Intelligent regulation and control method and system for electrodeionization system of hydrogen conductivity meter

Abstract

The invention discloses an electrodeionization system intelligent regulation and control method for a hydrogen conductivity meter, which is used for evaluating and intelligently regulating and controlling the structure state and the working state of an electrodeionization system, and comprises the following steps: carrying out automatic working mode parameter identification on the electrodeionization system based on a random subspace identification algorithm and a grid clustering algorithm; the structural state of the electrodeionization system is evaluated and intelligently regulated and controlled based on the automatic working mode parameter recognition result; based on the fact that the structural state of the electrodeionization system is evaluated as healthy or the structural state of the electrodeionization system is evaluated as healthy after intelligent regulation and control, the working state of the electrodeionization system is intelligently regulated and controlled; the working state affects the cation removal efficiency of the electrodeionization system and is related to the resin capacity of a concentrated water chamber, the resin capacity of a fresh water chamber, the ratio of the sectional area to the height of the concentrated water chamber, the ratio of the sectional area to the height of the fresh water chamber, membrane stack voltage, membrane stack current and module inlet water flow. The invention further discloses a system, electronic equipment and a computer readable storage medium.

Description

Technical Field

[0001] The present invention relates to the technical field of electrodeionization system regulation and hydrogen conductivity measurement, and in particular to an intelligent regulation method and system of an electrodeionization system for a hydrogen conductivity meter. Background Art

[0002] The working principle of the hydrogen conductivity meter is to calculate the value of hydrogen conductivity in water based on the measured current. Hydrogen conductivity refers to the conductivity under a unit concentration gradient, which reflects the size of the hydrogen ion concentration in water. In actual use, hydrogen conductivity is widely used in water quality monitoring, environmental protection and other fields, and can reflect the process of ion exchange and the migration of chemical substances in water bodies. Therefore, the accurate determination of hydrogen conductivity is of great significance to water environment research and industrial production. Hydrogen conductivity is one of the important indicators of water quality monitoring, and its test results are used to characterize the process of ion exchange and the migration of chemical substances in water. The hydrogen conductivity electrodeionization (EDI) module cleverly integrates electrodialysis technology and ion exchange technology. Through the selective permeation of ion exchange membranes and the exchange effect of exchange resins on ions in sample water, the directional migration of ions in sample water is achieved under the action of electric fields, thereby achieving the removal of cations in sample water.

[0003] The structural health and electrical property health of the electrodeionization system directly affect the working efficiency and ion removal effect of the electrodeionization system and the related hydrogen conductivity meter. However, the existing hydrogen conductivity electrodeionization (EDI) module does not have any diagnostic and control system. Until the hydrogen conductivity electrodeionization (EDI) module fails, the technicians will not notice the reduction in its working efficiency or the inability to remove cations. Due to the lack of real-time performance, the sample water has already been contaminated at this time. In addition, the maintenance process after the failure is cumbersome and the cost is increased. It can be seen that repair is inferior to maintenance.

[0004] Therefore, it is necessary to set up an evaluation and control system for the structural status and working status of the electrodeionization system to solve the current situation where repairs can only be carried out after a fault occurs, improve the maintenance and optimization rate during use, and thus improve the working efficiency of the overall system.

[0005] Operational working mode analysis (OMA) is an important means of structural health monitoring. It identifies the natural frequency, damping ratio and working mode shape of the structure through the output dynamic signal under environmental excitation. As the research basis of vibration response analysis, health monitoring, damage detection and structural dynamics optimization design, OMA has played an important role in the field of health monitoring of mechanical components. Compared with traditional input-output working mode analysis, OMA avoids the complexity and high cost of external excitation. However, in the existing OMA process, the distinction between physical working mode and pseudo working mode, the stability screening of working mode parameters and the final confirmation of representative working mode often rely on manual judgment, which not only increases the analysis cost but also may affect the accuracy of analysis.

[0006] In recent years, with the development of automation technology, automated operation mode analysis (AOMA) has gradually become a research hotspot, among which the stochastic subspace identification (SSI) algorithm has been widely used. However, the degree of automation of AOMA is still limited, especially in the automatic elimination of pseudo-operating modes, the identification of stable pole sequences, and the parameter screening under complex excitation conditions. Therefore, there is an urgent need for a more efficient, stable and intelligent automated analysis method to further improve the efficiency and reliability of complex structure operating mode analysis. Summary of the invention

[0007] The object of the present invention is to provide an intelligent control method and system for an electrodeionization system for a hydrogen conductivity meter, which automatically identifies working mode parameters of the electrodeionization system based on a random subspace recognition algorithm and a grid clustering algorithm, and evaluates and intelligently controls the structural state of the electrodeionization system based on the automatic working mode parameter recognition result; based on the structural state of the electrodeionization system being evaluated as healthy or the structural state of the electrodeionization system being evaluated as healthy after intelligent control, the working state of the electrodeionization system is intelligently controlled; the working state affects the cation removal efficiency of the electrodeionization system, and the working state is related to the resin capacity of the concentrated water chamber, the resin capacity of the fresh water chamber, the cross-sectional area to height ratio of the concentrated water chamber, the cross-sectional area to height ratio of the fresh water chamber, the membrane stack voltage, the membrane stack current and the module inlet water flow rate of the electrodeionization system. The premise of intelligent control of the structural state is to automatically identify the working mode parameters of the electrodeionization system based on the random subspace recognition algorithm and the grid clustering algorithm. The constructed automatic control framework can effectively solve the key problems such as working mode separation, parameter selection and pseudo working mode screening, and is suitable for working state health monitoring and working mode analysis of complex structure electrodeionization systems. Starting from the basic principles of stability graph, binary K-means clustering and grid clustering, the automatic algorithm framework defines the improved working mode distance based on MACXP to control the aliasing problem between spatially dense working modes; then construct a new k-dist graph, and finally automatically determine the key parameters of grid clustering through binary K-means clustering; binary K-means clustering is performed on the candidate working modes according to the normalized clustering dimension and working mode energy to realize the automatic selection of complex working modes or weakly excited working modes; the synergistic effect of MOF and MAC is used to intelligently control the working mode splitting phenomenon. The premise of intelligent control of the working state is to determine the optimal parameters after quantitative / non-quantitative experiments on each parameter.

[0008] The first aspect of the present invention is to provide an intelligent control method for an electrodeionization system for a hydrogen conductivity meter, which is used to evaluate and intelligently control the structural state and working state of the electrodeionization system, comprising:

[0009] S1, automatically identifying working mode parameters of the electrodeionization system based on a random subspace recognition algorithm and a grid clustering algorithm, and evaluating and intelligently regulating the structural state of the electrodeionization system based on the automatic working mode parameter recognition result;

[0010] S2, based on the structural state of the electrodeionization system being evaluated as healthy or the structural state of the electrodeionization system being evaluated as healthy after intelligent regulation, the working state of the electrodeionization system is intelligently regulated; the working state affects the cation removal efficiency of the electrodeionization system, and the working state is related to the resin capacity of the concentrate chamber, the resin capacity of the fresh water chamber, the ratio of the cross-sectional area to the height of the concentrate chamber, the ratio of the cross-sectional area to the height of the fresh water chamber, the membrane stack voltage, the membrane stack current and the module inlet water flow rate of the electrodeionization system.

[0011] Preferably, the S1 comprises:

[0012] S11, arranging a structural sensor group on the electrodeionization system, wherein the structural sensor group includes a MEMS displacement sensor, a MEMS velocity sensor, and a MEMS acceleration sensor, and the structural sensor group is used to obtain a structural dynamic response;

[0013] S12, the working mode of the structural dynamic response is calculated based on the covariance random subspace method in the time domain to obtain the original stability diagram;

[0014] S13, cleaning the original stability map based on a cleaning criterion to obtain a stability map after preliminary cleaning, calculating the average regularized power spectrum density in the frequency domain of each degree of freedom according to the structural dynamic response signal, and using the average regularized power spectrum density as a reference to overlay and plot on the stability map after preliminary cleaning;

[0015] S14, using all the poles in the stability map after the preliminary cleaning as input samples of the grid clustering, and determining the number of minimum clustering units of the grid clustering according to the total number of samples;

[0016] S15, calculating the normalized working mode distance between any two adjacent clustering units in the stability graph after the preliminary cleaning based on the frequency, damping ratio and working mode vibration type as independent variables for working mode distance calculation;

[0017] S16, calculating the working mode distance between the cluster unit where each pole is located and its kth nearest neighbor cluster unit, determining the working mode distance distribution based on the working mode distance, and determining the neighborhood radius Eps value according to the working mode distance distribution;

[0018] S17, substituting the obtained minimum number of neighborhood samples, neighborhood radius, and normalized working mode distance into the grid clustering algorithm to obtain a stable graph with noise removed;

[0019] S18, Linearly normalized working mode energy MEL and cluster dimension N cn, forming multiple groups of two-dimensional input samples, substituting the multiple groups of two-dimensional input samples into binary K-means clustering, and automatically screening the clusters where the candidate physical working modes are located;

[0020] S19, effectively evaluating whether the structure of the electrodeionization system has a working mode splitting phenomenon according to the synergistic effect of the working mode confidence factor MAC and the working mode overlap factor MOF, wherein the working mode confidence factor MAC is evaluated from the perspective of vibration mode coefficient, and the working mode overlap factor MOF is evaluated from the perspective of frequency and damping ratio;

[0021] S110, extracting a representative working mode indexed by the median value of the damping ratio from all the candidate working modes and identifying working mode parameters based on the representative working mode, and determining the structural state of the electrodeionization system based on the identified working mode parameters, and performing the intelligent regulation if there is a problem with the structural state; wherein the working mode parameter identification includes frequency, damping ratio, linear normalized working mode energy MEL and cluster dimension N cn .

[0022] Preferably, the S12 includes:

[0023] S121, construct the Toeplitz matrix, including:

[0024] Assuming that the output data is ergodic, the output covariance matrix R i And according to the covariance matrix R i Constructed Toeplitz matrix T 1|i Respectively expressed as:

[0025]

[0026] In the formula, represents the observation vector of the N-DOF system at discrete time k, that is, the measurement signal time history can be one or more of acceleration, velocity and displacement, where i is the number of test channels; j corresponds to the total length of the measurement data; l represents the comprehensive structural parameter dimension of the electrodeionization system, which can be selected from length, width, height or aspect ratio, aspect ratio, length-to-height ratio, that is, the dimension of l is one-dimensional, two-dimensional to six-dimensional, and the value of l is 1, 2, 3, 4, 5 or 6;

[0027]

[0028] In the formula, parameter i represents the number of rows of the Toeplitz matrix, and the matrix T 1|i The dimension is Smaller than the dimension of the original Hankel matrix This reduces the amount of computation and memory requirements; similarly, T 2|i+1It is expressed as:

[0029]

[0030] S122, performs singular value decomposition on the Toeplitz matrix, including:

[0031] The model order N is another parameter that needs to be defined by the user, by ignoring small singular values, which are equal to the small singular values ​​obtained from the Toeplitz matrix T 1|i The number of singular values ​​obtained by the singular value decomposition of (Equation (5));

[0032]

[0033] In the formula, and is an orthogonal matrix, is a diagonal matrix; is a non-singular matrix, usually taken as the unit matrix, that is, T = I, we can get:

[0034]

[0035] In the formula, and They are the extended observable matrix and the extended controllable matrix respectively;

[0036] S123, identifying a system matrix of the electrodeionization system, including:

[0037] The expression of the system matrix A is:

[0038] A=S1 -1 / 2 U1 T T 2|i+1 V1S1 -1 / 2 (8);

[0039] The working mode parameters of the discrete time system are obtained from the system matrix A and the output matrix C, where the output matrix C is equal to O in equation (6): i The first l lines of ;

[0040] S124, determining working mode parameters, including:

[0041] Decompose the system matrix A through the eigenvalue to obtain:

[0042] A=ψΛψ -1 ,Aψ i =λ i ψ i (9);

[0043] In formula (9), is the system pole λ of the discrete-time system iThe diagonal matrix composed of i corresponds to the right eigenvector of the matrix A; while the continuous-time eigenvalue λ ci and the eigenvalue λ in the discrete state i The following relations are satisfied:

[0044] λ ci =lnλ i / Δt (10);

[0045] where Δt represents the sampling time interval; finally, the natural frequency f of the system is i , damping ratio ξ i and vibration mode φ i It can be determined by formula (11) as:

[0046] f i =|λ ci | / 2π,ξ i = -Re(λ ci ) / |λ ci |,φ i =Cψ i (11);

[0047] S125, parameter settings, including:

[0048] In the matrix T 1|i When performing singular value decomposition, the matrix T 1|i The number of row blocks i and the model order N must satisfy the following relationship in the algorithm:

[0049] li ≥ N (12);

[0050] According to the fundamental frequency f0 and sampling frequency f of the structure s , the lower limit of i determined based on the empirical rule is:

[0051]

[0052] Preferably, the S13 includes:

[0053] S131, the candidate working modes that do not simultaneously satisfy the cleaning criteria in formula (14) are eliminated as mathematical false working modes:

[0054]

[0055] In the formula, λ i is the eigenvalue obtained by eigenvalue decomposition of the system matrix A in the discrete state, is j conjugation of;

[0056] S132, calculate the average regularized power spectral density ANPSD, which is calculated as shown in the following formula (15):

[0057]

[0058] Where m is the number of degrees of freedom of the test; PSD i is the PSD of the ith test degree of freedom signal; f k is the kth discrete frequency; n is the total number of discrete frequencies.

[0059] Preferably, the S15 includes:

[0060] Based on the frequency, damping ratio and working mode vibration type as independent variables for working mode distance calculation, the normalized working mode distance d(j, k) between any two adjacent clustering units in the stability graph after preliminary cleaning is calculated; the calculation formula is as follows:

[0061]

[0062] In the formula, φ j refers to the jth vibration mode vector, λ cj Refers to the eigenvalue of the jth continuous state space equation, and its calculation expression is formula (17):

[0063]

[0064] The working mode vibration confidence criterion MAC of the traditional real working mode space is extended to the complex working mode space MACXP (φ j ,φ k ), the calculation formula is shown in formula (18):

[0065]

[0066] In formula (18), is λ cj The conjugate of Yes j The transpose of Yes j The conjugate transpose of .

[0067] Preferably, the S16 includes:

[0068] S161, calculating the k-th working mode distance between the cluster unit where each data point is located and its k-th nearest neighbor cluster unit based on the working mode distance d(j,k) between the cluster unit where each pole is located and its k-th nearest neighbor cluster unit in the stability graph after the preliminary cleaning;

[0069] S162, constructing a k-dist function graph using binary K-means clustering, wherein the k-dist function graph is used for adaptive determination of parameters; wherein the basic calculation formula of the binary K-means clustering is shown in the following formula (19):

[0070]

[0071] In formula (19), C1 and C2 represent the potential physical working mode and false working mode, respectively, where dX represents the working mode distance of the one-dimensional working mode from the k-th neighbor of the data set calculated according to formula (16), μ k represents the initial ideal cluster center that can be objectively determined, that is, the initial cluster center representing the k-th working mode distance of the physical working mode and the false working mode is 0 and 1 respectively.

[0072] Preferably, the S18 includes:

[0073] S181, the normalized working mode energy MEL and cluster dimension N cn Constructing a two-dimensional input sample data set and objectively determining the initial clustering center of the two-dimensional input sample data set, that is, the initial clustering centers of the two-dimensional samples representing the physical working mode and the false working mode are vectors {1,1} and {0,0} respectively;

[0074] The calculation formula of MEL is shown in formula (20):

[0075]

[0076] In formula (20), ψ i is the matrix A(Aψ i =λ i ψ i )’s i-th right eigenvector; (ψ -1 ) i is the matrix ψ -1 The i-th row of i Refers to the system matrix A in the continuous state space c The i-th eigenvalue of c Δt);() * is a conjugate operation; matrices C and G are the output matrix and the output covariance matrix of the next state, respectively, where matrix C is the formula O i The first l rows of the matrix G can be expanded by the controllable matrix Γ i The last l columns are determined;

[0077] S182, retaining candidate polar axes with higher cluster dimensions and working mode energies based on the iterative optimization process of the binary K-means clustering, thereby automatically screening the clusters where the candidate physical working modes are located.

[0078] Preferably, the S19 includes:

[0079] S191, determining to use the working mode confidence factor MAC and the working mode overlap factor MOF as indicators for quantifying the degree of overlap between any two working modes, wherein the working mode confidence factor MAC is quantified from the perspective of vibration mode coefficient, and the working mode overlap factor MOF is quantified from the perspective of frequency and damping ratio;

[0080] The calculation formulas of the working mode confidence factor MAC and the working mode overlap factor MOF are respectively shown in the following formulas (21) and (22):

[0081]

[0082] In formula (21), H represents the Hermitian transpose, φ i and φ j Represent the vibration modes of the i-th and j-th order working modes respectively;

[0083]

[0084] In formula (22), f i and f j Represent the frequencies of a pair of working modes to be quantified, ξ i Refers to the damping ratio of the i-th order working mode;

[0085] S192, if the working mode overlap factor MOF of any pair of working modes exceeds 0.1 and the working mode confidence factor MAC exceeds 0.9, the two are automatically identified as belonging to the same order working mode. At this time, the candidate working mode cluster with the smaller cluster dimension will be eliminated.

[0086] Preferably, S2 includes:

[0087] S21, based on the current parameters of the electrodeionization system and the application scenario of the hydrogen conductivity meter, the resin capacity of the concentrated water chamber, the resin capacity of the deionization chamber, the ratio of the cross-sectional area to the height of the concentrated water chamber, and the ratio of the cross-sectional area to the height of the deionization chamber are set as constants; the membrane stack voltage, the membrane stack current, and the module inlet water flow rate are set as variables;

[0088] S22, controlling the water flow rate at the module inlet to remain unchanged, gradually increasing the membrane stack voltage and continuously monitoring the change of the membrane stack current, determining the mutation point of the membrane stack current as the current maximum point based on the change of the membrane stack current, and determining the optimal membrane stack voltage range based on the change of the membrane stack current;

[0089] S23, based on the determined current maximum point and the optimal membrane stack voltage range, controlling the membrane stack current by a constant current mode during the operation of the electrodeionization system, and judging whether the membrane stack voltage is within the optimal membrane stack voltage range; when the membrane stack voltage is not within the optimal membrane stack voltage range, executing step S231; when the membrane stack voltage is within the optimal membrane stack voltage range, executing step S232;

[0090] S231, controlling the membrane stack voltage within the initial optimal membrane stack voltage range by a constant voltage method, and determining the optimal membrane stack current by adjusting the module inlet water flow rate; then making the electrodeionization system constant at the optimal membrane stack current value, and making the module inlet water flow rate within the flow threshold range by adjusting the membrane stack voltage;

[0091] S232, determining that the membrane stack voltage at this time is the working membrane stack voltage of the electrodeionization system, and the corresponding membrane stack current is the working membrane stack current of the electrodeionization system, and continuing to determine the water inlet flow rate at the working module inlet; making the electrodeionization system constant at the working membrane stack voltage or the working membrane stack current, and making the water inlet flow rate at the module inlet be within the range of the flow threshold by adjusting the membrane stack current or the membrane stack voltage, and the water inlet flow rate at the module inlet at this time is the water inlet flow rate at the working module inlet of the electrodeionization system;

[0092] S24, determining a maintenance period and a diagnosis suggestion for the electrodeionization system based on the determined working membrane stack voltage, working membrane stack current and the working module inlet water flow rate and the percentage of the flow threshold.

[0093] Preferably, the method further comprises:

[0094] S25, determining an inlet conductivity range based on the determined working membrane stack voltage, working membrane stack current and working module inlet water flow.

[0095] The second aspect of the present invention is to provide an intelligent control system for an electrodeionization system for a hydrogen conductivity meter, which is used to implement the method of the first aspect, evaluate and intelligently control the structural state and working state of the electrodeionization system, including:

[0096] A structural state evaluation and intelligent control module (101) is used to automatically identify working mode parameters of the electrodeionization system based on a random subspace identification algorithm and a grid clustering algorithm, and to evaluate and intelligently control the structural state of the electrodeionization system based on the automatic working mode parameter identification result;

[0097] A working state evaluation and intelligent control module (102) is used to intelligently control the working state of the electrodeionization system based on the structural state of the electrodeionization system being evaluated as healthy or the structural state of the electrodeionization system being evaluated as healthy after intelligent control; the working state affects the cation removal efficiency of the electrodeionization system, and the working state is related to the resin capacity of the concentrated water chamber, the resin capacity of the fresh water chamber, the ratio of the cross-sectional area to the height of the concentrated water chamber, the ratio of the cross-sectional area to the height of the fresh water chamber, the membrane stack voltage, the membrane stack current and the water flow rate at the module inlet of the electrodeionization system.

[0098] A third aspect of the present invention provides an electronic device, comprising a processor and a memory, wherein the memory stores a plurality of instructions, and the processor is configured to read the instructions and execute the method described in the first aspect.

[0099] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a plurality of instructions, and the plurality of instructions can be read by a processor and execute the method described in the first aspect.

[0100] Beneficial effects of the method and system of the present invention:

[0101] (1) The structural health monitoring is organically combined with the working status monitoring to ensure that the failure rate of the electrodeionization system is greatly reduced, and the efficiency of cation removal in the sample water is improved, and the measurement results of hydrogen conductivity are more accurate.

[0102] (2) The present invention performs binary K-means clustering on the candidate working modes according to the normalized clustering dimension and the working mode energy MEL, which can effectively deal with the situation where the candidate working mode clustering "only contains physical working modes", thereby avoiding the need to manually define a certain number of empty sets to realize automatic screening of working modes in this situation.

[0103] (3) The present invention can effectively distinguish the influence of spatially dense working modes; through the synergistic effect of MOF and MAC, the working mode splitting phenomenon can be accurately identified.

[0104] (4) The present invention realizes the accurate identification of physical working mode and false working mode, and provides a more efficient, stable and intelligent automated analysis method. It has significant advantages in improving the efficiency, accuracy and applicability of structural operation working mode analysis, and provides strong technical support for the field of structural health monitoring.

[0105] (5) Through the variable / invariant alternating verification method, the most accurate control quantity for the parameters is provided, so that by changing the various parameter variables that affect the cation removal efficiency, the membrane stack voltage and membrane stack current of the electrodeionization module and the module inlet water flow rate are intelligently adjusted to determine and adjust to the optimal working state of the module. BRIEF DESCRIPTION OF THE DRAWINGS

[0106] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the related technologies, the drawings required for use in the specific embodiments or the related technical descriptions will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0107] Figure 1 A flow chart of an intelligent control method for an electrodeionization system of a hydrogen conductivity meter provided according to an embodiment of the present invention;

[0108] Figure 2 This is a flow chart of step S1 of an intelligent control method for an electrodeionization system of a hydrogen conductivity meter provided according to an embodiment of the present invention;

[0109] Figure 3 This is a flow chart of step S2 of the intelligent control method for an electrodeionization system of a hydrogen conductivity meter provided according to an embodiment of the present invention;

[0110] Figure 4 This is a diagram of the architecture of an intelligent control system for an electrodeionization system for a hydrogen conductivity meter provided according to an embodiment of the present invention;

[0111] Figure 5 A structural diagram of an electronic device provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0112] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0113] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0114] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0115] Embodiment 1

[0116] like Figure 1 As shown, this embodiment provides an intelligent control method for an electrodeionization system for a hydrogen conductivity meter, which is used to evaluate and intelligently control the structural state and working state of the electrodeionization system, including:

[0117] S1, automatically identifying working mode parameters of the electrodeionization system based on a random subspace recognition algorithm and a grid clustering algorithm, and evaluating and intelligently regulating the structural state of the electrodeionization system based on the automatic working mode parameter recognition result;

[0118] S2, based on the structural state of the electrodeionization system being evaluated as healthy or the structural state of the electrodeionization system being evaluated as healthy after intelligent regulation, the working state of the electrodeionization system is intelligently regulated; the working state affects the cation removal efficiency of the electrodeionization system, and the working state is related to the resin capacity of the concentrate chamber, the resin capacity of the fresh water chamber, the ratio of the cross-sectional area to the height of the concentrate chamber, the ratio of the cross-sectional area to the height of the fresh water chamber, the membrane stack voltage, the membrane stack current and the module inlet water flow rate of the electrodeionization system.

[0119] like Figure 2 As shown, as a preferred embodiment, the S1 includes:

[0120] S11, arranging a structural sensor group on the electrodeionization system, wherein the structural sensor group includes a MEMS displacement sensor, a MEMS velocity sensor, and a MEMS acceleration sensor, and the structural sensor group is used to obtain a structural dynamic response;

[0121] S12, the working mode of the structural dynamic response is calculated based on the covariance random subspace method in the time domain to obtain the original stability diagram;

[0122] S13, cleaning the original stability map based on a cleaning criterion to obtain a stability map after preliminary cleaning, calculating the average regularized power spectrum density in the frequency domain of each degree of freedom according to the structural dynamic response signal, and using the average regularized power spectrum density as a reference to overlay and plot on the stability map after preliminary cleaning;

[0123] S14, using all the poles in the stability map after the preliminary cleaning as input samples of the grid clustering, and determining the number of minimum clustering units of the grid clustering according to the total number of samples;

[0124] In this embodiment, grid-based clustering is to divide the data space into a limited number of units and then perform clustering on these units, which is suitable for processing high-dimensional data and is therefore suitable for the application scenario of the present invention.

[0125] S15, calculating the normalized working mode distance between any two adjacent clustering units in the stability graph after the preliminary cleaning based on the frequency, damping ratio and working mode vibration type as independent variables for working mode distance calculation;

[0126] S16, calculating the working mode distance between the cluster unit where each pole is located and its kth nearest neighbor cluster unit, determining the working mode distance distribution based on the working mode distance, and determining the neighborhood radius Eps value according to the working mode distance distribution;

[0127] S17, substituting the obtained minimum number of neighborhood samples, neighborhood radius, and normalized working mode distance into the grid clustering algorithm to obtain a stable graph with noise removed;

[0128] S18, Linearly normalized working mode energy MEL and cluster dimension N cn , forming multiple groups of two-dimensional input samples, substituting the multiple groups of two-dimensional input samples into binary K-means clustering, and automatically screening the clusters where the candidate physical working modes are located;

[0129] S19, effectively evaluating whether the structure of the electrodeionization system has a working mode splitting phenomenon according to the synergistic effect of the working mode confidence factor MAC and the working mode overlap factor MOF, wherein the working mode confidence factor MAC is evaluated from the perspective of vibration mode coefficient, and the working mode overlap factor MOF is evaluated from the perspective of frequency and damping ratio;

[0130] S110, extracting a representative working mode indexed by the median value of the damping ratio from all the candidate working modes and identifying working mode parameters based on the representative working mode, and determining the structural state of the electrodeionization system based on the identified working mode parameters, and performing the intelligent regulation if there is a problem with the structural state; wherein the working mode parameter identification includes frequency, damping ratio, linear normalized working mode energy MEL and cluster dimension N cn .

[0131] As a preferred implementation, the S12 includes:

[0132] S121, construct the Toeplitz matrix, including:

[0133] Assuming that the output data is ergodic, the output covariance matrix R i And according to the covariance matrix R i Constructed Toeplitz matrix T 1|i Respectively expressed as:

[0134]

[0135] In the formula, Represents the observation vector of the N-DOF system at discrete time k, that is, the measurement signal time history can be one or more of acceleration, velocity and displacement, wherein i is the number of test channels; j corresponds to the total length of the measurement data; l represents the comprehensive structural parameter dimension of the electrodeionization system; the length, width, height or aspect ratio, aspect ratio, length-to-height ratio or other parameters can be selected, all within the protection scope of the present invention, that is, the dimension of l is one-dimensional, two-dimensional to six-dimensional, and the value of l is 1, 2, 3, 4, 5 or 6;

[0136]

[0137] In the formula, parameter i represents the number of rows of the Toeplitz matrix, and the matrix T 1|i The dimension is Smaller than the dimension of the original Hankel matrix This reduces the amount of computation and memory requirements; similarly, T 2|i+1 It is expressed as:

[0138]

[0139] S122, performs singular value decomposition on the Toeplitz matrix, including:

[0140] The model order N is another parameter that needs to be defined by the user, by ignoring small singular values, which are equal to the small singular values ​​obtained from the Toeplitz matrix T 1|i The number of singular values ​​obtained by the singular value decomposition of (Equation (5));

[0141]

[0142] In the formula, and is an orthogonal matrix, is a diagonal matrix; is a non-singular matrix, usually taken as the unit matrix, that is, T = I, we can get:

[0143]

[0144] In the formula, and They are the extended observable matrix and the extended controllable matrix respectively;

[0145] S123, identifying a system matrix of the electrodeionization system, including:

[0146] The expression of the system matrix A is:

[0147] A=S1 -1 / 2 U1 T T 2|i+1 B1S1 -1 / 2 (8);

[0148] The working mode parameters of the discrete time system are obtained from the system matrix A and the output matrix C, where the output matrix C is equal to O in equation (6): i The first l lines of ;

[0149] S124, determining working mode parameters, including:

[0150] Decompose the system matrix A through the eigenvalue to obtain:

[0151] A=ψΛψ -1 ,Aψ i =λ i ψ i (9);

[0152] In formula (9), is the system pole λ of the discrete-time system i The diagonal matrix composed of i corresponds to the right eigenvector of the matrix A; while the continuous-time eigenvalue λ ciand the eigenvalue λ in the discrete state i The following relations are satisfied:

[0153] λ ci =lnλ i / Δt (10);

[0154] where Δt represents the sampling time interval; finally, the natural frequency f of the system is i , damping ratio ξ i and vibration mode φ i It can be determined by formula (11) as:

[0155] f i =|λ ci | / 2π,ξ i = -Re(λ ci ) / |λ ci |,φ i =Cψ i (11);

[0156] S125, parameter settings, including:

[0157] In the matrix T 1|i When performing singular value decomposition, the matrix T 1|i The number of row blocks i and the model order N must satisfy the following relationship in the algorithm:

[0158] li ≥ N (12);

[0159] According to the fundamental frequency f0 and sampling frequency f of the structure s , the lower limit of i determined based on the empirical rule is:

[0160]

[0161] As a preferred implementation, the S13 includes:

[0162] S131, the candidate working modes that do not simultaneously satisfy the cleaning criteria in formula (14) are eliminated as mathematical false working modes:

[0163]

[0164] In the formula, λ i is the eigenvalue obtained by eigenvalue decomposition of the system matrix A in the discrete state, is j conjugation of;

[0165] S132, calculate the average regularized power spectral density ANPSD, which is calculated as shown in the following formula (15):

[0166]

[0167] Where m is the number of degrees of freedom of the test; PSD i is the PSD of the ith test degree of freedom signal; f k is the kth discrete frequency; n is the total number of discrete frequencies.

[0168] As a preferred implementation, the S15 includes:

[0169] Based on the frequency, damping ratio and working mode vibration type as independent variables for working mode distance calculation, the normalized working mode distance d(j, k) between any two adjacent clustering units in the stability graph after preliminary cleaning is calculated; the calculation formula is as follows:

[0170]

[0171] In the formula, φ j refers to the jth vibration mode vector, λ cj Refers to the eigenvalue of the jth continuous state space equation, and its calculation expression is formula (17):

[0172]

[0173] The working mode vibration confidence criterion MAC of the traditional real working mode space is extended to the complex working mode space MACXP (φ j ,φ k ), the calculation formula is shown in formula (18):

[0174]

[0175] In formula (18), is λ cj The conjugate of Yes j The transpose of Yes j The conjugate transpose of .

[0176] As a preferred implementation, the S16 includes:

[0177] S161, calculating the k-th working mode distance between the cluster unit where each data point is located and its k-th nearest neighbor cluster unit based on the working mode distance d(j,k) between the cluster unit where each pole is located and its k-th nearest neighbor cluster unit in the stability graph after the preliminary cleaning;

[0178] S162, constructing a k-dist function graph using binary K-means clustering, wherein the k-dist function graph is used for adaptive determination of parameters; wherein the basic calculation formula of the binary K-means clustering is shown in the following formula (19):

[0179]

[0180] In formula (19), C1 and C2 represent the potential physical working mode and false working mode, respectively, where dX represents the working mode distance of the one-dimensional working mode from the k-th neighbor of the data set calculated according to formula (16), μ k represents the initial ideal cluster center that can be objectively determined, that is, the initial cluster center representing the k-th working mode distance of the physical working mode and the false working mode is 0 and 1 respectively.

[0181] As a preferred implementation, the S18 includes:

[0182] S181, the normalized working mode energy MEL and cluster dimension N cn Constructing a two-dimensional input sample data set and objectively determining the initial clustering center of the two-dimensional input sample data set, that is, the initial clustering centers of the two-dimensional samples representing the physical working mode and the false working mode are vectors {1,1} and {0,0} respectively;

[0183] The calculation formula of MEL is shown in formula (20):

[0184]

[0185] In formula (20), ψ i is the matrix A(Aψ i =λ i ψ i )’s i-th right eigenvector; (ψ -1 ) i is the matrix ψ -1 The i-th row of i Refers to the system matrix A in the continuous state space c The i-th eigenvalue of c Δt);() * is a conjugate operation; matrices C and G are the output matrix and the output covariance matrix of the next state, respectively, where matrix C is the formula O i The first l rows of the matrix G can be expanded by the controllable matrix Γ i The last l columns are determined;

[0186] S182, retaining candidate polar axes with higher cluster dimensions and working mode energies based on the iterative optimization process of the binary K-means clustering, thereby automatically screening the clusters where the candidate physical working modes are located.

[0187] As a preferred implementation, the S19 includes:

[0188] S191, determining to use the working mode confidence factor MAC and the working mode overlap factor MOF as indicators for quantifying the degree of overlap between any two working modes, wherein the working mode confidence factor MAC is quantified from the perspective of vibration mode coefficient, and the working mode overlap factor MOF is quantified from the perspective of frequency and damping ratio;

[0189] The calculation formulas of the working mode confidence factor MAC and the working mode overlap factor MOF are respectively shown in the following formulas (21) and (22):

[0190]

[0191] In formula (21), H represents the Hermitian transpose, φ i and φ j Represent the vibration modes of the i-th and j-th order working modes respectively;

[0192]

[0193] In formula (22), f i and f j Represent the frequencies of a pair of working modes to be quantified, ξ i Refers to the damping ratio of the i-th order working mode;

[0194] S192, if the working mode overlap factor MOF of any pair of working modes exceeds 0.1 and the working mode confidence factor MAC exceeds 0.9, the two are automatically identified as belonging to the same order working mode. At this time, the candidate working mode cluster with the smaller cluster dimension will be eliminated.

[0195] In this embodiment, the technical purpose of implementing S2 is to intelligently adjust the membrane stack voltage and membrane stack current of the electrodeionization module and the module inlet water flow rate by changing various parameter variables that affect the cation removal efficiency to determine and adjust to the optimal working state of the module.

[0196] The principle is:

[0197] (1) Membrane stack voltage: When the flow rate is constant, if the membrane stack voltage is too low, it will not generate a large enough driving force to migrate ions from the fresh water chamber to the concentrated water chamber, affecting the quality of the fresh water outlet, that is, affecting the cation removal efficiency; if the membrane stack voltage is too high, the excess potential difference will cause more water to dissociate, resulting in a decrease in the current efficiency of the membrane stack and a significant change in the water environment inside the membrane stack, thereby affecting the cation removal efficiency.

[0198] (2) Water flow rate at the inlet of the working module: When the voltage of the membrane stack is constant, the water flow rate at the inlet of the working module is increased, and more ions in the fresh water chamber will participate in the load current. In addition, the turbulence of the solution in the fresh water chamber is enhanced, the diffusion boundary layer thickness on the surface of the cation resin and the cation exchange membrane is reduced, the ion transfer resistance is reduced, and the membrane stack resistance is relatively small. At this time, the membrane stack current increases; when the water flow rate at the inlet of the working module continues to increase and gradually exceeds the maximum ion exchange capacity in the fresh water chamber, some cations will penetrate the fresh water chamber, and the cation removal efficiency will decrease; the water flow rate at the inlet of the working module cannot be too low. When the water flow rate at the inlet of the working module is too low, the membrane stack current is large, the degree of hydrolysis is severe, and scaling tends to appear inside the membrane stack.

[0199] like Figure 3 As shown, as a preferred embodiment, the S2 includes:

[0200] S21, based on the current parameters of the electrodeionization system and the application scenario of the hydrogen conductivity meter, the resin capacity of the concentrated water chamber, the resin capacity of the deionization chamber, the ratio of the cross-sectional area to the height of the concentrated water chamber, and the ratio of the cross-sectional area to the height of the deionization chamber are set as constants; the membrane stack voltage, the membrane stack current, and the module inlet water flow rate are set as variables;

[0201] S22, controlling the water flow rate at the module inlet to remain unchanged, gradually increasing the membrane stack voltage and continuously monitoring the change of the membrane stack current, determining the mutation point of the membrane stack current as the current maximum point based on the change of the membrane stack current, and determining the optimal membrane stack voltage range based on the change of the membrane stack current;

[0202] S23, based on the determined current maximum point and the optimal membrane stack voltage range, controlling the membrane stack current by a constant current mode during the operation of the electrodeionization system, and judging whether the membrane stack voltage is within the optimal membrane stack voltage range; when the membrane stack voltage is not within the optimal membrane stack voltage range, executing step S231; when the membrane stack voltage is within the optimal membrane stack voltage range, executing step S232;

[0203] S231, controlling the membrane stack voltage within the initial optimal membrane stack voltage range by a constant voltage method, and determining the optimal membrane stack current by adjusting the module inlet water flow rate; then making the electrodeionization system constant at the optimal membrane stack current value, and making the module inlet water flow rate within the flow threshold range by adjusting the membrane stack voltage;

[0204] S232, determining that the membrane stack voltage at this time is the working membrane stack voltage of the electrodeionization system, and the corresponding membrane stack current is the working membrane stack current of the electrodeionization system, and continuing to determine the water inlet flow rate at the working module inlet; making the electrodeionization system constant at the working membrane stack voltage or the working membrane stack current, and making the water inlet flow rate at the module inlet be within the range of the flow threshold by adjusting the membrane stack current or the membrane stack voltage, and the water inlet flow rate at the module inlet at this time is the water inlet flow rate at the working module inlet of the electrodeionization system;

[0205] S24, determining a maintenance period and a diagnosis suggestion for the electrodeionization system based on the determined working membrane stack voltage, working membrane stack current and the working module inlet water flow rate and the percentage of the flow threshold.

[0206] As a preferred embodiment, it also includes:

[0207] S25, determining an inlet conductivity range based on the determined working membrane stack voltage, working membrane stack current and working module inlet water flow.

[0208] Table 1 below shows a template for recording data.

[0209] Table 1

[0210] Concentrate chamber resin capacity Fresh water chamber resin capacity Two-chamber cross-sectional area Height ratio between two rooms Membrane stack voltage range Membrane stack current range flow Inlet conductivity range

[0211] Embodiment 2

[0212] like Figure 4 As shown, this embodiment provides an intelligent control system for an electrodeionization system for a hydrogen conductivity meter, which is used to evaluate and intelligently control the structural state and working state of the electrodeionization system, including:

[0213] A structural state evaluation and intelligent control module 101 is used to automatically identify working mode parameters of the electrodeionization system based on a random subspace recognition algorithm and a grid clustering algorithm, and to evaluate and intelligently control the structural state of the electrodeionization system based on the automatic working mode parameter recognition result;

[0214] The working state evaluation and intelligent control module 102 is used to intelligently control the working state of the electrodeionization system based on the structural state of the electrodeionization system being evaluated as healthy or the structural state of the electrodeionization system being evaluated as healthy after intelligent control; the working state affects the cation removal efficiency of the electrodeionization system, and the working state is related to the resin capacity of the concentrated water chamber, the resin capacity of the fresh water chamber, the ratio of the cross-sectional area to the height of the concentrated water chamber, the ratio of the cross-sectional area to the height of the fresh water chamber, the membrane stack voltage, the membrane stack current and the module inlet water flow rate of the electrodeionization system.

[0215] The present invention also provides a memory storing a plurality of instructions, wherein the instructions are used to implement the method as in the first embodiment.

[0216] like Figure 5 As shown, the present invention also provides an electronic device, including a processor 301 and a memory 302 connected to the processor 301, the memory 302 stores a plurality of instructions, and the instructions can be loaded and executed by the processor so that the processor can execute the method as in the first embodiment.

[0217] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent control method for an electrodeionization system of a hydrogen conductivity meter, used for evaluating and intelligently controlling the structural state and working state of the electrodeionization system, characterized in that: include: S1, automatically identifying working mode parameters of the electrodeionization system based on a random subspace recognition algorithm and a grid clustering algorithm, and evaluating and intelligently regulating the structural state of the electrodeionization system based on the automatic working mode parameter recognition result; S2, based on the structural state of the electrodeionization system being evaluated as healthy or the structural state of the electrodeionization system being evaluated as healthy after intelligent regulation, the working state of the electrodeionization system is intelligently regulated; the working state affects the cation removal efficiency of the electrodeionization system, and the working state is related to the resin capacity of the concentrate chamber, the resin capacity of the fresh water chamber, the ratio of the cross-sectional area to the height of the concentrate chamber, the ratio of the cross-sectional area to the height of the fresh water chamber, the membrane stack voltage, the membrane stack current and the module inlet water flow rate of the electrodeionization system.

2. The intelligent control method for an electrodeionization system for a hydrogen conductivity meter according to claim 1, characterized in that: The S1 includes: S11, arranging a structural sensor group on the electrodeionization system, wherein the structural sensor group includes a MEMS displacement sensor, a MEMS velocity sensor, and a MEMS acceleration sensor, and the structural sensor group is used to obtain a structural dynamic response; S12, the working mode of the structural dynamic response is calculated based on the covariance random subspace method in the time domain to obtain the original stability diagram; S13, cleaning the original stability map based on a cleaning criterion to obtain a stability map after preliminary cleaning, calculating the average regularized power spectrum density in the frequency domain of each degree of freedom according to the structural dynamic response signal, and using the average regularized power spectrum density as a reference to overlay and plot on the stability map after preliminary cleaning; S14, using all the poles in the stability map after the preliminary cleaning as input samples of the grid clustering, and determining the number of minimum clustering units of the grid clustering according to the total number of samples; S15, calculating the normalized working mode distance between any two adjacent clustering units in the stability graph after the preliminary cleaning based on the frequency, damping ratio and working mode vibration type as independent variables for working mode distance calculation; S16, calculating the working mode distance between the cluster unit where each pole is located and its kth nearest neighbor cluster unit, determining the working mode distance distribution based on the working mode distance, and determining the neighborhood radius Eps value according to the working mode distance distribution; S17, substituting the obtained minimum number of neighborhood samples, neighborhood radius, and normalized working mode distance into the grid clustering algorithm to obtain a stable graph with noise removed; S18, Linearly normalized working mode energy MEL and cluster dimension N cn , forming multiple groups of two-dimensional input samples, substituting the multiple groups of two-dimensional input samples into binary K-means clustering, and automatically screening the clusters where the candidate physical working modes are located; S19, effectively evaluating whether the structure of the electrodeionization system has a working mode splitting phenomenon according to the synergistic effect of the working mode confidence factor MAC and the working mode overlap factor MOF, wherein the working mode confidence factor MAC is evaluated from the perspective of vibration mode coefficient, and the working mode overlap factor MOF is evaluated from the perspective of frequency and damping ratio; S110, extracting a representative working mode indexed by the median value of the damping ratio from all the candidate working modes and identifying working mode parameters based on the representative working mode, and determining the structural state of the electrodeionization system based on the identified working mode parameters, and performing the intelligent regulation if there is a problem with the structural state; wherein the working mode parameter identification includes frequency, damping ratio, linear normalized working mode energy MEL and cluster dimension N cn .

3. The intelligent control method for an electrodeionization system for a hydrogen conductivity meter according to claim 2, characterized in that: The S12 includes: S121, construct the Toeplitz matrix, including: Assuming that the output data is ergodic, the output covariance matrix R i And according to the covariance matrix R i Constructed Toeplitz matrix T 1|i Respectively expressed as: In the formula, represents the observation vector of the N-DOF system at discrete time k, that is, the measurement signal time history can be one or more of acceleration, velocity and displacement, where i is the number of test channels; j corresponds to the total length of the measurement data; l represents the comprehensive structural parameter dimension of the electrodeionization system, which can be selected from length, width, height or aspect ratio, aspect ratio, length-to-height ratio, that is, the dimension of l is one-dimensional, two-dimensional to six-dimensional, and the value of l is 1, 2, 3, 4, 5 or 6; In the formula, parameter i represents the number of rows of the Toeplitz matrix, and the matrix T 1|i The dimension is Smaller than the dimension of the original Hankel matrix This reduces the amount of computation and memory requirements; similarly, T 2|i+1 It is expressed as: S122, performs singular value decomposition on the Toeplitz matrix, including: The model order N is another parameter that needs to be defined by the user, by ignoring small singular values, which are equal to the small singular values ​​obtained from the Toeplitz matrix T 1|i The number of singular values ​​obtained by the singular value decomposition of (Equation (5)); In the formula, and is an orthogonal matrix, is a diagonal matrix; is a non-singular matrix, usually taken as the unit matrix, that is, T = I, we can get: In the formula, and They are the extended observable matrix and the extended controllable matrix respectively; S123, identifying a system matrix of the electrodeionization system, including: The expression of the system matrix A is: A=S1 -1 / 2 U1 T T 2|i+1 V1S1 -1 / 2 (8); The operating mode parameters of the discrete-time system are obtained from the system matrix A and the output matrix C, where the output matrix C is equal to O in equation (6): i The first l lines of ; S124, determining working mode parameters, including: Decompose the system matrix A through the eigenvalue to obtain: A=ψLψ -1 ,Asp i =λ i ψ i (9); In formula (9), is the system pole λ of the discrete-time system i The diagonal matrix composed of i corresponds to the right eigenvector of the matrix A; while the continuous-time eigenvalue λ ci and the eigenvalue λ in the discrete state i The following relations are satisfied: l ci =lnλ i / Δt (10); where Δt represents the sampling time interval; finally, the natural frequency f of the system is i , damping ratio ξ i and vibration mode φ i It can be determined by formula (11) as: f i =|λ ci | / 2π,ξ i =-Re(λ ci ) / |λ ci |,φ i =Cψ i (11); S125, parameter settings, including: In the matrix T 1|i When performing singular value decomposition, the matrix T 1|i The number of row blocks i and the model order N must satisfy the following relationship in the algorithm: li ≥ N (12); According to the fundamental frequency f0 and sampling frequency f of the structure s , the lower limit of i determined based on the empirical rule is:

4. The intelligent control method for an electrodeionization system for a hydrogen conductivity meter according to claim 3, characterized in that: The S13 includes: S131, the candidate working modes that do not simultaneously satisfy the cleaning criteria in formula (14) are eliminated as mathematical false working modes: In the formula, λ i is the eigenvalue obtained by eigenvalue decomposition of the system matrix A in the discrete state, is j conjugation of; S132, calculate the average regularized power spectral density ANPSD, which is calculated as shown in the following formula (15): Where m is the number of degrees of freedom of the test; PSD i is the PSD of the ith test degree of freedom signal; f k is the kth discrete frequency; n is the total number of discrete frequencies.

5. The intelligent control method for an electrodeionization system for a hydrogen conductivity meter according to claim 4, characterized in that: The S15 includes: Based on the frequency, damping ratio and working mode vibration type as independent variables for working mode distance calculation, the normalized working mode distance d(j, k) between any two adjacent clustering units in the stability graph after preliminary cleaning is calculated; the calculation formula is as follows: In the formula, φ j refers to the jth vibration mode vector, λ cj Refers to the eigenvalue of the jth continuous state space equation, and its calculation expression is formula (17): The working mode vibration confidence criterion MAC of the traditional real working mode space is extended to the complex working mode space MACXP (φ j ,φ k ), the calculation formula is shown in formula (18): In formula (18), is λ cj The conjugate of Yes j The transpose of Yes j The conjugate transpose of .

6. The intelligent control method for an electrodeionization system for a hydrogen conductivity meter according to claim 5, characterized in that: The S16 includes: S161, calculating the k-th working mode distance between the cluster unit where each data point is located and its k-th nearest neighbor cluster unit based on the working mode distance d(j,k) between the cluster unit where each pole is located and its k-th nearest neighbor cluster unit in the stability graph after the preliminary cleaning; S162, constructing a k-dist function graph using binary K-means clustering, wherein the k-dist function graph is used for adaptive determination of parameters; wherein the basic calculation formula of the binary K-means clustering is shown in the following formula (19): In formula (19), C1 and C2 represent the potential physical working mode and false working mode, respectively, where dX represents the working mode distance of the one-dimensional working mode from the k-th neighbor of the data set calculated according to formula (16), μ k represents the initial ideal cluster center that can be objectively determined, that is, the initial cluster center representing the k-th working mode distance of the physical working mode and the false working mode is 0 and 1 respectively.

7. The intelligent control method for an electrodeionization system for a hydrogen conductivity meter according to claim 6, characterized in that: The S18 includes: S181, the normalized working mode energy MEL and cluster dimension N cn Constructing a two-dimensional input sample data set and objectively determining the initial clustering center of the two-dimensional input sample data set, that is, the initial clustering centers of the two-dimensional samples representing the physical working mode and the false working mode are vectors {1,1} and {0,0} respectively; The calculation formula of MEL is shown in formula (20): In formula (20), ψ i is the matrix A(Aψ i =λ i ψ i )’s i-th right eigenvector; (ψ -1 ) i is the matrix ψ -1 The i-th row of i Refers to the system matrix A in the continuous state space c The i-th eigenvalue of c Δt);() * is a conjugate operation; matrices C and G are the output matrix and the output covariance matrix of the next state, respectively, where matrix C is the formula O i The first l rows of the matrix G can be expanded by the controllable matrix Γ i The last l columns are determined; S182, retaining candidate polar axes with higher cluster dimensions and working mode energies based on the iterative optimization process of the binary K-means clustering, thereby automatically screening the clusters where the candidate physical working modes are located.

8. The intelligent control method for an electrodeionization system for a hydrogen conductivity meter according to claim 7, characterized in that: The S19 includes: S191, determining to use the working mode confidence factor MAC and the working mode overlap factor MOF as indicators for quantifying the degree of overlap between any two working modes, wherein the working mode confidence factor MAC is quantified from the perspective of vibration mode coefficient, and the working mode overlap factor MOF is quantified from the perspective of frequency and damping ratio; The calculation formulas of the working mode confidence factor MAC and the working mode overlap factor MOF are respectively shown in the following formulas (21) and (22): In formula (21), H represents the Hermitian transpose, φ i and φ i Represent the vibration modes of the i-th and j-th order working modes respectively; In formula (22), f i and f j Represent the frequencies of a pair of working modes to be quantified, ξ i Refers to the damping ratio of the i-th order working mode; S192, if the working mode overlap factor MOF of any pair of working modes exceeds 0.1 and the working mode confidence factor MAC exceeds 0.9, the two are automatically identified as belonging to the same order working mode. At this time, the candidate working mode cluster with the smaller cluster dimension will be eliminated.

9. The intelligent control method for an electrodeionization system for a hydrogen conductivity meter according to claim 8, characterized in that: The S2 includes: S21, based on the current parameters of the electrodeionization system and the application scenario of the hydrogen conductivity meter, the resin capacity of the concentrated water chamber, the resin capacity of the deionization chamber, the ratio of the cross-sectional area to the height of the concentrated water chamber, and the ratio of the cross-sectional area to the height of the deionization chamber are set as constants; the membrane stack voltage, the membrane stack current, and the module inlet water flow rate are set as variables; S22, controlling the water flow rate at the module inlet to remain unchanged, gradually increasing the membrane stack voltage and continuously monitoring the change of the membrane stack current, determining the mutation point of the membrane stack current as the current maximum point based on the change of the membrane stack current, and determining the optimal membrane stack voltage range based on the change of the membrane stack current; S23, based on the determined current maximum point and the optimal membrane stack voltage range, controlling the membrane stack current by a constant current mode during the operation of the electrodeionization system, and judging whether the membrane stack voltage is within the optimal membrane stack voltage range; when the membrane stack voltage is not within the optimal membrane stack voltage range, executing step S231; when the membrane stack voltage is within the optimal membrane stack voltage range, executing step S232; S231, controlling the membrane stack voltage within the initial optimal membrane stack voltage range by a constant voltage method, and determining the optimal membrane stack current by adjusting the module inlet water flow rate; then making the electrodeionization system constant at the optimal membrane stack current value, and making the module inlet water flow rate within the flow threshold range by adjusting the membrane stack voltage; S232, determining that the membrane stack voltage at this time is the working membrane stack voltage of the electrodeionization system, and the corresponding membrane stack current is the working membrane stack current of the electrodeionization system, and continuing to determine the water inlet flow rate at the working module inlet; making the electrodeionization system constant at the working membrane stack voltage or the working membrane stack current, and making the water inlet flow rate at the module inlet be within the range of the flow threshold by adjusting the membrane stack current or the membrane stack voltage, and the water inlet flow rate at the module inlet at this time is the water inlet flow rate at the working module inlet of the electrodeionization system; S24, determining a maintenance period and a diagnosis suggestion for the electrodeionization system based on the determined working membrane stack voltage, working membrane stack current and the working module inlet water flow rate and the percentage of the flow threshold.

10. The intelligent control method for an electrodeionization system for a hydrogen conductivity meter according to claim 9, characterized in that: The method further comprises: S25, determining an inlet conductivity range based on the determined working membrane stack voltage, working membrane stack current and working module inlet water flow.

Images