Constructed wetland intelligent scheduling control method oriented to energy conservation and consumption reduction

Abstract

The invention relates to the technical field of multi-equipment control, in particular to an energy-saving and consumption-reducing artificial wetland intelligent scheduling control method, which comprises the following steps of: acquiring a water inlet flow monitoring value, a water inlet end liquid level monitoring value and a water outlet end liquid level monitoring value of a wetland unit, and acquiring a difference value between the water inlet end liquid level monitoring value and the water outlet end liquid level monitoring value; calculating and generating a matrix blocking impedance coefficient value, and mapping the matrix blocking impedance coefficient value. According to the method, the aeration duration is determined according to the real dissolved oxygen demand in the working unit, matching of air supply and oxygen demand is achieved, electric energy loss caused by blind aeration is completely eradicated, finally, discrete air supply demands of all branches are subjected to serialized peak shifting arrangement on a time axis, a smooth and predictable total air volume demand curve is formed, and the overall air volume demand curve is optimized. And the curve is matched with the energy efficiency characteristic of the central air blower, so that the air blower always operates in the most efficient working interval, and collaborative energy conservation and consumption reduction are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of multi-device control, and particularly relates to an artificial wetland intelligent scheduling control method for energy saving and consumption reduction. BACKGROUND

[0002] The artificial wetland scheduling technology is used for the operation management and process optimization control of a sewage treatment project simulating a natural wetland ecological system throughout the whole life cycle, covering the adjustment of key hydraulic parameters such as the inflow and outflow of the wetland system, the hydraulic load, the hydraulic retention time and the water level, and also including the collaborative operation and maintenance of auxiliary systems such as oxygenation aeration equipment, reflux devices, dredging and sludge removal and plant harvesting, and the decontamination process inside the wetland is regulated by real-time monitoring of multi-dimensional data such as water quality, water quantity and environmental factors (such as temperature and dissolved oxygen).

[0003] In the existing artificial wetland scheduling technology, the scheduling strategy is often based on preset static logic or simple feedback mechanism. In this mode, the control system manages all execution units as ideal objects with homogeneous and constant states without discrimination, for example, in industrial processes such as aeration oxygenation or fluid distribution, even if part of the processing units has lost processing capacity due to physical blockage or wear, energy and materials will still be allocated to them according to the fixed program, which not only causes direct waste of energy, but also exacerbates the deterioration of the faulty units, and at the same time, unnecessary burden is brought to the normally operating units. Therefore, improvement is needed. SUMMARY

[0004] The purpose of the present application is to solve the problems existing in the prior art and to provide an artificial wetland intelligent scheduling control method for energy saving and consumption reduction.

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme, an artificial wetland intelligent scheduling control method for energy saving and consumption reduction, comprising the following steps: Collecting the wetland unit inflow flow monitoring value, the inflow end liquid level monitoring value and the outflow end liquid level monitoring value, obtaining the difference between the inflow end liquid level monitoring value and the outflow end liquid level monitoring value, calculating the generated substrate block impedance coefficient value, mapping the substrate block impedance coefficient value, and generating a unit health state score; Comparing the substrate block impedance coefficient value with the set high impedance threshold value, locking the corresponding electric valve array of the unit exceeding the high impedance threshold value as the shutdown target, and the unit below the high impedance threshold value is a low impedance unit, accumulating the unit health state score of the remaining low impedance units to obtain a total distribution weight, generating a flow path reconstruction guide instruction according to the proportion of the individual unit health state score in the total distribution weight; According to the flow path reconstruction guide instruction, a list of units in a working state is extracted, a regional dissolved oxygen monitoring value inside a corresponding unit is collected, a gas supply duration is obtained, a proportion value of the gas supply duration in a total gas supply cycle is calculated, and an aeration array duty cycle parameter is generated; The aeration array duty cycle parameter is called, the total air volume demand under each time slice is calculated, the total air volume demand is matched with the energy efficiency characteristic curve of the central air blower, the operation frequency of the central air blower and the action timing of each valve are determined, and a multi-device cooperative scheduling time sequence instruction is generated.

[0006] Preferably, the unit health state score obtaining step is: According to the wetland unit inflow flow monitoring value, the inflow end liquid level monitoring value and the outflow end liquid level monitoring value, records with missing or abnormal monitoring are removed, the inflow end liquid level monitoring value is subtracted from the outflow end liquid level monitoring value to obtain a liquid level difference value; According to the liquid level difference value, a substrate blockage impedance coefficient value is calculated; According to the substrate blockage impedance coefficient value, standardization mapping and range clipping are performed to form a unit health state score.

[0007] Preferably, the total distribution weight obtaining step is: The substrate blockage impedance coefficient value of each wetland unit is compared with a high impedance threshold value, the electric valve array that meets the greater than or equal condition is registered as a shutdown target, the units that meet the less than condition are summarized as low impedance units according to the original number, and a low impedance unit is obtained; According to the low impedance unit, the corresponding unit health state score is extracted according to the index, the missing record is removed and the effective value is retained, all effective unit health state scores are added item by item according to a unified order, and a total distribution weight is obtained.

[0008] Preferably, the flow path reconstruction guide instruction obtaining step is: According to the total distribution weight, the proportion of the unit health state score of each low impedance unit is calculated, the total hydraulic load is converted into an electric valve array opening parameter according to the proportion, and the low impedance unit index is paired to form a flow path reconstruction guide instruction.

[0009] Preferably, the gas supply duration obtaining step is: According to the flow path reconstruction guide instruction, the electric valve array opening parameter and the action flag of each unit are analyzed, the units with the action flag turned on and the opening parameter greater than zero are selected and de-duplicated according to the unit index, and a list of units in a working state is obtained. According to the unit list in the working state, the regional dissolved oxygen monitoring value is collected according to the unit index, the target setting value is read, the numerical deviation of the regional dissolved oxygen monitoring value minus the target setting value is calculated, the missing record is removed, the numerical deviation is matched to the unique output item according to the condition item of the fuzzy logic rule set, and the required air supply time length of the blower branch valve is obtained.

[0010] Preferably, the obtaining step of the aeration array duty cycle parameter is: According to the required air supply time length of the blower branch valve, the branch ratio value is converted with the total air supply period length as the denominator, the oxygen demand level code is divided according to the median and upper quartile of the branch ratio value, and the unit index is paired to generate the aeration array duty cycle parameter.

[0011] Preferably, the obtaining step of the total air volume requirement is: According to the aeration array duty cycle parameter, the duty cycle of each blower branch valve and the corresponding total air supply period are extracted, the opening time length is calculated according to the duty cycle and the total air supply period, and the opening time coordinates are arranged in sequence, the adjacent opening time intervals are detected whether they overlap, if overlapping is found, the opening time coordinate of the latter valve is sequentially extended to the end point of the former interval, and a non-overlapping start sequence is generated. According to the non-overlapping start sequence, the time-sliced total air volume requirement is calculated.

[0012] Preferably, the obtaining step of the multi-device collaborative scheduling time sequence instruction is: According to the total air volume requirement, the corresponding air volume point is found in the central blower efficiency characteristic curve, the running frequency value is read, the running frequency is synchronized with the opening time coordinate of each blower branch valve in the non-overlapping start sequence, and the multi-device collaborative scheduling time sequence instruction is formed.

[0013] Compared with the prior art, the advantages and positive effects of the present application are: The present application collects the hydraulic parameters of the wetland unit in real time, identifies and shuts off the high impedance unit, actively avoids water flow to the unit that has occurred obstruction, and distributes to the remaining unobstructed units according to the health degree, which not only avoids the energy waste of delivering water to invalid units, but also prolongs the effective operation life of the entire wetland system through intelligent load balancing. On the basis of completing the flow path optimization, the aeration time length is further determined according to the real dissolved oxygen demand in the working unit, the matching of air supply and oxygen demand is realized, the power consumption caused by blind aeration is eliminated, and finally the discrete air supply demand of each branch is sequenced and staggered on the time axis to form a smooth and predictable total air volume requirement curve. The curve is matched with the energy efficiency characteristics of the central blower, so that the blower always operates in the most efficient working interval, realizing collaborative energy saving and consumption reduction. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 The schematic diagram of the steps of the present application. DETAILED DESCRIPTION

[0015] The specific embodiments of the present application are further described in detail below in conjunction with the accompanying drawings and examples.

[0016] Referring to Figure 1 The present application provides a technical solution, an artificial wetland intelligent scheduling control method for energy saving and consumption reduction, comprising the following steps: Collecting wetland unit inflow flow monitoring value, inflow end liquid level monitoring value and outflow end liquid level monitoring value, obtaining the difference value of the inflow end liquid level monitoring value and the outflow end liquid level monitoring value, calculating the generated substrate block impedance coefficient value, mapping the substrate block impedance coefficient value, and generating unit health state score; Comparing the substrate block impedance coefficient value with the set high impedance threshold value, locking the corresponding electric valve array of the unit exceeding the high impedance threshold value as the shutdown target, and the unit below the high impedance threshold value as the low impedance unit, accumulating the unit health state score of the remaining low impedance units to obtain the total distribution weight, generating the flow path reconstruction guidance instruction according to the proportion of the individual unit health state score in the total distribution weight; According to the flow path reconstruction guidance instruction, extracting the unit list in the working state, collecting the regional dissolved oxygen monitoring value inside the corresponding unit, obtaining the gas supply time length, calculating the proportion value of the gas supply time length in the total gas supply cycle, and generating the air array duty cycle parameter; Calling the air array duty cycle parameter, calculating the total air demand under each time slice, matching the total air demand with the energy efficiency characteristic curve of the central air blower, determining the running frequency of the central air blower and the action time of each valve, and generating the multi-device collaborative scheduling time sequence instruction.

[0017] The unit health state score acquisition step is: According to the wetland unit inflow flow monitoring value, the inflow end liquid level monitoring value and the outflow end liquid level monitoring value, eliminating the records with missing or abnormal monitoring, subtracting the outflow end liquid level monitoring value from the inflow end liquid level monitoring value to obtain the liquid level difference value; According to the liquid level difference value, calculating the substrate block impedance coefficient value, the calculation formula is: ; Wherein, z is the substrate block impedance coefficient value, d is the liquid level difference value, q is the wetland unit inflow flow monitoring value, is the reference liquid level difference, is the reference flow, is the flow state characteristic index, is the cumulative running time since the last dredging, is the running time decay function, and the expression is , As the attenuation effect factor, Standard maintenance cycle; Based on the matrix blocking impedance coefficient value, a standardized mapping and range trimming are performed to form a cell health status score.

[0018] Specifically, based on the monitored values ​​of the wetland unit's influent flow rate, influent level, and effluent level, ultrasonic level gauges and electromagnetic flow meters are configured. The sampling frequency is set to 60 data acquisitions per minute to obtain raw time-series monitoring data. For the acquired raw data, a data cleaning process is performed. Using a sliding window algorithm with a window size of 5 minutes, the mean and standard deviation of the data within the window are calculated. For flow monitoring values, if the value at a certain moment deviates from the window mean by more than twice the standard deviation, or if the flow reading exceeds the maximum rated flow value of the physical pumping station (e.g., 1.2 times the design flow), it is identified as an abnormal abrupt change and removed. Simultaneously, the continuity of the level monitoring data is checked. If the rate of change of the influent or effluent level exceeds 5 cm per minute within three consecutive sampling periods, this is an impossible condition in actual physical hydraulic processes and is identified as sensor jump interference. This segment of the record is marked as invalid. For the missing time points that were removed, cubic spline interpolation was used to fit curves using 10 valid data points before and after the missing time points, thus completing the flow rate and liquid level values ​​at the missing time points and ensuring the temporal integrity of the data. Subsequently, the liquid level monitoring values ​​at the inlet and outlet ends after cleaning and completion were extracted at the same time, and synchronous differential calculation was performed, that is, the absolute elevation value at the inlet end was subtracted from the absolute elevation value at the outlet end. In this process, considering the basic elevation error of the sensor installation position, a fixed calibration bias was introduced (this bias was obtained by measuring the difference between the readings of the liquid level gauges at both ends in a static clean water test, for example, 0.02 meters). The calculated original difference was subtracted from this calibration bias, thereby eliminating the systematic deviation caused by the installation error. Finally, a liquid level difference value sequence reflecting the actual hydraulic resistance characteristics inside the wetland unit was obtained. This sequence accurately eliminated the false fluctuations caused by wave disturbance, sensor noise and electrical interference, and obtained the liquid level difference value.

[0019] In the formula for calculating the matrix blockage impedance coefficient, by introducing the flow characteristic index and the operating time decay function, the simple physical liquid level difference is transformed into a dimensionless impedance coefficient that reflects the changes in the matrix pore structure. This eliminates the interference of hydraulic load fluctuations on blockage judgment and, combined with the natural sedimentation pattern in the time dimension, improves the accuracy of assessing the health status of wetland units. d represents the liquid level difference in meters. This parameter is obtained by calculating the difference between the inlet and outlet liquid levels at each sampling point, based on the cleaned and calibrated data from the previous steps. To smooth out instantaneous fluctuations, the average liquid level difference over the most recent 10 minutes is used as the current input parameter d. For example, in a monitoring session, if the average inlet liquid level is 1.5 meters and the average outlet liquid level is 1.1 meters, the calculated difference is... rice; q represents the monitored influent flow rate of the wetland unit, in cubic meters per hour. This parameter is obtained by reading the instantaneous reading from the electromagnetic flowmeter in the influent pipe and then integrating and averaging it over the control cycle to eliminate the influence of fluid pulsation. For example, the average influent flow rate read is... cubic meters per hour; The reference water level difference, expressed in meters, is obtained as follows: In the initial state of a newly constructed or thoroughly dredged wetland unit, clean water is introduced, the influent flow rate is adjusted to the design rated value, and the unit is run continuously for 24 hours. The influent and effluent water level differences are recorded every hour, and the average of these 24 sets of data is taken as the reference water level difference. For example, if the initial reference water level difference is determined to be... rice; The reference flow rate, expressed in cubic meters per hour, is obtained by consulting the constructed wetland process design specifications and selecting the flow rate corresponding to the optimal hydraulic load for the wetland unit under design conditions. For example, the optimal treatment flow rate specified in the design specifications is... cubic meters per hour; The flow characteristic index is dimensionless. The steps for obtaining this parameter are as follows: During the initial commissioning phase of the wetland, a variable flow hydraulic test is conducted. The influent flow rate is gradually increased from 50% to 150% of the design flow rate, with a total of 10 flow rate gradients set. Each gradient is run stably for 2 hours, and the corresponding flow rates are recorded. and liquid level difference ,Establish Using a coordinate system, a linear regression is performed using the least squares method. The slope of the regression equation is the flow regime characteristic index, which can be obtained through regression calculation. ; This is the cumulative operating time since the last dredging, in days. The steps to obtain this parameter are as follows: retrieve the most recent timestamp marked "dredging completed" or "media replacement" from the database's operation and maintenance log table; read the current system clock; and calculate the time difference between the two. For example, if the last dredging was 200 days ago, then... sky; The attenuation impact factor is dimensionless. The steps to obtain this parameter are as follows: select data from three historical wetland units of the same type but with different operating years, extract their impedance rise factor at the end of the standard maintenance cycle, and calculate the formula as follows: ,in The impedance at the end of the period, The initial impedance is taken as the average of the calculation results of three units. For example, based on historical data analysis, the impedance at the end of the period is usually 2.5 times the initial value. Substituting this into the calculation yields... For ease of calculation, here we take ; The standard maintenance cycle is expressed in days. This parameter is obtained by determining the theoretical cleaning cycle required for the wetland packing material (e.g., gravel, expanded clay) from new to completely clogged, based on the material characteristics of the packing and the influent suspended solids concentration load, referring to industry design specifications. For example, it can be set as follows: Days (i.e., 2 years); Calculations based on parameters: First, calculate the hydraulic load correction term: Next, calculate the liquid level difference ratio: Calculate the normalized basic impedance value: Calculate the runtime decay function value: Finally, the matrix blocking impedance coefficient value was calculated: The results indicate that the current wetland unit has a matrix blockage resistance coefficient of 4.69, which is much greater than 1. This means that after deducting the normal effects of flow rate changes and natural time decay, there is obvious abnormal physical blockage or siltation inside the unit. The resistance has reached more than 4 times that of the initial state. If no intervention is taken, it will seriously affect the water flow capacity and treatment effect.

[0020] Based on the matrix blockage impedance coefficient value, a pre-built health state mapping model is invoked to first determine the physical boundary of the impedance coefficient. Then, based on historical extreme operating condition data analysis, an upper limit threshold for impedance when the matrix completely fails is set (e.g., ...). ) and the lower limit threshold of impedance under ideal clean conditions (e.g. ), for the calculated impedance coefficient value Perform reverse normalization to construct a linear mapping formula, i.e. ,in The current calculated value is 4.69. Substituting this value into the calculation yields the preliminary score. To prevent numerical values ​​from exceeding limits due to sensor drift, a range clipping operation is performed on the initial score, and logical judgment conditions are set. If the value is greater than 100, then the value will be forcibly assigned to 100. If the score is less than 0, it is forcibly assigned a value of 0 to ensure that the score always falls within the valid range of 0 to 100. In addition, a non-linear penalty mechanism is introduced. For the critical region below 60 points, a piecewise correction strategy is adopted, that is, when... When, execute To amplify the risk signal of low health units, in this example, 59.0 is less than 60, so the final score is adjusted to... The lower the score, the more severe the blockage, forming the unit's health status score.

[0021] The steps to obtain the total allocation weight are as follows: The matrix blocking impedance coefficient value of each wetland unit is compared with the high impedance threshold one by one. Electric valve arrays that meet the condition of being greater than or equal to the threshold are registered as shut-off targets, and units that meet the condition of being less than the threshold are summarized as low impedance units according to their original numbers, thus obtaining low impedance units. Based on the low impedance unit, extract the corresponding unit health status score by index, remove missing test records and retain valid values, and add up all valid unit health status scores in a uniform order to obtain the total allocation weight.

[0022] Specifically, the matrix blocking impedance coefficient value of each wetland unit is compared with the high impedance threshold. First, a specific value for the high impedance threshold is set, based on statistical analysis of historical operational data. Impedance monitoring data recorded over the past three years during wetland unit overflows or severe blockages are retrieved, and 50 impedance value samples at the time of failure are extracted. The mean (e.g., 9.5) and standard deviation (e.g., 0.5) of this sample set are calculated. To ensure timely intervention before complete blockage occurs, a statistical method of subtracting three times the standard deviation from the mean is used to determine the safety boundary, resulting in the calculated threshold. The value 8.0 is used as the critical criterion for determining whether a unit needs to be shut down. Then, a traversal program is started to sequentially read the real-time matrix blockage impedance coefficient values ​​of all online wetland units. For example, if the impedance value of unit A is 4.69 and the impedance value of unit B is 9.2, these two values ​​are logically compared with the set threshold 8.0. For unit B, since its impedance value of 9.2 is greater than the set threshold 8.0, it indicates that the internal pores of this unit are severely clogged and cannot handle normal water flow, triggering the shutdown protection logic. The hardware address of the corresponding inlet valve of this unit is queried, and the control register corresponding to that address is marked as "shutdown pending maintenance" and registered as the target object requiring shutdown to prevent sewage from continuing to be injected and causing an overflow accident. For unit A, since its impedance value of 4.69 is less than the set threshold 8.0, it indicates that this unit still has effective hydraulic conduction capacity and is judged to be in a healthy or sub-healthy state. Its index number is retained and added to the available resource list in sequence. All unit indices that meet this condition are summarized to obtain low-impedance units.

[0023] Based on the low-impedance unit, using the available resource list generated in the preceding steps as the index key, the unit status table in the real-time database is accessed one by one to read the unit health status score corresponding to each index. For example, if the first unit retrieved in the list has a score of 53.1 and the second unit has a score of 88.2, a data validity verification process is executed during this process. This checks whether each read score is accompanied by a valid sensor status flag and verifies whether the timestamp of the data is within the allowable deviation range of the current control cycle (e.g., within the last 10 minutes). If it is found that a unit is in the list, but its score data is displayed as "Null" due to a communication failure or the value exceeds the logical definition range of 0 to 100 (e.g., -1), the verification will be performed. If a record is missing or abnormal, it is discarded from the calculation buffer and not included in subsequent statistics to eliminate the interference of dirty data on the scheduling strategy. For values ​​that have been verified as valid, an accumulation operation is performed. A double-precision floating-point variable is initialized as an accumulator. The first valid score of 53.1 is added to the accumulator in the order of the list, followed by the second valid score of 88.2, and so on. The entire list of low-impedance units is traversed, and the health scores of all units that are in working condition and have normal data are mathematically summed. For example, the final calculated accumulation result is 450.5. This value represents the sum of the overall hydraulic acceptance capacity evaluation of all currently available wetland units, and the total allocation weight is obtained.

[0024] The steps for obtaining the flow path reconfiguration guidance instruction are as follows: Based on the total allocation weight, the proportion of the health status score of each low-impedance unit is calculated. The total hydraulic load is then converted into electric valve array opening parameters according to the proportion and paired with the low-impedance unit index to form a flow path reconfiguration guidance command.

[0025] Specifically, based on the total allocation weight, the current total influent flow rate setpoint of the wastewater treatment system is read, for example, 1000 cubic meters per hour. Using the previously calculated total allocation weight (e.g., 450.5) as the denominator, the health status score of each low-impedance unit is extracted as the numerator, and a division operation is performed to determine the load allocation ratio of each unit. For example, the allocation ratio of unit A is calculated as follows: The allocation ratio of computing unit B is: Similarly, the calculated proportional coefficients are then multiplied by the total influent flow rate setpoint to obtain the target flow rate that each unit should handle, i.e., the flow rate that unit A needs to allocate. For unit B, a flow rate of 195.8 cubic meters per hour is allocated. Then, a pre-set electric valve flow characteristic model is invoked. This model is built based on the Cv value curve provided by the valve manufacturer or on-site calibration data (for example, for valves with equal percentage characteristics, the relationship between flow rate and opening degree satisfies a specific exponential function). Based on the target flow rate value calculated by each unit, the corresponding valve physical opening percentage is found in the characteristic curve. For example, by looking up the table, it is found that the valve opening degree corresponding to the flow rate of 117.9 cubic meters per hour for unit A under the current pressure difference is 45%. This opening degree value is paired and mapped with the hardware control address of the unit, and encapsulated into a control message containing the device ID, target opening degree and action priority, forming a flow path reconfiguration guidance instruction.

[0026] The steps to obtain the gas supply duration are as follows: Based on the flow path reconstruction guidance instructions, the opening parameters and action flags of the electric valve array of each unit are analyzed. Units with the action flag "open" and the opening parameter greater than zero are selected and deduplicated by unit index to obtain a list of units in working state. Based on the list of units in operation, the dissolved oxygen monitoring values ​​of the area are collected by unit index and the target set value is read. The numerical deviation between the regional dissolved oxygen monitoring value and the target set value is calculated and missing records are eliminated. The numerical deviation is matched to the unique output item according to the condition item of the fuzzy logic rule set to obtain the gas supply time required by the blower branch valve.

[0027] Specifically, based on the flow path reconfiguration guidance instruction, the instruction parser is activated to read the control fields encapsulated in the instruction package. These fields contain a set of configuration metadata for each wetland unit, specifically consisting of a unique unit index, an action flag, and an opening setting value. Each data item in the instruction package is traversed, and its action flag and opening setting value are extracted for logical judgment. The valid state definition of the action flag is set; for example, a binary value "1" represents an opening request, and "0" represents a closing request. Simultaneously, the valid range of the opening setting value is set to 0% to 100%. A dual-condition filtering is performed on each instruction; only when the action flag is equal to "1" and the opening setting value is strictly greater than 0% is it determined that the unit has been assigned an actual hydraulic load task. For nodes that need to be put into operation, for example, in a single parsing process, if the action flag of the unit with index #05 is "1" and the opening degree is 30%, it is marked as a candidate object, while the action flag of the unit with index #08 is "0" or the opening degree is 0%, it is ignored. To prevent duplicate records caused by communication retransmission or instruction redundancy, a hash set data structure is introduced. The unit index is used as a unique hash key value. The selected candidate objects are tried to be inserted into the set in turn. If the key value already exists, it is automatically overwritten or skipped, ensuring that each physical unit appears only once in the list. After completing the traversal and deduplication of all instructions, the elements in the set are exported and sorted by index from smallest to largest to construct a clean dynamic running list, resulting in a list of units in working state.

[0028] Based on the list of operating units, each wetland unit in the list is polled using the Modbus-TCP communication protocol. Its associated water quality monitoring substation is accessed to read the real-time register value of the dissolved oxygen (DO) sensor. Simultaneously, the corresponding unit's process target setpoint is retrieved from the central control database. This setpoint is typically determined based on the wetland plant root respiration requirements and microbial degradation characteristics, for example, set to 2.0 mg / L. A numerical deviation calculation is performed by subtracting the target setpoint from the real-time monitoring value. For example, if a unit's monitoring value is 1.2 mg / L, the deviation is 1.2 - 2.0 = -0.8 mg / L. If a sensor disconnection or numerical overflow occurs during monitoring, the record is marked as invalid and discarded. For the calculated valid numerical deviation, a predefined model is introduced. Matching is performed using a fuzzy logic rule set, which is constructed based on long-term operational experience. The deviation is divided into four fuzzy intervals: "Severe Insufficiency," "Slight Insufficiency," "Suitable," and "Excessive." The critical thresholds for each interval are set as follows: During the historical commissioning phase, it was observed that when the DO deviation was below -1.0, the pollutant removal rate decreased by 50%, so -1.0 was set as the lower bound for "Severe Insufficiency." When the deviation was between -0.2 and +0.2, the effluent quality was stable, so this range was set as "Suitable." The specific matching process is as follows: If the deviation is less than -1.0, the matching output is "Full-cycle Gas Supply," meaning the gas supply duration equals the total gas supply cycle; if the deviation is between -1.0 and -0.2 (as mentioned above, -0.8), the matching output is "Incremental Gas Supply," with the duration calculated through linear interpolation using the following formula: ,in For the total cycle, The absolute value of the deviation, for example, if the total cycle is 60 minutes, is substituted into the calculation to obtain... Minutes; if the deviation is between -0.2 and 0.2, the matching output item is "maintain air supply", and the duration is set to 10% of the cycle to maintain stirring; if the deviation is greater than 0.2, the matching output item is "stop air supply", and the duration is 0. Through this logic, the required air supply duration of the blower branch valve is obtained.

[0029] The steps for obtaining the duty cycle parameter of the aeration array are as follows: Based on the required air supply time of the blower branch valves, the proportion of each branch is calculated using the total air supply cycle length as the denominator. The oxygen demand level is coded according to the median and upper quartile of the proportion and matched with the unit index to generate the aeration array duty cycle parameters.

[0030] Specifically, based on the required air supply duration of the blower branch valves, the currently set total air supply cycle length is read, for example, 60 minutes. A division operation is then performed, dividing the calculated air supply duration for each branch by the total cycle length to obtain the dimensionless duty cycle value. For example, if the air supply duration for a branch is 52.5 minutes, then the duty cycle is... A statistical sample set is constructed by collecting the duty cycle values ​​of all work units. This sample set is then sorted in ascending order, and the median (Q2) and upper quartile (Q3) of the sample set are determined using the position index method. For example, if the sample set is [0.1, 0.2, 0.5, 0.8, 0.9], the median is 0.5 and the upper quartile is 0.8. Based on these two statistical characteristics, a standard for classifying oxygen demand levels is constructed: if a unit's duty cycle is less than or equal to the median, it is classified as "low oxygen demand" and coded as L1; if the duty cycle is between the median and the upper quartile, it is classified as "medium oxygen demand" and coded as L2; ​​if the duty cycle is greater than the upper quartile, it is classified as "high oxygen demand" and coded as L3. This dynamic classification method based on statistical distribution can adapt to changes in the overall load in different seasons. The index of each unit is associated and combined with its calculated duty cycle value and corresponding level code, and encapsulated into a structured parameter package, for example, {Index: #05, Duty Cycle: 0.875, Level:}. L3} is used so that subsequent scheduling algorithms can perform grouping and staggered peak control according to the level, and generate the duty cycle parameters of the aeration array.

[0031] The steps to obtain the total air volume requirement are as follows: Based on the duty cycle parameters of the aeration array, the duty cycle and corresponding total air supply cycle of each blower branch valve are extracted. The opening duration is calculated according to the duty cycle and the total air supply cycle, and the opening time coordinates are arranged in sequence. The overlapping of adjacent opening time intervals is detected. If an overlap is found, the opening time coordinate of the next valve is extended to the end point of the previous interval to generate a non-overlapping start sequence. Based on the non-overlapping startup sequence, the total air volume requirement for the time slice is calculated using the following formula: in, Let M be the total air volume requirement for the j-th time slice, M be the number of blower branch valves, and r be the total air volume requirement for the j-th time slice. i The instantaneous air supply is adjusted for the back pressure of the i-th blower branch valve. It is a unit step function. Let s be the center time of the j-th time slice. i Let be the opening time coordinate of the valve in the i-th blower branch. Let be the opening duration of the valve in the i-th blower branch, and , Let be the duty cycle of the i-th blower branch valve, T be the total air supply cycle, and r be the instantaneous air supply rate corrected for back pressure. i The calculation formula is , Let be the design air supply capacity of the i-th blower branch valve under nominal operating conditions. This refers to the main pipe pressure at the blower outlet under nominal operating conditions. Let g be the density of water, g be the acceleration due to gravity, and h be the acceleration due to gravity. i Let i be the actual water depth of the aeration head of the valve in the i-th blower branch. This is the reference water depth for measuring the nominal gas supply.

[0032] Specifically, based on the duty cycle parameters of the aeration array, the duty cycle and corresponding total air supply cycle of each blower branch valve are extracted. A structure array containing the valve's physical address, the calculated duty cycle value, and the set cycle is constructed in memory. This array is iterated through, and a multiplication operation is performed, multiplying the duty cycle by the total air supply cycle to obtain the actual opening time required for each valve in the current control cycle (e.g., a duty cycle of 0.5 multiplied by a cycle of 60 minutes equals 30 minutes). Then, a time axis coordinate list is initialized, temporarily setting the default start time of all valves to the zero point of the cycle. To avoid excessive blower start-up load or a surge in pipeline pressure caused by all valves opening simultaneously, a greedy sorting algorithm is used to rearrange the valve list in descending order of opening time. After sorting, a conflict detection and time axis laying loop is executed, reading the opening time interval of the first valve in the sorted sequence. The system locks the valve at the beginning of the time axis, then reads the next valve and checks if its default start time (zero point) falls within the time interval already occupied by the previous valve. That is, it determines whether the expected opening time of the current valve is less than the end time of the previous valve. If an overlap conflict is found, a delay operation is performed, modifying the opening time coordinate of the current valve to the end time of the previous valve plus a small hydraulic buffer protection interval. This interval is set based on the mechanical action delay characteristics of the pneumatic valve. The physical travel time of the valve from fully closed to fully open is captured by a high-speed camera (e.g., 2 seconds), and a 50% margin is added to eliminate the risk of water hammer effect. The calculated buffer interval is 3 seconds. This process is repeated, adjusting the opening coordinates of subsequent valves one by one until the time intervals of all valves are non-interfering and linearly arranged on the time axis. Finally, the parameter set with the adjusted start and end times is encapsulated to generate a non-overlapping start sequence.

[0033] In the formula for calculating the total air volume demand of time slices, a time window filter is constructed by introducing a unit step function to identify the set of valves in the open state under each tiny time slice. By calculating the square root of the ratio of the actual effective driving pressure difference to the nominal pressure difference, the nonlinear effect of water level change on aeration air volume is quantified, ensuring that the calculation of the system air supply is consistent with the real physical working conditions. M represents the number of valves in the blower branch circuit, expressed in units of one. The steps to obtain this parameter are as follows: read the hardware configuration file of the industrial control system and count the number of effective control valve nodes connected to the current blower system. For example, if the system is configured with 4 independent aeration branches, then M=4. The design air supply capacity of the i-th blower branch valve under nominal operating conditions is expressed in cubic meters per minute. The steps to obtain this parameter are as follows: Based on the aerator product manual, select the unit flow rate under standard test conditions (usually constant water depth and pressure), and multiply it by the actual number of aerator heads installed in that branch. For example, if the unit nominal flow rate is 0.1... If 100 branch lines are installed, then ; This is the total pipe pressure (gauge pressure) at the blower outlet under nominal operating conditions, expressed in Pascals (Pa). To obtain this parameter, consult the blower's nameplate or performance curve to find its rated exhaust pressure (relative to atmospheric pressure). For example, a Roots blower's rated exhaust gauge pressure is 60 kPa. ; The density of water is expressed in kilograms per cubic meter (kg / m³). ); g is the acceleration due to gravity, measured in meters per second squared (m²). The steps to obtain this parameter are as follows: take the standard gravitational acceleration value, i.e. ; h i Let h be the actual water depth of the aeration head of the i-th blower branch valve, in meters (m). This parameter is obtained by collecting the real-time water column height using a pressure level gauge installed at the bottom of the pool, and subtracting the aeration head installation height. For example, if the level gauge reading is 1.8 meters and the aeration head is 0.2 meters from the bottom, then h... i =1.6m; The reference water depth, measured in meters (m), is used to measure the nominal gas supply. This parameter is obtained by setting the test conditions according to the manufacturer's calibration report. For example, if the nominal flow rate is measured under back pressure at a water depth of 4.0 meters, then... =0.4m; t j Let t be the center time of the j-th time slice, in minutes. The steps to obtain this parameter are as follows: Discretize the control cycle, for example, to calculate the state at the 10th minute, then t... j =10; s i Let be the opening time coordinate of the i-th blower branch valve, in minutes (min). This parameter is obtained by directly extracting it from the non-overlapping start-up sequence. For example, if valve 1 opens at time 0, then s... i =0; θ i Let θ be the opening duration of the valve in the i-th blower branch, in minutes (min). This parameter is obtained by multiplying the duty cycle by the period. For example, if the period is 30 minutes, then θ is...i =30; Calculations based on parameters: For example, the system has two branch valves (M=2), and the total gas supply cycle is T=60 minutes.

[0034] Valve 1 parameters: Actual water depth m.

[0035] Valve 2 parameters: Actual water depth m.

[0036] Common parameters: .

[0037] The calculation time is t. j =10 minutes (during the valve 1 opening period): First, calculate the effective driving pressure difference under nominal operating conditions (denominator): ; Next, calculate the effective driving pressure difference (numerator) for each valve under actual operating conditions: For valve 1: ; For valve 2: ; Calculate the back pressure correction instantaneous gas supply and : ; ; Calculate the state of the step function (time). ): For valve 1: The state is 1.

[0038] For valve 2: The state is 0.

[0039] Calculate total air volume requirements : ; The results indicate that at the 10-minute mark, due to the shallow water level (1.6 meters) in the wetland unit where valve 1 is located, the actual back pressure is much lower than the nominal back pressure (4.0 meters), resulting in a significant increase in the actual gas flow rate through the aerator head to 14.61 cubic meters per second under the same air supply pressure. Therefore, the central control system needs to identify this surge in flow demand and adjust the fan frequency accordingly to match the physical conditions, preventing dissolved oxygen oversaturation or energy waste due to excessive air volume.

[0040] The steps for obtaining multi-device collaborative scheduling timing instructions are as follows: Based on the total air volume demand, the corresponding air volume point is found in the energy efficiency characteristic curve of the central blower and the operating frequency value is read in time slices. The operating frequency is synchronously marked with the opening time coordinate of each blower branch valve in the non-overlapping start sequence to form a multi-device collaborative scheduling timing instruction.

[0041] Specifically, based on the total air volume demand, the corresponding air volume point is found in the energy efficiency characteristic curve of the central blower for each time slice, and the operating frequency value is read. The three-dimensional mapping relationship table of flow rate-frequency-power is obtained from the blower manufacturer's technical manual or laboratory test report. A continuous energy efficiency characteristic function is constructed using a linear interpolation algorithm. The total air volume demand value of each time slice calculated in the previous steps is used as the input variable. The minimum operating frequency that can meet the air volume output is looked up in the characteristic function. For example, for a demand of 14.61 cubic meters per minute, the corresponding operating frequency of the strain gauge is found to be 42. 5Hz. During this process, frequency safety upper and lower limits are set, for example, the minimum should not be lower than 20Hz to prevent motor overheating, and the maximum should not exceed 50Hz to protect the mechanical structure. If the query result exceeds the range, forced clamping is performed. Subsequently, a composite instruction frame containing timestamp, blower target frequency, and the on / off status of each branch valve is constructed. The calculated frequency value of 42.5Hz is mapped to the corresponding coordinate point on the time axis, and the logical status (open / closed) of each valve at that moment is synchronously associated. These discrete instruction frames are concatenated in time sequence to form a continuous scheduling curve. The curve is sent to the field PLC controller via Modbus TCP protocol to form a multi-device collaborative scheduling timing instruction.

Claims

1. A method for intelligent scheduling and control of constructed wetlands aimed at energy conservation and emission reduction, characterized in that, Includes the following steps: Collect the influent flow rate, influent level, and effluent level of the wetland unit, obtain the difference between the influent level and effluent level, calculate the matrix blocking impedance coefficient, and map the matrix blocking impedance coefficient to generate a unit health status score. The matrix blocking impedance coefficient value is compared with the set high impedance threshold. The electric valve array corresponding to the unit that exceeds the high impedance threshold is locked as the shutdown target, while the unit that is below the high impedance threshold is a low impedance unit. The unit health status score of the remaining low impedance units is accumulated to obtain the total allocation weight. According to the proportion of the individual unit health status score to the total allocation weight, the flow path reconfiguration guidance instruction is generated. According to the flow path reconstruction guidance command, extract the list of units in working state, collect the regional dissolved oxygen monitoring value inside the corresponding unit, obtain the air supply duration, calculate the proportion of the air supply duration in the total air supply cycle, and generate the aeration array duty cycle parameter. The duty cycle parameters of the aeration array are called to calculate the total air volume requirement for each time slice. The total air volume requirement is matched with the energy efficiency characteristic curve of the central blower to determine the operating frequency of the central blower and the timing of the action of each valve, and a multi-device collaborative scheduling sequence instruction is generated.

2. The intelligent scheduling and control method for constructed wetlands aimed at energy conservation and emission reduction according to claim 1, characterized in that, The steps for obtaining the unit's health status score are as follows: Based on the monitoring values ​​of the influent flow rate, influent level, and effluent level of the wetland unit, records with missing or abnormal measurements are removed, and the effluent level is subtracted from the influent level to obtain the level difference. Calculate the matrix blockage resistance coefficient based on the liquid level difference. Based on the matrix blocking impedance coefficient value, a standardized mapping and range clipping are performed to form a cell health status score.

3. The intelligent scheduling and control method for constructed wetlands aimed at energy conservation and emission reduction according to claim 1, characterized in that, The steps for obtaining the total allocation weight are as follows: The matrix blocking impedance coefficient value of each wetland unit is compared with the high impedance threshold one by one. Electric valve arrays that meet the condition of being greater than or equal to the threshold are registered as shut-off targets, and units that meet the condition of being less than the threshold are summarized as low impedance units according to their original numbers to obtain low impedance units. Based on the low impedance unit, extract the corresponding unit health status score by index, remove missing test records and retain valid values, and add all valid unit health status scores one by one in a uniform order to obtain the total allocation weight.

4. The intelligent scheduling and control method for constructed wetlands aimed at energy conservation and emission reduction according to claim 1, characterized in that, The steps for obtaining the flow path reconfiguration guidance instruction are as follows: Based on the total allocation weight, the proportion of the health status score of each low-impedance unit is calculated. The total hydraulic load is converted into electric valve array opening parameters according to the proportion and matched with the low-impedance unit index to form a flow path reconfiguration guidance command.

5. The intelligent scheduling and control method for constructed wetlands aimed at energy conservation and emission reduction according to claim 1, characterized in that, The steps for obtaining the gas supply duration are as follows: According to the flow path reconstruction guidance instruction, the opening parameters and action flags of the electric valve array of each unit are parsed, and units with the action flag of "open" and the opening parameter greater than zero are filtered out and deduplicated by unit index to obtain a list of units in working state. Based on the list of units in operation, the dissolved oxygen monitoring values ​​of the area are collected by unit index and the target set value is read. The numerical deviation between the regional dissolved oxygen monitoring value and the target set value is calculated and missing records are eliminated. The numerical deviation is matched to the unique output item according to the condition item of the fuzzy logic rule set to obtain the gas supply time required by the blower branch valve.

6. The intelligent scheduling and control method for constructed wetlands aimed at energy conservation and emission reduction according to claim 1, characterized in that, The steps for obtaining the duty cycle parameter of the aeration array are as follows: Based on the required air supply duration of the blower branch valves, the proportion of each branch is calculated using the total air supply cycle length as the denominator. The oxygen demand level is coded according to the median and upper quartile of the proportion and matched with the unit index to generate the aeration array duty cycle parameters.

7. The intelligent scheduling and control method for constructed wetlands aimed at energy conservation and emission reduction according to claim 1, characterized in that, The steps for obtaining the total air volume requirement are as follows: Based on the duty cycle parameters of the aeration array, the duty cycle and corresponding total air supply cycle of each blower branch valve are extracted. The opening duration is calculated according to the duty cycle and the total air supply cycle, and the opening time coordinates are arranged in sequence. The overlapping of adjacent opening time intervals is detected. If an overlap is found, the opening time coordinate of the next valve is extended to the end point of the previous interval to generate a non-overlapping start sequence. Calculate the total air volume requirement for the time slice based on the non-overlapping startup sequence.

8. The intelligent scheduling and control method for constructed wetlands aimed at energy conservation and emission reduction according to claim 1, characterized in that, The steps for obtaining the multi-device collaborative scheduling timing instructions are as follows: Based on the total air volume requirement, the corresponding air volume point is found in the energy efficiency characteristic curve of the central blower and the operating frequency value is read in time slices. The operating frequency is synchronously marked with the opening time coordinate of each blower branch valve in the non-overlapping start sequence to form a multi-device collaborative scheduling timing instruction.

Images