Water purification system of direct drinking water dispenser

Abstract

The invention relates to the technical field of intelligent monitoring of filter element states, in particular to a water purification system of a direct drinking water dispenser, a space pressure situation sensing unit divides independent monitoring sectors through annular array pressure sensors, captures dynamic pressure gradient distribution data in real time, and sends the dynamic pressure gradient distribution data to a central processing unit; a life dynamic modeling unit extracts a pressure change trend and an instantaneous characteristic, activates blockage tracking when the pressure change trend and the instantaneous characteristic exceed a threshold value, constructs a non-linear attenuation model containing water temperature and turbidity compensation, maps a filter element fatigue accumulation factor, and calculates a filter element fatigue accumulation factor; a service life space distribution map is generated through cross-sector coupling, the end of the service life is judged, finally, a collaborative response execution unit generates a directional water flow signal balance load according to a map attenuation hot spot, a judgment result and a hot spot coordinate are transmitted to a user terminal, and visual early warning is driven. The problems that filter element monitoring is inaccurate, service life judgment deviation exists, and local blockage affects the whole are solved, and the monitoring precision and the filter element utilization rate are improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent monitoring technology for filter cartridge status, and more specifically, to a water purification system for a direct drinking water dispenser. Background Technology

[0002] Intelligent monitoring of filter cartridge status is an important technology. With the increasing demand for direct drinking water and the rising requirements for drinking water safety and cost-effectiveness, this technology is a key support for overcoming the limitations of traditional water dispenser filter cartridge management. It can monitor changes in filter cartridge filtration capacity in real time, avoiding reduced water flow and water quality deterioration caused by filter cartridge clogging. It can also accurately determine the lifespan of the filter cartridge, preventing resource waste caused by premature replacement or health risks caused by exceeding the expiration date. At the same time, it can extend the overall service life of the filter cartridge by dynamically adjusting the water flow to balance the filter cartridge load. Existing direct drinking water dispenser water purification systems face core problems in practical applications: inaccurate filter condition monitoring, large deviations in lifespan prediction, and the tendency for localized blockages to lead to overall performance degradation. Traditional systems often use a single pressure sensor to monitor the overall pressure of the filter element, without segmenting the circumferential direction of the filter element for monitoring. This makes it impossible to identify localized blockage areas. When a certain area of ​​the filter element becomes blocked due to impurities, the overall pressure change is not obvious and is easily overlooked, leading to a continuous worsening of the blockage in that area. This causes water to concentrate in other areas, accelerating the accumulation of impurities and fatigue damage in non-blocked areas. Furthermore, pressure monitoring fails to separate water flow pulsation noise from the actual blockage trend and does not... Environmental compensation factors such as inlet water temperature and turbidity often misjudge pressure fluctuations caused by water flow fluctuations or environmental changes as blockages, or miss true blockage signals. Lifespan determination often relies on fixed usage time or simple pressure thresholds, without considering the fatigue accumulation characteristics and local attenuation differences of filter material. This either triggers replacement instructions too early, resulting in filter waste, or delays in judgment, leading to excessive clogging of the filter, resulting in substandard water quality or even damage to the filter structure. These defects accumulate, not only affecting the safety of direct drinking water quality, but also increasing user costs and maintenance burdens. To solve this technical problem, we have provided a water purification system for direct drinking water dispensers. Summary of the Invention

[0003] The purpose of this invention is to provide a water purification system for a direct drinking water dispenser to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, a water purification system for a direct drinking water dispenser is provided, comprising: The spatial pressure situational awareness unit consists of multiple pressure sensors arranged in a ring array on the inner wall of the PP cotton pre-filter housing. Each sensor forms an independent monitoring sector in the circumferential direction, and captures the dynamic pressure gradient distribution data of the independent monitoring sector in real time. The lifetime dynamic modeling unit is connected to the space pressure situational awareness unit and is used to perform the following lifetime analysis mechanisms: Based on the dynamic pressure gradient distribution data, the pressure change trend of the independent monitoring sector is extracted, and the instantaneous characteristic quantity of pressure change over time is analyzed. When a specific sector experiences a sudden increase in the pressure change rate and continuously exceeds the preset sensitivity threshold, the blockage evolution tracking mode of that sector is activated. A nonlinear decay function model is constructed based on the growth rate and duration of the pressure change rate, and the dynamic pressure parameters are mapped to the fatigue accumulation factor of the filter material. Through cross-sector fatigue factor coupling calculation, a spatial distribution map of the effective life of the filter element is generated, and the life end determination is triggered based on the critical state of the sector with the largest decay in the spatial distribution map. The collaborative response execution unit receives the output instructions from the lifetime dynamic modeling unit and generates a directional water flow suppression signal based on the attenuation hot spot location of the spatial distribution map to balance the filter cartridge load. At the same time, it transmits the lifetime end judgment result and the spatial attenuation hot zone coordinates of the spatial distribution map to the user terminal to drive the generation of a visual early warning interface.

[0005] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention utilizes a spatial pressure situational awareness unit with multiple pressure sensors arranged in a ring array to divide the system into independent monitoring sectors. Combined with multi-level filtering and fusion technology, it separates water flow pulsation noise and impurity accumulation trends, identifies abnormal pressure islands with localized blockages, and corrects the pressure benchmark. A lifespan dynamic modeling unit constructs a nonlinear decay function model with environmental compensation factors based on pressure change trends and instantaneous characteristics. This model maps pressure parameters to a three-dimensional fatigue accumulation factor and generates a filter cartridge lifespan spatial distribution map through cross-sector coupling calculations. The lifespan is determined based on the critical state of the maximum decay sector. A collaborative response execution unit generates directional water flow suppression signals based on decay hotspots to balance the filter cartridge load and transmits the determination results and hotspot coordinates to the user terminal for visual early warning. This achieves precise zoned monitoring of filter cartridge status, accurate dynamic lifespan determination, timely control of localized blockages, and load balancing, effectively solving the problems of inaccurate filter cartridge status monitoring, large lifespan determination deviations, and overall performance degradation caused by localized blockages in direct drinking water dispenser water purification systems. Attached Figure Description

[0006] Figure 1 This is an overall block diagram of the present invention.

[0007] The meanings of the labels in the diagram are as follows: 1. Space pressure situational awareness unit; 2. Lifetime dynamic modeling unit; 3. Cooperative response execution unit. Detailed Implementation

[0008] 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.

[0009] This invention provides a water purification system for a direct drinking water dispenser. Please refer to [link / reference]. Figure 1 As shown, it includes: The spatial pressure situational awareness unit 1 consists of multiple pressure sensors arranged in a ring array on the inner wall of the PP cotton pre-filter housing. Each sensor forms an independent monitoring sector in the circumferential direction and captures the dynamic pressure gradient distribution data of the independent monitoring sector in real time. Lifetime dynamic modeling unit 2 is connected to the space pressure situational awareness unit and is used to perform the following lifetime analysis mechanism: Based on the dynamic pressure gradient distribution data, the pressure change trend of the independent monitoring sector is extracted, and the instantaneous characteristic quantity of pressure change over time is analyzed. When a specific sector experiences a sudden increase in the pressure change rate and continuously exceeds the preset sensitivity threshold, the blockage evolution tracking mode of that sector is activated. A nonlinear decay function model is constructed based on the growth rate and duration of the pressure change rate, and the dynamic pressure parameters are mapped to the fatigue accumulation factor of the filter material. Through cross-sector fatigue factor coupling calculation, a spatial distribution map of the effective life of the filter element is generated, and the life end determination is triggered based on the critical state of the sector with the largest decay in the spatial distribution map. The collaborative response execution unit 3 receives the output instructions from the lifetime dynamic modeling unit 2, and generates a directional water flow suppression signal based on the attenuation hot spot location of the spatial distribution map to balance the filter cartridge load. At the same time, it transmits the lifetime end judgment result and the spatial attenuation hot zone coordinates of the spatial distribution map to the user terminal to drive the generation of a visual early warning interface.

[0010] In the space pressure situational awareness unit 1, the extraction of dynamic pressure gradient distribution data employs a multi-level filtering and fusion technique, specifically including: The original pressure signal of each independent monitoring sector is decomposed in time to separate the high-frequency noise component caused by water flow pulsation and the low-frequency trend component reflecting impurity accumulation. By establishing a pressure gradient coupling matrix between adjacent sectors, abnormal pressure island areas caused by local blockage are identified. Based on the direction of pressure difference change between this area and the surrounding sectors, the pressure reference value of each sector is dynamically corrected to eliminate the interference of pipeline pressure fluctuations on trend analysis.

[0011] When extracting instantaneous features of pressure changing over time, the life dynamic modeling unit 2 specifically includes: The low-frequency trend component after multi-stage filtering is input into the transient feature extraction module. This module uses a sliding time window to capture the derivative change point of the pressure trajectory. When a series of inflection points are detected in the pressure curve, the acceleration of pressure change between the inflection points is extracted as the core instantaneous feature quantity, and the time span of the continuous action of the feature quantity is recorded to form a pressure impact intensity-duration combination parameter.

[0012] The mechanism for determining when a specific sector experiences a sudden increase in the rate of pressure change that continuously exceeds a preset sensitivity threshold includes: A sudden increase in pressure change rate is defined as the acceleration value in the pressure impact intensity-duration combination parameter exceeding the adaptive threshold boundary. This boundary value is dynamically calibrated by the standard deviation of pressure fluctuation in the initial clean state of the filter element. The preset sensitivity threshold is constructed by multiplying the pressure impact intensity and duration to form a dynamic judgment interval. When the product value continues to increase and exceeds the upper limit of the interval in multiple consecutive sampling periods, the blockage evolution tracking mode is triggered.

[0013] A nonlinear decay function model is constructed based on the growth rate and duration of the pressure change rate, specifically including: Using pressure impact intensity as the independent variable input and cumulative duration as the attenuation coefficient adjustment factor, an exponential material fatigue equation is constructed. This equation quantifies the molecular chain fracture effect under high pressure impact as the pore collapse rate of the fiber structure, so that the pore collapse rate increases exponentially for every order of magnitude increase in pressure impact intensity per unit time.

[0014] An environmental compensation factor is introduced into the nonlinear decay function model, specifically including: The calculation model for pore collapse rate is corrected by real-time water temperature data obtained by temperature sensors. The coefficient of change of the exponential equation is automatically reduced in low-temperature environments to compensate for the influence of water viscosity. At the same time, the influent turbidity detection signal is integrated, and the time integral domain of the decay function is dynamically expanded when turbidity changes abruptly to avoid misjudgment of instantaneous pressure shock caused by high turbidity raw water.

[0015] Mapping pressure dynamic parameters to fatigue accumulation factors for filter material specifically includes: A layered mapping architecture is adopted. The base layer converts the corrected pore collapse rate into the microcrack density increment of PP cotton fiber. The decision layer maps the two-dimensional plane pressure parameter into a three-dimensional fatigue accumulation factor based on the diffusion depth of the microcrack density in the thickness direction of the filter element. The weight in the depth direction increases stepwise from the inner layer to the outer layer.

[0016] Spatial distribution maps are generated through cross-sector fatigue factor coupling calculations, specifically including: Based on the three-dimensional fatigue accumulation factor, a sector interference model is established to identify the gradient transfer effect of fatigue factors in adjacent sectors. By calculating the energy transfer equivalent from high fatigue sector to low fatigue sector, the actual attenuation value of low fatigue sector is corrected. Finally, the correction values ​​of each sector are integrated to generate a spatial distribution map with spatial continuity. Its color distribution reflects the remaining life intensity in the circumferential direction of the filter element.

[0017] The lifetime termination determination is triggered based on the critical state of the sector with the largest decay in the spectrum, specifically including: When an isolated extreme attenuation zone appears in the spatial distribution map, the fatigue factor gradient change rate at the edge of the zone is detected. If the gradient change rate exceeds the material stress diffusion critical value, it is determined to be irreversible structural damage and a replacement command is triggered. Otherwise, a local regeneration program is started, and the load in the zone is reduced by directional water flow suppression signal until the gradient change rate returns to the normal range.

[0018] A directional water flow suppression signal is generated based on the attenuation hotspot locations in the spatial distribution map, specifically including: The coordinates of the attenuation hotspots with abnormal gradient change rate in the spatial distribution map are analyzed, and the water flow impact angle in the tangential direction of the coordinates is calculated. By controlling the water inlet solenoid valve to generate a pulse phase modulation wave opposite to the water flow impact angle, a vortex barrier zone is formed in front of the attenuation hotspot. At the same time, the water inlet flow rate of the adjacent sector is adjusted to form a pressure balance zone on both sides of the barrier zone, forcing large particles of impurities to bypass and achieve dynamic load balance.

[0019] Further explanation is needed regarding the specific implementation of the multi-level filtering and fusion technology in the spatial pressure situational awareness unit. After the spatial pressure situational awareness unit 1 captures the original pressure signals of each independent monitoring sector through the ring array pressure sensor, in order to accurately extract dynamic pressure gradient distribution data, a multi-level filtering and fusion technology is required to eliminate interference from water flow pulsations and pipeline fluctuations, and to identify local blockage signals. The specific implementation method is as follows: First, the original pressure signal of each independent monitoring sector is decomposed into time-series components using wavelet decomposition technology. The db4 wavelet basis function is selected to balance time and frequency domain resolution and adapt to the non-stationary characteristics of the pressure signal. The original pressure signal is decomposed into five layers of wavelet coefficients. Layers 1-2 correspond to high-frequency components (12.5-50Hz), while layers 3-5 correspond to low-frequency components (0-6.25Hz). The high-frequency noise component refers to the pulsating signal caused by turbulence and pipe vibration during water flow through the filter cartridge. This component is irregular and unrelated to filter cartridge blockage. The low-frequency trend component, reflecting impurity accumulation, refers to the slow pressure change signal caused by the gradual accumulation of impurities in the filter cartridge. This component is continuous and is the core basis for judging blockage. During separation, layers 3-5 of the low-frequency coefficients are directly retained and the signal is reconstructed, while layers 1-2 of the high-frequency coefficients are discarded, completing the high and low frequency decomposition. Preliminary separation of frequency components is achieved by establishing a pressure gradient coupling matrix between adjacent sectors to identify abnormal pressure island regions caused by localized blockages. The pressure gradient coupling matrix is ​​a square matrix characterizing the pressure correlation between adjacent sectors. It is constructed with the number of monitored sectors as the dimension, and the matrix elements represent the real-time pressure gradient difference between two adjacent sectors, i.e., the pressure difference between sector i and sector j, ΔPij = Pi - Pj. Elements of non-adjacent sectors are set to 0. An abnormal pressure island region refers to a sector whose pressure is significantly higher than all adjacent sectors, and whose pressure gradient difference exceeds the normal range by 20%. During identification, the pressure gradient coupling matrix is ​​traversed. When the pressure difference between more than 3 adjacent elements in a row (corresponding to a sector) is greater than 0.5 kPa, the sector can be determined as an abnormal pressure island region. Finally, based on the direction of change in the pressure difference between this region and surrounding sectors, the pressure reference values ​​for each sector are dynamically adjusted. First, normal sectors surrounding the abnormal pressure island area are selected, and the average pressure value of these normal sectors is calculated. This average value is used as the correction benchmark. For the abnormal island area, its original benchmark reference value is corrected to the average pressure value plus the normal pressure gradient to avoid benchmark deviation caused by overall pipeline pressure fluctuations. For other normal sectors, the 70th percentile of its own historical normal pressure is used as the basis, and the average pressure fluctuation value of the surrounding sectors is superimposed as the corrected benchmark. Through this correction, the interference of pipeline pressure fluctuations on trend analysis can be eliminated. For example, when the overall pipeline pressure increases by 0.3 kPa, the benchmark of each sector is simultaneously adjusted upward by 0.3 kPa to ensure that the pressure change trend only reflects the accumulation of impurities, rather than external fluctuations.

[0020] After the low-frequency trend component of the spatial pressure situational awareness unit 1 is output and filtered through multiple stages, the life dynamic modeling unit 2 needs to extract the instantaneous characteristic of pressure changing with time from it to capture the pressure change signal in the early stage of filter blockage. The specific implementation method is as follows: The low-frequency trend component, after multi-stage filtering, is input into the transient feature extraction module. This module uses a sliding time window to capture the derivative abrupt change point of the pressure trajectory. The sliding time window refers to a fixed time interval, set to 5 seconds based on the pressure change rate, with a sliding step of 1 second, ensuring coverage of short-term pressure changes without missing details. For example, the first window covers 0-5 seconds, the second covers 1-6 seconds, and so on. The derivative abrupt change point refers to the slope of the pressure curve, that is, the time point when the pressure change rate suddenly changes significantly. During calculation, the pressure data within each sliding window is linearly fitted to obtain the average slope within the window, and then the difference in slope between adjacent windows is calculated. When the difference exceeds 0.1 kPa / s, the termination time point of that window is determined as the derivative abrupt change point. When consecutive inflection points are detected on the pressure curve, and derivative abrupt changes are detected in three consecutive sliding windows with the same direction of change (e.g., all with increasing slope), the pressure change acceleration between the inflection points is extracted as the core transient feature quantity. First, record the pressure change rate corresponding to two adjacent abrupt change points of derivative, such as the slope of the first abrupt change point being 0.1 kPa / s, the second 0.3 kPa / s, and the third 0.6 kPa / s. Then, calculate the acceleration between adjacent abrupt change points: Acceleration = (Later slope - Previous slope) / Time interval, where the time interval is a sliding step of 1 second, such as 0.3 - 0.1 = 0.2 kPa / s², 0.6 - 0.3 = 0.3 kPa / s². Take the average of these accelerations as the core instantaneous characteristic quantity (e.g., (0.2 + 0.3) / 2 = 0.25). The system records the duration of the pressure shock intensity (0.25 kPa / s²) and the duration of the shock. The duration of the shock is the total time from the appearance of the first derivative mutation point to the disappearance of the last derivative mutation point. For example, if three mutation points appear consecutively within 0-3 seconds, the duration is 3 seconds. The core instantaneous characteristic (0.25 kPa / s²) is combined with the duration (3 seconds) to form a pressure shock intensity-duration combination parameter (e.g., "0.25 kPa / s²-3s"). This parameter quantifies the intensity and duration of the pressure shock, providing a basis for subsequent judgment of blockage evolution.

[0021] The mechanism for determining sudden increases in pressure change rate is based on the pressure impact intensity-duration combination parameters generated above. A mechanism needs to be established to identify sudden increases in pressure change rate in specific sectors, avoiding misjudging instantaneous water flow disturbances as blockages. The specific implementation method is as follows: A sudden increase in the rate of pressure change is defined as the acceleration value in the pressure shock intensity-duration combination parameter exceeding the adaptive threshold boundary. This adaptive threshold boundary is dynamically calibrated by the standard deviation of pressure fluctuations in the initial clean state of the filter cartridge. During the initial installation phase (the first 7 days, considered the initial clean state), daily pressure change rate data is collected, and its standard deviation (e.g., 0.08 kPa / s²) is calculated. The threshold boundary is set to twice this standard deviation (0.16 kPa / s²). If the acceleration value in the combination parameter (e.g., 0.25 kPa / s²) exceeds this boundary, a sudden increase in the rate of pressure change is preliminarily identified. A preset sensitivity threshold is constructed by multiplying the pressure shock intensity and duration to create a dynamic judgment interval, thus controlling the pressure shock intensity. Multiply the degree (i.e., acceleration value, such as 0.25 kPa / s²) by the duration span (such as 3 seconds) to obtain the product value (0.75 kPa·s). Based on historical blockage data, set the upper limit of the dynamic judgment interval (such as 0.6 kPa·s, which is the 95th percentile of the product value under normal pressure fluctuation). When the product value continuously increases (such as 0.5→0.6→0.75) and exceeds the upper limit of the interval (0.6) within 3 consecutive sampling cycles (1 second per cycle), it indicates that the sudden increase in the pressure change rate is not an instantaneous disturbance, but a continuous change caused by local blockage of the filter element. At this time, the blockage evolution tracking mode of the sector is triggered, and the subsequent nonlinear decay function model construction and fatigue accumulation factor calculation are started.

[0022] The specific implementation of the nonlinear decay function model construction: After triggering the blockage evolution tracking mode in a specific sector, the life dynamic modeling unit 2 needs to construct a nonlinear decay function model based on the growth rate and duration of the pressure change rate to quantify the fatigue degradation process of the filter material. The pressure change rate alone cannot accurately reflect the cumulative damage to the PP cotton fiber structure caused by long-term high-pressure impact. Therefore, it is necessary to establish the correlation between pressure and degradation through an exponential material fatigue equation. The specific implementation is as follows: First, the core variables of the model are defined, with pressure impact intensity as the input independent variable. Pressure impact intensity is the extracted core instantaneous feature quantity, referring to the average acceleration value between consecutive inflection points of the pressure curve (e.g., 0.25 kPa / s²). Its magnitude directly reflects the force intensity of the high-pressure impact on PP cotton fibers. The cumulative duration is used as the attenuation coefficient adjustment factor. The cumulative duration refers to the total time after the pressure impact intensity exceeds the adaptive threshold boundary. For example, if it remains in the over-threshold state for 5 consecutive seconds, the cumulative duration is 5 seconds. The longer the duration, the more significant the cumulative damage to the fiber structure. Based on the above variables, an exponential material fatigue equation is constructed. The core logic of this equation is to quantify the molecular chain breakage effect under high-pressure impact as the fiber structure pore collapse rate. The molecular chain breakage effect refers to the breakage of covalent bonds between molecules in PP cotton fibers (mainly polypropylene) under continuous high pressure, leading to a decrease in fiber toughness and a loose structure. The fiber structure pore collapse rate refers to the rate of change of the covalent bonds between molecules in the filter element. To determine the rate at which the micron-sized pores of a filter collapse due to fiber breakage and loss of support, the equation was constructed by first calibrating initial parameters through experiments. In a laboratory environment, different pressure impact intensities (0.1-2 kPa / s²) were applied to a clean PP cotton filter element, and the corresponding pore collapse rates were recorded. It was found that when the pressure impact intensity increased from 0.2 kPa / s² (one order of magnitude) to 2 kPa / s² (the next order of magnitude), the pore collapse rate increased exponentially from 0.1 mm / d to 1 mm / d. That is, for every order of magnitude increase in pressure impact intensity, the pore collapse rate increased by approximately 10 times. Then, the cumulative duration was incorporated into the equation as an adjustment factor. For example, when the cumulative duration increased from 5 seconds to 10 seconds, the attenuation coefficient increased from 1.0 to 1.5, further increasing the pore collapse rate by 50%. Finally, a nonlinear attenuation function model that can simultaneously reflect pressure intensity and duration was formed, ensuring that the pore collapse rate output by the model is consistent with the actual degradation state of the filter element.

[0023] Specific Implementation of Introducing Environmental Compensation Factor into Nonlinear Decay Function Model The nonlinear decay function model constructed above is based on standard environmental conditions by default. However, in actual use, fluctuations in water temperature and sudden changes in influent turbidity can interfere with the calculation of pore collapse rate. For example, low temperature increases water viscosity, which can easily be misjudged as pore collapse caused by increased pressure. Instantaneous impurity impacts from high-turbidity raw water can easily be misjudged as blockage. Therefore, it is necessary to introduce an environmental compensation factor into the model. The specific implementation is as follows: First, the pore collapse rate calculation model is corrected using real-time water temperature data acquired by a temperature sensor. An NTC temperature sensor (measuring range 0-50℃, accuracy ±0.5℃) is installed on the outer wall of the filter cartridge's inlet pipe. Real-time water temperature data is collected every 10 seconds and transmitted to the lifespan dynamic modeling unit 2. The main effect of water temperature on the model is reflected in changes in water viscosity. In low-temperature environments (e.g., 10℃), water viscosity increases by approximately 50% compared to the standard 25℃, increasing the resistance to water flow through the filter cartridge and causing the pressure sensor to detect a higher pressure value. Directly substituting this into the model would overestimate the pore collapse rate. Therefore, a water temperature-viscosity-pressure correction coefficient mapping table is established for correction. For example, the correction coefficient is 1.0 at 25℃ (no correction required) and 0 at 20℃. The coefficient is 0.95 at 15℃ and 0.85 at 10℃. This coefficient is used to correct the detected pressure impact intensity. For example, if the detected pressure impact intensity at 10℃ is 0.25 kPa / s², the corrected value is 0.25 × 0.85 = 0.21 kPa / s². This value is then substituted into the model to calculate the pore collapse rate, avoiding calculation errors caused by water temperature. In low-temperature environments, the steepness coefficient of the exponential equation needs to be automatically reduced to further compensate for the influence of water viscosity. The steepness coefficient is a core parameter controlling the growth rate of the fatigue equation for exponential materials. The larger the coefficient, the more rapidly the pore collapse rate increases with pressure impact intensity. For example, a steepness coefficient of 1.2 results in a rapid rate increase, while a coefficient of 0.8 results in a slower increase. The specific reduction method is as follows: The preset water temperature-abrupt change coefficient correspondence is as follows: at 25℃, the abrupt change coefficient is 1.2, and for every 5℃ decrease in water temperature, the abrupt change coefficient decreases by 0.1. For example, at 10℃, the abrupt change coefficient drops from 1.2 to 0.9, making the pore collapse rate output by the model more closely match the actual degradation rate of fibers at low temperatures. At low temperatures, fiber molecular activity slows down, and the rate of molecular chain breakage decreases. It is necessary to suppress the excessively rapid increase in the rate by reducing the abrupt change coefficient. At the same time, the influent turbidity detection signal is integrated to avoid misjudgment caused by instantaneous pressure shocks from high turbidity raw water. An optical turbidity sensor is installed at the influent end of the filter element to monitor the turbidity of the influent in real time. A turbidity mutation is defined as a sudden increase in turbidity value from a stable range (e.g., 5-15 NTU) to above 30 NTU within 1 minute (e.g., a sudden rainstorm causing an increase in the turbidity of municipal water supply). At this time, a large number of suspended impurities in the raw water will instantaneously... Clogged filter surface pores cause a sudden pressure surge, but this pressure surge is temporary. Impurities only adhere to the surface and do not penetrate deep into the filter to cause pore collapse. If calculated using a conventional model, this would be misjudged as a clogging evolution. To address this situation, the system automatically and dynamically expands the time integration domain of the decay function. The time integration domain refers to the time range within which the model calculates the pore collapse rate (default 10 seconds). When turbidity changes abruptly, the integration domain is expanded from 10 seconds to 30 seconds. By extending the integration time, the system judges whether the pressure continues to rise. If the pressure drops back to the normal range within 30 seconds (indicating a transient impurity impact), the pore collapse rate is not updated. If the pressure continues to rise (indicating that impurities have penetrated deep into the filter), the rate is calculated based on the expanded integration result. This effectively avoids misjudgments caused by high turbidity raw water and ensures the accuracy of model calculations.

[0024] The specific implementation method for mapping pressure dynamic parameters to the fatigue accumulation factor of filter material involves calculating the corrected pore collapse rate using a nonlinear decay function model (including environmental compensation factor). This pressure dynamic parameter then needs to be further mapped to the fatigue accumulation factor of the filter material. The fatigue accumulation factor is a core indicator for quantifying the decline in filtration performance of filter material (PP cotton) due to continuous degradation. Its value ranges from 0 to 1, where 0 represents no fatigue damage and 1 represents complete failure. The mapping process employs a layered mapping architecture, taking into account both planar pressure distribution and damage differences in the thickness direction. The specific implementation method is as follows: The hierarchical mapping architecture is divided into a base layer and a decision layer. The core task of the base layer is to convert the corrected pore collapse rate into the microcrack density increment of PP cotton fibers. The microcrack density increment of PP cotton fibers refers to the increase in the number of microcracks (10-100μm in length) caused by pore collapse within a unit volume of PP cotton fibers (unit: cracks / mm³·d). The higher the microcrack density, the worse the integrity of the fiber structure. The specific conversion process is as follows: First, an experimental correlation table between pore collapse rate and microcrack density increment was established. The number of microcracks corresponding to different pore collapse rates (0.05-2 mm / d) was observed under a microscope. It was found that when the pore collapse rate was 0.1 mm / d, the microcrack density increment was 0.05 cracks / mm³·d; when the rate was 0.2 mm / d, the increment was 0.12 cracks / mm³·d. Then, based on the real-time calculated corrected pore collapse rate (e.g., 0.15 mm / d), the corresponding microcrack density increment (e.g., 0.08 cracks / mm³·d) was determined by interpolation, completing the parameter transformation of the base layer. The decision layer then maps the two-dimensional planar pressure parameters to a three-dimensional fatigue accumulation factor based on the diffusion depth of the microcrack density in the filter element thickness direction. PP cotton filter elements are typically cylindrical structures (thickness 5-10 mm). The filter cartridge consists of a pre-filtration layer and a fine filtration layer, arranged sequentially from the outermost layer to the innermost layer. The diffusion depth of microcracks refers to the distance (in mm) from the outermost layer (pre-filtration layer) to the innermost layer (fine filtration layer). The outermost layer is in direct contact with the raw water and has more impurities attached. Microcracks first form in the outermost layer and gradually diffuse inward. Microcracks in the inner fine filtration layer have a greater impact on filtration performance. During mapping, the diffusion depth is first determined by detecting the microcrack density at different thicknesses of the filter cartridge using an ultrasonic thickness gauge. For example, if microcracks are detected to have diffused from the outermost layer to a depth of 2 mm (inner layer direction), then the two-dimensional planar pressure parameters (pressure values ​​of each monitoring sector) are combined with the diffusion depth to construct a three-dimensional coordinate system (X / Y are the sector planar coordinates, Z is the thickness depth). The key is that the weight of the depth direction increases progressively from the innermost layer to the outermost layer. The specific implementation method is as follows: The filter element thickness is divided into three layers (outer layer 0-3mm, middle layer 3-6mm, inner layer 6-10mm). The outer layer is pre-weighted at 0.2, the middle layer at 0.3, and the inner layer at 0.5. For every 3mm increase in thickness towards the inner layer, the weight increases by 0.15, ensuring that microcracks in the inner layer (fine filtration layer) contribute more to the fatigue accumulation factor. Finally, the three-dimensional fatigue accumulation factor is calculated by multiplying the microcrack density increment of each thickness layer by its corresponding weight, and then adding this to the normalized value of the planar pressure parameter (pressure value / maximum pressure threshold). The weighted summation is as follows: for example, the microcrack increment in the outer layer is 0.08 × 0.2 = 0.016, in the middle layer it is 0.05 × 0.3 = 0.015, and in the inner layer it is 0.03 × 0.5 = 0.015. The normalized value of the plane pressure is 0.6 × 0.4 = 0.24, and the total fatigue accumulation factor is 0.016 + 0.015 + 0.015 + 0.24 = 0.286. This value can reflect both the pressure difference in the plane sector and the damage distribution in the thickness direction, providing core data for the subsequent generation of spatial distribution maps.

[0025] The specific implementation method for constructing the sector interference model and generating the spatial distribution map is as follows: After obtaining the three-dimensional fatigue accumulation factor of each sector, the life dynamic modeling unit 2 needs to consider the fatigue transmission effect between adjacent sectors. The blockage of a single sector will affect the surrounding sectors through water pressure transmission. For example, the pressure increase in a high-fatigue sector will force the water flow to shift towards a low-fatigue sector, accelerating the accumulation of impurities in the low-fatigue sector. Ignoring this interference will lead to distortion of the spatial distribution map. Therefore, it is necessary to establish a sector interference model and correct the attenuation value. The specific implementation method is as follows: First, a sector interference model is established based on the three-dimensional fatigue accumulation factor. This model takes eight monitoring sectors of a ring array as the research object. The core is to identify the gradient transmission effect of fatigue factors between adjacent sectors. The gradient transmission effect refers to the pressure gradient of high fatigue sectors (fatigue accumulation factor > 0.5) being transmitted to low fatigue sectors (fatigue accumulation factor < 0.3), causing the actual attenuation value of low fatigue sectors to be higher than the calculated result alone. The identification process requires first calculating the difference in fatigue factors between adjacent sectors (e.g., sector 3 has a fatigue factor of 0.6, adjacent sector 2 has 0.2, the difference is 0.4), and then setting the gradient transmission threshold, which is experimentally calibrated to 0.3. That is, when the difference exceeds 0.3, there is a significant transmission effect. If the difference exceeds... The threshold is used to further detect the pressure flow direction between the two sectors. The pressure difference between adjacent sensors determines that the high-pressure sector flows to the low-pressure sector, confirming the high-fatigue sector as the source and the low-fatigue sector as the receiver, thus identifying the gradient transfer effect. The energy transfer equivalent from the high-fatigue sector to the low-fatigue sector is calculated to correct the attenuation value. The energy transfer equivalent is a parameter that quantifies the additional attenuation from the high-fatigue sector to the low-fatigue sector. The calculation requires considering two key factors: the difference in fatigue factors between adjacent sectors (the larger the difference, the greater the transfer equivalent), and the distance between the sectors (in a ring array, the distance between adjacent sectors is fixed at 1 / 4 of the filter element radius; the closer the distance, the more significant the transfer). The specific calculation process is as follows: First, the fatigue factor difference is normalized (e.g., a difference of 0.4 is normalized to 0.4 / 1.0=0.4, where 1.0 is the maximum possible difference). Then, it is multiplied by a distance correction factor, which is fixed at 0.8. Since the spacing is fixed, the energy transfer equivalent is obtained. This equivalent is then superimposed on the original attenuation value of the low-fatigue sector. For example, the original attenuation value of sector 2 is 0.2, and after correction, it becomes 0.2+0.32=0.52. This ensures that the corrected attenuation value can reflect the interference effect of adjacent sectors. Finally, the corrected values ​​of each sector are fused to generate a spatial distribution map with spatial continuity. During fusion, the corrected attenuation value of each sector is first marked on a two-dimensional plane in circular coordinates and polar coordinates, with the angle corresponding to the sector position and the radius corresponding to the attenuation value. Then, a bilinear interpolation algorithm is used. To fill in the blank areas between sectors, such as between sector 2 and sector 3, attenuation data at the intermediate position is generated by interpolation based on the attenuation values ​​of the two sectors, ensuring the spatial continuity of the spectrum. The color gradient distribution of the spectrum uses a red-yellow-green three-color gradient to map the remaining life intensity of the filter element in the circumferential direction. The red area represents an attenuation value > 0.7 (remaining life < 30%, requiring close attention), the yellow area represents an attenuation value 0.3-0.7 (remaining life 30%-70%, normal use), and the green area represents an attenuation value < 0.3 (remaining life > 70%, good condition). The color gradient transition is smooth and the specific attenuation value range is marked. Maintenance personnel can intuitively judge the degradation distribution in the circumferential direction of the filter element through the spectrum, providing accurate hotspot location basis for subsequent directional water flow adjustment.

[0026] The specific implementation method for determining the end of life based on spatial distribution maps is as follows: After generating the spatial distribution map, the life dynamic modeling unit 2 needs to trigger the end of life determination based on the critical state of the sector with the largest attenuation in the map. This is to prevent irreversible damage in a single sector from spreading to the entire filter element. For example, local structural collapse can cause impurities to directly penetrate the filter element, affecting the quality of the effluent. The specific implementation method is as follows: First, identify isolated extreme attenuation regions in the spatial distribution map. These regions refer to sectors where the attenuation value is significantly higher than all adjacent sectors (difference > 0.5), and they are not connected to other high attenuation regions (e.g., sector 5 has an attenuation value of 0.9, adjacent sectors 4 and 6 both have 0.3, and there are no other sectors with attenuation values ​​> 0.7). These regions are often caused by concentrated damage due to localized blockage and are the core basis for determining whether the filter needs replacement. Detect the fatigue factor gradient change rate at the edge of this region. The fatigue factor gradient change rate refers to the ratio of the change in fatigue factor from the edge of the extreme attenuation region to the distance to the adjacent normal sector. For example, if the attenuation value at the edge point is 0.8, the adjacent normal point is 0.3, and the distance between the two points is 5mm, the gradient change rate = (0.8 - 0.3) / 5 = 0.1 / mm. This indicator reflects whether the damage has a diffusion trend. The larger the gradient change rate, the steeper the damage boundary, and the higher the possibility of structural damage. The critical value for material stress diffusion is... The threshold value calibrated by the mechanical test of the PP cotton filter element is 0.08 / mm, which represents the maximum stress diffusion rate that the filter element material can withstand. Exceeding this value means that the fiber structure has undergone irreversible breakage. If the molecular chain breaks and cannot be restored by water flow adjustment, if the gradient change rate (e.g., 0.1 / mm) exceeds the critical value, it is judged as irreversible structural damage, and the filter element replacement command is immediately triggered. At the same time, the coordinates and attenuation value of the area are recorded as the key inspection area during replacement. Conversely, if the gradient change rate (e.g., 0.06 / mm) is lower than the critical value, it means that the damage is still within the adjustable range. The local regeneration program is started. The directional water flow inhibition signal is generated by the collaborative response execution unit 3 to reduce the influent flow rate in the area, such as from the normal 1L / min to 0.5L / min, to reduce the continued accumulation of impurities until the gradient change rate returns to the normal range in subsequent tests, so as to avoid resource waste caused by excessive replacement.

[0027] The specific implementation method for generating directional water flow suppression signals and load balancing is as follows: After determining the location of attenuation hotspots and the coordinates of the red isolated extreme attenuation areas in the spatial distribution map, the collaborative response execution unit 3 needs to generate directional water flow suppression signals. By adjusting the water flow distribution to balance the filter cartridge load, the attenuation rate of the hotspot areas is slowed down. The specific implementation method is as follows: First, analyze the coordinates of the attenuation hotspots with abnormal gradient change rates in the spatial distribution map. Establish a polar coordinate system with the center of the annular filter cartridge as the origin. Convert the circumferential angle of the hotspot area (e.g., 67.5°, corresponding to the center of the second sector) and the radial depth (e.g., 2mm, corresponding to the middle layer of the filter cartridge) into specific physical coordinates (X = filter cartridge radius × sin67.5°, Y = filter cartridge radius × cos67.5°, Z = 2mm). Clarify the precise location of the hotspot on the filter cartridge. Calculate the tangential water flow impact angle at this coordinate. The water flow impact angle refers to the direction of the influent water flow at the hotspot location relative to the hotspot. The tangential angle, perpendicular to the radial direction, is determined by data from flow and pressure sensors in the inlet pipe. This data is used to fit the water flow trajectory on the filter element surface. For example, if the water flows clockwise along the circumference, the angle between the flow direction and the tangential angle at the hot spot is 30°. This angle directly determines the direction of subsequent water flow suppression. The inlet solenoid valve generates a pulse phase modulation wave opposite to the water flow impact angle. The inlet solenoid valve uses PWM (Pulse Width Modulation) control and is installed on the annular pipe at the filter element inlet. One solenoid valve is installed for each of the eight sectors. The control is based on the calculated water flow impact angle. The solenoid valve in the corresponding sector outputs a reverse pulse, forming a pulse phase-modulated wave opposite to the water flow direction. This wave creates a vortex barrier zone in front of the attenuation hotspot (upstream of the water flow direction). The vortex barrier zone is a rotating vortex formed by the reverse pulse driving the water flow, which can block most impurities from directly impacting the hotspot area with the water flow, forcing the impurities to change their flow path. At the same time, the inlet flow rate of adjacent sectors is adjusted to create a pressure balance zone on both sides of the barrier zone. Through the flow regulation function of the solenoid valve, the inlet flow rate of adjacent sectors (such as sectors 1 and 3) is increased from the normal 1L / min to 1.2L / min. With the flow rate reduced to L / min, the flow rate in the hot spot sector (sector 2) is reduced to 0.5 L / min, keeping the pressure difference between the two sides of the barrier zone within 0.1 kPa (pressure balance zone). This pressure balance prevents the water flow from shifting excessively due to the pressure difference, ensuring that impurities are evenly distributed across multiple sectors and achieving dynamic balance of filter cartridge load. For example, large particles (diameter > 10 μm) that would normally accumulate in sector 2 will be redirected to sectors 1 and 3 by the vortex barrier and pressure balance, reducing the input of impurities in the hot spot area, slowing down their decay rate, and extending the overall service life of the filter cartridge.

[0028] In this invention, the spatial pressure situational awareness unit 1 divides the independent monitoring sectors through a ring array pressure sensor, captures dynamic pressure gradient distribution data in real time, and uses multi-level filtering to separate water flow noise and impurity accumulation trends and correct the pressure benchmark. The lifespan dynamic modeling unit 2 extracts the pressure change trend and instantaneous characteristics, activates blockage tracking when the threshold is exceeded, and constructs a nonlinear attenuation model with water temperature and turbidity compensation to map the filter element fatigue accumulation factor. Through cross-sector coupling, a lifespan spatial distribution map is generated and the end of lifespan is determined. Finally, the collaborative response execution unit 3 generates directional water flow signals to balance the load according to the attenuation hotspots in the map, transmits the judgment results and hotspot coordinates to the user terminal, drives visual early warning, solves the problems of inaccurate filter element monitoring, lifespan judgment deviation, and the impact of local blockage on the overall system, and improves monitoring accuracy and filter element utilization.

[0029] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A water purification system for a direct drinking water water fountain, characterized in that, Comprise: The space pressure situation awareness unit (1) is composed of a plurality of pressure sensors arranged in a ring array on the inner wall of the PP cotton pre-filter cartridge shell, each sensor forms an independent monitoring sector in the circumferential direction, and real-time captures the dynamic pressure gradient distribution data of the independent monitoring sector; The life dynamic modeling unit (2) is connected with the space pressure situation awareness unit, and is used for executing the following life analysis mechanism: According to the dynamic pressure gradient distribution data, the pressure change trend of the independent monitoring sector is extracted, and the instantaneous characteristic quantity of the pressure change with time is analyzed. When the pressure change rate of a specific sector suddenly increases and continuously exceeds the preset sensitivity threshold, the clogging evolution tracking mode of the sector is activated. Based on the growth amplitude and duration of the pressure change rate, a nonlinear decay function model is constructed, the pressure dynamic parameters are mapped into the fatigue accumulation factor of the filter material, the fatigue factor coupling operation is performed across the sectors, the spatial distribution map of the effective life of the filter cartridge is generated, and the life termination judgment is triggered according to the critical state of the maximum decay sector in the spatial distribution map; The cooperative response execution unit (3) receives the output instruction of the life dynamic modeling unit (2), generates a directional water flow suppression signal to balance the filter cartridge load according to the decay hot spot direction of the spatial distribution map, and transmits the life termination judgment result and the spatial decay hot zone coordinates of the spatial distribution map to the user terminal to drive the visualization warning interface to generate.

2. A water purification system for a direct drinking water dispenser according to claim 1, characterized in that: In the space pressure situation awareness unit (1), the extraction of dynamic pressure gradient distribution data adopts multi-stage filtering fusion technology, specifically including: The original pressure signal of each independent monitoring sector is time-decomposed to separate the high-frequency noise component caused by water flow pulsation and the low-frequency trend component reflecting impurity accumulation. By establishing a pressure gradient coupling matrix between adjacent sectors, abnormal pressure island regions caused by local clogging are identified. Based on the pressure difference change direction between the region and the surrounding sectors, the pressure reference value of each sector is dynamically corrected to eliminate the interference of pipeline pressure fluctuation on trend analysis.

3. A water purification system for a direct drinking water dispenser according to claim 2, characterized in that: When the life dynamic modeling unit (2) extracts the instantaneous characteristic quantity of the pressure change with time, specifically including: The low-frequency trend component after multi-stage filtering is input into the transient feature extraction module. The module uses a sliding time window to capture the derivative sudden change point of the pressure trajectory. When a continuous inflection point is detected in the pressure curve, the acceleration of pressure change between the inflection points is extracted as the core instantaneous characteristic quantity, and the time span of the continuous action of the characteristic quantity is recorded to form the pressure impact strength-time length combination parameter.

4. A water purification system for a direct drinking water dispenser according to claim 3, characterized in that: The judgment mechanism when the pressure change rate of a specific sector suddenly increases and continuously exceeds the preset sensitivity threshold, specifically including: The sudden increase of the pressure change rate is defined as the acceleration value in the pressure impact strength-time length combination parameter breaking through the adaptive threshold boundary, and the boundary value is dynamically calibrated by the pressure fluctuation standard deviation under the initial clean state of the filter cartridge. The preset sensitivity threshold is constructed by the product operation of the pressure impact strength and the duration. When the product value continuously increases and exceeds the upper limit of the interval in a plurality of consecutive sampling periods, the clogging evolution tracking mode is triggered.

5. A water purification system for a direct drinking water dispenser according to claim 4, characterized in that: Based on the growth amplitude and duration of the pressure change rate, a nonlinear decay function model is constructed, specifically including: Take the pressure impact strength as the independent variable input, and take the cumulative duration as the attenuation coefficient adjustment factor to construct an exponential material fatigue equation. The equation quantifies the molecular chain rupture effect under high pressure impact as the pore collapse rate of the fiber structure, so that the pore collapse rate increases exponentially with each order of magnitude increase in the pressure impact strength per unit time.

6. A water purification system for a direct drinking water dispenser according to claim 5, characterized in that: An environmental compensation factor is introduced in the nonlinear attenuation function model, which specifically includes: The water temperature data obtained by the temperature sensor in real time is used to correct the pore collapse rate calculation model. In a low temperature environment, the steepness coefficient of the exponential equation is automatically reduced to compensate for the influence of water viscosity. At the same time, the turbidity detection signal is integrated. When the turbidity suddenly changes, the time integral domain of the attenuation function is dynamically expanded to avoid misjudgment of instantaneous pressure impact caused by high turbidity raw water.

7. A water purification system for a direct drinking water dispenser according to claim 6, characterized in that: The pressure dynamic parameters are mapped to the fatigue cumulative factor of the filter core material, which specifically includes: A hierarchical mapping architecture is adopted. The base layer converts the corrected pore collapse rate to the PP cotton fiber micro-crack density increment. The decision layer maps the two-dimensional plane pressure parameters to the three-dimensional fatigue cumulative factor according to the diffusion depth of the micro-crack density in the thickness direction of the filter core, and the depth direction weight gradually increases from the inner layer to the outer layer.

8. A water purification system for a direct drinking water water dispenser according to claim 7, characterized in that: A spatial distribution map is generated through cross-sector fatigue factor coupling calculation, which specifically includes: A sector interference model is established based on the three-dimensional fatigue cumulative factor to identify the gradient transmission effect of adjacent sector fatigue factors. The actual attenuation value of the low fatigue sector is corrected by calculating the energy transmission equivalent from the high fatigue sector to the low fatigue sector. Finally, the spatial distribution map with spatial continuity is generated by fusing the corrected values of each sector. The color scale distribution reflects the residual life intensity in the circumferential direction of the filter core.

9. A water purification system for a direct drinking water dispenser as defined in claim 1, characterized in that: The life termination judgment is triggered according to the critical state of the maximum attenuation sector in the map, which specifically includes: When an isolated extreme attenuation region appears in the spatial distribution map, the gradient change rate of the fatigue factor at the edge of the region is detected. If the gradient change rate exceeds the critical value of material stress diffusion, it is determined as irreversible structural damage, and the replacement instruction is triggered. Otherwise, the local regeneration program is started, and the load in the region is reduced through the directional water flow suppression signal until the gradient change rate returns to the normal range.

10. A water purification system for a direct drinking water dispenser as defined in claim 1, characterized in that: The directional water flow suppression signal is generated according to the orientation of the attenuation hot spot in the spatial distribution map, which specifically includes: The coordinates of the attenuation hot spot with abnormal gradient change rate in the spatial distribution map are analyzed, and the water flow impact angle in the tangential direction of the coordinates is calculated. A pulse phase modulation wave opposite to the water flow impact angle is generated by controlling the water inlet electromagnetic valve to form a vortex barrier zone in front of the attenuation hot spot. At the same time, the water inlet flow of the adjacent sector is adjusted to form a pressure balance zone on both sides of the barrier zone, forcing large particles to bypass to achieve dynamic load balance.

Images