Water ecological restoration management method and system based on real-time monitoring

Abstract

The invention relates to the field of intelligent management, and particularly discloses a water ecological restoration management method and system based on real-time monitoring, and the method comprises the steps: carrying out the timestamp alignment and filtering smoothing of real-time monitoring data, calculating a dynamic correction parameter through combining with a microbial dynamic parameter set, and precisely predicting a biological natural restoration curve and a lagging response moment. On the basis, an MPC optimization model containing a cost function and physical constraints is constructed, a control increment sequence considering the standard water quality and the optimal cost is generated through numerical solution, and the control increment sequence is converted into a field equipment instruction through dual-scale mapping. Through the mode, time delay in water ecological management is effectively compensated, intelligent coordination and refined regulation and control of the equipment group are realized, and the operation cost of the system is remarkably reduced on the premise of ensuring that the water quality reaches the standard.

Description

Technical Field

[0001] This application relates to the field of intelligent management, and more specifically, to a method and system for water ecological restoration management based on real-time monitoring. Background Technology

[0002] With the increasing complexity of water pollution problems, the restoration and reconstruction of aquatic ecosystems has become a core task of environmental governance. Constructing scientific and efficient water ecological restoration management solutions is of great significance for restoring the self-purification capacity of water bodies, maintaining the ecological balance of watersheds, and ensuring the sustainable use of water resources. Faced with complex water bodies such as large rivers or lakes, relying solely on human experience or static governance strategies is no longer sufficient to meet the needs of precise pollution control. Therefore, the transformation towards data-driven intelligent management has become an inevitable trend in the industry.

[0003] While existing water ecosystem restoration management solutions have achieved a degree of automation in monitoring and control, they still have significant limitations in practical industrial-scale applications. Current technologies primarily employ a single-point control model. This model, when facing the critical transition from "automation" to "intelligent control," exposes problems such as the lag in multi-device collaborative control under spatiotemporal coupling and energy consumption optimization issues. Specifically, existing systems often struggle to uniformly and coordinately command dozens of devices on-site, taking into full account the spatial impact of water flow direction and the time lag required for the effectiveness of chemicals or oxygen. Due to the lack of effective prediction and compensation for the lag characteristics of biochemical reactions, existing solutions fail to minimize operating costs while ensuring water quality meets standards, resulting in high electricity and chemical costs.

[0004] Therefore, there is a need for optimized water ecosystem restoration and management methods based on real-time monitoring. Summary of the Invention

[0005] To address the aforementioned technical issues, this application provides a water ecological restoration management method and system based on real-time monitoring.

[0006] According to one aspect of this application, a water ecosystem restoration and management method based on real-time monitoring is provided, comprising: The acquired raw water body real-time monitoring data stream is timestamped and filtered and smoothed to obtain the water body real-time monitoring vector; Based on the temperature component in the real-time water body monitoring vector and the preset benchmark microbial kinetic parameter set, the kinetic correction parameters for the current moment are determined. The system's initial state vector is obtained by concatenating the real-time water monitoring vector and the dynamic correction parameters at the current moment. Biological effect hysteresis prediction is performed on the initial state vector of the system and the dynamic correction parameters at the current moment to obtain the biological natural repair prediction curve and the predicted biological hysteresis response time. Based on the predicted biological lag response time, the predicted biological natural remediation curve, the current actual water quality and the target water quality, an MPC optimization problem is constructed. The MPC optimization problem consists of a cost function, a system state prediction model and a set of physical constraints for system operation. Numerical optimization is performed on the MPC optimization problem to obtain the optimal control increment sequence; The first action in the optimal control increment sequence is subjected to dual-scale instruction mapping and execution to obtain the execution instructions for the field equipment.

[0007] According to another aspect of this application, a water ecosystem restoration management system based on real-time monitoring is provided, comprising: The timestamp alignment and filtering smoothing module is used to perform timestamp alignment and filtering smoothing on the acquired raw water body real-time monitoring data stream set to obtain the water body real-time monitoring vector; The kinetic correction module is used to determine the kinetic correction parameters at the current moment based on the temperature component in the real-time water body monitoring vector and the preset benchmark microbial kinetic parameter set. The vector splicing module is used to splice the real-time water monitoring vector and the dynamic correction parameters at the current moment to obtain the initial state vector of the system. The biological effect lag prediction module is used to perform biological effect lag prediction on the system's initial state vector and the dynamic correction parameters at the current moment to obtain the biological natural repair prediction curve and the predicted biological lag response time. The MPC optimization problem construction module is used to construct an MPC optimization problem based on the predicted biological lag response time, the predicted biological natural remediation curve, the current actual water quality, and the target water quality. The MPC optimization problem consists of a cost function, a system state prediction model, and a set of physical constraints for system operation. The numerical optimization solution module is used to perform numerical optimization solutions on the MPC optimization problem to obtain the optimal control increment sequence; The dual-scale instruction mapping and execution module is used to perform dual-scale instruction mapping and execution on the first action in the optimal control increment sequence to obtain the execution instructions for the field equipment.

[0008] Compared with existing technologies, the water ecological restoration management method and system based on real-time monitoring provided in this application accurately predicts the biological natural restoration curve and lag response time by aligning and filtering real-time monitoring data with timestamps, and calculating kinetic correction parameters using a microbial kinetic parameter set. Based on this, an MPC optimization model incorporating cost functions and physical constraints is constructed. Numerical solutions generate a control increment sequence that balances water quality compliance with cost optimization, which is then transformed into field equipment commands through dual-scale mapping. This approach effectively compensates for time lags in water ecological governance, enabling intelligent collaboration and refined control of equipment clusters, and significantly reducing system operating costs while ensuring water quality compliance. Attached Figure Description

[0009] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0010] Figure 1 This is a flowchart of a water ecosystem restoration and management method based on real-time monitoring according to an embodiment of this application; Figure 2 This is a schematic diagram of data flow in the water ecological restoration and management method based on real-time monitoring according to an embodiment of this application; Figure 3 This is a block diagram of a water ecological restoration management system based on real-time monitoring according to an embodiment of this application. Detailed Implementation

[0011] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0012] As indicated in this application and claims, unless the context clearly indicates otherwise, the words "a," "an," "and," and / or "the" do not specifically refer to the singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of explicitly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0013] While this application makes various references to certain modules of the systems according to embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The modules described are merely illustrative, and different aspects of the systems and methods may use different modules.

[0014] Flowcharts are used in this application to illustrate the operations performed by the system according to embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, various steps can be processed in reverse order or simultaneously, as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0015] The technical solution of this application proposes a water ecological restoration and management method based on real-time monitoring. Figure 1 This is a flowchart of a water ecological restoration and management method based on real-time monitoring according to an embodiment of this application. Figure 2 This is a system architecture diagram of a water ecosystem restoration and management method based on real-time monitoring, according to an embodiment of this application. Figure 1 and Figure 2 As shown, the water ecological restoration management method based on real-time monitoring according to an embodiment of this application includes the following steps: S1, performing timestamp alignment and filtering smoothing on the acquired original water body real-time monitoring data stream set to obtain a water body real-time monitoring vector; S2, determining the dynamic correction parameters at the current moment based on the temperature component in the water body real-time monitoring vector and a preset set of benchmark microbial dynamic parameters; S3, concatenating the water body real-time monitoring vector and the dynamic correction parameters at the current moment to obtain a system initial state vector; S4, performing biological effect lag prediction on the system initial state vector and the dynamic correction parameters at the current moment to obtain a biological natural restoration prediction curve and a predicted biological lag response time; S5, constructing an MPC optimization problem based on the predicted biological lag response time, the biological natural restoration prediction curve, the current actual water quality, and the target water quality, wherein the MPC optimization problem consists of a cost function, a system state prediction model, and a set of system operation physical constraints; S6, performing numerical optimization to solve the MPC optimization problem to obtain an optimal control increment sequence; S7, performing dual-scale instruction mapping and execution on the first action in the optimal control increment sequence to obtain the execution instructions for the field equipment.

[0016] Specifically, in step S1, the acquired raw water body real-time monitoring data stream set is timestamped and filtered for smoothing to obtain a water body real-time monitoring vector. The raw water body real-time monitoring data stream set refers to the unprocessed time-series data received from the field's underlying sensing devices. Specifically, it includes temperature data streams, pH data streams, and dissolved oxygen data streams with independent acquisition timestamps, reflecting the physical and chemical state of the water body. Considering that the raw water body real-time monitoring data stream set is typically generated asynchronously by multiple sensors (such as temperature sensors, pH sensors, and dissolved oxygen sensors) deployed at different locations within the water body, the timestamps of their data points often cannot be precisely matched due to their respective sampling periods, communication delays, or clock differences. Directly using unaligned data for vectorization or subsequent calculations will introduce time dimension misalignment errors, leading to misjudgments of the system state. Furthermore, factors such as electromagnetic interference in the field environment and fluctuations in the sensors themselves can cause the raw monitoring data to contain high-frequency noise. This noise can interfere with the accuracy of subsequent calculations of microbial kinetic parameters and model predictions, and may even cause unstable actions in the control system. Therefore, in the technical solution of this application, eliminating time deviation through timestamp alignment and suppressing random noise through filtering and smoothing are necessary prerequisites to ensure the accuracy and robustness of the entire management system's decision-making.

[0017] It is worth mentioning that timestamp alignment refers to the process of mapping the aforementioned multi-source heterogeneous time series data to the same standard time grid point, ensuring that vector components at the same moment logically belong to the same physical moment; while filtering and smoothing refers to using digital signal processing algorithms to remove high-frequency noise and outliers from the data, while retaining the true trend of data change.

[0018] In practice, the system first aligns the acquired raw real-time water monitoring data streams with timestamps to obtain an aligned set of real-time water monitoring data streams. Specifically, the system receives raw data streams from different monitoring devices, each containing a timestamp and a monitored value (such as temperature or pH). During this process, the system sets a unified time reference point (e.g., the start time of each control cycle) and an alignment time window. For each type of monitoring parameter (e.g., temperature), the system retrieves data points for that parameter reported by all sensors within the alignment time window. Next, the system uses a linear interpolation algorithm to calculate an aligned value for each parameter at the unified time reference point. Specifically, the unified time reference point is first determined, and then, for each parameter requiring alignment, the two closest valid data points before and after the time reference point are found. Then, based on the proportional position of the time reference point relative to the times of these two known data points, the parameter value at the time reference point is estimated proportionally. In other words, the reference point time is placed within the interval formed by the times of the two preceding and following data points. Its time position is calculated as a percentage, and this percentage is then applied to the numerical difference between the two data points to estimate the value of the reference point. This process ensures that all parameters have valid, synchronized values ​​at the same point in time, forming a preliminary aligned data snapshot.

[0019] Next, after timestamp alignment, the initially aligned data snapshots are filtered to suppress noise. Specifically, a digital low-pass filter, such as a first-order infinite impulse response filter, is used to smooth each aligned parameter sequence (considering its historical data) to preserve the long-term trend of the data while filtering out short-term fluctuations. During this process, based on the idea of ​​exponentially weighted moving average, different weights are given to the current observation and the filtered output values ​​from past times, and a new, smoother output value is generated through weighted averaging. Specifically, for each parameter, the system maintains a historical filtered value. When a new, timestamp-aligned current observation is obtained, the system weights and fuses the current observation with the historical filtered value to generate a new current filtered value. This new filtered value contains information from the latest observation and inherits the smoothing trend of the historical data. A key smoothing coefficient determines the weight of the new observation in the fusion process. This coefficient ranges from zero to one. A larger coefficient indicates that the system relies more heavily on new observations, resulting in weaker smoothing and a faster response to data changes. Conversely, a smaller coefficient indicates that the system depends more on historical data trends, resulting in stronger smoothing and better noise suppression, but a delayed response to real-world changes. Adjusting this coefficient balances noise suppression and signal response speed. After this filtering and smoothing process, each parameter receives a smoothed value. Finally, the filtered and smoothed values ​​of all parameters at a unified time reference point are combined in a predetermined order to obtain the real-time water monitoring vector.

[0020] Specifically, S2 determines the corrected kinetic parameters for the current moment based on the temperature component in the real-time water monitoring vector and a preset set of baseline microbial kinetic parameters. It should be understood that in traditional aquatic ecological restoration management techniques, assessing microbial activity typically relies on single-dimensional kinetic calculations based on the modified Arrhenius equation. This method considers temperature as the sole determining environmental factor affecting microbial growth rate, which is an oversimplified and easily invalidated core assumption in the dynamic real aquatic environment. It ignores the strong nonlinear coupling and inhibitory relationship between the aquatic chemical environment (e.g., pH) and biochemical environment (e.g., dissolved oxygen DO) on microbial activity. Therefore, this method suffers from fundamental defects of model unidimensionality and environmental fragmentation. Specifically, core restoration microbial communities (e.g., nitrifying bacteria) have narrow optimal pH ranges. When the actual pH deviates from this range, even if the temperature is within the optimal range, the enzyme activity of microorganisms will decrease sharply or even become inactive. Simultaneously, for aerobic bacteria, dissolved oxygen itself is a reaction substrate; when its concentration falls below a certain threshold, it becomes a rate-limiting step in the entire biochemical reaction, restricting the actual metabolic rate of microorganisms. Therefore, traditional mechanisms, failing to couple the nonlinear inhibitory effects of two key environmental factors—pH and dissolved oxygen—calculate a growth rate that is merely a potential maximum growth rate, rather than the actual effective growth rate. This leads to significant deviations under complex water quality conditions such as extreme pH or low dissolved oxygen, lacking robustness for industrial applications. To address these shortcomings, this application proposes a dynamic calibration method for the effective growth rate of microorganisms based on the nonlinear inhibition of multidimensional environmental factors. By introducing a computational mechanism that couples the Gaussian inhibition effect of pH and the Monod saturation effect of dissolved oxygen, it generates kinetic correction parameters with high context awareness, ensuring that the system accurately reflects the current self-purification capacity of aquatic organisms.

[0021] In practice, firstly, real-time temperature is extracted from the real-time water monitoring vector and secondly, from the preset set of baseline microbial dynamic parameters. The quasi-maximum specific growth rate and temperature correction factor, based on real-time temperature, The quasi-maximum specific growth rate and temperature correction factor are used to calculate the basic thermodynamic growth rate to obtain the thermodynamically corrected maximum specific growth rate. This process involves extracting the real-time monitored water temperature. And using the preset 20℃ baseline maximum specific growth rate and temperature correction factor Substituting into the classic modified Arrhenius equation for calculation, the specific calculation formula is as follows: in, This represents the maximum specific growth rate after thermodynamic correction. This represents the baseline maximum specific growth rate at a standard temperature of 20°C. This is a dimensionless temperature correction factor, while This refers to the real-time water temperature. This step calculates the ideal activity ceiling that microorganisms can achieve at the current temperature, providing a scalable and standardized starting point for the subsequent introduction of inhibitory effects into the real environment, ensuring the logical completeness of the correction.

[0022] Next, based on real-time pH, real-time dissolved oxygen, optimal pH, pH tolerance width coefficient, and dissolved oxygen half-saturation constant, a comprehensive environmental inhibition coefficient is determined. It should be understood that an idealized model considering only temperature cannot accurately assess microbial activity in real aquatic ecosystems. Therefore, in the technical solution of this application, a correction mechanism that comprehensively reflects the inhibitory effects of pH and dissolved oxygen is introduced to ensure the prediction and control accuracy of the entire management system.

[0023] In this process, firstly, the pH inhibition coefficient is determined based on the real-time pH value, the optimum pH value, and the pH tolerance width coefficient. Then, a Gaussian function is introduced to calculate the pH inhibition coefficient. Specifically, the pH inhibition coefficient is determined using the following formula: in, It is the pH inhibition coefficient. For real-time pH value, The optimal pH value for the target bacterial group, This represents the pH tolerance width coefficient. This Gaussian function accurately simulates the sensitivity of microbial enzyme activity to pH: when the real-time pH value equals the optimum, the function output is 1 (no inhibition); once it deviates to either side, the function value rapidly and non-linearly decays to zero. The purpose and effect of this design is to accurately characterize the near-veto-growth inhibition effect of extreme acidic and alkaline environments on microorganisms. Secondly, based on real-time dissolved oxygen and the dissolved oxygen half-saturation constant, the dissolved oxygen saturation coefficient is determined. Specifically, the dissolved oxygen saturation coefficient is calculated using the Monod equation with the following formula. The formula is: in, It is the dissolved oxygen saturation coefficient. This is the real-time dissolved oxygen concentration. This is the dissolved oxygen half-saturation constant. This step treats dissolved oxygen as a gaseous substrate, and when the oxygen supply is sufficient ( Much larger When oxygen is scarce, the coefficient approaches 1 (unrestricted); while when oxygen is scarce, the coefficient approaches 0, indicating that microbial activity is severely restricted by the hypoxic state. This allows for precise quantification of the metabolic rate of aerobic bacteria at different oxygen concentrations, solving the problem that the original model could not reflect the hypoxic limitation.

[0024] Finally, the dissolved oxygen saturation coefficient and the pH inhibition coefficient are multiplied to obtain the comprehensive environmental inhibition coefficient. Specifically, this process is expressed by the formula: in, This is the final dimensionless comprehensive environmental inhibition coefficient. This multiplicative synthesis reflects the "weakest link" effect among different environmental pressures; that is, if any single environmental factor falls into the severely inhibited range, the final comprehensive coefficient will approach zero. In this way, a highly integrated single index that can reflect the final inhibition intensity under the synergistic effect of multiple factors can be constructed.

[0025] Furthermore, based on the comprehensive environmental inhibition coefficient and the thermodynamically corrected maximum specific growth rate, the kinetic correction parameters for the current moment are generated. That is, the ideal, temperature-dependent upper limit of activity is combined with the real, multi-dimensional environmental inhibition effect to output an effective parameter that can guide control decisions under complex operating conditions. In this process, the thermodynamically corrected maximum specific growth rate is used... With the overall environmental inhibition coefficient The multiplication process can be expressed by the following formula: in, This is the final output, which is the actual effective growth rate dynamically calibrated by multidimensional environmental factors. In this way, the predictive control module of the entire water ecological restoration management system is provided with a core input that can truly reflect the current biological self-purification capacity of the water body, thereby completing the accurate and dynamic calibration of microbial activity.

[0026] In summary, this mechanism overcomes the environmental fragmentation inherent in traditional techniques that use single-dimensional Arrhenius models to assess microbial activity, providing a calibration method that accurately and dynamically reflects the actual effective growth rate of microorganisms under real water quality conditions. Specifically, by creatively introducing a comprehensive environmental inhibition coefficient that couples the Gaussian inhibition effect of pH and the Monod saturation effect of dissolved oxygen, the previously isolated temperature correction model is upgraded into a multi-dimensional nonlinear model capable of dynamically responding to the synergistic effects of water chemistry and the biochemical environment. Thus, the system output is no longer a potential maximum growth rate that would produce significant deviations under complex operating conditions, but rather an actual effective growth rate with high robustness and accuracy. This high-precision parameter significantly improves the accuracy of downstream biodegradation prediction models, enabling the entire aquatic ecosystem restoration management system to make more precise control decisions. This effectively avoids problems such as over-aeration or chemical dosing caused by misjudging biological activity, ultimately achieving the comprehensive technical goals of reducing operating costs, enhancing system stability, and better protecting the ecological balance of aquatic bodies.

[0027] Specifically, in step S3, the real-time water monitoring vector and the current-moment kinetic correction parameters are concatenated to obtain the initial state vector of the system. It should be understood that the real-time water monitoring vector only characterizes the physicochemical properties of the water (such as temperature, pH, and dissolved oxygen), while the kinetic correction parameters characterize the actual metabolic capacity of microorganisms under current environmental stress (i.e., the actual effective growth rate). These two data streams logically belong to state variables and rate parameters, respectively, and are generated asynchronously by different computational modules. Without a unified mathematical encapsulation of both, subsequent differential equation solvers or discrete state-space models will be unable to obtain complete boundary conditions. Therefore, in the technical solution of this application, the real-time water monitoring vector and the current-moment kinetic correction parameters are concatenated to fuse environmental state data and biodynamic characteristic data into a high-dimensional vector, thereby enabling accurate prediction of future water quality evolution trajectories.

[0028] Vector concatenation refers to the dimensional expansion operation performed in a vector space, which involves appending a scalar or vector as a new dimensional element to the end of another vector.

[0029] In practical implementation, firstly, the system needs to confirm the specific content and dimensions of the real-time water monitoring vector. Assume the real-time water monitoring vector is a column vector containing several components, each corresponding to a pre-processed water quality parameter, such as temperature, pH, and dissolved oxygen concentration. The dimension of this vector is denoted as n. Secondly, the system confirms that the current kinetic correction parameter is a scalar, which can be considered a single-element vector with a dimension of 1. Then, the system performs a vector concatenation operation. Specifically, a new empty array or list is created with a length (i.e., dimension) equal to the original monitoring vector's dimension plus one. All elements from the original real-time water monitoring vector are sequentially filled into the first n positions of this new array, according to their original order. Finally, the scalar value of the current kinetic correction parameter is filled into the (n+1)th position (the last position) of the new array. This generates a new vector with expanded dimensions, namely the system's initial state vector. This vector is an augmented vector, mathematically containing both the physical quantities determining the reaction environment and the kinetic quantities determining the reaction rate, serving as a global state representation for subsequent prediction time windows.

[0030] Specifically, in step S4, the biological effect lag prediction is performed on the system's initial state vector and the current kinetic correction parameters to obtain the biological natural remediation prediction curve and the predicted biological lag response time. It should be understood that in actual water ecological restoration engineering, biochemical reactions exhibit significant inertia and lag characteristics. Microorganisms need to undergo an induction phase and a logarithmic growth phase from contact with pollutants and adaptation to the environment to reaching maximum metabolic activity. Therefore, the system needs to be guided by the current kinetic correction parameters, but knowing only the current instantaneous rate is insufficient to support long-term optimization of model predictive control (MPC). If the controller ignores the time difference of biological action, it may misjudge the treatment as ineffective before the microorganisms have entered a full-rate degradation state, leading to excessive dosage of chemicals or over-aeration, resulting in resource waste and secondary pollution. Therefore, in the technical solution of this application, the future pollutant degradation trajectory is extrapolated through numerical simulation, and the specific moment when biological function reaches its peak (i.e., the lag response time) is accurately identified, thereby providing the MPC controller with key constraints on the time dimension to ensure that the control strategy is synchronized with the growth rhythm of the microorganisms.

[0031] In practice, firstly, the system initializes the predicted state based on the initial state vector to obtain an initialized pollutant concentration prediction array and an initialized degradation rate prediction array. During this process, the system extracts the initial pollutant concentration value from the initial state vector and obtains the initial actual effective growth rate from the kinetic correction parameters at the current moment. Subsequently, the system creates two arrays to store the prediction results: a pollutant concentration prediction array and a degradation rate prediction array. During initialization, the initial pollutant concentration value is filled into the first position of the pollutant concentration prediction array, corresponding to the start time of the simulation. Simultaneously, based on the principles of microbial degradation kinetics, the initial degradation rate needs to be calculated through the functional relationship between the initial concentration and the growth rate. Specifically, the degradation rate of pollutants by microorganisms is related to their own growth rate and the current concentration of the pollutants, typically following the saturation kinetic characteristics described by the Monod equation, i.e., the degradation rate increases with increasing concentration, but there is an upper limit. Therefore, the initial degradation rate can be expressed as a function of the initial effective growth rate, the initial pollutant concentration, and a half-saturation constant representing the microorganism's affinity for the pollutant. Filling the calculated initial degradation rate value into the first position of the degradation rate prediction array completes the initialization of the prediction array.

[0032] Next, based on the kinetic correction parameters at the current moment, the system performs iterative forward extrapolation of the pollutant degradation process on the initialized pollutant concentration prediction array and the initialized degradation rate prediction array to obtain the bioremediation prediction curve. During this process, the system sets a simulation step size, for example, 1 hour. In each iteration, the system extrapolates the concentration value for the next time step based on the pollutant concentration and the current degradation rate at the current simulation moment. Its core calculation principle is the law of conservation of mass, which states that the decrease in pollutant concentration per unit time is equal to the rate at which it is degraded by microorganisms. Specifically, the pollutant concentration at the next moment is equal to the concentration at the current moment minus the product of the current degradation rate and the simulation step size. Simultaneously, as the pollutant concentration decreases, the degradation rate also changes. The system needs to update the degradation rate for the next time step based on the new pollutant concentration and the unchanged effective growth rate (this extrapolation assumes unchanged environmental conditions), using the aforementioned degradation rate calculation logic (such as the Monod equation). Then, the newly calculated concentration value and degradation rate value are stored in the corresponding prediction arrays. This iterative process continues until the pollutant concentration falls below a certain preset threshold or the maximum simulation duration is reached. Ultimately, the data sequences of the pollutant concentration prediction array and the degradation rate prediction array changing over time together constitute the bioremediation prediction curve. Here, the bioremediation prediction curve refers to a time series curve reflecting the trend of pollutant concentration changes and the corresponding degradation rate changes over a future period, based on the current system state, under the assumption of no additional external control actions (such as no additional drug administration). In terms of data structure, it is usually represented as a dual-track dataset containing the pollutant concentration prediction array and the degradation rate prediction array.

[0033] Furthermore, based on the simulation step size, the lag response time feature of the bioremediation prediction curve is extracted to obtain the predicted biolag response time. In this process, firstly, the peak rate of the degradation rate prediction array in the bioremediation prediction curve is searched to obtain the maximum value; secondly, the position where the maximum value first appears in the degradation rate prediction array is determined to obtain the peak index; finally, based on the peak index and the simulation step size, the predicted biolag response time is determined. Specifically, the system first scans the entire degradation rate prediction array to find the maximum value, which represents the peak of microbial degradation activity during the prediction period. Next, the system locates the sequence number where the maximum value first appears in the array, i.e., the peak index. Finally, the predicted biolag response time is obtained by converting the peak index into actual time. Here, the predicted biolag response time refers to the time point when the metabolic activity or degradation rate of pollutants by the microbial population reaches its maximum after adapting to the current aquatic environment; this indicator quantifies the system's response delay characteristics.

[0034] Specifically, in step S5, based on the predicted biological lag response time, the predicted natural biological restoration curve, the current actual water quality, and the target water quality, an MPC optimization problem is constructed. This MPC optimization problem consists of a cost function, a system state prediction model, and a set of physical constraints for system operation. It should be understood that in the complex dynamic process of aquatic ecological restoration, there is a significant time lag and nonlinearity between the treatment action and the water quality response. If control is based solely on the current water quality deviation (such as PID control), it often leads to excessive addition of chemicals or over-aeration during the lag period before the biological community is activated, resulting in energy waste; while during the period of explosive biological effects, overshoot occurs due to the failure to reduce intervention in time. Therefore, in the technical solution of this application, a Model Predictive Control (MPC) optimization problem is constructed to utilize this biological evolution information at future moments to pre-imagine the water quality change trajectory on the time axis, thereby finding an optimal control sequence. This approach can fully utilize the natural degradation capacity of organisms to reduce artificial treatment costs while ensuring that the water quality strictly meets the standards at the target time, achieving the best balance between energy consumption and effectiveness.

[0035] The MPC optimization problem is a mathematical programming model aimed at finding the optimal solution for the control variables within a finite prediction time domain. The cost function is a measure of the effectiveness of the control strategy, typically including a water quality deviation penalty (to ensure compliance) and a control energy penalty (to ensure energy conservation). The system state prediction model is a set of mathematical equations describing the evolution of water quality over time. In this application, the scheme creatively integrates the biological natural remediation prediction curve as a natural disturbance term within the system. The set of physical constraints for system operation defines the physical boundaries of the field equipment, such as the maximum frequency range of the aerator, the maximum flow rate of the dosing pump, and water quality non-negativity constraints.

[0036] In practical implementation, firstly, a system state prediction model is established. This model adopts a discrete-time state-space equation form to predict the biological natural repair curve. Specifically, the prediction model is constructed as follows: in, For the first Pollutant concentration at any given time For the first Real-time control inputs (such as aeration increments). The first one extracted from the bio-natural repair prediction curve The rate of natural degradation at any given time. Sampling time, To control the response coefficient of the input.

[0037] Next, based on the predicted biological lag response time A cost function is constructed based on the target water quality. Specifically, weighting coefficients are introduced to balance the trade-off between achieving water quality standards and operating costs. In particular, the system utilizes... To dynamically adjust the weights: before the lag response time, allow appropriate control intensity to assist microorganisms in overcoming the adaptation period; after the lag response time, due to the enhanced biological self-purification capacity, increase the penalty weight on control energy consumption to utilize natural repair forces. The cost function is then expressed as: in, To predict the time domain, For the target water quality, As the water quality error weight, To control energy consumption weight, Let be the cost function.

[0038] Furthermore, a set of physical constraints for system operation is defined, transforming equipment capability limitations into inequality constraints. Specifically, the constraints are as follows: in, and The physical output limits for the on-site actuators are defined by the lower and upper limits, while ensuring that the pollutant concentration is non-negative. The final set of constraints collectively defines the feasible solution space of the optimization problem, ensuring that the optimal control sequence is physically realizable and conforms to the system's safe operation specifications.

[0039] Specifically, S6 involves numerically optimizing the MPC optimization problem to obtain the optimal control increment sequence. It should be understood that the MPC optimization problem is an MPC optimization problem that includes a cost function, a system state prediction model, and a set of physical constraints on system operation; however, it is merely a static mathematical description and has not yet been transformed into specific instructions that can directly drive hardware actions. Therefore, it is necessary to use computer algorithms to find an optimal path in the multidimensional solution space. This path must simultaneously satisfy three stringent conditions: first, strictly adhere to physical constraints (e.g., equipment cannot be overloaded); second, ensure that the predicted future water quality trajectory is as close as possible to the set target; and third, minimize the energy consumption cost of the entire process. Traditional rule-based control cannot handle such time-varying and complex nonlinear relationships. This application uses numerical optimization to transform these biological prior knowledge into specific, time-segmented, precise control strategies, thereby achieving human-machine collaboration and on-demand governance at the mathematical level. It is worth mentioning that numerical optimization refers to the process of using computer algorithms (such as quadratic programming (QP) or nonlinear programming (NLP)) to iteratively calculate a defined mathematical model to find the set of decision variables that minimize the cost function.

[0040] In practice, the MPC optimization problem constructed in the previous steps is transformed into a standard matrix form. If the prediction model is linear and the cost function is quadratic, the system encapsulates it into the standard form of quadratic programming (QP): that is, defining the Hessian matrix H to represent the quadratic terms of the cost function (involving error weights and energy consumption weights), defining the gradient vector f to represent the linear terms, and defining the constraint matrix A and boundary vectors. .

[0041] Then, an embedded optimization solver (such as OSQP, QPOASES, or IPOPT) is invoked. Starting from the current system state, the solver initiates an iterative search process. In each iteration, the algorithm calculates the gradient direction of the objective function and projects or corrects the search direction according to the constraints. The solver continuously adjusts the future control increment sequence within the solution space, attempting to find a specific sequence combination that minimizes the total cost J of the prediction model output under the action of that sequence. This process is repeated until a preset convergence condition (such as the KKT condition) is met or the maximum number of iterations is reached. Once the solution converges, the decision variables extracted by the system are the optimal control increment sequence in the prediction time domain N. The process can be expressed by the following formula: in, Indicating the future At any given moment, this is the optimal control increment that the device should execute. Specifically, although the solver calculates the actions for the next N steps, according to the rolling optimization principle of model predictive control, this step ensures that the entire sequence is globally optimal at the current moment.

[0042] Specifically, in S7, a dual-scale instruction mapping and execution is performed on the first action in the optimal control increment sequence to obtain the execution instructions for the field equipment. It should be understood that although the numerical optimization solver calculates a mathematically optimal control increment sequence, this data is essentially an abstract mathematical solution, usually existing in the form of dimensionless or physical quantity increments, and cannot be directly understood or executed by the underlying actuators (such as pump frequency converters and aerator controllers). Furthermore, the core mechanism of Model Predictive Control (MPC) lies in rolling optimization, that is, although the actions of the next N steps are predicted, only the first action at the current moment is implemented to cope with random disturbances and model biases within the system, and resampling and calculation are performed at the next moment. Therefore, in the technical solution of this application, a dual-scale instruction mapping and execution is performed on the first action in the optimal control increment sequence to transform the abstract algorithm output into a specific electrical signal, and a rolling time-domain mechanism is used to realize the implementation of closed-loop control, ensuring that the control strategy can accurately drive the hardware equipment in the physical world.

[0043] In practice, firstly, the optimal control increment sequence is read from memory, and only the first element of the sequence is extracted. Since this value is the increment, the system needs to read the actual operating status value of the current device. (For example, the current operating frequency of the aerator), the increment is added to the current state to calculate the target absolute control quantity at the current moment. The process can be expressed by the following formula: in, This represents the absolute control quantity of the target at the current moment. This represents the current actual operating status value of the device.

[0044] Next, based on the preset device characteristic curve function This process maps target physical quantities (such as the target dissolved oxygen replenishment rate) to electrical control parameters of the equipment (such as the target frequency of the frequency converter or the valve opening percentage). This process also includes amplitude limiting logic to ensure that the generated instructions do not exceed the safe operating boundaries of the hardware (such as maximum / minimum frequency limits). Specifically, this mapping process is expressed by the formula: in, For the final device instruction value, This is the function for the device characteristic curve.

[0045] Finally, the calculated The data is encoded according to the fieldbus protocol specifications (e.g., converting floating-point numbers to hexadecimal register values) and sent to the underlying control unit via the I / O interface to drive hardware actions.

[0046] In summary, the water ecological restoration management method based on real-time monitoring according to the embodiments of this application is explained. It accurately predicts the biological natural restoration curve and lag response time by aligning and filtering real-time monitoring data with timestamps, calculating kinetic correction parameters using a microbial kinetic parameter set, and then performing this process. Based on this, an MPC optimization model incorporating cost functions and physical constraints is constructed. Numerical solutions generate a control increment sequence that balances water quality compliance with cost optimization, which is then converted into field equipment commands via dual-scale mapping. This approach effectively compensates for time lags in water ecological governance, achieving intelligent collaboration and refined control of equipment clusters, and significantly reducing system operating costs while ensuring water quality compliance.

[0047] Furthermore, a water ecosystem restoration management system based on real-time monitoring is also provided.

[0048] Figure 3 This is a block diagram of a water ecosystem restoration management system based on real-time monitoring according to an embodiment of this application. Figure 3As shown, the water ecological restoration management system 300 based on real-time monitoring according to an embodiment of this application includes: a timestamp alignment and filtering smoothing module 310, used to perform timestamp alignment and filtering smoothing on the acquired original water body real-time monitoring data stream set to obtain a water body real-time monitoring vector; a kinetic correction module 320, used to determine the kinetic correction parameters at the current moment based on the temperature component in the water body real-time monitoring vector and a preset set of benchmark microbial kinetic parameters; a vector splicing module 330, used to splice the water body real-time monitoring vector and the kinetic correction parameters at the current moment to obtain a system initial state vector; and a biological effect lag prediction module 340, used to perform vector splicing on the system initial state vector and the kinetic correction parameters at the current moment. Positive parameters are used to predict the lag of biological effects to obtain the biological natural remediation prediction curve and the predicted biological lag response time; the MPC optimization problem construction module 350 is used to construct the MPC optimization problem based on the predicted biological lag response time, the biological natural remediation prediction curve, the current actual water quality and the target water quality. The MPC optimization problem consists of a cost function, a system state prediction model and a set of physical constraints for system operation; the numerical optimization solution module 360 ​​is used to perform numerical optimization solution on the MPC optimization problem to obtain the optimal control increment sequence; the dual-scale instruction mapping and execution module 370 is used to perform dual-scale instruction mapping and execution on the first action in the optimal control increment sequence to obtain the execution instructions of the field equipment.

[0049] As described above, the water ecosystem restoration management system 300 based on real-time monitoring according to the embodiments of this application can be implemented in various wireless terminals, such as servers with water ecosystem restoration management algorithms based on real-time monitoring. In possible implementations, the water ecosystem restoration management system 300 based on real-time monitoring according to the embodiments of this application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the water ecosystem restoration management system 300 based on real-time monitoring can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the water ecosystem restoration management system 300 based on real-time monitoring can also be one of many hardware modules of the wireless terminal.

[0050] Alternatively, in another example, the real-time monitoring-based water ecological restoration management system 300 and the wireless terminal can also be separate devices, and the real-time monitoring-based water ecological restoration management system 300 can connect to the wireless terminal via wired and / or wireless networks and transmit interactive information in accordance with an agreed data format.

[0051] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A water ecological restoration and management method based on real-time monitoring, characterized in that, include: The acquired raw water body real-time monitoring data stream is timestamped and filtered and smoothed to obtain the water body real-time monitoring vector; Based on the temperature component in the real-time water body monitoring vector and the preset benchmark microbial kinetic parameter set, the kinetic correction parameters for the current moment are determined. The system's initial state vector is obtained by concatenating the real-time water monitoring vector and the dynamic correction parameters at the current moment. Biological effect hysteresis prediction is performed on the initial state vector of the system and the dynamic correction parameters at the current moment to obtain the biological natural repair prediction curve and the predicted biological hysteresis response time. Based on the predicted biological lag response time, the predicted biological natural remediation curve, the current actual water quality and the target water quality, an MPC optimization problem is constructed. The MPC optimization problem consists of a cost function, a system state prediction model and a set of physical constraints for system operation. Numerical optimization is performed on the MPC optimization problem to obtain the optimal control increment sequence; The first action in the optimal control increment sequence is subjected to dual-scale instruction mapping and execution to obtain the execution instructions for the field equipment.

2. The water ecological restoration and management method based on real-time monitoring according to claim 1, characterized in that, Based on the temperature component in the real-time water body monitoring vector and a preset set of baseline microbial kinetic parameters, the kinetic correction parameters for the current moment are determined, including: Real-time temperature is extracted from real-time water monitoring vectors and from a pre-defined set of baseline microbial dynamics parameters. Quasi-maximum specific growth rate and temperature correction factor; Based on real-time temperature, The quasi-maximum specific growth rate and temperature correction factor are used to calculate the basic thermodynamic growth rate to obtain the thermodynamically corrected maximum specific growth rate. The comprehensive environmental inhibition coefficient is determined based on real-time pH value, real-time dissolved oxygen, optimal pH value, pH tolerance width coefficient, and dissolved oxygen half-saturation constant. The kinetic correction parameters for the current moment are generated based on the comprehensive environmental inhibition coefficient and the thermodynamically corrected maximum specific growth rate.

3. The water ecological restoration and management method based on real-time monitoring according to claim 1, characterized in that, Based on real-time pH, real-time dissolved oxygen, optimal pH, pH tolerance width coefficient, and dissolved oxygen half-saturation constant, a comprehensive environmental inhibition coefficient is determined, including: The pH inhibition coefficient is determined based on real-time pH value, optimal pH value, and pH tolerance width coefficient. The dissolved oxygen saturation coefficient is determined based on real-time dissolved oxygen and dissolved oxygen half-saturation constant. The dissolved oxygen saturation coefficient and the pH inhibition coefficient are multiplied to obtain the comprehensive environmental inhibition coefficient.

4. The water ecological restoration and management method based on real-time monitoring according to claim 3, characterized in that, The pH inhibition coefficient is determined based on the real-time pH value, the optimum pH value, and the pH tolerance width coefficient, including: determining the pH inhibition coefficient using the following formula, wherein the formula is: in, It is the pH inhibition coefficient. For real-time pH value, The optimal pH value for the target bacterial group, This represents the pH tolerance width coefficient.

5. The water ecological restoration and management method based on real-time monitoring according to claim 3, characterized in that, Based on real-time dissolved oxygen and the dissolved oxygen half-saturation constant, the dissolved oxygen saturation coefficient is determined, including: determining the dissolved oxygen saturation coefficient using the following formula, wherein the formula is: in, It is the dissolved oxygen saturation coefficient. This is the real-time dissolved oxygen concentration. That is the dissolved oxygen half-saturation constant.

6. The water ecological restoration and management method based on real-time monitoring according to claim 1, characterized in that, Biological effect hysteresis prediction is performed on the system's initial state vector and the dynamic correction parameters at the current moment to obtain the biological natural repair prediction curve and the predicted biological hysteresis response time, including: Predictive state initialization is performed based on the system's initial state vector to obtain an initialized pollutant concentration prediction array and an initialized degradation rate prediction array. Based on the kinetic correction parameters at the current moment, an iterative forward extrapolation of the pollutant degradation process is performed on the initialized pollutant concentration prediction array and the initialized degradation rate prediction array to obtain the bioremediation prediction curve. Based on the simulation step size, the hysteresis response time feature of the biological natural repair prediction curve is extracted to obtain the predicted biological hysteresis response time.

7. The water ecological restoration and management method based on real-time monitoring according to claim 6, characterized in that, Based on the simulation step size, hysteresis response time features are extracted from the predicted bioremediation curve to obtain the predicted bioremediation hysteresis time, including... Peak rates were searched in the degradation rate prediction array of the bioremediation prediction curve to obtain the maximum value; Determine the position where the maximum value first appears in the degradation rate prediction array to obtain the peak index; The predicted biological hysteresis response time is determined based on the peak index and the simulation step size.

8. A water ecological restoration management system based on real-time monitoring, characterized in that, include: The timestamp alignment and filtering smoothing module is used to perform timestamp alignment and filtering smoothing on the acquired raw water body real-time monitoring data stream set to obtain the water body real-time monitoring vector; The kinetic correction module is used to determine the kinetic correction parameters at the current moment based on the temperature component in the real-time water body monitoring vector and the preset benchmark microbial kinetic parameter set. The vector splicing module is used to splice the real-time water monitoring vector and the dynamic correction parameters at the current moment to obtain the initial state vector of the system. The biological effect lag prediction module is used to perform biological effect lag prediction on the system's initial state vector and the dynamic correction parameters at the current moment to obtain the biological natural repair prediction curve and the predicted biological lag response time. The MPC optimization problem construction module is used to construct an MPC optimization problem based on the predicted biological lag response time, the predicted biological natural remediation curve, the current actual water quality, and the target water quality. The MPC optimization problem consists of a cost function, a system state prediction model, and a set of physical constraints for system operation. The numerical optimization solution module is used to perform numerical optimization solutions on the MPC optimization problem to obtain the optimal control increment sequence. The dual-scale instruction mapping and execution module is used to perform dual-scale instruction mapping and execution on the first action in the optimal control increment sequence to obtain the execution instructions for the field equipment.

Images