A post-disaster drinking water quality evaluation and treatment optimization method and system

Abstract

The application discloses a kind of post-disaster drinking water quality evaluation management optimization method and system, belong to water quality evaluation management optimization technical field.The method includes the following steps: S1, water quality detection and comprehensive equivalent load calculation;S2, membrane equipment operating condition and flux reduction factor calculation;S3, equipment operation control parameter extraction and reverse correction;S4, multi-water source mixed reaction interference matrix calculation;S5, water source deployment optimization solution and instruction issue.The application directly converts water quality evaluation results to operating control parameters, dynamically adjusts the water production capacity of the treatment equipment with cumulative interception load, and includes ion precipitation risk as a constraint condition in the generation stage of the water source mixing scheme, improving the reliability of the water supply system in the whole-link operation, solving the problem that the data cannot be connected between the evaluation, treatment and configuration in the prior art, resulting in the output management scheme cannot be reliably operated.

Description

Technical Field

[0001] This invention relates to the field of water quality assessment, treatment and optimization technology, and in particular to a method and system for post-disaster drinking water quality assessment, treatment and optimization. Background Technology

[0002] Post-disaster drinking water supply involves three stages: water source quality assessment, emergency treatment, and optimized allocation of water resources. The water quality assessment stage involves sampling and testing multiple temporary water intake points or damaged water sources to obtain data on the concentration of various pollutants. The emergency treatment stage uses processes such as electrolytic flocculation and membrane filtration to purify raw water exceeding standards into compliant water. The water resource allocation stage involves mixing multiple water sources in a specific ratio before supplying them uniformly, ensuring that the overall quality of the mixed water meets national drinking water hygiene standards.

[0003] In existing related technologies, the water quality assessment process uses machine learning algorithms to output water quality category labels and concentration levels; the treatment equipment operates according to rated parameters; and the configuration process uses iterative algorithms to mix substandard water sources with compliant water sources in a certain proportion so that the components after mixing meet the standard limits.

[0004] However, the discrete semantic labels of the evaluation output in the above scheme cannot drive the continuous physical parameter setting of the treatment equipment. During the operation of equipment such as reverse osmosis membranes, the permeate flux gradually decreases due to the interception of pollutants, and the configuration process treats its supply as a fixed upper limit. The calculation of various heavy metal loads is prone to distortion. Each water source may meet the standards when tested individually, but precipitation may occur after mixing, leading to actual water quality deterioration. In addition, the system lacks network hydraulic constraints and abnormal fault tolerance mechanisms.

[0005] In summary, current technologies suffer from a lack of data flow between the assessment, processing, and configuration stages, leading to unreliable operation of the output governance solution. To address these issues, this invention proposes a solution. Summary of the Invention

[0006] This application provides a method and system for post-disaster drinking water quality assessment and treatment optimization, solving the problems in existing technologies where data cannot be integrated across the assessment, treatment, and configuration stages, and where the lack of quantitative assessment of heavy metal and precipitation risks leads to unreliable operation of the output treatment plan. It achieves direct conversion of water quality assessment results into operational control parameters, dynamically adjusts the water production capacity of treatment equipment based on cumulative interception load, and incorporates ion precipitation risk and network hydraulic constraints into a unified-dimensional planning solution during the water source mixing scheme generation stage, improving the reliability and robustness of the water supply system throughout its entire lifecycle.

[0007] This application provides a method for post-disaster drinking water quality assessment, treatment, and optimization. This method is applied to a post-disaster drinking water quality assessment, treatment, and optimization system, and includes: After acquiring water quality test data from multiple drinking water sources and performing data anomaly correction and fault-tolerant cleaning, the water quality test data is converted into an equivalent load sequence of the target pollutants through the chemical oxygen demand equivalent rule corrected for heavy metal toxicity. The system acquires the operating condition data of the water treatment equipment, calculates the membrane resistance that distinguishes between reversible and irreversible fouling based on the equivalent load sequence and operating condition data, further calculates the dynamic treatment flux reduction factor of the water treatment equipment, and executes the backwashing procedure when the dynamic treatment flux reduction factor reaches the preset backwashing threshold. Based on the equivalent load sequence and dynamic processing flux reduction factor, multi-dimensional state vector fusion is performed to extract the operation control parameters of the water treatment equipment. The operation control parameters are sent to the water treatment equipment, and the reagent dosing frequency is corrected in reverse according to the mixing ratio. Interactive sensitive ion concentration data are extracted from water quality testing data. Combined with the hazard level weights of different precipitates, a multi-source mixed reaction interference matrix is ​​constructed based on the interactive sensitive ion concentration data. Based on the dynamic processing flux reduction coefficient, the multi-source mixing reaction interference matrix, the pipeline flow and pressure constraints, and the target water demand, a unified water resource allocation optimization solution is executed to generate data management and scheduling instructions containing the mixing ratio of each drinking water source, and an emergency backup strategy is triggered in emergency situations.

[0008] Furthermore, in acquiring water quality testing data from multiple drinking water sources and converting the water quality testing data into equivalent load sequences of the target pollutants: Analyze water quality testing data to separate the concentration sequences of individual pollutants; Retrieve the preset pollutant equivalent conversion and toxicity correction rule table, and assign the chemical oxygen demand equivalent coefficient and heavy metal toxicity correction coefficient to each independent pollutant according to the rules; A comprehensive equivalent load sequence is generated based on the concentration numerical sequence, the chemical oxygen demand equivalent coefficient, and the heavy metal toxicity correction coefficient.

[0009] Furthermore, in generating the equivalent load sequence: Three-standard-deviation filtering was applied to the water quality test data to remove outliers from the sensor; if missing values ​​were found, the moving median was used to fill them in. Based on the sampling timestamps of water quality testing data, the concentration sequences of individual pollutants are time-series aligned; The sliding window step size is adaptively adjusted according to the severity of water quality fluctuations. The equivalent load sequence within the time sliding window is output by summing the products of the concentration numerical sequence after aggregation and alignment within the time sliding window and the corresponding chemical oxygen demand equivalent coefficient and heavy metal toxicity correction coefficient of the pollutant.

[0010] Furthermore, in acquiring the operating condition data of the water treatment equipment, and calculating the dynamic treatment flux reduction factor of the water treatment equipment based on the equivalent load sequence and operating condition data: Extract the current membrane inlet pressure and current inlet flow rate of the water treatment equipment from the operating condition data; Calculate the time integral of the equivalent load sequence over the historical operating cycle to obtain the cumulative retained load. The current actual membrane resistance is decomposed into initial clean membrane resistance, reversible fouling resistance and irreversible fouling resistance. The current inlet pressure, current influent flow rate and cumulative interception load are substituted into the preset membrane fouling attenuation model to calculate the dynamic treatment flux reduction coefficient. When the dynamic processing flux reduction factor is less than or equal to the preset hydraulic backwash threshold, hydraulic backwashing is triggered. When the flux does not recover to the set range after hydraulic backwashing, chemical backwashing is triggered.

[0011] Furthermore, by substituting the current inlet pressure, current influent flow rate, and cumulative retention load into a preset membrane fouling attenuation model, the dynamic treatment flux reduction factor is calculated: Based on the current membrane pressure, current influent flow rate, and fluid viscosity at the current water temperature, the actual membrane resistance is calculated using Darcy's law. Calculate the ratio of the difference between the current actual membrane resistance value and the initial clean membrane resistance value; The dynamic processing throughput reduction factor is output based on the nonlinear fitting relationship between the difference ratio and the cumulative intercepted load.

[0012] Furthermore, in extracting the operating control parameters of the water treatment equipment based on the equivalent load sequence and the dynamic processing flux reduction factor: The equivalent load sequence is fused with the dynamic processing flux reduction factor to generate a multi-dimensional operating condition state vector. Retrieve the reference coordinates that have the smallest Euclidean distance to the current multidimensional operating condition state vector from the preset two-dimensional control mapping table; The operating voltage and reagent dosing frequency values ​​corresponding to the reference coordinates are extracted as operating control parameters. After optimizing the mixing ratio of each drinking water source, the reagent dosing frequency is reversed based on the actual mixing flow rate.

[0013] Furthermore, in extracting interactively sensitive ion concentration data from water quality testing data, and constructing a multi-source mixing reaction interference matrix based on the interactively sensitive ion concentration data: The concentration values ​​of metal cations and acid anions in water quality test data are identified as interactively sensitive ion concentration data. For any two different drinking water sources, calculate the ion product of the cross-combination of metal cations in one source with acid anions in the other source. Retrieve the solubility product constant and hazard level weight of the corresponding precipitate from the pre-set database, compare the ion product value with the solubility product constant, and map the comparison result to the intervention weight value in the multi-source mixing reaction interference matrix: When the ion product value is greater than the solubility product constant, the logarithm of the excess portion is multiplied by the precipitate hazard level weight and converted into a positive intervention weight value. When the ion product value is less than or equal to the solubility product constant, the intervention weight value is set to zero; The intervention weights of all water sources are combined into a symmetric multi-source mixing reaction interference matrix.

[0014] Furthermore, based on the dynamic processing flux reduction factor, the multi-source mixing reaction interference matrix, and the target water demand, water resource allocation optimization is performed to generate data management and scheduling instructions that include the mixing ratio of each drinking water source: The objective function is to maximize the water quality safety margin of all drinking water sources after relative margin normalization. The product of the dynamic processing flux reduction factor and the rated water production of each treatment device is used as the upper limit constraint of the supply of each water source. The maximum allowable flow rate of the pipeline network and the range of hydraulic pressure at the nodes of the pipeline network are used as hydraulic constraints for the pipeline network. The product of the multi-source mixing reaction interference matrix and the quadratic product of the mixing ratio of each drinking water source, multiplied by the dynamic penalty coefficient, is used as a dimensionless mixing penalty constraint and incorporated into the objective function for planning and solving. The output is the mixing ratio of each drinking water source that satisfies the target water demand and maximizes the objective function. In offline or faulty conditions, an emergency backup strategy is triggered to cut off the faulty water source and redistribute the remaining water source proportions.

[0015] Furthermore, in the objective function of maximizing the water quality safety margin of all drinking water sources: Calculate the expected concentration value of the mixture after all drinking water sources are mixed according to the set mixing ratio; Calculate the difference between the national standard limit and the expected concentration of the mixture, and divide by the national standard limit to obtain the dimensionless relative margin; Multiply the dimensionless relative margin by the health hazard weights of each pollutant, which are calculated from the reciprocal of the World Health Organization's health guidance values, and sum the results. Subtract the dimensionless sedimentation penalty term to obtain the objective function characterizing the water quality safety margin.

[0016] This application provides a post-disaster drinking water quality assessment, treatment and optimization system, which is used to implement a post-disaster drinking water quality assessment, treatment and optimization method, including: an equivalent conversion calculation module, a flux reduction calculation module, a control parameter extraction module, an interference matrix construction module, and a water source allocation optimization module; The equivalent conversion calculation module is used to acquire water quality test data from multiple drinking water sources, and after cleaning, convert the water quality test data into an equivalent load sequence of the target pollutants after heavy metal toxicity correction. The flux reduction calculation module is used to acquire the operating condition data of the water treatment equipment, calculate the dynamic treatment flux reduction coefficient of the water treatment equipment based on the equivalent load sequence and operating condition data, and link it to backwash control. The control parameter extraction module is used to extract the operating control parameters of the water treatment equipment after performing multi-dimensional state vector fusion based on the equivalent load sequence and dynamic processing flux reduction coefficient, and to realize reverse correction of water distribution. The interference matrix construction module is used to extract interactive sensitive ion concentration data from water quality testing data and construct a multi-source mixed reaction interference matrix by combining precipitation hazard level weights. The water source allocation optimization module is used to perform water resource allocation optimization based on the dynamic processing flux reduction coefficient, the multi-source mixing reaction interference matrix, the hydraulic constraints of the pipeline network, and the target water demand, and to generate data management and scheduling instructions and emergency backup scheduling instructions containing the mixing ratio of each drinking water source.

[0017] This application provides a method and system for post-disaster drinking water quality assessment, treatment, and optimization. The method involves acquiring water quality testing data from multiple drinking water sources and converting this data into an equivalent load sequence of target pollutants using a chemical oxygen demand (COD) equivalent coefficient. It also involves acquiring operating condition data from water treatment equipment and calculating a dynamic treatment flux reduction coefficient based on the equivalent load sequence and operating condition data. Furthermore, it involves extracting operating control parameters from the water treatment equipment based on the equivalent load sequence and dynamic treatment flux reduction coefficient, and sending these parameters to the equipment. Finally, it involves extracting interactive sensitive ion concentration data from the water quality testing data and constructing a multi-source mixing reaction interference matrix based on this data. Finally, it involves optimizing water resource allocation based on the dynamic treatment flux reduction coefficient, the multi-source mixing reaction interference matrix, and the target water demand, generating data management and scheduling instructions that include the mixing ratio of each drinking water source.

[0018] In this process, the concentrations of various pollutants in the water quality monitoring data are converted into a unified continuous equivalent load sequence after being converted by the chemical oxygen demand equivalent coefficient. This sequence serves as an intermediate numerical variable and directly participates in the extraction of operating control parameters and the calculation of flux reduction coefficients, overcoming the shortcomings of traditional assessment methods that cannot directly drive equipment due to the mismatch between the output form and the treatment side requirements.

[0019] Furthermore, when calculating the dynamic treatment flux reduction factor, the cumulative intercepted load of the equipment is represented by the time integral value of the equivalent load sequence within the historical operating cycle. The current inlet pressure and current influent flow rate are substituted into the membrane fouling attenuation model to calculate the flux reduction factor, which can accurately reflect the actual water production capacity of the equipment. In the configuration optimization stage, the product of this factor and the rated water production is used as the upper limit of the supply of each water source, avoiding the risk of water supply exceeding the actual treatment capacity of the equipment due to the use of constant rated values.

[0020] Furthermore, during the configuration phase, the concentrations of metal cations and acid anions are extracted from water quality testing data. The ion product value of any two water sources is calculated, and the ion product value is compared with the solubility product constant in the standard thermodynamic database. When the ion product is greater than the solubility product, the logarithm of the excess is used as an intervention weight value and is used as a penalty term in the objective function calculation of configuration optimization. This allows the optimization solver to automatically avoid water source combinations with high precipitation risk during iteration. Simultaneously, the flux reduction coefficient is used as a supply upper limit constraint, and the solution output includes the mixing ratio of each water source and equipment control parameters, realizing direct and safe control of water quality and treatment equipment of multiple water sources by the water supply scheduling system. Attached Figure Description

[0021] Figure 1 A flowchart of a post-disaster drinking water quality assessment, treatment, and optimization method provided in this application embodiment; Figure 2 This is a schematic diagram of a post-disaster drinking water quality assessment, treatment and optimization system provided in an embodiment of this application. Detailed Implementation

[0022] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0023] like Figure 1 As shown in the figure, this application provides a method for post-disaster drinking water quality assessment, treatment, and optimization. This method is applied to a post-disaster drinking water quality assessment, treatment, and optimization system, and includes: S1, Water quality testing and comprehensive equivalent load calculation.

[0024] In the step of acquiring water quality testing data from multiple drinking water sources and converting the water quality testing data into equivalent load sequences of target pollutants: First, the system acquires water quality monitoring data from various drinking water sources in the disaster-stricken area through online sensors. To address data anomalies caused by sensor contamination under extreme post-disaster conditions, the system implements data error-tolerant processing on the raw monitoring data: a three-standard-deviation filtering algorithm is used to remove abnormal noise points exceeding the normal statistical range; if short-term sensor data loss occurs, the median within a sliding window is used to fill in the missing values.

[0025] Then, the water quality test data after cleaning was analyzed to separate the concentration sequences of individual pollutants. Among them, the target pollutant refers to an independent pollutant in the water quality test data whose concentration exceeds the limit of the national drinking water hygiene standard (e.g., GB 5749-2022) and whose chemical oxygen demand equivalent coefficient is greater than zero, covering organic pollution indicators and heavy metal indicators.

[0026] To address the issue of heavy metal toxicity distortion caused by directly calculating total load using chemical oxygen demand (COD) in traditional technologies, this solution retrieves a pre-defined pollutant equivalent conversion and toxicity correction rule table. The COD equivalent coefficient in the rule table is determined based on the stoichiometric relationship of the potassium dichromate oxidation method in ISO 15705:2002, characterizing the amount of oxygen required for the complete oxidation of each gram of pollutant. Simultaneously, a heavy metal toxicity correction coefficient determined based on the US Environmental Protection Agency (USEPA) toxicological evaluation method is introduced, assigning a toxicity amplification weight greater than 1.0 to highly biotoxic heavy metals (such as lead, cadmium, and arsenic), thus eliminating load calculation distortion in heavy metal scenarios. The pre-defined pollutant COD equivalent coefficients and toxicity correction coefficients are shown in Table 1. Table 1: Comparison Table of Preset Pollutant Chemical Oxygen Demand Equivalent Coefficient and Toxicity Correction Coefficient

[0027] The specific formula for generating the equivalent load sequence based on the concentration numerical sequence, the chemical oxygen demand equivalent coefficient, and the heavy metal toxicity correction coefficient is as follows: ;in, Indicates a time sliding window The comprehensive equivalent load value within, in units of ; This represents the total number of independent pollutant species detected, and is a positive integer. Indicates a time sliding window Inner Concentration values ​​of each individual pollutant, in units of ; Indicates the first Preset chemical oxygen demand equivalent coefficients for each independent pollutant; Indicates the first Correction factor for heavy metal toxicity of each independent pollutant.

[0028] Based on the sampling timestamps, the concentration sequences of independent pollutants are time-aligned on the time axis. To adapt to the drastic fluctuations in water quality after a disaster, the system establishes an adaptive sliding window adjustment mechanism: when the water quality change rate is less than 10% over three consecutive sampling periods, a default 60-minute sliding window step size is used; when the change rate of water quality indicators (such as permanganate index and ammonia nitrogen concentration) exceeds 30% over three consecutive sampling periods, the system adaptively shortens the sliding window step size to 15 minutes, ensuring that the shortened step size is no less than 1 times the sensor sampling period, thus enabling real-time capture of sudden changes in water quality. The comprehensive equivalent load sequence within the sliding window is output based on the sum of the products after aggregation and alignment.

[0029] The lower limit of the step size of the time sliding window is no less than three times the sampling period of the water quality sensor, and no less than 5 minutes. Sliding window aggregation and alignment refers to unifying the window boundaries to integer scales of natural time when performing sliding window statistics, so that the window divisions of different tasks and different nodes are consistent, making the results more predictable and easier to share and compare.

[0030] S2, Calculation of membrane equipment operating conditions and flux reduction factor.

[0031] In the step of acquiring operating condition data of water treatment equipment and calculating the dynamic treatment flux reduction factor of water treatment equipment based on equivalent load sequence and operating condition data: The current inlet pressure and feed flow rate are obtained from sensors in water treatment equipment (such as reverse osmosis membrane modules or ultrafiltration modules). Simultaneously, the time integral of the equivalent load sequence over historical operating cycles is calculated to obtain the cumulative load cutoff. The time integral is performed using a trapezoidal integral formula for discrete calculation.

[0032] In this embodiment, the membrane module can be any one of a polyamide composite reverse osmosis membrane, a polyvinylidene fluoride (PVDF) ultrafiltration membrane, or a polyethersulfone (PES) ultrafiltration membrane. The effective filtration area A and the initial clean membrane resistance of the membrane module are... The values ​​are calibrated by the manufacturer for specific membrane materials at the time of shipment. The resistance influence coefficient β and load cumulative damage coefficient γ have different values ​​for different membrane materials and need to be calibrated separately. In this embodiment, β=0.12~0.25 and γ=0.0015~0.0040 are representative values ​​for polyamide composite reverse osmosis membranes. The corresponding values ​​for PVDF membranes and PES membranes need to be calibrated separately with reference to their respective membrane material technical specifications and then calculated using the same formula.

[0033] To address the limitation of traditional membrane fouling models that fail to differentiate between fouling types, this approach explicitly decomposes membrane resistance into initial clean membrane resistance, reversible fouling resistance recoverable through hydraulic backwashing, and irreversible fouling resistance recoverable only through chemical cleaning. The formula for calculating membrane resistance and the preset membrane fouling attenuation model used to obtain the dynamic flux reduction coefficient are as follows: ; ;in, This represents the current actual membrane resistance value, in units of... ; This indicates the current pressure value before the membrane, in kPa. This indicates the current inflow rate, in units of... ; This indicates the effective filtration area of ​​the membrane module, in units of... ; This represents the dynamic viscosity coefficient of water at the current water temperature, expressed in units of... ; This represents the dynamic processing flux reduction factor, which is between [value missing]. arrive The dimensionless ratio between them; Indicates the coefficient of influence of membrane material resistance; This represents the initial clean membrane resistance value, in units of... ; Indicates the cumulative damage coefficient under load; Indicates the first The comprehensive equivalent load value within a discrete time step, in units of ; Represents the discrete time step, in units of ; This represents the total number of time steps within the historical operating cycle, and is a positive integer. The range of constant values ​​for the membrane fouling attenuation model and the calibration test conditions are shown in Table 2. Table 2: Range of Constants for Membrane Fouling Attenuation Model and Calibration Test Conditions

[0034] During operation, when the dynamic processing flux reduction factor drops to the hydraulic backwash threshold of 0.65, the system automatically triggers the water treatment equipment to perform a 3-minute hydraulic backwash. If the flux reduction factor does not recover to above 0.80 after hydraulic backwashing and further drops to the chemical backwash threshold of 0.50, the system triggers an online chemical cleaning program to remove the irreversible fouling layer.

[0035] The aforementioned thresholds of 0.65 and 0.50 are determined based on the membrane module's factory standard test procedure. The calibration method is as follows: the initial flux is measured using deionized water; the membrane module is continuously operated at rated pressure using standard prepared water (containing 100 mg / L humic acid and 50 mg / L colloidal silica) suitable for its water source, with flux recorded every 10 minutes. The reduction factor of 0.80 corresponding to the flux dropping to 80% of the initial flux is the target value to be restored after chemical cleaning; the reduction factor of 0.65 corresponding to the flux dropping to 65% of the initial flux is determined as the hydraulic backwash trigger threshold; and the reduction factor of 0.50 corresponding to the flux dropping to 50% of the initial flux is determined as the chemical backwash trigger threshold. This calibration test needs to be performed separately for different membrane materials (polyamide composite membrane, PVDF, PES), and the calibration results will vary depending on the membrane material.

[0036] S3, Extraction and reverse correction of equipment operation control parameters.

[0037] In the step of extracting the operating control parameters of the water treatment equipment based on the equivalent load sequence and the dynamic treatment flux reduction factor: The system fuses the continuous comprehensive equivalent load sequence with the dynamic processing flux reduction coefficient into a multi-dimensional state vector to construct and generate the current multi-dimensional operating condition state vector.

[0038] Retrieve reference coordinates from the preset two-dimensional control mapping table. The discretization criterion for the two-dimensional control mapping table is: to divide the comprehensive equivalent load value range ( ) divided into Each discrete interval will dynamically process the flux reduction factor ( ) divided into The system consists of discrete intervals, forming 2500 grid nodes. The equipment supplier, using a hardware-in-the-loop simulation platform, calculated the operating voltage (in units) of the equipment that achieves the required water quality and lowest energy consumption for each node under various operating conditions. The values ​​(e.g., 220V or 380V) and the corresponding drug dosing frequency (in Hz) are calibrated and stored in the mapping table.

[0039] The system calculates the Euclidean distance between the multi-dimensional operating condition state vector and the coordinates of each grid node, extracts the equipment operating voltage value and reagent dosing frequency value corresponding to the reference coordinate with the smallest Euclidean distance, and sends them to the water treatment equipment. After the water source configuration stage outputs the mixing ratio of each drinking water source, the system establishes a full-link reverse correction logic: based on the actual allocated mixing flow of each water source, the reagent dosing frequency is proportionally gained and corrected. The correction formula is that the actual dosing frequency equals the initial mapped dosing frequency multiplied by the ratio of the actual mixing flow to the rated treatment flow, realizing the interaction of water quality assessment, parameter extraction, and water distribution control.

[0040] S4, Calculation of interference matrix for mixed water source reaction.

[0041] In the step of extracting interactively sensitive ion concentration data from water quality testing data and constructing a multi-source mixing reaction interference matrix based on the interactively sensitive ion concentration data: The system identifies metal cations (such as metal cations) in the detection data. ) and acid radical anions (such as For any two water sources and Calculate water source Metal cations in water source The actual ion product value after cross-mixing of acid radicals and anions.

[0042] To address the significant differences in the hazards posed by various sediments to water supply networks and human health, this solution incorporates a weighted approach to sediment hazard levels within its sediment intervention logic. The weighting is based on a comprehensive assessment of the risk of sediment clogging the network and the toxicity of heavy metals: ordinary hardness scale (such as calcium sulfate and calcium carbonate) is weighted at 1.0, highly clogging sediments (such as magnesium hydroxide) at 1.5, and harmful heavy metal sediments (such as ferric hydroxide and lead sulfate) at 3.0 or higher.

[0043] The hazard level weighting coefficients for precipitates are assigned based on the degree of impact of various precipitates on human health and the risk of pipe scaling. The grading rules are as follows: heavy metal toxicity precipitates pose the greatest hazard and are assigned the highest weight, such as PbSO4. The toxicity of lead is referenced to the WHO's Guidelines for Drinking-water Quality (0.01 mg / L) and its neurotoxic and nephrotoxic effects. Heavy metal precipitates are next, such as Fe(OH)3, where excessive iron intake may cause hemochromatosis. Hydroxide scaling is next, followed by carbonate scaling. Sulfate scaling is the least toxic; for example, while magnesium and calcium precipitates are not significantly toxic, pipe scaling can affect water delivery capacity. The hazard level weighting coefficients are used to quantify the risk priority of different precipitation reactions in multi-source mixing optimization. A higher hazard level weighting coefficient indicates that the precipitation reaction should be prioritized for avoidance in the optimization solution.

[0044] Preset standard solubility product constant of sparingly soluble substances at 25℃ ( The hazard levels and their weights are shown in Table 3: Table 3: Standard solubility product constant of pre-specified sparingly soluble substances at 25℃ ( ) and Hazard Level Weight Table

[0045] The formula for calculating the intervention weight value is as follows: ; ;in, Indicates water source Metal cations and water source The intervention weight value when acid radicals and anions cross-mix; This indicates the hazard level weighting coefficient of the precipitate; This represents the actual ion product of ions in the mixed water, in units of . ; Indicates water source The molar concentration of the metal cation, in units of ; Indicates water source The molar concentration of the anion in the middle acid radical, in units of ; The stoichiometric coefficient of the cation in the precipitate's chemical formula is a positive integer; The stoichiometric coefficient of the anion in the precipitate chemical formula is a positive integer; Represents the standard solubility product constant, in units of .

[0046] When the actual ion product is less than or equal to the solubility product constant, the intervention weight value is set to zero. The system only selects ion pairs whose actual ion product exceeds the solubility product constant for calculation, and combines the intervention weight values ​​of each pair of water sources into a symmetrical multi-source mixing reaction interference matrix.

[0047] S5, water source allocation optimization solution and command issuance.

[0048] In the step of optimizing water resource allocation based on the dynamic processing flux reduction factor, the multi-source mixing reaction interference matrix, and the target water demand, and generating data management and scheduling instructions containing the mixing ratio of each drinking water source: The goal of configuration optimization is to find the optimal mixing ratio of various water sources that maximizes water quality safety margin while mitigating sedimentation risks. Addressing the issue of inconsistent dimensions between water quality margin and sedimentation penalty in existing technologies, this scheme employs dimensionless normalization: the water quality margin is converted into a dimensionless relative margin by dividing the difference between the national standard limit and the expected mixed concentration by the national standard limit; health hazard weights are calculated based on the health guidance values ​​for each pollutant in the WHO Guidelines for Drinking-water Quality, Fourth Edition (2022). The conversion rule is: the health hazard weight is inversely proportional to the health guidance value; the lower the health guidance value, the more toxic the substance, and the higher the weight. The specific numerical correspondences are as follows: Lead (0.01 mg / L) and arsenic (0.01 mg / L) are the most toxic and have the highest weights, set at 0.35; copper (2.0 mg / L) is second, set at 0.15; ammonia nitrogen (1.5 mg / L) is third, set at 0.10; the sum of health hazard weights is 1, therefore the health hazard weight of the permanganate index is set at 0.05. Health hazard weights of target pollutants ( The limits for drinking water quality and the national drinking water standards are shown in Table 4: Table 4: Health Hazard Weights of Target Pollutants ( Table of Limits for National Drinking Water Standards

[0049] The objective function and constraint system for configuration optimization are as follows: ; in, The numerical value of the dimensionless objective function characterizing the water quality safety margin; This represents the total number of pollutant indicators to be assessed, and is a positive integer. Indicates the pollutant indicator number; Indicates the first National standard limits for each pollutant, in units of ; This indicates the expected concentration of the mixture, in units of... ; Indicates the first Health hazard weighting values ​​for each pollutant; This represents the dynamic penalty coefficient for precipitation intervention; This represents the total number of water sources, and is a positive integer. Indicates the intervention weight value; They represent water sources respectively. With water source The mixing ratio is a dimensionless ratio value between 0 and 1.

[0050] The constraints include: 1. Water source proportion normalization constraint: ,and ; 2. Equipment water supply upper limit constraint: ,in This represents the target water demand, in units of... ; Indicates water source The rated water production capacity of the treatment equipment, in units of ; Indicates water source The dynamic throughput reduction factor of the equipment; 3. Pipeline hydraulic constraints: The calculated pressure at the pipeline nodes satisfies... The unit is kPa, and the pipeline flow rate does not exceed the maximum allowable overflow rate of the pipeline network. . This is the minimum allowable water pressure limit for pipeline nodes, which is a threshold for system safety and service assurance, ensuring that the water supply pressure for end users or high-rise buildings is not lower than the minimum standard. This is the maximum allowable water pressure at a pipeline node, belonging to the physical pressure resistance safety threshold of the pipeline network, to prevent excessive pressure from causing pipeline rupture or equipment damage. Calculating pressure at nodes refers to setting the mixing ratio (or proportioning parameter) of each water source as follows: At that time, the actual water pressure values ​​at each node of the pipeline network were calculated using a hydraulic model.

[0051] Dynamic penalty coefficient Based on the maximum value of the non-zero weights in the intervention matrix Compared with the average calculate: The base scaling factor The convergence threshold for the solution algorithm is set to be the change in the objective function being less than a certain value in 50 consecutive iterations. .

[0052] To ensure continuous water supply under extreme post-disaster conditions, this solution includes an emergency backup strategy: when a single water source goes offline, equipment fails, or the disaster area experiences network or power outages, the system automatically switches to offline local backup mode, removes the faulty water source, substitutes the upper limit constraint of the remaining available water source into the simplified model, and recalculates and outputs the emergency mixing ratio according to the principle of maximizing water quality safety margin.

[0053] To demonstrate the technical effectiveness of this invention, a quantitative performance comparison test was conducted. In a simulated post-disaster water supply scenario involving three different water sources (high organic matter, high hardness, and slightly polluted heavy metals), this scheme was compared with traditional fixed-parameter water distribution schemes and traditional constant-flux control schemes. The test results are shown in Table 5. Table 5: Comparison of Quantitative Performance Test Data Between This Solution and Traditional Solutions

[0054] Example 1 The disaster site includes three temporary drinking water sources: Source A (a river polluted by surface runoff, with high concentrations of organic matter and suspended solids), Source B (high-hardness well water, with high concentrations of calcium and magnesium cations), and Source C (water from a damaged reservoir, with high concentrations of sulfate and other acid anions). Three sets of ultrafiltration / reverse osmosis water treatment equipment are deployed on-site, corresponding to each water source. The actual water demand of the target water supply area is set as follows: .

[0055] Step 1: Water quality testing and comprehensive equivalent load calculation.

[0056] The system acquires multimodal water quality monitoring data (measured in units of mass concentration) from water sources A, B, and C in real time via on-site data processing terminals. ).

[0057] After fault-tolerant cleaning, the water quality data collected by the sensors shows the following concentrations of major pollutants: Water source A: ammonia nitrogen , ; Substitute the parameters in Table 1 (ammonia nitrogen) ;lead ): Comprehensive equivalent load of water source A: .

[0058] Water source B: ,Right now ; ,Right now ammonia nitrogen Comprehensive equivalent load of water source B: .

[0059] Water source C: Phenol Comprehensive equivalent load of water source C: .

[0060] Step 2: Calculation of membrane equipment operating conditions and flux reduction factor.

[0061] Water source A treatment equipment (rated water production) Current membrane pressure ,flow membrane area The dynamic viscosity coefficient of water .

[0062] Calculate the current membrane resistance: .

[0063] Known initial cleanliness resistance drag coefficient Damage coefficient The historical cumulative load integral term equals 20.0.

[0064] Substitute into the formula to calculate the dynamic processing flux reduction factor: .

[0065] because The system automatically sends a hydraulic backwash command to water source A.

[0066] Similarly, the flux reduction factors for water source B and water source C are calculated as follows: , .

[0067] Step 3: Extraction and reverse correction of equipment operation control parameters.

[0068] Water source A and Composition of multi-dimensional working condition state vector The grid node with the smallest Euclidean distance is retrieved in the two-dimensional control mapping table, and the baseline operating control parameters are extracted: operating voltage 220V, reagent dosing frequency 30.0Hz.

[0069] Step 4: Calculation of interference matrix for mixed water source reaction.

[0070] Analysis of cross-mixing between water source A and water source B: Water source B concentration , Water source B concentration .

[0071] Cross-ion product .

[0072] According to Table 3 of Harm weight .

[0073] because Calculate the intervention weight value: .

[0074] When water source A and water source C are mixed, there are no excessive ions, and the corresponding intervention weight is 0.

[0075] Build Multi-source mixing reaction interference matrix : ; Step 5: Solve the water source allocation optimization and issue instructions.

[0076] The maximum water supply capacity of each water source is as follows: Water source A: ; Water source B (rated) ): ; Water source C (rated) ): .

[0077] Substituting the constraints and disturbance matrix into the configuration optimization objective function, and calling the solver to perform planning calculations, the optimal mixing ratio that maximizes water quality safety margin and minimizes sedimentation penalty is output as follows: Water Source A Proportion (flow ); Water Source B Proportion (flow ); Water source C proportion (flow ).

[0078] Total water supply The demand is fully met, and the flow rates of each water source do not exceed their reduced capacity limits.

[0079] Finally, reverse water distribution correction is performed: actual flow rate of water source A. Rated flow rate The system adjusts the dosing frequency of the reagents in water source A from the initial 30.0Hz to 15.0Hz, and generates a data management and scheduling instruction containing the final control parameters and proportions for execution.

[0080] like Figure 2 As shown, this application provides a post-disaster drinking water quality assessment, treatment, and optimization system to implement the aforementioned post-disaster drinking water quality assessment, treatment, and optimization method, comprising: The equivalent conversion calculation module is used to acquire water quality test data from multiple drinking water sources, perform data fault-tolerant cleaning, and convert the water quality test data into an equivalent load sequence of target pollutants after heavy metal toxicity correction. The flux reduction calculation module is used to acquire the operating condition data of the water treatment equipment, calculate the membrane resistance and dynamic treatment flux reduction coefficient that distinguish between reversible and irreversible fouling based on the equivalent load sequence and operating condition data, and link backwash control when the flux is too low. The control parameter extraction module is used to extract the operating control parameters of the water treatment equipment after performing multi-dimensional state vector fusion based on the equivalent load sequence and dynamic processing flux reduction coefficient, and to realize the reverse correction of the reagent dosing frequency based on the water distribution results. The interference matrix construction module is used to extract interactive sensitive ion concentration data from water quality testing data and construct a multi-source mixed reaction interference matrix by combining the hazard level weight of sediments. The water resource allocation optimization module is used to perform dimensionally unified water resource allocation optimization based on dynamic processing flux reduction coefficient, multi-source mixed reaction interference matrix, pipeline hydraulic constraints, and target water demand, and to generate data management scheduling instructions and offline emergency backup scheduling instructions.

[0081] All characters in the above formulas have been clearly explained and labeled with standard units below the formulas. The formulas are all derived from existing mature physical and chemical theories. The preset parameters are set by those skilled in the art based on the actual working conditions on site.

[0082] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for post-disaster drinking water quality assessment, treatment, and optimization, characterized in that: Includes the following steps: After acquiring water quality test data from multiple drinking water sources and performing data anomaly correction and fault-tolerant cleaning, the water quality test data is converted into an equivalent load sequence of the target pollutants through the chemical oxygen demand equivalent rule corrected for heavy metal toxicity. The system acquires the operating condition data of the water treatment equipment, calculates the membrane resistance that distinguishes between reversible and irreversible fouling based on the equivalent load sequence and operating condition data, further calculates the dynamic treatment flux reduction factor of the water treatment equipment, and executes the backwashing procedure when the dynamic treatment flux reduction factor reaches the preset backwashing threshold. Based on the equivalent load sequence and dynamic processing flux reduction factor, multi-dimensional state vector fusion is performed to extract the operation control parameters of the water treatment equipment. The operation control parameters are sent to the water treatment equipment, and the reagent dosing frequency is corrected in reverse according to the mixing ratio. Interactive sensitive ion concentration data are extracted from water quality testing data. Combined with the hazard level weights of different precipitates, a multi-source mixed reaction interference matrix is ​​constructed based on the interactive sensitive ion concentration data. Based on the dynamic processing flux reduction coefficient, the multi-source mixing reaction interference matrix, the pipeline flow and pressure constraints, and the target water demand, a unified water resource allocation optimization solution is executed to generate data management and scheduling instructions containing the mixing ratio of each drinking water source, and an emergency backup strategy is triggered in emergency situations.

2. The method for post-disaster drinking water quality assessment, treatment, and optimization as described in claim 1, characterized in that, The process involves acquiring water quality testing data from multiple drinking water sources and converting the data into equivalent load sequences of the target pollutants. Analyze water quality testing data to separate the concentration sequences of individual pollutants; Retrieve the preset pollutant equivalent conversion and toxicity correction rule table, and assign the chemical oxygen demand equivalent coefficient and heavy metal toxicity correction coefficient to each independent pollutant according to the rules; A comprehensive equivalent load sequence is generated based on the concentration numerical sequence, the chemical oxygen demand equivalent coefficient, and the heavy metal toxicity correction coefficient.

3. The method for post-disaster drinking water quality assessment, treatment, and optimization as described in claim 2, characterized in that, The generated integrated equivalent load sequence: Three-standard-deviation filtering was applied to the water quality test data to remove outliers from the sensor; if missing values ​​were found, the moving median was used to fill them in. Based on the sampling timestamps of the water quality testing data, the concentration sequences of individual pollutants are time-series aligned; The sliding window step size is adaptively adjusted according to the severity of water quality fluctuations. The comprehensive equivalent load sequence within the time-sliding window is output by summing the products of the aggregated and aligned concentration numerical sequence after time-sliding window aggregation, the corresponding pollutant's chemical oxygen demand equivalent coefficient, and the heavy metal toxicity correction coefficient. The specific formula is as follows: ; in, Indicates a time sliding window The comprehensive equivalent load value within, in units of ; This represents the total number of independent pollutant species detected, and is a positive integer. Indicates a time sliding window Inner Concentration values ​​for each individual pollutant, in units of ; Indicates the first Preset chemical oxygen demand equivalent coefficients for each independent pollutant; Indicates the first Correction factor for heavy metal toxicity of each independent pollutant.

4. The method for post-disaster drinking water quality assessment, treatment, and optimization as described in claim 1, characterized in that, The process involves acquiring operating condition data of the water treatment equipment and calculating the dynamic treatment flux reduction factor of the water treatment equipment based on the equivalent load sequence and operating condition data. Extract the current membrane inlet pressure and current inlet flow rate of the water treatment equipment from the operating condition data; Calculate the time integral of the equivalent load sequence over the historical operating cycle to obtain the cumulative retained load. The current actual membrane resistance is decomposed into initial clean membrane resistance, reversible fouling resistance and irreversible fouling resistance. The current inlet pressure, current influent flow rate and cumulative interception load are substituted into the preset membrane fouling attenuation model to calculate the dynamic treatment flux reduction coefficient. When the dynamic processing flux reduction factor is less than or equal to the preset hydraulic backwash threshold, hydraulic backwashing is triggered. When the flux does not recover to the set range after hydraulic backwashing, chemical backwashing is triggered.

5. The method for post-disaster drinking water quality assessment, treatment, and optimization as described in claim 4, characterized in that, The current inlet pressure, current influent flow rate, and cumulative retention load are substituted into a preset membrane fouling attenuation model to calculate the dynamic treatment flux reduction factor. Based on the current membrane pressure, current influent flow rate, and fluid viscosity at the current water temperature, the actual membrane resistance is calculated using Darcy's law. Calculate the ratio of the difference between the current actual membrane resistance value and the initial clean membrane resistance value; The dynamic processing throughput reduction factor is output based on the nonlinear fitting relationship between the difference ratio and the cumulative intercepted load.

6. The method for post-disaster drinking water quality assessment, treatment, and optimization as described in claim 1, characterized in that, The operating control parameters of the water treatment equipment are extracted based on the equivalent load sequence and the dynamic processing flux reduction factor. The equivalent load sequence is fused with the dynamic processing flux reduction factor to generate a multi-dimensional operating condition state vector. Retrieve the reference coordinates that have the smallest Euclidean distance to the current multidimensional operating condition state vector from the preset two-dimensional control mapping table; The operating voltage and reagent dosing frequency values ​​corresponding to the reference coordinates are extracted as operating control parameters. After optimizing the mixing ratio of each drinking water source, the reagent dosing frequency is reversed based on the actual mixing flow rate.

7. The method for post-disaster drinking water quality assessment, treatment, and optimization as described in claim 1, characterized in that, The process involves extracting interactively sensitive ion concentration data from water quality testing data, and constructing a multi-source mixed reaction interference matrix based on this data. The concentration values ​​of metal cations and acid anions in water quality test data are identified as interactively sensitive ion concentration data. For any two different drinking water sources, calculate the ion product of the cross-combination of metal cations in one source with acid anions in the other source. Retrieve the solubility product constant and hazard level weight of the corresponding precipitate from the pre-set database, compare the ion product value with the solubility product constant, and map the comparison result to the intervention weight value in the multi-source mixing reaction interference matrix: When the ion product value is greater than the solubility product constant, the logarithm of the excess is multiplied by the precipitate hazard level weight and converted into a positive intervention weight value. When the ion product value is less than or equal to the solubility product constant, the intervention weight value is set to zero; The intervention weights of all water sources are combined into a symmetric multi-source mixing reaction interference matrix.

8. The method for post-disaster drinking water quality assessment, treatment, and optimization as described in claim 1, characterized in that, The process involves optimizing water resource allocation based on dynamic processing flux reduction coefficients, multi-source mixing reaction interference matrices, and target water demand, generating data management and scheduling instructions that include the mixing ratios of each drinking water source. The objective function is to maximize the water quality safety margin of all drinking water sources after relative margin normalization. The product of the dynamic processing flux reduction factor and the rated water production of each treatment device is used as the upper limit constraint of the supply of each water source. The maximum allowable flow rate of the pipeline network and the range of hydraulic pressure at the nodes of the pipeline network are used as hydraulic constraints for the pipeline network. The product of the multi-source mixing reaction interference matrix and the quadratic product of the mixing ratio of each drinking water source, multiplied by the dynamic penalty coefficient, is used as a dimensionless mixing penalty constraint and incorporated into the objective function for planning and solving. The output is the mixing ratio of each drinking water source that satisfies the target water demand and maximizes the objective function. In offline or faulty conditions, an emergency backup strategy is triggered to cut off the faulty water source and redistribute the remaining water source proportions.

9. The method for post-disaster drinking water quality assessment, treatment, and optimization as described in claim 8, characterized in that, The objective function is to maximize the water quality safety margin of all drinking water sources. Calculate the expected concentration value of the mixture after all drinking water sources are mixed according to the set mixing ratio; Calculate the difference between the national standard limit and the expected concentration of the mixture, and divide by the national standard limit to obtain the dimensionless relative margin; Multiply the dimensionless relative margin by the health hazard weights of each pollutant, which are calculated from the reciprocal of the World Health Organization's health guidance values, and sum the results. Subtract the dimensionless sedimentation penalty term to obtain the objective function characterizing the water quality safety margin.

10. A post-disaster drinking water quality assessment, treatment, and optimization system, used to implement the post-disaster drinking water quality assessment, treatment, and optimization method according to any one of claims 1-9, characterized in that, include: The module includes: equivalent conversion calculation module, flux reduction calculation module, control parameter extraction module, interference matrix construction module, and water source allocation optimization module. The equivalent conversion calculation module is used to acquire water quality test data from multiple drinking water sources, perform data fault-tolerant cleaning, and convert the water quality test data into an equivalent load sequence of target pollutants after heavy metal toxicity correction. The flux reduction calculation module is used to acquire the operating condition data of the water treatment equipment, calculate the membrane resistance and dynamic treatment flux reduction coefficient that distinguish between reversible and irreversible fouling based on the equivalent load sequence and operating condition data, and link backwash control when the flux is too low. The control parameter extraction module is used to extract the operating control parameters of the water treatment equipment after performing multi-dimensional state vector fusion based on the equivalent load sequence and dynamic processing flux reduction coefficient, and to realize the reverse correction of the reagent dosing frequency based on the water distribution results. The interference matrix construction module is used to extract interactive sensitive ion concentration data from water quality testing data and construct a multi-source mixed reaction interference matrix by combining the hazard level weight of sediments. The water resource allocation optimization module is used to perform dimensionally unified water resource allocation optimization based on dynamic processing flux reduction coefficient, multi-source mixed reaction interference matrix, pipeline hydraulic constraints, and target water demand, and to generate data management scheduling instructions and offline emergency backup scheduling instructions.

Images