method, with a detection range of 0-100 CFU / mL and a detection accuracy of ±1 CFU / mL. The water quality monitoring points are connected to the central controller to upload monitoring data in real time, which serves as the basis for judging whether the water quality meets the standards and for backflow control.

[0052] The reflux valve is installed on the branch pipe of the high-quality water supply network and the connecting pipe of the collection and pretreatment unit. It is an electric three-way valve and is electrically connected to the central controller. When the water quality monitoring point detects that the turbidity of the effluent is greater than 5 NTU or the total number of bacteria is greater than 10 CFU / mL, the central controller immediately controls the reflux valve to switch to the reflux path, and returns the substandard water to the collection and pretreatment unit for reprocessing to ensure the safety of the recycled water quality.

[0053] The building linkage unit is the control core of the system, used to realize multi-source data fusion, intelligent decision-making and collaborative control. It includes a sensing module and a central controller, which are connected through industrial bus communication to form a closed-loop control link of data acquisition-processing-decision-execution.

[0054] The sensing module includes an internal sensor array and an external data interface for comprehensively collecting system operation data and external environmental data. The internal sensor array includes multiple level gauges, water quality sensors, and soil moisture sensors. The level gauges are submersible hydrostatic sensors, installed in key locations such as water storage facilities and water tanks in the treatment loop to monitor real-time water levels. In addition to the aforementioned turbidity and total bacterial count monitoring devices, the water quality sensors also include pH and residual chlorine sensors installed at the outlet of the deep treatment loop to comprehensively monitor the effluent quality. The soil moisture sensor uses the frequency domain reflectance principle to monitor soil moisture to optimize irrigation strategies.

[0055] The external data interface adopts an industrial-grade edge gateway, supports 5G / NB-IoT dual-mode communication, and has data preprocessing and local caching functions (it can cache 7 days of key data in the absence of network). This interface is used to access external weather forecast data (probability of rainfall in the next 2-24 hours, rainfall amount) and building automation system data (building water use mode, cooling tower operation status, time-of-use electricity price signal). The data update frequency is 1 hour / time to ensure the timeliness of decision-making. When the external data interface is disconnected from the network, the central controller automatically switches to locally cached historical data and real-time internal sensor data. The offline model based on reinforcement learning algorithm continues to generate scheduling strategies, and the system can maintain continuous operation for 7 days in the absence of network. When a sensor (such as a water quality sensor or level gauge) fails, the controller uses data interpolation of similar sensors to complete the system or switches to a backup sensor (if available). At the same time, it sends a fault prompt to the management personnel through the early warning module to ensure that the core functions of the system are not affected. When a sudden change in rainwater quality is detected (such as turbidity exceeding 100 NTU instantaneously), the central controller immediately controls the drain valve of the vortex diversion device to remain open and suspends the operation of subsequent treatment circuits to avoid contaminating the water storage facilities and treatment components.

[0056] The central controller adopts an industrial-grade PLC controller, which has multi-protocol compatibility, high reliability and scalability. Its hardware includes a CPU module, analog input / output module, digital input / output module, communication module and power supply module, which are integrated into the control cabinet through rack mounting. The controller has a built-in AI decision engine, which integrates LSTM prediction model and reinforcement learning algorithm to provide core algorithm support for intelligent scheduling.

[0057] The LSTM prediction model has 20-40 hidden layer neurons and 300-800 iterations. Input factors include weather forecast data, real-time water level data, real-time water quality data, soil moisture data, building water usage time period characteristics data, time-of-use electricity price signals, and cooling tower operation status data. After data standardization, the data is input into the model, and the output is the water supply and demand prediction results and water quality change trends for the next 6-24 hours, with a prediction error of no more than 10%. The reinforcement learning algorithm takes "maximizing water resource utilization and minimizing operating costs" as the optimization objective, constructs a reward function, and generates the optimal control command by continuously learning scheduling strategies under different scenarios.

[0058] The core control logic of the central controller includes: predicting rainfall based on meteorological forecast data, controlling the water level of water storage facilities in advance to free up storage capacity to maximize rainwater collection; controlling the switching of electric control valves in the treatment loop based on building water usage patterns and real-time water level and water quality data to prepare reclaimed water of different qualities as needed; optimizing the execution time of high-energy-consuming operations such as deep treatment and backwashing based on time-of-use electricity price signals; receiving feedback data from water quality monitoring points and various sensors, adjusting the water supply strategy in real time, and triggering backflow or fault alarm mechanisms.

[0059] The energy recovery unit is the energy efficiency optimization module of the system. It is used to recover the gravitational potential energy of rainwater during the falling process and convert it into electrical energy to power the low-power devices of the system. It includes a micro turbine generator and a supercapacitor, which are electrically connected by wires to form an energy cycle link of power generation-energy storage-power supply.

[0060] The micro turbine generator adopts an axial flow structure, which is suitable for the installation space of rainwater downpipes in urban buildings. Its impeller is made of stainless steel, which is highly corrosion resistant. The generator is fixedly installed at the bottom of the rainwater downpipe through a flange. When rainwater falls, it impacts the impeller to rotate, which drives the generator rotor to cut magnetic field lines and generate electricity.

[0061] The supercapacitor uses a double-layer capacitor, which features fast charging and discharging speed, long cycle life (≥100,000 times), and high reliability. The supercapacitor is connected to the output terminal of the micro turbine generator to store the electrical energy generated by the generator. At the same time, it supplies power to low-power devices such as sensors in the sensing module and communication modules of the central controller through a voltage regulator module, reducing the system's dependence on municipal power supply.

[0062] Predictive maintenance units are used to monitor the status of system equipment and provide early warnings of faults, thereby reducing operation and maintenance costs and downtime risks. They include status monitoring sensors and early warning modules, both of which are connected to the central controller of the building linkage unit.

[0063] The condition monitoring sensors include differential pressure sensors and vibration sensors. The differential pressure sensors are installed on the inlet and outlet pipes of the quartz sand filter assembly in the primary treatment loop and the ultrafiltration membrane assembly in the deep treatment loop. They are diffused silicon pressure sensors used to monitor the degree of filter clogging. When the inlet and outlet pressure difference exceeds a preset threshold, a backwashing or filter element replacement warning is triggered. The vibration sensors are installed on rotating equipment such as pumps and agitators in the system. They are piezoelectric accelerometers used to monitor the vibration amplitude during equipment operation and to determine whether there are faults such as imbalance or bearing wear.

[0064] The early warning module is integrated into the central controller. It uses software algorithms to perform trend analysis on the data collected by the status monitoring sensors. Based on the equipment's historical operating data and fault feature database, it predicts the component failure time 48-72 hours in advance. The early warning information is sent to the remote operation and maintenance platform and the management personnel's mobile APP through the central controller's communication interface. The early warning content includes the name of the faulty component, the predicted failure time, the suggested maintenance measures, and the maintenance window (avoiding peak water usage and building operation periods). This allows management personnel to formulate maintenance plans in advance and reduce interference with system operation and building operation.

[0065] Specific application examples

[0066] To more clearly illustrate how this system can be implemented, the following uses a 100,000㎡ urban residential community (2,000 households, 30,000㎡ of green space, including 3 500RT air conditioning cooling towers) as an application scenario, providing specific configuration and operational details (this embodiment is only one application form; those skilled in the art can adjust parameters and configurations according to the actual scenario, without departing from the scope of protection of this invention):

[0067] 1. System Core Configuration

[0068] Collection and pretreatment unit: The runoff collection pipeline network uses DN150 PE pipe with a total length of 800m; 2 sets of cyclone diversion devices, flow sensor range 0-50L / min, turbidity sensor range 0-100NTU, and sewage valve nominal diameter DN50.

[0069] Treatment and storage unit: The water storage facility adopts a combination of "300m³ concrete regulating tank + 200m³ vertical water storage module". The vertical water storage module is 600×600×600mm in size, with a total of 1389 modules and a stacking height of 2.4m. The filter media particle size of the primary treatment loop is 0.8-1.2mm, and the chelating agent dosage of the deep treatment loop is 1L / h.

[0070] Building linkage unit: The central controller adopts an industrial-grade programmable logic controller (PLC), the LSTM model has 30 hidden layer neurons and 500 iterations; the external data interface accesses the meteorological data of the municipal meteorological bureau through NB-IoT, with an update frequency of 1 hour / time;

[0071] Energy recovery unit: 4 micro turbine generators are installed at the bottom of the rainwater downpipe of the residential building, and the supercapacitor has a capacity of 500F;

[0072] Predictive maintenance unit: differential pressure sensor with a range of 0-1MPa, vibration sensor installed on 3 water supply pump sets.

[0073] 2. System Operation Coordination Logic

[0074] Pre-rainfall management: The weather forecast predicts 15mm of rainfall in the next 2 hours. The central controller will reduce the water level of the water storage facility from 80% to 45%, freeing up 180m³ of storage capacity.

[0075] Initial rainwater diversion: After rainfall, the sensor detects "flow rate 8L / min, turbidity 50NTU". After 3 minutes, the drain valve opens to divert the water. After 15 minutes, the drain valve closes when the water quality meets the standard.

[0076] Graded treatment and reuse: Green irrigation water undergoes primary treatment (turbidity 2.8 NTU) and is supplied through the secondary water network from 10:00 to 11:00; toilet flushing and cooling tower water undergoes advanced treatment (total bacterial count 3 CFU / mL) and is prepared 2 hours before the morning peak.

[0077] Energy efficiency optimization: The ultrafiltration membrane backwashing is scheduled to be performed during the off-peak electricity price period from 02:00 to 04:00. The electricity generated by the energy recovery unit can partially meet the power needs of low-power devices such as sensors and controllers in the system.

[0078] Predictive maintenance: When the pressure difference between the inlet and outlet of the quartz sand filter reaches 0.2MPa, the system predicts that backwashing is required after 12 hours and sends an early warning to the property management. The recommended maintenance window is 09:00-11:00 the next day.

[0079] This embodiment, through the above configuration and operating logic, fully realizes the whole process control of rainwater collection, pretreatment, graded treatment, precise allocation and equipment maintenance, with smooth linkage between units, meeting the needs of diverse water use scenarios in urban communities.

[0080] Second implementation method:

[0081] Figure 3 This paper presents an intelligent rainwater harvesting and reuse method. This method achieves efficient utilization of rainwater resources through multi-source data fusion, AI intelligent scheduling, modular hierarchical processing, and precise allocation and reuse. Specifically, it includes the following steps:

[0082] S1. Data Acquisition and Fusion:

[0083] After the system is started, the internal sensor array of the sensing module continuously collects real-time water level of the water storage facility, inlet and outlet water quality (turbidity, pH, residual chlorine, etc.) of the treatment circuit, soil moisture, pressure difference between filter inlet and outlet, equipment vibration amplitude, and other internal operating data at a sampling frequency of 10Hz. At the same time, the external data interface regularly accesses weather forecast data (probability of rainfall, rainfall amount, and rainfall period) for the next 2-24 hours, building water use pattern data output by the building automation system (difference in water use on weekdays / weekends, peak water use periods in the morning and evening), cooling tower operating status data (operation / stop, water replenishment demand), and municipal time-of-use electricity price signals through the 5G / NB-IoT communication network. The data update frequency is once every hour.

[0084] The collected multi-source data is transmitted to the edge gateway of the central controller for preprocessing, including outlier removal (using the 3σ criterion), data standardization (mapping data of different dimensions to the 0-1 range), and data fusion (using the weighted average method to integrate data from similar sensors). The preprocessed data is stored in the local cache module (retaining 7 days of historical data) and transmitted in real time to the AI ​​decision engine of the central controller to provide data support for subsequent prediction and scheduling.

[0085] S2, AI Prediction and Scheduling:

[0086] The central controller's AI decision engine, based on preprocessed multi-source data, first uses an LSTM prediction model to predict water supply and demand and water quality changes over the next 6-24 hours: combining rainfall data from weather forecasts with current water levels in storage facilities, it predicts rainwater collection volume and storage capacity requirements; combining building water usage patterns and cooling tower operating status, it predicts reclaimed water demand at different times; and combining initial rainwater quality monitoring data with historical treatment results, it predicts whether the treated rainwater will meet water quality standards.

[0087] Based on the prediction results, the reinforcement learning algorithm generates the optimal scheduling strategy, specifically including:

[0088] Storage capacity optimization scheduling: 1-6 hours before rainfall, when the predicted rainfall is not less than 10mm, the central controller controls the water outlet valve of the water storage facility to open, reducing the water level to 40%-50% of its safe capacity, emptying the storage capacity to maximize rainwater collection and avoid overflow during rainfall.

[0089] Treatment loop scheduling: Based on the predicted demand for recycled water and water quality requirements, the corresponding treatment loops are started in advance to prepare water. For example, if a large amount of high-quality recycled water is predicted to be needed during the peak period of toilet flushing in the building, the deep treatment loop is started 2 hours in advance to ensure sufficient water supply.

[0090] Energy consumption optimization scheduling: By combining municipal time-of-use electricity price signals, high-energy-consuming operations such as deep treatment and ultrafiltration membrane backwashing are scheduled to be performed during off-peak electricity price periods (such as 00:00-06:00) to reduce system operating costs;

[0091] Water supply priority scheduling: When the total amount of recycled water is insufficient, water volume is allocated according to the principle of "priority to key scenarios", giving priority to core scenarios such as cooling tower water replenishment and toilet flushing, and then supplying non-core scenarios such as greening irrigation and road washing.

[0092] After the scheduling strategy is generated, the central controller converts it into specific control commands, including valve opening and closing signals, pump group operating frequency signals, and chemical dosage signals, and transmits them to the corresponding actuators.

[0093] S3, Modular hierarchical processing:

[0094] After rainwater flows into the vortex diversion device through the runoff collection network of the collection and pretreatment unit, the central controller controls the opening and closing of the drain valve based on the data collected by the flow sensor and turbidity sensor and the duration of rainfall, using a three-parameter comprehensive judgment logic: when the influent flow rate is greater than 5-10 L / min, the turbidity is greater than 40-60 NTU and the duration reaches 2-5 minutes, it is determined to be initial polluted rainwater, and the drain valve is opened to divert the flow; when the detection data is lower than the above threshold, it is determined to be high-quality rainwater, the drain valve is closed, and the rainwater flows into the water storage facility of the treatment and storage unit.

[0095] Based on instructions generated by AI prediction and scheduling steps, the central controller controls the switching of electric control valves in the storage unit to guide rainwater in the storage facility to the corresponding treatment loops: If the reuse scenario is greening irrigation or road washing, the electric control valve of the primary treatment loop is opened, and the rainwater passes through the quartz sand filter to remove suspended particulate matter, and the activated carbon adsorption component to adsorb organic matter and odors. After treatment, it is stored in the clear water area of ​​the storage facility. If the reuse scenario is toilet flushing, landscape water replenishment, or cooling tower water replenishment, the electric control valve of the deep treatment loop is opened, and the rainwater passes through the heavy metal chelation reaction tank to remove complexed heavy metals, the ultrafiltration membrane component to remove microorganisms and suspended solids, and the ultraviolet sterilizer to kill residual bacteria. After treatment, it is stored in the clear water area of ​​the storage facility.

[0096] During the treatment process, the central controller receives real-time feedback data from the water quality sensors in the treatment loop. If the water quality in a certain treatment stage is found to be substandard, the corresponding process parameters are immediately adjusted (such as increasing the dosage of chemicals or reducing the water flow rate) to ensure that the final effluent meets the reuse requirements. Simultaneously, the central controller triggers a backwashing procedure based on the differential pressure sensor data at the filter inlet and outlet: For the quartz sand filter module in the primary treatment loop, when the inlet and outlet differential pressure is ≥0.2MPa, the inlet and outlet valves are closed, and the backwash valve is opened. Treated water meeting the standards is used for backwashing for 15 minutes (washing water flow rate 1.2m / s). The backwash wastewater is discharged into the municipal sewage network via the sewage pipe. For the ultrafiltration membrane module in the advanced treatment loop, when the inlet and outlet differential pressure is ≥0.3MPa, a combined air-water backwash is performed. First, 0.4MPa compressed air is introduced for 5 minutes, followed by a 1.5m / s water flow for 15 minutes. After rinsing, normal operation is restored. Backwashing operations are prioritized during off-peak electricity prices or low water usage periods to avoid affecting water supply.

[0097] S4. Precise Allocation and Reuse:

[0098] The central controller controls the opening and closing of the electric control valves of the distribution and reuse unit and adjusts the opening degree according to the scheduling strategy and the real-time demand of each water point, so as to achieve precise delivery of water with different quality treatments: primary treated water is delivered to water points with low water quality demand such as green irrigation sprinklers and road washing devices through the secondary water network; and deep treated water is delivered to water points with high water quality demand such as toilet flushing equipment, landscape water features, and cooling tower water inlets through the high-quality water network.

[0099] During the greening irrigation process, the central controller combines data collected by soil moisture sensors to start irrigation when the soil moisture is lower than 60% of the field capacity. The irrigation timing avoids peak periods of personnel activity inside the building (such as 7:00-9:00 and 18:00-20:00) to avoid affecting personnel activities. During the cooling tower water replenishment process, the water supply flow is adjusted in real time according to the operating status and water level feedback of the cooling tower to ensure the stability of the cooling tower water level.

[0100] The water quality monitoring point at the end of the high-quality water supply network monitors the effluent water quality in real time. When the turbidity is greater than 5 NTU or the total number of bacteria is greater than 10 CFU / mL, the central controller immediately controls the reflux valve to open, returning the substandard water to the collection and pretreatment unit for reprocessing. After the water quality monitoring data returns to the standard, the reflux valve is controlled to close, restoring normal water supply and ensuring the safety and reliability of the recycled water.

[0101] In light of current practical needs, the above-described embodiments of this invention are not limited to these specific implementations. Any changes made within the scope of knowledge possessed by those skilled in the art, without departing from the concept of this invention, still fall within the protection scope of this invention.

Claims

1. An intelligent rainwater harvesting and recycling system characterized in that, Comprise: a collection pre-treatment unit comprising a runoff collection pipe network and a cyclone reject device in communication therewith, the cyclone reject device comprising a cyclone separation bin, a flow sensor and a turbidity sensor disposed on the bin or an inlet pipe, and a controlled blowdown valve; a treatment and storage unit comprising a water storage facility in communication with an outlet of the reject device, and a primary treatment circuit and a deep treatment circuit arranged in parallel; a distribution and reuse unit comprising a secondary water pipe network in communication with an outlet of the primary treatment circuit, a high-quality water pipe network in communication with an outlet of the deep treatment circuit, and electrically controlled valves disposed on each pipe; a building linkage unit comprising: a sensing module comprising a plurality of liquid level meters, water quality sensors disposed in the system, and a communication interface for accessing external weather forecast data and building automation system data; a central controller in signal connection with the sensing module, the electrically controlled valves, and the blowdown valve, and configured to predict the water balance of the system in a future period based on the weather forecast data and building water data, and to control the electrically controlled valves and the blowdown valve accordingly to proactively reduce the water level of the water storage facility before rainfall and to distribute reused water of different qualities to corresponding water points.

2. The intelligent rainwater harvesting and reuse system of claim 1, wherein, The treatment and storage unit further comprises a vertical water storage module, which is a standard cubic structure with dovetail groove interlocking mechanisms and sealing pads disposed on the sides to support vertical stacking and horizontal splicing; the monomer size of the water storage module is 500-700mm x 500-700mm x 500-700mm, and the maximum stacking height is not greater than 3 meters.

3. The intelligent rainwater harvesting and reuse system of claim 2, wherein, The controller of the cyclone reject device is configured to control the opening and closing of the blowdown valve based on the flow rate, turbidity, and time; when the inlet flow rate is greater than 5-10 L / min, the turbidity is greater than 40-60 NTU, and the duration reaches 2-5 minutes, the blowdown valve is controlled to open for reject flow.

4. The intelligent rainwater harvesting and reuse system of claim 1, wherein, The primary treatment circuit comprises a quartz sand filtration assembly and an activated carbon adsorption assembly; the deep treatment circuit comprises a heavy metal chelation reaction tank, an ultrafiltration membrane assembly, and an ultraviolet sterilizer; the heavy metal chelation reaction tank is added with an amino phosphonic acid type heavy metal chelating agent.

5. The intelligent rainwater harvesting and reuse system of claim 1, wherein, The central controller has an AI decision engine built-in, which integrates an LSTM prediction model and a reinforcement learning algorithm; the input factors of the LSTM prediction model include weather forecast, real-time water level, real-time water quality, soil moisture, building water period characteristics, time-of-use electricity price signals, and cooling tower operating status; the number of hidden layer neurons of the LSTM prediction model is 20-40, the number of iterations is 300-800, and the prediction error is not greater than 10%.

6. The intelligent rainwater harvesting and reuse system of claim 1, wherein, The system further comprises an energy recovery unit, which comprises a micro-turbine generator installed at the bottom of a rainwater vertical pipe and a super capacitor connected thereto, the capacity of the super capacitor is 400-600F, which is used to convert the potential energy of falling rainwater into electrical energy to power the sensors and controllers in the system.

7. The intelligent rainwater harvesting and reuse system of claim 1, wherein, The system further comprises a predictive maintenance unit, which comprises a differential pressure sensor arranged at the inlet and outlet of the filter and / or a vibration sensor arranged on the pump set; the central controller is further configured to predict component failure 48-72 hours in advance and issue a maintenance warning based on the data trend of the sensor, and the warning information contains a maintenance window suggestion avoiding water use and activity peaks.

8. The intelligent rainwater harvesting and reuse system of claim 1, wherein, In the distribution and reuse unit, a water quality monitoring point is arranged at the end water use point of the high-quality water pipe network; when the monitored turbidity is greater than 5 NTU or the total number of bacteria is greater than 10 CFU / mL, the central controller controls the start of the backflow valve to return the water to the collection and pretreatment unit.

9. A method of intelligent rainwater harvesting and reuse, characterized by, The intelligent rainwater collection and reuse system based on any one of claims 1-8 comprises the following steps: S1, data acquisition and fusion: internal operation data is collected by a sensing module, and external meteorological forecast data and building automation system data are accessed; S2, AI prediction and scheduling: the central controller predicts water supply and demand and water quality changes in future periods based on the fused data, and generates a scheduling strategy, which includes actively emptying the storage facility capacity before rainfall occurs and preparing water in advance according to building water demand; S3, modular and hierarchical treatment: after the rainwater is discarded by the cyclone discarding device, it is guided to the primary treatment loop or the advanced treatment loop for treatment according to the water quality requirements of the final reuse scene; S4, precise distribution and reuse: the central controller operates the electric control valve according to the scheduling strategy and control instructions to deliver treated water of different qualities through the corresponding pipe network to the irrigation, toilet flushing, landscape water replenishment, or cooling tower water replenishment water use points.

10. The intelligent rainwater harvesting and reuse method as claimed in claim 9, wherein, In step S2, the generation of the scheduling strategy also combines the municipal time-of-use electricity price signal, and arranges the high-energy-consumption water treatment or backwashing operation to be performed at the electricity price valley period; and 1-6 hours before the rainfall, when the predicted rainfall is not less than 10 mm, the water level of the storage facility is lowered to 40%-50% of its safe capacity.

Images