A differential pressure-based coarse filter intelligent control method

Abstract

The application discloses a kind of rough filter intelligent control method based on differential pressure, comprising: through chlor-alkali filter operating condition pressure difference intelligent research and judge platform, the real-time pressure difference of rough filter import and export, filter layer operation time length, filter material initial filling state and filter medium flow and other multidimensional data are collected, dynamic tracking filter layer blockage evolution process is carried out using filter layer blockage progressive deduction model, quantifies filter material loss change using filter material loss dynamic fitting algorithm and establishes the mapping relationship with pressure difference, determines pressure difference threshold interval and filter material loss critical value after platform processing, the optimal backwashing starting time and duration are calculated by intelligent backwashing timing scheduling algorithm, execute backwashing operation and real-time feedback pressure difference change data.The application is replaced by traditional control mode through dynamic research and closed-loop control mechanism, realizes the accurate adaptation of filter layer state and backwashing operation, improves the stability and efficiency of rough filter operation, meets the stringent requirements of chlor-alkali production.

Description

Technical Field

[0001] This invention relates to the field of filter control technology in chlor-alkali production, and in particular to an intelligent control method for coarse filters based on differential pressure. Background Technology

[0002] In the chlor-alkali production process, the coarse filter, as a key pretreatment device, directly affects the stability of subsequent production systems and product quality. During filtration, the filter bed is prone to clogging due to impurities, and the filter media wears down over time. Both of these factors lead to abnormal changes in the pressure difference between the filter inlet and outlet. Improper control can result in increased energy consumption, decreased filtration efficiency, or even shutdown. Therefore, precise control of the coarse filter based on pressure difference signals is crucial for ensuring the continuous and stable operation of chlor-alkali production. Related control methods must consider filter bed clogging identification, filter media wear monitoring, and reasonable scheduling of backwashing operations to adapt to complex and changing production conditions.

[0003] Currently, differential pressure control in coarse filters for chlor-alkali production largely employs a traditional fixed threshold triggering mode. This involves setting a single differential pressure threshold, and initiating backwashing when the measured differential pressure reaches that threshold. Some technologies combine manual experience with adjusting the backwashing duration, or periodically monitoring filter media filling levels to assess media wear and optimize backwashing frequency. During control, data acquisition relies heavily on single sensors, lacking dynamic tracking of filter clogging progression. The correlation between media wear and differential pressure changes is not systematically quantified, and the timing and duration of backwashing operations lack intelligent algorithm support, relying primarily on fixed settings based on historical operating data or static parameters.

[0004] The existing technology has two main drawbacks: First, it does not adequately consider the synergistic effects of filter layer clogging and filter media wear. It relies solely on a single differential pressure threshold to determine the timing of backwashing, failing to dynamically predict the gradual process of filter layer clogging or accurately fit the dynamic changes in filter media wear. This results in backwashing operations starting too early or too late, either wasting water and energy resources or failing to completely remove filter layer blockages. Second, the level of intelligence in control is low. It lacks a dedicated operating condition analysis platform to integrate and process multi-dimensional data. The backwashing timing is not dynamically adjusted in conjunction with real-time operating conditions, and a complete status feedback mechanism has not been established. This makes it difficult to achieve precise matching between filter layer status and backwashing operations, and fails to meet the stringent requirements of chlor-alkali production for efficient and stable operation of the filtration system. Summary of the Invention

[0005] In order to overcome the shortcomings and deficiencies of the existing technology, the present invention provides an intelligent control method for coarse filters based on differential pressure.

[0006] The technical solution adopted in this invention is an intelligent control method for a coarse filter based on differential pressure, comprising the following steps: S1, collecting real-time differential pressure data at the inlet and outlet of the coarse filter, filter bed running time data, initial filling state data of the filter media, and filter medium flow rate data through a chlor-alkali filtration operating condition differential pressure intelligent judgment platform; S2, based on the collected multi-dimensional data, dynamically extrapolating the degree of filter bed clogging using a progressive extrapolation model of filter bed clogging, and generating filter bed clogging evolution trend data by analyzing the correlation between the rate of change of differential pressure and the porosity of the filter bed; S3, using a dynamic fitting algorithm for filter media loss to fit and analyze the loss state of the filter media during the filtration process, combined with the filter bed clogging... S4. Using the filter media evolution trend data, establish a mapping relationship between filter media wear and pressure difference changes; S5. Analyze and process the extrapolated and fitted data through the intelligent differential pressure judgment platform for chlor-alkali filtration conditions to determine the differential pressure threshold range and the critical value of filter media wear; S6. Based on the determined differential pressure threshold range and critical value of filter media wear, calculate the optimal backwash start time and backwash duration using an intelligent backwash timing scheduling algorithm; S7. Perform backwashing operation on the coarse filter according to the calculated optimal backwash start time and backwash duration, while simultaneously feeding back the differential pressure change data during the backwashing process to the intelligent differential pressure judgment platform for chlor-alkali filtration conditions in real time, forming a dynamic adaptation between the filter layer state and the backwashing operation.

[0007] Furthermore, the expression for the progressive deduction model of filter layer clogging is as follows: , in, The local clogging coefficient of the filter layer. Let i be the weighting coefficient for the influence of pressure difference in the i-th group. Let be the pressure difference change during the i-th sampling period. The time influence weighting coefficient for the i-th acquisition period is given by [reference]. Let i be the runtime of the i-th acquisition cycle. The total clogging coefficient of the filter bed. For local congestion coordination coefficient, This is the actual pressure difference correction factor. This is the real-time measured differential pressure value. denoted as the initial porosity of the filter layer, and n as the total number of sampling cycles.

[0008] Furthermore, the expression for the dynamic fitting algorithm for filter media loss is: , in, This refers to the amount of filter media lost per unit time. This represents the weighting coefficient for filter media loss in stage k. The average pressure difference in stage k. Let k be the duration of the k-th stage. This is the cumulative filter media loss. This is the integral correction factor for loss. Total runtime The correlation coefficient for blockage loss is... The total clogging coefficient of the filter bed. is the current porosity of the filter layer, and m is the total number of operating stages divided by filter media loss.

[0009] Furthermore, the expression for the intelligent backwashing timing scheduling algorithm is: , in, These are the backwash startup timing parameters. These are the main correction coefficients for the scheduling algorithm. Let be the real-time pressure difference at time j. To preset the differential pressure threshold, Let be the filter media loss at time j. This is a preset critical value for filter media loss. The congestion state influence coefficient. This represents the total clogging coefficient of the filter layer.

[0010] Furthermore, the intelligent differential pressure judgment platform for chlor-alkali filtration is equipped with a multi-channel differential pressure sensing module, a high-speed data acquisition card, and an edge computing unit. The sensing module uses a high-precision piezoresistive differential pressure sensor with a sampling frequency set to 50Hz to 100Hz. The data acquisition card has a resolution of no less than 16 bits. The edge computing unit has a built-in FPGA chip for real-time data preprocessing. The platform software sets the differential pressure acquisition cycle to 2s to 5s, the filter layer blockage evolution trend data update cycle to 30s, the filter media loss fitting analysis interval to be consistent with the differential pressure acquisition cycle, the intelligent backwashing timing scheduling algorithm calculation cycle to be no more than 1s, the backwashing operation execution signal response delay to be no more than 50ms, and the differential pressure threshold range can be dynamically adjusted according to the filter media type and production load, with an adjustment step size of 0.01 to 0.05.

[0011] Further, step S2 includes the following sub-steps: S21, sorting the differential pressure data and runtime data collected by the chlor-alkali filtration operating condition differential pressure intelligent judgment platform according to the time series, dividing it into multiple continuous data acquisition cycles, and extracting the maximum, minimum, and average differential pressure values ​​within each cycle; S22, calling the filter layer clogging progressive deduction model, inputting the differential pressure characteristic data and runtime data of each acquisition cycle, and determining the differential pressure influence weight coefficient and time influence weight coefficient corresponding to each cycle; S23, calculating the local clogging coefficient of each acquisition cycle through the model, and deriving the total clogging coefficient of the filter layer by combining the real-time measured differential pressure value and the initial porosity of the filter layer; S24, based on the changes in the total clogging coefficient of multiple continuous cycles, analyzing the data change pattern using the sliding window method, and generating filter layer clogging evolution trend data.

[0012] Further, step S3 includes the following sub-steps: S31, extracting historical differential pressure data, initial filling state data of filter media, and filter layer clogging evolution trend data generated in S2 from the chlor-alkali filtration operating condition differential pressure intelligent judgment platform, and dividing the data into groups according to the operating stage; S32, starting the filter media loss dynamic fitting algorithm, and setting the corresponding filter media loss weight coefficient according to the differential pressure change characteristics of different operating stages; S33, calculating the filter media loss per unit time in each operating stage through the algorithm, and performing integral calculation in combination with the total operating time to obtain the cumulative filter media loss; S34, establishing a mapping relationship model between the cumulative filter media loss and the total clogging coefficient of the filter layer, inputting the current filter layer porosity data, correcting the mapping relationship parameters, and outputting the correlation results between filter media loss and differential pressure change.

[0013] Further, step S4 includes the following sub-steps: S41, receiving the filter layer clogging evolution trend data from S2 and the correlation results of filter media loss and differential pressure change from S3, and transmitting them to the data processing module of the chlor-alkali filtration operating condition differential pressure intelligent judgment platform; S42, the platform performs outlier removal processing on the input data, retains valid data and sorts it according to priority, with priority determined based on the interval between the data acquisition time and the current time; S43, based on the valid data, analyzing the synergistic relationship between the filter layer clogging rate and the differential pressure growth rate, and combining the critical value reference range of filter media loss, initially defining the differential pressure threshold range; S44, through the platform's built-in threshold calibration module, combined with the optimal control parameters in historical operating data, adjusting the initially defined differential pressure threshold range, and finally determining the precise differential pressure threshold range and the critical value of filter media loss.

[0014] Further, S5 includes the following sub-steps: S51, retrieving the determined differential pressure threshold range and filter media loss critical value from the chlor-alkali filtration operating condition differential pressure intelligent judgment platform, and simultaneously collecting the real-time differential pressure data, filter media loss data, and total filter layer clogging coefficient at the current moment; S52, initializing various parameters of the intelligent backwashing timing scheduling algorithm, including the scheduling algorithm main correction coefficient and clogging state influence coefficient, and adjusting the parameter values ​​according to the current production operating conditions; S53, inputting the real-time collected data and determined threshold parameters into the algorithm, and calculating the backwashing start-up timing parameters through product calculation and coefficient correction; S54, based on the backwashing start-up timing parameters and combined with the filter layer clogging evolution trend, deriving the backwashing duration to ensure accurate matching between the backwashing operation and the filter layer contamination state.

[0015] A pressure differential-based intelligent control method for a coarse filter is disclosed. This method is implemented through a pressure differential-based intelligent control system for a coarse filter in chlor-alkali production. The system includes: a multi-dimensional data acquisition unit for chlor-alkali filtration operation, a filter bed clogging progressive calculation unit, a filter media loss dynamic fitting analysis unit, a pressure differential threshold and loss critical value judgment unit, an intelligent backwashing timing scheduling calculation unit, and a backwashing execution and status feedback unit. The multi-dimensional data acquisition unit for chlor-alkali filtration operation is connected to the pressure differential sensing component and flow detection component of the coarse filter, and is used to collect inlet and outlet pressure differentials, operating time, initial state of the filter media, and media flow data, and transmit them to the filter bed clogging progressive calculation unit. The filter bed clogging progressive calculation unit receives the collected data and, through a built-in module... The filter media loss dynamic fitting analysis unit simulates the filter layer clogging trend and sends it to the filter media loss dynamic fitting analysis unit. This unit combines the clogging trend data with an algorithm to fit the filter media loss state and transmits it to the differential pressure threshold and loss critical value judgment unit. The differential pressure threshold and loss critical value judgment unit processes the received data, determines the threshold range and critical value, and sends it to the intelligent backwashing timing scheduling calculation unit. The intelligent backwashing timing scheduling calculation unit calculates the optimal backwashing parameters based on the threshold data and sends them to the backwashing execution and status feedback unit. The backwashing execution and status feedback unit receives the parameters and controls the coarse filter to perform backwashing, while simultaneously collecting differential pressure data during the backwashing process and feeding it back to the chlor-alkali filtration operating condition multi-dimensional data acquisition unit, forming a complete control link.

[0016] Beneficial Effects: This invention proposes an intelligent control method for coarse filters based on differential pressure. By integrating multi-dimensional operational data and utilizing technologies related to progressive filter clogging simulation, dynamic fitting of filter media loss, and intelligent backwashing timing scheduling, coupled with a dedicated intelligent differential pressure analysis platform, a full-process intelligent control mechanism is constructed, encompassing data acquisition, status analysis, and backwashing control. Addressing the issue that existing technologies do not consider the synergistic effects of filter clogging and filter media loss, this method dynamically simulates the progressive process of filter clogging, accurately fits the dynamic changes in filter media loss, and quantifies the correlation between these two factors and differential pressure. This replaces the single fixed differential pressure threshold judgment mode, avoiding backwashing that is too early or too late. This method addresses the issues of resource waste and incomplete clogging caused by late cleaning. To mitigate the shortcomings of existing technologies with low levels of intelligence, a dedicated operating condition analysis platform integrates multi-dimensional data. Intelligent algorithms dynamically schedule backwashing sequences, coupled with a complete status feedback mechanism, to achieve precise adaptation between backwashing operations and the real-time status of the filter layer, thus improving the level of intelligent control. Furthermore, by refining the control process step-by-step and optimizing hardware configuration and software parameters, this method further ensures the stability and efficiency of the coarse filter operation, reduces equipment energy consumption, minimizes downtime, and meets the stringent operational requirements of chlor-alkali production for the filtration system, providing reliable assurance for production continuity and product quality. Attached Figure Description

[0017] Figure 1This is a flowchart illustrating the overall process of the method of the present invention.

[0018] Figure 2 This is a flowchart of method step S2 of the present invention;

[0019] Figure 3 This is a flowchart of method step S3 of the present invention;

[0020] Figure 4 This is a flowchart of method step S4 of the present invention;

[0021] Figure 5 This is a flowchart of step S5 of the method of the present invention. Detailed Implementation

[0022] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0023] like Figure 1 As shown, a pressure differential-based intelligent control method for a coarse filter includes the following steps: S1, collecting real-time pressure differential data at the inlet and outlet of the coarse filter, filter bed running time data, initial filling state data of the filter media, and filter medium flow rate data through a chlor-alkali filtration operating condition pressure differential intelligent judgment platform; S2, based on the collected multi-dimensional data, dynamically extrapolating the degree of filter bed clogging using a filter bed clogging progressive extrapolation model, and generating filter bed clogging evolution trend data by analyzing the correlation between the pressure differential change rate and the filter bed porosity; S3, using a filter media loss dynamic fitting algorithm to fit and analyze the loss state of the filter media during the filtration process, combined with the filter bed clogging evolution trend. S4. Data is used to establish a mapping relationship between filter media wear and pressure difference changes; S5. The deduced and fitted data is analyzed and processed through the intelligent differential pressure judgment platform for chlor-alkali filtration conditions to determine the differential pressure threshold range and the critical value of filter media wear; S6. Based on the determined differential pressure threshold range and the critical value of filter media wear, the optimal backwashing start time and backwashing duration are calculated through an intelligent backwashing timing scheduling algorithm; S7. Backwashing operation is performed on the coarse filter according to the calculated optimal backwashing start time and backwashing duration, while the differential pressure change data during the backwashing process is fed back to the intelligent differential pressure judgment platform for chlor-alkali filtration conditions in real time, forming a dynamic adaptation between the filter layer state and the backwashing operation.

[0024] The implementation process of step S1 involves building a multi-dimensional data acquisition system through the intelligent judgment platform for differential pressure in chlor-alkali filtration operation. The multi-channel differential pressure sensing module on the platform adopts a high-precision piezoresistive differential pressure sensor, with a sampling frequency set to 80Hz and a data acquisition card resolution set to 16 bits to ensure the accuracy and real-time performance of data acquisition. The core data collected includes real-time differential pressure data at the inlet and outlet of the coarse filter, filter bed runtime data, initial filling status data of the filter media, and filter medium flow rate data. The differential pressure data collection cycle is set to 3 seconds, recording the instantaneous value of the inlet and outlet differential pressure every 3 seconds. The filter bed runtime is accumulated from the start-up time of the filter, timed and stored in 1-second increments. The initial filling status data of the filter media is collected by detection components preset at different depths of the filter media layer, recording the initial bulk density and distribution uniformity of the filter media. The filter medium flow rate data collection frequency is consistent with that of the differential pressure data, once every 3 seconds. All collected data is transmitted in real time to the edge computing unit of the platform for preliminary data caching and format standardization. This step provides comprehensive and accurate basic data support for subsequent filter bed clogging simulation, filter media loss fitting, and backwashing scheduling, ensuring that subsequent algorithm models can be calculated based on real and effective data. The parameter values ​​are set by comprehensively considering the dynamic change rate of chlor-alkali production conditions and the response capability of the sensors, avoiding data redundancy caused by excessively high sampling frequency and preventing data lag caused by excessively low sampling frequency.

[0025] Step S2 involves using the multi-dimensional data collected and preliminarily organized in S1 to initiate a progressive deduction model for filter layer clogging to dynamically predict the degree of clogging. First, the collected differential pressure data is sorted by time series, dividing it into continuous collection cycles. Each cycle is set to 30 seconds. The maximum, minimum, and average values ​​of the differential pressure within each cycle are extracted. Simultaneously, the filter layer runtime data for the corresponding cycle is extracted. These data are used as input parameters for the model. The model analyzes the correlation between the rate of change of differential pressure and the porosity of the filter layer, and combines the preset weighting coefficients for the influence of differential pressure and time in each cycle to gradually predict the change process of the filter layer clogging state. During the simulation, a local clogging coefficient of the filter layer is generated every 30 seconds, and a total clogging coefficient of the filter layer is generated every 5 acquisition cycles (i.e., 150 seconds). At the same time, based on the total clogging coefficient data of 10 consecutive acquisition cycles, a filter layer clogging evolution trend curve is generated. This step captures the gradual change pattern of filter layer clogging, breaking the limitation of traditional methods that judge the clogging state only by a single differential pressure threshold. Through dynamic simulation, the development trend of filter layer clogging can be predicted in advance, providing forward-looking data support for subsequent filter media loss analysis and backwashing timing judgment. In terms of parameter settings, the matching of acquisition cycle and simulation cycle can balance simulation accuracy and computational efficiency, ensuring that clogging changes are reflected in a timely manner while avoiding excessive computational load.

[0026] Step S3 involves activating the dynamic fitting algorithm for filter media loss. Based on the initial filling state data and flow rate data of the filter media collected in S1, and the filter layer clogging evolution trend data generated in S2, a filter media loss analysis model is constructed. First, the data is divided into groups according to the filter's operating stages, with each stage set to 1800 seconds (30 minutes). For each stage, corresponding filter media loss weighting coefficients are set based on the pressure difference change characteristics, flow rate fluctuations, and clogging evolution data. The coefficient values ​​are dynamically adjusted according to the production load of different stages, ranging from 0.02 to 0.15. The algorithm analyzes the synergistic relationship between pressure difference change and flow rate data in each stage, combined with the filter layer clogging evolution trend, to calculate the filter media loss per unit time in each stage. The cumulative filter media loss is then obtained through cumulative calculation. Simultaneously, a mapping relationship model between the cumulative filter media loss and pressure difference change is established. The parameters of the mapping relationship model are corrected every 300 seconds to ensure the model's adaptability. This step quantifies the loss status of filter media during the filtration process, determines the intrinsic relationship between filter media loss and pressure difference changes, and makes up for the shortcomings of traditional methods in accurately monitoring filter media loss. It provides key data support for subsequent determination of critical values ​​for filter media loss and backwash scheduling. In terms of parameter settings, the division of operating stages and the setting of correction cycles fully consider the gradual nature of filter media loss and the dynamic changes in operating conditions in chlor-alkali production, ensuring the accuracy and timeliness of the fitting results.

[0027] Step S4 involves using the data analysis and processing module of the chlor-alkali filtration operating condition differential pressure intelligent judgment platform to perform in-depth processing on the filter layer clogging evolution trend data generated in S2 and the data on the mapping relationship between filter media loss and differential pressure change obtained in S3. First, outliers are removed from the input data. The 3σ criterion is used to judge and remove data points exceeding a reasonable range, retaining valid data and prioritizing it according to the order of acquisition time, with data closer to the current moment having higher priority. Then, based on the valid data, the synergistic change law between the filter layer clogging rate and the differential pressure growth rate is analyzed. Combined with the process requirements of chlor-alkali production, the upper and lower limits of the differential pressure threshold range are initially defined. The lower limit reference value is set as the average differential pressure value based on the initial state of the filter media and normal operating conditions, while the upper limit reference value is set as the differential pressure value corresponding to ensuring that the filtration efficiency is not lower than 90% of the design value. Next, the platform's built-in threshold calibration module calls upon the optimal control parameters from historical operating data to iteratively adjust the initially defined differential pressure threshold range. The adjustment step size is set to 0.03. Simultaneously, combined with the cumulative filter media loss calculated by S3, the critical value for filter media loss is determined. This critical value is set at 30% of the initial filter media filling amount. This step, through scientific data analysis and calibration, determines precise control threshold parameters, providing a definite judgment standard for subsequent backwashing scheduling and avoiding the limitations of traditional fixed threshold settings. In terms of parameter settings, the determination of outlier removal criteria, adjustment step size, and critical value ratio are all based on statistical analysis of a large amount of actual operating data from chlor-alkali production, ensuring the rationality and applicability of the threshold parameters.

[0028] Step S5 involves using the differential pressure threshold range and filter media wear threshold determined in S4 to initiate an intelligent backwashing timing and scheduling algorithm to calculate the optimal backwashing start time and duration. First, real-time data on the inlet and outlet differential pressure of the coarse filter, filter media wear, and total filter layer clogging coefficient are collected. This real-time data is compared with the threshold parameters determined in S4. When the real-time differential pressure reaches the upper limit of the differential pressure threshold range, or the cumulative filter media wear reaches the filter media wear threshold, the algorithm initiates the calculation process. The algorithm integrates data from multiple sources, including the difference between real-time differential pressure and the threshold, the difference between filter media loss and the critical value, the filter layer clogging rate, and the filter media flow rate. Combined with a preset scheduling algorithm master correction coefficient and a clogging state influence coefficient, it calculates the backwash start-up timing parameters through multi-parameter collaborative computation. These parameters directly correspond to the specific time of backwash initiation. Simultaneously, based on the total filter layer clogging coefficient and cumulative filter media loss, the backwash duration is derived. The calculated backwash duration ranges from 30s to 120s, with the specific value dynamically adjusted according to the degree of clogging and loss. More severe clogging and greater loss result in a longer duration, but not exceeding the upper limit of 120s. This step achieves precise timing control of the backwash operation, breaking away from the traditional fixed-time or fixed-differential-pressure triggering mode. It ensures that the backwash operation starts at the most suitable time and has a reasonable duration, guaranteeing both backwash effectiveness and avoiding resource waste. The parameter setting limits the duration range based on a balance analysis between filter media regeneration efficiency and backwash resource consumption.

[0029] Step S6 involves using the optimal backwash start-up timing and duration calculated in S5. The control output module of the chlor-alkali filtration operating condition differential pressure intelligent judgment platform sends a control signal to the backwash actuator of the coarse filter. Upon receiving the signal, the backwash actuator immediately initiates the backwash operation. The backwash operation employs a combination of air washing followed by water washing. The gas flow rate during the air washing stage is set to 15 m / s, and the liquid flow rate during the water washing stage is set to 8 m / s. The switching between air washing and water washing is automatically controlled by a solenoid valve, with a switching time set to 5 seconds. During the backwash operation, the platform's differential pressure sensing module continuously collects differential pressure change data at the inlet and outlet of the coarse filter at a frequency of 80 Hz, transmitting it in real-time to the edge computing unit. The edge computing unit analyzes the data in real-time, determining the rate and trend of differential pressure decrease during backwashing. When the differential pressure drops to the lower limit of the differential pressure threshold range and remains stable for more than 10 seconds, the backwash operation can be terminated early. If this threshold is not reached, the backwash is completed for the calculated duration. After backwashing, the platform stores all data from the backwashing process and updates the data on filter media loss and filter layer clogging status, providing updated basic data for the next round of filtration and backwashing control. This step effectively removes filter layer blockages and restores the filtration performance of the filter media by precisely executing the backwashing operation. At the same time, a closed-loop control system is built through real-time feedback and data updates to ensure the continuity and adaptability of the entire control process. In terms of parameter settings, the flow rate and switching time of air washing and water washing are determined based on the material characteristics of the filter media and the need to remove blockages, taking into account both backwashing effect and equipment protection.

[0030] Preferably, the expression for the progressive deduction model of filter layer clogging is: , in, The local clogging coefficient of the filter layer. Let i be the weighting coefficient for the influence of pressure difference in the i-th group. Let be the pressure difference change during the i-th sampling period. The time influence weighting coefficient for the i-th acquisition period is given by [reference]. Let i be the runtime of the i-th acquisition cycle. The total clogging coefficient of the filter bed. For local congestion coordination coefficient, This is the actual pressure difference correction factor. This is the real-time measured differential pressure value. denoted as the initial porosity of the filter layer, and n as the total number of sampling cycles.

[0031] Specifically, the progressive derivation model for filter bed clogging is based on the progressive characteristics of filter bed clogging in chlor-alkali production. Combining the synergistic effects of multi-cycle pressure difference changes and operating time, it first analyzes the correlation between pressure difference changes and filter bed clogging within different sampling cycles to determine the pressure difference influence weighting coefficient for each cycle. Simultaneously, based on the attenuation effect of operating time on filter bed porosity, a time influence weighting coefficient is set. An integral operation is used to integrate the synergistic effects of each cycle, forming a calculation expression for the local clogging coefficient of the filter bed. This derivation logic conforms to the evolutionary law of filter bed clogging from local to overall. Because the clogging effects of different cycles have a cumulative effect, the integral operation can accurately reflect the degree of clogging after the superposition of multiple factors. Based on this, an actual pressure difference correction coefficient and the initial porosity of the filter bed are introduced to construct an expression for the total clogging coefficient of the filter bed, realizing the correlation conversion between local and overall clogging. The derivation process fully considers the reference value of measured pressure differences and the benchmark role of initial porosity in actual operating conditions. The weighting coefficients for differential pressure and time are set according to the fluctuation range of chlor-alkali production load, both ranging from 0.05 to 0.3. The local clogging synergy coefficient is set to 0.85, and the actual differential pressure correction coefficient is set to 1.2. These values ​​are based on statistical analysis of a large amount of historical operating data to ensure that the model is adaptable to different production conditions. The implementation method involves inputting the collected multi-cycle differential pressure data and runtime data into the model, calculating the local and total clogging coefficients step by step according to the derivation logic, quantifying the dynamic changes in filter layer clogging, providing a scientific basis for subsequent backwashing decisions, and solving the problem that traditional methods cannot dynamically track the evolution of clogging.

[0032] Preferably, the expression for the dynamic fitting algorithm for filter media loss is: , in, This refers to the amount of filter media lost per unit time. This represents the weighting coefficient for filter media loss in stage k. The average pressure difference in stage k. Let k be the duration of the k-th stage. This is the cumulative filter media loss. This is the integral correction factor for loss. Total runtime The correlation coefficient for blockage loss is... The total clogging coefficient of the filter bed. is the current porosity of the filter layer, and m is the total number of operating stages divided by filter media loss.

[0033] Specifically, the dynamic fitting algorithm for filter media loss is based on the coupling relationship between filter media loss and pressure difference changes and operating time. First, the data is divided into groups according to operating stages. The contribution of the average pressure difference and duration of each stage to filter media loss is analyzed, and corresponding weighting coefficients for filter media loss are set for each stage. A summation operation is used to accumulate the filter media loss per unit time for each stage. This derivation is based on the differential characteristics of filter media loss in different operating stages, and the summation operation can accurately aggregate the cumulative loss effect of each stage. Further considering the continuity of loss within the total operating time, a loss integral correction coefficient is introduced. Through integration, the loss per unit time is converted into the cumulative filter media loss. Simultaneously, combining the correlation between the total clogging coefficient of the filter layer and the current porosity, a clogging loss correlation coefficient is introduced to construct a complete expression for calculating the cumulative loss. The calculation process takes into account both the stage-based and continuous nature of loss, ensuring the comprehensiveness of the fitting results. The filter media loss weighting coefficient ranges from 0.03 to 0.2, the loss integral correction coefficient is set to 0.9, and the clogging loss correlation coefficient is set to 1.15. These values ​​were obtained through experimental data calibration and are adapted to the loss characteristics of filter media in chlor-alkali production. The implementation method involves extracting relevant data according to the operation stage, calculating the loss amount by summation and integration operations in sequence, fitting the dynamic changes of filter media loss, determining the intrinsic relationship between loss and differential pressure and clogging status, providing data support for setting critical values ​​for filter media loss, and making up for the shortcomings of traditional methods that cannot quantify filter media loss.

[0034] Preferably, the expression for the intelligent backwashing timing scheduling algorithm is: , in, These are the backwash startup timing parameters. These are the main correction coefficients for the scheduling algorithm. Let be the real-time pressure difference at time j. To preset the differential pressure threshold, Let be the filter media loss at time j. This is a preset critical value for filter media loss. The congestion state influence coefficient. This represents the total clogging coefficient of the filter layer.

[0035] Specifically, the intelligent backwashing timing scheduling algorithm is based on the matching relationship between backwashing start-up timing and differential pressure threshold, and filter media wear threshold. First, it analyzes the ratio of real-time differential pressure to the preset differential pressure threshold, and the ratio of real-time filter media wear to the preset threshold. These two ratios reflect the degree of differential pressure exceeding limits and the degree of wear exceeding limits, respectively. A product operation is used to integrate the synergistic influence of these two ratios. This derivation logic is based on the practical need to consider both differential pressure and filter media wear when determining backwashing timing; the product operation can reflect the superposition effect of these two factors. A main correction coefficient for the scheduling algorithm is introduced to adjust the influence weight of the two ratios. Simultaneously, combined with the correction effect of the total filter layer clogging coefficient on the backwashing timing, a clogging state influence coefficient is introduced to construct the calculation expression for the backwashing start-up timing parameters. The derivation process fully considers the comprehensive influence of multiple factors on the backwashing effect in chlor-alkali production, ensuring the accuracy of the scheduling results. The main correction coefficient for the scheduling algorithm is set to 0.75, and the clogging state influence coefficient is set to 0.25. These values ​​are determined through numerical simulation optimization, balancing the influence weights of differential pressure, wear, and clogging state. The implementation method involves collecting real-time differential pressure, filter media loss, and total filter layer clogging coefficient, inputting these data into an algorithm to calculate backwash start-up timing parameters, thereby achieving intelligent scheduling of the backwash sequence. This ensures that the backwash operation is precisely matched with the filter layer status, solves the lag problem of traditional fixed timing scheduling, and improves backwash efficiency and resource utilization.

[0036] Preferably, the intelligent differential pressure judgment platform for chlor-alkali filtration is equipped with a multi-channel differential pressure sensing module, a high-speed data acquisition card, and an edge computing unit. The sensing module uses a high-precision piezoresistive differential pressure sensor with a sampling frequency set to 50Hz to 100Hz. The data acquisition card has a resolution of no less than 16 bits. The edge computing unit has a built-in FPGA chip for real-time data preprocessing. The platform software sets the differential pressure acquisition cycle to 2s to 5s, the filter layer blockage evolution trend data update cycle to 30s, the filter media loss fitting analysis interval to be consistent with the differential pressure acquisition cycle, the intelligent backwashing timing scheduling algorithm calculation cycle to be no more than 1s, the backwashing operation execution signal response delay to be no more than 50ms, and the differential pressure threshold range can be dynamically adjusted according to the filter media type and production load, with an adjustment step size of 0.01 to 0.05.

[0037] Preferred, such as Figure 2As shown, step S2 includes the following sub-steps: S21, sorting the differential pressure data and runtime data collected by the chlor-alkali filtration operating condition differential pressure intelligent judgment platform according to the time series, dividing into multiple continuous data acquisition cycles, and extracting the maximum, minimum, and average differential pressure values ​​within each cycle; S22, calling the filter layer clogging progressive deduction model, inputting the differential pressure characteristic data and runtime data of each acquisition cycle, and determining the differential pressure influence weight coefficient and time influence weight coefficient corresponding to each cycle; S23, calculating the local clogging coefficient of each acquisition cycle through the model, and deriving the total clogging coefficient of the filter layer by combining the real-time differential pressure measurement value and the initial porosity of the filter layer; S24, based on the changes in the total clogging coefficient of multiple continuous cycles, analyzing the data change pattern using the sliding window method, and generating filter layer clogging evolution trend data.

[0038] Specifically, step S2 includes: S21 First, the differential pressure data and runtime data collected by the intelligent differential pressure analysis platform for chlor-alkali filtration operation are arranged in an orderly manner according to the time series, dividing into continuous data acquisition cycles, with each acquisition cycle set to 30 seconds. The maximum, minimum, and average values ​​of the differential pressure within each cycle are extracted through a data filtering mechanism to ensure the integrity of the data characteristics; S22 Call the filter layer clogging progressive extrapolation model, inputting the differential pressure characteristic data and runtime data of each acquisition cycle into the model according to a preset format. The model dynamically adjusts the differential pressure influence weight coefficient and time influence coefficient corresponding to each cycle according to the chlor-alkali production load. The impact weighting coefficients range from 0.05 to 0.3, adapting to the clogging impact patterns under different operating conditions. S23 calculates the local clogging coefficient for each acquisition cycle according to the model's calculation logic, and then, combined with the real-time differential pressure measurement value and the initial porosity data of the filter layer, derives the total clogging coefficient of the filter layer through coefficient correction. The correction coefficient is set to 1.2 to ensure calculation accuracy. S24 uses the sliding window method to perform data trend analysis based on the total clogging coefficient change data over 10 consecutive acquisition cycles. The window size is set to 5 cycles, generating filter layer clogging evolution trend data and storing it in the platform database.

[0039] Preferred, such as Figure 3 As shown, step S3 includes the following sub-steps: S31, extracting historical differential pressure data, initial filling state data of filter media, and filter layer clogging evolution trend data generated in S2 from the intelligent differential pressure judgment platform for chlor-alkali filtration operation, and dividing the data into groups according to the operating stage; S32, starting the dynamic fitting algorithm for filter media loss, and setting the corresponding filter media loss weight coefficient according to the differential pressure change characteristics of different operating stages; S33, calculating the filter media loss per unit time in each operating stage through the algorithm, and performing integral calculation in combination with the total operating time to obtain the cumulative filter media loss; S34, establishing a mapping relationship model between the cumulative filter media loss and the total clogging coefficient of the filter layer, inputting the current filter layer porosity data, correcting the mapping relationship parameters, and outputting the correlation results between filter media loss and differential pressure change.

[0040] Specifically, step S3 consists of four sub-steps: S31 Extracts historical differential pressure data, initial filter media filling status data, and filter layer clogging evolution trend data generated by weight 6 from the database of the chlor-alkali filtration operating condition differential pressure intelligent judgment platform. The data is then divided into groups according to the filter's operating stage, with each stage set to 1800 seconds (30 minutes) to ensure the data grouping matches the operating condition change cycle; S32 Activates the filter media loss dynamic fitting algorithm. Based on the differential pressure change amplitude and flow fluctuation characteristics of different operating stages, the algorithm automatically sets the corresponding filter media loss weight coefficients, with the coefficient values ​​ranging from... The coefficient ranges from 0.03 to 0.2, remaining stable within each stage. S33, through algorithmic logic, first calculates the filter media loss per unit time in each operating stage, then, using the total operating time as the integration interval, converts the loss per unit time into the cumulative filter media loss through integration. The integration step size is set to 30 seconds to balance computational efficiency and accuracy. S34, based on the correlation analysis between the cumulative filter media loss and the total filter layer clogging coefficient, establishes a mapping relationship model. Inputting the current filter layer porosity data, the mapping relationship parameters are dynamically corrected every 300 seconds, outputting the correlation results between filter media loss and pressure difference changes. This step-by-step approach achieves accurate fitting of filter media loss, with parameter values ​​experimentally calibrated. The implementation process ensures real-time data transmission at each step, providing a scientific basis for determining the critical value of filter media loss and overcoming the shortcomings of traditional methods in quantifying loss.

[0041] Preferred, such as Figure 4 As shown, step S4 includes the following sub-steps: S41, receiving the filter layer clogging evolution trend data from S2 and the correlation results of filter media loss and differential pressure change from S3, and transmitting them to the data processing module of the chlor-alkali filtration operating condition differential pressure intelligent judgment platform; S42, the platform performs outlier removal processing on the input data, retains valid data and sorts it according to priority, with priority determined based on the interval between the data acquisition time and the current time; S43, based on the valid data, analyzing the synergistic relationship between the filter layer clogging rate and the differential pressure growth rate, and combining the critical value reference range of filter media loss, initially defining the differential pressure threshold range; S44, through the platform's built-in threshold calibration module, combined with the optimal control parameters in historical operating data, adjusting the initially defined differential pressure threshold range, and finally determining the precise differential pressure threshold range and the critical value of filter media loss.

[0042] Specifically, step S4 includes: S41 Receiving filter layer clogging evolution trend data and the correlation results between filter media loss and differential pressure change (weight 7), transmitting them to the data processing module of the chlor-alkali filtration operating condition differential pressure intelligent judgment platform via a data transmission interface, with the transmission rate set to 10Mbps to ensure data transmission efficiency; S42 The platform data processing module uses the 3σ criterion to process outliers in the input data, removing data points exceeding the mean plus or minus 3 times the standard deviation, retaining valid data, and determining priority according to the interval between the acquisition time and the current time, with data with an interval less than 60s set as the highest priority; S43 Based on the valid data, using... Correlation analysis algorithms were used to analyze the synergistic relationship between filter media clogging rate and differential pressure growth rate. Combined with the minimum filtration efficiency requirements of the chlor-alkali production process, a preliminary differential pressure threshold range was defined. The lower limit was set at 1.2 times the average differential pressure under normal operating conditions, and the upper limit was set at the differential pressure required to ensure a filtration efficiency of no less than 90%. The S44 platform, through its built-in threshold calibration module, used the optimal control parameters from nearly 30 days of historical operating data to iteratively adjust the initially defined differential pressure threshold range. The adjustment step size was set to 0.03, ultimately determining the precise differential pressure threshold range and the critical value for filter media loss. The critical value was set at 30% of the initial filter media filling volume. Through this threshold assessment process, parameter settings were based on process requirements and historical data. The implementation method emphasized the rigor of data processing, ensuring that the determined threshold parameters were practical and adaptable, providing a definitive basis for backwashing scheduling.

[0043] Preferred, such as Figure 5 As shown, S5 includes the following sub-steps: S51, retrieve the determined differential pressure threshold range and filter media loss critical value from the chlor-alkali filtration operating condition differential pressure intelligent judgment platform, and simultaneously collect the real-time differential pressure data, filter media loss data, and total filter layer clogging coefficient at the current moment; S52, initialize various parameters of the intelligent backwashing timing scheduling algorithm, including the scheduling algorithm main correction coefficient and clogging state influence coefficient, and adjust the parameter values ​​according to the current production operating conditions; S53, input the real-time collected data and determined threshold parameters into the algorithm, and calculate the backwashing start-up timing parameters through product calculation and coefficient correction; S54, based on the backwashing start-up timing parameters and combined with the filter layer clogging evolution trend, deduce the backwashing duration to ensure that the backwashing operation is accurately matched with the filter layer contamination state.

[0044] Specifically, the implementation process of step S5 is as follows: S51 retrieves the differential pressure threshold range and filter media wear threshold determined by weight 8 from the real-time database of the intelligent differential pressure judgment platform for chlor-alkali filtration operation. Simultaneously, it collects real-time differential pressure data, filter media wear data, and total filter layer clogging coefficient at the current moment through the data acquisition module, with the acquisition delay controlled within 50ms; S52 initializes various parameters of the intelligent backwashing timing scheduling algorithm, including the main correction coefficient and clogging state influence coefficient. The main correction coefficient is set to 0.75, and the clogging state influence coefficient is set to 0.25. The adjustment coefficient is determined based on the current production load. When the load exceeds 110% of the rated value, the main correction coefficient is adjusted accordingly. The correction coefficient is increased by 0.1; S53 inputs the real-time collected data and determined threshold parameters into the intelligent backwash timing scheduling algorithm according to the algorithm requirements. Multiple factors are integrated through product calculation, and after coefficient correction, the backwash start-up timing parameters are calculated, with the calculation cycle controlled within 1 second; S54, based on the backwash start-up timing parameters and the filter layer clogging evolution trend curve generated by weight 6, a trend prediction algorithm is used to deduce the backwash duration. Historical backwash effect data is referenced during the derivation process, and the duration is set within the range of 30 to 120 seconds. For every 10% increase in clogging degree, the duration is extended by 10 seconds to ensure precise matching between the backwash operation and the filter layer contamination state. Through step-by-step intelligent calculation of the backwash timing, parameter values ​​are optimized through simulation. The implementation process emphasizes real-time performance and prediction accuracy, solving the problem of lag in traditional backwash scheduling and improving the targeting and efficiency of backwash operations.

[0045] A pressure differential-based intelligent control method for a coarse filter is disclosed. This method is implemented through a pressure differential-based intelligent control system for a coarse filter in chlor-alkali production. The system includes: a multi-dimensional data acquisition unit for chlor-alkali filtration operation, a filter bed clogging progressive calculation unit, a filter media loss dynamic fitting analysis unit, a pressure differential threshold and loss critical value judgment unit, an intelligent backwashing timing scheduling calculation unit, and a backwashing execution and status feedback unit. The multi-dimensional data acquisition unit for chlor-alkali filtration operation is connected to the pressure differential sensing component and flow detection component of the coarse filter, and is used to collect inlet and outlet pressure differentials, operating time, initial state of the filter media, and media flow data, and transmit them to the filter bed clogging progressive calculation unit. The filter bed clogging progressive calculation unit receives the collected data and, through a built-in module... The filter media loss dynamic fitting analysis unit simulates the filter layer clogging trend and sends it to the filter media loss dynamic fitting analysis unit. This unit combines the clogging trend data with an algorithm to fit the filter media loss state and transmits it to the differential pressure threshold and loss critical value judgment unit. The differential pressure threshold and loss critical value judgment unit processes the received data, determines the threshold range and critical value, and sends it to the intelligent backwashing timing scheduling calculation unit. The intelligent backwashing timing scheduling calculation unit calculates the optimal backwashing parameters based on the threshold data and sends them to the backwashing execution and status feedback unit. The backwashing execution and status feedback unit receives the parameters and controls the coarse filter to perform backwashing, while simultaneously collecting differential pressure data during the backwashing process and feeding it back to the chlor-alkali filtration operating condition multi-dimensional data acquisition unit, forming a complete control link.

[0046] A pressure differential-based intelligent control method for coarse filters dynamically tracks the gradual process of filter media clogging through a specialized extrapolation model. It quantifies the dynamic changes in filter media wear using a fitting algorithm and systematically analyzes the intrinsic correlation between these factors and pressure differential changes, replacing the traditional judgment mode based on a single fixed pressure differential threshold. Instead of relying on static parameters or isolated data for backwashing decisions, it forms a dynamic and comprehensive state assessment based on multi-dimensional real-time data. This ensures that backwashing is neither too early, wasting water and energy, nor too late, preventing complete removal of filter media clogging, thus achieving a match between backwashing operations and the state of filter media fouling.

[0047] This method utilizes a dedicated intelligent differential pressure analysis platform for chlor-alkali filtration operations, integrating multi-channel data acquisition, high-speed processing, and edge computing capabilities to achieve efficient processing of multi-dimensional data such as differential pressure, runtime, and filter media status. Through an intelligent backwashing timing scheduling algorithm, combined with real-time operating conditions, the timing and duration of backwashing initiation are dynamically adjusted, eliminating the traditional control methods that rely on manual experience or historical static data. Simultaneously, a complete status feedback mechanism is constructed, transmitting differential pressure changes during backwashing back to the analysis platform in real time, forming a closed-loop control system. This ensures continuous adaptation between the filter bed status and backwashing operations, significantly improving the stability and efficiency of the coarse filter operation and meeting the stringent operational requirements of chlor-alkali production for the filtration system.

[0048] In the description of this invention, it should be noted that, unless otherwise specified and limited, the terms "set," "install," "connect," "link," and "fix" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0049] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent control of a coarse filter based on differential pressure, characterized in that, Includes the following steps: S1 collects real-time differential pressure data at the inlet and outlet of the coarse filter, filter bed running time data, filter media initial filling status data, and filter medium flow data through the chlor-alkali filtration operating condition differential pressure intelligent judgment platform. S2, based on the collected multi-dimensional data, uses a progressive inference model of filter layer clogging to dynamically infer the degree of filter layer clogging. By analyzing the correlation between the rate of change of pressure difference and the porosity of the filter layer, filter layer clogging evolution trend data is generated. S3. The filter media loss dynamic fitting algorithm is used to fit and analyze the loss state of the filter media during the filtration process. Combined with the filter layer clogging evolution trend data, a mapping relationship between the filter media loss and the pressure difference change is established. S4 uses the intelligent differential pressure judgment platform for chlor-alkali filtration operating conditions to analyze and process the data after simulation and fitting to determine the differential pressure threshold range and the critical value of filter media loss. S5. Based on the determined differential pressure threshold range and filter media wear threshold, the optimal backwash start time and backwash duration are calculated using an intelligent backwash timing scheduling algorithm. S6 performs backwashing on the coarse filter according to the calculated optimal backwashing start time and backwashing duration. At the same time, it feeds back the differential pressure change data during the backwashing process to the chlor-alkali filtration operating condition differential pressure intelligent judgment platform in real time, forming a dynamic adaptation between the filter layer status and the backwashing operation.

2. The intelligent control method for a coarse filter based on differential pressure according to claim 1, characterized in that, The expression for the progressive deduction model of filter layer clogging is: ， in, The local clogging coefficient of the filter layer. Let i be the weighting coefficient for the influence of pressure difference in the i-th group. Let be the pressure difference change during the i-th sampling period. The time influence weighting coefficient for the i-th acquisition period is given by [reference]. Let i be the runtime of the i-th acquisition cycle. The total clogging coefficient of the filter bed. For local congestion coordination coefficient, This is the actual pressure difference correction factor. This is the real-time measured differential pressure value. denoted as the initial porosity of the filter layer, and n as the total number of sampling cycles.

3. The intelligent control method for a coarse filter based on differential pressure according to claim 1, characterized in that, The expression for the dynamic fitting algorithm for filter media loss is: ， in, This refers to the amount of filter media lost per unit time. This represents the weighting coefficient for filter media loss in stage k. The average pressure difference in stage k. Let k be the duration of the k-th stage. This is the cumulative filter media loss. This is the integral correction factor for loss. Total runtime The correlation coefficient for blockage loss is... The total clogging coefficient of the filter bed. is the current porosity of the filter layer, and m is the total number of operating stages divided by filter media loss.

4. The intelligent control method for a coarse filter based on differential pressure according to claim 1, characterized in that, The expression for the intelligent backwashing timing scheduling algorithm is: ， in, These are the backwash startup timing parameters. These are the main correction coefficients for the scheduling algorithm. Let be the real-time pressure difference at time j. To preset the differential pressure threshold, Let be the filter media loss at time j. This is a preset critical value for filter media loss. The congestion state influence coefficient. This represents the total clogging coefficient of the filter layer.

5. The intelligent control method for a coarse filter based on differential pressure according to claim 1, characterized in that, The intelligent differential pressure judgment platform for chlor-alkali filtration is equipped with a multi-channel differential pressure sensing module, a high-speed data acquisition card, and an edge computing unit. The sensing module uses a high-precision piezoresistive differential pressure sensor with a sampling frequency set to 50Hz to 100Hz. The data acquisition card has a resolution of no less than 16 bits. The edge computing unit has a built-in FPGA chip for real-time data preprocessing. The platform software sets the differential pressure acquisition cycle to 2s to 5s, the filter layer clogging evolution trend data update cycle to 30s, and the filter media loss fitting analysis interval to be consistent with the differential pressure acquisition cycle. The intelligent backwashing timing scheduling algorithm has a calculation cycle of no more than 1s, and the backwashing operation execution signal response delay is no more than 50ms. The differential pressure threshold range can be dynamically adjusted according to the filter media type and production load, with an adjustment step size of 0.01 to 0.

05.

6. The intelligent control method for a coarse filter based on differential pressure according to claim 1, characterized in that, S2 includes: The differential pressure data and runtime data collected by the chlor-alkali filtration operating condition differential pressure intelligent analysis platform are sorted by time series, divided into multiple continuous data collection cycles, and the maximum, minimum and average differential pressure values ​​in each cycle are extracted; the filter layer blockage progressive inference model is called, and the differential pressure characteristic data and runtime data of each collection cycle are input to determine the differential pressure influence weight coefficient and time influence weight coefficient corresponding to each cycle; The local clogging coefficient for each acquisition cycle is obtained by model calculation. Combined with the real-time measured differential pressure and the initial porosity of the filter layer, the total clogging coefficient of the filter layer is derived. Based on the changes in the total clogging coefficient over multiple consecutive periods, the sliding window method is used to analyze the data variation patterns and generate data on the evolution trend of filter layer clogging.

7. The intelligent control method for a coarse filter based on differential pressure according to claim 1, characterized in that, S3 includes: Historical differential pressure data, initial filling status data of filter media, and filter layer clogging evolution trend data generated by S2 are extracted from the chlor-alkali filtration operating condition differential pressure intelligent judgment platform and divided into data groups according to the operating stage; the filter media loss dynamic fitting algorithm is started, and the corresponding filter media loss weight coefficient is set according to the differential pressure change characteristics of different operating stages. The algorithm calculates the filter media loss per unit time in each operating stage, and performs integral calculations based on the total operating time to obtain the cumulative filter media loss. A mapping relationship model between the cumulative filter media loss and the total clogging coefficient of the filter layer is established. The current filter layer porosity data is input, the mapping relationship parameters are corrected, and the correlation results between filter media loss and pressure difference change are output.

8. The intelligent control method for a coarse filter based on differential pressure according to claim 1, characterized in that, S4 includes: The system receives filter layer clogging evolution trend data from S2 and the correlation results of filter media loss and differential pressure change from S3, and transmits them to the data processing module of the chlor-alkali filtration operating condition differential pressure intelligent judgment platform. The platform performs outlier removal processing on the input data, retains valid data, and sorts it according to priority, which is determined based on the interval between the data acquisition time and the current time. Based on the valid data, the system analyzes the synergistic relationship between the filter layer clogging rate and the differential pressure growth rate, and, in conjunction with the critical value reference range of filter media loss, initially delineates the differential pressure threshold range. By using the platform's built-in threshold calibration module and combining the optimal control parameters from historical operating data, the initially defined differential pressure threshold range is adjusted, and the differential pressure threshold range and the critical value for filter media loss are finally determined.

9. The intelligent control method for a coarse filter based on differential pressure according to claim 1, characterized in that, S5 includes: The system retrieves the established differential pressure threshold range and critical value of filter media loss from the chlor-alkali filtration operating condition differential pressure intelligent judgment platform, and simultaneously collects the real-time differential pressure data, filter media loss data, and total filter layer clogging coefficient at the current moment. Initialize the parameters of the intelligent backwashing timing scheduling algorithm, including the main correction coefficient of the scheduling algorithm and the blockage state influence coefficient, and adjust the parameter values ​​according to the current production conditions; input the real-time collected data and the determined threshold parameters into the algorithm, and calculate the backwashing start-up timing parameters through product operation and coefficient correction; based on the backwashing start-up timing parameters and the filter layer blockage evolution trend, deduce the backwashing duration to ensure that the backwashing operation matches the filter layer contamination state.

10. A coarse filter intelligent control method based on differential pressure according to any one of claims 1-9, characterized in that, This method is implemented through a pressure differential-based intelligent control system for a coarse filter in chlor-alkali production, comprising: The system includes a multi-dimensional data acquisition unit for chlor-alkali filtration operation, a progressive calculation unit for filter layer clogging, a dynamic fitting analysis unit for filter media loss, a differential pressure threshold and loss critical value judgment unit, an intelligent backwashing timing scheduling calculation unit, and a backwashing execution and status feedback unit. The multi-dimensional data acquisition unit for chlor-alkali filtration is connected to the differential pressure sensing component and flow detection component of the coarse filter. It collects inlet and outlet differential pressure, runtime, initial state of the filter media, and media flow data, and transmits this data to the filter layer clogging progressive calculation unit. The filter layer clogging progressive calculation unit receives the collected data, uses a built-in model to deduce the filter layer clogging trend, and sends this data to the filter media loss dynamic fitting analysis unit. The filter media loss dynamic fitting analysis unit combines the clogging trend data with an algorithm to fit the filter media loss state and transmits this data to the differential pressure threshold and loss critical value judgment unit. The differential pressure threshold and loss critical value judgment unit processes the received data, determines the threshold range and critical value, and sends this data to the intelligent backwashing timing scheduling calculation unit. The intelligent backwashing timing scheduling calculation unit calculates the optimal backwashing parameters based on the threshold data and sends them to the backwashing execution and status feedback unit. The backwashing execution and status feedback unit receives the parameters and controls the coarse filter to perform backwashing, while simultaneously collecting differential pressure data during the backwashing process and feeding it back to the multi-dimensional data acquisition unit for chlor-alkali filtration, forming a complete control link.

Images