Filter tank backwashing water saving adaptive method

Abstract

The application relates to the technical field of water treatment control, and discloses a filter tank backwashing water-saving adaptive method; the method is used for solving the problem that in a traditional method, triggering, queuing and air-water parameter distribution of backwashing cannot be included in unified closed-loop control; the method firstly establishes a station-level operation constraint period update, including a clear water tank liquid level, water-air available capacity, drainage margin, outlet water turbidity, primary filtration water reflux or discharge quota; under a unified time base, each filter tank pressure difference, turbidity and filtration flow are synchronously collected, combined with historical backwashing records, backwashing urgency and backwashing demand intensity are obtained; when the constraint is satisfied, a backwashing sequence and concurrent quota are generated, occupation registration and access check are completed, air flushing, two-stage air-water joint and water flushing parameter linkage primary filtration water direction are overall arranged, backwashing is reviewed, written back and gradual adjustment of triggering threshold and parameter boundary is performed after backwashing, and backwashing water consumption is stable and unchanged under parallel filter tanks, and water quality and water supply constraints are controllable.

Description

Technical Field

[0001] This invention relates to the field of water treatment control technology, specifically to an adaptive water-saving method for backwashing a filter tank. Background Technology

[0002] As a common unit in water treatment and advanced wastewater treatment, filter beds typically require periodic air flushing or combined air-water backwashing during operation to restore filter bed permeability and maintain interception capacity. To reduce backwashing water consumption, current engineering practices often use operating parameters such as filter bed pressure difference or effluent turbidity as the trigger for backwashing, and set the backwashing stage sequence, air volume, water volume intensity, and duration based on experience. To further improve operating efficiency, some existing solutions introduce optimization control or intelligent control technologies. For example, the published invention patent application CN112915641B discloses a filter bed backwashing control method, which optimizes the filtration cycle and backwashing parameter set and implements backwashing control accordingly to increase effluent output and reduce energy costs. Another example is the published invention patent application CN108163967B, which discloses an intelligent control system and method for a contact oxidation-aerated biological filter bed, which monitors key operating parameters and adjusts operating conditions through differential pressure sensors, backwashing pumps, and controllers to maintain effluent compliance. However, existing technologies often rely on individual filter parameters such as pressure differential and turbidity as the basis for backwashing decisions. They lack a comprehensive closed-loop constraint modeling and collaborative control mechanism to address the challenges of multiple parallel filters and limited station-level resources. Furthermore, parallel filters cannot integrate station-level resource constraints such as clear water tank level, backwash pump and blower capacity margins, and inlet / outlet flow balance with water quality and hydraulic constraints such as effluent turbidity fluctuation control and primary filtrate discharge or return load control into the same decision-making framework. In actual operation, this often leads to water supply fluctuations and resource contention caused by concentrated backwashing of parallel filters at similar times. Alternatively, during water-saving periods, increased primary filtrate discharge amplifies effluent quality fluctuations, forcing an increase in backwash intensity to suppress these fluctuations. The degree of backwashing or the duration of backwashing make it difficult to sustain water-saving effects. Furthermore, existing technologies lack verification mechanisms for post-backwash recovery and station-level write-back correction mechanisms, making it difficult to continuously adjust trigger thresholds and stage parameters under operating disturbances and long-term drift. This results in uncertain filter layer recovery and filtration cycle drift, leading operators to revert to conservative decision-making patterns to prevent effluent risks, ultimately resulting in persistently high backwashing water consumption. Therefore, an adaptive water-saving method for filter backwashing is needed. This method should unify triggering, queuing, and closed-loop coordination of air and water parameters under conditions of multiple parallel filters and limited station-level resources. It should simultaneously meet water-saving requirements while controlling effluent quality changes, water supply continuity, and initial filtration discharge, achieving a sustained and reliable effect of stable backwashing water consumption over the long term. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a water-saving adaptive method for backwashing of filter tanks, which solves the problem in traditional methods of making it difficult to incorporate backwashing triggering, queuing, and air-water parameter allocation into a unified closed-loop control.

[0004] To achieve the above objectives, the present invention provides the following technical solution: An adaptive water-saving method for backwashing a filter tank includes: S1: Establish station-level operational constraints and determine the clear water tank level, available backwash water flow rate, available blower air volume, drainage capacity, effluent turbidity limit, and primary filtrate return or discharge limit. S2: Simultaneously acquire the pressure difference, effluent turbidity, filtration flow rate and historical backwash records of each filter, and determine the urgency and intensity of backwashing requirements respectively; S3: Determine the sequence of filters to be backwashed when the station-level constraints are met, and determine the number of filters that can be backwashed simultaneously. S4: Start backwashing sequentially according to the sequence, coordinate the intensity and duration of the air flushing, air-water combined and water flushing stages of each filter, and simultaneously switch the destination and return or discharge quota of the primary filtrate. S5: After backwashing is completed, perform a recovery review and record the backwashing process and result parameters. Based on the records, adjust the subsequent backwashing trigger conditions, sorting rules, and stage parameter ranges.

[0005] Furthermore, establish station-level operational constraints and determine the clear water tank level, available backwash water flow rate, available blower air volume, drainage capacity, effluent turbidity limits, and primary filtrate return or discharge limits, including: The liquid level uses a dual threshold and is subject to situation correction; The number of concurrent users for backwash water and blowers is calculated based on dynamic margin. Drainage adopts high-level and rise rate gating, and the external discharge is corrected for equivalent capacity according to the reduction factor; The effluent quality adopts dual limits, and the primary filtration water is subject to dual limits for return and discharge, combined with quota constraints. Station-level parameters are updated periodically and tightened based on abrupt changes in adjacent update values.

[0006] Furthermore, the differential pressure, effluent turbidity, filtration flow rate, and historical backwash records of each filter are acquired simultaneously, including: A unified clock within the station is used to sample differential pressure, effluent turbidity, and filtration flow rate in the same time frame, and all sampled values ​​are aligned to the same time reference. Before entering the judgment, the pressure difference and effluent turbidity are subjected to short-window stabilization processing. The stabilization is based on the joint constraint of the short-window mean and the short-window change rate. When the short-window change rate exceeds the preset threshold, the short-window median is used to replace the short-window mean. Maintain a backwash logbook, which should record at least the start-up time, background, stage configuration parameters, steady-state pressure difference after backwashing, and the peak width of the initial filtrate for each backwash.

[0007] Furthermore, the urgency and intensity of backwashing demand are determined separately, including: The urgency and intensity of demand are determined using a tiered approach.

[0008] The degree of urgency is determined by a combination of criteria consisting of pressure differential proximity, the trend of rising background turbidity in effluent, and disturbance sensitivity; the intensity of demand is determined by a combination of criteria consisting of the pressure differential growth rate per unit of water production and historical recovery performance. Set a confidence level threshold. When the differential pressure shows abnormal jumps or abnormal reverse fluctuations, or when the turbidity shows isolated spikes, reduce the confidence level of the corresponding measurement. When the confidence level is reduced, the judgment will be dominated by the measurement that still maintains confidence and the historical recovery performance.

[0009] Furthermore, when the station-level constraints are satisfied, the sequence of filter beds to be backwashed is determined, including: The station-level constraint satisfaction determination adopts a joint threshold consisting of liquid level boundary, liquid level drop rate, minimum water and gas demand, drainage threshold, and controllable water quality decline range. When the liquid level is in an early warning state or the water quality is in a tight state, the candidate set is tightened according to the necessary backwashing conditions. The backwash sequence is sorted by urgency as the first priority key, and then switched between priority keys corresponding to the initial filtration water peak risk, backwash duration, or historical water consumption based on the station-level situation.

[0010] Furthermore, the number of filter beds that can be backwashed simultaneously is determined, including: The number of concurrent users is constrained by both the maximum limit and tightened risk management. The upper limit of the quota is obtained by rounding down the smaller value between the water resource quota and the gas resource quota. When there is a condition that requires backwashing, the lower limit of the concurrent quota will be released. When water quality becomes tight, the quota for primary filtration becomes tight, the quota for return flow becomes tight, or the quota for drainage becomes tight, the concurrent quota will be reduced and new startups will be frozen, and the tightness will be lifted by continuous window criteria.

[0011] Furthermore, backwashing is initiated sequentially, with the intensity and duration of the air flushing, combined air-water flushing, and water flushing stages of each filter tank being uniformly coordinated, including: Before starting backwashing, register the occupancy of backwash water, blower air, peak drainage and primary filtration water quotas, and complete access verification at the station-level boundary; After approval, the intensity and duration of each stage of air flushing, air-water combination, and water flushing are configured according to the required intensity, and the latter stage of air-water combination or water flushing is finely adjusted once within the preset range.

[0012] Furthermore, the destination and recirculation or discharge quota of the pre-filtered water can be switched simultaneously, including: After backwashing, the primary filtrate observation window is entered. The destination of the primary filtrate is determined based on the effluent turbidity and continuous duration criteria. When the continuity criteria are met, the primary filtrate state is switched to the normal effluent state. The quota for return or discharge is managed at the station level. The quota for return or discharge of primary filtrate is controlled based on the quota occupancy status and is updated synchronously with the quota occupancy of concurrent backwashing.

[0013] Furthermore, after backwashing is completed, a recovery verification is performed and the backwashing process and result parameters are recorded, including: After backwashing, recovery verification is performed within a preset short-term window and a steady-state window. The short-term verification marks risk events based on the continuous effluent turbidity criterion and the stable filtration flow rate criterion, while the steady-state verification marks abnormal events based on the stable differential pressure recovery rate and the steady-state turbidity fluctuation criterion. The risk classification of the filter pool for the marked event is updated according to the peak risk classification adjustment rule, and the peak offset constraint is executed according to the updated risk classification in the subsequent sequence generation.

[0014] Furthermore, based on the records, the subsequent backwashing triggering conditions, sorting rules, and stage parameter ranges are adjusted, including: Based on anti-laundering records and resource release ledgers, continuous event triggering rules are adopted and small-step and prohibition period constraints are set; Continuous event triggering uses filtration cycle, differential pressure re-increase rate, and effluent approaching the upper limit as triggering conditions, and adjusts the candidate triggering threshold according to the differential pressure approach threshold. While satisfying the constraints of small step size and prohibition period, the sequence sorting weights and parameter boundaries of the gas-water combined stage and water flushing stage are updated synchronously.

[0015] Compared with the prior art, the present invention provides a water-saving adaptive method for backwashing a filter tank, which has the following beneficial effects: 1. This invention incorporates backwashing startup, queuing, and air-water parameter allocation under parallel operation of multiple filters into a station-level closed-loop constraint. It combines clear water tank level, available water and air capacity, drainage channel margin, effluent turbidity boundary, and primary filtrate return or discharge quotas for collaborative decision-making. This transforms backwashing scheduling from localized adaptation of a single tank to station-level control consistent with the hydraulic and water quality constraints of the entire plant. Based on the generation of executable backwashing sequences and concurrent quotas, the invention determines executable backwashing sequences and concurrent quotas by separating urgency and demand intensity, and combines occupancy registration and access verification. The system employs a two-stage air-water combined system with limited fine-tuning, a linkage between the primary filtrate observation window and quotas, and a small-step adjustment mechanism for verification and write-back. This avoids resource occupation and water quality / hydraulic disturbances caused by concentrated backwashing, suppresses the superposition of primary filtrate and ineffective discharge, and improves the verifiability of recovery after backwashing and the ability to continuously correct parameters. Ultimately, it ensures that backwashing water consumption remains constant while effluent water quality fluctuates, water supply is continuous, and the amount of primary filtrate treated is controlled at the station-level boundary. This solves the problem in traditional methods of incorporating backwashing triggering, queuing, and air-water parameter allocation into a unified closed-loop control.

[0016] 2. This invention records each state of the backwashing process in a verifiable log, and summarizes the triggering, queuing, occupancy, stage parameters, initial filtrate treatment, and short-term and steady-state verification results into a station-level ledger. It uses continuous event triggering, small-stepping, and prohibition period constraints to progressively correct the triggering threshold, sorting weight, and stage parameter boundaries, making strategy adjustments traceable, comparable, and reversible. This reduces the large fluctuations and frequent reversals in the strategy caused by a filter backwashing anomaly or instrument disturbance, and achieves self-adaptation and improved consistency of operation and control of the backwashing strategy during long-term operation. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the water-saving adaptive method for backwashing a filter tank according to the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1: Figure 1 An adaptive method for water-saving backwashing of filter tanks is presented, including: S1: Establish station-level operational constraints and determine the clear water tank level, available backwash water flow rate, available blower air volume, drainage capacity, effluent turbidity limit, and initial filtration water return or discharge limit. Specific implementation details are as follows: In water treatment or deep wastewater treatment processes where multiple filters operate in parallel, backwashing consumes backwash water, aeration air, and drainage channels, and causes fluctuations in the quality of primary filtrate and short-term permeate. To ensure that the occurrence, queuing, and allocation of stage parameters of backwashing are uniformly constrained by the station-level boundary, a station-level operational constraint set is established and updated at a fixed cycle. The station-level operational constraint set includes at least the following fields: clear water tank level, available backwash water flow rate, available aeration air volume, drainage capacity, effluent water quality limit, and primary filtrate return or discharge limit. These fields are generated from observable or calculable quantities and are solidified in the form of thresholds or upper and lower limits for the purpose of unifying subsequent judgment criteria and concurrent control criteria. The clear water tank level boundary uses a dual threshold to represent the allowable backwash range, setting a safety line and a warning line. The safety line is set at 15% to 25% of the effective volume of the clear water tank, with 20% being an option. This is because the clear water tank needs to retain a minimum buffer volume to cover level drops caused by pump unit switching, valve switching, and short-term fluctuations in the pipeline network. The warning line is set at 5% to 10% above the safety line, with 7% being an option. This is used to tighten and initiate backwashing in advance before approaching the safety line, and to reserve a margin for the treatment of backwash water and primary filtrate. The liquid level threshold is calculated by overlaying a load condition correction term onto the base value. This correction term is determined by the liquid level change rate and the factory flow rate deviation. The liquid level change rate is taken as the linear trend over the past 5 to 20 minutes, with a 10-minute window being preferable because this window can cover the main impacts of a pump unit switchover or water usage fluctuations and suppress instantaneous noise. The factory flow rate deviation is the percentage difference between the current factory flow rate and the average value over the past 30 minutes. The 30-minute average is used to cover short-cycle load fluctuations and avoid baseline shifts caused by single switches. If the factory flow rate is missing, it can be replaced by the factory pump unit frequency and valve position-converted flow rate, or the factory main pipe metering value. When the liquid level change rate is negative and its absolute value exceeds 2% of the effective volume of the clear water tank per hour, or when the factory flow rate deviation exceeds 10%, the liquid level warning line is raised by 2% to 5%, with a 3% option. The 10% threshold is used to identify significant load changes, which is different from the common short-term fluctuations within 5%. This is because when the load increases, the combined effect of backwash water consumption and production water decrease is more likely to trigger water supply risks. When the liquid level change rate is positive and the factory flow deviation does not exceed 5%, the liquid level warning line is adjusted to 1% to 3%, and can be selected as 2%, to release the backwash window during off-peak periods. To prevent threshold drift, the cumulative magnitude of the correction item does not exceed ±6% of the basic warning line or ±5% of the effective volume, whichever is smaller. This cumulative upper limit is used to limit the threshold from deviating too much from the basic operating range due to multiple corrections, thereby maintaining consistency with the design boundary of the clear water tank regulating volume. The relevant fields include at least the effective volume of the clear water tank, the liquid level safety line, the liquid level warning line, the liquid level change rate, the factory flow deviation, and the cumulative magnitude of the correction item. The backwash water resource boundary is characterized by the available backwash water flow rate. The available backwash water flow rate is determined by the number of backwash pump sets that can operate and the rated flow rate of a single pump set, and the flow rate occupied by the filter being backwashed is deducted to obtain the upper limit of incremental allocation for the current cycle. The backwash water intensity of a single tank is 12 to 18 cubic meters per square meter of filter surface per hour, and can be selected as 14 to 16 cubic meters per hour. This range refers to the commonly used backwash intensity range of sand filters or double-layer filter media filters to take into account the risks of filter layer expansion, scouring and carryover, and filter media loss. The boundary of blower air resources is characterized by available air volume, which is determined by the number of blowers that can operate and the rated air volume of each blower. The air volume occupied by the filter currently being backwashed is deducted to obtain the upper limit of incremental allocation for the current cycle. The air flushing intensity per tank is set at 15 to 25 cubic meters per square meter of filter surface per hour, with a selectable range of 18 to 22 cubic meters per hour. This range refers to the commonly used intensity range for air flushing backwashing to meet the bubble shearing required for filter layer loosening and to suppress the risk of excessive agitation. Both backwash water and blower air are decomposed into startable volume and dynamic margin, with the dynamic margin being the corresponding available volume. The range is 5% to 15%, with 10% being optional. This percentage is based on the margin requirements for pump operating condition fluctuations, valve characteristic deviations, and transient fluctuations during stage switching, and is used for stage fine-tuning and deviation compensation. The concurrent quota calculation uses the target intensity optional value as a benchmark to generate the target backwash water flow rate and target gas volume for a single pool. The station-level concurrent quota is determined by the smaller value between the water resource concurrent quota and the gas resource concurrent quota. The relevant fields include at least the available backwash water flow rate, available gas volume, dynamic margin ratio, target backwash water flow rate for a single pool, target gas volume for a single pool, and station-level concurrent quota. The drainage capacity boundary is primarily characterized by the backwash drainage ditch level. When using drainage pumps for external discharge, the equivalent drainage capacity is taken as the number of operational drainage pumps multiplied by the rated flow rate of a single pump, and corrected by a reduction factor of 0.7 to 1.0 according to the back pressure condition. This is because an increase in back pressure will reduce the actual flow rate of the pump, and this reduction range can cover common operating condition differences. A drainage high-level line is set, and when the drainage ditch level reaches the high-level line, it is prohibited to start new backwashing. The high-level line is taken as 75% to 90% of the ditch depth, and 85% can be selected. This is because the drainage ditch needs to retain a buffer to prevent short-term drainage fluctuations and overflow, and a high water level will reduce the drainage slope and affect stable discharge. To identify congestion before the high-level line, a drainage rise rate gating is introduced. The drainage channel level rise rate is taken as the average rate of change within a 2-minute to 8-minute window, with a 5-minute window optional. When the rise rate exceeds 3% per minute of the channel depth, drainage is considered to be tight, and the station-level concurrent capacity is reduced by one. This is based on the fact that drainage tightness usually occurs before reaching the high-level line, and early tightening can reduce the risk of drainage peaks being superimposed due to concurrent backwashing. When the channel depth exceeds the range of 0.8 meters to 2.5 meters, the threshold is recalculated using the equivalent volume change rate to maintain consistency and feasibility under different channel conditions. The relevant fields include at least the drainage channel level, drainage high-level line, drainage channel depth, number of operational drainage pumps, rated flow rate of a single pump, back pressure status, reduction factor, rise rate window length, and station-level concurrent capacity. The effluent water quality limit boundaries are used to constrain water quality disturbances caused by backwashing. When used for water supply treatment, target upper limits and absolute upper limits are set for effluent turbidity. The target upper limit is 0.3 to 0.5, and the absolute upper limit is 0.8 to 1.0, in turbidity units. This range refers to the common effluent control targets and regulatory limit ranges for water supply, and reserves a risk margin for short-term fluctuations during backwashing. When used for advanced wastewater treatment, the control indicators of filtered turbidity or suspended solids are used as the corresponding limit caliber. The target upper limit is preferentially taken as the design operation control indicator, and the absolute upper limit is preferentially taken as the regulatory limit or discharge limit. If the design indicator and operation control are inconsistent, the operation control target is taken as the target upper limit, and the regulatory limit is taken as the absolute upper limit to ensure that the limit sources are consistent and verifiable. The relevant fields include at least the target upper limit, absolute upper limit, indicator type, design control indicator, operation control target, and regulatory limit or discharge limit. The boundary for primary filtration water return or discharge adopts a dual-diameter constraint; the upper limit of return per unit time is 1% to 4% of the plant influent flow rate, which can be selected as 2%, based on the fact that excessive return load will increase the load of the upstream coagulation sedimentation or biological system and induce water quality fluctuations; the upper limit of unit backwash discharge volume is 0.3 to 1.0 times the effective water volume of a single tank, which can be selected as 0.6 times, based on the fact that excessive discharge will offset the water-saving benefits and increase the external discharge pressure; the effective water volume of a single tank is calculated based on the tank volume under the operating water level before filter backwashing, determined by the tank geometry and operating water level, or obtained from the water level and volume curve table; To support backwashing peak shifting, the recirculation limit is converted into a primary filtration water quota constraint. The quota is the maximum number of filters that can be in primary filtration water recirculation state at the same time. The limit is 10% to 30% of the total number of parallel filters, and can be selected as 20%. The reason is that the impact of primary filtration water recirculation on the front end increases with the increase of the recirculation ratio, and a scheduling margin needs to be reserved. The quota is rounded down and must be implemented for no less than one filter, and is also subject to the recirculation limit per unit time. The relevant fields include at least the plant influent flow rate, recirculation limit per unit time, effective water volume per filter, maximum backwash discharge volume per unit, quota limit and quota rounding rules. The station-level operational constraint set is updated at a fixed period, ranging from 10 to 60 seconds, with a 20-second option. This is based on the fact that changes in the clear water tank level, pump availability, and drainage ditch level are relatively slow, and 20 seconds allows for a boundary recalculation before resource fluctuations caused by concurrent startups and phase switching. Resource mutations are compared between two adjacent updates. When the update period is 20 seconds, if the available backwash water flow or blower air volume decreases by more than 30%, it is considered a resource mutation period, and the station-level concurrent quota is reduced by 1 to 2 stations, with a 1-station reduction option. This is based on the fact that pump switching or single-unit tripping usually causes a step drop in available resources, often exceeding 20%, and the 30% threshold is used to distinguish between normal fluctuations and fault mutations. When the update period exceeds 30 seconds, the resource mutation threshold is adjusted to 20% to 25% to match the cumulative fluctuations under longer sampling intervals. When the effluent quality approaches the target upper limit and shows an upward trend, the primary filtration quota will be temporarily reduced by 1 unit to suppress the accumulation of primary filtration. The relevant fields include at least the update cycle, available backwash water flow rate, available blower gas volume, resource mutation threshold, station-level concurrent quota, effluent quality indicators and primary filtration quota.

[0020] S2: Simultaneously acquire the pressure difference, effluent turbidity, filtration flow rate, and historical backwash records of each filter tank, and determine the urgency and intensity of backwashing requirements respectively. The specific implementation is as follows: After the station-level operating constraints are updated to the latest cycle, differential pressure, effluent turbidity and filtration flow rate are acquired under the same time reference and aligned to the same sampling cycle to form status input; the time reference adopts the unified clock in the station, and the sampling cycle is 1 to 5 seconds, which can be selected as 2 seconds. This is based on the fact that the refresh cycle of the differential pressure transmitter and the turbidity meter and the communication cycle of the control system are usually within this range, and 2 seconds can take into account both real-time performance and noise suppression. Before entering the judgment, the differential pressure and turbidity are subjected to short-window steady-state processing. The short-window length is 30 to 120 seconds, and 60 seconds can be selected. The steady-state processing adopts the joint constraint of window mean and rate of change. When the rate of change exceeds the threshold within the window, the median is used instead of the mean. The rate of change threshold is 2% per minute of the full scale of differential pressure and 0.2 times the upper limit of the turbidity target. For example, the threshold benchmark is obtained by statistically analyzing the steady-state data from 24 to 72 hours of continuous operation and taking the larger common value of the rate of change as the threshold benchmark, so that about 95% of the normal rate of change does not exceed the threshold. Then, the valve switching disturbance margin is added to suppress peak interference and preserve the trend. The relevant fields include at least the sampling period, short-window length, full scale of differential pressure, upper limit of turbidity target, rate of change threshold and steady-state data statistical window. Simultaneously, a historical backwashing record database is maintained. The database retains data from the most recent 3 to 10 backwashes for each filter bed, with the most recent 6 being selectable. This is based on the inertia of changes in filter media gradation, mud film adhesion, and uniformity of air and water distribution, and short- to medium-term records better reflect the current recovery pattern. Each record should include at least the following fields: pressure difference and turbidity background at backwash start-up, backwash stage duration, backwash end time, stable pressure difference after backwash, stabilization time, width of the initial filtrate turbidity peak, total backwash duration, and backwash water and air intensity level. The stable pressure difference is characterized by the average value of the window from 10 to 30 minutes after the backwash is completed, and 20 minutes can be selected, based on the fact that the filter bed and hydraulic distribution tend to be stable within this time range and can avoid short-term unstable periods; the peak width of the turbidity of the initial filtrate is measured by the time when the effluent turbidity is continuously higher than the target upper limit and recorded in minutes, which is used for subsequent peak shifting and risk classification input. Based on unified data collection and recording, the urgency and intensity of backwashing demand are given respectively. Both the urgency and intensity of demand are divided into at least low, medium and high levels. High level is determined by the simultaneous occurrence of any two types of criteria signals or the satisfaction of a preset combination threshold. Medium level is determined by the occurrence of a single type of criteria signal. Low level is determined by the failure of any two criteria signals. The urgency is driven by the degree of pressure difference approaching the upper limit, the rising trend of effluent turbidity background, and disturbance sensitivity. When any two criteria are met simultaneously, it enters the high-level candidate area. The upper limit of differential pressure is determined based on the maximum allowable head loss or operational experience, ranging from 2.0 meters to 3.5 meters, with 2.8 meters being a possible option. This is because exceeding this range significantly reduces filtration capacity and increases the risk of short-circuiting and penetration. The degree of differential pressure proximity is characterized by the ratio of the steady-state differential pressure to the upper limit, and an acceleration detection mechanism is introduced. Acceleration correction is triggered when the differential pressure growth rate over the past 30 minutes exceeds 1.5 times the average growth rate over the past 3 hours. Both the 30-minute and 3-hour average growth rates are calculated as the ratio of the steady-state differential pressure increment to the cumulative permeable volume during the same period, using the same sampling rate and integration caliber. This threshold is used to identify the acceleration phase of pollution growth and distinguish it from regular fluctuations. The effluent turbidity background value is calculated using a 60-second window average and compared with a baseline over the past 30 minutes to 2 hours. The baseline window can be selected as 1 hour. The system is designed to cover short-cycle raw water fluctuations within one hour and reflect changes in the filter bed's resistance to disturbances. A rise in background turbidity relative to the baseline exceeding 0.05 to 0.10 and lasting for more than 5 minutes is considered a rising state. Disturbance sensitivity is determined by whether the effluent turbidity response significantly amplifies when influent water quality or filtration flow fluctuates. Significant amplification is defined as an increase in effluent turbidity relative to the background value exceeding 0.05 to 0.10 or exceeding 30% to 60% of the background value, occurring more than twice within 30 minutes and each instance lasting more than 2 minutes, indicating a deteriorating state. This threshold is used to exclude occasional noise and match the common frequency of short-term disturbance events. Relevant fields include at least the upper limit of differential pressure, differential pressure proximity, differential pressure growth rate, turbidity background value, turbidity baseline, disturbance sensitivity criteria, urgency level, and demand intensity level. Demand intensity is used to characterize the backwash energy and duration requirements, and is determined by the pressure differential growth rate and historical recovery performance. The pressure differential growth rate is characterized by the pressure differential growth rate per unit production volume. The pressure differential growth rate is the ratio of the increment of steady-state pressure differential within the integration window to the cumulative production volume during the same period. The cumulative production volume is obtained by integrating the filtration flow rate over the same time base. The integration window ranges from 1 hour to 6 hours, and can be selected as 3 hours, based on the fact that 3 hours can take into account both short-term noise suppression and accelerated pollution identification. When the growth rate exceeds 1.3 times the median growth rate of the filter in the last 6 backwash cycles, the growth rate is determined to be in a deviation state. The median value is used to suppress the impact of a single abnormal backwash on the baseline, and 1.3 times is used to identify significant deviations. Historical recovery performance is constrained by both the stable differential pressure after backwashing and the peak width of the initial filtrate. The stable differential pressure recovery rate is the ratio of the difference between the differential pressure before backwashing and the steady-state verification differential pressure to the differential pressure before backwashing. When the stable differential pressure recovery rate is less than 0.75 and the peak width of the initial filtrate is greater than 12 minutes, the demand intensity is judged as high, based on the fact that this combination often corresponds to insufficient backwashing energy or deviation in air and water distribution, requiring an increase in the air-water combination or water flushing parameter range. When the stable differential pressure recovery rate is high and the peak width is less than 6 minutes, the demand intensity is judged as low, allowing the use of a lower intensity parameter range. Relevant fields include at least the steady-state differential pressure, integral window, cumulative water production, growth rate, median growth rate, stable differential pressure recovery rate, and initial filtrate peak width. To reduce misjudgments caused by instrument drift, a reliability gating system is introduced and the judgment caliber is switched. For differential pressure, when the sampling period is selectable to be 2 seconds, the reliability is reduced when the jump amplitude exceeds 10% of the full scale or there are 5 consecutive abnormal fluctuations in opposite directions. For turbidity, the reliability is reduced when an isolated peak appears and the peak width is less than 6 seconds. These thresholds are set based on typical abnormal patterns of bubble interference, signal jitter, and transient impacts. After the reliability decreases, the urgency determination prioritizes signals that still maintain high reliability, combined with historical record consistency constraints. The demand intensity determination prioritizes a combination of historical recovery performance and the normalized growth rate of permeable water production. Relevant fields include at least the differential pressure sampling period, differential pressure full scale, differential pressure jump amplitude, number of abnormal fluctuations, turbidity peak width, reliability level, and judgment caliber selection identifier. After data collection and judgment are completed, a backwash profile with station-level constraint update time is generated. The backwash profile includes at least the following fields: urgency, demand intensity, differential pressure proximity, turbidity background deviation, differential pressure growth rate classification, stable differential pressure recovery rate classification, initial filtrate peak width classification, total backwash duration classification, and backwash water-air intensity. These fields serve as inputs for subsequent candidate backwashing filters, sequence generation, and concurrent control. Based on the separation of urgency and demand intensity judgment, the switching of credibility gating logic, and historical recovery performance constraints, the backwash profile remains continuously verifiable and is integrated with subsequent scheduling processes.

[0021] S3: When the station-level constraints are satisfied, determine the sequence of filters to be backwashed and the number of filters that can be backwashed simultaneously. The specific implementation is as follows: After obtaining the urgency and intensity of backwashing demand for each filter and generating a backwashing profile, the station-level constraint satisfaction is determined based on the latest updated value of the station-level operation constraints. When the station-level constraints are satisfied, a sequence of filters to be backwashed and concurrent slots are generated. The station-level constraint satisfaction determination and station-level constraint update are performed using the same periodic caliber. The determination conditions include at least the following: the clear water tank level is higher than the warning line and the rate of decline does not exceed the preset upper limit; the available flow rate of backwash water and the available air volume of blower meet the minimum start-up requirements; the drainage capacity has not triggered the high-level line or tightened the threshold; and the effluent quality is lower than the target upper limit or is in a controllable decline range. The upper limit of the rate of decline is 1% to 3% of the effective volume of the clear water tank per hour, and can be selected as 2% per hour. The reason is that this threshold is used to identify the clear water tank entering a continuous consumption state and to distinguish it from the short-term fluctuations caused by pump switching. At the same time, it matches the buffer volume corresponding to the level warning line. The controllable decline range is defined as follows: the effluent quality is higher than the target upper limit but lower than the absolute upper limit, and the water quality change rate is negative for at least 2 minutes within the past 3 to 8 minutes window (5 minutes is optional). The water quality change rate is calculated using the difference of the 60-second moving average, and this change rate caliber is used for subsequent judgments on rise or fall. The minimum start-up requirement for a single pool is set according to the minimum feasible intensity of the backwash process. The minimum intensity of backwash water is set at 10 cubic meters per square meter of filter surface per hour, and the minimum intensity of blower air is set at 12 cubic meters per square meter of filter surface per hour. This threshold is based on the fact that it is lower than the commonly used target intensity lower limit and is used to determine whether the minimum feasible conditions for starting backwash are met, so as to avoid backwash interruption or incomplete cleaning when resources are insufficient. The drainage high level line and tightening threshold adopt the station-level operation constraint parameters. The relevant fields include at least the backwash profile identifier, station-level constraint update time, clear water pool level, level decline rate, available backwash water flow rate, available blower air volume, drainage status, effluent water quality indicators, and concurrent quota. When the station-level constraints are met, a candidate set of filters to be backwashed is determined and a sequence of filters to be backwashed is generated. The candidate set is cross-screened based on urgency and station-level risk status. Filters whose urgency reaches the trigger threshold are included in the candidate set. When the clear water tank is in an early warning state or the effluent water quality is close to the target upper limit, only filters that meet the mandatory backwash threshold are retained in the candidate set. The mandatory backwash threshold is either high urgency and differential pressure closeness of not less than 0.9, or high urgency and effluent turbidity background rise meeting the duration threshold. This is based on the fact that when the station-level margin is insufficient, the candidate scale needs to be tightened to ensure controllable execution. The effluent water quality close to the target upper limit is 0.8 to 0.95 times the target upper limit, and can be selected as 0.9 times. This is based on reserving adjustment margin before approaching the target upper limit to tighten the candidate and concurrent processes in advance. After the candidate set is formed, a sequence of filters to be backwashed is generated. The sequence is sorted using a multi-key rule; the first sorting key is the urgency level, arranged from high to low; the second sorting key switches according to the station-level risk situation. When the effluent quality is close to the target upper limit or the primary filtration water return quota is close to the upper limit, the historical primary filtration water peak risk level is used; when the clear water tank level is close to the warning line or the rate of level drop is close to the upper limit, the total backwash time level is used; when station-level resources are abundant, the historical backwash water consumption level is used; the primary filtration water return quota approaching the upper limit is taken as 80% to 95% of the occupied quota reaching the upper limit, and 90% can be selected, based on reserving adjustment margin before the quota is exhausted to avoid primary filtration water accumulation; the primary filtration water peak width risk level is divided into three levels, less than 6 minutes is low risk, 6 minutes is high risk, and 6 minutes is low risk. The risk level is defined as 10-12 minutes, and the risk level is defined as more than 12 minutes. The total backwash time is divided into three levels: less than 8 minutes is short, 8-15 minutes is medium, and more than 15 minutes is long. The above ranges are set with reference to the statistical range of common primary filtration duration and the engineering time configuration range of air flushing and combined air-water flushing. The historical backwash water consumption is obtained from the statistical data of the last 6 backwashes. The threshold is taken as the historical median or 75th percentile. The basis is that the use of percentile statistics can stably reflect the differences between filters and suppress the impact of a few abnormal backwashes on the ranking. The relevant fields include at least the urgency level, pressure difference proximity, effluent water quality indicators, primary filtration return quota occupancy, primary filtration peak width, total backwash time, and historical backwash water consumption. After the sequence is determined, the number of filters allowed to backwash concurrently is determined simultaneously. The number of concurrent filters is constrained by both the upper limit and risk tightening. The upper limit is determined by the smaller of the backwash water resource quota and the blower air resource quota, and rounded down. The lower limit of the number of concurrent filters is 0 or 1. When there are filters that meet the mandatory backwash threshold, the lower limit is 1. Risk tightening includes situations where the effluent quality is close to the target upper limit and is rising or not significantly declining. In these cases, the number of concurrent filters is reduced by 1 to 2 filters from the upper limit, with a possible reduction of 1 filter. When the utilization rate of the primary filtrate return quota exceeds 80% or the return flow rate is close to the return upper limit per unit time, the number of concurrent filters is reduced by 1 filter. When the return flow rate is close to the return upper limit, the value is 0.8 times to 0 times the return upper limit. 95 times, optional 0.9 times, 80% threshold is used to reserve adjustment margin before the quota is exhausted to suppress backflow exceeding the limit or front-end impact caused by the superposition of primary filtration water; rise or slight fall is determined by water quality change rate greater than or equal to 0 and established for 2 consecutive update cycles, water quality change rate adopts the 60-second moving average difference caliber; when drainage tightening threshold is triggered, the concurrent number value is additionally reduced by 1 and the new backwash start is suspended, tightening is determined by drainage channel liquid level rise rate falling back to below the threshold for 2 consecutive update cycles; relevant fields include at least concurrent number, backwash water resource quota, blower air resource quota, effluent water quality threshold, primary filtration water return quota utilization rate, return flow rate ratio, drainage tightening status and update cycle, etc.

[0022] S4: Initiate backwashing sequentially, coordinating the intensity and duration of the air flushing, combined air-water flushing, and water flushing stages in each filter, and simultaneously switching the destination and recirculation or discharge quota of the pre-filtered water. Specific implementation is as follows: After obtaining the sequence of backwashed filters and determining the concurrent slots, backwash access and execution control are performed according to the sequence, with the concurrent slots serving as the upper limit for simultaneous execution, and station-level operational constraints serving as the boundaries for water volume, water and air resources, drainage capacity, and water quality. Backwashing startup adopts a sequential access approach. When it is a filter's turn to start, station-level occupancy registration and station-level verification are completed first. After verification, the backwashing phase is switched. If the verification fails, the startup round of that filter is temporarily suspended and remains in the original sequence. Instead, the next filter in the sequence that meets the verification conditions and has an urgency level no less than the current filter is selected for occupancy registration and startup, in order to avoid resource over-allocation or drainage congestion and to maintain the original sequence ordering logic. The occupancy registration should include at least the following fields: backwash water flow occupancy, blower air volume occupancy, peak drainage occupancy, and primary filtration water return or discharge quota occupancy. Backwash water flow occupancy is calculated based on the filter surface area and planned backwash water intensity; blower air volume occupancy is calculated based on the filter surface area and planned air jet intensity; peak drainage occupancy is determined by multiplying the backwash water flow rate by a drainage margin coefficient, which ranges from 1.05 to 1.20, with a possible value of 1.10. This coefficient is chosen to account for short-term peak values ​​and measurement errors caused by coverage phase switching, such as a momentary peak value of approximately 10% coverage. The initial filtration water return or discharge quota is calculated based on the initial filtration water quota and the upper limit of return per unit time. When discharge is adopted, the upper limit of the unit backwash discharge volume is reserved. The station-level verification includes at least the following conditions: the remaining available backwash water flow after the registration is not lower than the lower limit of the dynamic margin; the remaining available blower gas volume is not lower than the lower limit of the dynamic margin; the drainage ditch level is lower than the high level line and the rising rate has not triggered the tightening threshold; and there is a vacancy in the initial filtration water return quota. The drainage high level line, tightening threshold and release criterion values ​​are taken as a unified standard in the station-level operation constraints and concurrent determination stage. After entering the backwashing stage, the intensity and duration of the air flushing stage, air-water combined stage, and water flushing stage are configured according to the required intensity level. The relevant intensity values ​​are all within the allowable range of a single pool under the station-level operational constraints. The air flushing stage is set according to the required intensity level. When the required intensity is high, the air flushing intensity is 20 to 25 cubic meters per square meter per hour and the duration is 2 to 4 minutes, with a 3-minute option. When the required intensity is medium, the air flushing intensity is 18 to 22 cubic meters per square meter per hour and the duration is 1.5 to 3 minutes. When the required intensity is low, the air flushing intensity is 15 to 20 cubic meters per square meter per hour and the duration is 1 to 2 minutes. The upper limit of the air flushing intensity is no more than 25 cubic meters per square meter per hour. This range is set with reference to the commonly used intensity range for air flushing loose filter layers and is used to suppress the risk of filter media loss and uneven air distribution caused by excessive agitation. The air-water combined stage adopts a two-stage configuration. The first stage has a water intensity of 14 to 18 cubic meters per square meter per hour, an air intensity of 18 to 24 cubic meters per square meter per hour, and a duration of 1 to 3 minutes. The second stage has a water intensity of 12 to 16 cubic meters per square meter per hour, an air intensity of 15 to 20 cubic meters per square meter per hour, and a duration of 2 to 6 minutes. This is based on the fact that the initial backwash needs to use a higher intensity to complete the loosening and stripping in a short time, and the later stage uses a medium intensity for a longer time to complete the carrying and discharge, thereby reducing ineffective water consumption and peak drainage. The water flushing stage has a water intensity of 12 to 16 cubic meters per square meter per hour and a duration of 2 to 6 minutes, which can be selected as 3 to 4 minutes, referring to the commonly used flushing intervals for filter layer stabilization and residue removal. The relevant fields should include at least the required intensity level, air flushing intensity, air flushing duration, first stage water intensity, first stage air intensity, first stage duration, second stage water intensity, second stage air intensity, second stage duration, water flushing intensity, and water flushing duration. To accommodate differences in pollution levels within a single pool and maintain station-level plan stability, a one-time staged fine-tuning is permitted during backwashing. This fine-tuning occurs only within the second stage of the combined air-water and water flushing process and is limited to one adjustment. The adjustment range is restricted to within ±20% of the original plan. This range is used to cover common pollution differences and execution deviations without altering concurrent slots and occupancy boundaries. Fine-tuning is triggered by a combined criterion of effluent water quality trend and differential pressure decline trend. The effluent water quality trend is characterized by the direction of change of the average value within a 30- to 90-second window, with a 60-second window being a viable option. This is because 60 seconds can cover short-term fluctuations caused by stage switching and suppress turbidity measurement noise. Effluent water quality is primarily obtained from the turbidity meter in the backwash drainage channel. If this is unavailable, alternative methods include, in sequence, the turbidity of the backwash drainage return pipeline, the turbidity of the backwash effluent, or the water quality indicators at the drainage channel return end. The low-level value of the wastewater quality is defined as the turbidity of the wastewater being 30% to 50% below the peak value for two consecutive short windows, with a selectability of 40%. The high-level value of the wastewater quality is defined as the turbidity of the wastewater being 50% to 70% above the peak value for two consecutive short windows, with a selectability of 60%. These ratios are used to quantify the decline and high-level states and avoid misjudgment due to single-point noise. The pressure difference decline trend is characterized by the ratio of the pressure difference before backwashing to the current pressure difference relative to the pressure difference before backwashing. The recovery ratio follows the definition of stable pressure difference recovery rate. When the recovery ratio reaches 0.6 to 0.8 and the wastewater quality is at a low level, a selectability of 0.7 can be used, and the second stage or water flushing stage can be shortened. When the wastewater quality is at a high level and the recovery ratio is below 0.6, the second stage or water flushing stage can be extended. When a fine-tuning occurs, the changes in water and air occupation and wastewater occupation are written back to the station-level occupation ledger. The relevant fields include at least the fine-tuning trigger flag, short window length, wastewater quality peak value, low-level threshold, high-level threshold, recovery ratio, adjustment direction, adjustment range, and occupation change amount. The switching of the primary filtrate destination and the quota for return or discharge are simultaneously incorporated into the station-level occupancy management at the end of backwashing. After backwashing, the primary filtrate observation window is entered, with a window value ranging from 2 to 12 minutes, selectable as 6 minutes, based on the fact that this range covers the common duration of the primary filtrate turbidity peak. Within the window, the destination is controlled by the effluent turbidity and continuity criteria. When the effluent turbidity is higher than the target upper limit, return or discharge is maintained. When the effluent turbidity is lower than the target upper limit for 30 to 90 seconds consecutively, the primary filtrate mode ends and switches to normal effluent. The continuity criteria value is based on suppressing false switching caused by instrument noise and hydraulic pulsation. The recirculation or discharge quota is managed at the station level, using the station-level operational constraint set's unit-time recirculation limit and primary filtration quota limit. Risk grading and proximity criteria follow the unified standards used in the sequence generation and concurrency determination stages. Low risk and high risk correspond to the low-risk and high-risk tiers of the primary filtration peak width, respectively, with tier boundaries set at 6 minutes and 12 minutes. This is based on matching the engineering statistical range of primary filtration duration and facilitating peak-shaving control. The proximity threshold ranges from 0.8 to 0.95 times the upper limit, with a selectable value of 0.9. Quota proximity is set at 80% to 95% of the quota limit, with a selectable value of 90%. The percentage is used to reserve adjustment margin before the quota is exhausted; when the return limit or quota reaches the threshold, the return duration of low-risk filters is shortened or the return is ended in advance, and the start-up of high-risk filters is delayed; when discharge is adopted, the discharge volume is constrained by the upper limit of the unit backwash discharge volume. When the drainage becomes tight or the external discharge is restricted, the low-risk filters are switched from discharge to return and the observation window is shortened to reduce the superposition of drainage peaks. The relevant fields include at least the primary filtration observation window, target upper limit, continuous criterion duration, return limit, primary filtration quota, risk classification boundary, threshold approach, quota approaching threshold and destination switching status. During the backwashing process, the station-level risk situation is continuously monitored. The risk situation includes at least the following fields: clear water tank level, drainage ditch level, effluent turbidity, and return quota utilization rate. When the clear water tank level is below the warning line, the drainage ditch level is close to the high level line, or the effluent turbidity is close to the absolute upper limit, the station-level tightening is triggered. The value of "close to the absolute upper limit" is 0.8 to 0.95 times the absolute upper limit, and 0.9 can be selected. The basis is to reserve adjustment margin before exceeding the limit to suppress the risk of water quality fluctuations. Station-level tightening adopts a method of freezing new startups and conservatively completing ongoing operations. Freezing new startups means pausing the registration of subsequent filters in the sequence for occupancy. Filters that are backwashing maintain the minimum intensity of the second stage of the combined air-water and water flushing stage up to the lower limit of the planned intensity level to maintain basic flushing and stable paving. When drainage becomes tight, the second stage is shortened first and the lower limit of water flushing is maintained. When the clear water tank level becomes tight, the water intensity is reduced to the lower limit and the necessary air volume is maintained. When the effluent turbidity becomes tight, the water flushing is extended to the upper limit and new concurrent startups are suspended. The stage parameters, fine-tuning events, duration of initial filtration, and changes in occupancy during the execution process are recorded as inputs for subsequent recovery verification and record writing.

[0023] S5: After backwashing is completed, a recovery review is performed, and the backwashing process and result parameters are recorded. Based on the records, the subsequent backwashing trigger conditions, sorting rules, and stage parameter ranges are adjusted. The specific implementation is as follows: After backwashing, recovery verification and record writing begin. Recovery verification is used to generate verifiable recovery conclusions for the filter layer, while record writing is used to convert the verification conclusions and execution process parameters into inputs for subsequent triggering conditions, sorting rules, and stage parameter boundaries. Verification and record writing are performed separately for each filter, and risk events and resource usage and release status are summarized at the station level. At the same time, available resource boundaries and risk classification are updated, which must include fields such as risk event type, risk level, resource usage and release amount, filter risk level, and update timestamp. Peak staggering is used to reduce the concurrency probability of high-risk filters and extend their start-up period during subsequent sequence generation. Recovery verification includes at least short-term verification and steady-state verification. Short-term verification begins after backwashing and the system enters primary filtrate management, with a duration of 3 to 10 minutes (optional, 5 minutes), based on the operating characteristic that the turbidity peak of primary filtrate typically appears and declines within a few minutes. Short-term verification uses a continuous window criterion to assess the downward trend of effluent turbidity, with a window value of 30 to 90 seconds (optional, 60 seconds), based on the requirements for suppressing instrument noise and hydraulic pulsation. A sustained increase is defined as a 60-second moving average difference greater than 0 for 2 consecutive minutes. Turbidity exceeding the target upper limit for more than 2 minutes is also considered a water quality risk event. Filtration flow stability uses a relative mean deviation criterion, with the flow fluctuation amplitude defined as the maximum flow rate within a 60-second window. The difference between the minimum and maximum values ​​is divided by the mean of the window. When the fluctuation exceeds 10% and lasts for more than 1 minute, it is judged as a hydraulic stability abnormal event, referring to the operating characteristics that the flow rate should maintain a small fluctuation under the fixed valve position condition. When a water quality risk event or a hydraulic stability abnormal event occurs, the event is registered as a high-priority risk event, and the risk level of the filter is updated according to the peak risk level update rule. The peak risk level takes three levels: low risk, medium risk, and high risk. The upward adjustment rule is from low to medium, medium to high, and the high risk level is no longer adjusted. It is used for subsequent staggered sorting. The relevant fields include at least the duration of short-term review, turbidity window length, turbidity rise criterion, cumulative duration of exceeding the limit, flow fluctuation range, duration, event flag, and peak risk level. Steady-state verification is performed within a 10-30 minute window after backwashing, with 20 minutes being a suitable option. This is because filter bed reconstruction and hydraulic distribution tend to stabilize within this timeframe, reducing the impact of transient disturbances on the evaluation. The stable pressure difference is taken as the average pressure difference within the steady-state verification window, and a recovery rate evaluation is formed by comparing it with the pressure difference before backwashing. The recovery rate is expressed as the difference between the pressure difference before backwashing and the steady-state verification pressure difference, divided by the pressure difference before backwashing. During the steady-state period, the average effluent turbidity and its fluctuation amplitude are recorded. The fluctuation amplitude is expressed as the difference between the maximum and minimum turbidity values ​​within the steady-state window. When the fluctuation amplitude exceeds 0.3 times the target upper limit and persists, the water quality is considered to be in a steady state. Unstable events are defined as those where two consecutive steady-state verification windows exceed the threshold, or where the cumulative duration of exceeding the threshold within the same steady-state window is not less than 5 minutes. This is based on the premise that the turbidity during the steady-state filtration stage should be in a low-fluctuation zone, and continuous exceedance indicates uneven filter layer distribution or an increased risk of short circuits. When a steady-state water quality instability event occurs, the risk classification is updated according to the peak-shaped risk classification upward adjustment rule, and peak staggering is implemented. At the same time, the upper limit of the high-intensity segment or the lower limit of the stable cleaning segment is tightened in the stage parameter boundary. Relevant fields include at least the steady-state verification window, stable pressure difference, recovery rate, steady-state turbidity mean, fluctuation amplitude, event marker, and risk classification. After completing short-term and steady-state reviews, backwash records are generated and written back. The backwash records use a fixed caliber and fixed units. The record fields include at least the urgency level and demand intensity level at the trigger time, the station-level constraint status when backwashing is started, the sequence position, the number of concurrent users and the actual number of concurrent users, the planned and actual values ​​of backwash water flow rate and blower air volume, the air flushing intensity and duration, the intensity and duration of the first and second stages of the air-water combined process, the water flushing intensity and duration, the stage fine-tuning mark and direction, the time of switching the destination of the primary filtrate, the duration of the primary filtrate, the peak value and peak width of the primary filtrate, the short-term review event mark, the stable pressure difference and recovery rate of the steady-state review, the steady-state turbidity mean and fluctuation range, the pressure difference regeneration rate or the filtration cycle length, etc. Among them, the peak width of the primary filtrate is in minutes, the peak value is in turbidity units, the pressure difference is in meters of head, the flow rate is in cubic meters per hour, and the duration is in minutes, which facilitates cross-filter comparison. The differential pressure regeneration rate is calculated by dividing the steady-state differential pressure increment within a 3-hour window after backwashing by the cumulative water production during the same period. The 3-hour window is used to cover the initial sensitive growth phase after backwashing. A significant increase in the differential pressure regeneration rate is defined as an increase of more than 20% to 40% relative to the median of the last 6 times, with 30% being an option, to identify accelerated growth deviating from the baseline. After the backwash record is generated, the resource occupancy ledger is written back and the occupancy fields are released. The released fields include at least the backwash water resource occupancy, aeration air resource occupancy, peak drainage occupancy, and primary filtration quota occupancy. The release granularity is consistent with the station-level update cycle to ensure the accuracy of the available resource boundaries for the next round of sequence generation. When making subsequent adjustments based on backwash records, a combination of small-step and prohibition period rules is adopted. Small-step is used to gradually adjust the boundary of stage parameters, with each step increment not exceeding 10% of the stage duration and 15% of the stage intensity. This is based on the fact that the backwash process has a non-linear response, and excessive single changes will amplify the fluctuations in results and affect the stability of station-level scheduling. The prohibition period is used to freeze aggressive adjustments in the event of water quality risk events, hydraulic stability anomalies, or steady-state water quality instability events. The prohibition period ranges from 2 to 12 hours, with 6 hours being an option. This is based on the fact that filter layer reconstruction is not yet stable in a short period of time, and frequent adjustments can easily lead to risk superposition. Relevant fields include at least the stage duration step limit, stage intensity step limit, prohibition period duration, risk event type, and adjustment enable status. The trigger condition adjustment is based on the differential pressure proximity threshold and adopts a continuous event triggering caliber. The differential pressure proximity is the ratio of the steady-state differential pressure to the upper limit of the differential pressure, with a value range of 0 to 1, used to characterize the degree to which the differential pressure approaches the upper limit. The filter cycle is significantly shortened when it decreases by more than 15% to 30% relative to the median filter cycle of the last 6 times, with 20% being an option. The basis for this is that using the median value can suppress the impact of individual abnormal cycles on the baseline, and 15% to 30% is used to distinguish between normal fluctuations and continuous deterioration trends. When three consecutive backwashes result in a significant shortening of the filtration cycle or a significant increase in the differential pressure regeneration rate, the differential pressure proximity threshold is lowered by 0.05 to 0.10 to trigger candidate judgment in advance. When three consecutive backwashes result in an excessively wide peak in the initial filtrate and the turbidity of the station's effluent repeatedly approaches the target upper limit during the concentrated backwash period, the differential pressure proximity threshold is raised by 0.05 to 0.10, and the risk classification is updated according to the peak risk level adjustment rule for subsequent peak staggering sorting. The value of approaching the target upper limit is 0.8 to 0.95 times the target upper limit of effluent turbidity, which can be selected as 0.9 times, and occurs more than twice within a single backwash period. This is based on reserving adjustment margin before exceeding the limit and excluding occasional spike interference. The relevant fields include at least the differential pressure proximity threshold, median filtration cycle value, filtration cycle decrease ratio, differential pressure regeneration rate, initial filtrate peak width, effluent turbidity proximity threshold, and risk classification status. The sorting rules are adjusted based on risk grading and duration grading priorities. When the turbidity of the station-level effluent approaches the target upper limit multiple times during concurrent backwashing periods, the priority of the low-risk peak-shaped filter sequence is increased and the probability of concurrent high-risk peak-shaped filters is reduced. When the clear water tank level approaches the warning line multiple times or the rate of level drop approaches the upper limit multiple times, the priority of the short-term backwash filter sequence is increased and the start-up window of the long-term backwash filter is tightened. The adjustment of the stage parameter range is based on the upper and lower limits of the two-stage air-water combined and water flushing stages. When the recovery rate is consistently low or the differential pressure regeneration rate is consistently high, the lower limit of the second stage of air-water combined or water flushing is raised. When the peak shape of the primary filtrate is consistently wide or there are frequent steady-state water quality instability events, the lower limit of water flushing is raised and the upper limit of the intensity of the first stage of air-water combined is tightened. The updated content is written into the backwash profile and the station-level risk grading table. The written content includes at least the fields of differential pressure proximity threshold, peak shape risk level, concurrent tightening suggestion, stage intensity upper and lower limits, and stage duration upper and lower limits.

[0024] In this embodiment, taking a water treatment plant with six rapid filter beds operating in parallel as an example, station-level operational constraints are set before operation. These constraints include clear water tank level boundaries, backwash water and blower availability, drainage capacity, effluent turbidity control boundaries, and primary filtrate return or discharge limits, and are updated periodically. During operation, each filter bed simultaneously collects differential pressure, effluent turbidity, and filtration flow rate, and generates a backwash profile based on the most recent backwash profile, obtaining the urgency and intensity of backwash demand. Once the station-level constraints are met, a backwash sequence is generated based on urgency, combined with the risk of primary filtrate peak shape and backwash duration. The sequence is then determined according to water and air resource quotas and water quality risk tightening rules. The number of concurrent processes; during the execution phase, access is granted sequentially, and the peak values ​​of backwash water, blower air, drainage, and primary filtration water quota occupancy are first registered and verified; after verification, the parameters of air flushing, two-stage air-water combination, and water flushing are set according to the intensity of demand, and a one-time fine adjustment is made within a certain range; after the backwashing is completed, the primary filtration water observation window is entered, and the return or discharge is switched according to the continuous criteria and subject to station-level quota constraints. Then, short-term verification and steady-state verification are performed, the backwashing process and result parameters are recorded, and resource occupancy is released. At the same time, the subsequent trigger thresholds, sorting weights, and stage parameter boundaries are updated according to continuous events and small-step rules, so as to achieve coordinated operation of backwashing water saving and controlled water quality fluctuations when multiple filters are connected in parallel.

[0025] It should be noted that this invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0026] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented in software, the above embodiments can be implemented in whole or in part by a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions of the embodiments of this application are implemented in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted wirelessly or wiredly from one website, computer, server, or data center to another website, computer, server, or data center. Wired methods include optical fiber, twisted pair, coaxial cable, etc. Wireless methods include infrared, microwave, etc. Available media include any available media that can be accessed by a computer or data storage devices such as servers and data centers that contain one or more sets of available media. Available media can be magnetic media (floppy disks, hard disks, magnetic tapes), optical media (DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0027] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0028] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A water-saving adaptive method for backwashing a filter tank, characterized in that, include: S1: Establish station-level operational constraints and determine the clear water tank level, available backwash water flow rate, available blower air volume, drainage capacity, effluent turbidity limit, and primary filtrate return or discharge limit. S2: Simultaneously acquire the pressure difference, effluent turbidity, filtration flow rate and historical backwash records of each filter, and determine the urgency and intensity of backwashing requirements respectively; S3: Determine the sequence of filters to be backwashed when the station-level constraints are met, and determine the number of filters that can be backwashed simultaneously. S4: Start backwashing sequentially according to the sequence, coordinate the intensity and duration of the air flushing, air-water combined and water flushing stages of each filter, and simultaneously switch the destination and return or discharge quota of the primary filtrate. S5: After backwashing is completed, perform a recovery review and record the backwashing process and result parameters. Based on the records, adjust the subsequent backwashing trigger conditions, sorting rules, and stage parameter ranges.

2. The adaptive water-saving method for backwashing a filter tank according to claim 1, characterized in that, Establish station-level operational constraints and determine the clear water tank level, available backwash water flow rate, available blower air volume, drainage capacity, effluent turbidity limits, and primary filtrate return or discharge limits, including: The liquid level uses a dual threshold and is subject to situation correction; The number of concurrent users for backwash water and blowers is calculated based on dynamic margin. Drainage adopts high-level and rise rate gating, and the external discharge is corrected for equivalent capacity according to the reduction factor; The effluent quality adopts dual limits, and the primary filtration water is subject to dual limits for return and discharge, combined with quota constraints. Station-level parameters are updated periodically and tightened based on abrupt changes in adjacent update values.

3. The adaptive water-saving method for backwashing a filter tank according to claim 1, characterized in that, Simultaneously acquire differential pressure, effluent turbidity, filtration flow rate, and historical backwash records for each filter, including: A unified clock within the station is used to sample differential pressure, effluent turbidity, and filtration flow rate in the same time frame, and all sampled values ​​are aligned to the same time reference. Before entering the judgment, the pressure difference and effluent turbidity are subjected to short-window stabilization processing. The stabilization is based on the joint constraint of the short-window mean and the short-window change rate. When the short-window change rate exceeds the preset threshold, the short-window median is used to replace the short-window mean. Maintain a backwash logbook, which should record at least the start-up time, background, stage configuration parameters, steady-state pressure difference after backwashing, and peak width of the initial filtrate for each backwash.

4. The adaptive water-saving method for backwashing a filter tank according to claim 1, characterized in that, Determine the urgency and intensity of backwashing requirements separately, including: The urgency and intensity of demand are determined using a tiered approach. The urgency is determined based on a combination of criteria including the proximity of pressure differentials, the upward trend of effluent turbidity, and the sensitivity to disturbances. The intensity of demand is determined based on a combination of criteria including the growth rate of pressure differentials per unit of permeable water production and historical recovery performance. Set a confidence level threshold. When the differential pressure shows abnormal jumps or abnormal reverse fluctuations, or when the turbidity shows isolated spikes, reduce the confidence level of the corresponding measurement. When the confidence level is reduced, the judgment will be dominated by the measurement that still maintains confidence and the historical recovery performance.

5. The adaptive water-saving method for backwashing a filter tank according to claim 1, characterized in that, When the station-level constraints are met, the sequence of filters to be backwashed is determined, including: The station-level constraint satisfaction determination adopts a joint threshold consisting of liquid level boundary, liquid level drop rate, minimum water and gas demand, drainage threshold, and controllable water quality decline range. When the liquid level is in an early warning state or the water quality is in a tight state, the candidate set is tightened according to the condition that backwashing is necessary. The backwash sequence is sorted by urgency as the first priority key, and then switched between priority keys corresponding to the initial filtration water peak risk, backwash duration, or historical water consumption based on the station-level situation.

6. The adaptive water-saving method for backwashing a filter tank according to claim 1, characterized in that, Determine the number of filter beds that can be backwashed simultaneously, including: The number of concurrent users is constrained by both the maximum limit and tightened risk management. The upper limit of the quota is obtained by rounding down the smaller value between the water resource quota and the gas resource quota. When there is a condition that requires backwashing, the lower limit of the concurrent quota will be released. When water quality becomes tight, the quota for primary filtration becomes tight, the quota for return flow becomes tight, or the quota for drainage becomes tight, the concurrent quota will be reduced and new startups will be frozen, and the tightness will be lifted by continuous window criteria.

7. The adaptive water-saving method for backwashing a filter tank according to claim 1, characterized in that, Backwashing is initiated sequentially, with unified coordination of the intensity and duration of the air flushing, combined air-water flushing, and water flushing stages in each filter, including: Before starting backwashing, register the occupancy of backwash water, blower air, peak drainage and primary filtration water quotas, and complete access verification at the station-level boundary; After approval, the intensity and duration of each stage of air flushing, air-water combination, and water flushing are configured according to the required intensity, and the latter stage of air-water combination or water flushing is finely adjusted once within the preset range.

8. The adaptive water-saving method for backwashing a filter tank according to claim 1, characterized in that, Synchronously switch the destination and recirculation or discharge quota of the pre-filtered water, including: After backwashing, the primary filtrate observation window is entered. The destination of the primary filtrate is determined based on the effluent turbidity and continuous duration criteria. When the continuity criteria are met, the primary filtrate state is switched to the normal effluent state. The quota for return or discharge is managed at the station level. The quota for return or discharge of primary filtered water is limited based on the quota occupancy status and is updated synchronously with the quota occupancy of concurrent backwashing.

9. The adaptive water-saving method for backwashing a filter tank according to claim 1, characterized in that, After backwashing is completed, a recovery check should be performed and the backwashing process and result parameters should be recorded, including: After backwashing, recovery verification is performed within a preset short-term window and a steady-state window. The short-term verification marks risk events based on the continuous effluent turbidity criterion and the stable filtration flow rate criterion, while the steady-state verification marks abnormal events based on the stable differential pressure recovery rate and the steady-state turbidity fluctuation criterion. The risk classification of the filter pool for the marked event is updated according to the peak risk classification adjustment rule, and the peak offset constraint is executed according to the updated risk classification in the subsequent sequence generation.

10. The adaptive water-saving method for backwashing a filter tank according to claim 1, characterized in that, Based on the records, adjust the subsequent backwashing trigger conditions, sorting rules, and stage parameter ranges, including: Based on anti-laundering records and resource release ledgers, continuous event triggering rules are adopted and small-step and prohibition period constraints are set; Continuous event triggering uses filtration cycle, differential pressure re-increase rate, and effluent approaching the upper limit as triggering conditions, and adjusts the candidate triggering threshold according to the differential pressure approach threshold. While satisfying the constraints of small step size and prohibition period, the sequence sorting weights and parameter boundaries of the gas-water combined stage and water flushing stage are updated synchronously.

Images