Energy-saving control method and system for electrostatic precipitator

Abstract

The present invention relates to the technical field of electrostatic precipitator control, and discloses an energy-saving control method and system for an electrostatic precipitator. The method comprises: obtaining dust data to generate a spatial distribution map, and calculating a baseline value in combination with a concentration threshold; calculating dust change data based on the distribution map, and generating a difference value in combination with the baseline value and the change threshold; comparing the difference value with the difference threshold, generating a boundary adjustment signal, and dividing the control area to form a coordinate set; obtaining real-time dust data based on the coordinate set, and generating power supply parameters in combination with the distribution threshold; determining the power supply parameter set through deviation calculation and optimization processing; obtaining efficiency parameters based on the power supply parameter set, generating an adjustment parameter set in combination with the efficiency threshold, extracting regional characteristic parameters, and re-dividing the boundary in combination with the characteristic threshold to generate an update plan; according to the update plan, re-acquiring dust data and performing cluster analysis, and outputting electric field adjustment instructions to complete energy-saving control. This method achieves optimal dust removal efficiency and energy consumption through multi-region coordinated power supply.

Description

Technical Field

[0001] The present invention relates to the technical field of electrostatic precipitator control, and in particular to an energy-saving control method and system for an electrostatic precipitator. Background Art

[0002] As industrial flue gas purification standards continue to improve and dynamic operating fluctuations become more common, electrostatic precipitator systems must simultaneously maximize dust removal efficiency and dynamically optimize energy consumption, despite the increasing spatiotemporal heterogeneity of parameters such as dust concentration and particle size distribution. However, differences in dust charge characteristics, the nonlinear nature of electric field intensity distribution, and multi-region coupling effects make it difficult to accurately map real-time operating data to dynamic control targets. Traditional control systems face the bottleneck of insufficient global optimization capabilities.

[0003] One existing technology uses a rigid control framework based on fixed geometric partitions for electrostatic precipitators. This approach uses predefined horizontal or vertical boundaries to define partition sizes and apply uniform voltage and current parameters to each partition. This approach relies on historical experience to determine partition size and power supply intensity. While this simplifies system design, it fails to fully consider the characteristics of dust, leading to a mismatch between power supply strategies and dust removal requirements in highly dynamic scenarios.

[0004] The existing technology uses fixed partitions and unified power supply parameters, which cannot adapt to the dynamic distribution of dust and the differences in real-time working conditions, resulting in low dust removal efficiency and high energy consumption. There is a problem that the system cannot dynamically generate a multi-area collaborative power supply solution that takes into account both dust removal efficiency and optimal energy consumption under complex working conditions. Summary of the Invention

[0005] The present invention provides an energy-saving control method and system for an electrostatic precipitator, aiming to solve the problem of being unable to dynamically generate a multi-region collaborative power supply solution that takes into account both dust removal efficiency and optimal energy consumption.

[0006] In a first aspect, in order to solve the above technical problems, the present invention provides an energy-saving control method for an electrostatic precipitator, comprising:

[0007] Obtaining an original dust data set, generating a spatial distribution map of the dust based on the original data set and performing regional division, and calculating a characteristic baseline data set based on a preset concentration threshold;

[0008] Based on the spatial distribution map, calculating the change data of dust distribution, combining the characteristic baseline data set with a preset change threshold, and generating a regional difference value;

[0009] Comparing the regional difference value with a preset difference threshold, generating an adjustment signal for the regional boundary according to the comparison result, dividing the electric field control region in combination with the spatial distribution map, and generating a boundary coordinate set;

[0010] Acquire real-time dust data based on the boundary coordinate set, calculate differential distribution data of each electric field control area, generate a voltage adjustment signal based on a preset distribution threshold, and determine a current distribution value;

[0011] performing deviation calculation based on the voltage adjustment signal and the current distribution value to obtain voltage deviation distribution data, and performing optimization processing to determine a power supply parameter set;

[0012] Obtaining dust removal efficiency parameters according to the power supply parameter set, and adjusting the power supply parameters in combination with a preset efficiency parameter threshold to obtain an adjusted power supply parameter set;

[0013] Acquire characteristic parameters according to the adjusted power supply parameter set, and redivide the boundary coordinates in combination with a preset characteristic threshold to obtain an updated division scheme;

[0014] According to the updated division scheme, dust data is reacquired and cluster analysis is performed to determine the operating status of the dust collector and obtain adjustment instructions for the electric field control.

[0015] As an optional implementation, generating a spatial distribution map of dust based on the original data set and performing regional division, and calculating the characteristic baseline value of dust in each area in combination with a preset concentration threshold, to obtain a characteristic baseline data set, includes:

[0016] The original data set includes dust concentration data and particle size distribution data, historical concentration data, historical particle size distribution data and corresponding timestamps;

[0017] Extracting a certain amount of sample data from the original data set, gridding the sample data and performing interpolation calculations using a Kriging algorithm to obtain estimated values ​​for each grid point, and visualizing the estimated values ​​to generate a spatial distribution map;

[0018] Based on the concentration data and particle size distribution values ​​in the spatial distribution map, combined with a preset concentration threshold and a preset particle size threshold, regional division is performed to obtain multiple characteristic zones of dust;

[0019] If the concentration data within the characteristic partition is greater than the concentration threshold, a spatial distribution map corresponding to the characteristic partition is obtained, and a convolutional neural network is used to perform feature extraction on the spatial distribution map to obtain distribution characteristics of the abnormal area;

[0020] According to the distribution characteristics of the abnormal area, the sampling data of the characteristic partition is matched with the historical operating condition database to calculate the compensation baseline value of the concentration;

[0021] If the concentration data in the characteristic partition is less than the concentration threshold, obtain the historical concentration data of the characteristic partition, and calculate the concentration mean in the moving time window as the statistical baseline value of the concentration;

[0022] The moving time window is obtained by setting the window length and sliding step size according to the historical concentration data and timestamp;

[0023] The compensated baseline value and the statistical baseline value are combined to form a characteristic baseline data set of dust.

[0024] As an optional implementation, the calculation of dust distribution change data based on the spatial distribution map, combining the characteristic baseline data set with a preset change threshold, and generating a regional difference value includes:

[0025] According to the dust data in the spatial distribution map, the increase or decay rate of the concentration data per unit time is calculated based on the difference method to generate concentration change data, and the absolute difference of the particle size in adjacent time periods is counted to generate particle size offset data;

[0026] The concentration change data and the particle size deviation data together constitute the dust distribution change data;

[0027] If the distribution change data is greater than a preset change threshold or a baseline value corresponding to the offset, the corresponding area is determined to be an abnormal area, and a regional difference value of the abnormal area is calculated;

[0028] If the distribution change data is smaller than a preset change threshold or is within the corresponding baseline value, the corresponding area is determined to be a normal area, and the area difference value is 0.

[0029] As an optional implementation manner, the comparing the regional difference value with a preset difference threshold, generating an adjustment signal for the regional boundary according to the comparison result, dividing the electric field control area in combination with the spatial distribution map, and generating a boundary coordinate set includes:

[0030] Comparing the regional difference value of each region with a preset difference threshold, if the regional difference value is greater than the difference threshold, generating an adjustment signal for the region, and recording the coordinates of the region as abnormal coordinates;

[0031] If the regional difference value is less than the difference threshold, an adjustment signal for the region is generated, and the coordinates of the region are recorded as normal coordinates;

[0032] According to the adjustment signal, the system extracts all the abnormal coordinates and the normal coordinates, divides the electric field control area in combination with the spatial distribution map, and generates boundary coordinate sets of multiple independent electric field control areas.

[0033] As an optional implementation, acquiring real-time dust data based on the boundary coordinate set, calculating differential distribution data of each electric field control area, generating a voltage adjustment signal in combination with a preset distribution threshold, and determining a current distribution value include:

[0034] The real-time dust data includes real-time dust concentration value and real-time particle size distribution value;

[0035] Preprocessing the real-time dust data to eliminate noise interference to obtain noise-reduced dust data, and calculating regional difference values ​​based on the noise-reduced dust data to obtain difference distribution data;

[0036] If the difference distribution data is greater than a preset difference threshold, a voltage adjustment mechanism is triggered and a voltage adjustment signal is generated;

[0037] Calculating the current distribution value of each electric field control area according to the voltage adjustment signal and the preset current distribution rule;

[0038] If the difference distribution data is less than the preset difference threshold, the dust collector continues to work normally;

[0039] Calculating the current distribution value of each electric field control area according to the voltage adjustment signal and the preset current distribution rule;

[0040] The current distribution value of each electric field control area when the dust collector is working normally is calculated based on the rated voltage and rated power of the dust collector.

[0041] As an optional implementation manner, performing deviation calculation based on the voltage adjustment signal and the current distribution value to obtain voltage deviation distribution data, and performing optimization processing to determine a power supply parameter set includes:

[0042] Acquire monitoring data of the actual working condition of the dust collector, combine the voltage adjustment signal and the current distribution value, calculate the power supply voltage and power supply current of each electric field control area, and obtain a power supply parameter set;

[0043] Calculating the deviation value of the power supply voltage based on the power supply parameter set and the operating condition change rate in the monitoring data to determine regional voltage deviation distribution data;

[0044] Determine the fluctuation range data of the power supply current based on the deviation distribution data, judge the matching trend of the power supply parameters and the working conditions in combination with a preset fluctuation threshold, and generate trend distribution data;

[0045] Based on the trend distribution data, establishing a mapping relationship between power supply matching and power supply parameters to obtain a first mapping data set;

[0046] If the first mapping data set is greater than a preset first mapping threshold, a genetic algorithm is used to optimize the power supply voltage and the power supply current to obtain a preliminary power supply parameter set.

[0047] As an optional implementation manner, obtaining the dust removal efficiency parameter according to the power supply parameter set, adjusting the power supply parameter in combination with a preset efficiency parameter threshold, and obtaining the adjusted power supply parameter set includes:

[0048] If the dust removal efficiency parameter is higher than a preset efficiency parameter threshold, the power supply parameter set is updated using a particle swarm optimization algorithm to obtain an updated power supply parameter set;

[0049] According to the updated power supply parameter set, new dust removal efficiency parameters are obtained and recorded as efficiency update parameters, and the efficiency update parameters are analyzed using a support vector machine algorithm to obtain efficiency change data;

[0050] Based on the efficiency change data, a mapping relationship between the power supply parameters of each area and the dust removal efficiency is established to obtain a second mapping data set;

[0051] If the second mapping data set is greater than a preset second mapping threshold, the power supply parameters are iteratively adjusted according to the change direction to obtain an adjusted power supply parameter set.

[0052] As an optional implementation, the method of reacquiring dust data and performing cluster analysis according to the updated partitioning scheme to determine the operating status of the dust collector and obtain adjustment instructions for the electric field control includes:

[0053] Based on the updated division scheme, dust concentration, particle size distribution data and environmental interference values ​​of each area of ​​the electrostatic precipitator are collected in real time to generate a real-time data set, obtain real-time power supply current and power supply voltage, and generate electric field management parameters;

[0054] Calculating the power supply matching degree and regional difference value of each area according to the real-time data set to obtain matching degree data;

[0055] If the power supply matching degree in the matching degree data is higher than a preset matching degree threshold, adjusting the electric field management parameters based on an adaptive particle swarm optimization algorithm to generate a management parameter set;

[0056] Obtain updated dust concentration and particle size distribution data based on the management parameter set, and generate an updated data set in combination with the environmental interference value;

[0057] Using a K-means clustering algorithm to classify the regional difference values ​​in the updated data set to obtain a classification result set;

[0058] According to the classification result set, a fuzzy logic control algorithm is used to determine the change trend of the dust collector operation state to obtain a state data set;

[0059] According to the state data set, a supply voltage adjustment amount and a supply current adjustment amount are generated, and an adjustment instruction is output at the same time to perform electric field control on the dust collector to complete energy-saving control.

[0060] In a second aspect, the present invention provides an energy-saving control system for an electrostatic precipitator, comprising:

[0061] A data acquisition module is used to obtain an original dust data set, generate a spatial distribution map of the dust based on the original data set, perform regional division, and calculate a characteristic baseline data set based on a preset concentration threshold;

[0062] A regional difference value acquisition module is used to calculate the change data of dust distribution based on the spatial distribution map, and generate a regional difference value by combining the characteristic baseline data set and a preset change threshold;

[0063] a boundary coordinate set acquisition module, configured to compare the regional difference value with a preset difference threshold, generate an adjustment signal for the regional boundary according to the comparison result, divide the electric field control area in combination with the spatial distribution map, and generate a boundary coordinate set;

[0064] an adjustment signal generation module for acquiring real-time dust data based on the boundary coordinate set, calculating differential distribution data of each electric field control area, generating a voltage adjustment signal based on a preset distribution threshold, and determining a current distribution value;

[0065] A power supply parameter set module is used to perform deviation calculation based on the voltage adjustment signal and the current distribution value to obtain voltage deviation distribution data, and perform optimization processing to determine a power supply parameter set;

[0066] A power supply parameter set adjustment module is used to obtain dust removal efficiency parameters according to the power supply parameter set, and adjust the power supply parameters in combination with a preset efficiency parameter threshold to obtain an adjusted power supply parameter set;

[0067] An updating partitioning scheme module is used to obtain characteristic parameters according to the adjusted power supply parameter set, and to re-divide the boundary coordinates in combination with a preset characteristic threshold to obtain an updated partitioning scheme;

[0068] The energy-saving control module reacquires the dust data and performs cluster analysis according to the updated partitioning scheme, determines the operating status of the dust collector, and obtains adjustment instructions for the electric field control.

[0069] In a third aspect, the present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the energy-saving control method of an electrostatic precipitator described in the first aspect above.

[0070] Compared with the prior art, the present invention has the following beneficial effects:

[0071] (1) The present invention constructs a spatial distribution map of dust in an electrostatic precipitator, uses a differential algorithm to analyze the concentration change rate and particle size offset, and combines the intelligent comparison of dynamic thresholds and characteristic baselines to accurately generate regional difference values. This technology breaks through the limitations of traditional monitoring methods and achieves precise positioning of abnormal dust deposition areas through dynamic analysis of concentration and particle size characteristics. The difference value generation integrates historical data trends and fluctuation characteristics, effectively avoiding misjudgments caused by monitoring lags, and adaptively matches different operating parameters to provide accurate support for the zoning power supply strategy, ultimately achieving a coordinated improvement in electrostatic precipitator energy consumption optimization and dust removal efficiency.

[0072] (2) The present invention obtains dust distribution data in real time based on the boundary coordinate set, constructs a voltage adjustment signal set for multiple electric field partitions through differential distribution data, and accurately calculates the current distribution value of each region. This technology breaks through the extensive mode of unified power supply for the entire traditional electrostatic precipitator, and adopts a partition control algorithm that links the dust concentration gradient with the threshold to convert real-time monitoring data into a quantifiable voltage / current parameter combination. By dynamically matching the dust distribution characteristics, an adaptive enhanced power supply signal is generated for the high-load area to maintain the dust removal efficiency, and the low-load area outputs the optimal energy-saving parameter combination. This mechanism achieves millimeter-level response to the power supply strategy and dust distribution, achieving a precise balance between improving dust removal efficiency and reducing ineffective energy consumption.

[0073] (3) The present invention dynamically updates the power supply parameter set through the particle swarm optimization algorithm, combines it with the support vector machine to analyze the performance change data, and constructs a quantitative mapping model between the power supply parameters and the dust removal efficiency. When the mapping data exceeds the threshold, an iterative algorithm is used to continuously optimize the voltage / current combination to form a closed-loop control mechanism of "parameter update-performance feedback-strategy iteration". This technology associates discrete parameters with dust removal efficiency and models them, which not only ensures the scientific nature of parameter adjustment, but also accurately matches the dynamic changes of working conditions, achieves the optimal energy consumption configuration while maintaining emission stability, and ultimately achieves intelligent optimization of the power supply strategy of the electrostatic precipitator and refined energy efficiency management.

[0074] (4) The present invention reacquires dust data and performs cluster analysis to calculate the regional power supply matching and difference values ​​to accurately identify the operating status. When the matching degree is higher than the threshold, the adaptive particle swarm optimization algorithm is used to dynamically adjust the electric field management parameters, and the updated regional difference values ​​are classified in combination with K-means clustering to form a classification result set. The classification results are further analyzed by the fuzzy logic control algorithm to determine the changing trend of the dust collector operating status, and finally the optimization parameters and control instructions such as the power supply voltage are generated to achieve precise adjustment of the electric field parameters. The multi-algorithm linkage converts environmental interference and real-time operating condition data into dynamic control strategies, improves dust removal efficiency and power utilization efficiency, and enhances the system's adaptability and stability to complex working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 This is a flow chart of an energy-saving control method for an electrostatic precipitator provided by an embodiment of the present invention;

[0076] Figure 2 The present invention provides an energy-saving control system for an electrostatic precipitator. DETAILED DESCRIPTION

[0077] The following will clearly and completely describe the technical solutions in the embodiments of the present invention 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. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0078] To solve the above problems, refer to Figure 1 The first embodiment of the present invention provides an energy-saving control method for an electrostatic precipitator, comprising the following steps:

[0079] S1, obtaining an original dust dataset, generating a spatial distribution map of the dust based on the original dataset and performing regional division, and calculating a characteristic baseline dataset based on a preset concentration threshold;

[0080] S2, calculating the change data of dust distribution based on the spatial distribution map, combining the characteristic baseline data set with a preset change threshold, and generating a regional difference value;

[0081] S3, comparing the regional difference value with a preset difference threshold, generating an adjustment signal for the regional boundary according to the comparison result, dividing the electric field control region in combination with the spatial distribution map, and generating a boundary coordinate set;

[0082] S4, acquiring real-time dust data according to the boundary coordinate set, calculating differential distribution data of each electric field control area, generating a voltage adjustment signal in combination with a preset distribution threshold, and determining a current distribution value;

[0083] S5, performing deviation calculation based on the voltage adjustment signal and the current distribution value to obtain voltage deviation distribution data, and performing optimization processing to determine a power supply parameter set;

[0084] S6, obtaining dust removal efficiency parameters according to the power supply parameter set, and adjusting the power supply parameters in combination with a preset efficiency parameter threshold to obtain an adjusted power supply parameter set;

[0085] S7, acquiring characteristic parameters according to the adjusted power supply parameter set, and re-dividing the boundary coordinates in combination with a preset characteristic threshold to obtain an updated division scheme;

[0086] S8, according to the updated division scheme, reacquire the dust data and perform cluster analysis to determine the operating status of the dust collector and obtain adjustment instructions for the electric field control.

[0087] In step S1, the original data set of dust is obtained, a spatial distribution map of dust is generated according to the original data set and regional division is performed, and a characteristic baseline data set is calculated in combination with a preset concentration threshold.

[0088] In one implementation, obtaining an original dust dataset, generating a spatial distribution map of the dust based on the original dataset and performing regional division, and calculating a characteristic baseline dataset in combination with a preset concentration threshold may include:

[0089] Obtain a raw dust dataset, including dust concentration and particle size distribution data, historical concentration data, historical particle size distribution data, and corresponding timestamps. Dust concentration data is obtained using a laser scattering sensor and is characterized by the current concentration value, concentration change rate, and historical average concentration. Particle size distribution data is measured using an optical particle size analyzer and is characterized by the median particle size, particle size distribution width, and particle fraction. Historical concentration and particle size distribution data are obtained from the dust collector database. Timestamps are recorded in the database using the system clock.

[0090] Extracting a certain amount of sample data from the original data set, gridding the sample data and performing interpolation calculations using a Kriging algorithm to obtain estimated values ​​for each grid point, and visualizing the estimated values ​​to generate a spatial distribution map;

[0091] Based on the concentration data and particle size distribution values ​​in the spatial distribution map, combined with a preset concentration threshold and a preset particle size threshold, regional division is performed to obtain multiple characteristic zones of dust;

[0092] Furthermore, the concentration data and particle size distribution values ​​in the spatial distribution map are combined with a preset concentration threshold and a preset particle size threshold to perform regional division to obtain multiple characteristic zones of dust, including:

[0093] Each grid point is traversed, and the partition category to which it belongs is determined based on the current point's estimated concentration and particle size values, combined with the preset concentration threshold and particle size threshold. The partition categories include low concentration, medium concentration, high concentration, fine particle, medium particle, and coarse particle areas. The estimated concentration value is denoted as C1, the estimated particle size value is denoted as D1, the concentration threshold range is denoted as C2-C3, and the particle size threshold is denoted as D2-D3. When C1 is less than C2 and D1 is less than D2, the grid point is classified as a low-concentration, fine particle area. When C1 is less than C2 and D1 is between D2 and D3, the grid point is classified as a low-concentration, medium particle area. When C1 is less than C2 and D1 is greater than D3, the grid point is classified as a low-concentration, coarse particle area. When C1 is between C2 and C2 and D1 is less than D2, the grid point is classified as a medium-concentration, fine particle area. When C1 is between C2 and C2, and D1 is between D2 and D3, the grid point is divided into the medium-concentration medium-particle area. When C1 is between C2 and C2, and D1 is greater than D3, the grid point is divided into the medium-concentration coarse-particle area. When C1 is greater than C3, and D1 is less than D2, the grid point is divided into the high-concentration fine-particle area. When C1 is greater than C3, and D1 is between D2 and D3, the grid point is divided into the high-concentration medium-particle area. When C1 is greater than C3, and D1 is greater than D3, the grid point is divided into the high-concentration coarse-particle area.

[0094] If the concentration data within the characteristic partition is greater than the concentration threshold, a spatial distribution map corresponding to the characteristic partition is obtained, and a convolutional neural network is used to perform feature extraction on the spatial distribution map to obtain distribution characteristics of the abnormal area;

[0095] Specifically, when the concentration data within a characteristic partition exceeds the preset concentration threshold, a spatial distribution image of the concentration data is extracted from the spatial distribution map corresponding to the characteristic partition. The image resolution is consistent with the grid division, and each pixel value corresponds to the concentration estimate after Kriging interpolation. The convolutional neural network processing process is as follows:

[0096] The spatial distribution image is fed into the convolutional and pooling layers of the neural network to extract global features from the feature map. The feature vector is mapped to a 1024-dimensional hidden layer with a Reluctant Unit (ReLU) activation function. Using semantic segmentation, a pixel-level anomaly probability map is output, marking the coordinates and geometric extent of the anomaly region. Consecutive anomaly pixels are analyzed to generate independent bounding boxes for the anomaly region. The output is the distribution characteristics of the anomaly region, including its coordinates, concentration mean / variance, gradient direction, and duration.

[0097] According to the distribution characteristics of the abnormal area, the sampling data of the characteristic partition is matched with the historical operating condition database to calculate the compensation baseline value of the concentration;

[0098] If the concentration data in the characteristic partition is less than the concentration threshold, obtain the historical concentration data of the characteristic partition, and calculate the concentration mean in the moving time window as the statistical baseline value of the concentration;

[0099] The compensated baseline value and the statistical baseline value are taken as a vector in a vector set to form a characteristic baseline data set of dust.

[0100] Furthermore, the sampled data is gridded and interpolated using the Kriging algorithm to obtain an estimated value for each grid point, including:

[0101] According to the target area to which the sampling data belongs, the monitoring space is divided into equally spaced rectangular grids (e.g., 1m×1m), with each grid point corresponding to a unique geographic coordinate, forming a regularized grid matrix covering the entire area of ​​the electrostatic precipitator;

[0102] For each grid point, a preset number (e.g., 10) of sampling points in its neighborhood are selected to construct a covariance matrix, and the weight coefficients are solved based on the Kriging equations.

[0103] The weight coefficient is weighted and summed with the observation value of the corresponding sampling point to obtain the estimated value of the grid point, and this process is repeated to complete the interpolation calculation of all grid points;

[0104] The estimated value is calculated as:

[0105]

[0106] in, Represented as sampling point x i The estimated value of λ i Represented as x i The corresponding Kriging weight coefficient, Z(x i ) is represented by x i The observed value of , n represents the number of sampling points;

[0107] It should be noted that the estimated values ​​are visualized and a spatial distribution map is generated using visualization tools such as a three-dimensional surface map, a contour map, and a heat map.

[0108] Furthermore, the calculation of the concentration compensation baseline value based on the distribution characteristics of the abnormal area, in combination with the sampling data of the characteristic partition and matching with the historical operating condition database, includes:

[0109] Extracting characteristic parameters of the abnormal area from the spatial distribution map, wherein the characteristic parameters include concentration change rate, concentration time series data and morphological parameters;

[0110] Label the historical working condition database and construct a multi-dimensional feature index: working condition characteristics, dust characteristics and morphological space characteristics;

[0111] For concentration time series data, the DTW algorithm is used to calculate the similarity between the characteristic parameters of the current abnormal area and the historical operating conditions, and the Mahalanobis distance is used to quantify the multi-dimensional feature similarity;

[0112] Select the top K historical operating conditions with the highest similarity, obtain the distribution weights based on the similarity, and use the weighted average result as the compensation baseline value:

[0113]

[0114] The weight calculation formula is:

[0115]

[0116] Among them, C com Expressed as the concentration-compensated baseline value, λ i represents the weight of the i-th historical data, C his It is expressed as the baseline concentration value of the i-th matching historical operating condition, d i It is expressed as the Mahalanobis distance between the current abnormal area and the i-th historical operating condition, and α is the adjustment coefficient, which controls the sensitivity of the weight to the distance.

[0117] It should be noted that the K value is determined by analyzing the similarity distribution and data volume of historical working conditions, combined with the elbow method, cross-validation or statistical calculation (such as taking the square root of the total number of historical working conditions).

[0118] Furthermore, historical concentration data of the characteristic partition is obtained, and the concentration mean value within the moving time window is calculated as the statistical baseline value of the concentration, including:

[0119] Sorting the historical concentration data into historical concentration time series data by timestamp, and setting the length of the moving time window and the sliding step size;

[0120] For each time point, a data subset within the window is intercepted and the arithmetic mean of the concentration within the window is calculated as the statistical baseline value:

[0121]

[0122] Among them, C b (t i ) represents the time point t i The statistical baseline value, C jIt is represented as the concentration value at the jth time point in the window, and n is represented as the number of valid data points in the window.

[0123] In step S2, based on the spatial distribution map, the change data of the dust distribution is calculated, and the regional difference value is generated by combining the characteristic baseline data set with a preset change threshold.

[0124] In one implementation, calculating the dust distribution change data based on the spatial distribution map, combining the characteristic baseline data set with a preset change threshold, and generating a regional difference value includes:

[0125] According to the dust data in the spatial distribution map, the increase or decay rate of the concentration data per unit time is calculated based on the difference method to generate concentration change data, and the absolute difference of the particle size in adjacent time periods is counted to generate particle size offset data;

[0126] The concentration change data and the particle size deviation data together constitute the dust distribution change data;

[0127] If the distribution change data is greater than a preset change threshold or a baseline value corresponding to the offset, the corresponding area is determined to be an abnormal area, and a regional difference value of the abnormal area is calculated;

[0128] If the distribution change data is smaller than a preset change threshold or is within the corresponding baseline value, the corresponding area is determined to be a normal area, and the area difference value is 0.

[0129] Furthermore, the increase or decrease rate of the concentration data per unit time is calculated based on the difference method. The calculation formula is expressed as:

[0130]

[0131] Among them, R ij,t Expressed as the concentration change rate per unit time at the grid point (i, j), C ij,t It is represented as the dust concentration of grid point (i, j) at time t, Δt is the data sampling time interval, and k is the number of time windows for smoothing processing;

[0132] Furthermore, the formula for calculating the regional difference value of the abnormal area is:

[0133]

[0134] Among them, V diff (i, j) represents the quantitative value of the regional anomaly at the grid point (i, j), that is, the difference value, ε represents the weight coefficient of the concentration change rate in the difference value, T RIt is expressed as the concentration change rate threshold, β is expressed as the weight coefficient of particle size deviation in the difference value, and the weight coefficients ε and β satisfy ε+β=1, ΔD i,j,t It is expressed as the deviation of the real-time particle size distribution at the grid point (i, j) from the historical baseline, B D Expressed as historical baseline value, ΔT D Expressed as the particle size deviation tolerance threshold.

[0135] In step S3, the regional difference value is compared with a preset difference threshold to generate an adjustment signal for the regional boundary, and the electric field control region is divided in combination with the spatial distribution map to generate a boundary coordinate set.

[0136] In one implementation, the comparing the regional difference value with a preset difference threshold, generating an adjustment signal for the regional boundary, dividing the electric field control region in combination with the spatial distribution map, and generating a boundary coordinate set includes:

[0137] Comparing the regional difference value of each region with a preset difference threshold, if the regional difference value is greater than the difference threshold, generating an adjustment signal for the region, and recording the coordinates of the region as abnormal coordinates;

[0138] If the regional difference value is less than the difference threshold, an adjustment signal for the region is generated, and the coordinates of the region are recorded as normal coordinates;

[0139] It should be noted that the adjustment signal is a control instruction generated by the system based on the comparison result of the regional difference value and the threshold value, which is used to guide the division and boundary adjustment of the electric field control area and includes coordinates and status identification;

[0140] According to the adjustment signal, the system extracts all the abnormal coordinates and the normal coordinates, divides the electric field control area in combination with the spatial distribution map, and generates boundary coordinate sets of multiple independent electric field control areas.

[0141] In one example, the steps for dividing the electric field control area are as follows:

[0142] The discrete points in the abnormal coordinate set are spatially grouped to generate multiple independent abnormal area clusters. For example, if multiple grid points are marked as abnormal due to excessive pollution concentration, these points are aggregated into one or more clusters, each representing an independent abnormal area.

[0143] The minimum convex polygon boundary is calculated for the coordinate points of each abnormal cluster to generate a closed polygon vertex sequence. This step converts discrete abnormal points into clear geometric boundaries, such as connecting scattered dust points into a regular or polygonal area.

[0144] For normal areas, the scattered normal coordinate points are merged into a continuous large area, and then the outer contour boundary is extracted to form a closed polygon, for example, the scattered normal grid points are integrated into a complete area boundary;

[0145] For abnormal areas, each abnormal cluster is independently divided into a high-voltage power supply area, and for normal areas, it is divided into a low-energy operation area.

[0146] It should be noted that the abnormal coordinates include the spatial coordinates of all grid points in the abnormal area, and the normal coordinates include the spatial coordinates of all grid points in the normal area.

[0147] In step S4, real-time dust data is acquired according to the boundary coordinate set, differential distribution data of each electric field control area is calculated, a voltage adjustment signal is generated in combination with a preset distribution threshold, and a current distribution value is determined.

[0148] In one implementation, acquiring corresponding real-time dust data according to the boundary coordinate set, calculating differential distribution data of each electric field control area, generating a voltage adjustment signal in combination with a preset differential threshold, and determining a current distribution value include:

[0149] The real-time dust data includes real-time dust concentration value and real-time particle size distribution value;

[0150] Preprocessing the real-time dust data to eliminate noise interference to obtain noise-reduced dust data, and performing real-time regional difference value calculation based on the noise-reduced dust data to obtain real-time difference distribution data;

[0151] If the difference distribution data is greater than a preset difference threshold, a voltage adjustment mechanism is triggered and a voltage adjustment signal is generated;

[0152] Calculating the current distribution value of each electric field control area according to the voltage adjustment signal and the preset current distribution rule;

[0153] If the difference distribution data is less than the preset difference threshold, the dust collector continues to work normally;

[0154] The current distribution value of each electric field control area when the dust collector is working normally is calculated based on the rated voltage and rated power of the dust collector.

[0155] It should be noted that the voltage adjustment signal is a voltage adjustment amount generated according to the ratio of the difference value exceeding the threshold value. The preset rule is an on-demand allocation rule, that is, the current is allocated according to the voltage adjustment amount to meet the power balance. The calculation formula of the current allocation value is expressed as:

[0156]

[0157] Among them, I i,jExpressed as the current distribution value of region (i, j), P b Expressed as baseline power, ΔU i,j It is expressed as the voltage adjustment amount of region (i, j), I ref Expressed as electric field rated current;

[0158] In one example, the steps of preprocessing the real-time dust data to eliminate noise interference are as follows: the dust concentration value and particle size distribution value in the electric field control area are collected through the real-time dust monitoring system, and the data is preprocessed using a sliding average filtering algorithm, with the sliding window set to 5 minutes and the step length set to 30 seconds to eliminate noise interference.

[0159] In step S5, deviation calculation is performed based on the voltage adjustment signal and the current distribution value to obtain voltage deviation distribution data, and optimization processing is performed to determine a power supply parameter set.

[0160] In one implementation, performing deviation calculation based on the voltage adjustment signal and the current distribution value to obtain voltage deviation distribution data, and performing optimization processing to determine a power supply parameter set includes:

[0161] Acquire monitoring data of the actual working condition of the dust collector, combine the voltage adjustment signal and the current distribution value, calculate the power supply voltage and power supply current of each electric field control area, and obtain a power supply parameter set;

[0162] It should be noted that the power supply parameter set is a combination of supply voltage and supply current, such as [3.1kV / 1.4A, 3.4kV / 1.7A, 2.9kV / 1.2A]

[0163] The deviation value of the supply voltage is calculated based on the power supply parameter set and the operating condition change rate in the monitoring data to determine the deviation distribution data of the regional voltage. Furthermore, the calculation formula of the deviation value of the supply voltage is expressed as:

[0164]

[0165] Among them, ΔU ％ Indicates the deviation value of the supply voltage, U actual,i represents the supply voltage of the ith region, U nom,i represents the detection voltage of the i-th region, and the deviation distribution data is D={ΔU ％,1 ,ΔU ％,2 ,…,ΔU ％,n}.

[0166] Determine the fluctuation range data of the power supply current based on the deviation distribution data, judge the matching trend of the power supply parameters and the working conditions in combination with a preset fluctuation threshold, and generate trend distribution data;

[0167] Furthermore, the calculation formula of the fluctuation range data is expressed as:

[0168] ΔI range,i =m·ΔU ％,i

[0169] Among them, ΔI range,i represents the current fluctuation in the i-th region, m represents the current-voltage sensitivity coefficient (describing the effect of voltage deviation on current fluctuation), ΔU ％,i Represents the deviation distribution data of the i-th region.

[0170] It should be noted that the deviation distribution data indicates the distribution formed in different areas or time points, reflecting the degree of matching between the power supply parameters and the working conditions. The fluctuation range data indicates that when the power supply voltage deviates due to changes in the working conditions (such as fluctuations in dust concentration) (such as the voltage is set too high or too low), the current will directly change accordingly. For example, if the voltage in a certain area is increased due to the deviation, and the dust in the actual working conditions is reduced, the current may fluctuate greatly due to the excessively strong electric field (such as a sudden increase from 100A to 150A). The system captures this current change caused by the voltage deviation; further, the trend distribution data is obtained as follows: the actual fluctuation range of the current is compared with the preset fluctuation threshold (such as the allowed fluctuation of ±10%). If the fluctuation exceeds the threshold, it means that the power supply parameters (voltage / current) do not match the current working conditions (such as the voltage setting is too high, resulting in unstable current). Through continuous monitoring, observe whether this mismatch is gradually worsening (such as the fluctuation range is widening) or improving (such as the fluctuation is shrinking), thereby generating trend distribution data;

[0171] In one example, the generation of trend distribution data is represented as follows: if the fluctuation range of subsequent time points is greater than the fluctuation range of previous time points, then the matching degree is considered to be gradually deteriorating, such as: If the fluctuation range at subsequent time points is smaller than that at previous time points, the matching degree is considered to be gradually improving, such as:

[0172] Record the changing trend of the fluctuation range in the form of time series to form trend distribution data:

[0173]

[0174] Among them, T D Represents trend distribution data, It represents the fluctuation range data of the i-th region at time t1. Status1 represents the matching status at the corresponding time, and its value can be "deterioration", "improvement" or "stable".

[0175] Based on the trend distribution data, establishing a mapping relationship between power supply matching and power supply parameters to obtain a first mapping data set;

[0176] It should be noted that the establishment of the mapping relationship is specifically manifested in converting the relationship between power supply parameters (voltage, current) and working condition matching into a queryable "parameter-matching" data set; for example, in the power supply parameter combination (8kV / 120A), the fluctuation range is ±10A (within the standard, stable trend), and the score is 90 points;

[0177] If the first mapping data set is greater than a preset first mapping threshold, a genetic algorithm is used to optimize the power supply voltage and the power supply current to obtain a preliminary power supply parameter set.

[0178] It should be noted that if the first mapping data set is smaller than a preset first mapping threshold, the power supply parameter set is a preliminary power supply parameter set.

[0179] It should be noted that the optimization process using a genetic algorithm can be expressed as follows: converting voltage and current parameters into "chromosomes" (e.g., binary or numerical form), generating multiple sets of random parameter combinations (e.g., 100 sets), and forming an initial population. For each set of parameters, its operating performance under the current operating conditions is simulated (e.g., predicting the current fluctuation range), and combined with a preset threshold (e.g., allowing fluctuations of ±10%), a matching score is calculated (e.g., the smaller the fluctuation, the higher the score). Parameter combinations with high matching degrees are retained (e.g., the top 20% high-scoring combinations are retained), two sets of parameters are randomly selected, some parameter values ​​are exchanged to generate new combinations, and some parameters are fine-tuned (e.g., adjusting the voltage from 8kV to 8.2kV) to explore new solutions. The above process is repeated until the termination condition is met (e.g., the number of iterations reaches 100 or the matching degree no longer improves). The parameter combination with the highest matching degree is selected from the final population to form a preliminary power supply parameter set.

[0180] Furthermore, according to the voltage adjustment signal and the current distribution value, combined with the monitoring data of the actual working condition of the dust collector, the power supply voltage and power supply current of each electric field control area are calculated, including:

[0181] The monitoring data of the actual working conditions include real-time dust concentration, ambient temperature and ambient humidity, etc.;

[0182] The supply voltage calculation formula is expressed as:

[0183] U(i,j)=U base +ΔU i,j f(T,H)

[0184] The power supply current calculation formula is expressed as:

[0185]

[0186] Among them, U(i,j) represents the supply voltage of area (i,j), U baseis the reference voltage value, f(T,H) is the temperature and humidity correction factor, I′(i,j) is the power supply current of region (i,j), C base Expressed as the reference concentration value, C scale Expressed as a concentration scaling factor.

[0187] In step S6, the dust removal efficiency parameters are obtained according to the power supply parameter set, and the power supply parameters are adjusted in combination with the preset efficiency parameter threshold to obtain an adjusted power supply parameter set.

[0188] In one implementation, obtaining the dust removal efficiency parameter according to the power supply parameter set, adjusting the power supply parameter in combination with a preset efficiency parameter threshold, and obtaining the adjusted power supply parameter set includes:

[0189] If the dust removal efficiency parameter is higher than a preset efficiency parameter threshold, the power supply parameter set is updated using a particle swarm optimization algorithm to obtain an updated power supply parameter set;

[0190] According to the updated power supply parameter set, new dust removal efficiency parameters are obtained and recorded as efficiency update parameters, and the efficiency update parameters are analyzed using a support vector machine algorithm to obtain efficiency change data;

[0191] It should be noted that the dust removal efficiency parameters include dust deposition and electric field response time. The dust deposition is measured by a dust concentration sensor installed at the outlet of the dust collector to monitor the dust content in the exhaust gas in real time. Combined with the dust content before entering the dust collector (a sensor is also installed at the inlet), the amount of captured dust is calculated. Dust deposition = (inlet dust concentration - outlet dust concentration) × processed air volume. The electric field response time uses a monitoring system to record the time point of each working condition change and the time point when the dust collector starts to respond. The difference between the two is the electric field response time.

[0192] The support vector machine algorithm's analysis process is as follows: The parameters of the updated power supply parameter set (e.g., voltage 8.2kV, current 130A) and the corresponding dust removal performance (e.g., dust reduction 15%) are input into the SVM model. The SVM is trained using historical dust removal performance data and power supply parameters to identify patterns in the correlation between the parameters and performance.

[0193] When a new set of power supply parameters is introduced, these new parameters are fed into the trained SVM model to predict the expected dust removal performance under this parameter configuration. Based on the SVM model's output, the performance indicators (e.g., dust reduction percentage) before and after the adjustment are compared, and the change in performance is calculated, representing the performance change data. For example, if the original dust reduction rate was 15% and it increased to 20% after the parameter adjustment, the performance change data might include "+5%."

[0194] Based on the efficiency change data, a mapping relationship between the power supply parameters of each area and the dust removal efficiency is established to obtain a second mapping data set;

[0195] It should be noted that the establishment of the mapping relationship is specifically manifested in converting the relationship between power supply parameters (voltage, current) and efficiency change data into a queryable "power supply parameter-efficiency data" data set; for example, in the power supply parameter combination (8.2kV / 130A), the corresponding dust efficiency parameter is 5%, that is, {(8.2kV / 130A), 5%}.

[0196] If the second mapping data set is greater than a preset second mapping threshold, iteratively adjusting the power supply parameters according to the change direction to obtain an adjusted power supply parameter set;

[0197] It should be noted that if the predicted performance of the second mapping data set exceeds the preset threshold, it means there is still room for optimization. Based on the SVM prediction trend, the parameter adjustment direction is determined (for example, the voltage needs to be further increased to further reduce dust); combined with safety limits (such as the voltage upper limit), the adjustment range is controlled (for example, ±0.1kV each time), and the process is repeated until the performance change stabilizes or reaches the target, and the adjusted power supply parameter set is output;

[0198] Furthermore, the power supply parameter set is updated using a particle swarm optimization algorithm to obtain an updated power supply parameter set, including:

[0199] Based on the power supply parameter set, m groups of candidate parameters (U, I) are randomly generated. Each parameter combination is called a "particle". The dust removal efficiency of the particle is evaluated, with priority given to reducing dust deposition and shortening electric field response time.

[0200] If the dust deposition amount or response time exceeds the preset threshold, the score will be reduced. The higher the score, the better the parameter combination (U, I). The best parameter combination of each particle in the historical iteration is called the individual optimal, and the parameter combination with the highest score in the entire particle group is called the global optimal.

[0201] The particles move towards the individual optimal and global optimal directions, with the speed controlled by the inertia weight and learning factor. The voltage and current values ​​are adjusted according to the updated speed to generate new candidate parameters.

[0202] After the iteration is completed, the global optimal parameter is the updated power supply parameter set (U new ,I new ).

[0203] In step S7, characteristic parameters of the region are acquired according to the adjusted power supply parameter set, and the boundary coordinates are re-divided in combination with a preset characteristic threshold to obtain an updated division scheme.

[0204] In one implementation, acquiring characteristic parameters of the region according to the adjusted power supply parameter set and re-dividing the boundary coordinates in combination with a preset characteristic threshold to obtain an updated division scheme includes:

[0205] Based on the adjusted power supply parameter set, real-time dust concentration, particle size distribution and boundary displacement data of the area are recollected to generate a characteristic parameter set;

[0206] Comparing the feature parameter set with a preset feature threshold to identify different areas;

[0207] It should be noted that the characteristic parameters and thresholds are compared as follows: the real-time dust concentration value is compared with the preset concentration threshold (such as "≤50mg / m 3 ”) by comparison, if the dust concentration in a certain area exceeds the threshold (e.g. 60mg / m 3 ), it is marked as the “concentration exceeds the standard” area;

[0208] Compare the real-time particle size distribution data with the preset distribution range (e.g., "0.1-10 μm"). If the particle size in a certain area is abnormal (e.g., a large number of 15 μm particles appear), mark it as an "abnormal particle size" area.

[0209] Compare the real-time boundary displacement data with the displacement threshold (e.g., "±5cm"). If the displacement exceeds the range (e.g., a certain area has a displacement of 8cm), it is marked as an "exceeding displacement" area.

[0210] If multiple parameters are abnormal in a certain area at the same time (such as dust concentration exceeding the standard and boundary displacement exceeding the limit), it will be marked as a high-priority difference area;

[0211] According to the difference area, a dynamic boundary partitioning algorithm is used to adjust the coordinate boundary of the electric field control area to generate an updated partitioning scheme;

[0212] It should be noted that the dynamic boundary division algorithm includes adaptive region segmentation based on gradient change or boundary offset compensation based on fuzzy logic.

[0213] Furthermore, the step of adjusting the coordinate boundary of the electric field control area using a dynamic boundary partitioning algorithm based on the difference area to generate an updated partitioning scheme includes:

[0214] According to the difference area, the high gradient boundary is identified by the difference in dust concentration between adjacent difference areas, and the concentration gradient distribution characteristics are obtained. For example, the dust concentrations in area A and area B are 60 mg / m 3 and 30 mg / m 3 The concentration gradient at the boundary between the two is 30 mg / m 3 , which is much higher than other areas, is marked as a high gradient boundary.

[0215] Based on the boundaries of the differential region, the direction, magnitude, and temporal trend of the boundary displacement are analyzed to determine whether it is caused by airflow disturbances, temperature changes, or equipment vibration. The boundary displacement pattern is then determined. The distribution pattern of the differential region is composed of the concentration gradient distribution characteristics and the boundary displacement pattern characteristics.

[0216] Spatial statistical methods (such as gradient analysis) are used to quantify the distribution of the difference areas. Based on the distribution of the difference areas, the gradient threshold method and fuzzy logic compensation method are used to adjust the coordinate boundaries of the electric field control area.

[0217] The gradient threshold method and fuzzy logic compensation method include:

[0218] Perform gradient calculation on the spatial distribution map of dust concentration to extract areas with drastic concentration changes (such as the junction of the capture and displacement zone and the separation zone). Set the concentration gradient threshold based on historical operating data or empirical formulas, use the center line of the high gradient area as the new boundary, and generate smooth boundary coordinates through linear interpolation or spline curve fitting. Convert the boundary displacement (such as ±5% coordinate offset), displacement direction (X / Y axis component), and displacement rate into fuzzy sets (such as "small offset", "medium offset", and "large offset"). Based on the fuzzy set, establish fuzzy rules. For example, if "the displacement is a large offset" and "the direction is perpendicular to the separation interface", then "boundary compensation = displacement × preset weight parameter" to obtain the fuzzy reasoning result.

[0219] The fuzzy inference result is converted into a specific boundary coordinate compensation value to generate an updated boundary coordinate.

[0220] The boundary coordinates of the dynamic boundary partitioning algorithm are combined with power supply parameters to form an updated partitioning scheme.

[0221] In step S8, according to the updated division scheme, dust data is reacquired and cluster analysis is performed to determine the operating status of the dust collector and obtain adjustment instructions for electric field control.

[0222] In one implementation, performing cluster analysis based on the updated partitioning scheme and the environmental interference value to determine the operating state of the dust collector, outputting an adjustment instruction for fine electric field control, and completing energy-saving control includes:

[0223] Based on the updated partitioning scheme, dust concentration, particle size distribution data and environmental interference values ​​of each area of ​​the electrostatic precipitator are collected in real time to generate a real-time data set;

[0224] Calculating the power supply matching degree and regional difference value of each area according to the real-time data set to obtain matching degree data;

[0225] It should be noted that the regional difference value quantifies the differences in dust concentration and particle size distribution between adjacent or different electric field control areas. Concentration difference is calculated as the absolute value or normalized value of the concentration difference between adjacent areas. Particle size difference is obtained using the standard deviation of the particle size distribution or a chi-square test. The regional difference value is calculated by weighted summation of the concentration and particle size differences. Matching data includes the power supply matching degree of each area and the regional difference value.

[0226] If the power supply matching degree in the matching degree data is higher than a preset matching degree threshold, adjusting the electric field management parameters based on an adaptive particle swarm optimization algorithm to generate a management parameter set;

[0227] It should be noted that the steps for adjusting the electric field management parameters based on the adaptive particle swarm optimization algorithm are as follows: based on the electric field management parameters, m groups of candidate parameters (U, I, F, etc.) are randomly generated. Each parameter combination is called a "particle". The electric field control capability of the system is evaluated, and the priority is to shorten the electric field response time.

[0228] If the electric field response time exceeds the preset threshold, the score is reduced. The higher the score, the better the parameter combination (U, I, F, etc.). The best parameter combination of each particle in the historical iteration is called the individual optimal, and the parameter combination with the highest score in the entire particle group is called the global optimal.

[0229] The particles move towards the individual optimal and global optimal directions, and the speed is controlled by the inertia weight and learning factor. The supply voltage, supply current, supply frequency, etc. are adjusted according to the updated speed to generate new candidate parameters;

[0230] After the iteration is completed, the global optimal parameter is the updated power supply parameter set (U new ,I new ,F new ,...);

[0231] It should be noted that the management parameters mentioned in the adaptive particle swarm optimization algorithm to adjust the management parameters refer to the electric field control parameters that need to be optimized during the operation of the electrostatic precipitator, such as power supply voltage, power supply current, power supply frequency, dust concentration, particle size distribution data and temperature compensation coefficient.

[0232] Obtain updated dust concentration and particle size distribution data based on the management parameter set, and generate an updated data set in combination with the environmental interference value;

[0233] It should be noted that after obtaining the updated dust concentration and particle size distribution data and environmental interference values, the power supply matching degree and regional difference value of each electric field control area are recalculated;

[0234] The K-means clustering algorithm is used to classify the recalculated regional difference values ​​of each region to obtain a classification result set;

[0235] According to the classification result set, a fuzzy logic control algorithm is used to determine the change trend of the dust collector operation state to obtain a state data set;

[0236] According to the state data set, a supply voltage adjustment amount and a supply current adjustment amount are generated, and an adjustment instruction is output at the same time to perform electric field control on the dust collector to complete energy-saving control.

[0237] In one example, the use of a fuzzy logic control algorithm to determine the changing trend of the dust collector operating state can be described as follows: setting a fuzzy rule as "if the difference value is less than 0.2, the state is optimization; if the difference value is greater than 0.2, the state is adjustment", and outputting the state of each area [optimization, adjustment, optimization, optimization, ...];

[0238] It should be noted that when the state is optimization, the adjustment direction is to maintain the current voltage or make fine adjustments, and the adjustment amount is optimized in a small range according to the "optimal voltage range" in historical data; when the state is optimization, the adjustment direction is dynamically adjusted according to the difference value direction (for example, when the concentration increases, the voltage is increased, and when the concentration decreases, the voltage is decreased). The larger the difference value, the larger the adjustment range, and the final value is the current value plus the adjustment amount;

[0239] Generate specific instruction formats for different electric field control areas, send PWM signals through PLC or embedded controller to adjust the voltage, and input the final power supply value to the corresponding area;

[0240] Furthermore, the calculation formula of the power supply matching degree is:

[0241]

[0242] Among them, M represents the power supply matching degree, V actual,i Represented as the actual power supply parameter value of the i-th dimension, V optimal,i It is represented as the theoretical optimal power supply parameter value of the i-th dimension, and n is represented as the power supply parameter dimension;

[0243] Furthermore, the K-means clustering algorithm is used to classify the recalculated regional difference values ​​of each region to obtain a classification result set, including:

[0244] Dividing the regional difference value of each region into K clusters (such as high difference region, medium difference region, and low difference region) to guide electric field zoning control;

[0245] Randomly select K samples as the initial cluster centers, and for each sample x i Calculate the distance between it and all center points, assign it to the nearest cluster, and get the assignment result;

[0246] According to the distribution results, the center points of each cluster are recalculated. If the center point change is less than the threshold or the maximum number of iterations is reached, the iteration is stopped; otherwise, steps 2-3 are repeated; the output results of the iterative calculation are used as the classification result set.

[0247] Reference Figure 2 The second embodiment of the present invention provides an energy-saving control system for an electrostatic precipitator, comprising:

[0248] 101 data acquisition module, used to obtain the original data set of dust, generate the spatial distribution map of dust and perform regional division based on the original data set, and calculate the characteristic baseline data set in combination with the preset concentration threshold;

[0249] 102 a regional difference value acquisition module, configured to calculate the change data of dust distribution based on the spatial distribution map, and generate a regional difference value by combining the characteristic baseline data set with a preset change threshold;

[0250] 103 a boundary coordinate set acquisition module, configured to compare the regional difference value with a preset difference threshold, generate an adjustment signal for the regional boundary according to the comparison result, divide the electric field control area in combination with the spatial distribution map, and generate a boundary coordinate set;

[0251] 104 adjustment signal generation module, used to obtain real-time dust data according to the boundary coordinate set, calculate the difference distribution data of each electric field control area, generate a voltage adjustment signal in combination with a preset distribution threshold, and determine a current distribution value;

[0252] 105 a power supply parameter set module, configured to perform deviation calculation based on the voltage adjustment signal and the current distribution value to obtain voltage deviation distribution data, perform optimization processing, and determine a power supply parameter set;

[0253] 106, a power supply parameter set adjustment module, configured to obtain dust removal efficiency parameters according to the power supply parameter set, adjust the power supply parameters in combination with a preset efficiency parameter threshold, and obtain an adjusted power supply parameter set;

[0254] 107, a partitioning scheme updating module, configured to obtain characteristic parameters according to the adjusted power supply parameter set, and re-divide the boundary coordinates in combination with a preset characteristic threshold value to obtain an updated partitioning scheme;

[0255] The energy-saving control module 108 reacquires the dust data and performs cluster analysis according to the updated partitioning scheme, determines the operating status of the dust collector, and obtains adjustment instructions for the electric field control.

[0256] It should be noted that the energy-saving control system of an electrostatic precipitator provided in an embodiment of the present invention is used to execute all the process steps of the energy-saving control method of an electrostatic precipitator in the above embodiment. The working principles and beneficial effects of the two correspond one to one, so they will not be repeated here.

[0257] In summary, the present invention provides an energy-saving control method for an electrostatic precipitator. Compared with the prior art, the present invention has the following beneficial effects:

[0258] This method constructs a spatial distribution map of dust from an electrostatic precipitator (ESP), combines a differential algorithm to analyze the dynamic changes in concentration and particle size, and utilizes intelligent comparison of dynamic thresholds with characteristic baselines to accurately generate regional difference values. This method overcomes the limitations of traditional monitoring methods and achieves millimeter-level localization of abnormal dust deposition areas. Based on real-time dust distribution data, a multi-zone voltage / current parameter combination is constructed. A zone control algorithm, combining concentration gradients with thresholds, enhances power supply in high-load areas and optimizes energy conservation in low-load areas, forming a dynamic matching mechanism between dust distribution characteristics and power supply strategies. A particle swarm optimization algorithm and support vector machine modeling are used to quantitatively map power supply parameters to dust removal efficiency. Combined with an iterative algorithm, a closed-loop control system of "parameter update-efficiency feedback-strategy iteration" is formed. Cluster analysis and fuzzy logic control are then introduced to predict operating status trends and optimize classification. Multiple algorithms collaborate to transform environmental interference and operating condition data into dynamic control strategies. While ensuring emission stability, these algorithms achieve optimal energy consumption, improved dust removal efficiency, and adaptability to complex operating conditions. Ultimately, this achieves a multi-dimensional synergistic improvement in ESP zoning power supply optimization, refined energy efficiency management, and system stability.

[0259] An embodiment of the present invention further provides an electronic device. The electronic device includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as an energy-saving control program for an electrostatic precipitator. When the processor executes the computer program, the steps in the above-mentioned embodiments of the energy-saving control method for an electrostatic precipitator are implemented, such as Figure 1 Alternatively, when the processor executes the computer program, the functions of the modules / units in the above-mentioned device embodiments are realized, such as the data acquisition module 101 .

[0260] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, and the instruction segments are used to describe the execution process of the computer program in the electronic device.

[0261] The electronic device may be a computing device such as a desktop computer, notebook, PDA, or smart tablet. The electronic device may include, but is not limited to, a processor and memory. Those skilled in the art will appreciate that the aforementioned components are merely examples of electronic devices and do not constitute a limitation of the electronic device. The electronic device may include more or fewer components than those described above, or a combination of certain components, or different components. For example, the electronic device may also include input / output devices, network access devices, buses, and the like.

[0262] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the electronic device, connecting various parts of the entire electronic device using various interfaces and lines.

[0263] The memory can be used to store the computer programs and / or modules, and the processor realizes various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function (such as a sound playback function, an image playback function, etc.); the data storage area can store data created based on the use of the mobile phone (such as audio data, a phone book, etc.). In addition, the memory can include a high-speed random access memory and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0264] Wherein, if the module / unit integrated in the electronic device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by a computer program to instruct the relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, it can implement the steps of each of the above-mentioned method embodiments. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0265] It should be noted that the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement the present invention without inventive effort.

[0266] The specific embodiments described above further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. An energy-saving control method for an electrostatic precipitator, characterized in that: include: Obtaining an original dust data set, generating a spatial distribution map of the dust based on the original data set and performing regional division, and calculating a characteristic baseline data set based on a preset concentration threshold; Based on the spatial distribution map, calculating the change data of dust distribution, combining the characteristic baseline data set with a preset change threshold, and generating a regional difference value; Comparing the regional difference value with a preset difference threshold, generating an adjustment signal for the regional boundary according to the comparison result, dividing the electric field control region in combination with the spatial distribution map, and generating a boundary coordinate set; Acquire real-time dust data based on the boundary coordinate set, calculate differential distribution data of each electric field control area, generate a voltage adjustment signal based on a preset distribution threshold, and determine a current distribution value; performing deviation calculation based on the voltage adjustment signal and the current distribution value to obtain voltage deviation distribution data, and performing optimization processing to determine a power supply parameter set; Obtaining dust removal efficiency parameters according to the power supply parameter set, and adjusting the power supply parameters in combination with a preset efficiency parameter threshold to obtain an adjusted power supply parameter set; Acquire characteristic parameters according to the adjusted power supply parameter set, and redivide the boundary coordinates in combination with a preset characteristic threshold to obtain an updated division scheme; According to the updated partitioning scheme, dust data is reacquired and cluster analysis is performed to determine the operating status of the dust collector and obtain adjustment instructions for the electric field control; The spatial distribution map of dust is generated based on the original data set and regional division is performed. Combined with the preset concentration threshold, the characteristic baseline data set is calculated, including: The original data set includes dust concentration data and particle size distribution data, historical concentration data, historical particle size distribution data and corresponding timestamps; Extracting a certain amount of sample data from the original data set, gridding the sample data and performing interpolation calculations using a Kriging algorithm to obtain estimated values ​​for each grid point, visualizing the estimated values, and generating a spatial distribution map; Based on the concentration data and particle size distribution values ​​in the spatial distribution map, combined with a preset concentration threshold and a preset particle size threshold, regional division is performed to obtain multiple characteristic zones of dust; If the concentration data in the characteristic partition is greater than the concentration threshold, a spatial distribution map corresponding to the characteristic partition is obtained, and a convolutional neural network is used to perform feature extraction on the spatial distribution map to obtain distribution characteristics of the abnormal area; According to the distribution characteristics of the abnormal area, the sampling data of the characteristic partition is matched with the historical operating condition database to calculate the compensation baseline value of the concentration; If the concentration data in the characteristic partition is less than the concentration threshold, obtain the historical concentration data of the characteristic partition, and calculate the concentration mean in the moving time window as the statistical baseline value of the concentration; The moving time window is obtained by setting the window length and sliding step size according to the historical concentration data and timestamp; Combining the compensated baseline value with the statistical baseline value to form a characteristic baseline data set of dust; The method includes acquiring real-time dust data according to the boundary coordinate set, calculating differential distribution data of each electric field control area, generating a voltage adjustment signal in combination with a preset distribution threshold, and determining a current distribution value, including: The real-time dust data includes real-time dust concentration value and real-time particle size distribution value; Preprocessing the real-time dust data to eliminate noise interference to obtain noise-reduced dust data, and calculating regional difference values ​​based on the noise-reduced dust data to obtain difference distribution data; If the difference distribution data is greater than a preset distribution threshold, a voltage adjustment mechanism is triggered and a voltage adjustment signal is generated; Calculating the current distribution value of each electric field control area according to the voltage adjustment signal and the preset current distribution rule; If the difference distribution data is less than the preset distribution threshold, the dust collector continues to operate normally; Calculate the current distribution value of each electric field control area when the dust collector is working normally according to the rated voltage and rated power of the dust collector; The method includes performing deviation calculation based on the voltage adjustment signal and the current distribution value to obtain voltage deviation distribution data, performing optimization processing, and determining a power supply parameter set, including: Acquire monitoring data of the actual working condition of the dust collector, combine the voltage adjustment signal and the current distribution value, calculate the power supply voltage and power supply current of each electric field control area, and obtain a power supply parameter set; Calculating the deviation value of the power supply voltage based on the power supply parameter set and the operating condition change rate in the monitoring data to determine regional voltage deviation distribution data; Determine the fluctuation range data of the power supply current based on the deviation distribution data, judge the matching trend of the power supply parameters and the working conditions in combination with a preset fluctuation threshold, and generate trend distribution data; Based on the trend distribution data, establishing a mapping relationship between power supply matching and power supply parameters to obtain a first mapping data set; If the first mapping data set is greater than a preset first mapping threshold, a genetic algorithm is used to optimize the power supply voltage and the power supply current to obtain a preliminary power supply parameter set.

2. The energy-saving control method for an electrostatic precipitator according to claim 1, characterized in that: The step of calculating the change data of dust distribution based on the spatial distribution map and generating a regional difference value by combining the characteristic baseline data set with a preset change threshold comprises: According to the dust data in the spatial distribution diagram, the increase or decay rate of the concentration data per unit time is calculated based on the difference method to generate concentration change data, and the absolute difference of the particle size in adjacent time periods is counted to generate particle size offset data; The concentration change data and the particle size deviation data together constitute the dust distribution change data; If the distribution change data is greater than a preset change threshold or a baseline value corresponding to the offset, the corresponding area is determined to be an abnormal area, and a regional difference value of the abnormal area is calculated; If the distribution change data is smaller than a preset change threshold or is within the corresponding baseline value, the corresponding area is determined to be a normal area, and the area difference value is 0.

3. The energy-saving control method for an electrostatic precipitator according to claim 1, characterized in that: The step of comparing the regional difference value with a preset difference threshold, generating an adjustment signal for the regional boundary according to the comparison result, dividing the electric field control region in combination with the spatial distribution map, and generating a boundary coordinate set includes: Comparing the regional difference value of each region with a preset difference threshold, if the regional difference value is greater than the difference threshold, generating an adjustment signal for the region, and recording the coordinates of the region as abnormal coordinates; If the regional difference value is less than the difference threshold, an adjustment signal for the region is generated, and the coordinates of the region are recorded as normal coordinates; According to the adjustment signal, the system extracts all the abnormal coordinates and the normal coordinates, divides the electric field control area in combination with the spatial distribution map, and generates boundary coordinate sets of multiple independent electric field control areas.

4. The energy-saving control method for an electrostatic precipitator according to claim 1, characterized in that: The step of obtaining the dust removal efficiency parameter according to the power supply parameter set and adjusting the power supply parameter in combination with a preset efficiency parameter threshold to obtain an adjusted power supply parameter set includes: If the dust removal efficiency parameter is higher than a preset efficiency parameter threshold, the power supply parameter set is updated using a particle swarm optimization algorithm to obtain an updated power supply parameter set; According to the updated power supply parameter set, new dust removal efficiency parameters are obtained and recorded as efficiency update parameters, and the efficiency update parameters are analyzed using a support vector machine algorithm to obtain efficiency change data; Based on the efficiency change data, a mapping relationship between the power supply parameters of each area and the dust removal efficiency is established to obtain a second mapping data set; If the second mapping data set is greater than a preset second mapping threshold, the power supply parameters are iteratively adjusted according to the change direction to obtain an adjusted power supply parameter set.

5. The energy-saving control method for an electrostatic precipitator according to claim 1, characterized in that: The method of reacquiring dust data and performing cluster analysis according to the updated division scheme to determine the operating status of the dust collector and obtain adjustment instructions for electric field control includes: Based on the updated division scheme, dust concentration, particle size distribution data and environmental interference values ​​of each area of ​​the electrostatic precipitator are collected in real time to generate a real-time data set, obtain real-time power supply current and power supply voltage, and generate electric field management parameters; Calculating the power supply matching degree and regional difference value of each area according to the real-time data set to obtain matching degree data; If the power supply matching degree in the matching degree data is higher than a preset matching degree threshold, adjusting the electric field management parameters based on an adaptive particle swarm optimization algorithm to generate a management parameter set; Obtain updated dust concentration and particle size distribution data based on the management parameter set, and generate an updated data set in combination with the environmental interference value; Using a K-means clustering algorithm to classify the regional difference values ​​in the updated data set to obtain a classification result set; According to the classification result set, a fuzzy logic control algorithm is used to determine the change trend of the dust collector operation state to obtain a state data set; According to the state data set, a supply voltage adjustment amount and a supply current adjustment amount are generated, and an adjustment instruction is output at the same time to perform electric field control on the dust collector to complete energy-saving control.

6. An energy-saving control system for an electrostatic precipitator, characterized in that: The energy-saving control method for an electrostatic precipitator according to any one of claims 1 to 5 comprises: A data acquisition module is used to obtain an original dust data set, generate a spatial distribution map of the dust based on the original data set, perform regional division, and calculate a characteristic baseline data set based on a preset concentration threshold; A regional difference value acquisition module is used to calculate the change data of dust distribution based on the spatial distribution map, and generate a regional difference value by combining the characteristic baseline data set and a preset change threshold; a boundary coordinate set acquisition module, configured to compare the regional difference value with a preset difference threshold, generate an adjustment signal for the regional boundary according to the comparison result, divide the electric field control area in combination with the spatial distribution map, and generate a boundary coordinate set; an adjustment signal generation module for acquiring real-time dust data based on the boundary coordinate set, calculating differential distribution data of each electric field control area, generating a voltage adjustment signal based on a preset distribution threshold, and determining a current distribution value; A power supply parameter set module is used to perform deviation calculation based on the voltage adjustment signal and the current distribution value to obtain voltage deviation distribution data, and perform optimization processing to determine a power supply parameter set; A power supply parameter set adjustment module is used to obtain dust removal efficiency parameters according to the power supply parameter set, and adjust the power supply parameters in combination with a preset efficiency parameter threshold to obtain an adjusted power supply parameter set; An updating partitioning scheme module is used to obtain characteristic parameters according to the adjusted power supply parameter set, and to re-divide the boundary coordinates in combination with a preset characteristic threshold to obtain an updated partitioning scheme; The energy-saving control module reacquires the dust data and performs cluster analysis according to the updated partitioning scheme, determines the operating status of the dust collector, and obtains adjustment instructions for the electric field control.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the energy-saving control method for the electrostatic precipitator according to any one of claims 1 to 5.

Images