An energy-saving sewage treatment method and system based on ecological simulation

Abstract

The present application belongs to the technical field of sewage treatment, and provides an energy-saving sewage treatment method and system based on ecological simulation, which comprises: obtaining and based on the historical operation data of a target sewage treatment system, using a time series analysis method, the continuous operation time of the system is initially divided into multiple regulation and control periods with different ecological characteristics and management targets. By introducing a dynamically divided and self-adaptively adjusted regulation and control period, the ecological characteristics and management needs of different stages can be automatically identified based on historical data and real-time information, thereby realizing fine management in the time dimension, flexibly responding to seasonal changes, climate fluctuations and sudden pollution events, and enabling the system to have macro stability and micro adaptability, thereby improving the long-term operation stability and overall energy efficiency of the sewage treatment system under the premise of ensuring that the effluent water quality meets the standards, and providing an intelligent time management basis for realizing energy-saving, low-maintenance and ecological sewage treatment.

Description

Technical Field

[0001] This invention belongs to the field of wastewater treatment technology, specifically an energy-saving wastewater treatment method and system based on ecological simulation. Background Technology

[0002] Wastewater treatment based on ecological simulation refers to the technology of purifying wastewater through the synergistic effects of physical, chemical and biological processes by artificially constructing or enhancing simulated natural ecosystems (such as wetlands, ponds, soil infiltration systems, etc.) and utilizing composite ecological units composed of plants, microorganisms and substrates in the system. Its core characteristics are low energy consumption (mainly relying on solar energy and gravity flow), low maintenance costs, and the combination of landscape creation and ecological service functions, and it belongs to the category of green infrastructure.

[0003] However, a prominent problem in actual operation is that key ecological management operations within the system, such as decisions on aquatic plant harvesting and sediment cleaning, often fall into an irreconcilable dilemma of conflicting multiple objectives. These objectives must be balanced simultaneously: pollutant removal efficiency, system carbon sequestration function, and ecosystem structural stability. Existing technologies typically rely on fixed calendars or empirical judgments, adopting a one-size-fits-all approach. This leads to several problems: premature and excessive harvesting and dredging in pursuit of immediate water quality, while temporarily removing pollutants, severely damages plant roots and benthic microbial communities, interrupts carbon sequestration, and causes ecosystem degradation and biodiversity decline due to frequent disturbances. In the long run, this reduces the system's sustainable treatment capacity and overall environmental benefits. Conversely, delaying necessary interventions to protect the ecosystem or fix carbon can easily lead to secondary release of pollutants from decaying matter, anaerobic deterioration of sediment, system blockage, and treatment function failure. This contradiction keeps the system in a state of long-term imbalance, where treatment efficiency, carbon sequestration gain, and ecological health are mutually exclusive, hindering the efficient and stable performance of the system and impeding the quantification and realization of the value of ecological products such as carbon sequestration.

[0004] Therefore, the present invention provides an energy-saving wastewater treatment method and system based on ecological simulation. Summary of the Invention

[0005] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0006] The technical solution adopted by this invention to solve its technical problem is: an energy-saving wastewater treatment method based on ecological simulation, comprising:

[0007] Based on the historical operation data of the target wastewater treatment system, the continuous operation time of the system is initially divided into multiple control cycles with different ecological characteristics and management objectives using time series analysis.

[0008] Based on the initial division of the control cycle, during the system operation, the division of the control cycle is dynamically evaluated and adjusted based on real-time updated operating data.

[0009] During the regulation cycle, real-time operational data is collected and analyzed. Based on the analysis results and the management objectives of the corresponding management cycle, the priority order of multi-objective decisions is determined, with the constraint being the minimization of multi-objective disturbances. Among these multi-objectives are pollutant removal efficiency, system carbon sink function, and ecosystem.

[0010] Based on the multi-objective decision priority order, a multi-objective optimization algorithm is used to solve the problem, generate and output ecological management operation instructions for the current dynamic management stage, which include at least plant harvesting and sediment cleaning strategies.

[0011] An energy-saving wastewater treatment system based on ecological simulation, the system comprising:

[0012] Initial control cycle design module: Based on the historical operation data of the target wastewater treatment system, the continuous operation time of the system is initially divided into multiple control cycles with different ecological characteristics and management objectives using time series analysis.

[0013] Control cycle update module: Based on the initially defined control cycle, the module dynamically evaluates and adjusts the control cycle division based on real-time updated operating data during system operation.

[0014] Multi-objective sequential output module: During the control period, it collects and analyzes real-time operating data within the control period. Based on the analysis results and the management objectives of the corresponding management period, it determines the priority order of multi-objective decisions, with the constraint being the minimization of multi-objective disturbances. Among these, the multi-objectives are pollutant removal efficiency, system carbon sink function, and ecosystem.

[0015] Management optimization and control module: Based on the priority order of multi-objective decisions, it uses a multi-objective optimization algorithm to solve the problem, generate and output ecological management operation instructions for the current dynamic management stage, which includes at least plant harvesting and sediment cleaning strategies.

[0016] The beneficial effects of this invention are as follows:

[0017] This invention can intelligently and dynamically adjust the focus of work at different times and under different ecological conditions. For example, it can prioritize carbon accumulation during the critical period of carbon sequestration, prioritize strengthening decontamination when water quality safety is threatened, and prioritize avoiding disturbance during the ecologically fragile period. In the long-term operation, the three major goals of pollutant removal efficiency, system carbon sequestration function and ecosystem health can be synergistically optimized rather than mutually hindered, thereby improving the overall environmental benefits and sustainability of the system.

[0018] This invention can sense deviations between its own state and the external environment in real time and automatically trigger the optimization of the control cycle framework. This makes the management strategy no longer static, but can evolve with changes in internal and external conditions. It has strong adaptability and resilience. Whether it is a gradual ecological succession or a sudden shock event, it can adjust the management rhythm in time, maintain the stability and efficiency of operation, and reduce the risk of functional failure due to sudden environmental changes.

[0019] This invention introduces a dynamic division and adaptive adjustment control cycle, which can automatically identify the ecological characteristics and management needs of different stages based on historical data and real-time information, thereby achieving refined management in the time dimension. It can flexibly respond to seasonal changes, climate fluctuations and sudden pollution events, giving the system macroscopic stability and microscopic adaptability. Under the premise of ensuring that the effluent water quality meets the standards, it improves the long-term operational stability and overall energy efficiency of the sewage treatment system, and provides an intelligent time management foundation for achieving energy-saving, low-maintenance ecological sewage treatment.

[0020] This invention constructs a multi-objective decision-making mechanism that incorporates three major objectives—pollutant removal efficiency, system carbon sequestration function, and minimization of ecosystem disturbance—into a unified optimization framework. Based on real-time data and contextualized priority ranking, it can dynamically coordinate conflicts between different objectives, avoiding system performance imbalances or ecological degradation caused by prioritizing a single objective. Combined with multi-objective optimization algorithms, it can generate optimal instructions for key management operations such as plant harvesting and sediment cleaning. This improves wastewater treatment efficiency while enhancing the system's carbon sequestration capacity and ecological resilience, achieving intelligent synergy among treatment efficiency, carbon sequestration gain, and ecological health. This promotes the evolution of wastewater treatment systems towards a more efficient, low-carbon, and eco-friendly sustainable direction. Attached Figure Description

[0021] The invention will now be further described with reference to the accompanying drawings.

[0022] Figure 1 This is a flowchart of the steps of an energy-saving wastewater treatment method based on ecological simulation according to the present invention;

[0023] Figure 2 This is a framework diagram of an energy-saving wastewater treatment system based on ecological simulation according to the present invention. Detailed Implementation

[0024] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0025] Example 1

[0026] One of the core inventive points of this invention is that, addressing the shortcomings of current technologies, it proposes an energy-saving wastewater treatment method based on ecological simulation. This method replaces fixed calendar management by dynamically dividing and adaptively adjusting the system's control cycle based on historical data and real-time information. Within each dynamically determined control cycle, it comprehensively analyzes management objectives that match the cycle's ecological characteristics, prioritizing three main objectives—pollutant removal efficiency, system carbon sink function, and minimization of ecosystem disturbance—in a contextualized manner. Finally, based on this priority order, it drives a multi-objective optimization algorithm to solve the problem, generating and outputting optimal instructions for key ecological management operations such as plant harvesting and sediment cleaning. This ensures that water quality meets standards while intelligently and collaboratively improving treatment efficiency, carbon sink gain, and ecological health, achieving long-term, stable, and efficient adaptive operation of the system.

[0027] Please see Figure 1 As shown in the embodiment of the present invention, an energy-saving wastewater treatment method based on ecological simulation includes the following steps:

[0028] Step 1: Based on the historical operation data of the target wastewater treatment system, use time series analysis to initially divide the continuous operation time of the system into multiple control cycles with different ecological characteristics and management objectives;

[0029] In step one, the historical operation dataset of the target wastewater treatment system is obtained. This historical operation dataset includes multi-dimensional time series data, and at least includes:

[0030] Water quality parameter sequence: chemical oxygen demand, total nitrogen, total phosphorus, and ammonia nitrogen concentrations of influent and effluent;

[0031] Climate-driving sequences: daily average temperature, precipitation, and sunshine duration;

[0032] Ecological status sequence: vegetation indices (such as NDVI), key plant phenological records, and sediment thickness monitoring values ​​obtained through remote sensing or periodic sampling;

[0033] Operation log sequence: Records of the time, intensity, and subsequent effects of historical plant harvesting, sediment cleaning, and other management operations;

[0034] The acquired historical dataset is preprocessed, including data cleaning, alignment, and noise reduction.

[0035] Extract time-series features characterizing the system's processing efficiency, carbon cycle status, and ecological stability, and construct a time-series feature set;

[0036] Examples of pollutant removal efficiency characteristics include, for instance, the average total phosphorus removal rate over the past 7 days.

[0037] Carbon sink potential characteristics: such as the weekly growth rate of plant biomass;

[0038] Ecological disturbance characteristics: such as the number of days since the last major management operation (such as harvesting);

[0039] A clustering algorithm suitable for multivariate time series is adopted to perform unsupervised learning on the time series feature set. The continuous running time axis is divided into K time periods, so that the change patterns of each feature within the same time period are similar. The corresponding K time periods are the initial control cycles. Each control cycle is automatically assigned a typical ecological feature vector defined by the cluster center and associated with the basic management target tendency based on historical statistics.

[0040] Optionally, a K-Means clustering algorithm based on dynamic time warping can be used, the specific process of which is as follows:

[0041] The preprocessed time series feature set is represented as a multivariate time series matrix, where each row of the matrix corresponds to a time point and each column corresponds to a feature dimension. The matrix dimension is T×D (T is the total number of time points and D is the number of feature dimensions).

[0042] Set the sliding window length to L (e.g., L=30 days, corresponding to approximately one month of ecosystem cycle) and the sliding step size to S (e.g., S=7 days, corresponding to one week), and use this sliding window to extract a set of subsequences from the multivariate time series matrix;

[0043] Each subsequence is an L×D matrix segment, representing D feature change patterns over L consecutive days. A total of N subsequences are extracted to form a subsequence set.

[0044] Set the number of control cycles K to be divided. K is determined by the elbow rule or the silhouette coefficient method: by trying different K values, calculate the total sum of squared deviations of the clustering results, and select the K value corresponding to the elbow inflection point or the K value with the largest silhouette coefficient as the final number of control cycles.

[0045] The k-means++ initialization algorithm is used to select initial cluster centers: First, a subsequence is randomly selected as the first cluster center; then, the minimum dynamic time regularization distance between each subsequence and the selected cluster centers is calculated, and the next cluster center is selected according to the probability distribution of the squared distance; this process is repeated until K initial cluster centers are selected.

[0046] Set the maximum number of iterations (e.g., 100) and the convergence threshold (e.g., 0.001).

[0047] Perform iterative calculations. For the t-th iteration (t starts from 0):

[0048] For each extracted subsequence, calculate its dynamic time-warped distance with each current cluster center. Since each subsequence contains multiple feature dimensions, the dynamic time-warped distance of each dimension needs to be calculated separately when calculating the distance. Then, the subsequence is weighted and summed according to the preset importance weight of each dimension to obtain the comprehensive distance metric between the subsequence and each cluster center. After completing all distance calculations, each subsequence is assigned to the category corresponding to the cluster center with the smallest comprehensive distance, thus forming K different subsequence sets.

[0049] For each subsequence set formed in the previous step, the corresponding cluster centers need to be updated. The update process uses the dynamic time warping centroid averaging algorithm: taking the current cluster center as the initial reference sequence, each subsequence in the set is dynamically time warped and aligned with this reference sequence to find the optimal correspondence path between the two. Based on the alignment results of all subsequences and the reference sequence, each feature value at each time point of the reference sequence is recalculated. The new feature value is the weighted average of the corresponding feature values ​​of all subsequences aligned to that point, thereby generating a new center sequence. The alignment and recalculation of the average value are repeated until the newly generated center sequence changes very little compared to the previous round, or reaches the preset maximum number of repetitions. At this point, the updated cluster centers are obtained.

[0050] After a complete allocation and update step is completed, the dynamic time-normalized distance between all K updated cluster centers and their corresponding unupdated cluster centers is calculated, and the average of these distances is obtained. If this average distance is less than a preset minimum threshold, or the number of iterations of the entire clustering process has reached a preset upper limit, then the clustering is considered to have converged, the iteration is stopped, and the final cluster centers and subsequence partitioning results are output. Otherwise, the newly updated cluster centers will be used to start the allocation, update, and judgment loop again from the beginning.

[0051] For each time point on the original timeline, identify all subsequences containing that time point, count the cluster labels of these subsequences, and use majority voting to determine the final cluster labels for each time point.

[0052] By merging consecutive time points with the same cluster label into one time period, the entire time axis is divided into K consecutive and non-overlapping time periods, which are the K control cycles obtained.

[0053] For each regulation cycle, the mean vector of the feature vectors at all time points within the regulation cycle is calculated, which is used as the typical ecological feature vector of the corresponding regulation cycle.

[0054] Analyze the historical operational log sequences corresponding to the regulation cycle, and combine them with typical ecological feature vectors to determine the basic management target tendencies:

[0055] If the pollutant removal efficiency characteristic value is higher than the historical average, and the historical management operations mainly focused on strengthening decontamination, then the basic objective of prioritizing efficient decontamination should be adopted.

[0056] If the carbon sequestration potential characteristic value is at a high level and shows an upward trend, and the historical management operation is to delay harvesting to accumulate biomass, then the associated carbon sequestration gain priority basic target.

[0057] If a large-scale management operation is performed immediately afterward, and the ecological disturbance characteristic value indicates that the system is in the recovery period, then the ecological restoration priority basic objective should be associated.

[0058] If multiple features are prominent, then the basic objective is to be balanced across multiple objectives.

[0059] Step 2: Based on the initially defined control period, the control period is dynamically evaluated and adjusted during system operation based on real-time updated operating data.

[0060] In step two, during system operation, the real-time running data stream corresponding to step one above is continuously collected, and the real-time feature set is calculated;

[0061] Set the sliding time window to M, and at each predetermined evaluation time point, calculate the similarity between the real-time feature set and the typical ecological feature vector of the current regulation cycle.

[0062] Optional similarity calculations can use cosine similarity or Euclidean distance formulas. Calculating similarity is used to quantify the consistency in change patterns and trends.

[0063] Based on the calculated similarity value, the degree of deviation of the real-time state from the current control cycle is evaluated. The degree of deviation is defined as the complement of the similarity. The higher the degree of deviation, the greater the difference between the actual behavior of the system and the expected characteristics of the control cycle.

[0064] Based on preset trigger conditions, it is determined whether to initiate a dynamic adjustment procedure for the control cycle. The trigger conditions include:

[0065] Timed triggering: When the system running time reaches the preset end point of the current control cycle, the evaluation process is automatically triggered to prepare for possible cycle switching;

[0066] Offset Trigger: If the average deviation calculated within the sliding window continues to exceed the preset tolerance threshold, it is determined that the system's ecological state has significantly deviated from the typical description of the current cycle, and the cycle division needs to be re-examined and adjusted immediately.

[0067] Event Trigger: When a severe shock to the influent load is detected (such as an abnormally sharp increase in pollutant concentration) or an extreme weather event (such as sustained high temperature or heavy rainfall), an emergency assessment will be immediately triggered to deal with the emergency, regardless of the current degree of deviation.

[0068] If any of the triggering conditions is activated, a dynamic adjustment process for the control cycle will be executed based on the trigger type and real-time operational data, including:

[0069] Periodic boundary redraw (triggered by major response offset or major event):

[0070] Real-time data sequences that deviate from the original cycle characteristics in the recent period (especially within the sliding window M) are taken as new time series segments and fused with the original historical operation dataset. Based on the fused dataset, the same time series clustering analysis as in step one is re-executed to generate a new control cycle division scheme.

[0071] The new scheme reflects the time structure after incorporating the latest ecological behavior patterns. The quantity, time boundaries, and typical characteristic vectors of each cycle may change. This operation is used to deal with the failure of the original cycle framework caused by the gradual change of ecosystem state or major external shocks.

[0072] Periodic attribute updates (primarily triggered by minor offsets or as a regular optimization for timed triggering):

[0073] Without changing the time boundaries of the current and subsequent control cycles, the descriptive attributes of the control cycle are optimized using real-time operational data;

[0074] Specifically, the typical ecological characteristic vector of the current regulation cycle (or the next cycle to which it is about to switch) is recalculated, and real-time characteristic data is included in the calculation scope so that the characteristic vector can better reflect the recent actual state. The related basic management objectives are reassessed and updated. For example, if recent data shows that carbon sink capacity continues to increase, the weight of "carbon sink gain" in the basic objectives may be strengthened. The above operation is a fine-tuning to make the cycle definition more in line with reality without changing the overall time frame.

[0075] Special periodic insertion and processing (specifically designed to respond to specific event triggers):

[0076] When a clear short-term emergency occurs (such as extreme rainstorms lasting several days or known short-term discharge of high-concentration wastewater), a temporary special response sub-cycle is dynamically inserted within the current control cycle framework.

[0077] Special response subcycles have a clear and short start and end time (usually matching the duration of the event) and define highly specific management objectives that differ from regular cycles, such as operating safely at maximum hydraulic load during heavy rainfall and suspending all ecological interventions.

[0078] After a special response sub-cycle ends, it automatically exits and returns to the original cycle sequence, or triggers a new assessment based on the residual impact of the event;

[0079] Based on the above dynamic adjustment process, a new and internally consistent control cycle division scheme is generated.

[0080] Based on the integration of steps one and two, the essence of dividing the regulation cycle is the process of initialization and continuous optimization, which enables the perception and response to real-time ecological conditions and external problems. This breaks the pattern of operating according to a fixed calendar, and can both follow the long-term macro-ecological rhythms revealed by history (such as seasonal changes and plant annual cycles) and flexibly adapt to short-term and unpredictable micro-fluctuations (such as abnormal climate and sudden pollution loads).

[0081] The initial framework based on historical patterns provided in Step 1 ensures macro-level foresight and stability in management; the adjustment capability based on real-time feedback provided in Step 2 endows the system with micro-level flexibility and adaptability. This intelligent time management capability, characterized by macro-level stability and micro-level flexibility, is the fundamental prerequisite for contextualized multi-objective sequencing and optimization in the subsequent Step 3. It ensures that the system can identify the correct principal contradiction at the right time, thereby providing a temporal context for generating optimal management instructions that coordinate the three major objectives. This enables the continuous optimization and stable improvement of the system's processing efficiency, carbon sequestration gains, and comprehensive ecological health benefits throughout its entire lifecycle.

[0082] Step 3: During the regulation cycle, collect and analyze real-time operating data. Based on the analysis results and the management objectives of the corresponding management cycle, determine the priority order of multi-objective decisions, with the constraint being the minimization of multi-objective disturbances. The multi-objectives are pollutant removal efficiency, system carbon sink function, and ecosystem.

[0083] In step three, within any one control cycle;

[0084] Continuously collect the corresponding real-time running data stream, and extract and calculate the following three target current status index values ​​using the same method as in step one:

[0085] The relative difference in removal rates is used to characterize the decontamination efficiency index;

[0086] The process of obtaining the relative difference in removal rate is as follows: calculate the mass concentration difference of key pollutants (represented by total nitrogen and total phosphorus) in the current influent and effluent, divide it by the influent concentration to obtain the real-time removal rate, and obtain the expected removal rate benchmark determined by the characteristics of the current control cycle. Calculate the difference between the expected removal rate benchmark and the real-time removal rate, and then calculate the ratio of the difference to the expected removal rate benchmark as the relative difference in removal rate.

[0087] Based on the equipped eddy covariance instrument, net ecosystem exchange data is obtained, and negative values ​​of the exchange data are used as indicators of carbon sink function.

[0088] Indicators of ecosystem disturbance based on time recovery factors;

[0089] The process of obtaining the time recovery factor is as follows: record the completion time of the last execution of a specific type of ecological management operation, calculate the number of consecutive days that have passed from that time point to the current time as the recovery days, and calculate the ratio of the recovery days to the standard recovery cycle to obtain the time recovery factor.

[0090] The standard recovery period is determined based on the type of operation and historical system recovery data. The standard recovery period is the typical time required for the ecological function to recover to a stable state after completing one such operation. It is derived from historical data analysis. The closer the time recovery factor is to or greater than 1, the more fully the system has recovered.

[0091] Use the reciprocal of the time recovery factor as an indicator of ecosystem disturbance;

[0092] It should be noted that taking the reciprocal of the time recovery factor is to make the ecosystem disturbance index positively correlated with the disturbance risk. When the recovery days are short, the ecosystem disturbance index will be large, indicating a higher potential ecological disturbance risk.

[0093] Based on the fundamental management objectives associated with the current regulatory cycle;

[0094] Among them, the basic management objectives tend to be obtained from the aforementioned step one, such as prioritizing efficient decontamination, prioritizing carbon sink gains, prioritizing ecological restoration, or balancing multiple objectives;

[0095] The process of determining the priority order of multi-objective decisions based on rule matching and state fine-tuning is as follows:

[0096] Based on a pre-defined rule base, a default priority order for multiple objectives is defined under different basic management goal orientations. Examples of rule bases include:

[0097] When the preference is "high efficiency decontamination priority": the default priority order is: 1) pollutant removal efficiency, 2) system carbon sink function, 3) minimization of ecosystem disturbance;

[0098] When the preference is "carbon sink gain priority": the default priority order is: 1) system carbon sink function, 2) pollutant removal efficiency, 3) minimization of ecosystem disturbance;

[0099] When the preference is "ecological restoration first": the default priority order is: 1) minimizing ecosystem disturbance, 2) system carbon sink function, 3) pollutant removal efficiency;

[0100] When the preference is "multi-objective balance": the default priority order is: 1) minimizing ecosystem disturbance, 2) pollutant removal efficiency, 3) system carbon sink function (as an example balance strategy).

[0101] By reading the basic management objectives associated with the current control cycle and based on the aforementioned rule base, the default priority order of the three major objectives under the corresponding control cycle is directly mapped.

[0102] Based on the calculated pollution removal efficiency index, carbon sequestration function index, and ecological disturbance index, it is determined whether there are any anomalies.

[0103] If the decontamination efficiency index is greater than the corresponding abnormal threshold, it means that the pollutant removal rate is lower than expected, and the pollutant removal efficiency is judged to be in an abnormal state.

[0104] If the carbon sequestration function index is less than the corresponding abnormal threshold, it indicates that the system's carbon sequestration capacity is extremely weak or has become a carbon source, and the system's carbon sequestration function is judged to be in an abnormal state.

[0105] If the ecological disturbance index is greater than the corresponding abnormal threshold, it indicates that the recovery is extremely insufficient and the disturbance risk is extremely high, and the disturbance risk is judged as an abnormal state.

[0106] It should be noted that the decision-making meaning of abnormal states is as follows: when a certain target is judged to be in an "abnormal state", it indicates that the system is currently facing or about to face serious performance degradation or security risks in that target dimension. When making decisions, handling the abnormal state and restoring the system to safe and stable operation becomes the most urgent task.

[0107] Based on the implications of abnormal states in decision-making, fine-tuning rules are executed:

[0108] Principle: Any abnormal state of a target has the highest priority for immediate response;

[0109] Rule: Check if the target ranked first in the current default order is in its own "abnormal state";

[0110] If so, the current default order will remain unchanged (because the primary objective itself is in urgent need and needs to be dealt with first).

[0111] If not, check if any other targets are in an "abnormal state";

[0112] If so, the target that is in an "abnormal state" and ranks first in the default order will be temporarily promoted to first place, and the original first-place target will be moved up one place to form the final priority order.

[0113] If not, the default priority order will be maintained as the final order;

[0114] The final multi-objective decision priority order, determined through the above rule matching and state fine-tuning, will be output.

[0115] Step 3 combines the characteristics of the regulation cycle with real-time operational data to output a clear priority order. By judging and fine-tuning abnormal states, it embeds the principle of prioritizing abnormal states, thereby enabling timely blocking and priority handling of sudden risks, enhancing the system's robustness, and providing the optimal management strategy for the optimization in Step 4 below, which is in line with the current ecological scenario and realistic constraints.

[0116] Step 4: Based on the priority order of multi-objective decisions, use a multi-objective optimization algorithm to solve the problem, generate and output ecological management operation instructions for the current dynamic management stage, which include at least plant harvesting and sediment cleaning strategies.

[0117] In step four, based on the final multi-objective decision priority order output in step three, a constrained multi-objective optimization problem model is constructed to solve for the optimal ecological management operation parameters.

[0118] Decision variables (X): Define a set of variables representing the management operations to be optimized, including at least:

[0119] Plant harvesting related variables: such as the planned harvest time, the area to be harvested (which can be expressed as a percentage of the total area), the intensity of harvesting, or the stubble height;

[0120] Variables related to sediment removal: such as the planned removal time, the area to be removed, and the removal depth;

[0121] Other adjustable operating parameters: such as adjustable water level setpoint and hydraulic residence time target within the current cycle (if the system has the corresponding control capability).

[0122] Objective function (F): Establish three mathematical functions corresponding to the objective of step three, used to evaluate the expected effect of any operation plan X on the three objectives;

[0123] The objective function for decontamination efficiency is F_P(X): After executing operation plan X, the system is expected to achieve the average removal rate of key pollutants (such as total phosphorus) in the next evaluation period (such as the next control cycle or the next 30 days), or the predicted improvement in the removal rate relative to the baseline. The objective is to maximize F_P(X).

[0124] The objective function for carbon sink function is F_C(X): predicting the net change in net carbon sink (the integral of the negative net ecosystem exchange) or carbon storage in the system during the corresponding time period after implementing operation plan X, with the goal of maximizing F_C(X).

[0125] Ecological disturbance objective function F_E(X): quantifies the intensity of ecological disturbance directly caused by the operation scheme X itself. This can be the weighted sum of the disturbance indicators of all operations (harvesting, dredging, etc.) in the scheme (based on the reciprocal of the recovery factor corresponding to its operation time, etc.), or the predicted impact on the integrity of the system structure (such as vegetation cover loss rate). The goal is to minimize F_E(X).

[0126] The constraints include:

[0127] Hard constraints: conditions that must be met, such as: the quality of the effluent must meet regulatory standards; no operation may be carried out during clearly prohibited ecologically sensitive periods; the manipulated variables have corresponding physical upper and lower limits;

[0128] Soft constraint / preference guidance: Transforms the final priority order into guidance for the optimization search process;

[0129] For example, if the order is "1) pollutant removal efficiency, 2) system carbon sink function, 3) ecosystem disturbance minimization", then the algorithm can be set as follows: first find a solution set that satisfies F_P(X), then optimize F_C(X) in the solution set, and finally take F_E(X) into account. This can be achieved by lexicographical optimization, priority weighting method (giving high priority objectives a lot of weight) or setting a hierarchical optimization structure in the algorithm.

[0130] Perform multi-objective optimization solution: The above model is solved using a multi-objective optimization algorithm applicable to nonlinear and possibly nonconvex models;

[0131] Optional algorithms: Use multi-objective evolutionary algorithms such as non-dominated sorting genetic algorithm with elitist strategy (NSGA-II) or particle swarm optimization (PSO) to search in a complex decision space and generate a set of Pareto optimal solutions;

[0132] Using decision variable X as individuals, the population evolves continuously through iteration (selection, crossover, mutation). In each generation, the fitness of each individual is evaluated according to the objective functions F_P(X), F_C(X), and F_E(X). Based on the priority order-guided sorting or selection mechanism (such as comparing F_P(X) first, then comparing F_C(X)), the best individuals are retained, and a Pareto optimal solution set is output.

[0133] Select the final solution from the Pareto optimal solution set and generate operation instructions:

[0134] The final execution plan is selected based on decision preferences. The selection strategy can be consistent with the priority order of the final multi-objective decision. For example, on the Pareto front, the solution with the highest F_P(X) is preferred (if P has the highest priority), or a solution that is relatively balanced on the three objectives but strictly conforms to the priority order is selected.

[0135] Command generation and output: Decode the final selected execution plan into specific, executable ecological management operation commands. The commands must include at least:

[0136] Plant harvesting instructions: Clearly instruct on the harvesting time, the harvested area, and the harvesting operation to be carried out according to the required stubble height;

[0137] Sediment Removal Instructions: Clearly specify the time, area, and depth of sediment removal operations to be carried out.

[0138] Command format: Commands should be generated in a structured data format (such as JSON, XML) or natural language description to ensure that they can be parsed and executed by the automatic control system or clearly and accurately conveyed to on-site maintenance personnel.

[0139] Example 2

[0140] Based on the same inventive concept as the energy-saving wastewater treatment method based on ecological simulation in the foregoing embodiments, such as Figure 2 As shown, this application provides an energy-saving wastewater treatment system based on ecological simulation, wherein the system specifically includes:

[0141] Initial control cycle design module: Based on the historical operation data of the target wastewater treatment system, the continuous operation time of the system is initially divided into multiple control cycles with different ecological characteristics and management objectives using time series analysis.

[0142] Control cycle update module: Based on the initially defined control cycle, the module dynamically evaluates and adjusts the control cycle division based on real-time updated operating data during system operation.

[0143] Multi-objective sequential output module: During the control period, it collects and analyzes real-time operating data within the control period. Based on the analysis results and the management objectives of the corresponding management period, it determines the priority order of multi-objective decisions, with the constraint being the minimization of multi-objective disturbances. Among these, the multi-objectives are pollutant removal efficiency, system carbon sink function, and ecosystem.

[0144] Management optimization and control module: Based on the priority order of multi-objective decisions, it uses a multi-objective optimization algorithm to solve the problem, generate and output ecological management operation instructions for the current dynamic management stage, which includes at least plant harvesting and sediment cleaning strategies.

[0145] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. An energy-saving wastewater treatment method based on ecological simulation, characterized in that: include: Based on the historical operation data of the target wastewater treatment system, the continuous operation time of the system is initially divided into multiple control cycles with different ecological characteristics and management objectives using time series analysis. Based on the initial division of the control cycle, during the system operation, the division of the control cycle is dynamically evaluated and adjusted based on real-time updated operating data. During the regulation cycle, real-time operational data is collected and analyzed. Based on the analysis results and the management objectives of the corresponding management cycle, the priority order of multi-objective decisions is determined, with the constraint being the minimization of multi-objective disturbances. Among these multi-objectives are pollutant removal efficiency, system carbon sink function, and ecosystem. Based on the multi-objective decision priority order, a multi-objective optimization algorithm is used to solve the problem, generate and output ecological management operation instructions for the current dynamic management stage, which include at least plant harvesting and sediment cleaning strategies.

2. The energy-saving wastewater treatment method based on ecological simulation according to claim 1, characterized in that: The system's continuous operating time is initially divided into multiple regulation cycles with different ecological characteristics and management objectives, including: Obtain the historical operation dataset of the target wastewater treatment system. The historical operation dataset includes multi-dimensional time series, which includes: water quality parameter series, climate driving series, ecological state series, and operation log series. The historical operation dataset is preprocessed to extract time-series features that characterize the system's processing efficiency, carbon cycle status, and ecological stability, and a time-series feature set is constructed. A clustering algorithm suitable for multivariate time series is adopted to perform unsupervised learning on the time series feature set. The continuous running time axis is divided into K time periods, so that the change patterns of each feature within the same time period are similar. The K time periods are the initial control period. Each regulation cycle is automatically assigned a typical ecological characteristic vector defined by the cluster center and associated with the basic management target tendency based on historical statistics.

3. The energy-saving wastewater treatment method based on ecological simulation according to claim 1, characterized in that: The division of the regulatory cycle is dynamically evaluated and adjusted, including: During system operation, real-time operational data is continuously collected, and real-time feature sets are calculated. Set a sliding time window, calculate the similarity between the real-time feature set and the typical ecological feature vector of the current regulation cycle, and evaluate the degree of deviation of the real-time state of the system from the current cycle. Based on preset trigger conditions, determine whether to activate the dynamic adjustment program of the control cycle; When the trigger condition is activated, the corresponding control cycle adjustment operation is executed according to the trigger type and real-time running data; Based on the adjustment operation, the system's control cycle division scheme is generated and updated.

4. The energy-saving wastewater treatment method based on ecological simulation according to claim 3, characterized in that: The similarity calculation uses the cosine similarity or Euclidean distance formula; The degree of deviation is the complement of the similarity.

5. The energy-saving wastewater treatment method based on ecological simulation according to claim 3, characterized in that: The triggering conditions include: timed triggering, offset triggering, and event triggering; Timed triggering: When the system running time reaches the preset end point of the current control cycle, the evaluation process is automatically triggered; Offset Trigger: If the average deviation calculated within the sliding window continues to exceed the preset tolerance threshold, it is determined that the ecological state of the system has deviated from the typical description of the current regulation cycle, and the adjustment of the regulation cycle division must be triggered immediately. Event Trigger: When a severe impact on the inflow load or an extreme weather event is detected, an emergency assessment will be triggered immediately, regardless of the current degree of deviation.

6. The energy-saving wastewater treatment method based on ecological simulation according to claim 1, characterized in that: Determining the priority order of multi-objective decisions includes: During the regulation period, real-time operating data is collected and three target current status index values ​​are extracted, namely, pollution removal efficiency index, carbon sink function index, and ecological disturbance index. Based on the underlying management objectives associated with the current regulatory cycle, the corresponding default priority order is mapped from the pre-set rule base; Based on the calculated pollution removal efficiency index, carbon sink function index, and ecological disturbance index, it is determined whether each target is in an abnormal state. If an abnormal state exists, the default priority order is fine-tuned according to the decision meaning of the abnormal state to generate the final multi-objective decision priority order.

7. The energy-saving wastewater treatment method based on ecological simulation according to claim 6, characterized in that: The decontamination efficiency index is based on the relative difference in removal rates; Calculate the mass concentration difference of key pollutants in the current influent and effluent, divide it by the influent concentration to obtain the real-time removal rate, and obtain the expected removal rate benchmark determined by the characteristics of the current control cycle. Calculate the difference between the expected removal rate benchmark and the real-time removal rate, and then compare the difference with the expected removal rate benchmark to obtain the relative gap in removal rate. The carbon sink function index is based on the equipped eddy covariance instrument, which obtains net ecosystem exchange data, and uses the negative value of the exchange data as the carbon sink function index.

8. The energy-saving wastewater treatment method based on ecological simulation according to claim 6, characterized in that: The ecological disturbance index is characterized based on the time recovery factor; Record the completion time of the last specific type of ecological management operation, calculate the number of consecutive days that have passed from that time to the current time as the recovery days, and calculate the ratio of the recovery days to the standard recovery cycle to obtain the time recovery factor; The reciprocal of the time recovery factor is used as an indicator of ecological disturbance.

9. The energy-saving wastewater treatment method based on ecological simulation according to claim 1, characterized in that: Generate and output ecological management operation instructions for the current dynamic management phase, including at least plant harvesting and sediment cleanup strategies, including: Based on the priority order of multi-objective decision-making, the constrained multi-objective optimization problem model includes: Decision variables are used to represent the ecological management operational parameters to be optimized. The objective function is used to evaluate the expected effects of the operating scheme on three aspects: pollutant removal efficiency, system carbon sink function, and ecosystem disturbance. Constraints include effluent quality standards, upper and lower limits of manipulated variables, and optimization guidance based on priority order; The model is solved using a multi-objective optimization algorithm to generate a set of Pareto optimal solutions; The final execution plan is selected from the Pareto optimal solution set and decoded into specific, executable ecological management operation instructions.

10. An energy-saving wastewater treatment system based on ecological simulation, characterized in that, The system is used to perform the method according to any one of claims 1-9, the system comprising: Initial control cycle design module: Based on the historical operation data of the target wastewater treatment system, the continuous operation time of the system is initially divided into multiple control cycles with different ecological characteristics and management objectives using time series analysis. Control cycle update module: Based on the initially defined control cycle, the module dynamically evaluates and adjusts the control cycle division based on real-time updated operating data during system operation. Multi-objective sequential output module: During the control period, it collects and analyzes real-time operating data within the control period. Based on the analysis results and the management objectives of the corresponding management period, it determines the priority order of multi-objective decisions, with the constraint being the minimization of multi-objective disturbances. Among these, the multi-objectives are pollutant removal efficiency, system carbon sink function, and ecosystem. Management optimization and control module: Based on the priority order of multi-objective decisions, it uses a multi-objective optimization algorithm to solve the problem, generate and output ecological management operation instructions for the current dynamic management stage, which includes at least plant harvesting and sediment cleaning strategies.

Images