Acid rain component detection chip
Through the acid rain component detection chip, combined with thermodynamic model and machine learning algorithm, the rapid and accurate detection of acid rain ion concentration is achieved, and the problems of high detection cost and low accuracy in the existing technology are solved, and are suitable for large-scale acid rain monitoring.
Patent Information
- Application Number
- CN202411046458.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-31
- Publication Date
- 2024-12-06
AI Technical Summary
The existing acid rain component detection technology has problems such as high detection cost, low sensitivity and accuracy, which is difficult to meet the needs of large-scale and high-frequency monitoring, and lacks adaptability and stability.
The acid rain component detection chip is used to read and execute acid rain component detection instructions through the processor to obtain the pH value and conductivity of the sample. Combined with thermodynamic models and machine learning algorithms, ion equation systems and regression models are constructed, and iterative updates and data fusion are performed. Predict ion concentration.
It realizes fast and accurate acid rain ion component detection, reduces detection costs, improves detection accuracy and applicability, and is suitable for large-scale monitoring.
Smart Images

Figure CN119086741A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the technical field of acid rain detection, and in particular to an acid rain component detection chip. Background Art
[0002] Acid rain has become a major environmental problem worldwide, posing a serious threat to ecosystems, buildings and human health. Acid rain is mainly formed by the reaction of sulfur oxides and nitrogen oxides in the atmosphere with water. These pollutants mainly come from industrial emissions, motor vehicle exhaust and fossil fuel combustion. The detection of acid rain components is crucial for evaluating environmental quality, formulating pollution control strategies and studying the impact of acid rain on ecosystems. Accurate detection of acid rain components helps government departments formulate effective environmental protection policies, helps scientific researchers to gain a deep understanding of the formation mechanism and environmental impact of acid rain, and provides a basis for pollution control for industrial enterprises. Therefore, acid rain component detection is of great significance in many fields such as environmental protection, scientific research and industrial development. At present, the main technologies for acid rain component detection include ion chromatography, conductivity method and titration. Ion chromatography is the most commonly used method, which determines the acid rain composition by separating and quantitatively analyzing various ions in acid rain. The conductivity law indirectly estimates the acidity and ion concentration of acid rain by measuring its conductivity. Titration is mainly used to determine the pH value and acidity of acid rain. The working principles of these methods are based on chemical analysis and physical measurement, and can provide relatively accurate information on the composition of acid rain. However, these traditional methods have some problems: first, they usually require professional laboratory equipment and personnel, making it difficult to achieve rapid on-site detection; second, the sample collection and processing process may introduce errors; third, these methods are often time-consuming and difficult to meet the needs of large-scale, high-frequency monitoring; finally, the detection sensitivity and accuracy of some methods need to be improved, especially for the detection of low-concentration ions.
[0003] To solve the above problems, researchers have developed some new detection methods, such as spectral analysis, electrochemical sensors and biosensors. These methods have improved the shortcomings of traditional technologies in some aspects, such as improving detection speed or portability. However, these methods still have some defects. First, they can usually only detect a limited number of ions, making it difficult to fully analyze the composition of acid rain; second, the detection accuracy and stability of these methods are often greatly affected by environmental factors; third, they lack adaptive capabilities and are difficult to cope with complex and changing actual environments; finally, these methods usually require frequent calibration and maintenance, which increases the cost and difficulty of use.
[0004] Therefore, there is an urgent need for a technical solution that can reduce detection costs and improve detection accuracy and adaptability. Summary of the invention
[0005] In order to solve the deficiencies of the prior art, the embodiments of the present disclosure provide an acid rain component detection chip, which solves the technical problems of the prior art in acid rain ion concentration detection such as high cost, low detection sensitivity and accuracy.
[0006] The present disclosure provides an acid rain component detection chip, the chip includes one or more processors and one or more data interfaces, the processor reads and executes the acid rain component detection instruction stored on the target storage medium through the data interface, the acid rain component detection instruction includes: obtaining the pH value and conductivity of the target acid rain sample, and building an ion equation group in combination with the concentration balance and thermodynamic model of the preset ion types; calculating the residual by bringing the initialized ion concentration matrix into the ion equation group, and iterating and updating the ion concentration matrix once or multiple times to obtain an approximate solution of the ion equation group; building a regression model corresponding to each ion concentration, and inputting data including the pH value, conductivity and approximate solution into all regression models to obtain a predicted value of each ion concentration; fusing the approximate solution and the predicted value based on the pH value and conductivity to obtain the concentration value of each ion in the target acid rain sample.
[0007] According to the aspects described above and any possible implementation method, an implementation method is further provided, wherein an approximate solution to the ion equations is obtained by bringing an initialized ion concentration matrix into the ion equations to calculate the residual, and performing one or more iterative updates on the ion concentration matrix, including: initializing a parameter vector including initial estimated values of each ion concentration to obtain an ion concentration matrix; bringing the ion concentration matrix into the ion equations to calculate the residual; calculating the Jacobian matrix of the ion equations, and solving the incremental equations in combination with the residual; cumulatively updating the ion concentration matrix using the solved increments; performing one or more iterative updates on the ion concentration matrix, and stopping the iterative update when the absolute value of the residual is less than a first threshold or the increment is less than a second threshold, to obtain an ion concentration matrix that is an approximate solution to the ion equations.
[0008] In a possible implementation manner, a method for obtaining the first threshold includes: Among them, τ 1 represents the first threshold, α represents the scaling factor, which is 0.03, β represents the parameter controlling the convergence speed, which is 0.27, n represents the current number of iterations, pH represents the pH value of the target acid rain sample, σ represents the conductivity of the target acid rain sample, σ 0 represents the standard conductivity, with a value of 1 μS / cm, γ represents the pH adjustment parameter, with a value of 0.5, δ represents the conductivity adjustment parameter, with a value of 0.25, c i Represents the concentration value of the i-th ion in the ion concentration matrix.
[0009] In a possible implementation manner, a method for obtaining the second threshold includes: Among them, T 2 represents the second threshold, ε represents the basic threshold, which is 1e-6, σ represents the conductivity of the target acid rain sample, σ 0 represents the standard conductivity, with a value of 1μS / cm, λ represents the influence factor of conductivity, with a value of 0.53, θ represents the influence factor of concentration change, with a value of 0.46, Δc i represents the change in the concentration of the i-th ion in this iteration, c i represents the concentration value of the ion in the ion concentration matrix, c 0 represents the reference concentration, which is 1 mol / L; μ represents the influencing factor of total ion concentration, which is 0.3.
[0010] In a possible implementation, before inputting data including pH value, conductivity and approximate solution into all regression models to obtain the predicted value of each ion concentration, it also includes: collecting multiple acid rain samples, and constructing a training data set with the corresponding pH value, conductivity and ion concentration approximate solution as input features, and the actual ion concentration measured by high-precision ion chromatography as a label; constructing a loss function based on the chemical equilibrium equation to perform integrated training on all regression models.
[0011] In one possible implementation, a loss function is constructed based on a chemical equilibrium equation, including: L = MSE + λ 1 ·I∑(z i ·C i,predicted )-0| 2 +λ 2 ·L reg , where L represents the loss function, MSE represents the mean square error term, and λ 1 represents the first coefficient, z i represents the charge number of ion i, C i,predicted represents the predicted concentration of ion i, λ 2 represents the second coefficient, L reg represents the regularization term.
[0012] A possible implementation method also includes: using acid rain data from other regions or laboratories to improve the performance of the model in the target area through a transfer learning method; and / or simultaneously predicting ion concentration and other related parameters during the training process, wherein the related parameters include acid deposition flux, so as to perform multi-task learning training on the model.
[0013] In a possible implementation, the approximate solution and the predicted value are merged based on the pH value and the conductivity to obtain the concentration value of each ion in the target acid rain sample, including: determining a weighting coefficient based on the pH value and the conductivity to perform a weighted fusion of the approximate solution and the predicted value; wherein the weighting coefficient includes: Wherein, w represents the weighting coefficient, a and b represent the adjustment parameters, pH represents the pH value of the target acid rain sample, EC represents the conductivity of the target acid rain sample, pH 0 ,EC 0 represents the basic value; constructs the target optimization problem according to the ion non-negative constraint, charge balance constraint and conductivity constraint; transforms the target optimization problem into an equality constraint and constructs the Lagrangian function; uses the Newton method to iteratively solve the KKT condition to obtain the concentration value of each ion in the target acid rain sample.
[0014] In an acid rain component detection chip provided above, the embodiment of the present disclosure can realize rapid and accurate detection of the main ion components and their concentrations in acid rain by combining mathematical models and machine learning algorithms; the detection chip of the embodiment of the present disclosure has the advantages of fast detection speed, low cost, high accuracy, and wide applicability.
[0015] It should be understood that the contents described in the summary of the invention are not intended to limit the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, among which:
[0017] Figure 1 A schematic block diagram of an acid rain component detection chip provided in an embodiment of the present disclosure;
[0018] Figure 2 A schematic diagram of a flow chart of an acid rain component detection instruction provided by an embodiment of the present disclosure;
[0019] Figure 3 A schematic diagram of a flow chart of an instruction for approximate calculation of acid rain ion concentration provided by an embodiment of the present disclosure. DETAILED DESCRIPTION
[0020] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.
[0021] In addition, the term "and / or" in this article is only a description of the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.
[0022] Figure 1 A schematic block diagram of an acid rain component detection chip provided for an embodiment of the present disclosure. It should be understood that the chip shown in the figure is exemplary and not restrictive. This means that the chip architecture involved is not limited to a specific form or design, but is presented as an example. In other words, the architecture shown in the figure can be regarded as a way of expression to clearly describe related concepts and relationships, and does not exclude other forms of architecture. Therefore, when interpreting the architecture in the picture, it should be understood that the chip has flexibility and diversity, and its purpose is to provide an exemplary description rather than a restrictive provision on a specific form.
[0023] like Figure 1 As shown, an acid rain component detection chip 100 includes a processor and a data interface. The processor reads and executes acid rain component detection instructions stored on a target storage medium through the data interface. The acid rain component detection instructions include: obtaining the pH value and conductivity of the target acid rain sample, and constructing an ion equation group in combination with the concentration balance and thermodynamic model of the preset ion types; calculating the residual by bringing the initialized ion concentration matrix into the ion equation group, and iterating and updating the ion concentration matrix once or multiple times to obtain an approximate solution to the ion equation group; constructing a regression model corresponding to each ion concentration, and inputting data including pH value, conductivity and approximate solution into all regression models to obtain a predicted value of each ion concentration; fusing the approximate solution and the predicted value based on pH value and conductivity to obtain the concentration value of each ion in the target acid rain sample.
[0024] Figure 2 A schematic diagram of a flow chart of an acid rain component detection instruction provided in an embodiment of the present disclosure. The present disclosure proposes a chip for detecting acid rain ion components and concentrations based on pH value and conductivity. The chip measures the pH value and conductivity of rainwater samples, combines mathematical models and machine learning algorithms, and realizes rapid and accurate detection of the main ion components and their concentrations in acid rain. The detection chip in the embodiment of the present disclosure has the advantages of fast detection speed, low cost, high accuracy, and wide applicability, and provides an innovative technical means for acid rain monitoring and research.
[0025] Specifically, if Figure 2 As shown, at step S201, the pH value and conductivity of the target acid rain sample are obtained, and an ion equation group is constructed in combination with the concentration balance and thermodynamic model of the preset ion species.
[0026] In one embodiment, for sample collection and pretreatment, a clean polyethylene container can be used to collect rainwater samples to avoid contamination. Measure the sample volume V (L) with an accuracy of 0.1 mL. Filter the sample using a polytetrafluoroethylene (PTFE) filter membrane with a pore size of 0.45 μm to remove suspended matter. Transfer the filtered sample to a clean polyethylene bottle and seal it for storage. Control the sample storage and measurement environment temperature at 25 ± 1 ° C to ensure the accuracy of subsequent measurements.
[0027] Then the pH value and conductivity are measured. On the one hand, the pH value is measured: the pH value of the sample is measured using a pH meter that has been calibrated at three points (pH 4.01, 7.00, 10.01). The measurement accuracy should reach ±0.01 pH unit. The three-parallel determination method is used, with a 1-minute interval between each measurement, and the average value is taken as the final result. On the other hand, the conductivity is measured: the conductivity κ (μS / cm) of the sample is measured using a conductivity meter calibrated with a standard KC1 solution (0.01mol / L). The measurement accuracy should reach ±0.5%. The three-parallel determination method is used, with a 1-minute interval between each measurement, and the average value is taken as the final result.
[0028] Through the above operations, the pH value and conductivity of the target acid rain sample can be obtained and stored in the corresponding target storage medium for the detection chip to read through the data interface.
[0029] Next, the ion equations are constructed by combining the concentration balance of the preset ion types and the thermodynamic model. It should be noted that according to the principle of electroneutrality, the total equivalent concentration balance equation of cations and anions is established: ∑[Cation + ]=∑[Anion - ], considering that the main ions in acid rain include: H + , NH 4 + , Ca 2+ ,Mg 2+ , Na + , K + , OH - , F - , SO 4 2- , NO 3 - , Cl - , HCO 3 - , C 2 O 4 2- For each ion i, its equivalent concentration [i] eq It can be expressed as: [i] eq =c i ×|z i|, where c i is the ion molar concentration (mol / L), z i is the ion valence. The specific equilibrium equation can be expressed as:
[0030] The second step is to establish a thermodynamic model. In order to accurately describe the behavior of ions in solution, an extended Debye-Hckel equation can be established to calculate the activity coefficient γ of each ion. i : Among them, A and B are temperature-related constants, A = 0.5085, B = 0.3281 at 25°C. i is the ion valence, is the ion size parameter (nm), b is the empirical coefficient (taken as 0.1), I is the ion strength, and the calculation formula is: For the major ions, The possible values are as follows: H+: 0.9, NH4+: 0.25, Ca2+: 0.6, Mg2+: 0.8, Na+: 0.45, K+: 0.3, OH - :0.35、F - :0.35, SO42-:0.4, NO3-:0.3, Cl-:0.3, HCO3-:0.45, C 2 O 4 2- : 0.5.
[0031] Therefore, a nonlinear ionic equation system can be constructed based on pH value, conductivity, ion balance and thermodynamic model:
[0032] 1. pH = -log[H + ];
[0033] 2. Among them, Λ 0 is the limiting molar conductivity of the ion (S·cm 2 ·mol -1 ), the values are as follows: 0 (H + )=349.8、 Λ 0 (Ca 2+ )=119.0、Λ 0 (Mg 2+ )=106.0、Λ 0 (Na + )=50.1、 Λ 0 (Cl- )=76.3、
[0034] 3.
[0035] 4.K W =[H + ][OH - ], where K W is the ion product constant of water, K at 25°C W =1.0×10 -14 ;
[0036] 5. in, is the ionization constant of bicarbonate ion at 25℃
[0037] 6. in, is the first ionization constant of carbonic acid at 25°C
[0038] 7. For each ion i, calculate its activity coefficient γ i .
[0039] At step S202, the initialised ion concentration matrix is brought into the ion equations to calculate the residual, and the ion concentration matrix is iteratively updated once or multiple times to obtain an approximate solution to the ion equations. Figure 3 As shown, the specific instruction steps will be Figure 3 ), initialize a parameter vector including initial estimated values of each ion concentration to obtain an ion concentration matrix; bring the ion concentration matrix into the ion equation system to calculate the residual; calculate the Jacobian matrix of the ion equation system, and solve the incremental equation in combination with the residual; cumulatively update the ion concentration matrix using the solved increments; iteratively update the ion concentration matrix one or more times, and stop iterative updating when the absolute value of the residual is less than a first threshold or the increment is less than a second threshold, to obtain the ion concentration matrix as an approximate solution to the ion equation system.
[0040] In step S203, a regression model corresponding to each ion concentration is constructed, and data including pH value, conductivity and approximate solution are input into all regression models to obtain a predicted value of each ion concentration.
[0041] It should be noted that in one implementation scenario, before inputting data including pH value, conductivity and approximate solution into all regression models to obtain the predicted value of each ion concentration, it also includes: collecting multiple acid rain samples, and using the corresponding pH value, conductivity and ion concentration approximate solution as input features, and the actual ion concentration measured by high-precision ion chromatography as a label to construct a training data set; constructing a loss function based on the chemical equilibrium equation to perform integrated training on all regression models.
[0042] Among them, the loss function is constructed based on the chemical equilibrium equation, including:
[0043] L=MSE+λ 1 ·|∑(z i ·C i,predicted )-0| 2 +λ 2 ·L reg ,
[0044] Among them, L represents the loss function, MSE represents the mean square error term, and λ 1 represents the first coefficient, z i represents the charge number of ion i, C i,predicted represents the predicted concentration of ion i, λ 2 represents the second coefficient, L reg represents the regularization term.
[0045] Specifically, the regression model can use a random forest model with the following parameter settings: number of decision trees: 100, maximum tree depth: 10, minimum number of leaf node samples: 5, and the feature selection criterion can be a loss function based on the chemical equilibrium equation. The model training process uses k-fold cross validation (k=5) to evaluate performance and prevent overfitting. For each ion i, a separate regression model f is established. i :c i_optimized =f i (pH, κ, c 1 , c 2 , ..., c n ), where c i_optimized is the predicted value of ion concentration after optimization.
[0046] The specific steps are as follows: First, for data preparation, a large number of acid rain samples (n>1000) are collected, and the pH value and conductivity are measured at the same time, and the actual concentration of each ion is determined by ion chromatography. The preliminary ion concentration is calculated using the method described in the previous embodiment. The data set is randomly divided into a training set (80%) and a test set (20%). The second is feature engineering, input feature X: [pH, κ, c 1 , c 2 , ..., c n], target variable y: actual ion concentration. The next step is model training. For each ion i, a random forest regression model f is trained. i . You can use grid search and 5-fold cross validation to optimize hyperparameters. Evaluate model performance on the test set and calculate the loss function based on the chemical equilibrium equation.
[0047] In addition, in another implementation scenario, it also includes: using acid rain data from other regions or laboratories to improve the performance of the model in the target area through transfer learning methods; and / or, simultaneously predicting ion concentration and other related parameters during the training process, wherein the related parameters include acid deposition flux, to perform multi-task learning training on the model.
[0048] In step S204, the approximate solution and the predicted value are merged based on the pH value and the conductivity to obtain the concentration value of each ion in the target acid rain sample.
[0049] In one embodiment, a weighting coefficient is determined based on the pH value and the conductivity to perform weighted fusion on the approximate solution and the predicted value; wherein the weighting coefficient includes: Wherein, w represents the weighting coefficient, a and b represent the adjustment parameters, pH represents the pH value of the target acid rain sample, EC represents the conductivity of the target acid rain sample, pH 0 ,EC 0 represents the basic value; constructs the target optimization problem according to the ion non-negative constraint, charge balance constraint and conductivity constraint; transforms the target optimization problem into an equality constraint and constructs the Lagrangian function; uses the Newton method to iteratively solve the KKT condition to obtain the concentration value of each ion in the target acid rain sample.
[0050] This step aims to comprehensively utilize the approximate solution of the ion equations and the predicted value of the regression model to obtain a more accurate ion concentration estimate through data fusion technology. Specifically, first, for each ion i, there are two concentration estimates: the approximate solution of the ion equations and the predicted value c′ of the regression model i A weight needs to be assigned to the two estimates of pH and conductivity, depending on their reliability.
[0051] Define a weight function w(pH, EC) that determines the weights based on the measurement accuracy of pH and conductivity. Then, for each ion i, calculate the weighted average: c′ i , in order to ensure the physical rationality of the ion concentration, some constraints need to be applied. For example:
[0052] Non-negative constraint: All ion concentrations must be greater than or equal to zero;
[0053] Charge balance constraint: the total charge of cations should be equal to the total charge of anions;
[0054] Conductivity constraint: According to Kohlrausch's law, ion concentration and conductivity should satisfy a certain relationship;
[0055] These constraints can be implemented by optimizing the problem:
[0056]
[0057] Among them, λ i is the molar conductivity of ion i, and δ is the allowable error range.
[0058] This optimization problem can be solved by quadratic programming. For example, first convert the inequality constraints into equality constraints: Then construct the Lagrangian function: L(c, s, λ, μ) = f(c) - λ T (g(c)+s)-μ T s3), then solve the KKT condition: Next, the Newton method can be used to iteratively solve the KKT condition, and the final c i That is the concentration value of each ion in the target acid rain sample.
[0059] This fusion method makes full use of the theoretical basis of the ion equations and the data-driven characteristics of the regression model, while taking into account measurement errors and physical constraints, and can obtain more reliable ion concentration estimates. In practical applications, the weight function and constraints can be adjusted according to specific circumstances to adapt to different types of acid rain samples and measurement environments.
[0060] In addition, in order to further improve the detection accuracy, it is possible to consider introducing other relevant parameters, such as temperature, atmospheric pressure, etc., and using more complex machine learning models, such as deep neural networks or integrated learning methods. At the same time, by establishing a large-scale acid rain sample database, the model can be continuously optimized and updated to improve the applicability and accuracy of the detection chip.
[0061] The chip of the disclosed embodiment achieves high-precision detection of ion concentrations in acid rain samples by combining ion balance models, machine learning regression and data fusion technology. Compared with traditional ion chromatography, the detection chip has the advantages of low cost and high speed, and is suitable for large-scale acid rain monitoring. In practical applications, the parameters of each instruction step can be optimized and adjusted according to specific needs to obtain the best detection effect.
[0062] Figure 3 A schematic diagram of a flow chart of an instruction for approximate calculation of acid rain ion concentration provided by an embodiment of the present disclosure. Figure 3As shown, at step S301, a parameter vector including initial estimated values of each ion concentration is initialized to obtain an ion concentration matrix. That is, the initialized parameter vector β = β 0 , contains initial estimates of the concentrations of each ion.
[0063] At step S302, the ion concentration matrix is substituted into the ion equations to calculate the residual, where the residual vector r = yf(β).
[0064] In step S303, the Jacobian matrix of the ionic equations is calculated, and the incremental equation is solved in combination with the residual. The increment equation can be (J T J+λdiag(J T J))δ=J T r.
[0065] At step S304, the ion concentration matrix is cumulatively updated using the solved increments. In one embodiment, the update parameter β=β+δ.
[0066] At step S305, the ion concentration matrix is iteratively updated once or multiple times. When the absolute value of the residual is less than the first threshold or the increment is less than the second threshold, the iterative update is stopped to obtain the ion concentration matrix as an approximate solution to the ionic equations. Specifically, the above steps are repeated until convergence, and the convergence criterion is: ||δ||<τ 1 or ||r||<τ 2 , where τ 1 and τ 2 is the preset threshold.
[0067] In one embodiment, a method for obtaining the first threshold includes:
[0068]
[0069] Among them, τ 1 represents the first threshold, α represents the scaling factor, which is 0.03, β represents the parameter controlling the convergence speed, which is 0.27, n represents the current number of iterations, pH represents the pH value of the target acid rain sample, σ represents the conductivity of the target acid rain sample, σ 0 represents the standard conductivity, with a value of 1 μS / cm, γ represents the pH adjustment parameter, with a value of 0.5, δ represents the conductivity adjustment parameter, with a value of 0.25, c i Represents the concentration value of the i-th ion in the ion concentration matrix.
[0070] In another embodiment, a method for obtaining the second threshold includes:
[0071]
[0072] Among them, τ 2 represents the second threshold, ε represents the basic threshold, which is 1e-6, σ represents the conductivity of the target acid rain sample, σ 0 represents the standard conductivity, with a value of 1μS / cm, λ represents the influence factor of conductivity, with a value of 0.53, θ represents the influence factor of concentration change, with a value of 0.46, Δc i represents the change in the concentration of the i-th ion in this iteration, c i represents the concentration value of the ion in the ion concentration matrix, c 0 represents the reference concentration, which is 1 mol / L; μ represents the influencing factor of total ion concentration, which is 0.3.
[0073] In addition, it should be understood that the component detection chip disclosed in the present invention has been trained and tested on real data sets, and the detection effect has been improved by combining the mechanism of ion balance model, machine learning regression and data fusion technology. The following table 1 shows a practical application case, showing the detection process and results of the detection chip. A rainwater sample was collected in a certain area, and the pH value was measured to be 4.32 and the conductivity was 35.6μS / cm. The following results were obtained by using this detection chip to detect ion composition and concentration.
[0074] Table 1. Comparison results of real data set detection
[0075]
[0076]
[0077] It can be seen that the anion-cation balance error E = 2.3%, which is within the acceptable range (±5%). Compared with the results of ion chromatography (IC), the average relative error is 6.8%, the correlation coefficient R is 0.994, and the RMSE is 2.7 μmol / L. Regarding the linear range, for the main ions The linear range is 1-1000 μmol / L, R2>0.999, for minor ions (Ca 2+ ,Mg 2+ , Na + , K + , Cl - ), the linear range was 0.5-500μmol / L, R2>0.998.
[0078] In addition, in the inter-laboratory comparison, cross-validation was performed with 3 laboratories, and the Z scores were all within the range of ±2, indicating that the results were reliable. Standard reference material (SRM) test: NIST SRM 2694a (Simulated Rainwater) was used for testing, and the relative error RE of each ion was less than ±8%. The results show that the detection chip has high accuracy, precision and reliability, and can meet the requirements of acid rain monitoring.
[0079] In summary, the acid rain ion composition and concentration detection chip based on pH value and conductivity proposed in the present invention realizes fast, accurate and low-cost acid rain composition detection by combining mathematical models and machine learning technology. The detection chip has the following advantages:
[0080] Fast detection speed: Only by measuring pH value and conductivity can complete ion composition information be obtained, and the detection time can be shortened to 1 / 10 of the traditional method.
[0081] Low cost: No expensive equipment such as ion chromatograph is required, and the detection cost can be reduced by more than 80%.
[0082] High accuracy: Through nonlinear optimization and machine learning models, the main ion detection accuracy can be achieved within ±10%.
[0083] Wide applicability: It can be used for rapid field detection and large-scale monitoring, and is suitable for various types of precipitation samples.
[0084] It can be known from the description of the above implementation mode that those skilled in the art can clearly understand that all or part of the steps in the above-mentioned embodiment method can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present disclosure is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network communication device such as a media gateway, etc.) to execute the methods described in each embodiment of the present disclosure or some parts of the embodiments.
[0085] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.
[0086] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.
Claims
1. An acid rain component detection chip, characterized in that: The chip includes one or more processors and one or more data interfaces. The processor reads and executes the acid rain component detection instruction stored in the target storage medium through the data interface. The acid rain component detection instruction includes: Obtain the pH value and conductivity of the target acid rain sample, and construct an ion equation system based on the concentration balance and thermodynamic model of the preset ion species; The residual is calculated by bringing the initialized ion concentration matrix into the ion equation system, and the ion concentration matrix is iteratively updated once or multiple times to obtain an approximate solution to the ion equation system; Construct regression models corresponding to each ion concentration, and input data including pH value, conductivity and approximate solution into all regression models to obtain the predicted value of each ion concentration; The approximate solution and the predicted value are fused based on pH value and conductivity to obtain the concentration value of each ion in the target acid rain sample.
2. The acid rain component detection chip according to claim 1, characterized in that: in, The initial ion concentration matrix is brought into the ion equations to calculate the residual, and the ion concentration matrix is iteratively updated once or multiple times to obtain an approximate solution to the ion equations, including: Initializing a parameter vector including initial estimated values of concentrations of various ions to obtain an ion concentration matrix; Substitute the ion concentration matrix into the ion equations to calculate the residuals; Calculate the Jacobian matrix of the ionic equations and solve the incremental equations with the residuals; The ion concentration matrix is cumulatively updated using the solved increments; The ion concentration matrix is iteratively updated once or multiple times. When the absolute value of the residual is less than a first threshold or the increment is less than a second threshold, the iterative update is stopped to obtain the ion concentration matrix as an approximate solution to the ion equation system.
3. The acid rain component detection chip according to claim 2, characterized in that: The method for obtaining the first threshold comprises: Wherein, T1 represents the first threshold, α represents the scaling factor, which is 0.03, β represents the parameter for controlling the convergence speed, which is 0.27, n represents the current number of iterations, pH represents the pH value of the target acid rain sample, σ represents the conductivity of the target acid rain sample, σ0 represents the standard conductivity, which is 1 μS / cm, γ represents the pH adjustment parameter, which is 0.5, δ represents the conductivity adjustment parameter, which is 0.25, c i Represents the concentration value of the i-th ion in the ion concentration matrix.
4. The acid rain component detection chip according to claim 2, characterized in that: The method for obtaining the second threshold comprises: Among them, τ2 represents the second threshold, ε represents the basic threshold, which is 1e-6, σ represents the conductivity of the target acid rain sample, σ0 represents the standard conductivity, which is 1μS / cm, λ represents the influence factor of conductivity, which is 0.53, θ represents the influence factor of concentration change, which is 0.46, Δc i represents the change in the concentration of the i-th ion in this iteration, c i represents the concentration value of the i-th ion in the ion concentration matrix, c0 represents the reference concentration, which is 1 mol / L, and μ represents the influencing factor of the total ion concentration, which is 0.
3.
5. The acid rain component detection chip according to claim 1, characterized in that: in, Before entering data including pH, conductivity and approximate solutions into all regression models to obtain predicted values for each ion concentration, it also includes: Collect multiple acid rain samples, and use the corresponding pH value, conductivity and ion concentration approximate solution as input features, and the actual ion concentration measured by high-precision ion chromatography as labels to construct a training data set; A loss function is constructed based on the chemical equilibrium equation to perform ensemble training for all regression models.
6. The acid rain component detection chip according to claim 5, characterized in that: in, Construct a loss function based on the chemical equilibrium equation, including: L=MSE+λ1·|∑(z i ·C i,predicted )-0| 2 +λ2·L reg , Among them, L represents the loss function, MSE represents the mean square error term, λ1 represents the first coefficient, z i represents the charge number of ion i, C i,predicted represents the predicted concentration of ion i, λ2 represents the second coefficient, L reg represents the regularization term.
7. The acid rain component detection chip according to claim 5, characterized in that: Also includes: Using acid rain data from other regions or laboratories, the performance of the model in the target area can be improved through transfer learning methods; and / or Ion concentrations and other related parameters, including acid deposition flux, are simultaneously predicted during the training process to train the model for multi-task learning.
8. The acid rain component detection chip according to claim 1, characterized in that: in, The approximate solution and the predicted value are combined based on pH and conductivity to obtain the concentration value of each ion in the target acid rain sample, including: A weighting coefficient is determined based on the pH value and the conductivity to perform a weighted fusion of the approximate solution and the predicted value; wherein the weighting coefficient includes: Wherein, w represents the weighting coefficient, a and b represent the adjustment parameters, pH represents the pH value of the target acid rain sample, EC represents the conductivity of the target acid rain sample, and pH0 and EC0 represent the basic values; The target optimization problem is constructed according to the ion non-negativity constraint, charge balance constraint and conductivity constraint; Transform the target optimization problem into equality constraints and construct the Lagrangian function; The KKT conditions were iteratively solved using Newton's method to obtain the concentration value of each ion in the target acid rain sample.
Citation Information
Patent Citations
Iron precipitation process exit ion method and system thereof
CN110675918A
Spectrum technology and deep learning technology combined TOC rapid detection algorithm
CN117169166A
Acid rain observation quality evaluation method, device and equipment and storage medium
CN117931778A
Analysis method of ion concentration utilizing electric conductivity of solution, and ion concentration analyzer
JP2010096532A
Non-fixed manual shoulder orthosis for stroke patients
KR1020210001484A
