Industrial solid waste landfill leachate treatment failure prediction and health management system

Abstract

This invention relates to the field of industrial wastewater treatment and intelligent control technology, specifically to a fault prediction and health management system for leachate treatment in industrial solid waste landfills. The system includes a process monitoring module, a dosing control module, an edge computing gateway, and a prediction and health management hub. The process monitoring module collects process status data from the anaerobic and aerobic zones and controls fluid operation parameters. The dosing control module collects reagent dosing data and controls the reagent dosing equipment. The hub calculates reagent consumption rate, reagent accumulation concentration, and core health index based on process status data, reagent dosing data, and a kinetic mechanism model, and jointly generates fluid parameter adjustment instructions and dynamic reagent optimization instructions. This invention achieves comprehensive control over fluid treatment equipment and reagent dosing equipment, and maintains stable system operation under load fluctuations and communication anomalies.

Description

Technical Field

[0001] This invention relates to the field of industrial wastewater treatment and intelligent control technology, specifically to a fault prediction and health management system for leachate treatment in industrial solid waste landfills. Background Technology

[0002] In existing industrial solid waste landfill leachate treatment systems, there are usually anaerobic zones, aerobic zones, fluid transport equipment, and reagent dosing equipment. The anaerobic and aerobic zones work together through units such as reflux, aeration, lifting, and discharge to complete the denitrification treatment of the leachate. The reagent dosing equipment is used to add regulating agents to the treatment process to maintain the stable operation of short-cut nitrification and subsequent denitrification processes. During operation, existing systems typically monitor the treatment status using parameters such as dissolved oxygen, pH, oxidation-reduction potential, hydraulic retention time, sludge concentration, and reagent dosage. Based on these parameters, blowers, return pumps, valves, and metering pumps are adjusted to ensure continuous operation of the treatment system under conditions of fluctuating influent water quality, load changes, or external disturbances. However, existing control methods mostly rely on single point-in-time parameters or lagging water quality indicators as the basis for adjustment. It is difficult to identify the actual consumption and accumulation status of the reagents in the sludge system, and it is also difficult to distinguish in a timely manner the changes in the activity of ammonia-oxidizing bacteria, the degree of inhibition of nitrite-oxidizing bacteria, and the changes in denitrification potential. As a result, when the composition of leachate from industrial solid waste landfills exceeds the preset fluctuation threshold, problems such as mis-dosing of reagents, loss of short-range nitrification dominance, instability of bacterial communities under pressure, and untimely fault warnings are likely to occur, thereby causing a decrease in denitrification efficiency and a deterioration in system operation stability. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a fault prediction and health management system for leachate treatment in industrial solid waste landfills. Specifically, the technical solution of this invention includes: The system includes fluid handling equipment, reagent dosing equipment, process monitoring module, dosing control module, edge computing gateway, and prediction and health management center; the fluid handling equipment includes at least an anaerobic zone and an aerobic zone; the reagent dosing equipment stores regulating agents. The process monitoring module is connected to the fluid processing equipment to collect process status data of the fluid processing equipment and control the fluid operating parameters; The dosing control module is connected to the pesticide dosing equipment to collect pesticide dosing data and control the execution of pesticide dosing tasks. The prediction and health management center communicates with the process monitoring module and the dosing control module through an edge computing gateway; The edge computing gateway is used to perform outlier removal and local caching on process status data and drug dosing data before sending them to the prediction and health management center. The prediction and health management center is configured as follows: based on process status data and drug dosing data, and combined with the built-in kinetic mechanism model, the core health index is determined; and control instructions are generated based on the core health index and issued to the process monitoring module and the dosing control module respectively; among them, the control instructions include at least fluid parameter adjustment instructions and dynamic drug optimization instructions.

[0004] Preferably, the process state data includes at least hydraulic retention time, oxidation-reduction potential, pH, dissolved oxygen, ambient temperature, sludge concentration, and actual carbon-to-nitrogen ratio; The prediction and health management center is also used to determine the amount of simultaneous nitrification and denitrification nitrogen removal in the aerobic zone based on process status data.

[0005] Preferably, the core health index includes the ammonia-oxidizing bacteria activity index, the nitrite-oxidizing bacteria inhibition index, and the polysaccharide bacteria denitrification potential index.

[0006] Preferably, the prediction and health management center is also used to monitor the trend changes of the nitrite-oxidizing bacteria inhibition index; If the nitrite-oxidizing bacteria inhibition index shows a continuous downward trend within a preset time window containing multiple consecutive sampling cycles, and the simultaneous nitrification and denitrification nitrogen removal in the aerobic zone is lower than the preset nitrogen removal baseline, then a short-range nitrification destruction warning and fluid parameter adjustment command will be generated.

[0007] Preferably, the fluid parameter adjustment commands include residence time extension commands and dissolved oxygen fine-tuning commands; The process monitoring module responds to the residence time extension command by extending the hydraulic residence time in the anaerobic zone; and responds to the dissolved oxygen fine-tuning command by adjusting the dissolved oxygen concentration in the aerobic zone.

[0008] Preferably, the prediction and health management center dynamically generates dynamic drug optimization instructions within the preset dosage concentration range based on the core health index and the preset dosage concentration range. The dosing control module responds to dynamic reagent optimization commands, adjusts the reagent dosage of the reagent dosing equipment, and regulates the reagents, including hydroxylamine.

[0009] Preferably, the process monitoring module also obtains metagenomic feature values ​​from an external server through a data interface; the metagenomic feature values ​​include the abundance of ammonia assimilation genes, the abundance of ammonia oxidation genes, and the proportion of bacterial communities; The prediction and health management center binds metagenomic feature values ​​with redox potential and pH over time to construct a gene epigenetic mapping dictionary with a time dimension.

[0010] Preferably, the prediction and health management center predicts the current abundance of ammonia assimilation genes based on a gene epigenetic mapping dictionary, and performs abnormal diagnosis in conjunction with the actual carbon-nitrogen ratio: If the current abundance of ammonia assimilation genes is higher than the preset abundance threshold based on historical sequencing data, it is determined to be an enhanced ammonia assimilation state; if the current abundance of ammonia assimilation genes is lower than or equal to the preset abundance threshold, it is determined to be a precursor state of sludge disintegration.

[0011] Preferably, the process monitoring module further includes a communication subsystem that is connected to a cloud server; The prediction and health management center generates a strategy update instruction and sends it to the communication subsystem. After receiving the instruction, the communication subsystem transmits the relevant data to the cloud server and receives feedback data. Based on the feedback data, the prediction and health management center optimizes the generation strategy of the control instruction.

[0012] Compared with the prior art, the present invention has the following beneficial effects: 1. By adopting the above technical solution, this invention collaboratively constructs a closed-loop control system for leachate treatment in industrial solid waste landfills by integrating the process monitoring module, dosing control module, edge computing gateway, and prediction and health management center. A kinetic mechanism model is pre-installed in the center, and the consumption rate and accumulation concentration of the reagent are calculated based on process state data and reagent dosing data. This effectively solves the problem that existing control methods struggle to identify the actual consumption and accumulation status of reagents in the sludge system, easily leading to mis-dosing. 2. This invention further combines dissolved oxygen, hydraulic retention time, oxidation-reduction potential, pH, ambient temperature, sludge concentration, and actual carbon-nitrogen ratio to calculate the nitrogen removal capacity of simultaneous nitrification and denitrification. It also subdivides the core health index into an ammonia-oxidizing bacteria activity index, a nitrite-oxidizing bacteria inhibition index, and a polysaccharide bacteria denitrification potential index. This ensures that the system can identify the activity, inhibition degree, and denitrification capacity of the bacterial community in a more granular manner, thereby effectively overcoming the diagnostic lag and misjudgment caused by relying solely on a single point-in-time parameter or a lagging water quality indicator. 3. This invention, by jointly determining the continuous downward trend of the nitrite-oxidizing bacteria inhibition index within a preset time window and the simultaneous nitrification and denitrification nitrogen removal amount being lower than a preset nitrogen removal baseline, and generating instructions to extend residence time, fine-tune dissolved oxygen, and optimize hydroxylamine dynamic reagents, can intervene in advance at the initial stage of loss of short-range nitrification dominance and bacterial community instability under pressure, thereby significantly improving the timeliness of fault warning and the stability of system operation; 4. This invention constructs a gene epigenetic mapping dictionary with a time dimension by introducing metagenomic feature values, and optimizes and adjusts the control command generation strategy by combining cloud server feedback. At the same time, it is combined with a safe operation mechanism under sensor anomalies, execution limitations, and communication failures. This further improves the interpretability of abnormal ammonia nitrogen loss cause diagnosis, the adaptability of strategy updates under complex operating conditions, and the reliability of continuous on-site operation. Ultimately, it can achieve coordinated management of reagents, microbial community status, and hydraulic operation, improve denitrification efficiency, and reduce the risk of short-range nitrification instability. Attached Figure Description

[0013] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and are used to explain the invention. They do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram of the modules of the industrial solid waste landfill leachate treatment fault prediction and health management system in the embodiments of this application. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0015] The industrial solid waste landfill leachate treatment fault prediction and health management system includes: fluid processing equipment, reagent dosing equipment, process monitoring module, dosing control module, edge computing gateway, and prediction and health management center; the fluid processing equipment includes at least an anaerobic zone and an aerobic zone; the reagent dosing equipment stores regulating agents; The process monitoring module is connected to the fluid processing equipment to collect process status data of the fluid processing equipment and control the fluid operating parameters; The dosing control module is connected to the pesticide dosing equipment to collect pesticide dosing data and control the execution of pesticide dosing tasks. The prediction and health management center communicates with the process monitoring module and the dosing control module through an edge computing gateway; The edge computing gateway is used to perform outlier removal and local caching on process status data and drug dosing data before sending them to the prediction and health management center. The prediction and health management center is configured as follows: based on process status data and drug dosing data, and combined with the built-in kinetic mechanism model, the core health index is determined; and control instructions are generated based on the core health index and issued to the process monitoring module and the dosing control module respectively; among them, the control instructions include at least fluid parameter adjustment instructions and dynamic drug optimization instructions. This embodiment provides a mechanism for fault prediction and health management of leachate treatment in industrial solid waste landfills, such as... Figure 1 As shown; in specific implementation, the system is deployed in a leachate biochemical treatment device with an anaerobic-aerobic zone series arrangement, and the treatment object is leachate from industrial solid waste landfills with significant fluctuations in composition; In the main scenario set in this embodiment, the landfill enters a high-load period after rain on the 186th day of continuous operation. The influent ammonia nitrogen, biodegradable organic matter and accompanying inhibitory components all fluctuate. The system needs to maintain deep denitrification and avoid microbial instability without stopping operation. Furthermore, the process monitoring module is connected to the on-site fluid handling equipment, preferably by collecting online detection data of the anaerobic and aerobic zones through a programmable logic controller, and issuing adjustment commands to actuators such as reflux pumps, blowers, valves, and booster pumps; The dosing control module communicates with the hydroxylamine dosing equipment to read data such as the liquid level of the reagent storage tank, the frequency of the metering pump, the instantaneous dosing flow rate, and the cumulative dosing amount, and to execute start-up, shutdown, frequency adjustment, and interlocking control. The edge computing gateway is set up in the central control cabinet of the plant. On the one hand, it receives fieldbus data, and on the other hand, it performs data timestamp unification, outlier removal, breakpoint resume transmission, and local caching before sending it to the prediction and health management center. The predictive and health management center includes a pre-built kinetic mechanism model. This model is not limited to a single form of expression; its core lies in estimating the immediate fate of hydroxylamine in the sludge system based on real-time process state data and reagent dosage data entering the system. The kinetic mechanism model constructs a difference equation that includes reagent dosage, sludge concentration, and reagent consumption rate. This difference equation is specifically expressed as follows: in, for The increase in drug accumulation that is not consumed within the time step, expressed in mgN / L; The actual dosage of the agent at the current time step is expressed in mgN. The effective volume of the reaction zone is expressed in liters (L). The specific rate constant for reagent consumption per unit mass of sludge is expressed in mgN / (mgMLSS·h). Real-time sludge concentration, in mg / L; To control the cycle time step, the unit is h; the kinetic model outputs the increment of unconsumed drug accumulation through this equation; The central nervous system can generate a continuous time-series accumulation concentration curve, rather than relying solely on the dosage value at a single point in time for static judgment; after obtaining the drug accumulation concentration, the central nervous system further combines dissolved oxygen and hydraulic retention time to calculate the core health index; To illustrate the calculation logic, the core health index can be abstracted into a normalized score between 0 and 1. For example, when the stock concentration is 0.2, the dissolved oxygen in the aerobic zone is 0.5 mg / L, and the hydraulic retention time is 2.8 h, the central nervous system can calculate a health evaluation vector containing different bacterial community states or response potential components. in, Represents the activity index of ammonia-oxidizing bacteria. Represents the inhibition index of nitrite-oxidizing bacteria. The three components represent the denitrification potential index of polysaccharide bacteria. They are all dimensionless normalized scores between 0 and 1. For example, when the parameters for this period are input, the vector H can be specifically calculated as [0.82, 0.76, 0.71]. When the accumulated concentration rises to 0.7 in subsequent periods and the dissolved oxygen drops to 0.2 mg / L for a short time, the vector H may become [0.48, 0.39, 0.66], indicating that the system has changed from the short-range nitrification stable zone to the microbial community pressure zone; the core health index is calculated by normalizing the accumulated concentration of the reagent, the dissolved oxygen concentration and the hydraulic retention time through a preset weighted matrix; When generating control commands, the central system does not rely solely on a single indicator, but rather processes process status data, drug dosing data, and core health indices together. It can employ a combination of rule engines and optimization solutions: if the overall health index remains within the first preset range, the current operating state is maintained; if the indicator reflecting suppressed microbial communities is below the second preset threshold, fluid parameter adjustment commands are generated first. If the indicators show a decrease in drug utilization efficiency but the microbial community has not been significantly damaged, a dynamic drug optimization instruction will be generated first; if both types of anomalies occur at the same time, both types of instructions will be issued simultaneously and distributed to the process monitoring module and the dosing control module through the edge computing gateway to form a linkage closed loop. As a safety redundancy measure, when any online sensor has no new data within a preset time period, jumps beyond the physical possible range, or forms a logical conflict with adjacent sensors, the edge computing gateway first marks the channel with a reduced confidence level. Before the channel recovers, the central hub uses the smoothed value, redundant measurement point value, or mechanism estimate value from the most recent period to participate in the calculation. If the feedback from the dosing equipment is interrupted, the central system will stop executing the optimization amplification strategy and only retain the conservative dosage or directly enter the manual confirmation mode. If neither the process monitoring module nor the dosing control module can receive downlink commands, the system will switch to the pre-set safe operation curve to maintain the anaerobic and aerobic zones in a low-to-medium load stable state to avoid secondary instability caused by communication failure. In the above-mentioned 186th day of operation, the influent ammonia nitrogen fluctuated from 980 mg / L to 1120 mg / L, accompanied by drainage disturbances from the upstream industrial workshop; the data uploaded by the process monitoring module showed that although the dissolved oxygen in the aerobic zone remained at 0.5 mg / L, the reagent consumption capacity corresponding to the anaerobic zone return sludge had been declining continuously. Based on this, the central system determined that if the fixed hydroxylamine dosage continued to be used, the accumulated concentration would cross the toxicity risk line in the next few hours. Therefore, it first issued a dynamic agent optimization instruction to reduce hydroxylamine to below the upper boundary of the maintenance range, and at the same time issued an instruction to adjust fluid parameters to extend the anaerobic residence time and maintain the micro-aerobic state in the aerobic zone. As a result, the system completed feedforward intervention in the early stage of microecological damage before the total nitrogen in the effluent deteriorated. The purpose of this step is to transform the control of lagging apparent indicators in the leachate treatment process into predictive control based on microbial health, thereby achieving synergistic management of agents, microbial community status and hydraulic operation, and reducing the risk of short-range nitrification instability caused by hydroxylamine misdosing or sudden load changes.

[0016] Furthermore, the process status data should include at least the hydraulic retention time, oxidation-reduction potential, pH, dissolved oxygen, ambient temperature, sludge concentration, and actual carbon-to-nitrogen ratio. The prediction and health management center is also used to determine the amount of simultaneous nitrification and denitrification nitrogen removal in the aerobic zone based on process status data.

[0017] This embodiment provides a mechanism for expanding the acquisition of process status data and estimating the amount of nitrogen removed by simultaneous nitrification and denitrification. Specifically, in the aforementioned continuous operation scenario, although dissolved oxygen, hydraulic retention time and reagent dosage can reflect part of the health status, when the fluctuation range of the industrial solid waste leachate composition exceeds the preset change rate threshold, there may still be a situation where the apparent dissolved oxygen is normal but the actual reaction path deviates. For example, when toxic organic matter enters the system for a short time, the oxidation-reduction potential and pH will shift before the total nitrogen in the effluent. If these process variables are missing at this time, the central nervous system will not be sensitive enough in judging the health status. Therefore, this embodiment further introduces hydraulic retention time, oxidation-reduction potential, pH, dissolved oxygen, ambient temperature, sludge concentration and actual carbon-nitrogen ratio as basic process status data. Specifically, the process monitoring module is equipped with oxidation-reduction potential probes and pH probes in the anaerobic and aerobic zones, respectively, and a dissolved oxygen probe in the aerobic zone. A sludge concentration acquisition unit is installed on the return sludge pipeline, and the ambient temperature is obtained from the weather chamber or equipment room temperature sensor. The actual carbon-nitrogen ratio can be obtained from online chemical oxygen demand, ammonia nitrogen, or total nitrogen data according to a preset conversion relationship. In some scenarios, the data can also be obtained by fitting and compensating laboratory test data with online trends. After the central processing unit performs time alignment on the above multidimensional data, a unified state vector is formed. For example, within a 5-minute sampling period, the state vector can be simplified to a vector containing components corresponding to hydraulic retention time, oxidation-reduction potential, pH, dissolved oxygen, ambient temperature, sludge concentration, and actual carbon-nitrogen ratio. The formula for calculating the vector of the corresponding components is as follows: in, This indicates the hydraulic residence time, expressed in hours (h). This represents the redox potential, expressed in mV. Indicates acidity or alkalinity, dimensionless. This indicates the dissolved oxygen concentration, expressed in mg / L. This indicates the ambient temperature, expressed in °C. This indicates the sludge concentration, expressed in mg / L. This represents the actual carbon-to-nitrogen ratio, which is dimensionless. For example, within a 5-minute sampling period, this state vector is calculated to give V1 = [2.8h, -145mV, 7.6, 0.5mg / L, 31℃, 4200mg / L, 3.2]. Based on this, the central processing unit also calculates the nitrogen removal capacity of simultaneous nitrification and denitrification in the aerobic zone. This capacity reflects the net nitrogen removal contribution of the aerobic zone under microaerobic conditions, and is an important reference for subsequent judgment on whether short-cut nitrification remains stable. The specific calculation logic is as follows: based on the difference in total inorganic nitrogen concentration between the entry and exit points of the aerobic zone, the apparent total inorganic nitrogen removal capacity is calculated. The specific calculation formula is as follows: in, The apparent total inorganic nitrogen removal is expressed in mg / L. The total inorganic nitrogen concentration entering the aerobic zone, expressed in mg / L; The total inorganic nitrogen concentration upon leaving the aerobic zone, expressed in mg / L; The proportion of anoxic / micro-oxic microenvironments within the aerobic zone is dynamically assessed based on online dissolved oxygen and redox potential parameters. The specific assessment formula is as follows: in, This is the spatial proportion coefficient of hypoxic / micro-oxygen microenvironments, dimensionless; This is the real-time dissolved oxygen concentration, in mg / L. is the dissolved oxygen half-saturation constant, in mg / L; This is the actual redox potential, in mV; This is the redox reference potential in a micro-oxygen environment, expressed in mV. and These are dimensionless weighting coefficients fitted based on historical data. The biomass synthesis and assimilation nitrogen requirement is determined using ambient temperature and sludge concentration, and then subtracted from the apparent total inorganic nitrogen removal. The formula for calculating the biomass synthesis and assimilation nitrogen requirement is as follows: in, Nitrogen requirement for biomass synthesis and assimilation, expressed in mg / L; This is a dimensionless metabolic correction factor that depends on ambient temperature. This refers to sludge concentration, expressed in mg / L. The average nitrogen content of microbial cells is dimensionless. Based on the spatial proportion coefficient of anoxic / micro-oxygen microenvironment and the actual carbon-nitrogen ratio, the remaining removal amount is allocated proportionally to extract the absolute value of net nitrogen removal truly contributed by the short-cut simultaneous nitrification-denitrification pathway. In abnormal situations, if the actual carbon-nitrogen ratio is missing for a certain period, the moving average of the most recent consecutive effective periods can be used as a substitute; if the redox potential probe is contaminated and the reading is delayed for a long time, its weight will be reduced, and the pH slope and dissolved oxygen fluctuations will be used as the primary reference to judge the reaction section status; if the aerobic zone is missing online data for both ammonia nitrogen and total nitrogen, the absolute value of nitrogen removal will not be output synchronously for that period, only the trend label of rising, remaining flat or falling will be output to prevent erroneous values ​​from driving the control system in reverse. In the early morning of day 187 of the main scenario, due to the switching of the temporary storage tank, the actual carbon-to-nitrogen ratio of the leachate increased from 3.2 to 4.7 at a rate greater than the preset gradient. At the same time, the oxidation-reduction potential in the anaerobic zone rose from -145mV to -110mV. If only dissolved oxygen is considered, the apparent monitoring value of the system is still within the normal range. However, based on the expanded state vector calculation, the central system found that the nitrogen removal capacity of simultaneous nitrification and denitrification in the aerobic zone had decreased from 12mg / L to 7mg / L, indicating that the microecological pathway has begun to deviate from the design conditions. At this time, the central system uses this result as a leading signal for subsequent fault prediction to intervene and adjust in advance. The purpose of this step is to achieve earlier and finer-grained state identification of the short-cut nitrification system by adding process state variables with precursor significance and calculating the nitrogen removal rate of simultaneous nitrification and denitrification in the aerobic zone.

[0018] Furthermore, the core health indices include the ammonia-oxidizing bacteria activity index, the nitrite-oxidizing bacteria inhibition index, and the polysaccharide bacteria denitrification potential index.

[0019] This embodiment provides a mechanism for the detailed definition of core health index; specifically, in the previous embodiment, although it is possible to assess the overall health of the system using multi-source process state data and drug accumulation concentration, the single conclusion of overall health still has a drawback in complex leachate scenarios: different microbial communities may respond to operating condition disturbances in opposite directions; for example, at the same time, ammonia-oxidizing bacteria may have been slightly inhibited by hydroxylamine accumulation, while nitrite-oxidizing bacteria have not yet recovered; If only a total score is given, the control system cannot distinguish whether to reduce the dosage of the agent to protect ammonia-oxidizing bacteria or to maintain the inhibited state to prevent the rebound of nitrite-oxidizing bacteria; therefore, this embodiment breaks down the core health index into three parts: ammonia-oxidizing bacteria activity index, nitrite-oxidizing bacteria inhibition index, and polysaccharide bacteria denitrification potential index. Specifically, the central processing unit calculates three indices based on the accumulated concentration of the reagent, dissolved oxygen, hydraulic retention time, oxidation-reduction potential, pH, sludge concentration, and actual carbon-to-nitrogen ratio. The specific structured decomposition logic is as follows: For the ammonia-oxidizing bacteria activity index, dissolved oxygen and hydraulic retention time are used as positive basal metabolic support dependent variables. The accumulated concentration of the reagent is converted into a nonlinear inhibitory factor, and the basal metabolism is reduced and evaluated. That is, the activity index is obtained by multiplying the basal metabolic support dependent variables by the inverse proportional function of the nonlinear inhibitory factor. For the nitrite-oxidizing bacteria inhibition index, the ratio of the accumulated concentration of the agent to the sludge concentration is taken as the main driving force of chemical inhibition, and the deviation of the extreme value of pH fluctuation is extracted as an auxiliary variable of environmental stress. The two are weighted and fused to obtain the index. For the polysaccharide bacteria denitrification potential index, the redox potential time-series slope characteristics and the actual carbon-nitrogen ratio are mainly combined to measure the efficient carbon source retention capacity and the sufficiency of denitrification electron donors in the system. The three indices can also form a composite criterion; for example, when the ammonia oxidizing bacteria activity index is lower than the first threshold and the nitrite oxidizing bacteria inhibition index is higher than the second threshold, it indicates that the inhibition target is still present, but the negative impact on the main reacting bacteria is significant; when the ammonia oxidizing bacteria activity index is higher than the first threshold and the nitrite oxidizing bacteria inhibition index is lower than the second threshold, it indicates that the main reacting bacteria are normal, but the short-cut nitrification advantage is being lost; when the polysaccharide bacteria denitrification potential index continues to decline, it indicates that the carbon flow distribution capacity inside the system has deteriorated, and even if short-cut nitrification is maintained in the future, the total nitrogen removal may decrease due to the lack of synergistic denitrification. In abnormal situations, if one of the three indices cannot be calculated stably due to the lack of key input, the central system can mark it as an estimated state and reduce the weight of that index in the control decision. If all three show high fluctuations at the same time but the online instruments are diagnosed as normal, the central system will regard it as a real working condition switch rather than simple noise and shorten the decision cycle to improve the response speed. On the morning of day 187, the system monitored that the activity index of ammonia oxidizing bacteria dropped from 82 to 61, the inhibition index of nitrite oxidizing bacteria remained at 74, and the denitrification potential index of polysaccharide bacteria was 67. Based on this, the central system judged that the main problem at present was not that nitrite oxidizing bacteria had recovered, but that the mismatch between hydroxylamine utilization and sludge activity led to the inhibition of ammonia oxidizing bacteria. Therefore, the system did not directly increase hydroxylamine, but instead switched to a health management mode that prioritized the protection of ammonia oxidizing bacteria. The purpose of this mechanism is to refine the microecological status from a single comprehensive evaluation into three directional indicators that are distinguishable, interpretable, and controllable, thereby achieving more precise process diagnosis and control strategy allocation.

[0020] Furthermore, the prediction and health management center is also used to monitor the trend changes of the nitrite-oxidizing bacteria inhibition index; If the nitrite-oxidizing bacteria inhibition index shows a continuous downward trend within a preset time window containing multiple consecutive sampling cycles, and the simultaneous nitrification and denitrification nitrogen removal in the aerobic zone is lower than the preset nitrogen removal baseline, then a short-range nitrification destruction warning and fluid parameter adjustment command will be generated. Fluid parameter adjustment commands include residence time extension commands and dissolved oxygen fine-tuning commands; The process monitoring module responds to the residence time extension command by extending the hydraulic residence time in the anaerobic zone; and responds to the dissolved oxygen fine-tuning command by adjusting the dissolved oxygen concentration in the aerobic zone.

[0021] This embodiment provides a mechanism for short-range nitrification damage early warning and fluid parameter linkage adjustment; specifically, based on the aforementioned three-index evaluation, if only comparing whether the nitrite oxidizing bacteria inhibition index is lower than the threshold at a single time point, it may still be affected by short-term noise, instrument drift or transient fluctuations in water quality, leading to false alarms or missed alarms. Especially in industrial leachate treatment, a decrease at a single point in time does not necessarily mean that short-cut nitrification has been destroyed. What is truly dangerous is the combination of a continuous decrease and a deterioration in the actual nitrogen removal contribution. Therefore, this embodiment uses a method of judging the trend of time window and judging the nitrogen removal baseline to trigger an early warning. Specifically, the central processing unit continuously reads the nitrite-oxidizing bacteria inhibition index sequence within a preset time window. Assuming the time window is the most recent 6 sampling periods, if the sequence is [81, 79, 77, 75, 73, 71], it can be determined as a continuous decline; if it is [81, 79, 80, 76, 74, 73], since there is a rebound point in the middle, it is not treated as a strictly continuous decline, but can be recorded as a suspected decline. At the same time, the central processing unit compares the amount of nitrogen removed by simultaneous nitrification and denitrification within the same window with the preset nitrogen removal baseline; for example, if the baseline is set to 11.85 mg / L, and the average value of the current window is 8.2 mg / L, then the condition of being below the baseline is met; only when both of these conditions are met will the central processing unit generate a short-range nitrification destruction warning. After generating an early warning, the central control system further generates fluid parameter adjustment instructions. These adjustments are not simply about increasing or decreasing speed, but include at least instructions to extend residence time and fine-tune dissolved oxygen. The former mainly affects the anaerobic zone by changing the inflow and outflow rates or the return distribution, allowing hydroxylamine and the returned sludge to have more sufficient contact and buffer time. The latter mainly affects the aerobic zone by adjusting the blower frequency or valve opening according to a preset adjustment step size, so that dissolved oxygen enters a micro-aerobic zone that is more suitable for maintaining short-range nitrification without excessively stimulating the recovery of nitrite-oxidizing bacteria. In one specific embodiment, assuming that the nitrite-oxidizing bacteria inhibition index drops from 81 to 71 for six consecutive cycles during a certain period, and the simultaneous nitrification-denitrification nitrogen removal rate drops from 12.4 mg / L to 8.2 mg / L, the central control system triggers an early warning; two instructions are generated: first, extend the hydraulic retention time in the anaerobic zone from 2.6 h to 3.0 h; second, fine-tune the dissolved oxygen in the aerobic zone from 0.50 mg / L to 0.42 mg / L; after receiving the instructions, the process monitoring module controls the front-end pump group to reduce the instantaneous processing flow rate by a first preset step and simultaneously reduce the blower volume; if the nitrite-oxidizing bacteria inhibition index stops falling and stabilizes between 72 and 74 in the next window, and the simultaneous nitrification-denitrification nitrogen removal rate rises to above 10.8 mg / L, it indicates that the fluid parameter adjustment has achieved a repair effect; In abnormal circumstances, if the nitrite oxidizing bacteria inhibition index continues to decline but the simultaneous nitrification and denitrification nitrogen removal is not lower than the baseline, the central nervous system will not directly issue a damage warning, but will first mark it as a potential risk increase and shorten the next assessment cycle. If the denitrification rate is lower than the baseline but the nitrite-oxidizing bacteria inhibition index does not decrease, it indicates that the problem may be due to the suppression of ammonia-oxidizing bacteria, a mutation in the carbon source structure, or a sensor malfunction. The central nervous system will then call other diagnostic branches instead of misdiagnosing it as nitrite-oxidizing bacteria resurgence. If the residence time has reached the upper limit allowed by the equipment and cannot be extended further, the central nervous system will only perform a fine adjustment of dissolved oxygen and give a capacity-limited warning. If the dissolved oxygen is already below the safe lower limit, it will not be adjusted further to avoid the aerobic zone from turning into a hypoxic runaway state. On the afternoon of the 187th day of the main scenario, the nitrite-oxidizing bacteria inhibition index was 78, 76, 74, 72, 70, and 68 for six consecutive cycles, and the average nitrogen removal rate of simultaneous nitrification and denitrification in the aerobic zone dropped to 7.9 mg / L. Based on this, the central control issued a short-range nitrification damage warning and issued a linkage command to increase the hydraulic retention time in the anaerobic zone by 0.4 h and decrease the dissolved oxygen in the aerobic zone by 0.06 mg / L. After 4 hours of execution, the downward trend of the nitrite-oxidizing bacteria inhibition index was stopped, and it subsequently stabilized at around 70, indicating that the system had transitioned from an unstable and deteriorating state to a controlled recovery. The purpose of this step is to reduce misjudgments through dual verification of trends and baselines, and to achieve proactive intervention in the risk of short-range nitrification instability by adjusting fluid parameters in small increments.

[0022] Furthermore, the prediction and health management center dynamically generates dynamic drug optimization instructions within the preset dosage concentration range based on core health indices and pre-defined dosage concentration ranges. The dosing control module responds to dynamic reagent optimization commands, adjusts the reagent dosage of the reagent dosing equipment, and regulates the reagents, including hydroxylamine.

[0023] This embodiment provides a dynamic optimization dosing mechanism for hydroxylamine based on a core health index. Specifically, in the aforementioned early warning mechanism, fluid parameter adjustments can provide recovery space for the bacterial community. However, if the dosage remains fixed, two types of defects will still occur under certain operating conditions: one is that the dosage is too high, leading to hydroxylamine accumulation and continuous suppression of ammonia-oxidizing bacteria; the other is that the dosage is too low, resulting in insufficient inhibition of nitrite-oxidizing bacteria and loss of short-range nitrification advantage. Therefore, this embodiment further optimizes the hydroxylamine dosage within a preset concentration range, rather than using a single fixed value. Specifically, the preset dosage concentration range can be set to 0 to 6 mgN / L based on project design experience. The central control unit reads three core health indices in each control cycle and outputs a candidate dosage value. Specifically, the candidate dosage value output logic is as follows: the central control unit has a built-in multi-dimensional lookup table rule and step size regulator. When the ammonia oxidizing bacteria activity index is lower than the set safe activity threshold, the priority protection rule is triggered, and the step size regulator outputs a negative adjustment step size to reduce the candidate dosage value. When the ammonia oxidizing bacteria activity index is normal, but the nitrite oxidizing bacteria inhibition index is lower than the inhibition baseline, a positive adjustment step size is output to increase the candidate dosage value; when multiple indices conflict in their corresponding adjustment directions, the ammonia oxidizing bacteria activity index is strictly used as the highest priority, that is, the upper limit of dosage is locked and a negative adjustment is forced or a steady state is maintained to avoid irreversible system collapse. For ease of understanding, the optimization process can be simplified into a discrete process of trial-evaluation-correction; assuming that the current ammonia oxidizing bacteria activity index is 58, the nitrite oxidizing bacteria inhibition index is 73, and the polysaccharide bacteria denitrification potential index is 66, and the system identifies a state where ammonia oxidizing bacteria are prioritized, then the candidate dosage value is reduced from the current 2.0 mg N / L to 1.2 mg N / L; If the ammonia oxidizing bacteria activity index rises to 65 in the next cycle, while the nitrite oxidizing bacteria inhibition index remains above 70, then the dosage is finely adjusted to 1.0 mg N / L. If the nitrite oxidizing bacteria inhibition index suddenly drops below 60 after further adjustment, it indicates insufficient inhibition, and the central control will adjust the dosage back to 1.4 mg N / L. Through this small-step adjustment, the system can gradually approach the balance point that balances the protection of ammonia oxidizing bacteria and the inhibition of nitrite oxidizing bacteria. To avoid frequent oscillations, the central control system can also set an optimization dead zone and an upper limit for change. For example, if the difference between the two calculated optimal doses is less than 0.1 mgN / L, the current setting remains unchanged; if the difference is greater than 1.0 mgN / L, it will not jump to the target value all at once, but will adjust in stages in steps of 0.4 mgN / L to reduce the transient impact on the sludge system. After receiving the instruction, the dosing control module converts it into pump frequency or valve opening according to the metering pump calibration curve, and transmits the actual execution value back to the central control system for verification in real time. As an anomaly handling mechanism, if the current liquid level in the drug storage tank is lower than the minimum operating level, the central control will stop further increasing the dosage, retain only the minimum conservative dose, and issue a replenishment prompt; if the deviation between the metering pump's executed value and the target value continues to exceed the preset tolerance, for example, if the target is 1.4 mgN / L but the actual value is only 0.8 mgN / L, the central control will mark the device as having limited execution, and subsequent optimization will no longer be based on the ideal value, but will be remodeled based on the actual feedback value; if all three core health indices show high uncertainty at the same time, the central control will freeze the automatic optimization, switch to manual confirmation, or only maintain a low-risk steady-state dose. In the early morning of day 188 of the main scenario, after the previous stage of hydraulic retention time and dissolved oxygen fine-tuning, the ammonia-oxidizing bacteria activity index was still low. The central control gradually reduced the hydroxylamine dosage from the original fixed 2.0 mgN / L to 1.2 mgN / L, and then to 1.0 mgN / L. After several cycles, the ammonia-oxidizing bacteria activity index rebounded from 58 to 69, and the nitrite-oxidizing bacteria inhibition index remained above 71, indicating that the dosage of the agent had returned from the range that could cause cumulative toxicity to the range that maintained both inhibition and activity. If the influent disturbance intensifies again at this time, the central control can readjust the dosage within the range of 0 to 6 mgN / L without the need for long-term manual trial and error. The purpose of this mechanism is to make the hydroxylamine dosage adapt to the health status of the microbial community, thereby achieving a dynamic balance between drug consumption, microbial community stability and denitrification efficiency.

[0024] Furthermore, the process monitoring module also obtains metagenomic feature values ​​from an external server through a data interface; the metagenomic feature values ​​include the abundance of ammonia assimilation genes, the abundance of ammonia oxidation genes, and the proportion of bacterial communities. The prediction and health management center binds metagenomic feature values ​​with redox potential and pH over time to construct a gene epigenetic mapping dictionary with a time dimension. The prediction and health management center predicts the current abundance of ammonia assimilation genes based on a gene epigenetic mapping dictionary and performs abnormal diagnosis by combining the actual carbon-nitrogen ratio: If the current abundance of ammonia assimilation genes is higher than the preset abundance threshold based on historical sequencing data, it is determined to be an enhanced ammonia assimilation state; if the current abundance of ammonia assimilation genes is lower than or equal to the preset abundance threshold, it is determined to be a precursor state of sludge disintegration.

[0025] This embodiment provides an abnormality diagnosis mechanism that integrates metagenomic features and epigenetic parameters. Specifically, the aforementioned method mainly relies on online process data and mechanistic models, which can effectively determine short-cut nitrification stability. However, under certain late-stage operating conditions, there is still a challenge: when abnormal ammonia nitrogen loss occurs in the anaerobic zone, it is difficult to distinguish whether this is a beneficial enhancement of ammonia assimilation or a harmful precursor to sludge disintegration based solely on online instruments. Both may manifest as a decrease in ammonia nitrogen, but the subsequent treatment directions are completely different. It should be further explained that in the early stage of sludge disintegration, the disintegration of bacterial flocs and the loosening of cell structure usually expose the extracellular polymers and free adsorption sites that were originally encapsulated inside. As a result, the physical adsorption of ammonia nitrogen in the aqueous phase reaches the preset adsorption threshold or is non-specifically intercepted within the first preset time period. This leads to an abnormal decrease in the concentration of ammonia nitrogen in the aqueous phase on a macroscopic scale, i.e., an apparent abnormal loss of ammonia nitrogen. However, this is fundamentally different from the true ammonia assimilation controlled by the microbial biosynthetic metabolic pathway. To solve this problem, this embodiment introduces metagenomic feature values ​​from an external server into the online diagnostic link. Specifically, the process monitoring module periodically obtains metagenomic feature values ​​from an external server through a data interface. These feature values ​​can come from laboratory sequencing platforms, third-party testing platforms, or internal enterprise analysis servers, and at least include the abundance of ammonia assimilation genes, the abundance of ammonia oxidation genes, and the proportion of bacterial communities. Since the update frequency of metagenomic data is usually lower than that of online instruments, the central system needs to bind it to high-frequency epigenetic data such as redox potential and pH in a time series manner, thereby constructing a gene epigenetic mapping dictionary that includes the time dimension. For ease of explanation, it is assumed that three offline metagenomic results were obtained on days 180, 184, and 188, with corresponding ammonia assimilation gene abundances of 0.42, 0.57, and 0.73, respectively; the combined features of redox potential and pH components in the vicinity of the same time period can be represented as A1 = [-150mV, 7.4], A2 = [-132mV, 7.7], and A3 = [-118mV, 7.9], respectively. The central processing unit can establish a mapping dictionary: when the phenotypic feature is close to A1, the abundance of ammonia assimilation genes tends to be close to 0.42; when close to A2, it tends to be close to 0.57; when close to A3, it tends to be close to 0.73; if the current real-time phenotypic feature is [-122mV, 7.8], then it is located between A2 and A3, and the central processing unit can predict the current calculated abundance of ammonia assimilation genes to be 0.66; here, "closeness" can be achieved through distance comparison, interval matching, or trend correlation, and is not limited to a certain algorithm; After obtaining the current predicted abundance value of ammonia assimilation genes, the central nervous system performs abnormal diagnosis in combination with the actual carbon-nitrogen ratio. For example, if the preset abundance threshold is 0.60, the current predicted value is 0.66, and the actual carbon-nitrogen ratio is maintained within the range suitable for ammonia assimilation, it is determined to be an enhanced ammonia assimilation state, indicating that the decrease in ammonia nitrogen in the anaerobic zone is more likely to be the result of the metabolic redistribution of the microbial community. Conversely, if the current predicted value is only 0.48, which is not higher than the threshold, but the anaerobic zone continues to show abnormal ammonia nitrogen loss, combined with the fact that the microorganisms have not started large-scale anabolic metabolism at this time, the central nervous system is more likely to determine that the ammonia nitrogen drop at this time is due to the temporary physical adsorption and intracellular material imbalance caused by the initial damage to the sludge cell structure and the floc breakage, that is, the precursor state of sludge disintegration. In one specific embodiment, it is assumed that at a certain time on the 200th day, the ammonia nitrogen concentration in the influent to the anaerobic zone is 100 units, and the concentration in the effluent to the anaerobic zone is 90 units, resulting in an abnormal loss of 10 units. The central reading of the real-time oxidation-reduction potential and pH is [-120mV, 7.85], and the ammonia assimilation gene abundance is predicted to be 0.67 according to the mapping dictionary. Meanwhile, the actual carbon-nitrogen ratio is 4.5; since 0.67 is higher than the threshold of 0.60, the central nervous system judges this phenomenon as an enhanced ammonia assimilation state, rather than immediately identifying the system as unstable; if another period also shows a loss of 10 concentration units, but the predicted abundance corresponding to the real-time feature is only 0.46, and the sludge concentration shows a synchronous downward trend, then the central nervous system excludes the possibility of metabolic consumption, and instead judges it as a sludge disintegration precursor state dominated by floc breakage and apparent adsorption illusion, and activates the protection strategy accordingly. As an anomaly handling mechanism, if the external server does not return new metagenomic feature values ​​within a preset period, the central processing unit does not stop diagnosis but continues to use the most recent valid dictionary and lowers the diagnostic confidence. If the deviation between the current phenotypic feature and any known region in the dictionary exceeds the preset feature matching tolerance threshold, such as when redox potential and pH both exceed the limits, the central processing unit can treat it as an unknown phenotype and will not directly give a conclusion of enhanced ammonia assimilation or sludge disintegration. Instead, it will suggest that additional sequencing or manual verification is needed. If the predicted abundance falls within the fuzzy range near the threshold, such as 0.58 to 0.62, the central processing unit can combine auxiliary factors such as the sludge concentration change rate and the ammonia-oxidizing bacteria activity index change rate for secondary discrimination to reduce misclassification. On day 200 of the main scenario, the system experienced an abnormal loss of approximately 10 mg / L of ammonia nitrogen in the anaerobic zone. Due to significant changes in the composition of the leachate in the preceding days, on-site personnel could not determine the cause based solely on online water quality data. The central processing unit (CPU) invoked the previously accumulated gene epigenetic mapping dictionary, predicting that the current ammonia assimilation gene abundance was 0.68 and the actual carbon-to-nitrogen ratio was 4.6. Therefore, it was determined to be an enhanced ammonia assimilation state, and the system only performed a slight operating condition correction without triggering a shutdown for maintenance. A few days later, if the predicted abundance corresponding to another similar phenomenon dropped to 0.49, the CPU would identify the aforementioned physical adsorption artifact through gene phenotype mismatch, reclassify it as a precursor to sludge disintegration, and initiate the protection and diagnostic process ahead of schedule. The purpose of this mechanism is to couple low-frequency but high-information metagenomic features with high-frequency online epigenetic signals to achieve interpretable diagnosis of the causes of abnormal ammonia nitrogen loss in anaerobic zones and to accurately distinguish between metabolic fate and disintegration artifacts.

[0026] Furthermore, the process monitoring module also includes a communication subsystem that communicates with a cloud server; The prediction and health management center generates a strategy update instruction and sends it to the communication subsystem. After receiving the instruction, the communication subsystem transmits the relevant data to the cloud server and receives feedback data. Based on the feedback data, the prediction and health management center optimizes the generation strategy of the control instruction.

[0027] This embodiment provides a cloud-edge collaborative strategy update mechanism. Specifically, the aforementioned local prediction and health management center can already achieve closed-loop control on site, but during the long-term operation of the landfill, it will encounter a limitation: the local model is better at handling the known operating conditions of the current site, and the response delay to slow variables such as seasonal changes, replacement of upstream industrial components, and long-term microbial succession exceeds the preset time tolerance. If the system relies entirely on the fixed on-site strategy, it may react slowly under new disturbance modes. Therefore, this embodiment further introduces a communication subsystem to work in collaboration with the cloud server to achieve continuous updates to the control strategy. Specifically, the communication subsystem can be integrated into the process monitoring module to upload process status data and reagent dosing data to the cloud server according to the strategy update instructions issued by the central system. The cloud server can aggregate data from different time periods at this site, or further aggregate data from multiple similar leachate projects, and perform offline analysis on the operating samples to generate strategy feedback that is more suitable for the current stage. This feedback can be a threshold correction value, an optimization step size correction value, an early warning window length correction value, or a new mapping dictionary fragment. In one specific embodiment, it is assumed that the original warning window for the nitrite oxidizing bacteria inhibition index of the local central hub is 6 cycles, and the hydroxylamine optimization step size is 0.4 mgN / L; after receiving the operating data of the last 30 days, the cloud server found that the real turning point of the nitrite oxidizing bacteria inhibition index often occurs in the 8th cycle under the rainy season conditions, and when the drug step size exceeds the set step size limit, it will cause the ammonia oxidizing bacteria to be under short-term pressure. Therefore, the cloud provides a set of strategy parameters: the warning window is updated to 8 cycles, the drug step size is updated to 0.25 mgN / L, and the ammonia-oxidizing bacteria protection threshold is increased by 5 points; after verifying that the feedback does not exceed the safety boundary, the local control center writes it into the new strategy template and makes it effective in subsequent control. The communication and update process preferably adopts the method of uploading summary + necessary original fragments, which reduces bandwidth consumption and avoids meaningless full transmission; for example, the local system first uploads the daily average, extreme values, abnormal window fragments and command execution results; when the cloud identifies a special disturbance in a certain period of time, it then requests the retransmission of the corresponding original minute-level data. After receiving feedback from the cloud, the communication subsystem first writes it into the pending activation area, and the central system performs consistency and security checks. If the feedback parameters violate the constraints of the field equipment, such as the recommended lower limit of dissolved oxygen being lower than the minimum controllable value of the blower system, or the recommended upper limit of hydraulic retention time exceeding the allowable value of the pool volume, the central system refuses to activate it directly and only extracts the executable part. In abnormal situations, if cloud communication is interrupted, the local central hub will continue to operate independently using the effective strategy and will not interrupt on-site control due to external network failures. If the data returned by the cloud is incomplete, fails verification, or is obviously mismatched with the local operating conditions, the central hub will mark it as invalid feedback and will not participate in the current cycle decision. If there are conflicts between multiple cloud feedbacks, such as one group suggesting increasing the drug step size and another group suggesting decreasing the step size, the central hub will choose the more conservative solution according to the principle of safety priority, or keep the existing local strategy unchanged. On the 205th day after the main scenario entered the rainy season, the communication subsystem uploaded the process status data, pesticide dosing data and early warning trigger records of the past two weeks to the cloud server according to the strategy update instruction. After comparing samples from multiple sites, the cloud found that the original 6-cycle early warning window of this site was too short under the co-occurrence of high salt and high toxicity, which could easily mistake short-term fluctuations for an unstable trend. The cloud returned new feedback data, suggesting that the trend judgment window be extended to 8 periods and the hydroxylamine dose correction step size be changed from 0.4 mgN / L to 0.25 mgN / L. After the local central system completed the boundary verification, it adopted the update, which reduced the frequency of subsequent control adjustments while maintaining the sensitivity to responding to real risks. The purpose of this mechanism is to continuously optimize local control strategies by leveraging cross-cycle and cross-scenario knowledge accumulated in the cloud, while ensuring the safe and continuous operation of the field system through boundary verification and network outage autonomy.

[0028] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A fault prediction and health management system for leachate treatment in industrial solid waste landfills, characterized in that, include: Fluid handling equipment, reagent dosing equipment, process monitoring module, dosing control module, edge computing gateway, and prediction and health management center; The fluid processing equipment includes at least an anaerobic zone and an aerobic zone; the reagent dosing equipment stores a regulating agent. The process monitoring module is communicatively connected to the fluid processing equipment and is used to collect process status data of the fluid processing equipment and control fluid operating parameters. The dosing control module is communicatively connected to the drug dosing equipment, and is used to collect the drug dosing data of the drug dosing equipment and control the execution of the drug dosing task; The prediction and health management center is communicatively connected to the process monitoring module and the dosing control module through the edge computing gateway; The edge computing gateway is used to perform outlier removal and local caching on the process status data and the drug dosing data before sending them to the prediction and health management center. The prediction and health management center is configured as follows: based on the process status data and the drug dosing data, a core health index is determined in conjunction with a built-in kinetic mechanism model; and control instructions are generated based on the core health index and issued to the process monitoring module and the dosing control module respectively; wherein, the control instructions include at least fluid parameter adjustment instructions and dynamic drug optimization instructions.

2. The industrial solid waste landfill leachate treatment fault prediction and health management system according to claim 1, characterized in that, The process status data includes at least hydraulic retention time, oxidation-reduction potential, pH, dissolved oxygen, ambient temperature, sludge concentration, and actual carbon-to-nitrogen ratio; The prediction and health management center is also used to determine the amount of simultaneous nitrification and denitrification nitrogen removal in the aerobic zone based on the process status data.

3. The industrial solid waste landfill leachate treatment fault prediction and health management system according to claim 2, characterized in that, The core health indices include the ammonia-oxidizing bacteria activity index, the nitrite-oxidizing bacteria inhibition index, and the polysaccharide bacteria denitrification potential index.

4. The industrial solid waste landfill leachate treatment fault prediction and health management system according to claim 3, characterized in that, The prediction and health management center is also used to monitor the trend changes of the nitrite-oxidizing bacteria inhibition index; If the nitrite-oxidizing bacteria inhibition index shows a continuous downward trend within a preset time window containing multiple consecutive sampling cycles, and the simultaneous nitrification and denitrification nitrogen removal in the aerobic zone is lower than the preset nitrogen removal baseline, then a short-range nitrification destruction warning and a fluid parameter adjustment command are generated.

5. The industrial solid waste landfill leachate treatment fault prediction and health management system according to claim 4, characterized in that, The fluid parameter adjustment commands include residence time extension commands and dissolved oxygen fine-tuning commands; The process monitoring module responds to the residence time extension command by extending the hydraulic residence time in the anaerobic zone; and responds to the dissolved oxygen fine-tuning command by adjusting the dissolved oxygen concentration in the aerobic zone.

6. The industrial solid waste landfill leachate treatment fault prediction and health management system according to claim 3, characterized in that, The prediction and health management center dynamically generates the dynamic drug optimization instruction within the preset dosage concentration range based on the core health index and the preset dosage concentration range. The dosing control module responds to the dynamic agent optimization command and adjusts the agent dosage of the agent dosing device, wherein the regulating agent includes hydroxylamine.

7. The industrial solid waste landfill leachate treatment prognostics and health management system, as claimed in claim 2, wherein, The process monitoring module also obtains metagenomic feature values ​​from an external server through a data interface; the metagenomic feature values ​​include the abundance of ammonia assimilation genes, the abundance of ammonia oxidation genes, and the proportion of bacterial communities; The prediction and health management center binds the metagenomic feature values ​​with the redox potential and pH value in a time series to construct a gene epigenetic mapping dictionary with a time dimension.

8. The industrial solid waste landfill leachate treatment failure prediction and health management system, as claimed in claim 7, wherein, The prediction and health management center predicts the current abundance of ammonia assimilation genes based on the gene epigenetic mapping dictionary, and performs abnormal diagnosis in conjunction with the actual carbon-nitrogen ratio: If the current abundance of ammonia assimilation genes is higher than a preset abundance threshold calibrated based on historical sequencing data, it is determined to be an enhanced ammonia assimilation state; if the current abundance of ammonia assimilation genes is lower than or equal to the preset abundance threshold, it is determined to be a precursor state of sludge disintegration.

9. The industrial solid waste landfill leachate treatment prognostics and health management system, as claimed in claim 1, wherein, The process monitoring module also includes a communication subsystem that communicates with a cloud server. The prediction and health management center generates a strategy update instruction and sends it to the communication subsystem. After receiving the instruction, the communication subsystem transmits the relevant data to the cloud server and receives feedback data. The prediction and health management center optimizes the generation strategy of the control instruction based on the feedback data.

Images