Method and device for dynamic evaluation and correction of operation performance of sewage treatment plant and medium

Abstract

The present application relates to the sewage treatment technical field, discloses sewage treatment plant operation performance dynamic evaluation and correction method, device and medium, the present application constructs the water inlet working condition index set;According to the water inlet working condition index set, the comprehensive index for representing the overall state of water inlet is constructed;According to the change trend of comprehensive index with time, whether the water inlet working condition occurs phase change is identified, if phase change occurs, then different operation stages are divided;The first model of the same response curve shared by each operation stage and the second model of each operation stage respectively having respective response curve are constructed, and whether the response relationship of operation performance changes with the change of water inlet condition under different operation stages is judged;If reconstruction occurs, then the operation performance evaluation benchmark corresponding to different stages is determined, and the evaluation result and correction suggestion are output;Solve the technical problems of low accuracy of operation performance evaluation in the prior art, difficult to support fine regulation and control and low-carbon operation decision-making of sewage treatment plant under the background of plant network integration.

Description

Technical Field

[0001] This invention relates to the field of wastewater treatment technology, specifically to a method, apparatus, and medium for dynamic evaluation and correction of the operational performance of wastewater treatment plants. Background Technology

[0002] Urban wastewater treatment plants are crucial infrastructure for urban water pollution control and greenhouse gas emission reduction. As the wastewater treatment industry shifts from single-plant operation and management to a more systematic operation model encompassing plant-network collaboration, integrated plant-network-river management, and integrated water supply and drainage, the influent conditions of wastewater treatment plants are no longer solely determined by the plant's internal processes. Instead, they are simultaneously influenced by factors such as pipeline collection efficiency, external water infiltration, stormwater and sewage mixing remediation, pipeline scheduling, pump station operation, regional interception, and changes in the structure of wastewater dischargers. Particularly under the integrated plant-network operation model, pipeline management and scheduling strategies can significantly alter the influent concentration, quantity, and composition of wastewater, as well as their relative proportions, thereby affecting operational performance indicators such as energy consumption per unit of pollutant reduction and carbon emission reduction per unit of pollutant.

[0003] Currently, existing methods for evaluating the operational performance of wastewater treatment plants typically rely on indicators such as influent COD, NH3-N, TN, treated water volume, hydraulic load rate, electricity consumption per ton of water, electricity consumption per unit of COD reduction, and carbon emission intensity. These methods assess the plant's operational level through annual statistics, monthly comparisons, linear correlation analysis, benchmark comparisons, or static evaluation models. While these methods can reflect the average operational status over a certain period, they generally suffer from the following shortcomings: First, the characterization of influent operating conditions is incomplete. Existing methods often use single pollutant concentrations or single load indicators to characterize influent conditions, making it difficult to comprehensively depict the systemic changes in influent operating conditions under the context of integrated plant and grid systems. In actual operation, influent conditions are not only manifested in changes in concentrations of COD, TN, TP, and SS, but also in adjustments to physicochemical ratios such as COD / TN, TN / TP, SS / COD, and BOD5 / COD. Single indicators are insufficient to identify such multidimensional changes in operating conditions, easily leading to incomplete judgments of influent conditions.

[0004] Second, it is difficult to identify transitions between different influent operating conditions. Existing methods often treat the entire operation of a wastewater treatment plant as a whole, assuming that the relationship between influent conditions and operational performance remains stable during the study period. This makes it difficult to identify potential transitions between different influent operating conditions during integrated plant-network operation. As measures such as pipeline improvement, increased wastewater collection rates, external water control, and joint scheduling continue to be implemented, the influent status of the wastewater treatment plant may transition from one operational phase to another. If a uniform baseline is still used for performance evaluation, it may lead to biased assessment results.

[0005] Third, it fails to distinguish the root causes of performance changes. Existing operational performance evaluations typically only assess the level of energy consumption or carbon emission intensity, making it difficult to further differentiate whether performance changes stem from fluctuations in influent concentration, overall shifts in operational phases, or changes in the response relationship between influent conditions and operational performance. For example, under the same influent COD concentration, the energy consumption per unit of COD reduction may differ across different phases; similarly, under the same influent NH3-N concentration, the energy consumption per unit of NH3-N reduction may also vary across different phases. Without recognizing this response function reconstruction, it will be difficult to support the dynamic correction of operational strategies.

[0006] Fourth, there is a lack of phased correction mechanisms. Existing technologies lack phased correction mechanisms for synergistic optimization of pollution reduction and carbon reduction. When the system enters a new operational phase, the concentration-performance relationship established based on historical phases may no longer be applicable. If the original model is continued to be used for energy consumption, carbon emission intensity evaluation and operational optimization, the phase differences may be underestimated, thereby affecting the refined control and low-carbon operation decisions of integrated wastewater treatment plants. Summary of the Invention

[0007] This invention provides a method, device, and medium for dynamic evaluation and correction of the operating performance of wastewater treatment plants, in order to solve the technical problems in the prior art, such as incomplete influent operating condition characterization, difficulty in identifying stage transitions, inability to distinguish the root causes of performance changes, and lack of staged correction mechanisms, which lead to low accuracy in operating performance evaluation and difficulty in supporting refined control and low-carbon operation decisions of wastewater treatment plants in the context of plant-network integration.

[0008] In a first aspect, the present invention provides a method for dynamic evaluation and correction of the operational performance of a wastewater treatment plant, comprising: acquiring operational data of the wastewater treatment plant; constructing an influent operating condition index set based on the operational data, the influent operating condition index set including multi-dimensional influent operating condition indicators; constructing a comprehensive index for characterizing the overall state of the influent based on the influent operating condition index set; identifying whether the influent operating conditions have undergone phased changes based on the changing trend of the comprehensive index over time, and dividing the operation into different phases if phased changes occur; constructing a first model sharing the same response curve for each operational phase and a second model having its own response curve for each operational phase, comparing the fitting effect of the first model and the second model, and determining whether the response relationship of operational performance with changes in influent conditions under different operational phases has been reconstructed; if it is determined that a reconstruction has occurred, determining the corresponding operational performance evaluation benchmark according to different operational phases, and outputting evaluation results and correction suggestions.

[0009] This invention addresses the problem in existing technologies where a single indicator is insufficient to comprehensively depict the systematic changes in influent operating conditions under the integrated plant-network context by acquiring operational data from wastewater treatment plants and constructing a set of influent operating condition indicators. It identifies whether influent operating conditions have undergone phased changes and divides the operation into different phases based on the changing trends of the comprehensive indicators over time, avoiding misjudging the entire operation period as a single stable state. By constructing a first model sharing the same response curve across all operating phases and a second model with its own response curve for each phase, the fitting effects of the first and second models are compared to determine whether the response relationship of operational performance under different operating phases has been reconstructed. This solves the problem in existing technologies where it is difficult to distinguish whether performance changes originate from fluctuations in influent concentration, overall shifts in the operating phase, or changes in the response relationship itself. Furthermore, after determining that a reconstruction has occurred, corresponding operational performance evaluation benchmarks are determined for different operating phases, and evaluation results and correction suggestions are output. This realizes the transformation of operational performance evaluation from static benchmarking to dynamic correction, providing data support for refined control and low-carbon operation decisions of wastewater treatment plants under the integrated plant-network context.

[0010] In one optional implementation, the operational data includes at least the data required to construct the influent operating condition index set and determine the operational performance indicators: when the operational performance indicators include power consumption per unit of pollutant reduction, the operational data includes power consumption, treated water volume, and influent and effluent pollutant concentrations; when the operational performance indicators include pollutant reduction carbon emission intensity, the operational data includes carbon emission source data and pollutant reduction amount data. In this implementation, the operational data is not a pre-defined fixed list of data types, but rather the required data types are determined in reverse based on the actual selected operational performance indicators. That is, when calculating power consumption per unit of pollutant reduction, data such as power consumption, treated water volume, and influent and effluent pollutant concentrations are automatically associated; when calculating pollutant reduction carbon emission intensity, data such as carbon emission source data and pollutant reduction amount are automatically associated. This data acquisition method ensures that each technical solution has a complete data foundation, avoiding the inability to calculate subsequent indicators due to a lack of necessary data, and also avoids forcibly superimposing the data requirements of all evaluation indicators onto the same solution, thus increasing the data acquisition burden. Therefore, while ensuring the feasibility of the solution, it also takes into account the flexibility and adaptability of data collection.

[0011] In one optional implementation, the influent operating condition index set is constructed based on the influent pollutant concentration, influent flow rate, and concentration ratio between pollutants in the operating data. It includes influent pollutant concentration index, hydraulic load rate index, and physicochemical ratio index. The physicochemical ratio index includes one or more of COD / TN, TN / TP, SS / COD, and BOD5 / COD. Traditional methods often use single pollutant concentrations or single load indicators to characterize influent conditions, which are difficult to reflect the complex changes in influent water quality under the background of integrated plant and network. However, this implementation method constructs an influent operating condition index set by simultaneously introducing influent pollutant concentrations, hydraulic loads, and multiple physicochemical ratios such as COD / TN, TN / TP, SS / COD, and BOD5 / COD. This can comprehensively characterize the influent state from multiple dimensions such as concentration level, water load, and component structure. In particular, the ratios such as COD / TN and TN / TP reflect the carbon-nitrogen matching and nitrogen-phosphorus ratio, while SS / COD and BOD5 / COD reflect the suspended solids ratio and the biodegradability of organic matter. This information is crucial for judging the operating performance of wastewater treatment plants, avoiding information gaps and judgment biases caused by single indicators, thereby improving the completeness of influent operating condition characterization and the accuracy of subsequent stage identification.

[0012] In one optional implementation, a comprehensive index for characterizing the overall state of the influent is constructed based on a set of influent operating condition indicators. This includes standardizing the influent operating condition indicator set and then performing dimensionality reduction and fusion on the standardized multiple influent operating condition indicators to obtain the comprehensive index. This implementation integrates multi-dimensional influent operating condition indicators into a single comprehensive index, achieving a unified quantitative characterization of complex influent states under the background of integrated plant and grid systems. It avoids the one-sidedness of single-indicator evaluation and solves the problems of redundant information in multi-dimensional indicators and difficulty in directly using them for subsequent stage identification. It provides a quantifiable and comparable unified scale for time series analysis and stage transition identification of influent operating conditions, thereby improving the feasibility and accuracy of subsequent stage change identification.

[0013] In one optional implementation, the method identifies whether the influent operating conditions have undergone phased changes based on the changing trend of comprehensive indicators over time. This includes: constructing a sequence of comprehensive indicators in chronological order; performing breakpoint detection on the sequence; determining that a phased change in the influent operating conditions has occurred when a breakpoint is detected; and dividing the sequence into different operating stages based on the breakpoint. This implementation utilizes breakpoint detection technology to automatically identify phased changes in influent operating conditions, avoiding the subjectivity and lag of relying on human experience. When changes in the influent status occur during integrated plant and network operation due to measures such as pipeline management, external water control, or joint scheduling, it can objectively and promptly identify stage transitions and divide operating stages. This allows for targeted analysis of operational performance evaluation based on actual operating stage conditions, avoiding the distortion of evaluation results caused by mixing data from different stages, and providing an accurate stage division basis for subsequent phased determination of response relationships.

[0014] In one optional implementation, the fitting effects of the first model and the second model are compared to determine whether the response relationship between operating performance and changes in influent conditions has been reconstructed under different operating stages. This includes: if the fitting effect of the second model is better than that of the first model, then it is determined that the response relationship has been reconstructed. This implementation can objectively determine whether the response relationship between influent conditions and operating performance has undergone substantial changes under different operating stages. When the fitting effect of the second model is better than that of the first model, it indicates that the same influent conditions correspond to different operating performance levels at different stages, and the shape of the response curve between the two has changed. Based on this identification result, it is possible to further distinguish whether the performance change comes from the overall shift of the operating stage or the change in the response relationship itself, thereby avoiding misjudging the reconstruction of the response relationship as a simple fluctuation in the performance level, providing an objective basis for whether evaluation standard correction is needed, and improving the scientificity and accuracy of operating performance evaluation.

[0015] In one optional implementation, the first model is an additive generalized additive model, and the second model is an interactive generalized additive model. Comparing the fitting effects includes comparing the goodness of fit between the additive generalized additive model and the interactive generalized additive model, as well as the significance of the stage interaction terms. If the interactive generalized additive model outperforms the additive generalized additive model and the stage interaction terms meet the preset significance condition, then the response relationship is determined to have been reconstructed. This implementation uses an additive generalized additive model and an interactive generalized additive model for comparison. The generalized additive model can flexibly capture the nonlinear response law of operational performance changing with influent conditions, avoiding the underfitting problem of traditional linear models for complex response relationships. By introducing stage interaction terms and testing their significance, it is possible to objectively determine from a statistical perspective whether the response curves at different stages have significant differences, avoiding the uncertainty brought about by subjective judgment, thereby improving the accuracy and reliability of response relationship reconstruction judgment and providing statistical support for the establishment of subsequent staged operational performance evaluation benchmarks.

[0016] In one optional implementation, corresponding operational performance evaluation benchmarks are determined according to different operational stages, including: acquiring operational performance indicator data for each operational stage; using the data as samples, establishing response curves for operational performance at each stage as a function of influent conditions, and using the response curves as operational performance evaluation benchmarks for the corresponding stages; wherein, the operational performance evaluation benchmarks include benchmark values ​​for power consumption per unit of pollutant reduction at different operational stages, whereby power consumption per unit of pollutant reduction is used to characterize the amount of power consumed to treat a unit of pollutant, and is determined based on power consumption, treated water volume, and the difference in pollutant concentration between influent and effluent during the corresponding time period. This implementation method establishes corresponding response curves for different operating stages after determining that the response relationship has been reconstructed, serving as the performance evaluation benchmark for each stage. This ensures that each stage has an evaluation standard that matches its influent conditions, avoiding evaluation distortion caused by applying a single static model across stages. Furthermore, using power consumption per unit of pollutant reduction as a specific indicator of the evaluation benchmark directly reflects the energy efficiency of the wastewater treatment plant under specific influent conditions. When subsequent data is input, comparing the actual power consumption per unit of pollutant reduction with the benchmark value for that stage accurately determines whether the operating performance deviates from expectations, thus providing clear data support and diagnostic basis for staged operation optimization and low-carbon regulation.

[0017] Secondly, the present invention provides a dynamic evaluation and correction device for the operating performance of a wastewater treatment plant, comprising: a data acquisition module for acquiring operating data of the wastewater treatment plant; an index construction module for constructing an influent operating condition index set based on the operating data, wherein the influent operating condition index set includes multi-dimensional influent operating condition indicators; a comprehensive characterization module for constructing a comprehensive index for characterizing the overall state of the influent based on the influent operating condition index set; a stage identification module for identifying whether the influent operating condition has undergone stage changes based on the changing trend of the comprehensive index over time, and dividing the operating into different operating stages if stage changes occur; a reconstruction discrimination module for constructing a first model sharing the same response curve for each operating stage and a second model having its own response curve for each operating stage, comparing the fitting effect of the first model and the second model, and determining whether the response relationship of operating performance with changes in influent conditions under different operating stages has been reconstructed; and a correction output module for determining the corresponding operating performance evaluation benchmark according to different operating stages if reconstruction is determined, and outputting the evaluation results and correction suggestions.

[0018] Thirdly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the wastewater treatment plant operation performance dynamic evaluation and correction method described in the first aspect or any corresponding embodiment. Attached Figure Description

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the first process of the dynamic evaluation and correction method for the operating performance of a wastewater treatment plant according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the second process of the dynamic evaluation and correction method for the operating performance of a wastewater treatment plant according to an embodiment of the present invention; Figure 3 This is a diagram showing the PC1 load and interpretation degree in a specific embodiment of the present invention; Figure 4 This is a diagram showing the PC1 monthly sequence and breakpoint identification in a specific embodiment of the present invention; Figure 5 This is a three-stage distribution diagram of PC1 in a specific embodiment of the present invention; Figure 6 This is a stage effect diagram after controlling environmental variables in a specific embodiment of the present invention; Figure 7 This is a time-series variation diagram of the monthly COD concentration in the influent in a specific embodiment of the present invention; Figure 8 This is a time-series variation diagram of the monthly TP concentration in the influent in a specific embodiment of the present invention; Figure 9 This is a time-series variation diagram of the monthly SS concentration in the influent in a specific embodiment of the present invention; Figure 10 This is a time-series variation diagram of the monthly TN concentration in the influent in a specific embodiment of the present invention; Figure 11 This is a box plot showing the COD concentration distribution of the influent in the three stages of this invention in a specific embodiment. Figure 12 This is a box plot showing the TP concentration distribution in the three-stage influent in a specific embodiment of the present invention; Figure 13 This is a box plot showing the SS concentration distribution in the three-stage influent in a specific embodiment of the present invention; Figure 14 This is a box plot showing the TN concentration distribution in the three-stage influent in a specific embodiment of the present invention; Figure 15 This is a nonlinear response curve of power consumption per unit COD reduction with influent COD concentration in a specific embodiment of the present invention; Figure 16 This is a nonlinear response curve of pollutant reduction carbon emission intensity as a function of influent COD concentration in a specific embodiment of the present invention; Figure 17 This is a schematic diagram illustrating the interpretation bias of the GAM model in a specific embodiment of the present invention; Figure 18 This is a schematic diagram of the ΔAIC of the interactive GAM model relative to the additive GAM model in a specific embodiment of the present invention; Figure 19 This is a partial effect curve of the power consumption per unit COD reduction in the three stages as a function of the influent COD concentration in a specific embodiment of the present invention. Figure 20 This is a partial effect curve of the three-stage pollutant reduction carbon emission intensity as a function of influent COD concentration in a specific embodiment of the present invention; Figure 21 This is a partial effect curve of the three-stage unit NH3-N reduction power consumption as a function of the influent NH3-N concentration in a specific embodiment of the present invention; Figure 22 This is a structural block diagram of a wastewater treatment plant operation performance dynamic evaluation and correction device according to an embodiment of the present invention; Figure 23 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0023] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0024] Existing methods for evaluating the operational performance of wastewater treatment plants often use single indicators to characterize influent conditions, making it difficult to comprehensively depict the systematic changes in influent operating conditions under the context of plant-network integration. They also assume that the correspondence between influent conditions and operational performance remains stable during the study period, making it difficult to identify transitions in influent operating conditions. Furthermore, they cannot distinguish whether performance changes stem from fluctuations in influent concentration, overall shifts in the operational phase, or changes in the response relationship itself. Finally, they lack a phased correction mechanism for synergistic optimization of pollution reduction and carbon reduction. Therefore, this invention provides a method, apparatus, and medium for dynamic evaluation and correction of wastewater treatment plant operational performance to address the technical problems existing in the background art.

[0025] According to an embodiment of the present invention, a method for dynamic evaluation and correction of the operating performance of a wastewater treatment plant is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0026] This embodiment provides a method for dynamic evaluation and correction of the operational performance of a wastewater treatment plant. Figure 1 This is a flowchart of a method for dynamic evaluation and correction of wastewater treatment plant operation performance according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps: Step S101: Obtain the operating data of the wastewater treatment plant.

[0027] In this step, operational data refers to various monitoring and recording data generated during the actual operation of the wastewater treatment plant, including but not limited to data on water quality, water quantity, and power consumption. Specifically, this data can be read from the wastewater treatment plant's central control system, online monitoring instruments, or SCADA system via a data interface, or it can be obtained through manual input or data import; this embodiment does not impose any limitations on this method.

[0028] In the operational data, influent refers to the wastewater entering the wastewater treatment plant for treatment, and effluent refers to the water discharged after treatment.

[0029] This step provides a complete and reliable data foundation for subsequent influent condition characterization and operational performance evaluation. Under integrated plant-network operation conditions, the influent status of a wastewater treatment plant is not only determined by the plant's internal processes, but is also affected by factors such as pipeline collection efficiency, external water infiltration, pump station scheduling, rainfall disturbance, and the interception status of the area. By acquiring operational data covering multiple dimensions such as influent and effluent water quality, water quantity, and power consumption, evaluation biases caused by relying on a single data source can be avoided, thereby ensuring the accuracy of subsequent condition characterization and performance evaluation.

[0030] Step S102: Construct a set of water intake condition indicators based on the operating data. The set of water intake condition indicators includes multi-dimensional water intake condition indicators.

[0031] The influent operating condition index set in this step refers to a multi-dimensional set of indicators used to characterize the influent status of a wastewater treatment plant. It is a standardized combination of indicators formed after the original operating data has been structured and processed, and can depict the water quality characteristics, hydraulic load and component structure relationship of the influent from different perspectives.

[0032] This step transforms the raw operational data into a standardized indicator system that can be used for subsequent comprehensive characterization and model analysis. In the context of integrated plant and grid operation, changes in influent conditions are not only reflected in the rise and fall of single pollutant concentrations, but also in fluctuations in water load and adjustments in the proportions of components such as carbon, nitrogen, and phosphorus. By constructing an influent operating condition indicator set that includes multi-dimensional indicators, a more complete information foundation can be provided for the construction of subsequent comprehensive indicators, avoiding operational condition characterization biases caused by single indicators.

[0033] Step S103: Construct a comprehensive index to characterize the overall state of the water intake based on the water intake condition index set.

[0034] The comprehensive index in this step refers to a single quantitative value obtained by mathematically integrating multiple dimensions of influent operating condition indicators. It is used to reflect the overall influent status of the wastewater treatment plant within a certain period of time.

[0035] In this step, the comprehensive index is a sequence value that changes with the sampling time. Each time point corresponds to a comprehensive index value. By analyzing the trend of the comprehensive index over time, the overall evolution of the water intake condition can be tracked.

[0036] This step condenses multi-dimensional influent operating condition information into a traceable and comparable comprehensive indicator. Compared with using single indicators such as COD, TN, or water volume, the comprehensive indicator can more completely reflect the overall level of influent concentration, the magnitude of water volume load, and the comprehensive changes in component structure such as carbon-nitrogen ratio and nitrogen-phosphorus ratio under integrated plant and grid operation, providing a unified and quantifiable analytical basis for subsequent stage transition identification.

[0037] Step S104: Based on the trend of changes in comprehensive indicators over time, identify whether the water intake conditions have undergone phased changes. If phased changes have occurred, divide the operation into different phases.

[0038] In this step, the phased change refers to the transition of the influent operating conditions of the wastewater treatment plant from one relatively stable operating state to another relatively stable operating state in a time series. This change is different from the random fluctuations of daily life. It has the characteristics of continuity and system. It is usually caused by integrated plant and network operation measures such as pipeline network management, external water control, rainwater and sewage diversion transformation, and joint scheduling. Once it occurs, it will continue to affect the influent water quality and quantity of the wastewater treatment plant for a long period of time.

[0039] This step automatically identifies whether there have been continuous phase changes in the influent operating conditions based on the time-series changes of comprehensive indicators. During the integrated operation of the plant and network, systemic measures such as pipeline renovation, external water control, and joint scheduling may lead to long-term and substantial changes in the influent operating conditions of the wastewater treatment plant. By detecting breakpoints, it is possible to objectively determine whether such changes have actually occurred, and accordingly divide the entire operation period into multiple stages with different influent characteristics, thereby avoiding misjudging the entire operation period as the same stable state.

[0040] Step S105: Construct a first model that shares the same response curve for each operating stage and a second model that has its own response curve for each operating stage. Compare the fitting effects of the first model and the second model to determine whether the response relationship of operating performance with changes in influent conditions under different operating stages has been reconstructed.

[0041] In this step, the response relationship refers to the corresponding pattern of how the operating performance of a wastewater treatment plant changes with influent conditions; that is, under specific influent conditions, at what level of energy consumption and carbon emissions does the wastewater treatment plant achieve pollutant reduction? Reconfiguration refers to a substantial change in the form or characteristics of this response relationship between different operational stages; that is, the same influent conditions correspond to different levels of operating performance at different stages. Operating performance refers to the efficiency with which a wastewater treatment plant achieves pollutant reduction at the cost of energy consumption and carbon emissions under specific influent conditions. Influent conditions refer to the characteristics of the influent state that affect the operating effect of a wastewater treatment plant, including factors such as influent pollutant concentration, flow rate, and composition.

[0042] This step is used to determine whether the response relationship between influent conditions and operational performance has substantially changed under different operational stages. This distinction is crucial for determining whether to establish operational performance evaluation benchmarks in stages, and it is also the key difference between this invention and traditional static operational performance evaluation methods.

[0043] In step S106, if a reconfiguration is determined, the corresponding operational performance evaluation benchmarks are determined according to different operational stages, and the evaluation results and correction suggestions are output.

[0044] In this step, the operational performance evaluation benchmark refers to the reference standard used to measure whether the operational performance level of the wastewater treatment plant is reasonable. It is represented by a response curve showing the operational performance changing with influent conditions. This curve reflects the expected level of operational performance under different influent conditions within a specific operational phase. The response curve is a function curve plotted with influent conditions on the horizontal axis and operational performance on the vertical axis. Each point on the curve represents the expected value of operational performance under a certain influent condition. This curve is obtained by fitting historical data within this phase.

[0045] In this step, the evaluation results include the phase division results, key operating condition characteristics of each phase, response function reconstruction judgment results, and the degree of deviation between the actual operating performance and the current phase evaluation benchmark. Correction suggestions include prompts to check strategies related to aeration control, booster pump operation, sludge age, load matching, chemical dosing, sludge treatment, photovoltaic substitution, biogas utilization, or reclaimed water utilization when the operating performance deviates from the reasonable range of the current phase.

[0046] After determining that the response relationship has been reconstructed, this step establishes corresponding operational performance evaluation benchmarks for different stages, so that each stage has an evaluation standard that matches its influent operating conditions. This realizes the transformation of operational performance evaluation from static benchmarking to dynamic correction, providing data support for energy conservation, emission reduction, carbon reduction, and plant-network coordinated scheduling in wastewater treatment plants. When new data is input, the system can call the corresponding response curve according to its stage to conduct performance evaluation and compare the actual performance with the expected performance. If it deviates from the reasonable range of the current stage, it outputs diagnostic prompts and correction suggestions, so that the evaluation results have clear guiding significance.

[0047] The wastewater treatment plant operation performance dynamic evaluation and correction method provided in this embodiment acquires the operation data of the wastewater treatment plant and constructs a set of influent operating condition indicators. It then constructs a comprehensive indicator to characterize the overall state of the influent, solving the problem in existing technologies where a single indicator is insufficient to comprehensively depict the systematic changes in influent operating conditions under the background of plant-network integration. By identifying whether the influent operating conditions have undergone phased changes based on the changing trend of the comprehensive indicator over time and dividing different operating stages, it avoids misjudging the entire operation period as a single stable state. By determining whether the response relationship of operation performance to changes in influent conditions has been reconstructed under different operating stages, it solves the problem in existing technologies of difficulty in distinguishing whether performance changes originate from changes in influent concentration, overall shifts in operating stages, or changes in the response relationship itself. After determining that a reconstruction has occurred, it determines the corresponding operation performance evaluation benchmarks for different operating stages and outputs evaluation results and correction suggestions, realizing the transformation of operation performance evaluation from static benchmarking to dynamic correction. This provides data support for the refined control and low-carbon operation decision-making of wastewater treatment plants under the background of plant-network integration.

[0048] This embodiment provides a method for dynamic evaluation and correction of the operational performance of a wastewater treatment plant. Figure 2 This is a flowchart of a method for dynamic evaluation and correction of wastewater treatment plant operation performance according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps: Step S201: Obtain the operating data of the wastewater treatment plant.

[0049] In one alternative implementation, operational data within a preset time range is acquired. The preset time range can be set according to evaluation needs, such as the past month, a quarter, or a year.

[0050] Operational data includes one or more of the following: influent water quality data, effluent water quality data, influent water volume data, power consumption data, chemical dosing data, sludge treatment and disposal data, meteorological data, and pipeline operation data.

[0051] The influent and effluent water quality data may include COD, BOD5, SS, TN, NH3-N, TP, etc.; water volume data may include daily influent volume, daily treated water volume, hydraulic load rate, etc.; power consumption data may include total power consumption of the plant area, power consumption of booster pumps, power consumption of aeration system, power consumption of return system, etc.; chemical data may include carbon source, phosphorus removal agent, disinfectant, sludge dewatering agent, etc.; meteorological data may include rainfall, temperature, etc.; pipeline network operation data may include pump station scheduling, pipeline network liquid level, pipeline network flow, interception status, external water control records, and plant-network linkage scheduling records.

[0052] In one optional implementation, the operational data includes at least the data required to construct the influent operating condition index set and to determine the operational performance indicators: when the operational performance indicators include power consumption per unit of pollutant reduction, the operational data includes power consumption, treated water volume, and influent and effluent pollutant concentrations; when the operational performance indicators include the carbon emission intensity of pollutant reduction, the operational data includes carbon emission source data and pollutant reduction amount data.

[0053] The significance of acquiring the aforementioned multi-source operational data lies in providing a complete data foundation for subsequent dynamic assessments. Under integrated plant-network operation conditions, the influent status of a wastewater treatment plant is not only determined by the plant's internal processes but is also influenced by factors such as pipeline collection efficiency, external water infiltration, pump station scheduling, rainfall disturbance, and the interception status of the affected area. By acquiring multi-dimensional operational data covering influent and effluent water quality, quantity, power consumption, chemicals, sludge, meteorology, and pipeline operation, evaluation biases caused by relying solely on a single water quality or energy consumption indicator can be avoided, thereby ensuring the completeness and accuracy of subsequent influent condition characterization and operational performance evaluation.

[0054] In one optional implementation, a preprocessing step for operational data is also included. Specifically, after acquiring the operational data of the wastewater treatment plant, the data needs to be preprocessed to eliminate interference from different data sources, sampling frequencies, units of measurement, and abnormal observations on subsequent analysis. Without preprocessing, misjudgment of stage breakpoints, model fitting bias, or distortion of operational performance evaluation may occur.

[0055] In one alternative implementation, the preprocessing includes one or more of the following processing methods: (1) Time alignment: Time alignment of data from different sources is performed according to a unified time scale to ensure that each data sequence is consistent in time.

[0056] (2) Removal of invalid observations: Remove invalid observations caused by instrument failure, equipment maintenance, abnormal test results, shutdown, etc.

[0057] (3) Missing data processing: missing data can be filled by interpolation, nearest neighbor imputation, or mean imputation within the same period, or the missing records can be deleted directly.

[0058] (4) Outlier identification and processing: outlier identification and processing.

[0059] (5) Unit uniformity: unify the units for indicators such as concentration, water volume, power consumption, drug dosage and carbon emissions.

[0060] (6) Standardization: Standardize the variables that enter the comprehensive characterization and model analysis.

[0061] The significance of preprocessing lies in eliminating interference from different data sources, sampling frequencies, units of measurement, and outlier observations that could affect subsequent analysis. Without preprocessing, misjudgments of stage breakpoints, model fitting biases, or distorted performance evaluations may occur.

[0062] Step S202: Construct a set of water intake condition indicators based on the operating data. The set of water intake condition indicators includes multi-dimensional water intake condition indicators.

[0063] In one optional implementation, the influent operating condition index set is constructed based on the influent pollutant concentration, influent flow rate, and concentration ratio between pollutants in the operating data. It includes influent pollutant concentration index, hydraulic load rate index, and physicochemical ratio index. The physicochemical ratio index includes one or more of COD / TN, TN / TP, SS / COD, and BOD5 / COD.

[0064] The influent pollutant concentration refers to the measured concentration values ​​of various pollutants in the influent, reflecting the water quality level, including influent COD, BOD5, SS, TN, NH3-N, TP, etc. Hydraulic load refers to the volume of water entering the wastewater treatment plant per unit time, used to characterize the influent volume load intensity. Pollutant concentration ratios refer to the proportional relationships between different pollutant concentrations, such as the carbon-to-nitrogen ratio (COD / TN) and the nitrogen-to-phosphorus ratio (TN / TP), used to characterize the component structure characteristics of the influent water quality, specifically including COD / TN, TN / TP, SS / COD, BOD5 / COD, etc. Specifically, COD / TN characterizes the degree of carbon-to-nitrogen matching, reflecting the sufficiency of the carbon source; TN / TP characterizes the nitrogen-to-phosphorus ratio; SS / COD characterizes the structural relationship of suspended solids concentration relative to the organic pollution load; and BOD5 / COD characterizes the biodegradability of organic matter. These ratios collectively reflect the overall characteristics of the influent water quality from the perspective of component structure.

[0065] Furthermore, this embodiment also includes the step of constructing a set of operational performance indicators. The set of operational performance indicators characterizes the energy efficiency and low-carbon operation level of the wastewater treatment plant under specific influent conditions, including one or more of the following: electricity consumption per unit of COD reduction, electricity consumption per unit of NH3-N reduction, and carbon emission intensity of pollutant reduction. This set of operational performance indicators provides a quantitative basis for subsequently determining whether the response relationship of operational performance to changes in influent conditions has been restructured at different stages, and serves as the basis for establishing operational performance evaluation benchmarks in stages.

[0066] The unit COD reduction power consumption is used to characterize the amount of power consumed to process a unit of COD, and the calculation formula is as follows: ; In the formula, The energy consumption reduction per unit COD in time period t is expressed in kWh / kg COD. The power consumption during time period t is expressed in kWh. The volume of water processed in time period t is expressed in m³. and The influent and effluent COD concentrations are respectively for time period t, in mg / L; 10⁻³ is the conversion factor from g to kg.

[0067] The unit NH3-N reduction power consumption is used to characterize the amount of power consumed in treating a unit of NH3-N, and the calculation formula is: ; In the formula, The power consumption reduction per unit of NH3-N in time period t is expressed in kWh / kg NH3-N. and The values ​​represent the influent and effluent NH3-N concentrations at time t, respectively, in mg / L.

[0068] Pollutant reduction carbon emission intensity is used to characterize the carbon emission level corresponding to a unit of pollutant reduction, and is calculated using the following formula: ; In the formula, To reduce carbon emission intensity for pollutants in time period t; t represents the carbon emissions from the wastewater treatment plant during time period t, expressed in kgCO2-eq. This represents the reduction amount of pollutants or oxygen-consuming pollutants in time period t, expressed in kg. It is calculated based on COD equivalents, by converting each oxygen-consuming pollutant (including COD and NH3-N, etc.) into COD equivalents according to their respective oxygen requirements and then summing them up. Carbon emission factors include one or more of the following: carbon emissions from electricity consumption, carbon emissions from chemicals, carbon emissions from sludge treatment and disposal, and direct emissions from the process. Carbon substitution from one or more of the following methods may also be deducted: photovoltaic power generation, biogas utilization, and reclaimed water utilization. Specific carbon emission factors can be determined according to national, industry, group, or local standards.

[0069] The significance of constructing the aforementioned influent operating condition index set and operational performance index set lies in transforming raw operational data into standardized indicators that can be used for model identification and performance evaluation. The influent operating condition index is used to characterize the system input state, while the operational performance index is used to characterize the energy consumption and carbon emission costs at which the system achieves pollutant reduction. Together, these two types of indicators form the basis for subsequent response relationship modeling.

[0070] Step S203: Construct a comprehensive index to characterize the overall state of the water intake based on the water intake condition index set.

[0071] This step standardizes the set of influent operating condition indicators and then performs dimensionality reduction and fusion on the standardized influent operating condition indicators to obtain a comprehensive indicator.

[0072] Standardization refers to transforming raw indicator data of different dimensions and orders of magnitude into a comparable unified scale through mathematical transformation. This eliminates the problem of incomparability and inability to directly compare and integrate indicators due to differences in units and numerical ranges. Methods such as Z-score standardization are commonly used. Dimensionality reduction fusion refers to compressing multiple interrelated indicators into a few comprehensive indicators through mathematical methods. This reduces data dimensionality while preserving as much original information as possible, facilitating subsequent tracking, comparison, and analysis.

[0073] In one optional implementation, principal component analysis is used to integrate multiple influent operating condition indicators into a single comprehensive indicator, calculated using the following formula: ; in, Let be the comprehensive index value of the i-th sample; The standardized value of the influent condition index for the i-th sample and the j-th influent condition; is the loading coefficient of the j-th index in the principal component; p is the number of indices included in the comprehensive characterization.

[0074] In practical implementation, the construction of comprehensive indicators is not limited to principal component analysis; factor analysis, partial least squares, autoencoders, or weighted comprehensive indices can also be used. Principal component analysis has the advantages of simple calculation, strong interpretability, and the ability to output indicator loadings, making it suitable for operational management scenarios. Autoencoders are suitable for high-dimensional and complex data, but their interpretability is relatively weak. Those skilled in the art can flexibly choose appropriate dimensionality reduction and fusion methods based on the actual application scenario and data type.

[0075] When the explanatory power of the first principal component is insufficient, a weighted combination of the first two or more principal components can be used. In other implementations, factor analysis, partial least squares, autoencoders, cluster fusion, or weighted composite indices can be used to replace principal component analysis.

[0076] The significance of constructing the aforementioned comprehensive index lies in compressing multidimensional information on influent concentration, hydraulic load, and component structure into a traceable and comparable comprehensive index. Compared to using single indicators such as COD, TN, or water volume, the comprehensive index can more completely reflect the systematic changes in the influent status under integrated plant and grid operation, providing a foundation for identifying subsequent stage transitions.

[0077] Step S204: Based on the trend of changes in comprehensive indicators over time, identify whether the water intake conditions have undergone phased changes. If phased changes have occurred, divide the operation into different phases.

[0078] Specifically, step S204 above includes: Step S2041: The comprehensive indicators are arranged into a sequence in chronological order. The sequence is then broken down. When a break is detected, it is determined that the water intake condition has changed in stages.

[0079] Breakpoint detection refers to the statistical analysis of a time-series of comprehensive indicators to identify statistically significant points of change. These points represent the locations in the sequence where the distribution or trend of data undergoes a substantial shift, corresponding to different operational stages. In this step, the comprehensive indicators are arranged in chronological order to form a time series, with each time point corresponding to a comprehensive indicator value. When the level of the comprehensive indicator changes significantly and persistently before and after a certain position in the series, that position is considered a breakpoint.

[0080] In one alternative implementation, daily-scale composite indices are aggregated into monthly-scale sequences to reduce the impact of short-term fluctuations and single-day outliers on stage identification. Subsequently, candidate breakpoints are searched while satisfying the minimum stage length constraint, and the optimal breakpoint position is selected based on information criteria.

[0081] Preferably, the minimum BIC is used as the criterion for determining the optimal breakpoint. Different candidate breakpoint schemes are compared, and the scheme with the lowest BIC value is selected as the stage division result.

[0082] In practical implementation, the identification process of phased changes (i.e., phase transition identification) is not limited to the BIC candidate breakpoint search method. Other methods include Bai-Perron multiple breakpoint testing, the PELT algorithm, the CUSUM test, or the Bayesian change point model. Among these, the BIC candidate breakpoint search is suitable for scenarios with moderate sample sizes and where maintaining model interpretability is crucial; the PELT algorithm is suitable for rapid identification of long-term series; and the Bayesian change point model can provide breakpoint uncertainty, but its computational complexity is high. Those skilled in the art can flexibly choose the appropriate breakpoint detection method based on the actual application scenario and computational resources.

[0083] Step S2042: Divide the sequence into different running stages according to the breakpoints.

[0084] For example, after obtaining the optimal breakpoint, if there is only one breakpoint, the comprehensive index sequence is divided into a pre-breakpoint stage and a post-breakpoint stage, using this breakpoint as the boundary of the stage change; if there are two breakpoints, it can be divided into a first stage, a second stage, a third stage, and so on. Specifically, different stages can be divided and different stage names can be set according to the specific application scenario and the number of breakpoints. This embodiment does not impose any restrictions on this.

[0085] The significance of the aforementioned phase division lies in identifying whether the influent operating conditions of a wastewater treatment plant have transitioned from one operational state to another. During the integrated operation of the plant and network, changes in pipeline network improvement, external water control, joint scheduling, the efficiency of centralized wastewater collection, and rainfall disturbances can all lead to phased changes in the influent status. If phase identification is not performed and a uniform model covering all time periods is directly adopted, it is easy to mask the differences in operating conditions between different stages, resulting in biased assessment results.

[0086] Step S205: Construct a first model that shares the same response curve for each operating stage and a second model that has its own response curve for each operating stage. Compare the fitting effects of the first model and the second model to determine whether the response relationship of operating performance with changes in influent conditions under different operating stages has been reconstructed.

[0087] Specifically, step S205 includes: Step S2051: Construct a first model that shares the same response curve for each running stage, and a second model that has its own response curve for each running stage.

[0088] The first model assumes that all operating stages share the same response relationship, which is used to characterize the overall pattern of performance level changes with influent conditions. This pattern remains consistent across different stages, meaning that stage changes can only lead to an overall higher or lower performance level, but the shape of the response curve remains unchanged.

[0089] The second model allows for different response relationships in each operational phase. It is used to characterize the fact that the performance level may vary substantially between different phases as the influent conditions change, meaning that the shape of the response curve may change.

[0090] In one optional implementation, a generalized additive model (GAM model) is used as the nonlinear response model to characterize the nonlinear response relationship of operational performance indicators to key influent variables. Its basic form is as follows: ; in, The performance indicators for the t-th period can be the electricity consumption reduction per unit of COD, the electricity consumption reduction per unit of NH3-N, or the carbon emission intensity of pollutant reduction. For the intercept term; The core influent variable can be the influent COD concentration, influent NH3-N concentration, or other key influent indicators; For the smoothed response function corresponding to the core influent variable; The r-th control variable may include meteorological variables, hydraulic load rate variables, physicochemical ratio variables, or process operation variables; Here is the function form corresponding to the control variables; q is the number of control variables; This is the random error term.

[0091] In one optional implementation, the above-described generalized additive model uses operational performance indicators as response variables, assumes that the response variables follow a normal distribution, and employs an identity link function. The smoothing function uses thin-plate regression splines, and the smoothing parameters are selected using REML. Control variables include one or more of meteorological variables (such as rainfall and temperature) and hydraulic loading rate variables (such as hydraulic loading rate). Physicochemical ratio variables (such as COD / TN, SS / COD, etc.) can be included in the control variables or used as independent explanatory variables. After model fitting, the model fit can be tested using explained bias, adjusted R², and residual distribution. Explained bias reflects the degree to which the model explains the variation in the response variables; a higher explained bias indicates a better model fit.

[0092] In practical implementation, the construction of nonlinear response models is not limited to generalized additive models; random forests, gradient boosting trees, support vector regression, neural networks, or piecewise regression models can also be used. Among these, generalized additive models can simultaneously address nonlinear identification and response curve interpretation, making them suitable for analyzing the operating mechanisms of wastewater treatment plants; machine learning models have strong predictive capabilities but weaker interpretability. Those skilled in the art can flexibly choose the appropriate model based on the actual application scenario and the need for model interpretability.

[0093] In one embodiment, when the energy consumption per unit of COD reduction and the carbon emission intensity of pollutant reduction are used as response variables, the core influent variable is preferably the influent COD concentration; when the energy consumption per unit of NH3-N reduction is used as the response variable, the core influent variable is preferably the influent NH3-N concentration.

[0094] Based on the above generalized additive model, the marginal convergence interval can be identified by the slope of the response curve or the marginal rate of change, which can be used to determine whether the improvement in operating performance slows down after further increasing the influent concentration or load.

[0095] In practical implementation, the identification of the marginal convergence interval can employ the first derivative threshold method, curvature change method, piecewise linear fitting method, or plateau period identification method. Among these, the first derivative threshold method directly reflects the rate of change of the response curve and is suitable for identifying the marginal improvement interval of energy consumption or carbon emission intensity as a function of influent concentration. Those skilled in the art can flexibly choose the appropriate method based on actual needs.

[0096] The significance of the aforementioned nonlinear response model lies in identifying the nonlinear relationship between operational performance and influent conditions. The unit energy consumption and carbon emission intensity of a wastewater treatment system usually do not change linearly with the influent concentration, but may have a rapid decline zone, a slow decline zone, or a plateau zone. By modeling the nonlinear response, the operational performance improvement zone and the marginal convergence zone can be identified, providing a basis for optimizing low-carbon operation.

[0097] Based on this, the first model assumes that each operational stage shares the same response relationship, meaning that stage changes can only lead to an overall higher or lower performance level, but the shape of the response curve remains unchanged.

[0098] In one alternative implementation, the first model employs an additive generalized additive model, which allows different stages to have different overall performance levels, but assumes that the stages share the same influent response curve.

[0099] The second model allows for different response relationships at each operational stage, meaning the shape of the response curve may change between different stages. In one optional implementation, the second model employs an interactive generalized additive model, which allows different influent response curves at different stages, as follows: ; Where g(t) represents the operating phase of time period t; These are the main effect parameters for the corresponding stage; Let g(t) be the influent condition response function corresponding to the g(t) stage.

[0100] In practical implementation, the response function reconstruction discrimination is not limited to comparing additive and interactive models; it can also employ methods such as phased independent modeling, sliding window modeling, or online learning models. Among these, comparing additive and interactive models can directly distinguish between performance level deviations and changes in response curve shape, making it suitable for generating auditable and interpretable operational evaluation results. This embodiment preferably uses a comparison between additive generalized additive models and interactive generalized additive models.

[0101] Step S2052: Compare the fitting effects of the first model and the second model. If the fitting effect of the second model is better than that of the first model, it is determined that the response relationship has been reconstructed.

[0102] Among them, the fitting effect refers to the degree to which the model fits the actual data, reflecting the model's ability to capture the changes in operating performance with water inflow conditions. The better the fitting effect, the more accurately the model can reflect the true response relationship.

[0103] In one alternative implementation, the comparison of fit can be made using any one of the following as a quantitative criterion: information criterion comparison (such as AIC (Akaike Information Criterion) or BIC (Bayesian Information Criterion)), cross-validation error comparison, likelihood ratio test, or stage interaction term significance test.

[0104] In one optional implementation, the first model is an additive generalized additive model, and the second model is an interactive generalized additive model. The AIC (Akaike Information Criterion) is used as the primary evaluation index for model fit; a lower AIC value indicates a better model fit. If the AIC value of the interactive generalized additive model is lower than that of the additive generalized additive model (i.e., ΔAIC < 0), and the p-value of the significance test for the stage interaction term is less than 0.05, then the interactive generalized additive model is considered to have a better fit than the additive generalized additive model, and the response relationship is reconstructed. If the AIC value of the interactive generalized additive model is not lower than that of the additive generalized additive model, or the stage interaction term does not meet the preset significance condition (p ≥ 0.05), then the response relationship is considered not to have been reconstructed. In other implementations, cross-validation error or explanatory bias can also be used as evaluation indices for model fit.

[0105] The above comparison rules can objectively distinguish between two different situations: whether the stage change only leads to an overall shift in performance level (the first model holds true), or whether the response relationship itself undergoes a change in form (the second model holds true), thus providing a statistically meaningful objective basis for whether to establish evaluation benchmarks in stages.

[0106] The significance of the above reconstruction judgment lies in determining whether the response relationship between influent conditions and operational performance has substantially changed under different operational stages. If the second model, which allows different response curves for each stage, is significantly better than the first model, which assumes that each stage shares the same response curve, it indicates that the same influent conditions correspond to different operational performance levels under different stages, meaning the shape of the response relationship has changed. This judgment can distinguish between two different scenarios: the first is that stage changes only lead to an overall increase or decrease in performance level, meaning the shape of the response curve has not changed, and this can be evaluated by unifying the response relationship and adding the stage main effect; the second is that the same influent conditions correspond to different operational performances under different stages, meaning the shape of the response curve has changed, and it is necessary to re-establish the operational performance evaluation benchmark according to the stages. This step is the key link that distinguishes this invention from traditional static operational performance evaluation methods.

[0107] Step S206: If a reconfiguration is determined to have occurred, the corresponding operational performance evaluation benchmarks are determined according to different operational stages, and the evaluation results and correction suggestions are output.

[0108] In this step, corresponding operational performance evaluation benchmarks are determined according to different operational stages, including: Step a1: Obtain operational performance indicator data for each operational phase; Step a2: Using data as a sample, establish response curves for the operational performance of each stage as a function of influent conditions, and use the response curves as the benchmark for operational performance evaluation of the corresponding stage. The benchmark for operational performance evaluation includes the benchmark value of power consumption per unit of pollutant reduction under different operational stages. Power consumption per unit of pollutant reduction is used to characterize the amount of power consumed to treat a unit of pollutant, and is determined based on the power consumption, treated water volume, and the difference in pollutant concentration between the influent and effluent in the corresponding time period.

[0109] The electricity consumption within a corresponding time period refers to the total electricity consumed by the wastewater treatment plant within the same time period when calculating the electricity consumption per unit of pollutant reduction. The treated water volume refers to the actual volume of water treated by the wastewater treatment plant within the same time period. The pollutant concentration difference between influent and effluent refers to the difference in the concentration of a certain pollutant between the influent and effluent within the same time period, reflecting the amount of pollutant reduction per unit volume of water during that period.

[0110] In one alternative implementation, when no response function reconstruction is identified, a unified response function is used for performance evaluation; when response function reconstruction is identified, performance evaluation benchmarks are established for different stages, and the actual performance is compared with the expected performance for the corresponding stage, and deviation diagnosis and correction suggestions are output.

[0111] In one optional implementation, after determining the corresponding operational performance evaluation benchmark, the actual operational performance is evaluated based on the determined benchmark. Specifically: Substitute the current influent conditions into the response curve of the corresponding stage to calculate the expected performance value; The evaluation benchmark interval is determined based on the prediction interval or residual distribution of the response model at this stage. The actual performance is compared with the expected performance to obtain the deviation value; or it is determined whether the actual performance exceeds the evaluation benchmark range, and if it does, it is judged as a deviation. The evaluation results are output based on the degree of deviation.

[0112] The output evaluation results and correction suggestions may include one or more of the following: comprehensive index time series, stage transition breakpoints, stage division results, key operating condition characteristics of each stage, nonlinear response curves of operating performance, marginal convergence points or optimized response intervals, comparison results between the first model and the second model, response function reconstruction discrimination results, staged operating performance evaluation benchmarks, corrected performance evaluation results, and operational anomaly warnings and control suggestions.

[0113] Specifically, when the electricity consumption per unit of COD reduction is higher than the reasonable range for the current stage, it can prompt the verification of aeration control, booster pump operation, sludge age, load matching, or chemical dosing strategies; when the carbon emission intensity of pollutant reduction is higher than the reasonable range for the current stage, it can prompt the verification of electricity consumption structure, chemical dosing, sludge treatment, photovoltaic substitution, biogas utilization, or reclaimed water utilization.

[0114] The significance of the above assessment results and correction recommendations lies in transforming the model identification results into dynamic evaluation benchmarks, anomaly diagnosis results, and phased correction recommendations that can be used for operation management. This avoids using a single static model to evaluate all operational stages and makes the performance evaluation of wastewater treatment plants more adaptable to the dynamically changing influent conditions under the background of plant-network integration.

[0115] It should be noted that the calibration mentioned in this invention refers to the phased calibration of the operational performance evaluation benchmark. That is, corresponding operational performance evaluation benchmarks are determined according to different operational stages to ensure that the evaluation standards match the actual influent operating conditions, rather than the automatic control of wastewater treatment plant process operating parameters (such as aeration rate, chemical dosage, sludge discharge, etc.). If the operational performance deviates from the reasonable range of the current stage, the output verification directions (such as aeration control, booster pump operation, chemical dosage, etc.) are used as auxiliary diagnostic prompts to guide operation and management personnel in manual verification and control decisions, rather than automatically generating specific process control instructions.

[0116] In an optional implementation, this embodiment further includes a step of visualizing the dynamic evaluation results, specifically including: Based on the comprehensive indicators, stage division results, response curves, reconstruction judgment results, and evaluation results obtained from the above steps, visual display content is generated. The visual display content includes one or more of the following: (1) Stage division diagram: used to show the trend of the comprehensive index changing over time and the location of the breakpoints of each stage, and to intuitively present the stage transition of the water intake condition. (2) Response curve diagram: used to show the nonlinear response curve of the operating performance under each operating stage as the influent conditions change, including the shape of the response curve and the comparison of differences between stages; (3) Partial effect curve: used to show the independent impact of the core influent variable on the operating performance while keeping other variables constant, and to reflect the marginal impact of different influent conditions on the operating performance at each stage; (4) Statistical comparison chart: used to show the comparison of the fitting effect of the additive model and the interaction model, including the comparison of statistical indicators such as information criterion value and model interpretation bias; (5) Evaluation results charts and reports: used to display the phased performance evaluation results, the degree of deviation between actual performance and phase evaluation benchmark, early warning of operational anomalies and correction suggestions.

[0117] The above visualizations can intuitively present the changes in the influent operating conditions of the wastewater treatment plant, the nonlinear response law of its operating performance, and the basis for phased corrections, making it easier for operation and management personnel to quickly grasp the system's operating status and formulate targeted control strategies.

[0118] The core of this invention does not lie in simply calculating a certain performance indicator, but rather in: (1) Identify whether there are phased changes in the water intake conditions; (2) Identify whether the relationship between influent conditions and operational performance remains stable at different stages; (3) When the relationship changes, the original static performance evaluation benchmark is dynamically corrected, that is, the applicable performance evaluation benchmark for each stage is re-determined based on the stage identification results, so as to realize the staged update of the evaluation standard.

[0119] Therefore, this invention can overcome the limitation of the default evaluation standard being constant in the prior art, and is more suitable for the operation evaluation scenario of sewage treatment plants with continuously changing influent conditions and nonlinear and phased system responses in the context of plant-network integration.

[0120] This invention is implemented in the following order: acquiring operational data, preprocessing, constructing a set of influent operating condition indicators and a set of operational performance indicators, constructing comprehensive indicators, identifying phased changes and dividing the operation into stages, determining whether the response relationship needs to be reconstructed, and determining evaluation benchmarks in stages and outputting evaluation results and correction suggestions. This process can be used for single-plant integrated wastewater treatment plants, and can also be extended to the dynamic evaluation of operational performance of multi-plant, multi-region, or city-level smart water management systems.

[0121] The method for dynamic evaluation and correction of wastewater treatment plant operation performance provided in this embodiment has the following beneficial effects: First, it improves the completeness and accuracy of influent condition identification. This invention no longer relies solely on single indicators such as influent COD, TN, or treated water volume. Instead, it integrates influent pollutant concentration, hydraulic loading rate, and physicochemical ratios such as COD / TN, TN / TP, SS / COD, and BOD5 / COD to construct a comprehensive influent index. This index can simultaneously reflect changes in influent concentration, water volume loading, and component structure, avoiding biases caused by evaluation based on a single indicator.

[0122] Second, it can identify phased transitions in water intake conditions. By performing time series analysis on comprehensive water intake indicators, combined with candidate breakpoint search and information criterion comparison, this invention identifies whether water intake conditions undergo phased changes during the integrated operation of the plant and grid, and divides different operating phases accordingly. This avoids misjudging the entire operating period as the same stable state, and improves the pertinence of subsequent operational performance evaluations.

[0123] Third, it can determine whether the operational performance response relationship has been restructured. By comparing the fitting effects of the first model and the second model, this invention distinguishes between two situations: an overall shift in performance level and a change in the shape of the response curve. When the second model is better than the first model, it can be determined that the response relationship between influent conditions and operational performance has been restructured at different stages, thereby avoiding the direct extrapolation of the concentration-performance relationship established in the early stages to subsequent stages.

[0124] Fourth, it supports phased operational performance correction and low-carbon operation optimization. When a phased reconstruction of the response function is identified, this invention can output operational performance response curves, operational performance evaluation benchmarks, marginal convergence intervals, and correction suggestions according to different stages. This enables the transformation of operational performance evaluation from static benchmarking to dynamic correction, providing data support for energy conservation, emission reduction, carbon reduction, and plant-network coordinated scheduling in wastewater treatment plants.

[0125] In this embodiment, the method of the present invention is applied and verified using a second-phase project of a municipal wastewater treatment plant that implements integrated plant and network operation.

[0126] First, daily operational and meteorological data for the plant from 2022 to 2024 were acquired. Operational data included influent COD, TN, TP, SS, BOD5, NH3-N, treated water volume, power consumption, and chemical dosage; meteorological data included daily rainfall, daily average temperature, daily maximum temperature, and daily minimum temperature. The method of this invention first performs time alignment, outlier removal, missing value processing, and unit standardization on the above data to form a standardized operational database that can be used for subsequent analysis.

[0127] After data preprocessing, the method of this invention constructs an influent operating condition index set based on influent water quality, quantity, and physicochemical ratios. The influent operating condition index set includes influent COD, TN, TP, SS, BOD5, NH3-N, hydraulic loading rate, and 11 indicators such as COD / TN, TN / TP, SS / COD, and BOD5 / COD. This invention standardizes the above influent operating condition indicators and uses principal component analysis to extract the first principal component as a comprehensive influent index to characterize the overall change status of influent conditions at wastewater treatment plants. The first principal component loading and explanatory power, the monthly time series discontinuities of the first principal component, the distribution of the first principal component over three annual operating periods, and the annual operating period effects after controlling for environmental variables are as follows: Figure 3 , Figure 4 , Figure 5 , Figure 6 As shown in the figure, PC1 is the first principal component extracted by principal component analysis, with a variance explained of 37.06%. Although PC1 does not cover all the original index information, its loading structure is clear, its differences are significant during the operation period from 2022 to 2024, and its changes are consistent with those of individual indicators (positive loading with influent COD, TP, SS, COD / TN, BOD5, and TN, and negative loading with TN / TP, BOD5 / COD, and SS / COD), thus providing a reasonable basis for its use as a comprehensive index. Figure 3 The PC1 load and interpretation diagram shows the PC1 load coefficients corresponding to 11 influent operating conditions, intuitively reflecting the contribution of each individual influent indicator to the overall influent index. Figure 4For the PC1 monthly sequence and breakpoint identification diagram, the time series change curve of the comprehensive influent index was plotted on a monthly scale. The breakpoint of the influent condition transition was identified as 2023-06 by the BIC criterion. Figure 5 The figure shows the overall distribution differences of comprehensive influent indicators for PC1 during the three operational periods of 2022 (baseline period), 2023 (jump period), and 2024 (adjustment period). The overall operating conditions differ significantly between representative stages; Figure 6 To control for environmental variables, the annual operating period effect diagram was created by quantifying the overall deviation of the comprehensive inflow index in 2023 and 2024 relative to the 2022 baseline period after removing environmental disturbance variables such as rainfall and temperature. This further corroborates the existence of a phased shift in inflow conditions.

[0128] Subsequently, the daily-scale influent comprehensive indicators were aggregated into a monthly-scale sequence, and the candidate breakpoint search and BIC criterion were used to identify the stage transition locations. The identification results showed that the plant's influent operating conditions underwent a stage transition in 2023. Based on this breakpoint, the operational period was statistically divided into a pre-breakpoint stage and a post-breakpoint stage, indicating that the wastewater treatment plant's influent status did not remain stable throughout the study period but rather experienced identifiable structural changes. To further illustrate the annual evolution of the operational status before, during, and after the breakpoint, this embodiment sets three annual operational periods according to the natural year: the 2022 baseline period, the 2023 transition period, and the 2024 adjustment period. The influent comprehensive indicators, key influent indicators, and operational performance response relationships for each annual operational period were compared. To further verify the engineering implications of the stage division results, this invention simultaneously outputs monthly time-series curves and staged distribution box plots for four key influent pollutant indicators: COD, TP, SS, and TN. The results are shown below. Figure 7 , Figure 8 , Figure 9 , Figure 10 , Figure 11 , Figure 12 , Figure 13 , Figure 14 As shown. Among them, Figure 7 , Figure 8 , Figure 9 , Figure 10 The graphs show the monthly concentration changes of COD, TP, SS, and TN in the influent. The vertical dashed line marks the breakpoint of the operating conditions in June 2023. The three-color background areas correspond to the baseline period in 2022, the transition period in 2023, and the adjustment period in 2024, respectively, which intuitively reflect the fluctuation pattern of the concentration of each pollutant over time. During the transition period in 2023, the concentrations of all four pollutants showed significant peaks. Figure 11 , Figure 12 , Figure 13 , Figure 14The following are box plots showing the concentration distribution of influent COD, TP, SS, and TN during three annual operating periods. The numbers marked on the plots are the average values ​​of the indicators for each period. The concentrations of the indicators representing different stages showed highly statistically significant differences, while NS represented no significant statistical differences between stages. It is evident that the average values ​​of influent COD, TP, and SS during the 2023 transition period were significantly higher than those during the 2022 baseline period and the 2024 adjustment period, with highly significant differences between annual operating periods. Influent TN showed highly significant differences only between the 2022 and 2023 stages, with no significant difference between the 2023 and 2024 stages. These stage-specific differences in the four key influent indicators corroborate the fact that the influent operating condition stage transitions identified by the PC1 comprehensive index are supported by actual water quality data.

[0129] After identifying the influent operating conditions, operational performance indicators were further calculated. These indicators included power consumption per unit of COD reduction, power consumption per unit of NH3-N reduction, and carbon emission intensity for pollutant reduction. Power consumption per unit of COD reduction characterized the power efficiency of the organic matter removal process, power consumption per unit of NH3-N reduction characterized the energy efficiency related to nitrogen removal, and carbon emission intensity for pollutant reduction characterized the carbon emission level corresponding to a unit of pollutant reduction. Generalized additive models were established for power consumption per unit of COD reduction and carbon emission intensity for pollutant reduction as response variables, respectively, based on the influent COD concentration. The nonlinear response results of these models are shown below. Figure 15 , Figure 16 As shown. Figure 15 The curve shows the smoothing effect of energy consumption per unit of COD reduction as a function of influent COD concentration. The red solid line represents the smoothing effect of the GAM model, and the pink filled area represents the 95% confidence interval (95% CI). The model has an explained bias of 90.6% and an AIC value of -49.2. Within the data coverage of this embodiment, the marginal convergence point was identified at an influent COD concentration of 761.0 mg / L. It can be seen that as the influent COD concentration increases, the energy consumption per unit of COD reduction decreases rapidly. After exceeding 761.0 mg / L, the improvement in energy consumption slows down significantly, entering the marginal convergence interval. It should be noted that this convergence point is located in the high concentration range of the sample data distribution, and its applicability is limited to the data coverage of this embodiment. Applications outside this range need to be re-evaluated based on actual data. Figure 16The curves represent the smoothing effect of pollutant carbon emission reduction intensity as a function of influent COD concentration. The solid green line represents the smoothing effect, and the light green filled area represents the 95% confidence interval (95% CI). The model explains 85.4% of the bias, with an AIC value of 329.5. The marginal convergence point corresponds to an influent COD concentration of 588.0 mg / L. With increasing influent COD, the carbon emission reduction intensity per unit of pollutant continuously decreases, and the carbon reduction benefit tends to plateau after the influent COD exceeds 588.0 mg / L. Both sets of curves jointly demonstrate a significant nonlinear relationship between wastewater treatment plant energy consumption, carbon emission performance, and influent COD concentration, with marginal convergence characteristics observed in the high influent concentration range.

[0130] Furthermore, for three types of operational performance indicators—electricity consumption per unit of COD reduction, electricity consumption per unit of NH3-N reduction, and carbon emission intensity of pollutant reduction—an additive generalized additive model (additive GAM model) as the first model and an interactive generalized additive model (interactive GAM model) as the second model were constructed for comparative analysis. The comparative statistical results of the two types of models are as follows: Figure 17 , Figure 18 As shown. Figure 17 The bar chart shows the explanatory biases of the GAM models for the three performance indicators. The explanatory biases of the models for unit COD reduction power consumption, unit NH3-N reduction power consumption, and pollutant reduction carbon emission intensity are 90.6%, 94.3%, and 85.4%, respectively. Each model can explain more than 85% of the data variation, proving that the generalized additive model can effectively characterize the nonlinear correlation between water intake conditions and operational performance. Figure 18 The bar chart shows the ΔAIC of the interactive GAM model relative to the additive GAM model. ΔAIC is defined as the AIC of the interactive GAM model minus the AIC of the additive GAM model. The chart indicates that ΔAIC < 0 represents a better fit from the interactive GAM model. The ΔAIC values ​​for the three indices are -200.6, -41.1, and -57.2, all less than 0. This indicates that for any operational performance indicator, the interactive GAM model, which allows for independent response curves at each stage, significantly outperforms the additive GAM model, which shares the same response curve across all stages. Combined with the significance test results of the model interaction terms, it can be determined that the nonlinear response relationship between influent conditions and operational performance undergoes a phased reconstruction at different operational stages. Therefore, it can be confirmed that the non-overlapping stage curves are not merely a general shift in performance levels, but rather a substantial change in the shape of the response curve itself; that is, the same influent conditions correspond to different operational performance response patterns at different operational stages.

[0131] After determining that the response function has undergone phased reconstruction, to further visually demonstrate the morphological differences in the response relationship at different stages, based on the interactive generalized additive model (interactive GAM model), the performance partial effect curves and phased correction baselines for the 2022 baseline period, the 2023 transition period, and the 2024 adjustment period are output respectively. The phased partial effect results for the three types of operational performance indicators are as follows: Figure 19 , Figure 20 , Figure 21 As shown. Figure 19 The partial effect curves of power consumption per unit COD reduction as a function of influent COD concentration are shown for different operating stages. The blue, orange, and green curves correspond to the baseline period in 2022, the transition period in 2023, and the adjustment period in 2024, respectively. Figure 20 The partial effect curves of pollutant carbon emission reduction intensity as a function of influent COD concentration at different operational stages are shown. Each color curve corresponds to a different stage. Figure 19 Maintain consistency; Figure 21 The partial effect curves show the change in power consumption per unit of NH3-N reduction with the influent NH3-N concentration at different operating stages. Each color curve corresponds to a different stage. Figure 19 Maintaining consistency is crucial. As seen from the three sets of partial effect curves, for the same influent conditions, there are significant differences in the operational performance values ​​corresponding to the 2022 baseline period, the 2023 transition period, and the 2024 adjustment period. The curves for the three periods do not overlap, intuitively demonstrating the annual reconfiguration of the nonlinear response relationship between influent conditions and operational performance. This confirms that a unified and fixed response benchmark is not suitable for evaluating the operational performance of a wastewater treatment plant throughout its entire lifecycle; instead, dynamic operational performance evaluation benchmarks need to be established for each identified operational stage. The aforementioned partial effect curves are a visual representation of the model comparison results, used to intuitively present the specific form of reconfiguration—that is, the differentiated patterns of different operational performance levels corresponding to the same influent conditions at different stages—rather than serving as an independent basis for reconfiguration determination.

[0132] Based on the above results, this embodiment ultimately outputs the plant's dynamic evaluation results and phased correction suggestions, including the influent operating condition phase division results, key influent indicator change characteristics, nonlinear response curve of operating performance, marginal convergence interval, model comparison results, response function reconstruction discrimination results, and phased operating performance correction baseline. When subsequent operating data is input, the method of this invention first determines the operating phase to which the current period belongs, then calls the response function of the corresponding phase to calculate the expected performance level, and compares the actual performance with the expected performance. If the actual unit energy consumption or pollutant reduction carbon emission intensity is significantly higher than the reasonable range of the current phase, an anomaly warning is output, and corrections are suggested from aspects such as aeration control, booster pump operation, load matching, chemical dosing, sludge treatment, photovoltaic substitution, biogas utilization, or reclaimed water utilization.

[0133] Therefore, in this embodiment, the method of the present invention can transform daily operating data of a wastewater treatment plant into influent condition stage identification results, operating performance response functions, and staged operating performance evaluation benchmarks. When the integrated operation of the plant and network leads to staged changes in the influent status, the present invention can identify whether the original concentration-performance relationship is still applicable, and output new staged evaluation and correction results when the response function is reconstructed, thereby improving the accuracy, adaptability, and operability of wastewater treatment plant operating performance evaluation.

[0134] This embodiment also provides a device for dynamic evaluation and correction of the operating performance of a wastewater treatment plant. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0135] This embodiment provides a device for dynamic evaluation and correction of the operational performance of a wastewater treatment plant, such as... Figure 22 As shown, it includes: The data acquisition module 2201 is used to acquire the operating data of the wastewater treatment plant; The indicator construction module 2202 is used to construct a set of water intake condition indicators based on the operating data. The set of water intake condition indicators includes multi-dimensional water intake condition indicators. The comprehensive characterization module 2203 is used to construct a comprehensive index to characterize the overall state of the water intake based on the water intake condition index set; The phase identification module 2204 is used to identify whether the water intake condition has changed in stages based on the trend of changes in comprehensive indicators over time. If a stage change occurs, it divides the operation into different stages. The reconstruction discrimination module 2205 is used to construct a first model that shares the same response curve for each operating stage and a second model that has its own response curve for each operating stage. The fitting effect of the first model and the second model is compared to determine whether the response relationship of the operating performance with the change of water inlet conditions under different operating stages has been reconstructed. The calibration output module 2206 is used to determine the corresponding operational performance evaluation benchmarks according to different operational stages if a reconfiguration is determined to have occurred, and to output the evaluation results and calibration suggestions.

[0136] The wastewater treatment plant operation performance dynamic evaluation and correction device provided in this embodiment of the invention can execute the wastewater treatment plant operation performance dynamic evaluation and correction method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method. Further functional descriptions of the above modules and units are the same as in the corresponding embodiments described above, and will not be repeated here.

[0137] Figure 23 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0138] The following is a detailed reference. Figure 23 This diagram illustrates a suitable structural schematic for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 2301, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 2302 or a program loaded from memory 2308 into random access memory (RAM) 2303. The RAM 2303 also stores various programs and data required for the operation of the electronic device. The processor 2301, ROM 2302, and RAM 2303 are interconnected via a bus 2304. An input / output (I / O) interface 2305 is also connected to the bus 2304.

[0139] Typically, the following devices can be connected to I / O interface 2305: input devices 2306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 2307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 2308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 2309. Communication device 2309 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 23 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0140] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 2309, or installed from memory 2308, or installed from ROM 2302. When the computer program is executed by processor 2301, it performs the functions defined in the wastewater treatment plant operation performance dynamic evaluation and correction method of the embodiments of the present invention.

[0141] Figure 23 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0142] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the dynamic evaluation and correction method for the operating performance of wastewater treatment plants shown in the above embodiments is implemented.

[0143] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0144] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for dynamic evaluation and correction of the operational performance of a wastewater treatment plant, characterized in that, The method includes: Obtain operational data from wastewater treatment plants; A set of water intake condition indicators is constructed based on the operational data, and the set of water intake condition indicators includes multi-dimensional water intake condition indicators. Based on the aforementioned set of water intake condition indicators, a comprehensive index is constructed to characterize the overall state of the water intake. Based on the trend of the comprehensive indicators over time, identify whether the water intake conditions have undergone phased changes. If phased changes have occurred, divide the operation into different phases. A first model sharing the same response curve across all operating stages and a second model having their own response curves for each operating stage are constructed. The fitting effects of the first model and the second model are compared to determine whether the response relationship of operating performance with changes in influent conditions under different operating stages has been reconstructed. If a reconfiguration is determined, the corresponding operational performance evaluation benchmarks are determined according to different operational stages, and the evaluation results and correction suggestions are output.

2. The method for dynamic evaluation and correction of wastewater treatment plant operation performance according to claim 1, characterized in that, The operational data includes at least the data required to construct the influent operating condition index set and to determine the operational performance indicators: When the operational performance indicators include power consumption per unit of pollutant reduction, the operational data includes power consumption, treated water volume, and influent and effluent pollutant concentrations; When the operational performance indicators include the intensity of carbon emission reduction for pollutants, the operational data includes carbon emission source data and pollutant reduction amount data.

3. The method for dynamic evaluation and correction of wastewater treatment plant operation performance according to claim 1, characterized in that, The influent operating condition index set is constructed based on the influent pollutant concentration, influent flow rate, and concentration ratio between pollutants in the operating data. It includes influent pollutant concentration index, hydraulic load rate index, and physicochemical ratio index. The physicochemical ratio index includes one or more of COD / TN, TN / TP, SS / COD, and BOD5 / COD.

4. The method for dynamic evaluation and correction of wastewater treatment plant operation performance according to claim 1, characterized in that, The construction of a comprehensive index to characterize the overall state of the influent based on the influent operating condition index set includes: The set of water intake condition indicators is standardized, and the standardized water intake condition indicators are then dimensionality-reduced and fused to obtain the comprehensive indicator.

5. The method for dynamic evaluation and correction of wastewater treatment plant operation performance according to claim 1, characterized in that, The step of identifying whether the water intake conditions have undergone phased changes based on the changing trend of the comprehensive index over time includes: The comprehensive indicators are arranged into a sequence in chronological order. The sequence is then subjected to breakpoint detection. When a breakpoint is detected, it is determined that the water intake condition has undergone a phased change. Based on the breakpoint, the sequence is divided into different operating phases.

6. The method for dynamic evaluation and correction of wastewater treatment plant operation performance according to claim 1, characterized in that, The step of comparing the fitting effects of the first model and the second model to determine whether the response relationship of operating performance to changes in influent conditions under different operating stages has been reconstructed includes: If the second model fits better than the first model, then the response relationship is determined to have been reconstructed.

7. The method for dynamic evaluation and correction of wastewater treatment plant operation performance according to claim 6, characterized in that, The first model is an additive generalized additive model, and the second model is an interactive generalized additive model. The comparison of the fitting effects of the first model and the second model includes: comparing the goodness of fit between the additive generalized additive model and the interactive generalized additive model and the significance of the stage interaction terms. If the interactive generalized additive model is better than the additive generalized additive model and the stage interaction terms reach the preset significance condition, then it is determined that the response relationship has been reconstructed.

8. The method for dynamic evaluation and correction of wastewater treatment plant operation performance according to claim 1, characterized in that, The determination of corresponding operational performance evaluation benchmarks based on different operational stages includes: Obtain operational performance indicator data for each operational phase; Using the data as a sample, a response curve for the operational performance of each stage as a function of the influent conditions is established, and the response curve is used as the benchmark for evaluating the operational performance of the corresponding stage. The operational performance evaluation benchmark includes the benchmark value of power consumption per unit of pollutant reduction under different operating stages. The power consumption per unit of pollutant reduction is used to characterize the amount of power consumed to treat a unit of pollutant, and is determined based on the power consumption, treated water volume, and the difference in pollutant concentration in the influent and effluent during the corresponding time period.

9. A device for dynamic evaluation and correction of the operational performance of a wastewater treatment plant, characterized in that, The device includes: The data acquisition module is used to acquire operational data from the wastewater treatment plant. The indicator construction module is used to construct a set of water intake operating condition indicators based on the operating data. The set of water intake operating condition indicators includes multi-dimensional water intake operating condition indicators. The comprehensive characterization module is used to construct a comprehensive index to characterize the overall state of the water intake based on the water intake condition index set. The phase identification module is used to identify whether the water intake condition has undergone phased changes based on the changing trend of the comprehensive index over time. If phased changes occur, different operating phases are divided. The reconstruction discrimination module is used to construct a first model that shares the same response curve for each operating stage and a second model that has its own response curve for each operating stage. The fitting effect of the first model and the second model is compared to determine whether the response relationship of operating performance with changes in influent conditions under different operating stages has been reconstructed. The correction output module is used to determine the corresponding operational performance evaluation benchmarks according to different operational stages if a reconfiguration is determined to have occurred, and to output the evaluation results and correction suggestions.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the wastewater treatment plant operation performance dynamic evaluation and correction method according to any one of claims 1 to 8.

Images