Circulating water monitoring method and system, product and medium

Abstract

A circulating water monitoring method, system, product and medium. The method comprises the following steps: collecting real-time circulating water operation data; inputting the real-time circulating water operation data into a preset circulating water data prediction model; after the circulating water data prediction model generates a preliminary prediction result, calculating a real-time chemical equilibrium constant according to the preliminary prediction result and a preset chemical equilibrium equation; under the condition that the real-time chemical equilibrium constant difference value is larger than a preset chemical equilibrium constant difference value threshold value, the preliminary prediction result is adjusted, and predicted circulating water operation data are obtained; the predicted circulating water data is compared with a circulating water operation data range threshold value, and whether real-time abnormal operation data exceeding the circulating water operation data range threshold value exists or not is judged; and if yes, determining a real-time abnormal condition by combining the real-time abnormal data with the causal relationship Bayesian network. By implementing the technical scheme provided by the invention, the abnormal condition of the circulating water system can be found in time.

Description

Technical Field

[0001] This application relates to the technical field of circulating water treatment, and particularly to a circulating water monitoring method, system, product, and medium. Background Art

[0002] With the booming development of industry, circulating water systems play an indispensable role in many fields such as industrial production, energy supply, and building refrigeration. In industrial production, circulating water systems can effectively remove the heat generated by equipment operation and ensure the stable operation of equipment; in the field of energy supply, they can improve energy conversion efficiency and reduce energy consumption; in building refrigeration, circulating water systems achieve efficient heat exchange and create a comfortable environment for people.

[0003] Currently, circulating water monitoring and treatment involve real-time monitoring and recording of current water quality parameters. After real-time detection of problems in circulating water quality data, appropriate solutions will be formulated based on the abnormal situation of the circulating water to treat the circulating water.

[0004] However, in the actual process of circulating water monitoring, although circulating water data is continuously detected, before obvious abnormal conditions occur in the circulating water, relevant data often fluctuates and is difficult to be recognized as abnormal. As a result, when the data finally exceeds the threshold and is determined to be abnormal by the monitoring system, and then corresponding treatment solutions are taken, the best treatment opportunity is often missed, and the stable operation of the circulating water system cannot be guaranteed in time. Summary of the Invention

[0005] This application provides a circulating water monitoring method, system, product, and medium for timely detecting abnormal situations in circulating water systems.

[0006] In the first aspect of this application, a circulating water monitoring method is provided, and the method includes: Collect real-time circulating water operation data; the real-time circulating water operation data includes real-time pH value, real-time hardness value, real-time turbidity value, real-time conductivity, and real-time dissolved oxygen content value; input the real-time circulating water operation data into a preset circulating water data prediction model; the circulating water data prediction model is obtained by pre-training with a historical circulating water operation data set; after the preliminary prediction result is generated by the circulating water data prediction model, calculate the real-time chemical equilibrium constant according to the preliminary prediction result and the preset chemical equilibrium equation; in the case where the difference between the real-time chemical equilibrium constants is greater than the preset chemical equilibrium constant difference threshold, adjust the preliminary prediction result to obtain the predicted circulating water operation data; the difference between the real-time chemical equilibrium constants is the value obtained by taking the absolute value of the difference between the real-time chemical equilibrium constant and the preset standard chemical equilibrium constant; the predicted circulating water operation data is such that the standard chemical equilibrium constant can be calculated according to the chemical equilibrium equation; compare the predicted circulating water data with the circulating water operation data range threshold to determine whether there is real-time abnormal operation data exceeding the circulating water operation data range threshold; if so, determine the real-time abnormal situation through the real-time abnormal data combined with the causal relationship Bayesian network; the causal relationship Bayesian network is a Bayesian network that reflects the causal relationship between the various parameters of the circulating water and is generated in advance based on the historical circulating water parameter data sets.

[0007] In the above embodiment, by collecting various key operation data of the real-time circulating water and inputting them into a pre-trained prediction model, the model can learn the data law based on past experience and can preliminarily predict the operation trend of the circulating water; calculate and adjust the prediction result in combination with the chemical equilibrium equation. Since the circulating water system involves various chemical reactions, the chemical equilibrium equation can ensure that the prediction result conforms to the actual chemical law and makes the prediction more accurate; compare the prediction result with the threshold to determine whether there is an abnormality, and use the causal relationship Bayesian network to determine the abnormal situation and locate the root cause of the abnormality, so that the operation status of the circulating water system can be monitored in real time and accurately. Since the prediction model combined with the chemical equilibrium equation ensures the prediction accuracy, the threshold comparison can detect abnormalities in a timely manner, and the Bayesian network can clarify the cause of the abnormality, so potential abnormalities can be discovered in a timely manner, and early knowledge of the abnormal situation provides a basis for taking effective measures subsequently, thereby reducing the risks caused by circulating water abnormalities and ensuring the stable operation of the circulating water system.

[0008] Combined with some embodiments of the first aspect, in some embodiments, if so, after determining the real-time abnormal situation through the real-time abnormal data combined with the causal relationship Bayesian network, it further includes: Retrieve a preset number of historical abnormal solution plans; for each historical abnormal solution plan, calculate the solution score corresponding to the historical abnormal solution plan in combination with the real-time circulating water operation data; sort all the solution scores from high to low, and the historical abnormal solution plan corresponding to the highest solution score after sorting is the optimal abnormal solution plan; execute the optimal abnormal solution plan.

[0009] In the above embodiment, by retrieving a preset number of historical abnormal solution plans, calculating the scores of each plan in combination with the real-time circulating water operation data, and then sorting the scores to select the optimal abnormal solution plan and execute it. When facing the abnormality of the circulating water system, the solution plan that best suits the current situation can be quickly screened out from past experiences. This not only improves the efficiency of handling abnormalities, reduces the decision-making time, but also fully draws on historical successful experiences, enhances the accuracy and effectiveness of abnormality handling, and reduces production losses and equipment wear caused by abnormalities.

[0010] Combined with some embodiments of the first aspect, in some embodiments, for each historical abnormal solution plan, calculating the corresponding solution score in combination with the real-time circulating water operation data specifically includes: Retrieve the historical circulating water operation data corresponding to each historical abnormal solution plan; input each historical circulating water operation data and the real-time circulating water operation data into the first formula to calculate the solution score corresponding to each historical solution plan; where the first formula is: Where S i is the solution score of the i-th historical solution plan, n represents the number of index data in the real-time circulating water operation data. Here, n = 5, that is, the real-time pH value, real-time hardness value, real-time turbidity value, real-time conductivity, and real-time dissolved oxygen content value, ω j is the weight coefficient of the j-th index data, D j,r is the actual measured value of the index data in the j-th real-time circulating water operation data, D j,i is the value of the j-th index data in the i-th historical circulating water operation data, is the average value of the j-th index data in all historical circulating water operation data, α j 、β j and γ j are adjustment coefficients, Max j and Min j are the maximum and minimum values of the j-th index data in all historical circulating water operation data respectively.

[0011] In the above embodiments, the solution scores of each historical solution are calculated through formulas. First, the differences between the real-time data and individual historical data, and the differences between the real-time data and the historical average data are comprehensively considered, so as to reflect the similarity between the current and historical data from different levels. Then, the comprehensive difference is normalized to eliminate the influence brought by the differences in dimension and value range of different data indicators, ensuring that each indicator can participate in the calculation fairly. Then, weights are set according to the importance of each indicator to the circulating water system, weighted summation is performed, and by taking the reciprocal, the historical data with a smaller difference from the current situation gets a higher score. Through such a calculation structure, the similarity between the historical circulating water operation data and the current situation is comprehensively and accurately measured, making the calculation and evaluation of the solution score more comprehensive and accurate, and providing strong support for the stable operation of the circulating water system.

[0012] Combined with some embodiments of the first aspect, in some embodiments, after the circulating water data prediction model generates a preliminary prediction result and the preliminary prediction result is adjusted by applying the chemical equilibrium equation constraint to obtain the predicted circulating water operation data, it further includes: Obtain real-time actual ion concentration data; the real-time actual ion concentration data includes the real-time concentration data of all ions involved in the chemical equilibrium equation; according to the predicted circulating water operation data and the chemical equilibrium equation, calculate the real-time ion concentration data in the circulating water; the real-time ion concentration data includes the real-time concentration data of all ions involved in the chemical equilibrium equation; in the case where the concentration data difference of the same type of ions in the real-time ion concentration data and the real-time actual ion concentration data is greater than the preset ion concentration difference threshold, the predicted circulating water operation data is adjusted again according to the chemical equilibrium equation.

[0013] In the above embodiments, by obtaining the real-time actual concentration data of the ions involved in the chemical equilibrium equation in the circulating water, and at the same time calculating the real-time ion concentration data based on the predicted circulating water operation data and the chemical equilibrium equation, comparing the concentration data difference of the same type of ions between the two, when the difference is greater than the preset threshold, the predicted circulating water operation data is adjusted again according to the chemical equilibrium equation. This makes the predicted circulating water operation data more accurately reflect the actual situation. Because the ion concentration is a key indicator of the chemical state of the circulating water, by comparing the actual and calculated ion concentrations, prediction deviations can be found. Adjusting based on the chemical equilibrium equation is because the equation reflects the chemical reaction law of the circulating water, and readjustment can correct the deviation and make the predicted data fit the actual situation, providing a more reliable basis for the stable operation of the circulating water system.

[0014] Combined with some embodiments of the first aspect, in some embodiments, if there is any, after determining the real-time abnormal situation through the real-time abnormal data combined with the causal Bayesian network, it further includes: Extract multiple data related to the determined abnormal conditions from the real-time circulating water operation data and input them into a preset mathematical relationship function; the mathematical relationship function is a mathematical relationship function among multiple parameters constructed in advance based on historical circulating water operation data and historical abnormal conditions; calculate according to the mathematical relationship function to obtain the theoretical operation result corresponding to the real-time abnormal condition; calculate the difference index between the theoretical operation result and the real-time circulating water operation data; in the case that the difference index exceeds the preset verification threshold, add the theoretical operation result and the real-time circulating water operation data to the historical circulating water operation data set to obtain an updated historical circulating water operation data set; use the updated historical circulating water operation data set to train the circulating water data prediction model again.

[0015] In the above embodiments, by extracting data related to abnormalities from the real-time circulating water operation data, inputting the mathematical relationship function constructed based on historical data, calculating the theoretical operation result and comparing it with the real-time data, and adding the relevant data to the historical data set when the difference exceeds the threshold, and then training the prediction model with it. The circulating water data prediction model becomes more accurate and reliable. Because the mathematical relationship function can reflect the law among parameters, by comparing the real-time and theoretical results, the prediction deviation of the model can be found. Incorporating the deviation data into the historical data set allows the model to be exposed to more special situations, learn these abnormal features during training, and optimize the prediction logic, so as to more accurately judge the operating state of the circulating water in subsequent predictions.

[0016] Combined with some embodiments of the first aspect, in some embodiments, the collection of the real-time circulating water operation data specifically includes: Collect the circulating water operation data at a real-time collection frequency; calculate the fluctuation amplitude value between the real-time circulating water operation data and the nearest circulating water operation data; the real-time circulating water operation data is the latest collected circulating water operation data; the nearest circulating water operation data is the circulating water operation data collected in the previous time at the real-time collection frequency; in the case that the fluctuation amplitude value exceeds the fluctuation amplitude threshold, adjust the collection frequency of the circulating water operation data according to the real-time frequency multiple; the real-time frequency multiple is the difference between the fluctuation amplitude value and the fluctuation amplitude threshold.

[0017] In the above embodiments, by collecting the circulating water operation data at a fixed real-time frequency, calculating the fluctuation amplitude value between two adjacent collected data, and once the fluctuation amplitude exceeds the preset threshold, adjusting the collection frequency according to the difference between the exceeded amplitude and the threshold. This enables the system to flexibly adjust the data collection strategy according to the fluctuation of the circulating water operation data. Increasing the collection frequency when the data fluctuates greatly can capture the details of data changes in a timely manner and improve the monitoring sensitivity to abnormal conditions; maintaining a lower collection frequency when the data fluctuates little can reduce resource waste.

[0018] In some embodiments in combination with some embodiments of the first aspect, after executing the optimal exception solution, it further includes: Sending exception information to a preset mobile terminal device; the exception information includes real-time circulating water operation data and an optimal operation plan.

[0019] In the above embodiments, when an exception occurs in the circulating water system, the real-time circulating water operation data and the optimal operation plan for the exception are sent to a preset mobile terminal device. This enables relevant personnel, such as engineers and managers, to timely obtain the exception status and coping strategies of the circulating water system even if they are not at the monitoring site.

[0020] In a second aspect, an embodiment of the present application provides a circulating water monitoring system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the circulating water monitoring system to execute the method described in the first aspect and any possible implementation manner in the first aspect.

[0021] In a third aspect, an embodiment of the present application provides a computer program product containing instructions, when the above computer program product runs on a circulating water monitoring system, enabling the above circulating water monitoring system to execute the method described in the first aspect and any possible implementation manner in the first aspect.

[0022] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, including instructions, when the above instructions run on a circulating water monitoring system, enabling the above circulating water monitoring system to execute the method described in the first aspect and any possible implementation manner in the first aspect.

[0023] It can be understood that the circulating water monitoring system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the circulating water monitoring method provided in the embodiments of the present application. Therefore, the beneficial effects they can achieve can refer to the beneficial effects in the corresponding method, which will not be elaborated here.

[0024] One or more technical solutions provided in the embodiments of the present application at least have the following technical effects or advantages: 1. This application collects various key operating data of real-time circulating water, inputs them into a pre-trained prediction model, calculates and adjusts the prediction results in combination with chemical equilibrium equations, then compares with thresholds to determine whether there are abnormalities, and determines the abnormal conditions with the help of a causal relationship Bayesian network. This enables the operating status of the circulating water system to be monitored in real time and accurately, timely discovers potential abnormalities, provides a basis for subsequent effective measures, thereby reducing the risks caused by abnormal circulating water and ensuring the stable operation of the circulating water system.

[0025] 2. This application retrieves a preset number of historical abnormal solutions, calculates the scores of each solution in combination with real-time circulating water operating data, and then sorts the scores to select the optimal abnormal solution and execute it. This enables, when faced with abnormalities in the circulating water system, to quickly screen out the most suitable solution for the current situation from past experience. This not only improves the efficiency of handling abnormalities, reduces decision-making time, but also can fully draw on historical successful experiences, enhances the accuracy and effectiveness of abnormal handling, and reduces production losses and equipment wear caused by abnormalities.

[0026] 3. This application extracts data related to abnormalities from real-time circulating water operating data, inputs them into a mathematical relationship function constructed based on historical data, calculates the theoretical operating results and compares with real-time data. When the difference exceeds the threshold, the relevant data is added to the historical data set, and then the prediction model is trained with it. This makes the circulating water data prediction model more accurate and reliable. Because the mathematical relationship function can reflect the laws between parameters, by comparing the real-time and theoretical results, the prediction deviation of the model can be found. Incorporating the deviation data into the historical data set allows the model to encounter more special situations, learn these abnormal features during training, optimize the prediction logic, and thus more accurately judge the operating status of the circulating water in subsequent predictions. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 is a schematic diagram of a system architecture to which the circulating water monitoring method in the embodiment of this application can be applied; Figure 2 is a schematic flow diagram of the circulating water monitoring method in the embodiment of this application; Figure 3 is another schematic flow diagram of the circulating water monitoring method in the embodiment of this application; Figure 4 is a schematic diagram of an exemplary hardware structure of the circulating water monitoring system in the embodiment of this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] The terms used in the following embodiments of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. As used in the specification and appended claims of this application, the singular forms "a", "an", "the", "above-mentioned", "said", and "this" are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in this application refers to and includes any and all possible combinations of one or more of the listed items.

[0029] Hereinafter, the terms "first" and "second" are only used for descriptive purposes and should not be construed as implying or suggesting relative importance or implicitly indicating the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of this application, unless otherwise specified, the meaning of "a plurality" is two or more.

[0030] Figure 1 It is a schematic diagram of a system architecture applicable to the circulating water monitoring method in the embodiments of this application.

[0031] Please refer to Figure 1 , which includes a sensor group, a data storage device, a controller, an execution device, and a communication module in this system architecture.

[0032] The sensor group contains multiple sensors, which are used to collect the real-time operating data of the circulating water in real time, provide the basic circulating water quality data for the system, and help to grasp the quality status of the circulating water in real time.

[0033] The sensor group includes, but is not limited to, a pH sensor (such as the Hach HQ40d pH analyzer), a hardness sensor (such as the hardness detection module in the CM442 multi-parameter water quality analyzer), a turbidity sensor (such as the Hach 2100Q turbidimeter), a conductivity sensor (such as the Leici DDS-307A conductivity meter), and a dissolved oxygen sensor (such as the Mettler-Toledo InPro6800 dissolved oxygen sensor), etc. These sensors use different sensing principles to collect the real-time index data such as the pH, hardness, turbidity, conductivity, and dissolved oxygen content value of the circulating water during the operation process of the circulating water in real time. For example, the pH sensor determines the pH by measuring the hydrogen ion concentration in the water and converts the chemical signal into an electrical signal for output.

[0034] The data storage device is used to store a large amount of real-time data, historical data, system configuration information, various model training data, etc. It provides data support for the system, facilitating subsequent data query, analysis, and model training, and is the core component of the system's data management. Through built-in storage media (such as hard disks, flash memories, etc.) and corresponding data management software, the input data is classified, sorted, and stored. A relational database (such as MySQL) or a non-relational database (such as MongoDB) can be used to organize and manage the data, and an appropriate data storage structure and query method are selected according to the characteristics and usage requirements of the data.

[0035] The controller receives data from the sensor group, as well as information such as system preset rules and model analysis results, and performs logical judgment and decision-making. According to the judgment results, control instructions are sent to achieve the automatic control and adjustment of the circulating water monitoring system and the circulating water system, ensuring the stable operation of the circulating water system. The controller is usually based on hardware platforms such as programmable logic controllers (PLCs), industrial personal computers (IPCs), etc., and runs corresponding control software. By writing program codes (such as the ladder diagram program of the PLC or the control algorithm program running on the industrial computer), the functions of data processing, logical judgment, and instruction output are realized.

[0036] The execution device is an execution component that actually operates and adjusts the circulating water system according to the control instructions sent by the controller to achieve the system control target. The execution device usually consists of execution elements such as motors, solenoid valves, pumps, etc., as well as corresponding transmission mechanisms and control interfaces. When receiving the control signal (such as current, voltage signal) from the controller, the execution element acts according to the magnitude and direction of the signal, and realizes the physical adjustment of the circulating water system through the transmission mechanism. For example, the dosing pump drives the plunger or diaphragm to move through the motor, and quantitatively injects chemical agents into the circulating water; the electric valve drives the valve core to move through the motor, changing the opening of the valve, thereby adjusting the flow rate of the circulating water.

[0037] The communication module is used to realize data transmission and information interaction between various devices within the system and between the system and external remote terminal devices. It enables the system to remotely monitor and manage the circulating water system, facilitating operators to obtain system operation status information anytime and anywhere, and perform remote operation and control. The communication module can adopt wired communication technologies (such as Ethernet, RS-485, etc.) or wireless communication technologies (such as 4G / 5G, Wi-Fi, LoRa, etc.). According to the actual application scenario and requirements, an appropriate communication technology is selected, and the data is packaged, encoded, and then transmitted through the corresponding communication medium (such as network cable, wireless signal). At the receiving end, the communication module decodes and verifies the received data to ensure the accuracy and integrity of the data.

[0038] In the related art, although the water quality parameters of the circulating water are continuously monitored and recorded in real time, and solutions are formulated to deal with the circulating water when problems occur in the water quality data. However, before the circulating water actually shows obvious abnormal conditions, the water quality data will fluctuate, and it is difficult to be recognized as abnormal due to the unclear fluctuation characteristics. This leads to the situation that when the data finally exceeds the threshold and is determined to be abnormal by the monitoring system, the treatment plan is taken, but the best treatment opportunity has been missed, and the stable operation of the circulating water system cannot be guaranteed in time.

[0039] In the embodiment of the present application, after collecting the key real-time data of the circulating water, it is input into a prediction model trained by historical data. This model generates a preliminary prediction result by learning the laws of historical data. Then, the predicted value is adjusted in combination with the chemical equilibrium equation. The chemical reactions in the circulating water system are complex, and the chemical equilibrium equation can make the predicted value fit the actual chemical changes, improving the prediction accuracy. When an abnormality is found by comparing the predicted value with the threshold, the causal Bayesian network constructed in advance is used to determine the abnormal situation. The Bayesian network is constructed based on historical parameter data, presenting the causal relationships of various parameters and being used for the root cause of the abnormality. Thus, early detection and early treatment of the abnormalities in the circulating water system are realized, effectively guaranteeing the stable operation of the circulating water system, and effectively solving the problems in the related art that it is difficult to identify early data fluctuations and miss the best treatment opportunity.

[0040] Figure 2 It is a flow schematic diagram of using the circulating water monitoring method in the embodiment of the present application, including the following steps: S201. Collect the real-time operating data of the circulating water; Specifically, the real-time operating data of the circulating water includes the real-time pH value, real-time hardness value, real-time turbidity value, real-time conductivity, and real-time dissolved oxygen content value.

[0041] Collecting the real-time operating data of the circulating water is achieved by installing a variety of sensors in the circulating water system. These sensors include pH sensors, hardness sensors, turbidity sensors, conductivity sensors, and dissolved oxygen sensors, etc. Each sensor is responsible for real-time monitoring of specific parameters in the circulating water and converting the monitored data into electrical signals. These electrical signals are then sent to the controller or data processing unit through a data transmission module (such as a wired or wireless communication module).

[0042] S202. Input the real-time operating data of the circulating water into a preset circulating water data prediction model; Specifically, the circulating water data prediction model is trained in advance using a historical circulating water operation data set and can learn the operating laws of the circulating water system under different conditions.

[0043] Before inputting real-time circulating water operation data into the model, preprocessing is usually required. The preprocessing steps include data cleaning, normalization, denoising, etc. For example, the deviation normalization method is used to preprocess the sample data, and the data is scaled to a specific range, such as [0, 1], to improve the training and prediction efficiency of the model.

[0044] The circulating water data prediction model is trained in advance using a historical circulating water operation dataset. During the training process, various machine learning algorithms can be adopted, such as random forest, BP neural network, long short-term memory network (LSTM), and gated recurrent unit (GRU), etc. For example, random forest is an integrated machine learning method that makes predictions by generating multiple decision trees and combining their output values. LSTM and GRU are often used for time series prediction and can effectively process time-dependent data in the circulating water system.

[0045] The preprocessed real-time circulating water operation data is input into the trained model. The model uses its internal algorithms and trained parameters to predict future circulating water operation data.

[0046] S203. After the circulating water data prediction model generates a preliminary prediction result, calculate the real-time chemical equilibrium constant according to the preliminary prediction result and the preset chemical equilibrium equation; After the circulating water data prediction model generates a preliminary prediction result, first, retrieve the preset chemical equilibrium equation information, which is: Calculate the real-time chemical equilibrium constant according to the chemical equilibrium equation and the preliminary prediction result.

[0047] The chemical equilibrium constant (K) is a parameter that describes the concentration relationship of each substance in a chemical reaction at the equilibrium state. For this preset chemical equilibrium equation, Its chemical equilibrium constant calculation formula is: Among them, A, B, C, and D represent reactants and products in the chemical equation, and [A], [B], [C], and [D] respectively represent the concentrations of the corresponding reactants and products in the chemical equilibrium constant calculation formula, and a, b, c, and d are their stoichiometric coefficients.

[0048] For example, for the dissolution equilibrium of calcium carbonate (CaCO3), the chemical equilibrium equation is: Obtain the concentration values of each chemical substance from the preliminary prediction result. For example, obtain the concentration values of calcium ions (Ca2+) and carbonate ions (CO32-); substitute the concentration values in the preliminary prediction result into the chemical equilibrium equation to calculate the real-time chemical equilibrium constant. For example, for the above calcium carbonate dissolution equilibrium, the real-time chemical equilibrium constant K can be expressed as:

[0049] S204. When the difference between the real-time chemical equilibrium constant and the preset chemical equilibrium constant difference threshold is greater than zero, adjust the preliminary prediction result to obtain the predicted circulating water operation data; Specifically, the difference between the real-time chemical equilibrium constant and the preset standard chemical equilibrium constant is obtained by taking the absolute value of the difference between the real-time chemical equilibrium constant and the preset standard chemical equilibrium constant; the predicted circulating water operation data can be used to calculate the standard chemical equilibrium constant according to the chemical equilibrium equation.

[0050] First, calculate the real-time chemical equilibrium constant according to the preliminary prediction result and the preset chemical equilibrium equation. Then, take the absolute value of the difference between the real-time chemical equilibrium constant and the preset standard chemical equilibrium constant to obtain the difference between the real-time chemical equilibrium constant and the preset standard chemical equilibrium constant. The standard chemical equilibrium constant is calculated based on the thermodynamic data of the reaction and can also be obtained by referring to the literature or using a thermodynamic database. If the difference between the real-time chemical equilibrium constant and the preset chemical equilibrium constant difference threshold is greater than zero, it is considered that the preliminary prediction result needs to be adjusted.

[0051] In some embodiments of the present application, the specific steps for adjusting the preliminary prediction result are as follows: First, determine the adjustment direction according to the chemical equilibrium principle. If the real-time chemical equilibrium constant is less than the preset standard chemical equilibrium constant, it is necessary to increase the concentration of the product or decrease the concentration of the reactant to shift the equilibrium to the positive reaction direction; otherwise, it is necessary to decrease the concentration of the product or increase the concentration of the reactant to shift the equilibrium to the reverse reaction direction. Then, according to the chemical equilibrium equation, adjust the concentration values of each chemical substance in the preliminary prediction result. The specific adjustment method can be to increase or decrease the concentration of one or more chemical substances to make the real-time chemical equilibrium constant close to the preset standard chemical equilibrium constant. After adjusting the concentration value, recalculate the real-time chemical equilibrium constant to ensure that the difference between it and the preset standard chemical equilibrium constant is within the preset threshold. After adjusting the concentration, the index data value in the preliminary prediction result changes with the adjusted concentration. Compare the adjusted prediction result with the actual operation data to verify whether the adjusted prediction result meets the actual situation. If it meets, use the adjusted prediction result as the final prediction result. If it does not meet, it is necessary to adjust again.

[0052] In the above steps, by adjusting the preliminary prediction result when the difference between the real-time chemical equilibrium constant and the preset chemical equilibrium constant difference threshold is greater than zero, the system can ensure that the predicted circulating water operation data meets the requirements of the chemical equilibrium equation, making the prediction result more accurate and reliable, thereby improving the stability and safety of the circulating water system.

[0053] S205. Compare the predicted circulating water data with the threshold of the circulating water operation data range to determine whether there is real-time abnormal operation data exceeding the threshold of the circulating water operation data range; If so, execute the following step S206; If not, return to execute the above step S201; Compare each of the predicted index data of the circulating water, such as acidity and alkalinity, hardness, turbidity, conductivity, and dissolved oxygen content, etc., with the preset normal operation range threshold one by one. If a certain index in the predicted data exceeds its corresponding threshold range, the system determines that there is real-time abnormal operation data. In the case of abnormal operation data, execute the following steps to trace and determine the real-time abnormal situation; in the case of no abnormal operation data, return to the above data collection step, collect data again, and then determine whether there is abnormal operation data again.

[0054] S206. Determine the real-time abnormal situation based on the real-time abnormal data and the causal relationship Bayesian network.

[0055] Specifically, the causal relationship Bayesian network is generated in advance based on the historical circulating water parameter data sets and reflects the causal relationship between the circulating water parameters.

[0056] First, construct a causal relationship Bayesian network based on the historical circulating water parameter data sets. This network represents the causal relationship between the circulating water parameters through a directed acyclic graph (DAG), where each node represents a parameter and the edge represents the causal relationship. For example, acidity and alkalinity (pH value) may affect hardness and turbidity, and hardness may affect conductivity, etc. For each node, determine its conditional probability table (CPT), which represents the probability distribution of the value of this node given the values of its parent nodes (i.e., the cause nodes in the causal relationship). These probability distributions are obtained based on historical data statistics.

[0057] Input the real-time abnormal data into the causal relationship Bayesian network. These data include the real-time measured values of parameters such as acidity and alkalinity, hardness, turbidity, conductivity, and dissolved oxygen content. Using Bayes' theorem, update the probability distribution of each node according to the input real-time abnormal data. Specifically, by calculating the posterior probability of each node, determine which parameter changes are the main factors causing the abnormality. For example, if the real-time value of acidity and alkalinity deviates significantly from the normal range, the network will update the probability distribution of the nodes related to acidity and alkalinity to reflect this change. By analyzing the updated probability distribution, determine the specific cause of the abnormality. For example, if the posterior probability of acidity and alkalinity increases significantly and there are also significant changes in the posterior probabilities of hardness and turbidity, it can be inferred that the change in acidity and alkalinity may be the main cause of the abnormality of hardness and turbidity.

[0058] In the above steps, by combining real-time abnormal data with a causal Bayesian network, the system can accurately determine the specific causes of real-time abnormal situations. This enables the system not only to detect abnormalities but also to deeply analyze the root causes of the abnormalities, obtain accurate abnormal situations, and thus take more targeted measures for processing. This abnormal diagnosis method based on causal relationships improves the accuracy and efficiency of abnormal handling, avoids equipment damage and production interruptions caused by misjudgment or missed judgment, and further enhances the stability and operating efficiency of the circulating water system.

[0059] In the above embodiment, by collecting real-time circulating water operation data and inputting it into a preset circulating water data prediction model, using the preliminary prediction results generated by the model, and combining with a preset chemical equilibrium equation to calculate the real-time chemical equilibrium constant. When the difference in the real-time chemical equilibrium constant is greater than a preset threshold, the preliminary prediction results are adjusted to obtain predicted circulating water operation data that meets the chemical equilibrium requirements. Subsequently, the predicted data is compared with the circulating water operation data range threshold to determine whether there is abnormal data. If there is abnormal data, the specific cause of the abnormal situation is determined by combining real-time abnormal data with a causal Bayesian network. In the related art, the abnormal situation of the circulating water system can only be detected after relatively obvious abnormalities occur in the circulating water system, while the circulating water monitoring method proposed in this application enables the system to promptly discover and accurately diagnose the abnormal situation of the circulating water system, avoiding the further deterioration of the abnormal situation and ensuring the stable operation of the circulating water system.

[0060] In some other embodiments of the present application, after detecting an abnormal situation in the circulating water system, by using the circulating water monitoring method provided in the present application, the processing method corresponding to the abnormal situation can also be accurately matched.

[0061] As Figure 3 shown, it is another process schematic diagram provided by the embodiment of the present application. This method can be used in Figure 1 the system architecture shown, and includes the following steps: S301. Collect circulating water operation data at a real-time acquisition frequency; Collecting circulating water operation data at a real-time acquisition frequency means continuously obtaining the operation parameters of the circulating water system at a preset time interval using sensors and data acquisition devices.

[0062] S302. Calculate the fluctuation amplitude value between the real-time circulating water operation data and the most recent circulating water operation data; It can be understood that the real-time circulating water operation data is the most recently collected circulating water operation data; the most recent circulating water operation data is the circulating water operation data collected in the previous acquisition at the real-time acquisition frequency.

[0063] By comparing the recently collected real-time circulating water operation data with the most recently collected circulating water operation data from the previous time, calculate the difference between the two. The specific calculation method can adopt various ways. For example: Absolute difference method: Calculate the absolute difference between the real-time circulating water operation data and the most recently collected circulating water operation data, that is, "fluctuation amplitude value = real-time circulating water operation data - most recently collected circulating water operation data"; Relative difference method: Calculate the relative difference between the real-time circulating water operation data and the most recently collected circulating water operation data, that is, "fluctuation amplitude value = (real-time circulating water operation data - most recently collected circulating water operation data) / most recently collected circulating water operation data × 100%"; Moving average method: Calculate the moving average value of the real-time circulating water operation data and the most recently collected circulating water operation data, and then calculate the difference between the two.

[0064] S303. In the case where the fluctuation amplitude value exceeds the fluctuation amplitude threshold, adjust the acquisition frequency of the circulating water operation data according to the real-time frequency multiple; Specifically, the real-time frequency multiple is the difference between the fluctuation amplitude value and the fluctuation amplitude threshold.

[0065] In the case where the fluctuation amplitude value exceeds the fluctuation amplitude threshold, adjust the acquisition frequency of the circulating water operation data according to the real-time frequency multiple. First, calculate the real-time frequency multiple: The real-time frequency multiple is the difference between the fluctuation amplitude value and the fluctuation amplitude threshold. That is, real-time frequency multiple = fluctuation amplitude value - fluctuation amplitude threshold. Adjust the acquisition frequency: According to the real-time frequency multiple, adjust the acquisition frequency of the circulating water operation data. If the real-time frequency multiple is positive, it means the fluctuation amplitude value is greater than the fluctuation amplitude threshold, and the acquisition frequency needs to be increased; if the real-time frequency multiple is negative, it means the fluctuation amplitude value is less than the fluctuation amplitude threshold, and the acquisition frequency can be decreased.

[0066] The adjustment of the acquisition frequency can be achieved in various ways. For example, set a basic acquisition frequency, and then multiply or divide it according to the real-time frequency multiple. Assume the basic acquisition frequency is once per minute, and the real-time frequency multiple is 2, then the new acquisition frequency is once every 30 seconds; if the real-time frequency multiple is -1, then the new acquisition frequency is once every 2 minutes.

[0067] In some embodiments of the present application, real-time weather data can also be obtained through the Internet. In the case where the real-time weather data is within the range of preset severe weather data, adjust the acquisition frequency, and calculate the real-time acquisition frequency for severe weather according to the preset weather change acquisition frequency multiple and the real-time acquisition frequency, so as to more accurately monitor the water quality change in the case where the severe weather will affect the circulating water system.

[0068] In the above steps, by adjusting the acquisition frequency of circulating water operation data according to the real-time frequency multiple, the system can more flexibly respond to data fluctuations, ensuring that sufficient data can be obtained in a timely manner for analysis and processing in case of anomalies. This enables the system to detect potential problems earlier, take timely measures to avoid the further deterioration of anomalies, and ensure the stable operation of the circulating water system. In addition, this method of dynamically adjusting the acquisition frequency can also optimize the efficiency of data acquisition, reduce unnecessary data transmission and storage, and further improve the operation efficiency and economic benefits of the system.

[0069] S304. Input the real-time circulating water operation data into a preset circulating water data prediction model. S305. Obtain real-time actual ion concentration data. Specifically, the real-time actual ion concentration data includes the real-time concentration data of all ions involved in the chemical equilibrium equation.

[0070] The methods for obtaining real-time actual ion concentration data include instrumental analysis methods and chemical analysis methods, etc. Instrumental analysis methods include: ion chromatography, which utilizes the principle of ion exchange. Different ions have different affinities for the ion exchange resin. Through elution with an eluent, various ions are separated in sequence and the ion concentration is detected by a detector. For example, for anions such as chloride ions and sulfate ions, and cations such as sodium ions and calcium ions in circulating water, accurate detection can be achieved; atomic absorption spectrometry (AAS), which is mainly used to detect the concentration of metal ions. By atomizing the sample to make the metal ions in the ground state atoms, when irradiated with light of a specific wavelength, the ground state atoms absorb the energy of the light and transition to the excited state, and the ion concentration is calculated based on the relationship between absorbance and concentration. For example, detecting copper ions and iron ions in circulating water; inductively coupled plasma mass spectrometry (ICP-MS), which can simultaneously detect the concentrations of multiple trace element ions. After ionizing the sample, the mass-to-charge ratio of the ions is measured by a mass spectrometer, and the concentration is determined according to the ion intensity. When there are multiple trace element ions in circulating water, efficient analysis can be carried out. Chemical analysis methods include titration methods. Utilizing the quantitative relationship of chemical reactions, such as acid-base titration can measure the concentration of hydrogen ions or hydroxide ions, and complexometric titration can measure the concentration of metal ions. For example, using the EDTA titration method to determine the concentrations of calcium ions and magnesium ions in circulating water.

[0071] S306. Calculate the real-time ion concentration data in the circulating water according to the predicted circulating water operation data and the chemical equilibrium equation. Specifically, the real-time ion concentration data includes the real-time concentration data of all ions involved in the chemical equilibrium equation.

[0072] First, clarify the corresponding chemical equilibrium equation, extract the parameters related to the chemical equilibrium equation from the predicted circulating water operation data, and then combine chemical relationships such as charge conservation and material conservation to solve the simultaneous equations to calculate the real-time concentration data of the ions involved in all chemical equilibrium equations, which are the real-time ion concentration data.

[0073] S307. When the concentration data difference of the same type of ions in the real-time ion concentration data and the real-time actual ion concentration data is greater than the preset ion concentration difference threshold, adjust the predicted circulating water operation data again according to the chemical equilibrium equation; When the difference in the concentration of the same type of ions between the real-time calculated ion concentration data and the actually detected ion concentration data is greater than the preset threshold, it indicates that there is a deviation in the predicted circulating water operation data. Since the chemical equilibrium equation reflects the chemical relationship between the ions in the circulating water, the predicted circulating water operation data is adjusted again based on it.

[0074] In some embodiments of the present application, if the deviated ions participate in the precipitation-dissolution equilibrium, the assumption of the precipitate ion concentration can be adjusted, and the solubility product can be recalculated to adjust the prediction data; if it involves the redox equilibrium, the assumption of the concentration of the oxidized or reduced state substances is adjusted, and the electrode potential and the relevant ion concentrations are recalculated according to the Nernst equation to correct the prediction data.

[0075] In the above technical steps, by comparing the real-time calculated ion concentration data with the actually detected ion concentration data, when the difference exceeds the threshold, the predicted circulating water operation data is adjusted according to the chemical equilibrium equation. This makes the prediction data more in line with the actual situation, improves the accuracy of judging the state of the circulating water system, can timely detect and correct the prediction deviation, provides a reliable basis for the stable operation of the circulating water system, and avoids equipment failures or production abnormalities caused by incorrect judgments. S308. After the circulating water data prediction model generates a preliminary prediction result, calculate the real-time chemical equilibrium constant according to the preliminary prediction result and the preset chemical equilibrium equation; S309. When the difference in the real-time chemical equilibrium constant is greater than the preset chemical equilibrium constant difference threshold, adjust the preliminary prediction result to obtain the predicted circulating water operation data; S310. Compare the predicted circulating water data with the circulating water operation data range threshold to determine whether there is real-time abnormal operation data exceeding the circulating water operation data range threshold; If so, execute the following step S311; If not, return to execute the above step S301; S311. Determine the real-time abnormal situation through the real-time abnormal data combined with the causal relationship Bayesian network; S312. Extract multiple data related to the determined abnormal conditions from the real-time circulating water operation data and input them into a preset mathematical relation function; Specifically, the mathematical relation function is a mathematical relation function among multiple parameters constructed in advance based on historical circulating water operation data and historical abnormal conditions.

[0076] First, the circulating water monitoring system has determined the real-time abnormal conditions. Then, based on the understanding of the circulating water system and past experience, data closely related to the abnormal conditions are screened out from the real-time circulating water operation data (such as pH value, hardness value, turbidity value, etc.). Next, these data are input into the pre-constructed mathematical relation function framework. This mathematical relation function framework is obtained through data analysis and mathematical modeling methods using historical circulating water operation data and historical abnormal conditions, and it reflects the internal relationships among multiple parameters, such as linear relationships, non-linear relationships, etc.

[0077] If the mathematical relation function framework is a linear regression model, after inputting the data, calculate the result according to the linear equation of the function framework. For a complex circulating water system, there may be a combination of multiple sub-functions. For example, first use a function to analyze the relationship between water quality parameters, and then use another function to further analyze in combination with equipment operation parameters.

[0078] S313. Calculate according to the mathematical relation function to obtain the theoretical operation result corresponding to the real-time abnormal conditions; the mathematical relation function is constructed based on historical data and reflects the internal relationships among the parameters of the circulating water. It may be a linear equation, a non-linear function, or a complex function or a combination of functions. When inputting the data related to the real-time abnormal conditions, the mathematical relation function performs operations according to its established mathematical logic and parameter relationships. For example, in a linear regression model, by multiplying the input data by the coefficients in the model and summing them up, a theoretical value is obtained, and finally the theoretical operation result is output. This result is a prediction of the operating state that the circulating water system should have under the current abnormal conditions based on the learning of historical laws by the mathematical relation function.

[0079] S314. Calculate the difference index between the theoretical operation result and the real-time circulating water operation data; Calculating the difference index between the theoretical operation result and the real-time circulating water operation data is to evaluate the accuracy and reliability of the model by comparing the theoretical result calculated by the mathematical relation function with the real-time data of the actual system.

[0080] Specifically, the theoretical operation results obtained from the mathematical relationship function are compared one by one with the real-time circulating water operation data, which include parameters such as pH value, hardness, turbidity, conductivity, and dissolved oxygen content. There are various methods to calculate the difference between the theoretical operation results and the real-time data, such as the mean square error (MSE), mean absolute error (MAE), etc. For example, the calculation formula for the mean square error is: where n is the number of data points, and the theoretical value i and the real-time value i are the theoretical value and the real-time value of the i-th data point respectively. According to the calculated difference index, the accuracy and reliability of the model are evaluated. If the difference index exceeds the preset threshold, it indicates that there is a large deviation between the prediction result of the model and the actual operation data, and the model needs to be optimized and adjusted.

[0081] S315. In the case where the difference index exceeds the preset verification threshold, add the theoretical operation result and the real-time circulating water operation data to the historical circulating water operation dataset to obtain an updated historical circulating water operation dataset; Specifically, add the theoretical operation result and the real-time circulating water operation data to the historical circulating water operation dataset. This step can be achieved in various ways: direct addition, directly add the new data points to the historical dataset to form a new dataset; data fusion, use data fusion technology to fuse the new data points with the historical dataset, for example, through weighted average or other statistical methods to ensure the representativeness and accuracy of the dataset; data cleaning, before adding new data, perform data cleaning to remove outliers or noise data to ensure the quality of the dataset.

[0082] In some embodiments of the present application, adding the theoretical operation result and the real-time circulating water operation data to the historical circulating water operation dataset specifically means: before adding new data, perform data cleaning to remove outliers or noise data to ensure the quality of the dataset. The data points outside the normal range can be identified and eliminated by setting a reasonable threshold range. Use data fusion technology to fuse the new data points with the historical dataset. For example, through weighted average or other statistical methods, ensure the representativeness and accuracy of the dataset. The weighted average method can assign different weights according to the reliability and timeliness of the data, so as to more accurately reflect the actual operation state of the system. Store the cleaned and fused data in the historical circulating water operation dataset. This can be achieved through the INSERT statement of the database management system, inserting the new data points into the corresponding data table. For example, use the SQL INSERT INTO statement to insert the new data points into the historical data table.

[0083] S316. Use the updated historical circulating water operation dataset to retrain the circulating water data prediction model. Specifically, first, preprocess the updated historical circulating water operation dataset, including operations such as data cleaning, normalization, and standardization, to ensure the quality and consistency of the data. Data cleaning can remove outliers and noisy data, while normalization and standardization can transform the data to the same scale for easy model training.

[0084] Select appropriate features according to the requirements of the model and the characteristics of the data. For example, parameters such as pH value, hardness, turbidity, conductivity, and dissolved oxygen content can be selected as features. Feature selection can be carried out through methods such as correlation analysis and principal component analysis (PCA).

[0085] Select a suitable machine learning model for training. Commonly used models include BP neural network, random forest (RF), long short-term memory network (LSTM), and gated recurrent unit (GRU), etc. Each model has its own advantages and disadvantages. For example, the BP neural network has the ability of self-adaptation and self-learning, the random forest has a small computational amount and is not sensitive to data noise and outliers, and LSTM and GRU are suitable for time series prediction.

[0086] Use the preprocessed dataset to train the selected model. During the training process, the dataset can be divided into a training set and a test set. The training set is used for model training, and the test set is used for model validation and evaluation. The model performance can be optimized by adjusting the hyperparameters of the model (such as learning rate, number of hidden layer nodes, etc.) during the training process.

[0087] In some embodiments of the present application, if a certain pump in the circulating water system fails, resulting in an abnormal increase in water flow. At this time, the system will collect the operation data during the failure period and add it to the historical dataset. Retraining the circulating water data prediction model with the updated dataset can help the model learn the operation mode under fault conditions and improve the prediction accuracy and robustness of the model.

[0088] In the above steps, by using the updated historical circulating water operation dataset to retrain the circulating water data prediction model, the system can continuously optimize the prediction performance of the model, can more accurately predict the operation state of the circulating water system, and timely discover potential problems.

[0089] S317. Retrieve a preset number of historical abnormal solution plans; Specifically, the system pre-stores multiple historical abnormal solutions, which are generated based on past abnormal data and processing experience. Each solution contains detailed processing steps and applicable conditions. When the system detects real-time abnormal data, according to the type and characteristics of the abnormality, it retrieves the historical abnormal solutions that match the current abnormal situation from the stored solutions. For example, if the abnormality is caused by abnormal pH, the system will retrieve all solutions related to abnormal pH. The retrieved solutions will be sorted according to their matching degree with the current abnormal situation. The matching degree can be calculated by the similarity between the characteristics in the solution and the current abnormal data. For example, if the current abnormal data is very similar to the data in a certain solution, the matching degree of this solution will be higher. The system selects several solutions with the highest matching degree according to the preset quantity. These solutions will be used as candidate solutions for further analysis and selection.

[0090] S318. Retrieve the historical circulating water operation data corresponding to each historical abnormal solution; Specifically, the system stores the historical circulating water operation data corresponding to each solution, which are collected during the implementation of the solution. After the system retrieves the preset number of historical abnormal solutions, for the retrieved historical abnormal solutions, the system will retrieve the historical circulating water operation data corresponding to this solution. These data include parameters such as pH, hardness, turbidity, conductivity, and dissolved oxygen content.

[0091] S319. Input each historical circulating water operation data and the real-time circulating water operation data into the first formula to calculate the solution score corresponding to each historical solution; Specifically, the first formula is:

[0092] where S i is the solution score of the i-th historical solution, n represents the number of index data in the real-time circulating water operation data. Here, n = 5, that is, the real-time pH value, real-time hardness value, real-time turbidity value, real-time conductivity, and real-time dissolved oxygen content value. ω j is the weight coefficient of the j-th index data, D j,r is the actual measured value of the j-th index data in the real-time circulating water operation data, D j,i is the value of the j-th index data in the i-th historical circulating water operation data, is the average value of the j-th index data in all historical circulating water operation data, α j 、β j and γ j are adjustment coefficients, Max j and Min j are the maximum and minimum values of the j-th index data in all historical circulating water operation data respectively.

[0093] A detailed explanation of each part of the above first formula: |D j,r -D j,i |: This part calculates the absolute difference between the j-th real-time circulating water operation data index (such as the real-time pH value) and the corresponding index value in the i-th historical solution case. It reflects the degree of difference between the current real-time data and the specific historical case data. The smaller the difference, the more similar the historical case is to the current situation in this index, and the higher the reference value for solving the current problem may be.

[0094] In is the average value of the j-th data index in all historical cases. This part calculates the difference between the real-time data and the historical average data, and is adjusted by β j . Introducing the historical average value can reflect the position of the current real-time data in the historical data distribution. If the difference between the real-time data and the historical average data is large, it indicates that the current situation may be relatively special, and more caution is needed when selecting a solution. β j can adjust the importance of this difference in the calculation according to actual needs. If the fluctuation of this index has a greater impact on the circulating water system, the value of β j can be appropriately increased.

[0095] In, α j is an adjustment coefficient used to control the contributions of the difference between the real-time data and the single historical case data and the difference from the historical average data in the overall calculation. By adjusting α j , the weights of these two differences can be flexibly allocated according to the actual situation. For example, in some cases, if more attention is paid to the direct comparison between the real-time data and a single historical case, the value of α j for |D j,r -D j,i | can be appropriately increased; if more importance is attached to the deviation from the historical average level, α j can be adjusted to make play a more prominent role.

[0096] In γ j ×(Max j -Min j ), Max j and Min j are respectively the maximum and minimum values of the j-th data index in all historical cases. Dividing by γ j ×(Max j -Min j) is a normalization operation. Different data metrics (such as pH, hardness, turbidity, etc.) have different dimensions and value ranges. Without normalization, some metrics with larger values may dominate the calculation and affect the accuracy of the final result. For example, the value of conductivity may be much larger than the pH value. Without normalization, the differences in conductivity may mask the differences in other metrics during the calculation. γ j As an adjustment coefficient, it can further fine-tune the degree of normalization to adapt to the characteristics of different data metrics.

[0097] ω j is the weight coefficient of the jth metric data. In a circulating water system, different metric data have different degrees of influence on the system operation. For example, pH may directly affect the corrosion of equipment and is crucial for the safe operation of the system; while turbidity also has an impact, but relatively smaller. By setting different ω j values, the role of key metrics in evaluating historical circulating water operation data can be highlighted. If pH is a key metric, its corresponding ω j can be set larger. In this way, when calculating the score of historical data, the matching degree of the pH metric has a greater impact on the final score, ensuring that the selected optimal solution can more effectively solve key problems.

[0098] In , the purpose of taking the reciprocal is to make the scores of historical circulating water operation data with smaller differences higher. Because the smaller the difference value calculated previously, the more similar the historical data is to the current real-time situation. After taking the reciprocal, the historical data with high matching degree will contribute a greater score in the summation calculation, thus ranking higher in the final score ranking.

[0099] Summing up the weighted results of all metric data (a total of n, such as metrics like pH, hardness, turbidity, conductivity, and dissolved oxygen content value, etc.) comprehensively considers the influence of multiple metrics on the matching degree of historical circulating water operation data. This can comprehensively evaluate the overall similarity between each historical circulating water operation data and the current real-time circulating water operation status, avoid selecting solutions based on a single metric, and improve the accuracy and reliability of the screening results.

[0100] In the above steps, the differences between the real-time circulating water operation data and the historical solution case data, as well as the differences between the real-time data and the historical average data, are comprehensively considered. By performing weighted calculations on different indicators and normalizing the results, the matching degree between each historical solution case and the current real-time situation can be evaluated more comprehensively and accurately. Not only the differences between the real-time data and a single historical case are considered, but also the differences from the historical average data are combined to comprehensively measure the similarity between the historical case and the current situation; the weight coefficient and adjustment coefficient can be flexibly adjusted according to the actual situation to adapt to the characteristics of different circulating water systems and business requirements, improving the accuracy of screening the optimal solution; ensuring the comparability of different data indicators in the calculation and avoiding unreasonable screening results caused by differences in data characteristics.

[0101] S320. Sort all the solution scores from high to low, and the historical abnormal solution corresponding to the highest solution score with the highest ranking is the optimal abnormal solution; Specifically, sort the scores of all historical abnormal solutions from high to low. The sorting can be implemented using various algorithms, such as quick sort, merge sort, etc. After sorting, the solution with the highest score is ranked at the top. Select the solution with the highest score from the sorted list as the optimal abnormal solution.

[0102] S321. Execute the optimal abnormal solution; According to the specific content of the optimal solution, perform corresponding operations. For example, if the optimal solution is "increase the operation frequency of the side filter", the system will automatically adjust the operation parameters of the side filter to increase its operation frequency.

[0103] In some embodiments of the present application, during the process of executing the solution, the system will monitor the operation data of the circulating water system in real time to ensure the implementation effect of the solution. If data anomalies are found, the system will adjust the solution in a timely manner or take other measures. Based on the real-time monitored data, the system will evaluate the implementation effect of the solution. If the effect is not ideal, the system will feedback relevant information and adjust the solution or select other solutions as needed.

[0104] In the above steps, the system can quickly and accurately determine the most suitable abnormal handling solution based on historical experience and real-time data, thereby improving the efficiency and success rate of abnormal handling, reducing equipment damage or production interruption caused by improper abnormal handling, and further enhancing the stability of the circulating water system.

[0105] S322. Send abnormal information to the preset mobile terminal device.

[0106] Specifically, the abnormal information includes the real-time circulating water operation data and the optimal operation solution.

[0107] Package the real-time circulating water operation data and the optimal operation plan to form exception information. This can be achieved by creating a data structure or object and storing the relevant data as attributes or fields therein.

[0108] Select a suitable transmission method according to the preset mobile terminal device type and network environment. Common transmission methods include, but are not limited to: SMS, sending the exception information to the mobile terminal device through the SMS gateway; instant messaging tools, such as various social software, sending messages by calling their API interfaces; email, sending the exception information to the preset email box through the mail server; mobile application push, if a specific application is installed on the mobile terminal device, sending notifications through push services (such as APNs, FCM).

[0109] Establish a connection with the mobile terminal device according to the selected transmission method. For example, if sending by SMS, it is necessary to connect to the SMS gateway; if sending by instant messaging tools, it is necessary to call the corresponding API interface.

[0110] Send the packaged exception information to the mobile terminal device through the established connection.

[0111] In the above steps, by sending the exception information to the preset mobile terminal device, the system can timely convey the real-time circulating water operation data and the optimal operation plan to the relevant personnel. This enables the relevant personnel to quickly understand the abnormal situation and take corresponding measures for processing, thereby improving the efficiency and accuracy of exception handling.

[0112] Steps S304, S308 - S311 are similar to Figure 2 Steps S202 - S206 in the illustrated embodiment. Refer to the description in Steps S202 - S206, and details are not repeated here.

[0113] In the above embodiments, by collecting real-time circulating water operation data and using the circulating water data prediction model obtained from the experience of learning historical data for preliminary prediction, and then adjusting it in combination with the chemical equilibrium equation to make the prediction more in line with the actual chemical laws. Comparing the predicted value with the threshold can quickly identify anomalies, and using the causal relationship Bayesian network can accurately locate anomalies based on the historical parameter relationships. Calculate the historical solution scores, comprehensively evaluate the real-time data, select the optimal solution, and improve the pertinence of handling anomalies. Compare the calculated and actual ion concentrations, and adjust the predicted data again based on the chemical equilibrium to ensure the prediction accuracy. After verifying the mathematical relationship function, update the dataset to train the model so that it can adapt to more complex situations. Adjust the collection frequency according to the data fluctuations to capture anomaly changes in a timely manner. Send anomaly information after implementing the solution to facilitate relevant personnel to understand the situation. This enables the circulating water system to achieve real-time and accurate operation monitoring, effectively improve the accuracy of anomaly prediction and handling, enhance the stability and reliability of the circulating water system operation, reduce equipment failures and production losses caused by problems such as abnormal water quality, and facilitate managers to promptly understand the system status and make decisions.

[0114] The following introduces the exemplary circulating water monitoring system 400 provided by the embodiments of the present application. Figure 4 It is an exemplary hardware structure diagram of the circulating water monitoring system 400 provided by the embodiments of the present application.

[0115] In some embodiments, the circulating water monitoring system 400 includes a computer device. The computer device includes a processor, a memory, and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data. The network interface of the computer device is used to communicate with other external terminals or servers through a network connection. In some embodiments, the network interface can be a wired network interface, and in some embodiments, the network interface can also be a wireless network interface. When the computer program is executed by the processor, it implements the method in the embodiments of the present application.

[0116] Those skilled in the art can understand that Figure 4 the structure shown in

[0117] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the various embodiments of the present application.

[0118] In the above embodiments, depending on the context, the term "when..." can be interpreted to mean "if...", or "after...", or "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if detecting (the stated condition or event)" can be interpreted to mean "if determining...", or "in response to determining...", or "when detecting (the stated condition or event)", or "in response to detecting (the stated condition or event)".

[0119] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as coaxial cable, optical fiber, digital subscriber line) or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer, or a data storage device such as a server or data center that includes one or more integrated available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid-state drive), etc.

[0120] Those of ordinary skill in the art can understand all or part of the processes in the above embodiments of the method. These processes can be completed by hardware instructed by a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above method embodiments. The foregoing storage medium includes: ROM or random access memory RAM, magnetic disk, or optical disk, etc., which can store program code.

Claims

1. A circulating water monitoring method, characterized in that, Including: Collecting real-time circulating water operation data; The real-time circulating water operation data includes real-time pH value, real-time hardness value, real-time turbidity value, real-time conductivity, and real-time dissolved oxygen content value; Inputting the real-time circulating water operation data into a preset circulating water data prediction model; the circulating water data prediction model is obtained by training in advance using a historical circulating water operation data set; After the preliminary prediction result is generated by the circulating water data prediction model, calculating the real-time chemical equilibrium constant according to the preliminary prediction result and a preset chemical equilibrium equation; When the difference in the real-time chemical equilibrium constant is greater than a preset chemical equilibrium constant difference threshold, adjusting the preliminary prediction result to obtain predicted circulating water operation data; the difference in the real-time chemical equilibrium constant is the value obtained by taking the absolute value of the difference between the real-time chemical equilibrium constant and the preset standard chemical equilibrium constant; the predicted circulating water operation data is such that the standard chemical equilibrium constant can be calculated according to the chemical equilibrium equation; Comparing the predicted circulating water data with the circulating water operation data range threshold to determine whether there is real-time abnormal operation data exceeding the circulating water operation data range threshold; If so, determining the real-time abnormal situation through the real-time abnormal data combined with a causal relationship Bayesian network; the causal relationship Bayesian network is a Bayesian network generated in advance based on historical circulating water parameter data groups, reflecting the causal relationship between circulating water parameters.

2. The method according to claim 1, characterized in that, After the step of "If so, determining the real-time abnormal situation through the real-time abnormal data combined with a causal relationship Bayesian network", it further includes: Retrieving a preset number of historical abnormal solution plans; For each historical abnormal solution plan, calculating the solution score corresponding to the historical abnormal solution plan in combination with the real-time circulating water operation data; Sorting all the solution scores from high to low, and the historical abnormal solution plan corresponding to the highest solution score is the optimal abnormal solution plan; Executing the optimal abnormal solution plan.

3. The method according to claim 2, characterized in that, The step of "For each historical abnormal solution plan, calculating the corresponding solution score in combination with the real-time circulating water operation data" specifically includes: Retrieving the historical circulating water operation data corresponding to each historical abnormal solution plan; Inputting each historical circulating water operation data and the real-time circulating water operation into a first formula to calculate the solution score corresponding to each historical solution plan; Wherein, the first formula is: Among them, S i is the solution score of the i-th historical solution, n represents the number of index data in the real-time circulating water operation data. Here, n = 5, namely the real-time pH value, real-time hardness value, real-time turbidity value, real-time conductivity, and real-time dissolved oxygen content value. ω j is the weight coefficient of the j-th index data, D j,r is the actual measured value of the index data in the j real-time circulating water operation data, D j,i is the value of the j-th index data in the i-th historical circulating water operation data, is the average value of the j-th index data in all historical circulating water operation data, α j , β j and γ j are adjustment coefficients, Max j and Min j are respectively the maximum value and the minimum value of the j-th index data in all historical circulating water operation data.

4. The method according to claim 1, wherein After the preliminary prediction result is generated by the circulating water data prediction model and the preliminary prediction result is adjusted by applying chemical equilibrium equation constraints to obtain predicted circulating water operation data, it further includes: Obtaining real-time actual ion concentration data; the real-time actual ion concentration data includes the real-time concentration data of all ions involved in the chemical equilibrium equation; Calculating the real-time ion concentration data in the circulating water according to the predicted circulating water operation data and the chemical equilibrium equation; the real-time ion concentration data includes the real-time concentration data of all ions involved in the chemical equilibrium equation; In the case where the concentration data difference of the same type of ions in the real-time ion concentration data and the real-time actual ion concentration data is greater than the preset ion concentration difference threshold, the predicted circulating water operation data is adjusted again according to the chemical equilibrium equation.

5. The method according to claim 1, wherein If it exists, after determining the real-time abnormal situation through the real-time abnormal data combined with the causal relationship Bayesian network, it further includes: Extracting multiple data related to the determined abnormal situation from the real-time circulating water operation data and inputting them into a preset mathematical relationship function; the mathematical relationship function is a mathematical relationship function between multiple parameters constructed in advance based on historical circulating water operation data and historical abnormal situations; Calculating according to the mathematical relationship function to obtain the theoretical operation result corresponding to the real-time abnormal situation; Calculating the difference index between the theoretical operation result and the real-time circulating water operation data; In the case where the difference index exceeds the preset verification threshold, adding the theoretical operation result and the real-time circulating water operation data to the historical circulating water operation data set to obtain an updated historical circulating water operation data set; Using the updated historical circulating water operation data set to train the circulating water data prediction model again.

6. The method according to claim 1, wherein The collecting of the real-time circulating water operation data specifically includes: collecting the circulating water operation data at a real-time collection frequency; Calculating the fluctuation amplitude value between the real-time circulating water operation data and the nearest circulating water operation data; the real-time circulating water operation data is the latest collected circulating water operation data; the nearest circulating water operation data is the circulating water operation data collected last time at the real-time collection frequency; In the case where the fluctuation amplitude value exceeds the fluctuation amplitude threshold, adjusting the circulating water operation data collection frequency according to the real-time frequency multiple; the real-time frequency multiple is the difference between the fluctuation amplitude value and the fluctuation amplitude threshold.

7. The method according to claim 2, characterized in that After executing the optimal abnormal solution, it further includes: Sending abnormal information to a preset mobile terminal device; the abnormal information includes the real-time circulating water operation data and the optimal operation plan.

8. A circulating water monitoring system, characterized in that, The circulating water monitoring system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the circulating water monitoring system to execute the method according to any one of claims 1-7.

9. A computer program product comprising instructions, characterized in that, When the computer program product runs on the circulating water monitoring system, enabling the circulating water monitoring system to execute the method according to any one of claims 1-7.

10. A computer-readable storage medium, comprising instructions, characterized in that, When the instruction runs on the circulating water monitoring system, enabling the circulating water monitoring system to execute the method according to any one of claims 1-7.

Images