A water quality identification system and method based on multi-component ion mobility deviation characteristics

Abstract

The application discloses a water quality identification system and method based on multi-component ion mobility deviation characteristics. The system adopts a three-layer architecture of a hardware perception layer, an algorithm analysis layer and a performance application layer. The hardware perception layer collects TDS, pH, EC and temperature signals in real time through a multi-parameter integrated sensor array. The algorithm analysis layer includes a binary matrix model module, a nonlinear jump detection module and a dynamic thermal transient response module. The ion component type is identified by calculating the conductivity response deviation factor D of different ion components. The critical jump point of the performance mutation of the reverse osmosis membrane is captured as the filter core life warning criterion through the concentration gradient incremental test. The water samples with similar static conductivity but different ion compositions are distinguished through the thermal pulse excitation test. The performance application layer converts the analysis results into a multi-dimensional characteristic fingerprint radar chart and a performance evaluation report. The system is also provided with an adaptive correction module, which can automatically correct the judgment criteria when deployed across regions.

Description

Technical Field

[0001] This invention relates to the field of water quality monitoring and electrochemical analysis technology, specifically to a water quality identification system and method based on the deviation characteristics of multi-component ion mobility. Background Technology

[0002] With the widespread adoption of household and commercial reverse osmosis membrane water purification equipment, online monitoring of effluent water quality has become a fundamental requirement for ensuring drinking water safety. Currently, the common water quality monitoring method in the industry involves installing a TDS sensor at the outlet of the water purification equipment. The purification effect is reflected by measuring the total dissolved solids (TDS) value or conductivity (EC) value of the effluent. Some mid-to-high-end models also incorporate sensors at the inlet, using a comparison of inlet and outlet water data to assess desalination efficiency. Within this technical framework, the TDS sensor reading is directly used as a characterization parameter for the total mass concentration of dissolved ions in the water.

[0003] However, the aforementioned monitoring methods have inherent physical limitations. Commercially available TDS sensors mostly operate on the principle of AC-excited conductivity measurement, and their output essentially reflects the conductivity of the water sample, which is then converted to a TDS reading using an empirical conversion coefficient. Because the molar conductivity of different ion species varies significantly—for example, at the same mass concentration, monovalent sodium ions contribute far more to conductivity than divalent magnesium or calcium ions—using a single empirical coefficient for linear conversion inevitably introduces systematic bias. This means that when the ionic composition of a water sample changes, even if the TDS reading does not show abnormalities, the actual characteristics of the water quality may have already changed significantly, and this compositional change cannot be distinguished solely by the conductivity value.

[0004] From the perspective of electrochemical theory, the response of ion-selective electrodes to ion activity follows the Nernst equation: In the formula, E is the electrode potential. 0 Let be the standard electrode potential, R be the ideal gas constant, T be the absolute temperature, n be the ion valence, F be the Faraday constant, and a be the ion activity. This equation establishes a quantitative relationship between ion activity and electrode potential. However, in the real-time monitoring of reverse osmosis membrane water purification systems, existing technologies often simply apply the linear conductivity formula, neglecting the nonlinear deviation of ion migration velocity caused by the interference of the interfacial environment and multivalent components.

[0005] Regarding filter cartridge life monitoring, existing technologies primarily rely on cumulative usage time or cumulative water production to indicate cartridge replacement, with some solutions using an effluent TDS exceeding a fixed threshold as the alarm criterion. These methods are essentially based on linear prediction models with a single parameter, failing to consider the actual interception performance changes of reverse osmosis membranes under different ion load conditions. In practical use, when the influent ion load is high or the ion composition is complex, the interception performance of the reverse osmosis membrane does not decline linearly but may experience a sudden decrease under a certain concentration load condition. Traditional linear monitoring methods struggle to capture such abrupt changes, leading to delayed warnings.

[0006] In summary, existing water quality monitoring technologies for reverse osmosis membrane water purification equipment have the following shortcomings: First, relying on a single conductivity parameter for water quality assessment fails to distinguish the differences in the contribution of different ionic components to conductivity and lacks the ability to identify ionic composition characteristics; second, the filter cartridge life prediction uses a linear model, which cannot reflect the nonlinear changes in the interception performance of the reverse osmosis membrane under high load or complex ionic environments, resulting in insufficient accuracy and timeliness of early warnings; third, there is a lack of comprehensive analysis methods to correlate and analyze multi-dimensional sensor data and transform it into intuitive performance evaluation results. Summary of the Invention

[0007] In view of the above-mentioned shortcomings of the water quality monitoring of reverse osmosis membrane water purification equipment in the prior art, the present invention provides a water quality identification system and method based on the deviation characteristics of multi-component ion mobility.

[0008] This invention provides a water quality identification system based on the deviation characteristics of multi-component ion mobility. The system adopts a three-layer architecture consisting of a sensing layer, a parsing layer, and an application layer, including: The hardware sensing layer is equipped with a multi-parameter integrated sensor array for real-time acquisition of the total dissolved solids (TDS), pH, electrical conductivity (EC), and temperature (T) signals of the water sample being tested. The algorithm analysis layer, connected to the hardware sensing layer, receives the raw signals output by the multi-parameter integrated sensor array. The algorithm analysis layer includes a binary matrix model module and a nonlinear jump detection module. The binary matrix model module utilizes the nonlinear correlation between the response characteristics of conductivity sensors to ion mobility and the response characteristics of potential sensors to ion activity to construct a potential-conductivity binary matrix model. By calculating the conductivity response deviation of different ion components under the same concentration conditions, it identifies the characteristics of ion components in the tested water sample. The nonlinear jump detection module monitors the nonlinear abrupt changes in conductivity and TDS of the composite ion system during concentration gradient changes, capturing the critical jump point where the reverse osmosis membrane transitions from a linear interception state to a nonlinear synergistic permeation state, serving as a filter cartridge life warning criterion. The performance application layer is connected to the algorithm analysis layer and is used to convert the ionic component characteristic parameters and jump point criteria output by the algorithm analysis layer into multi-dimensional feature fingerprints and performance evaluation reports, so as to realize the quantitative expression of water quality characteristics and the visual output of filter cartridge status.

[0009] The aforementioned three-layer architecture enables the system of this invention to achieve simultaneous acquisition of multiple parameters at the hardware level, to leap from single-parameter monitoring to multi-dimensional correlation analysis at the algorithm level, and to transform from digital readings to intuitive feature representation at the application level. The sensor array in the hardware sensing layer simultaneously acquires signals from four dimensions: TDS, pH, EC, and temperature, providing a sufficient data foundation for the algorithm analysis layer. The binary matrix model module and nonlinear jump detection module in the algorithm analysis layer analyze water quality from two perspectives: ionic component identification and membrane performance mutation detection, respectively. These two modules complement each other, forming a complete water quality feature identification system.

[0010] Furthermore, the binary matrix model module identifies components by calculating the conductivity response deviation of the ionic components: under the same influent concentration conditions, the conductivity values ​​of the effluent after reverse osmosis filtration for different ionic components are obtained; the effluent conductivity EC of the component to be tested is then calculated. x Effluent conductivity EC compared to the reference component ref The deviation factor D is obtained through comparison. The deviation factor D is calculated using the following formula: In the formula, The conductivity of the effluent is based on the reference component (sodium chloride, NaCl). This refers to the effluent conductivity of the analyte (such as magnesium sulfate, MgSO4, etc.). The logical basis for this determination is that sodium chloride, as a typical monovalent strong electrolyte, has significantly higher ion mobility and effluent conductivity under the same conditions than divalent or polyvalent ion salts such as magnesium sulfate. Therefore, when the analyte is a polyvalent ion salt, due to EC... x Less than EC ref The calculated deviation factor D > 1. When D deviates from 1 by more than the preset deviation threshold, the system automatically determines that there are ionic components in the current water sample that have a significant difference in migration rate from the reference component.

[0011] The introduction of the deviation factor D provides a quantitative means for identifying ionic components. Different types of ions exhibit different conductivity responses after being filtered through a reverse osmosis membrane due to differences in their valence state, hydration radius, and migration rate in an electric field. The deviation factor D, obtained by comparing the conductivity of the reference component with that of the analyte, directly reflects the degree of difference in their mobility characteristics. When the D value is close to 1, it indicates that the mobility characteristics of the analyte and the reference component are similar; when the D value deviates significantly from 1, it indicates that there is a fundamental difference in the ion types of the two components.

[0012] Furthermore, sodium chloride was selected as the reference component, and the influent concentration was set within the range of 50 ppm to 1000 ppm. Within the influent concentration range of 200–600 ppm, the deviation factor D remained consistently between 1.8 and 2.8, verifying the closed-loop stability of this identification criterion. Using an influent concentration of 400 ppm as a typical calibration point, the effluent conductivity EC of sodium chloride was [data missing]. NaCl The conductivity (EC) of magnesium sulfate in effluent is 18~24 μS / cm. MgSO4 The hardness intensity ranges from 7 to 12 μS / cm, and the deviation factor D between the two ranges from 1.8 to 2.8. This deviation range constitutes a characteristic fingerprint for identifying monovalent or polyvalent ions, thereby distinguishing between water samples dominated by monovalent ions and water samples dominated by polyvalent hardness ions.

[0013] Sodium chloride was chosen as the reference component because sodium ions are typical monovalent strong electrolyte ions with high and stable molar conductivity in aqueous solutions. Its conductivity response exhibits good linearity and high repeatability, making it suitable as a comparison benchmark. Experimental data show that at an influent concentration of 400 ppm, the conductivity of sodium chloride effluent is more than twice that of magnesium sulfate effluent. This significant deviation allows the system to reliably distinguish between water samples dominated by monovalent ions and those dominated by polyvalent hardness ions.

[0014] Furthermore, the nonlinear jump detection module captures the critical jump point in the following manner: it performs a concentration gradient increase test on the water sample to be tested, and monitors the dynamic response curves and slopes of the effluent TDS and EC values ​​under each concentration gradient in real time. When the slope of the effluent TDS or EC value changes abruptly relative to the previous concentration gradient and the magnitude of the change exceeds the preset jump threshold, the corresponding influent concentration is calibrated as the critical jump point concentration.

[0015] The physical meaning of the critical breakpoint lies in the qualitative change that occurs in the interception mechanism of the reverse osmosis membrane under the specified concentration load conditions. Before the breakpoint, the membrane's interception performance changes relatively steadily with increasing feed water concentration; however, at the breakpoint, the membrane's interception performance decreases abruptly, and the solute content in the effluent rises sharply. The location of this breakpoint is closely related to the actual operating condition of the membrane. The feed water concentration corresponding to the breakpoint of a new membrane is relatively high, while as the membrane ages, the feed water concentration corresponding to the breakpoint will gradually decrease. Therefore, the breakpoint can serve as an objective indicator of the degree of membrane performance degradation.

[0016] Furthermore, the influent concentration range for the concentration gradient escalation test is 50 ppm to 1000 ppm, and the gradient step is 50 to 200 ppm. For a sodium chloride single-component water sample, the critical jump point concentration is located in the range of 850 to 950 ppm. At this jump point, the effluent EC value crosses the key physical characteristic threshold of 32 μS / cm, jumping from 25 to 32 μS / cm in the previous gradient to 34 to 40 μS / cm, exhibiting obvious nonlinear breakdown characteristics. For a composite water sample prepared by mixing sodium chloride and magnesium sulfate at a mass ratio of 1:1, the critical jump point concentration is located in the range of 700 to 900 ppm. At this jump point, the effluent TDS value jumps from the stable trend of the previous gradient to 9 to 14 ppm.

[0017] The experimental data revealed significant differences in the transition characteristics between single-component and composite water samples. Due to the synergistic effect of multiple ions, the influent concentration at the transition point in composite water samples was lower than that in single-component water samples. This indicates that the presence of multivalent ions accelerates the nonlinear transition process of membrane interception performance.

[0018] Furthermore, the signal acquisition frequency of the nonlinear jump detection module is not less than 1Hz, which is used to capture the instantaneous TDS pulse signal that appears near the critical jump point of the reverse osmosis membrane; the amplitude fluctuation range of the instantaneous TDS pulse signal is such that the difference between the peak value and the steady-state value reaches more than 50% of the steady-state value.

[0019] Near the critical transition point, the reverse osmosis membrane exhibits significant instability in its interception behavior; the effluent TDS does not rise smoothly but instead shows large-amplitude pulse-like fluctuations. Using a sampling frequency of at least 1 Hz, the waveform characteristics of the pulse signal can be fully recorded, providing sufficient time-domain data for accurately determining the transition point.

[0020] Furthermore, the algorithm parsing layer also includes a dynamic thermal transient response module. This module applies a controllable thermal pulse excitation to the water sample to raise its temperature by 8-15°C, and collects the conductivity values ​​EC before and after the thermal excitation. 常温 and EC 高温 Calculate the conductivity temperature rise rate ΔEC%; ΔEC% is defined as the percentage difference between high-temperature conductivity and room-temperature conductivity, relative to the room-temperature conductivity, calculated using the following formula: Based on the difference in the rate of temperature increase of different ionic components, water samples with similar conductivity responses but different ionic compositions under static monitoring conditions can be distinguished.

[0021] The introduction of the dynamic thermal transient response module solves the problem that the conductivity values ​​of different ionic components may be similar and difficult to distinguish under static measurement conditions. Different ions have different sensitivities to temperature changes in conductivity. By applying a temperature perturbation and comparing the ratio of conductivity changes before and after the perturbation, a characteristic quantity directly related to the ionic composition can be obtained, thus adding an independent dimension to the determination of ionic components.

[0022] In specific implementation, the controllable thermal pulse excitation can adopt microwave thermal excitation mode, high frequency induction heating mode or infrared radiation heating mode. By applying controlled energy to the water sample, rapid and uniform temperature rise control is achieved, ensuring the repeatability of the thermal excitation process.

[0023] Furthermore, there is an identifiable difference in the temperature rise rate between high-salinity water samples and high-hardness water samples, with the temperature rise rate of high-salinity water samples being 28%~32% and the temperature rise rate of high-hardness water samples being 35%~39%, and the difference between the two being no less than 2%.

[0024] The data shows that the conductivity increase rate of high-hardness water samples is higher than that of high-salinity water samples under the same temperature change conditions, and the difference between the two is sufficient for reliable identification.

[0025] Furthermore, the performance application layer uses at least two of the deviation factor D, critical jump point concentration, effluent pH value, and conductivity temperature rise increase ratio as feature dimensions to construct a multi-dimensional performance feature fingerprint radar map; the system also includes an adaptive correction module, which is used to automatically correct the baseline parameters of deviation factor D and jump threshold based on the baseline TDS value, baseline EC value, and baseline pH value of the local source water when deployed across regions.

[0026] The multi-dimensional feature fingerprint radar chart presents the analysis results from multiple dimensions in a visually intuitive way, allowing users to understand water quality and filter status without needing professional knowledge. The adaptive correction module automatically adjusts the judgment benchmark by collecting basic parameters of local source water, eliminating the impact of regional differences on recognition accuracy.

[0027] This invention also provides a water quality identification method based on the deviation characteristics of multi-component ion mobility, comprising the following steps: S1. Multi-parameter signal acquisition steps: Using a multi-parameter integrated sensor array, the TDS, pH, EC and temperature values ​​of water samples before and after reverse osmosis membrane filtration are acquired in real time. S2. Ionic component characteristic identification steps: Under the same influent concentration conditions, obtain the effluent conductivity EC values ​​of different ionic components, calculate the deviation factor D of each component relative to the reference component, and when the deviation factor D exceeds the preset deviation threshold, it is determined that there are ionic components in the water sample that are significantly different from the migration characteristics of the reference component, and the ionic component type of the water sample is determined accordingly. S3. Nonlinear jump point capture step: Perform concentration gradient increase test on the water sample to be tested, monitor the changing trend of TDS value and EC value of the effluent under each gradient, and when the slope of change changes abruptly and the magnitude of the change exceeds the preset jump threshold, the concentration is calibrated as the critical jump point of the reverse osmosis membrane, which serves as the criterion for membrane performance degradation and filter life warning. S4. Performance Feature Mapping Step: The ionic component feature parameters obtained in step S2 and the jump point criteria obtained in step S3 are mapped into a multi-dimensional feature fingerprint and performance evaluation report, realizing the quantitative and visual output of water quality characteristics and filter cartridge status.

[0028] The method described above achieves a complete processing flow from raw signal acquisition to final performance evaluation through the sequential execution of four steps. The logical progression and coherent data transmission between each step form a closed-loop water quality identification process.

[0029] Further, in step S2, sodium chloride is selected as the reference component. Sodium chloride water sample and magnesium sulfate water sample are prepared in the range of 200~600ppm in the influent concentration and filtered through the same reverse osmosis membrane to obtain the conductivity of their respective effluents and calculate the deviation factor D. When D is in the range of 1.8~2.8, it is determined that there is a significant difference in the migration rate between monovalent ions and polyvalent ions.

[0030] Further, in step S3, the operating conditions for the concentration gradient increase test are as follows: the influent concentration is gradually increased from 50 ppm to 1000 ppm in increments of 50 to 200 ppm, the stable acquisition time for each concentration gradient is not less than 30 seconds, and the signal acquisition frequency is not less than 1 Hz; the preset jump threshold is when the increase in the slope of the effluent EC change between adjacent gradients exceeds 20% of the slope of the previous gradient, or the absolute increase in the effluent TDS value exceeds 30% of the increment of the previous gradient.

[0031] The above operating conditions balance testing accuracy and efficiency. A stable acquisition time of at least 30 seconds for each concentration gradient ensures that the sensor output value reaches a steady state before recording. A 20% slope increase threshold and a 30% absolute increment threshold effectively distinguish between normal gradual trends and abrupt changes.

[0032] Furthermore, a dynamic thermal transient response step is included between step S2 and step S3, specifically: a controllable thermal pulse excitation is applied to the water sample to raise the water sample temperature by 8~15℃, and the conductivity values ​​before and after the thermal excitation are collected respectively, and the conductivity temperature rise increase ratio is calculated; the temperature rise increase ratio is used as an auxiliary dimension to be incorporated into the comprehensive determination of ion component type, in order to distinguish water samples with similar static conductivity responses but different ion compositions.

[0033] Further, in step S4, a multidimensional performance characteristic fingerprint radar map is constructed using at least two of the following coordinate dimensions: deviation factor D, critical jump point concentration, effluent pH value, and conductivity temperature rise increase ratio. Simultaneously, based on the historical change trend of the critical jump point concentration, a quantitative prediction model for the remaining life of the filter element is established, and the current jump point concentration is compared with the initial calibrated jump point concentration to output the evaluation result of the remaining life of the filter element.

[0034] Furthermore, the method also includes an adaptive correction step: when the system is deployed to a new water source environment, the basic TDS value, basic EC value and basic pH value of the local source water are first collected as background reference parameters. The background reference parameters are then deducted or normalized from the subsequent calculation of deviation factor D and determination of jump points to eliminate the interference of different regional source water background differences on the identification results.

[0035] The adaptive correction step ensures the universality of the method in different regional water source environments. By automatically collecting local benchmark parameters in the early stage of deployment and correcting subsequent calculations accordingly, the engineering goal of one-time deployment and adaptive operation is achieved.

[0036] Compared with the prior art, this application has the following beneficial effects: Firstly, the water quality identification system of this invention establishes a component identification criterion based on the difference in ion mobility by introducing a deviation factor D. This breaks through the limitations of traditional technologies that rely solely on a single conductivity or TDS value for water quality assessment, and achieves a technological leap from measuring total quantity to grouping. It can effectively distinguish between water samples dominated by monovalent ions and water samples dominated by polyvalent hardness ions, providing a quantitative means for the refined analysis of water quality characteristics.

[0037] Secondly, the nonlinear jump detection module of this invention captures the critical jump point where the reverse osmosis membrane interception performance changes abruptly through concentration gradient increasing test, and uses this as a filter life warning criterion. Compared with the traditional linear prediction method based on running time or cumulative water production, it can more realistically reflect the actual degradation state of membrane performance, and the timing of the warning signal is more accurate and advanced.

[0038] Third, the dynamic thermal transient response module of the present invention adds an auxiliary identification dimension independent of static conductivity by applying a controllable temperature perturbation and comparing the ratio of conductivity changes before and after the perturbation. This solves the problem that different ionic components have similar conductivity values ​​under static measurement conditions and are difficult to distinguish, thus improving the reliability and robustness of ionic component identification.

[0039] Fourth, the system of the present invention is equipped with an adaptive correction module, which automatically collects the basic parameters of the local water source when deployed across regions and corrects the judgment benchmark accordingly. It can adapt to different water source environments without manual recalibration, reducing the deployment and maintenance costs of the system and improving its versatility and convenience in practical engineering applications.

[0040] Fifth, the core logic of the filter cartridge life warning in this invention lies in the nonlinear breakdown point concentration drift evolution. Unlike the industry's traditional rough estimations based on "cumulative water flow" or "fixed TDS threshold," the nonlinear jump detection module dynamically tracks the downward trend of the critical jump point concentration (e.g., gradually decreasing from an initially calibrated 900 ppm to 700 ppm or lower), establishing a quantitative prediction model based on the degradation of membrane fiber microscopic interception performance. This logic of predicting lifespan by capturing the displacement of the "mechanical change point" can eliminate prediction bias caused by differences in water quality across regions, achieving truly individualized and precise filter cartridge management. Attached Figure Description

[0041] Figure 1 A comparison chart of the deviation of the slope between the ion concentration and conductivity of a single component; Figure 2 A fingerprint trend of decreasing pH in the effluent of a composite water sample as concentration increases; Figure 3 This is a nonlinear jump characteristic diagram of TDS synergistic permeation under high load conditions; Figure 4 This is a schematic diagram of the three-tier architecture logic flow of the system. Detailed Implementation

[0042] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and experimental data.

[0043] like Figure 4As shown, the water quality identification system of this invention consists of three parts: a hardware sensing layer, an algorithm analysis layer, and a performance application layer. The hardware sensing layer is installed on the outlet pipe of the reverse osmosis membrane water purification equipment and is equipped with a multi-parameter integrated sensor array. This sensor array includes at least one conductivity sensor, one potential sensor, and one temperature sensor. The conductivity sensor is used to measure the EC value and convert it to TDS value; the potential sensor is used to measure the pH value; and the temperature sensor is used to record the water sample temperature. The output signals of each sensor are transmitted to the algorithm analysis layer for processing after analog-to-digital conversion. It should be noted that the system of this invention acquires water quality parameters at the inlet and outlet ends through the sensor array. In a specific embodiment, the inlet parameters can be acquired in real time by sensors built into the pipeline, or by calibrated external measuring devices (such as high-precision conductivity testers) and input into the processor to ensure the accuracy of the algorithm calibration. During the algorithm calibration and verification phase, the influent water sample uses a manually prepared standard solution of known concentration (such as sodium chloride, magnesium sulfate, and their compound solutions). The influent concentration parameters are obtained through precise external weighing and calibration, rather than relying on readings from the equipment's built-in sensors. This provides traceable benchmark data for calculating the deviation factor D and locating the critical jump point. The algorithm parsing layer can be implemented using an embedded microprocessor, running the algorithm programs for the binary matrix model module, the nonlinear jump detection module, and the dynamic thermal transient response module. The performance application layer can connect to a mobile terminal via a display screen or wireless communication module, presenting the analysis results to the user in the form of a multi-dimensional feature fingerprint radar chart and a performance evaluation report.

[0044] The working process of the binary matrix model module is as follows. Sodium chloride is selected as the reference component, and a series of sodium chloride standard solutions and magnesium sulfate standard solutions of known concentrations are prepared, with a concentration range of 50~1000ppm. The standard solutions of each concentration are sequentially passed into the same reverse osmosis membrane water purification equipment, and the EC value and pH value collected by the sensor array at the outlet are recorded. Taking sodium chloride as an example, the outlet EC value and outlet TDS value at each concentration gradient are shown in Table 1.

[0045] Table 1. Sodium chloride single-component concentration gradient test data Table 2. Concentration gradient test data of magnesium sulfate single components like Figure 1As shown, based on the concentration-conductivity curves of the two sets of data above, the effluent EC of sodium chloride exhibits a higher slope with increasing influent concentration, while the effluent EC slope of magnesium sulfate is significantly lower than that of sodium chloride. At a calibration concentration of 400 ppm, the effluent EC of sodium chloride is 21.0 μS / cm, and the effluent EC of magnesium sulfate is 9.0 μS / cm. The deviation factor D = 21.0 / 9.0 ≈ 2.33, which falls within the judgment range of 1.8 to 2.8. Therefore, it is determined that the water sample contains a multivalent ion component with significantly different migration characteristics from sodium chloride.

[0046] Table 3. Concentration gradient test data of 1:1 sodium chloride and magnesium sulfate composite solution like Figure 3 As shown in the figure and combined with the data in Table 3, when the influent concentration increased from 600 ppm to 800 ppm, the effluent TDS jumped from 8 ppm to 11 ppm, the effluent EC jumped from 36.0 μS / cm to 48.0 μS / cm, and the effluent pH decreased from 6.65 to 6.50. The effluent TDS and EC exhibited significant nonlinear jumps relative to the previous gradient. The system calibrated 800 ppm as the critical jump point concentration for this composite water sample under the current membrane conditions. Figure 2 As shown, the pH of the effluent decreases in a stepwise manner with the increase of the influent concentration. Near the jump point, the pH decrease is significantly greater. This pH fingerprint feature can be used as an auxiliary reference for determining the jump point.

[0047] For the sodium chloride single-component water sample, as shown in Table 4, the high-concentration gradient test data indicates that at an influent concentration of 900 ppm, the effluent EC jumps significantly from approximately 29 μS / cm at the previous gradient (850 ppm) to 36 μS / cm, exhibiting a significant positive abrupt change in slope. Furthermore, the instantaneous TDS pulse signal was captured using 1 Hz high-frequency sampling, with a peak value reaching 32 ppm, while the steady-state value is approximately 11–13 ppm. The difference between the peak value and the steady-state value exceeds 50% of the steady-state value. The system calibrates 900 ppm as the critical jump point for the sodium chloride single-component water sample. At gradients of 950 ppm and 1000 ppm, the effluent EC is 31 μS / cm and 36 μS / cm, respectively, exhibiting characteristics of masking decline and resaturation after polarization saturation, further confirming the authenticity of the jump point.

[0048] Table 4. High concentration gradient test data for single components of sodium chloride It is noteworthy that at a concentration of 900 ppm, the EC curve of the effluent from the sodium chloride single-component water sample exhibits a significant upward slope abrupt change (jump characteristic), while the EC curve of the effluent from the composite water sample shows a passive and stable trend at this concentration (masking characteristic). The two curves show a clear difference in opening angle in the high-concentration region. This masking effect is due to the suppression of the jump signal of monovalent ions by multivalent ions under high load conditions, and the system needs to identify and distinguish this difference.

[0049] Table 5 Comparison data of ultimate load composite components E1 / E2 In the table above, E1 is prepared by diluting 180.00g NaCl + 80.00g MgSO4 + 100.00g NaHCO3 into 2000g distilled water; E2 is prepared by diluting 80.00g NaCl + 120.00g MgSO4 + 100.00g NaHCO3 into 2000g distilled water. The effluent EC of E1 is 29μS / cm, while that of E2 is only 14μS / cm, with a deviation factor D = 29 / 14 ≈ 2.07.

[0050] In the experiments described above, both formulations E1 and E2 contained sodium bicarbonate (NaHCO3). Experimental verification showed that the combination of sodium chloride (NaCl), magnesium sulfate (MgSO4), and sodium bicarbonate (NaHCO3) constituted a closed loop for comprehensive physical characterization, covering the monovalent ion response mode, the multivalent ion response mode, and the alkaline buffer ion response mode, respectively. The conductivity response characteristics of other salts could be logically derived based on this three-component model using a deviation factor D, without the need for exhaustive testing of all salts individually.

[0051] Table 6 Dynamic thermal transient response test data As shown in Table 6, after applying a thermal pulse excitation of approximately 10-12°C to the E1 and E2 water samples, the conductivity temperature rise increase rate of E1 (high salinity type) was 31.0%, while that of E2 (high hardness type) was 35.7%, a difference of 4.7%. Even though the static conductivity values ​​of the two water samples tend to be similar due to concentration adjustments, they can still be distinguished based on the difference in the temperature rise increase rate through thermal transient response testing.

[0052] Table 7. Cross-regional source water adaptive validation data Water source A and water source B represent two typical municipal tap water backgrounds with significant gradient differences, used to verify the algorithm's recognition accuracy and adaptive compensation capability under different initial ion loads.

[0053] As shown in Table 7, the TDS values ​​of the source water from water source A and water source B are 50 ppm and 224 ppm, respectively, showing a significant difference. When the system is deployed to various locations, the adaptive correction module first collects the baseline TDS, baseline EC, and baseline pH values ​​of the local source water as background reference parameters, subtracting the background contribution from subsequent measurement data, so that the calculation of deviation factor D and jump threshold are based on the normalized net signal. The experimental results show that after adaptive correction, the component identification results and jump point determination results of the system in both locations are consistent with the standard laboratory calibration results, verifying the effectiveness of the adaptive correction mechanism. In the field test of water source B, the measured TDS value of the source water was 224 ppm, and the measured EC value of the effluent after reverse osmosis membrane filtration was 16.5 μS / cm.

[0054] In terms of performance application layer implementation, the system maps data from four dimensions—deviation factor D, influent concentration corresponding to the critical jump point, effluent pH value, and conductivity temperature rise rate—to a multi-dimensional feature fingerprint radar chart. Each dimension corresponds to a radial axis of the radar chart. Different water quality characteristics are presented with different areas and shapes on the radar chart, allowing users to intuitively compare water quality changes between tests. Simultaneously, the system uses the initial jump point concentration as a baseline and records the trend of jump point concentration changes in subsequent periodic tests. When the jump point concentration drops to a preset warning value, a filter replacement reminder is issued, along with an assessment of the remaining lifespan.

[0055] The core logic of the filter cartridge life prediction system in this invention lies in the nonlinear concentration drift evolution at the breakdown point. Unlike the industry's traditional rough estimates based on "cumulative water flow" or "fixed TDS threshold," the algorithm of the nonlinear jump detection module dynamically tracks the downward trend of the concentration at the critical jump point (e.g., gradually decreasing from an initially calibrated 900 ppm to 700 ppm or lower), establishing a quantitative prediction model based on the degradation of the membrane fiber's microscopic interception performance. This logic, which predicts lifespan by capturing the displacement of the "mechanical change point," eliminates prediction biases caused by differences in water quality across regions, achieving truly individualized and precise filter cartridge management.

[0056] The above description is merely a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. All equivalent modifications and improvements made based on the technical solutions of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A water quality identification system based on the deviation characteristics of multi-component ion mobility, characterized in that, The system adopts a three-layer architecture consisting of a perception layer, a parsing layer, and an application layer, including: The hardware sensing layer is equipped with a multi-parameter integrated sensor array for real-time acquisition of the total dissolved solids (TDS), pH, electrical conductivity (EC), and temperature (T) signals of the water sample being tested. The algorithm analysis layer, connected to the hardware sensing layer, receives the raw signals output by the multi-parameter integrated sensor array. The algorithm analysis layer includes a binary matrix model module and a nonlinear jump detection module. The binary matrix model module utilizes the nonlinear correlation between the response characteristics of conductivity sensors to ion mobility and the response characteristics of potential sensors to ion activity to construct a potential-conductivity binary matrix model. By calculating the conductivity response deviation of different ion components under the same concentration conditions, it identifies the characteristics of ion components in the tested water sample. The nonlinear jump detection module monitors the nonlinear abrupt changes in conductivity and TDS of the composite ion system during concentration gradient changes, capturing the critical jump point where the reverse osmosis membrane transitions from a linear interception state to a nonlinear synergistic permeation state, serving as a filter cartridge life warning criterion. The performance application layer is connected to the algorithm analysis layer and is used to convert the ionic component characteristic parameters and jump point criteria output by the algorithm analysis layer into multi-dimensional feature fingerprints and performance evaluation reports, so as to realize the quantitative expression of water quality characteristics and the visual output of filter cartridge status.

2. The system according to claim 1, characterized in that, The binary matrix model module identifies components by calculating the conductivity response deviation of ionic components: under the same influent concentration conditions, the conductivity values ​​of the effluent after reverse osmosis membrane filtration of different ionic components are obtained respectively. The effluent conductivity EC of the component to be tested x Effluent conductivity EC compared to the reference component ref The deviation factor D is obtained by comparison; whereby the deviation factor D is calculated by the following formula: In the formula, The effluent conductivity of the reference component NaCl, The effluent conductivity is the analyte concentration. The logical basis for this determination is that sodium chloride, as a monovalent strong electrolyte, has a significantly higher ion mobility than polyvalent salts under the same conditions. When the analyte is a polyvalent salt, because… Less than The calculated deviation factor D > 1; when the degree of deviation of D from 1 exceeds the preset deviation threshold, the system automatically determines that there are ionic components in the current water sample that have a significant difference in migration rate from the reference component.

3. The system according to claim 2, characterized in that, The reference component is sodium chloride, and the influent concentration is set in the range of 50 ppm to 1000 ppm; within the range of 200 to 600 ppm influent concentration, the deviation factor D is in the range of 1.8 to 2.

8. Using an influent concentration of 400 ppm as a typical calibration point, the effluent conductivity EC of sodium chloride... NaCl The conductivity (EC) of magnesium sulfate in effluent is 18~24 μS / cm. MgSO4 The hardness intensity ranges from 7 to 12 μS / cm, and the deviation factor D between the two ranges from 1.8 to 2.

8. This deviation range constitutes a characteristic fingerprint for identifying monovalent or polyvalent ions, thereby distinguishing between water samples dominated by monovalent ions and water samples dominated by polyvalent hardness ions.

4. The system according to claim 1, characterized in that, The nonlinear jump detection module captures the critical jump point in the following way: it performs a concentration gradient increase test on the water sample, monitors the dynamic response curve and change slope of the effluent TDS value and effluent EC value under each concentration gradient in real time, and when the change slope of the effluent TDS value or effluent EC value changes abruptly relative to the previous concentration gradient and the change amplitude exceeds the preset jump threshold, the corresponding influent concentration is calibrated as the critical jump point concentration.

5. The system according to claim 4, characterized in that, The influent concentration range for the concentration gradient escalation test is 50 ppm to 1000 ppm, with a gradient step of 50 to 200 ppm. For a sodium chloride single-component water sample, the critical jump point concentration is in the range of 850 to 950 ppm. For a composite water sample prepared by mixing sodium chloride and magnesium sulfate at a mass ratio of 1:1, the critical jump point concentration is in the range of 700 to 900 ppm, at which point the effluent TDS value jumps from the stable trend of the previous gradient to 9 to 14 ppm. The signal acquisition frequency of the nonlinear jump detection module is not less than 1Hz, and it is used to capture the instantaneous TDS pulse signal that appears near the critical jump point of the reverse osmosis membrane; the amplitude fluctuation range of the instantaneous TDS pulse signal is such that the difference between the peak value and the steady-state value reaches more than 50% of the steady-state value.

6. The system according to claim 1, characterized in that, The algorithm parsing layer also includes a dynamic thermal transient response module. This module applies a controllable thermal pulse excitation to the water sample to raise its temperature by 8-15°C, and collects the conductivity values ​​EC before and after the thermal excitation. 常温 and EC 高温 Calculate the rate of increase in electrical conductivity temperature rise ΔEC% ΔEC% is defined as the percentage of the difference between high-temperature conductivity and room-temperature conductivity to the room-temperature conductivity, and is calculated using the following formula: Based on the difference in the rate of increase of temperature rise of different ionic components, water samples with similar conductivity responses but different ionic compositions under static monitoring conditions can be distinguished. There is an identifiable difference in the rate of temperature increase between high-salinity water samples and high-hardness water samples. The rate of temperature increase for high-salinity water samples is 28% to 32%, while that for high-hardness water samples is 35% to 39%, with a difference of not less than 2%.

7. The system according to claim 1, characterized in that, The performance application layer uses at least two of the deviation factor D, critical jump point concentration, effluent pH value, and conductivity temperature rise increase ratio as feature dimensions to construct a multi-dimensional performance feature fingerprint radar map; the system also includes an adaptive correction module, which is used to automatically correct the baseline parameters of deviation factor D and jump threshold based on the baseline TDS value, baseline EC value, and baseline pH value of local source water when deployed across regions.

8. A water quality identification method based on the deviation characteristics of multi-component ion mobility, characterized in that, Includes the following steps: S1. Using a multi-parameter integrated sensor array, the TDS, pH, EC and temperature values ​​of water samples before and after reverse osmosis membrane filtration are collected in real time. S2. Under the same influent concentration conditions, obtain the effluent conductivity EC values ​​of different ionic components, calculate the deviation factor D of each component relative to the reference component, and when the deviation factor D exceeds the preset deviation threshold, it is determined that there are ionic components in the water sample that are significantly different from the mobility characteristics of the reference component, and the ionic component type of the water sample is determined accordingly. S3. Conduct a concentration gradient increase test on the water sample to be tested, and monitor the changing trends of the TDS and EC values ​​of the effluent under each gradient. When the slope of change changes abruptly and the magnitude of the change exceeds the preset jump threshold, the concentration is calibrated as the critical jump point of the reverse osmosis membrane, which serves as a criterion for membrane performance degradation and filter life warning. S4. Map the ionic component characteristic parameters obtained in step S2 and the jump point criterion obtained in step S3 into a multi-dimensional feature fingerprint map and performance evaluation report to achieve quantitative and visual output of water quality characteristics and filter cartridge status.

9. The method according to claim 8, characterized in that, In step S2, sodium chloride is selected as the reference component. Sodium chloride water sample and magnesium sulfate water sample are prepared in the range of 200~600ppm of influent concentration and filtered through the same reverse osmosis membrane to obtain the conductivity of their respective effluents and calculate the deviation factor D. When D is in the range of 1.8 to 2.8, it is determined that there is a significant difference in the mobility between monovalent and multivalent ions; In step S3, the operating conditions for the concentration gradient escalation test are as follows: the influent concentration is gradually increased from 50 ppm to 1000 ppm in increments of 50-200 ppm, with a stable acquisition time of no less than 30 seconds for each concentration gradient and a signal acquisition frequency of no less than 1 Hz; the preset jump threshold is when the increase in the slope of the effluent EC change between adjacent gradients exceeds 20% of the slope of the previous gradient, or when the absolute increase in the effluent TDS value exceeds 30% of the increment of the previous gradient; Between step S2 and step S3, there is also a dynamic thermal transient response step, which is as follows: a controllable thermal pulse excitation is applied to the water sample to raise the water sample temperature by 8~15℃, and the conductivity values ​​before and after the thermal excitation are collected respectively, and the conductivity temperature rise increase ratio is calculated; the temperature rise increase ratio is used as an auxiliary dimension to be incorporated into the comprehensive determination of ion component type, which is used to distinguish water samples with similar static conductivity responses but different ion compositions. In step S4, a multidimensional performance characteristic fingerprint radar map is constructed using at least two of the following coordinate dimensions: deviation factor D, critical jump point concentration, effluent pH value, and conductivity temperature rise increase ratio. Simultaneously, based on the historical change trend of the critical jump point concentration, a quantitative prediction model for the remaining lifespan of the filter element is established. The current jump point concentration is compared with the initial calibrated jump point concentration, and the evaluation result of the remaining lifespan of the filter element is output.

10. The method according to claim 9, characterized in that, The method also includes an adaptive correction step: when the system is deployed to a new water source environment, the basic TDS value, basic EC value and basic pH value of the local source water are collected as background reference parameters. The background reference parameters are deducted or normalized from the subsequent deviation factor D calculation and jump point determination to eliminate the interference of the difference in the background of source water in different regions on the identification results.

Images