Intelligent control method and system for operation parameters of water treatment equipment

Abstract

The present application relates to the technical field of water treatment, and provides a water treatment equipment operation parameter intelligent control method and system, the method comprising: obtaining operation parameters of the water treatment equipment; performing preliminary verification on the operation parameters to determine the validity of the operation parameters; performing reliability evaluation on probe flow information in the operation parameters that have passed the preliminary verification; monitoring the change trend of a probe original signal of a flow measurement probe to obtain change information of water quality information; evaluating process operation risks based on the reliability evaluation result information, the change information of the water quality information, and the proximity of the water treatment equipment operation parameters to a preset minimum energy consumption critical point; and adjusting the redundancy of the water treatment equipment operation parameters according to the process operation risks, and sending control instructions corresponding to the adjusted water treatment equipment operation parameters to an execution mechanism. The present application has the effect of optimizing the balance between stable operation of the equipment and cost efficiency.

Description

Technical Field

[0001] This invention relates to the field of water treatment technology, and specifically to an intelligent control method and system for the operating parameters of water treatment equipment. Background Technology

[0002] In the field of water treatment, achieving a balance between stable equipment operation and cost-effectiveness has always been a challenge. Existing water treatment plants generally employ intelligent control systems, which use sensors to monitor water flow, level, and quality to ensure that effluent meets standards and optimize energy and chemical consumption. These systems typically have strict data verification rules that, based on long-term operating experience and industry standards, preliminarily identify sensor malfunctions or data transmission errors to ensure the reliability of control decisions.

[0003] In such an operating environment, the continuous operation of the water treatment process, especially the trace suspended solids and specific ions contained in the raw water, may cause slow and continuous wear or deposit accumulation on the surface of the flow measurement probe. This gradual physical change, due to its slow development and lack of perceptibility, is often difficult to detect in routine maintenance. Changes in the probe's surface characteristics can lead to subtle systematic deviations in its response curve to the actual water flow; for example, the probe may consistently overestimate or underestimate the actual flow rate. However, this initial deviation is small and usually remains within the error range allowed by the "sensor data validity verification rules" preset by the intelligent control system. Therefore, the system still considers it valid data and makes subsequent control decisions based on it. The system's internal self-diagnostic mechanism cannot identify this "hidden" deviation.

[0004] When upstream water sources are affected by seasonal rainfall, causing significant fluctuations in raw water turbidity and suspended solids content, the intelligent control system automatically responds according to preset logic, adjusting parameters such as chemical dosage and mixing pump operating frequency to maintain treatment effectiveness and operating costs. However, it is precisely during this automatic adjustment process to cope with changes in raw water quality that subtle but systemic deviations in the flow measurement probe are significantly amplified. The decision-making algorithm of the intelligent control system, especially the algorithm involving chemical dosage, heavily relies on accurate flow data to calculate the dosage ratio. If the flow measurement probe continuously overestimates the actual flow rate, the system will calculate and execute an insufficient dosage of chemicals relative to the true flow rate based on the erroneously overestimated flow data. This insufficient dosage, for treating high-turbidity raw water, means inadequate treatment intensity and poor coagulation effect.

[0005] This continuous operation of process parameters below the critical values ​​required for treating high-turbidity raw water leads to a gradual decline in water treatment efficiency, manifested as a slow deterioration in effluent turbidity or the removal rate of specific pollutants. This deterioration is cumulative and often does not immediately trigger existing alarm thresholds, causing the system to "feel good" while its actual process performance is quietly declining. Ultimately, although the intelligent control system reports operation in an "optimized" state and all real-time online sensor data shows "normal," periodic laboratory water quality test reports may reveal that the treated water quality is approaching or even occasionally exceeding discharge standards. Simultaneously, the decline in water treatment efficiency leads to increased load on downstream filtration units, faster filter media clogging, and increased backwashing frequency, thereby increasing energy consumption and maintenance costs. Therefore, how to effectively identify and compensate for this gradual sensor deviation within the normal calibration range in the intelligent control system, especially when the system is required to operate near energy consumption thresholds and automatically respond to complex and changing raw water conditions, becomes a pressing technical problem to be solved.

[0006] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0007] This application discloses an intelligent control method and system for the operating parameters of water treatment equipment, which aims to solve the problem that existing intelligent control systems for water treatment have difficulty in effectively identifying and compensating for progressive sensor deviations within the conventional calibration range when the operating parameters of the equipment are close to the minimum energy consumption critical point. This is especially true when dealing with complex and ever-changing raw water conditions, which can lead to a subtle decline in process performance and even affect the effluent water quality.

[0008] The technical solution of this application is as follows:

[0009] In a first aspect, this application discloses an intelligent control method for the operating parameters of water treatment equipment, specifically including:

[0010] Obtain the operating parameters of the water treatment equipment; the operating parameters include probe flow information and water quality information reported by the flow measurement probe;

[0011] Perform preliminary verification on the operating parameters to determine their validity;

[0012] A reliability assessment is performed on the probe flow rate information in the preliminarily verified operating parameters. The reliability assessment includes at least the following: First reference value is derived based on the physical principles of the water treatment process. This first reference value is a reference flow rate derived from the physical principles of the water treatment process, equipment structural parameters, and operating condition parameters. This reference flow rate is used to verify the probe flow rate information and is compared with the probe flow rate information to obtain first comparison information. Second process effect comparison information is generated based on water quality information, probe flow rate information, and reagent dosing information. This process effect comparison information is then compared with the actual process treatment effect to obtain second comparison information. Finally, a reliability assessment result is generated based on the first and second comparison information.

[0013] By monitoring the change trend of the original signal from the flow measurement probe, information on changes in water quality can be obtained.

[0014] Based on the reliability assessment results, changes in water quality information, and the degree to which the operating parameters of the water treatment equipment are close to the preset minimum energy consumption critical point, the risk of process operation is assessed.

[0015] Based on the operational risks of the process, the redundancy of the water treatment equipment's operating parameters is adjusted, and control commands corresponding to the adjusted operating parameters are sent to the actuators.

[0016] Through this technical solution, this application can effectively identify and handle potential deviations in the operating parameters of water treatment equipment. Especially when the equipment is running close to the minimum energy consumption threshold, it can intelligently adjust the operating parameters of the equipment through multi-dimensional reliability assessment and process operation risk assessment, thereby solving the problem of process performance degradation caused by latent sensor deviations in the prior art, ensuring that the effluent water quality meets the standards and optimizing operating costs.

[0017] Secondly, this application also discloses an intelligent control system for the operating parameters of water treatment equipment, used to perform intelligent control of the operating parameters of water treatment equipment, specifically including:

[0018] The operating parameter acquisition module is used to acquire the operating parameters of the water treatment equipment; the operating parameters include probe flow information and water quality information reported by the flow measurement probe;

[0019] The preliminary verification execution module is used to perform preliminary verification on the operating parameters to determine their validity.

[0020] The reliability assessment execution module is used to perform reliability assessment on the probe flow information in the preliminarily verified operating parameters. The reliability assessment includes at least: deriving a first reference value based on the physical principles of the water treatment process. The first reference value is a reference flow rate derived from the physical principles of the water treatment process, equipment structural parameters, and operating condition parameters. This reference flow rate is used to verify the probe flow information and compare it with the probe flow information to obtain first comparison information; generating process effect comparison information based on water quality information, probe flow information, and reagent dosing information, and comparing the process effect comparison information with the actual process treatment effect to obtain second comparison information; and generating reliability assessment result information based on the first and second comparison information.

[0021] The change monitoring module is used to monitor the change trend of the original signal of the flow measurement probe and obtain information on the change of water quality.

[0022] The risk assessment module is used to assess the operational risks of the process based on reliability assessment results, changes in water quality information, and the proximity of water treatment equipment operating parameters to the preset minimum energy consumption critical point.

[0023] The control command sending module is used to adjust the redundancy of the water treatment equipment's operating parameters based on the process operation risks, and to send the control commands corresponding to the adjusted water treatment equipment operating parameters to the actuators.

[0024] This application provides an intelligent control system for the operating parameters of water treatment equipment. Through modular design, it can effectively realize the above-mentioned intelligent control method for the operating parameters of water treatment equipment, thereby supporting intelligent control of the operating parameters of water treatment equipment at the hardware level. This solves the problem of process performance degradation caused by latent sensor deviations in the prior art, ensuring that the effluent water quality meets the standards and optimizing operating costs.

[0025] Beneficial Effects: This application effectively solves the technical problem of existing intelligent water treatment control systems failing to identify gradual sensor deviations within the conventional calibration range when equipment operating parameters approach the minimum energy consumption threshold, leading to a subtle decline in process performance. By introducing a multi-dimensional reliability assessment mechanism, this application can accurately identify hidden deviations in probe flow information, avoiding insufficient or excessive reagent dosage due to erroneous data. This ensures that the system maintains stable treatment results even when raw water quality fluctuates. Furthermore, through comprehensive assessment of process operation risks and intelligent adjustment of parameter redundancy, this application optimizes energy consumption and reagent consumption while ensuring effluent quality meets standards. This avoids additional maintenance costs due to efficiency degradation, achieving a balance between stable equipment operation and cost-effectiveness, demonstrating significant and superior technical effects. Attached Figure Description

[0026] Figure 1 This is a flowchart of a method for intelligent control of operating parameters of a water treatment device according to one embodiment of the present invention;

[0027] Figure 2 This is a flowchart of a method for intelligent control of operating parameters of a water treatment device according to another embodiment of the present invention;

[0028] Figure 3 This is a system block diagram of an intelligent control system for the operating parameters of a water treatment device according to another embodiment of the present invention;

[0029] Explanation of reference numerals in the attached figures:

[0030] 1. Intelligent control system for water treatment equipment operating parameters; 11. Operating parameter acquisition module; 12. Preliminary verification execution module; 13. Reliability assessment execution module; 14. Change monitoring module; 15. Operational risk assessment module; 16. Control command sending module. Detailed Implementation

[0031] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0032] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0033] This application proposes an intelligent control method for the operating parameters of water treatment equipment, combined with... Figure 1 As shown, it includes:

[0034] S1, Obtain the operating parameters of the water treatment equipment; the operating parameters include probe flow information and water quality information reported by the flow measurement probe;

[0035] S2, Perform preliminary verification on the operating parameters to determine their validity;

[0036] S3, perform a reliability assessment on the probe flow rate information in the preliminarily verified operating parameters; the reliability assessment includes at least: deriving a first reference value based on the physical principles of the water treatment process, the first reference value being a reference flow rate derived from the physical principles of the water treatment process, equipment structural parameters, and operating condition parameters, used to verify the probe flow rate information and compare it with the probe flow rate information to obtain first comparison information; generating process effect comparison information based on water quality information, probe flow rate information, and reagent dosing information, and comparing the process effect comparison information with the actual process treatment effect to obtain second comparison information; generating reliability assessment result information based on the first and second comparison information;

[0037] S4 monitors the trend of the original signal change of the flow measurement probe to obtain information on changes in water quality.

[0038] S5 assesses the operational risks of the process based on reliability assessment results, changes in water quality information, and the proximity of water treatment equipment operating parameters to the preset minimum energy consumption critical point.

[0039] S6 adjusts the redundancy of the water treatment equipment's operating parameters based on the process operation risks, and sends the control instructions corresponding to the adjusted water treatment equipment operating parameters to the actuator.

[0040] The term "water treatment equipment" as used in this application refers to various process units and their auxiliary actuators used in a water treatment plant to complete the raw water purification process, including but not limited to coagulation tanks, sedimentation tanks, filtration units, disinfection units, and their associated pumps, valves, agitators, metering devices, etc. The aforementioned equipment operates collaboratively according to a predetermined process flow, removing pollutants from the water body through physical, chemical, or biological treatment methods. "Operating parameters" refer to measurable data used to characterize the operating status of the equipment and the water quality during the water treatment process, including flow rate, pressure, temperature, pH value, turbidity, conductivity, dissolved oxygen, residual chlorine, COD, ammonia nitrogen, etc. Specifically, "probe flow information" refers to the instantaneous or cumulative flow value directly output by the flow measurement probe. This information is usually generated in the form of an electrical signal and enters the control system after analog-to-digital conversion. "Water quality information" covers the physical, chemical, or biological indicators of raw water, water samples from each treatment process stage, and effluent, which can be provided by online water quality analyzers or laboratory testing devices.

[0041] "Preliminary verification" refers to performing basic validity checks on the acquired operating parameters to eliminate obviously erroneous or distorted data. This preliminary verification includes at least checking data completeness, numerical format, measurement range, and physical reasonableness. For example, it involves determining whether there are missing values, non-numerical characters, whether the measurement exceeds the sensor's nominal range or the process's physical limits, and whether there are any transient changes that violate the equipment's operating rules. This step ensures that the data entering subsequent analysis stages has basic reliability.

[0042] "Reliability assessment" is a comprehensive evaluation of the authenticity and credibility of the probe flow rate information. It is achieved through a multi-source information cross-validation mechanism, rather than relying solely on a single threshold. The "first reference value" is a theoretical flow rate value derived from a physical model of the water treatment process. This first reference value can be calculated based on equipment structural dimensions (e.g., pipe inner diameter, effective cross-sectional area), pump head curve and speed parameters, valve opening and resistance coefficients, combined with the fluid continuity equation and the principle of energy conservation. Since this first reference value is derived from independent physical derivation, its deviation from the probe flow rate information can serve as an important basis for judging probe measurement errors. "Process effect comparison information" is obtained through coupled analysis of water quality information, probe flow rate information, and reagent dosage. For example, in a coagulation sedimentation process, the theoretical removal rate can be calculated based on influent turbidity, coagulant dosage, and probe flow rate information, and then compared with the actual effluent turbidity to form process effect deviation information. If the theoretical treatment efficiency deviates significantly from the actual treatment effect, it can be inferred that there is a risk of distortion in the probe flow rate information.

[0043] "Water quality information change information" refers to the trend information indirectly reflecting changes in suspended solids concentration, dissolved substances, or bubble disturbances in the water body through time-domain or frequency-domain analysis of the probe's raw signal. Since the microscopic fluctuations of the flow probe's raw signal are often affected by the fluid state, extracting signal mean drift, variance changes, spectral structure, or harmonic components can identify conditions such as scale buildup on the probe surface, drastic water quality fluctuations, or abnormal flow patterns. "Preset minimum energy consumption critical point" refers to the minimum energy consumption allowed for the water treatment equipment to operate while ensuring the effluent water quality meets standards. This critical point can be obtained through historical operating data statistics or optimization model calculations, and it corresponds to a boundary range of equipment operating parameters. When the operating parameters approach this critical range, the system's tolerance for probe errors or water quality fluctuations decreases significantly, as any slight error may lead to a decline in effluent quality or a rebound in energy consumption. "Redundancy" refers to the control margin reserved to cope with uncertainty risks, such as reserving a certain adjustment space for reagent dosage or pump operating frequency to improve system robustness.

[0044] In the specific implementation process, the operating parameters of the water treatment equipment are first acquired. Probe flow information can be collected through electromagnetic or ultrasonic flow meters, while water quality information can be collected through devices such as turbidity meters, pH meters, and online COD analyzers. After the sensors convert analog signals into digital signals, they are transmitted to the central control unit via an industrial communication network. Subsequently, the operating parameters are preliminarily verified, and invalid or abnormal data is eliminated.

[0045] After initial verification, a reliability assessment is performed on the probe flow rate information. First, a first reference value is calculated based on the physical principles of the water treatment process, and this first reference value is compared with the probe flow rate information to form first comparison information. Second, process effect comparison information is generated based on water quality information, probe flow rate information, and reagent dosage, and the theoretical treatment effect is compared with the actual treatment effect to form second comparison information. Finally, a reliability assessment result is generated based on the degree of deviation between the first and second comparison information. If both comparison results show significant deviations, the probe flow rate information is deemed to have low reliability.

[0046] Simultaneously, the system continuously monitors the changing trends of the probe's raw signal, generating information on changes in water quality. By performing mean analysis, fluctuation intensity analysis, or spectrum analysis on the raw signal, it can identify whether there is signal drift, abnormal oscillation, or contamination.

[0047] Next, based on the reliability assessment results, changes in water quality information, and the proximity of operating parameters to the preset minimum energy consumption threshold, a comprehensive assessment of process operation risk is conducted. When the reliability of probe flow information is low, water quality fluctuates drastically, and the operating status is close to the minimum energy consumption threshold, the process operation risk is determined to be high; otherwise, the risk is determined to be low.

[0048] Finally, based on the risk level of the process operation, the redundancy of the water treatment equipment's operating parameters is adjusted. When the risk is high, the redundancy of the reagent dosage or the pump operating frequency is appropriately increased to keep the system away from the critical boundary operating state; when the risk is low, the redundancy can be maintained or appropriately reduced to maintain energy-saving operation. The adjusted operating parameters are sent to actuators such as metering pumps, frequency converters, or agitators through control commands, thereby achieving risk-oriented adaptive operation control.

[0049] Optional, combined Figure 2 As shown, the steps for performing a reliability assessment on the probe flow information in the preliminarily verified operating parameters specifically include:

[0050] A1. Acquire contextual characteristic data; contextual characteristic data includes raw water dissolved organic matter characteristic data, probe micro-signal spectrum characteristic data, and upstream environmental macro-data;

[0051] A2. Based on contextual characteristic data, adjust the evidence weight information of the three types of evidence sources; the three types of evidence sources include flow rate back-inference verification evidence sources, process effect evidence sources, and probe health evidence sources.

[0052] A3. When the adjusted three types of evidence sources form a conflict, the conflict resolution is performed on the conflicting information based on the process operation risk information and priority rules to determine whether there is a probe deviation in the probe flow information and to obtain the conflict resolution result information.

[0053] A4. Based on the conflict resolution results, perform deviation isolation and data trust reconstruction to provide input for the assessment of subsequent process operation risks.

[0054] A5. When the conflict resolution result information indicates that the deviation is due to the inaccuracy of the process model, an adaptive adjustment is performed on the process model.

[0055] Contextual characteristic data refers to a multi-dimensional set of information used to characterize the current operating environment of water treatment equipment and the working status of the flow measurement probe itself. This contextual characteristic data includes not only internal process parameters but also external environmental variables and probe status characteristics. Specifically, raw water dissolved organic matter characteristic data is a set of data that quantitatively describes the types, concentration levels, and distribution structure of dissolved organic matter in the raw water entering the water treatment system. For example, characteristic peak intensities, absorbance ratios, or characteristic index parameters can be obtained using online ultraviolet-visible spectrometers, fluorescence spectrometers, or chromatographic analysis devices. This type of data reflects the complexity and fluctuation range of raw water quality, thus affecting the stability of process performance evaluation.

[0056] Microscopic signal spectral characteristic data of a flow measurement probe refers to the energy distribution, phase characteristics, and harmonic structure information of the original output signal of the probe at different time scales or frequency domains. This data can be obtained by performing Fast Fourier Transform, Wavelet Transform, or Short-Time Fourier Transform on the original electrical signal of the probe. Its purpose is to identify signal distortion characteristics caused by the formation of an adsorption layer on the probe surface, probe electrode contamination, aging of internal electronic components, or mechanical wear. For example, when a microscopic deposition layer forms on the probe surface, its high-frequency signal components may attenuate or abnormally enhance, thus forming an identifiable spectral characteristic pattern.

[0057] Upstream environmental macro data refers to external environmental variables that affect the operational stability of water treatment processes, such as changes in upstream reservoir water levels, regional rainfall, industrial discharge load changes, and seasonal water temperature variations. This type of data is used to construct operational background information, enabling the system to distinguish between process parameter deviations caused by changes in the external macro environment and anomalies caused by probe errors themselves.

[0058] Evidence weighting information consists of credibility or importance coefficients assigned to different evidence sources, used to weight and integrate multi-source evidence during reliability assessment. Flow rate back-calculation verification evidence is derived from the physical principles of water treatment processes, resulting in a first comparison information set by comparing a first reference value with probe flow rate information. Process effectiveness evidence is a second comparison information set between process effectiveness comparison information generated from water quality information, probe flow rate information, and reagent dosing information, and the actual treatment effect. Probe health evidence is a probe self-condition assessment information formed based on the probe's microscopic signal spectrum characteristic data.

[0059] The system dynamically adjusts the weighting of evidence based on contextual characteristic data. For example, when the raw water dissolved organic matter characteristic data indicates a significant increase in raw water complexity and drastic water quality fluctuations, the process effect may exhibit nonlinear fluctuations. In this case, the system reduces the weighting of evidence sources related to process effect and relatively increases the weighting of evidence sources related to flow rate back-calculation and probe health, thereby reducing the interference of water quality complexity on reliability judgments. Conversely, when the probe's microscopic signal spectrum characteristic data exhibits a stable structure and there are no obvious anomalies in the upstream environmental macroscopic data, the weighting of evidence sources related to process effect can be increased, thereby enhancing confidence in the overall consistency of the process.

[0060] Conflict of evidence refers to a situation where different sources of evidence reach inconsistent conclusions regarding the reliability of probe flow information. For example, flow back-calculation evidence may show a small deviation, while process effect evidence may indicate abnormal processing efficiency. Conflict resolution refers to the system's comprehensive evaluation of conflicting evidence based on preset priority rules and current process operation risk information after detecting a conflict. Priority rules can be dynamically adjusted according to the process risk level. For example, in high-risk conditions approaching the preset minimum energy consumption threshold, probe health evidence has a higher priority to ensure measurement accuracy; in low-risk conditions, process effect evidence may have a higher weight. The conflict resolution result clearly indicates whether probe flow information exhibits probe bias and the possible attribution categories of the bias.

[0061] When the conflict resolution results confirm a deviation in the probe flow information, deviation isolation processing is performed. Deviation isolation processing refers to logically separating the deviated flow data from the normal data stream, preventing it from participating in subsequent process operation risk assessment and control decisions. The isolated data can be marked as abnormal and stored in an abnormal data cache.

[0062] After isolating the deviation data, a data confidence reconstruction process is performed. This process involves recalculating the confidence score for the remaining valid data and generating alternative flow estimates through data fusion, interpolation, or model prediction. For example, a predictive model can be built based on a first reference value and historical stable operating ranges to generate corrected flow values, thereby ensuring the continuity and stability of subsequent process control.

[0063] When the conflict resolution results indicate that the evidence conflict stems from an inaccurate process model, the system performs adaptive adjustments. Adaptive adjustments refer to parameter corrections, structural optimizations, or model updates to the process model used to generate evidence for flow back-calculation or process effectiveness. For example, updating pipeline resistance coefficients, correcting pump efficiency curve parameters, or recalibrating the reagent reaction kinetic model improves the model's adaptability to current situational data. This adaptive adjustment mechanism enables the model to maintain accuracy and robustness under different raw water characteristics and environmental conditions, thereby reducing the probability of future evidence conflicts.

[0064] Optionally, the step of performing a reliability assessment on the probe flow information in the preliminarily verified operating parameters may also include:

[0065] Time-frequency analysis was performed on the probe's raw signal to identify and quantify the nonlinear signal distortion characteristics caused by the micro-adsorption layer or wear on the probe surface;

[0066] Establish decision tree rules based on multidimensional contextual feature parameters, and adjust the evidence weight information according to the combination state of contextual feature data in nonlinear signal distortion features;

[0067] Based on the adjusted evidence weight information, a multi-attribute decision analysis method is adopted to comprehensively evaluate the information on evidence conflicts by taking into account the contribution of evidence, the risk level of process operation, and the preset expert experience rules, and output the conflict adjudication result information.

[0068] Specifically, performing time-frequency analysis on the probe's raw signal involves using signal processing techniques such as Fourier transform, wavelet transform, or Hilbert-Huang transform to convert the probe's raw signal from the time domain to the frequency domain while preserving time information. This allows observation of frequency component changes at different time points. The aim is to identify and quantify the nonlinear signal distortion characteristics caused by microscopic adsorption layers or wear on the probe surface. For example, microscopic adsorption layers on the probe surface may cause signal attenuation, frequency shift, or harmonic distortion, while probe wear may cause increased signal noise, the appearance of specific frequency components, or broadening of the spectral width. Time-frequency analysis can capture these subtle, nonlinear signal characteristics that are difficult to fully capture using traditional single time-domain or frequency-domain analysis.

[0069] The establishment of decision tree rules based on multi-dimensional contextual feature parameters involves constructing a decision model that dynamically adjusts the evidence weights of three types of evidence sources (flow rate back-calculation evidence source, process effect evidence source, and probe health evidence source) based on various contextual feature parameters (such as raw water dissolved organic matter characteristic data, probe micro-signal spectrum characteristic data, and upstream environmental macro-data) and nonlinear signal distortion characteristics obtained through time-frequency analysis. Specifically, the decision tree rules can adjust the weights of the flow rate back-calculation evidence source based on the combination of contextual feature data in the nonlinear signal distortion characteristics. For example, when a specific frequency of signal distortion is detected and the raw water dissolved organic matter concentration is high, the weight of the flow rate back-calculation evidence source may be reduced, while the weight of the probe health evidence source may be increased. The aim is to make the evidence weight adjustment more intelligent and context-sensitive, thereby more accurately reflecting the reliability of different evidence sources under specific operating conditions.

[0070] In practical applications, the multi-attribute decision analysis method based on adjusted evidence weights means that when evidence conflicts occur, instead of relying solely on simple priority rules, it comprehensively considers multiple decision attributes, such as the contribution of each evidence source, the current process operation risk level, and pre-set expert experience rules, to quantitatively evaluate the information regarding evidence conflicts. For example, the Analytic Hierarchy Process (AHP), TOPSIS, or fuzzy comprehensive evaluation method can be used to assign a quantified contribution score to each evidence source. This score is then combined with the process operation risk level (e.g., high, medium, low risk) and pre-set expert adjudication rules to calculate a comprehensive conflict adjudication score or confidence level, thereby outputting a quantified conflict adjudication result. The aim is to provide a more objective, comprehensive, and refined conflict adjudication mechanism, avoiding misjudgments caused by single rules or subjective judgments, and ensuring more reasonable decisions in complex situations.

[0071] The steps for performing a reliability assessment on the probe flow information in the preliminarily verified operating parameters include:

[0072] Continuously acquire contextual feature data;

[0073] Time-domain and frequency-domain feature information is extracted from the original signal from the probe to capture instantaneous changes and spectral distribution information of the signal;

[0074] The steps for constructing and updating evidence weight adjustment rules include: monitoring the deviation of contextual feature data relative to historical data to identify new deviation patterns; and adjusting evidence weight information based on the new deviation patterns when they are identified.

[0075] Based on the adjusted evidence weighting information, calculate the evidence contribution of each source of evidence;

[0076] A conflict resolution mechanism based on case-based reasoning is adopted. The conflict resolution mechanism includes: when conflicting evidence occurs, retrieving historical cases from a pre-set case library that are identical to the current situational feature data and conflicting evidence; when identical historical cases are retrieved, outputting conflict resolution results based on the resolution results and actual effects of the historical cases; when identical historical cases are not retrieved, outputting preliminary resolution results based on the current process operation risk level, pre-set expert experience rules, and evidence contribution, and storing the current situational feature data, conflicting evidence information, and preliminary resolution results as new cases.

[0077] Based on the conflict resolution results, deviation isolation and data trust reconstruction are performed.

[0078] When the discrepancy in the conflict resolution results is attributed to an inaccurate process model, an adaptive adjustment is performed on the process model.

[0079] Specifically, continuously acquiring contextual characteristic data refers to the system's uninterrupted collection of multi-dimensional information related to the water treatment equipment's operating environment and probe status. This includes real-time changes in raw water quality, macroscopic data such as ambient temperature, humidity, and pressure, as well as microscopic signal characteristic data of the probe itself. This continuous acquisition ensures that the system can promptly perceive and respond to dynamic changes in the environment and equipment status.

[0080] Extracting time-domain and frequency-domain features from the probe's raw signal can be understood as conducting in-depth analysis of the raw electrical signal output by the flow measurement probe. Time-domain features aim to capture instantaneous changes in the signal along the time axis, such as amplitude, mean, variance, peak value, and zero-crossing rate. These features reflect the probe's response speed or the presence of instantaneous disturbances. Frequency-domain features, through methods such as Fourier transform, analyze the signal's distribution along the frequency axis, such as power spectral density, energy in specific frequency bands, and harmonic component intensity. These features help identify periodic or non-periodic signal distortions caused by mechanical wear, vibration, or adsorption layers within the probe. By fusing time-domain and frequency-domain features, the probe's operating state and potential deviations can be more comprehensively and meticulously characterized.

[0081] In practical applications, the purpose of constructing and updating evidence weight adjustment rules is to enable the system to dynamically adjust the importance of different evidence sources (such as flow rate back-calculation evidence sources, process effect evidence sources, and probe health evidence sources) in reliability assessment based on the evolution of the actual operating situation. Specifically, the system continuously monitors the currently acquired situational characteristic data and compares it with historical data to identify new deviation patterns. For example, when raw water turbidity or organic matter content exhibits continuous abnormal fluctuations, and this fluctuation pattern has not appeared in historical data or has a low frequency of occurrence, the system will identify it as a new deviation pattern. Once a new deviation pattern is identified, the system will adaptively adjust the evidence weight information of each evidence source based on this new deviation pattern. For example, under a certain new water quality fluctuation pattern, the weight of probe health evidence sources may be increased to focus more on the stability of the probe itself; while under another pattern, the weight of process effect evidence sources may be decreased to avoid excessive influence from external process disturbances. This dynamic adjustment mechanism makes reliability assessment more flexible and accurate.

[0082] Furthermore, based on the adjusted evidence weighting information, the evidentiary contribution of each source of evidence is calculated. This refers to quantifying the degree of influence of each source of evidence on the final reliability assessment result after the evidence weighting adjustment. This helps to more reasonably evaluate the supporting strength of different pieces of evidence in subsequent conflict adjudication.

[0083] As a preferred implementation, a case-based conflict resolution mechanism is employed to enhance the system's ability to handle complex and unknown evidence conflict scenarios. When the system detects a conflict between different evidence sources—for example, a significant discrepancy between the back-calculated flow rate and the probe flow rate information, coupled with abnormal process performance but normal probe health evidence—the system first retrieves historical cases from a pre-defined case library that are similar to the current situation's characteristic data (e.g., current water quality, operating conditions) and evidence conflict information (e.g., which evidences conflict and the degree of conflict). If a matching historical case is found, the system directly outputs the current conflict resolution result based on the resolution outcome of that historical case (i.e., how the probe deviation was determined at that time) and its corresponding actual effect information (i.e., whether the resolution was proven correct in subsequent actual operation). This approach leverages historical experience, improving the efficiency and accuracy of the resolution. However, when no matching historical case is found, it indicates that a new, unprocessed conflict scenario may be present. At this point, the system will output a preliminary ruling based on the current operational risk level (e.g., a more conservative judgment may be made in a high-risk situation), pre-defined expert experience rules (e.g., prioritizing certain types of evidence under specific conditions), and the previously calculated contribution of each piece of evidence. Simultaneously, to enable the system to learn, the current situational characteristic data, information on evidence conflicts, and this preliminary ruling will be stored as new cases, expanding the case library so that similar situations can be directly referenced in the future.

[0084] Therefore, based on the conflict resolution results, deviation isolation and data confidence reconstruction are performed. The purpose is to identify and separate deviations in the probe flow information according to the resolution results, and to reassess the confidence of the data. For example, if the resolution results indicate that the probe has a drift deviation, the deviation signal will be isolated, the probe flow data will be corrected, and the confidence of the probe data will be reduced to avoid its negative impact on subsequent control decisions.

[0085] Specifically, when the deviation in the conflict resolution results is attributed to an inaccurate process model, adaptive adjustments are made to the process model to continuously optimize it. For example, if the system repeatedly determines that the deviation between the probe flow rate information and the back-calculated value from the process model is due to the process model failing to accurately reflect the current operating conditions or water quality characteristics, the system will trigger adaptive adjustments to the parameters or structure of the process model to improve its accuracy and predictive ability.

[0086] Optionally, the steps for extracting time-domain and frequency-domain feature information from the probe's original signal include:

[0087] Continuously receive the raw probe signal transmitted by the flow measurement probe;

[0088] The original signal from the probe is sampled in real time and divided into continuous time windows;

[0089] Within each time window, the root mean square value, peak factor, waveform factor are calculated, and the instantaneous amplitude, rise rate, and fall rate are monitored to obtain time-domain characteristic information.

[0090] Spectral analysis is performed on the probe's raw signal within each time window to obtain frequency domain characteristics of power spectral density, energy in a specific frequency band, harmonic component intensity, and spectral centroid.

[0091] By fusing time-domain and frequency-domain feature information to form multi-dimensional feature vector information, noise suppression is performed on the multi-dimensional feature vector information to obtain the processed multi-dimensional feature vector information.

[0092] The nonlinear signal distortion characteristics indicating the micro-adsorption layer or wear on the probe surface are identified and quantified from the processed multidimensional feature vector information.

[0093] Specifically, continuously receiving the raw probe signal transmitted by the flow measurement probe refers to the system continuously acquiring unprocessed raw electrical or physical quantity signals from the flow measurement probe. These raw signals are electrical or optical representations of the physical quantities directly sensed by the probe, and they contain all information about the probe's operating status, including normal operating data, noise, and potential fault or deviation signals.

[0094] The process of real-time sampling and dividing the probe's raw signal into continuous time windows can be understood as converting a continuous analog signal into a discrete digital signal sequence, and then dividing these digital sequences into a series of interconnected data segments according to a preset time length or number of sampling points. Real-time sampling ensures timely capture of signal changes, while the division of time windows provides independent and manageable analysis units for subsequent time-domain and frequency-domain analysis, which helps to capture the local characteristics of the signal.

[0095] In practical applications, the root mean square (RMS) value, peak factor, and waveform factor are calculated within each time window, and the instantaneous amplitude, rise rate, and fall rate are monitored to obtain time-domain characteristic information. The RMS value reflects the effective energy of the signal, the peak factor indicates the impulsiveness of the signal, the waveform factor describes the waveform shape of the signal, while the instantaneous amplitude, rise rate, and fall rate capture the transient behavior and rate of change of the signal. These time-domain characteristics can comprehensively describe the characteristics of the probe's original signal in the time dimension, which is of great significance for identifying abnormal fluctuations, instantaneous impacts, or drifts in the signal.

[0096] Furthermore, spectral analysis is performed on the raw probe signal within each time window to obtain frequency domain characteristics such as power spectral density, energy in specific frequency bands, harmonic component intensity, and spectral centroid. Spectral analysis transforms the time-domain signal to the frequency domain using methods such as Fourier transform, revealing the frequency component distribution of the signal. Power spectral density represents the distribution of signal energy at different frequencies; energy in specific frequency bands can focus on specific frequency ranges related to probe fault modes; harmonic component intensity helps identify periodic interference or internal mechanical vibrations of the probe; and the spectral centroid reflects the average frequency of the signal energy. These frequency domain characteristics can effectively reveal the periodicity, harmonic characteristics, and potential resonance or attenuation phenomena of the probe signal, offering unique advantages for diagnosing internal structural problems or external interference of the probe.

[0097] Therefore, by fusing time-domain and frequency-domain feature information, a multi-dimensional feature vector is formed. Noise suppression is then applied to this multi-dimensional feature vector to obtain the processed multi-dimensional feature vector. Fusing time-domain and frequency-domain features provides a more comprehensive and richer description of the probe signal, forming a high-dimensional feature space. Noise suppression aims to remove or reduce random interference unrelated to the probe state in the original signal, such as environmental electromagnetic interference and internal sensor thermal noise, ensuring the accuracy of subsequent analysis. Noise suppression improves the signal-to-noise ratio of the feature vector, making the signal components that truly reflect probe deviation more prominent.

[0098] Specifically, this involves identifying and quantifying nonlinear signal distortion features indicating microscopic adsorption layers or wear on the probe surface from the processed multidimensional feature vector information. Microscopic adsorption layers or wear on the probe surface are common causes of probe performance degradation and measurement bias, often introducing nonlinear distortions into the raw signal that are difficult to judge using simple thresholds. Identifying and quantifying these nonlinear distortion features, for example through nonlinear dynamics analysis, higher-order statistical analysis, or machine learning models, can provide deeper and more diagnostic information for probe health assessment, thereby detecting potential probe failures earlier and more accurately.

[0099] Optionally, the noise suppression steps include:

[0100] Continuously acquire operating parameters that reflect the current operating conditions;

[0101] Based on operating parameters, evaluate noise statistical characteristics in real time;

[0102] Based on operating parameters and noise statistical characteristics, the noise suppression strategy is adjusted. When the noise is non-Gaussian or non-stationary, nonlinear filtering or adaptive thresholding is used. When the noise and the probe deviation signal overlap in the frequency domain, they are distinguished by combining the signal's time-domain sparsity or local energy characteristics.

[0103] During noise suppression, protective processing is performed on the probe deviation signal component in the multidimensional feature vector information that indicates probe deviation;

[0104] A consistency check is performed on the noise-suppressed multidimensional feature vector information to ensure that the multidimensional feature vector information is consistent with the physical laws and logical relationships of the water treatment process.

[0105] Specifically, continuously acquiring operating parameters reflecting the current working conditions means that the system constantly receives data from various sensors of the water treatment equipment. These operating parameters may include, but are not limited to, the flow rate, pressure, temperature, pH value, turbidity, conductivity, pump speed, valve opening, and reagent dosage of the water treatment equipment. These parameters can reflect the current operating status of the water treatment process and environmental conditions in real time, providing basic information for subsequent noise assessment and strategy adjustment.

[0106] Real-time evaluation of noise statistical characteristics based on operating parameters refers to the statistical description of unwanted components in the signal, such as the noise's mean, variance, power spectral density, probability distribution type (e.g., Gaussian, Poisson), and stationarity (e.g., whether it changes over time). Analyzing these characteristics allows us to understand the nature of the current noise, providing a basis for selecting appropriate suppression methods. For example, when operating parameters indicate that the equipment is in a stable operating state, the noise may exhibit some stationary characteristics; while when operating parameters indicate a change in operating conditions or equipment failure, the noise may exhibit non-stationary or abrupt characteristics.

[0107] In practical applications, adjusting noise suppression strategies based on operating parameters and noise statistical characteristics can be understood as the system dynamically selecting or adjusting the most suitable noise suppression method for the current operating conditions based on real-time evaluated noise statistical characteristics and current operating parameters. Specifically, when noise exhibits a non-Gaussian distribution or non-stationarity, nonlinear filtering or adaptive thresholding is employed. Nonlinear filtering methods can include median filtering, bilateral filtering, and wavelet transform denoising, which are highly effective in handling noise with spikes, impulses, or asymmetric distributions. Adaptive thresholding refers to dynamically adjusting the denoising threshold based on the local characteristics of the signal or noise. For example, a lower threshold is used in areas with high signal strength to preserve details, while a higher threshold is used in noise-dominant areas for effective denoising. When noise and probe deviation signals overlap in the frequency domain, they are distinguished by incorporating signal temporal sparsity or local energy characteristics. Specifically, signal temporal sparsity means that in the time domain, probe deviation signals may manifest as transient, discrete events, while noise may be more continuous or random. Local energy characteristics refer to the significant energy concentration of the probe deviation signal in a specific time or frequency region, while the energy distribution of noise is more diffuse. By analyzing these characteristics, even if there is overlap in the frequency domain, it is possible to effectively distinguish the true probe deviation signal from background noise.

[0108] Furthermore, during noise suppression, protective processing is performed on the probe deviation signal components in the multidimensional feature vector information that indicate probe deviation. This protective processing aims to prevent the critical probe deviation signal from being mistakenly damaged or weakened during denoising. This can be achieved by identifying and labeling these signal components before filtering or thresholding, or by enhancing or restoring them after denoising. The goal is to ensure that even in noisy environments, weak signals indicating probe deviation can be effectively preserved and extracted, providing accurate input for subsequent deviation identification.

[0109] Furthermore, a consistency check is performed on the noise-suppressed multidimensional feature vector information to ensure that it remains consistent with the physical laws and logical relationships of the water treatment process. Specifically, the consistency check compares the processed feature vector information with the known physical models, chemical reaction laws, and logical constraints of the water treatment process. For example, it checks whether features such as flow rate, pressure, and water quality parameters satisfy fundamental physical laws such as mass conservation and energy conservation, or whether they conform to the expected trends during normal equipment operation. The purpose is to further verify the effectiveness of noise suppression and prevent data distortion caused by excessive denoising or incorrect processing, thereby improving the overall reliability of the data.

[0110] Optionally, the step of performing protective processing on the probe deviation signal component in the multidimensional feature vector information that indicates probe deviation includes:

[0111] Continuously acquire operating parameters, including pump speed information, valve opening information, reaction tank stirring intensity information, and raw water turbidity change rate information;

[0112] Based on operating parameters, the signal recognition threshold information is adjusted; specifically, when the operating parameters indicate a stable operating state, the signal recognition threshold information is reduced to improve the sensitivity to weak deviation signals; when the operating parameters indicate a switching or disturbance state, the signal recognition threshold information is increased to suppress misjudgment.

[0113] Time-domain feature matching and frequency-domain feature matching are performed on the signal components in the multidimensional feature vector information, and when the matching degree reaches the signal recognition threshold information, the corresponding signal component is identified as the probe deviation signal component;

[0114] Protective enhancement processing is performed on the identified probe deviation signal components; the protective enhancement processing includes amplitude amplification, time-domain smoothing, and frequency-domain filtering;

[0115] The probe deviation signal component after protective enhancement processing is fused with the remaining feature vector information that was not identified as a probe deviation signal component to obtain the final feature vector information.

[0116] Specifically, continuously acquiring operating parameters refers to the system's real-time monitoring and collection of key operational status data from water treatment equipment, such as pump speed information, valve opening information, reaction tank stirring intensity information, and raw water turbidity change rate information. These operating parameters comprehensively reflect the current operating conditions of the water treatment process, providing a basis for subsequent signal identification threshold adjustments. Among them, pump speed information and valve opening information are directly related to hydraulic conditions and flow rate changes, reaction tank stirring intensity information affects mixing uniformity and reaction efficiency, while raw water turbidity change rate information indicates fluctuations in raw water quality.

[0117] Based on the aforementioned operating parameters, the signal recognition threshold is dynamically adjusted. Specifically, when operating parameters indicate that the water treatment equipment is in a stable operating state—for example, pump speed, valve opening, stirring intensity, and the rate of change of raw water turbidity all remain within a preset stable range—the system lowers the signal recognition threshold. This is to increase sensitivity to weak deviation signals, ensuring that even subtle probe anomalies are detected promptly. Conversely, when operating parameters indicate a change in operating conditions or the presence of external disturbances—for example, significant changes in pump speed or valve opening, or drastic fluctuations in raw water turbidity—the system raises the signal recognition threshold. This is to suppress misjudgments caused by changes in operating conditions or noise, avoiding the misidentification of normal process fluctuations as probe deviations.

[0118] In practical applications, time-domain feature matching and frequency-domain feature matching are performed on the signal components in the multi-dimensional feature vector information to accurately identify the probe deviation signal components from complex signals. Time-domain feature matching focuses on features such as the instantaneous amplitude, duration, rising edge, and falling edge of the signal, while frequency-domain feature matching focuses on the signal's spectral distribution, specific frequency energy, and harmonic components. When these matching degrees reach the dynamically adjusted signal identification threshold information, the corresponding signal component is confirmed as the probe deviation signal component.

[0119] Furthermore, protective enhancement processing is performed on the identified probe deviation signal components. This processing aims to strengthen these critical deviation signals, making them less susceptible to being overwhelmed or lost in subsequent processing. The protective enhancement processing specifically includes amplitude amplification, time-domain smoothing, and frequency-domain filtering. Amplification increases the strength of the deviation signal, making it easier to detect and analyze; time-domain smoothing removes high-frequency random noise from the deviation signal while preserving its main trend; and frequency-domain filtering specifically filters out interference within a particular frequency range, further purifying the deviation signal.

[0120] Finally, the probe deviation signal component after protective enhancement processing is fused with the remaining feature vector information that was not identified as probe deviation signal component to obtain the final feature vector information. This fusion ensures that while protecting and enhancing critical deviation information, other useful process operation information is also preserved, providing comprehensive and accurate input for subsequent process operation risk assessment.

[0121] Optionally, the protective enhancement process steps are optimized, specifically including:

[0122] Adjust the intensity of the enhanced processing based on the operating parameters;

[0123] Based on the adjusted enhancement intensity, perform the corresponding protective enhancement processing;

[0124] Real-time monitoring of the difference between the characteristic components of the probe deviation signal after protective enhancement processing and the normal reference characteristics;

[0125] Based on the degree of difference, feedback adjustments are made to the effectiveness of the protective enhancement treatment.

[0126] Specifically, enhanced processing intensity refers to the specific parameter settings when performing operations such as amplitude amplification, time-domain smoothing, and frequency-domain filtering on the probe deviation signal components. Examples include the amplitude amplification factor, the window size for time-domain smoothing, or the cutoff frequency of the filter. These parameters are adjusted based on continuously acquired operating parameters, which may include pump speed information, valve opening information, reaction tank stirring intensity information, and raw water turbidity change rate information. When operating parameters indicate changes in the operating conditions, such as a sharp increase in raw water turbidity or a significant adjustment in pump speed, the system will correspondingly adjust the enhanced processing intensity to adapt to the new operating conditions.

[0127] The normal baseline characteristics can be understood as the typical characteristic patterns exhibited by the probe's original signal when the probe is in a healthy state and the process is operating normally. These characteristic patterns can be established through historical data analysis, expert experience, or physical models. Real-time monitoring of the difference between the characteristic components of the probe deviation signal after protective enhancement treatment and the normal baseline characteristics aims to assess whether the enhancement treatment has achieved the expected effect, that is, protecting the deviation signal while avoiding the introduction of excessive distortion or significant deviation from the normal signal pattern.

[0128] In practical applications, the degree of difference can be measured using various quantitative indicators, such as Euclidean distance, correlation coefficient, or anomaly detection scores based on machine learning. Once the difference exceeds a preset range or exhibits a specific pattern, the system will perform feedback adjustments based on the effect of the protective enhancement processing. For example, if the difference is too large, it may indicate that the enhancement processing intensity is too high, in which case the system will reduce the amplitude amplification factor or adjust the filtering parameters; if the difference is too small, it may indicate that the enhancement processing intensity is insufficient, in which case the system will increase the enhancement processing intensity accordingly.

[0129] Optionally, the step of real-time monitoring of the difference between the characteristic components of the probe deviation signal after protective enhancement processing and the normal reference characteristics includes:

[0130] Continuously acquire operating parameters and raw probe signals;

[0131] Based on the operating parameters and the original probe signal, the characteristic components of the probe deviation signal after protective enhancement processing are extracted, and the normal reference features are also extracted.

[0132] The probe deviation signal feature components after protective enhancement processing are compared with normal reference features in real time to generate a degree of difference and identify difference pattern information.

[0133] When the difference pattern information presents a new pattern that is inconsistent with the historical case rules in the case library, the current operating parameters, the characteristic information of the probe's original signal, and the difference pattern information are recorded. Based on the recorded current operating parameters, the characteristic information of the probe's original signal, and the difference pattern information, the difference evaluation parameters are adjusted, and the normal baseline characteristics are updated.

[0134] Based on the adjusted difference assessment parameters and the updated normal baseline characteristics, a difference assessment is performed to obtain the difference assessment results.

[0135] Based on the current contextual feature data, difference pattern information, and difference degree assessment results, a similarity comparison is performed with the case database to obtain similar case retrieval results;

[0136] When the similar case search results indicate that no similar cases were found, the current contextual feature data and difference assessment results are stored as a new case.

[0137] Specifically, the steps for real-time monitoring of the difference between the characteristic components of the probe deviation signal after protective enhancement processing and the normal reference characteristics include the following steps:

[0138] First, operating parameters and raw probe signals are continuously acquired. Operating parameters may include, but are not limited to, pump speed, valve opening, reaction tank stirring intensity, and raw water turbidity change rate; these parameters reflect the current operating condition of the water treatment equipment. Raw probe signals are the unprocessed signals directly output by the flow measurement probe, containing rich time and frequency domain information. Continuously acquiring this data ensures the real-time nature and accuracy of the difference assessment, providing a foundation for subsequent feature extraction and comparison.

[0139] Secondly, based on the operating parameters and the original probe signal, the characteristic components of the probe deviation signal after protective enhancement processing are extracted, and normal reference features are also extracted. The characteristic components of the probe deviation signal after protective enhancement processing refer to the signal characteristics that more clearly reflect the probe deviation after processing such as amplitude amplification, time-domain smoothing, and frequency-domain filtering. Normal reference features refer to the signal characteristics under normal probe operation and without deviation, which can be established based on historical data, expert experience, or model predictions. Extracting these features aims to transform the complex information in the original signal into comparable and quantifiable indicators.

[0140] Next, the characteristic components of the probe deviation signal after protective enhancement processing are compared with the normal reference characteristics in real time to generate a difference degree and identify difference pattern information. The difference degree is a quantitative indicator that measures the degree of deviation between the current probe deviation signal characteristics and the normal reference characteristics, and can be calculated using various methods such as Euclidean distance, cosine similarity, and Mahalanobis distance. Difference pattern information refers to the specific manifestation of the difference degree in time, frequency, or feature space, such as periodic fluctuations, instantaneous spikes, or trend drifts. Identifying difference patterns helps to gain a deeper understanding of the nature and causes of probe deviation.

[0141] Furthermore, when the difference pattern information presents a new pattern that is inconsistent with the historical case rules in the case library, the current operating parameters, the characteristic information of the probe's original signal, and the difference pattern information are recorded. Based on the recorded current operating parameters, the characteristic information of the probe's original signal, and the difference pattern information, the difference assessment parameters are adjusted, and the normal benchmark features are updated. The case library stores historical probe deviation cases and their corresponding operating parameters, original signal characteristics, and difference patterns. When a new, unseen difference pattern appears, the system identifies it as a "new pattern" and automatically records the relevant data. Based on this newly recorded data, the system can adaptively adjust the weights, thresholds, and other parameters of the difference assessment to improve the ability to identify new patterns. At the same time, the normal benchmark features are also updated according to new operating conditions and data to ensure the dynamic adaptability of the benchmark.

[0142] Based on this, a difference assessment is performed using the adjusted difference assessment parameters and updated normal baseline characteristics to obtain the difference assessment results. This step ensures the accuracy and real-time nature of the difference assessment, reflecting the true state of the current probe deviation.

[0143] Subsequently, based on the current contextual characteristic data, difference pattern information, and difference assessment results, a similarity comparison is performed with the case database to obtain similar case retrieval results. Contextual characteristic data includes raw water dissolved organic matter characteristic data, probe microscopic signal spectrum characteristic data, and upstream environmental macroscopic data, providing comprehensive information about the current operating environment. By comparing the current deviation pattern with historical cases in the case database, it can be determined whether it is similar to known probe faults or anomalies, thus providing a basis for subsequent fault diagnosis and handling.

[0144] Finally, when the similar case search results indicate that no similar cases were found, the current contextual feature data and difference assessment results are stored as a new case. This mechanism enables the system to continuously learn and accumulate new experience, expand the case library, and thus improve its ability to identify and process unknown or complex probe deviation patterns, forming a continuously evolving intelligent monitoring system.

[0145] This application proposes an intelligent control system for the operating parameters of water treatment equipment, used to perform intelligent control of the operating parameters of water treatment equipment, combined with... Figure 3 As shown, the intelligent control system 1 for the operating parameters of the water treatment equipment includes:

[0146] The operating parameter acquisition module 11 is used to acquire the operating parameters of the water treatment equipment; the operating parameters include probe flow information and water quality information reported by the flow measurement probe;

[0147] The preliminary verification execution module 12 is used to perform preliminary verification on the operating parameters to determine the validity of the operating parameters;

[0148] The reliability assessment execution module 13 is used to perform reliability assessment on the probe flow information in the preliminarily verified operating parameters. The reliability assessment includes at least: deriving a first reference value based on the physical principles of the water treatment process. The first reference value is a reference flow rate derived from the physical principles of the water treatment process, equipment structural parameters, and operating condition parameters. This reference flow rate is used to verify the probe flow information and compare it with the probe flow information to obtain first comparison information; generating process effect comparison information based on water quality information, probe flow information, and reagent dosing information, and comparing the process effect comparison information with the actual process treatment effect to obtain second comparison information; and generating reliability assessment result information based on the first and second comparison information.

[0149] The change monitoring module 14 is used to monitor the change trend of the original signal of the flow measurement probe and obtain the change information of water quality information.

[0150] The risk assessment module 15 is used to assess the process operation risk based on the reliability assessment results, changes in water quality information, and the proximity of the water treatment equipment operating parameters to the preset minimum energy consumption critical point.

[0151] The control command sending module 16 is used to adjust the redundancy of the water treatment equipment operating parameters based on the process operation risk, and send the control command corresponding to the adjusted water treatment equipment operating parameters to the actuator.

[0152] The water treatment equipment mentioned in this application can refer to various process units and supporting devices in a water treatment plant used to complete the raw water purification process, including coagulation tanks, sedimentation tanks, filtration units, disinfection units, and related pumps, valves, agitators, metering devices, etc. These devices operate collaboratively according to a predetermined process flow, achieving water purification through physical, chemical, and biological processes. Operating parameters refer to measurable data collected in real time or periodically during the water treatment process that reflect the operating status of the equipment and the water quality, including flow rate, pressure, temperature, pH value, turbidity, conductivity, dissolved oxygen, residual chlorine, and reagent dosage. Among them, probe flow information refers to the instantaneous flow rate value or cumulative flow rate value directly output by the flow measurement probe, while water quality information covers the physical, chemical, and biological indicators of raw water, effluent from each process stage, and final effluent.

[0153] Preliminary verification refers to a basic validity check performed on operating parameters before they enter the control decision chain to eliminate obviously erroneous data. This verification includes not only range verification but also data continuity verification, timestamp consistency verification, and anomaly jump identification. For example, when traffic data undergoes drastic changes exceeding the physically possible range within a very short period, it can be marked as an anomaly candidate. This step prevents obviously erroneous data from entering the subsequent reliability assessment stage.

[0154] Reliability assessment is a key technical aspect of this application, which evaluates the authenticity and reliability of probe flow information through a multi-source cross-validation mechanism. The first reference value is the theoretical flow rate calculated based on a physical model of the water treatment process. This model comprehensively considers parameters such as pipe diameter, pipe length, local resistance coefficient, pump head curve, pump speed, and valve opening, and uses fluid dynamics formulas to inversely deduce the flow rate. The deviation between the first reference value and the probe flow information constitutes the first comparison information. The process effect comparison information is generated by analyzing water quality information, probe flow information, and reagent dosage to calculate the treatment efficiency or reagent utilization rate of the process section, and comparing it with the preset ideal effect or historical best effect to form the second comparison information. A comprehensive analysis of the first and second comparison information generates the reliability assessment result.

[0155] Information on changes in water quality is obtained by analyzing the time and frequency domain characteristics of the probe's raw signal. For example, by analyzing the signal mean drift trend, changes in spectral energy distribution, or anomalies in harmonic components, changes in suspended solids concentration, bubble interference, or scaling on the probe surface can be indirectly identified. This information is used to help determine whether probe deviation is caused by environmental factors or changes in its own condition.

[0156] The preset minimum energy consumption critical point refers to the theoretical minimum energy consumption state of the system under the condition that the effluent water quality meets the standards. This critical point is usually obtained through historical data statistics or optimization model calculation. When the operating parameters approach this critical point, the system's safety margin decreases, and its tolerance for parameter anomalies decreases, thus requiring a more rigorous risk assessment.

[0157] Redundancy is the operational space reserved by the control system to cope with uncertainties, such as the safety margin for reagent dosing or the pump frequency adjustment range. When the risk of process operation increases, increasing redundancy can improve the system's immunity to disturbances.

[0158] The implementation environment of this application is a distributed control system or programmable logic controller system that integrates sensor networks, actuators, and a central control unit. The modules interact with each other via industrial Ethernet or fieldbus to form a closed-loop control link.

[0159] The operating parameter acquisition module collects probe flow and water quality information, formats it uniformly, and transmits it to the central control unit. This module can acquire analog signals via a data acquisition card or digital data from a PLC or DCS system via a communication protocol. In addition to data acquisition, this module also performs timestamp processing on the data to ensure synchronous analysis of multi-source data.

[0160] The preliminary verification execution module receives the data output by the running parameter acquisition module, performs rule verification, trend verification, and continuity verification on it, and transmits the valid data to the reliability evaluation execution module, while marking and recording abnormal data.

[0161] The reliability assessment module performs physical model back-calculation and process effect comparison on the probe flow information, and outputs the reliability assessment results. This module may contain a physical calculation submodule, a process effect evaluation submodule, and a comprehensive judgment submodule. The comprehensive judgment submodule generates a reliability level score based on the first comparison information and the second comparison information.

[0162] The change monitoring module continuously analyzes the raw signals from the probe, generating information on changes in water quality. This information, along with the reliability assessment results, is then input into the operational risk assessment module. This continuous monitoring mechanism ensures that the system can promptly detect gradual anomalies, rather than simply responding to sudden events.

[0163] The operational risk assessment module generates a process operation risk level based on reliability assessment results, changes in water quality information, and the distance between operating parameters and the preset minimum energy consumption threshold. This risk assessment does not consider a single indicator but adopts a comprehensive evaluation method, such as logically combining reliability scores, water quality fluctuation indices, and energy consumption margins to obtain the final risk level.

[0164] The control command sending module adjusts redundancy based on the process operation risk level and generates corresponding control commands to send to the actuators. For example, when the risk level increases, the redundancy ratio of reagent dosing is appropriately increased or the pump operating frequency margin is increased; when the risk decreases, the system gradually returns to energy-saving operation. This module ensures that control commands undergo logical verification and safety checks before execution to prevent malfunctions.

[0165] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for intelligent control of operating parameters of water treatment equipment, characterized in that, include: Obtain the operating parameters of the water treatment equipment; the operating parameters include probe flow information and water quality information reported by the flow measurement probe; A preliminary verification is performed on the operating parameters to determine their validity; A reliability assessment is performed on the probe flow information in the preliminarily verified operating parameters; The reliability assessment includes at least the following steps: First, a first reference value is derived based on the physical principles of the water treatment process. This first reference value is a reference flow rate derived from the physical principles of the water treatment process, equipment structural parameters, and operating condition parameters. This reference flow rate is used to verify the probe flow rate information and is compared with the probe flow rate information to obtain first comparison information. Second, process effect comparison information is generated based on the water quality information, the probe flow rate information, and the reagent dosing information. This process effect comparison information is then compared with the actual process treatment effect to obtain second comparison information. Finally, a reliability assessment result is generated based on the first and second comparison information. By monitoring the change trend of the original signal from the flow measurement probe, information on changes in water quality can be obtained. Based on the reliability assessment results, changes in water quality information, and the degree to which the operating parameters of the water treatment equipment are close to the preset minimum energy consumption critical point, the risk of process operation is assessed. Based on the aforementioned process operation risks, the redundancy of the water treatment equipment's operating parameters is adjusted, and control commands corresponding to the adjusted water treatment equipment operating parameters are sent to the actuator.

2. The intelligent control method for operating parameters of a water treatment equipment according to claim 1, characterized in that, The steps for performing a reliability assessment on the probe flow information in the preliminarily verified operating parameters include: Acquire contextual characteristic data; the contextual characteristic data includes raw water dissolved organic matter characteristic data, probe micro-signal spectrum characteristic data, and upstream environmental macro-data; Based on the aforementioned contextual feature data, the evidence weight information of the three types of evidence sources is adjusted; the three types of evidence sources include flow rate back-calculation verification evidence sources, process effect evidence sources, and probe health evidence sources. When the adjusted three types of evidence sources form a conflict, a conflict resolution is performed on the conflicting information based on the process operation risk information and priority rules to determine whether there is a probe deviation in the probe flow information, and to obtain the conflict resolution result information. Based on the conflict resolution results, deviation isolation and data trust reconstruction are performed to provide input for the assessment of subsequent process operation risks. When the conflict resolution result information indicates that the deviation is due to inaccuracy in the process model, an adaptive adjustment is performed on the process model.

3. The intelligent control method for operating parameters of a water treatment equipment according to claim 2, characterized in that, The step of performing a reliability assessment on the probe flow information in the preliminarily verified operating parameters further includes: Time-frequency analysis was performed on the probe's raw signal to identify and quantify the nonlinear signal distortion characteristics caused by the micro-adsorption layer or wear on the probe surface; Establish decision tree rules based on multidimensional contextual feature parameters, and adjust the evidence weight information according to the combination state of contextual feature data in the nonlinear signal distortion features; Based on the adjusted evidence weight information, a multi-attribute decision analysis method is adopted to comprehensively evaluate the information on evidence conflicts by taking into account the contribution of evidence, the risk level of process operation, and the preset expert experience rules, and output the conflict adjudication result information.

4. The intelligent control method for operating parameters of a water treatment equipment according to claim 2, characterized in that, The steps for performing a reliability assessment on the probe flow information in the preliminarily verified operating parameters include: Continuously acquire contextual feature data; Time-domain and frequency-domain feature information is extracted from the original signal from the probe to capture instantaneous changes and spectral distribution information of the signal; The steps of constructing and updating evidence weight adjustment rules include: monitoring the deviation of the contextual feature data relative to historical data to identify new deviation patterns; and when a new deviation pattern is identified, adjusting the evidence weight information according to the new deviation pattern. Based on the adjusted evidence weighting information, calculate the evidence contribution of each source of evidence; A conflict resolution mechanism based on case-based reasoning is adopted. The conflict resolution mechanism includes: when conflicting evidence information occurs, retrieving historical cases from a preset case library that are identical to the current situational feature data and the conflicting evidence information; when identical historical cases are retrieved, outputting conflict resolution result information based on the resolution result information and actual effect information of the historical cases; when identical historical cases are not retrieved, outputting preliminary resolution result information based on the current process operation risk level, preset expert experience rules, and evidence contribution, and storing the current situational feature data, the conflicting evidence information, and the preliminary resolution result information as a new case. Based on the conflict resolution results, deviation isolation and data trust reconstruction are performed. When the discrepancy in the conflict resolution results is attributed to an inaccurate process model, an adaptive adjustment is performed on the process model.

5. The intelligent control method for operating parameters of a water treatment equipment according to claim 4, characterized in that, The steps for extracting time-domain and frequency-domain feature information from the probe's raw signal include: Continuously receive the raw probe signal transmitted by the flow measurement probe; The original signal from the probe is sampled in real time and divided into continuous time windows; Within each time window, the root mean square value, peak factor, waveform factor are calculated, and the instantaneous amplitude, rise rate, and fall rate are monitored to obtain time-domain characteristic information. Spectral analysis is performed on the probe's raw signal within each time window to obtain frequency domain characteristics of power spectral density, energy in a specific frequency band, harmonic component intensity, and spectral centroid. The time-domain feature information and the frequency-domain feature information are fused to form a multi-dimensional feature vector information, and noise suppression is performed on the multi-dimensional feature vector information to obtain the processed multi-dimensional feature vector information; The nonlinear signal distortion characteristics indicating the micro-adsorption layer or wear on the probe surface are identified and quantified from the processed multidimensional feature vector information.

6. The intelligent control method for operating parameters of a water treatment equipment according to claim 5, characterized in that, The noise suppression steps include: Continuously acquire operating parameters that reflect the current operating conditions; Based on the aforementioned operating parameters, noise statistical characteristics are evaluated in real time; Based on the operating parameters and the noise statistical characteristics, the noise suppression strategy is adjusted; when the noise is non-Gaussian distributed or non-stationary, nonlinear filtering or adaptive thresholding is used; when the noise and the probe deviation signal overlap in the frequency domain, they are distinguished by combining the signal time-domain sparsity or local energy characteristics. During noise suppression, protective processing is performed on the probe deviation signal component in the multidimensional feature vector information that indicates probe deviation; A consistency check is performed on the noise-suppressed multidimensional feature vector information to ensure that the multidimensional feature vector information is consistent with the physical laws and logical relationships of the water treatment process.

7. The intelligent control method for operating parameters of a water treatment equipment according to claim 6, characterized in that, The step of performing protective processing on the probe deviation signal component indicating probe deviation in the multidimensional feature vector information includes: Continuously acquire operating parameters, including pump speed information, valve opening information, reaction tank stirring intensity information, and raw water turbidity change rate information; Based on the operating parameters, the signal recognition threshold information is adjusted; wherein, when the operating parameters indicate a stable operating state, the signal recognition threshold information is reduced to improve the sensitivity to weak deviation signals; when the operating parameters indicate a switching or disturbance state, the signal recognition threshold information is increased to suppress misjudgment. Time-domain feature matching and frequency-domain feature matching are performed on the signal components in the multidimensional feature vector information, and when the matching degree reaches the signal recognition threshold information, the corresponding signal component is identified as the probe deviation signal component; Protective enhancement processing is performed on the identified probe deviation signal components; the protective enhancement processing includes amplitude amplification, time-domain smoothing, and frequency-domain filtering; The probe deviation signal component after protective enhancement processing is fused with the remaining feature vector information that was not identified as a probe deviation signal component to obtain the final feature vector information.

8. The intelligent control method for operating parameters of a water treatment equipment according to claim 7, characterized in that, The steps of the protective enhancement process include: Adjust the enhanced processing intensity based on the aforementioned operating parameters; Based on the adjusted enhancement intensity, perform the corresponding protective enhancement processing; Real-time monitoring of the difference between the characteristic components of the probe deviation signal after protective enhancement processing and the normal reference characteristics; Based on the degree of difference, feedback adjustments are made to the effectiveness of the protective enhancement treatment.

9. The intelligent control method for operating parameters of a water treatment equipment according to claim 8, characterized in that, The step of real-time monitoring of the difference between the characteristic components of the probe deviation signal after protective enhancement processing and the normal reference characteristics includes: Continuously acquire operating parameters and raw probe signals; Based on the operating parameters and the original probe signal, the characteristic components of the probe deviation signal after protective enhancement processing are extracted, and the normal reference features are extracted. The probe deviation signal feature components after protective enhancement processing are compared with normal reference features in real time to generate a degree of difference and identify difference pattern information. When the difference pattern information presents a new pattern that is inconsistent with the historical case rules in the case library, the current operating parameters, the feature information of the probe's original signal, and the difference pattern information are recorded. Based on the recorded current operating parameters, the feature information of the probe's original signal, and the difference pattern information, the difference degree evaluation parameters are adjusted, and the normal benchmark features are updated. Based on the adjusted difference assessment parameters and the updated normal baseline characteristics, a difference assessment is performed to obtain the difference assessment results. Based on the current contextual feature data, the difference pattern information, and the difference degree evaluation result information, a similarity comparison is performed with the case library to obtain similar case retrieval results; When the similar case retrieval result indicates that no similar cases were found, the current contextual feature data and difference assessment result information are stored as a new case.

10. An intelligent control system for the operating parameters of a water treatment equipment, used to perform intelligent control of the operating parameters of the water treatment equipment, characterized in that, include: The operating parameter acquisition module is used to acquire the operating parameters of the water treatment equipment; The operating parameters include probe flow information and water quality information reported by the flow measurement probe; The preliminary verification execution module is used to perform a preliminary verification on the operating parameters to determine the validity of the operating parameters; The reliability assessment execution module is used to perform a reliability assessment on the probe flow information in the preliminarily verified operating parameters; The reliability assessment includes at least the following steps: First, a first reference value is derived based on the physical principles of the water treatment process. This first reference value is a reference flow rate derived from the physical principles of the water treatment process, equipment structural parameters, and operating condition parameters. This reference flow rate is used to verify the probe flow rate information and is compared with the probe flow rate information to obtain first comparison information. Second, process effect comparison information is generated based on the water quality information, the probe flow rate information, and the reagent dosing information. This process effect comparison information is then compared with the actual process treatment effect to obtain second comparison information. Finally, a reliability assessment result is generated based on the first and second comparison information. The change monitoring module is used to monitor the change trend of the original signal of the flow measurement probe and obtain information on the change of water quality. The risk assessment module is used to assess the operational risk of the process based on the reliability assessment results, changes in water quality information, and the proximity of the water treatment equipment operating parameters to the preset minimum energy consumption critical point. The control command sending module is used to adjust the redundancy of the water treatment equipment operating parameters based on the process operation risk, and send the control command corresponding to the adjusted water treatment equipment operating parameters to the actuator.

Images