Intelligent control method and system for boiler wastewater treatment

Abstract

The invention relates to the technical field of electric digital data processing, in particular to an intelligent control method and system for boiler wastewater treatment, and the method comprises the steps: obtaining the fluctuation degree of boiler wastewater equipment at a current time node, the fluctuation degree is positively correlated with the abnormal degree of the boiler wastewater equipment at the current time node and is negatively correlated with the dynamic lag time of the boiler wastewater equipment at the current time node; and inputting the fluctuation degree into an improved machine learning model, inputting an operation state of the boiler wastewater equipment at the current time node, and controlling start and stop of the boiler wastewater equipment based on the operation state. The problem that an existing machine learning model is not high in prediction accuracy is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic digital data processing, and more particularly to an intelligent control method and system for boiler wastewater treatment. Background Art

[0002] During the operation of industrial boiler systems, the generation of wastewater is inevitable. Its quality and quantity are affected by a variety of factors, including the boiler's operating load, fuel properties, heat transfer efficiency, and water replenishment. Boiler wastewater typically contains high temperature, high pressure, high hardness, and a variety of impurity particles and chemical substances. If improperly handled, it can easily cause equipment corrosion, scaling, and excessive pollutant emissions, thereby affecting the stable operation of the entire thermal system and the achievement of environmental protection indicators. To improve the efficiency and intelligence of boiler wastewater treatment, an increasing number of companies are attempting to introduce advanced data-driven methods and intelligent control systems into the wastewater treatment process to achieve real-time monitoring and automatic adjustment of wastewater quality, emissions, and operating status.

[0003] In recent years, with the continuous development of sensors, data acquisition, and edge computing technologies, the dimensions of wastewater treatment-related data available at industrial sites have become increasingly rich, including flow, pressure, temperature, chemical concentration, conductivity, inlet and outlet pressure differentials, and other parameters, forming a typical time series structure. Based on this, researchers have widely introduced machine learning models, particularly deep learning-based neural network models, attempting to train high-precision predictive models using historical operating data to proactively identify potential abnormal behavior in wastewater systems, thereby providing feedforward decision support for control systems.

[0004] However, existing machine learning models still have many limitations in practical applications, especially in terms of prediction accuracy and generalization ability, which are still difficult to meet the high requirements of boiler wastewater treatment under typical complex working conditions. First, many models focus on fitting single or static features during design, and lack sufficient modeling of the dynamic response process, hysteresis characteristics and nonlinear fluctuation behavior of the equipment, resulting in large deviations in prediction results in industrial environments with severe actual fluctuations or frequent anomalies. Secondly, most models fail to effectively introduce a dynamic evaluation mechanism related to working conditions in the design of loss functions, and are unable to identify the evolution of key features that appear normal but are actually abnormal. This makes the model's prediction ability insufficient in the early stages of abnormal conditions, and may result in missed detections or misjudgments, which in turn leads to low prediction accuracy. Summary of the Invention

[0005] In order to solve the problem of low prediction accuracy raised in the above background technology, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides an intelligent control method for boiler wastewater treatment, comprising: obtaining the degree of fluctuation of the boiler wastewater equipment at the current time node, the degree of fluctuation is positively correlated with the degree of abnormality of the boiler wastewater equipment at the current time node, and negatively correlated with the dynamic lag time of the boiler wastewater equipment at the current time node; inputting the degree of fluctuation into an improved machine learning model, inputting the operating status of the boiler wastewater equipment at the current time node, and controlling the start and stop of the boiler wastewater equipment based on the operating status; wherein the improved machine learning model includes a loss function, the loss function is positively correlated with the difference between the degree of fluctuation at the current time node and the mean of the degree of fluctuation at multiple set time nodes in history, and negatively correlated with the standard deviation of the degree of fluctuation at multiple set time nodes in history.

[0007] This technical solution uses the fluctuation degree formed by combining the current abnormality level of the equipment with the dynamic lag time as the core input. It then uses an improved machine learning model to control the start and stop of the equipment in conjunction with the current operating status. During the model training process, a loss function is designed based on the deviation of the fluctuation degree from the historical average level and the historical fluctuation amplitude. This enhances the model's sensitivity to abnormal fluctuations and suppresses the interference of normal fluctuations. This structural logic effectively integrates the dynamic characteristics of equipment operation and historical behavioral patterns, enabling the control system to more accurately identify and respond to abnormal conditions, improving operational stability and control precision.

[0008] Furthermore, the degree of fluctuation for: , For boiler wastewater equipment The degree of abnormality at the time node, For boiler wastewater equipment Dynamic lag time of time nodes, The natural constant An exponential function with base .

[0009] The above technical solution combines the indicator representing the degree of abnormality with the lag time factor reflecting the system response speed, and uses an exponential decay function to dynamically modulate the degree of abnormality. When the response lag is small, the fluctuation degree is more significant, and when the response lag is large, the fluctuation effect is automatically weakened, thereby achieving the enhancement of the real abnormal signal and the suppression of delayed noise.

[0010] Furthermore, the abnormality for: , For boiler wastewater equipment The inlet and outlet pressure difference at the time node, It is the reference pressure difference of boiler wastewater equipment under normal operating conditions.

[0011] The above technical solution constructs an abnormality indicator that can dynamically reflect the degree to which the equipment deviates from the normal state by performing ratio differentiation processing on the current inlet and outlet pressure difference of the equipment and the reference pressure difference under normal operating conditions, thereby quantifying the degree of abnormality of the current operating state of the equipment.

[0012] Furthermore, the dynamic lag time for: , is the lag time of the boiler wastewater equipment in standby state, is the lag time of the boiler wastewater equipment under full load condition, For boiler wastewater equipment Wastewater flow at the time node, The maximum wastewater flow rate carried by the boiler wastewater equipment, To set the time window size, is the normalization function.

[0013] The above technical solution combines the changes in wastewater flow within a set time window and normalizes it with the equipment carrying capacity to construct a lag time indicator that can dynamically change with the load state. The higher the equipment load, the faster the response and the shorter the lag time, thereby more realistically reflecting the response characteristics of the system under different operating conditions.

[0014] Furthermore, the loss function for: , For boiler wastewater equipment The degree of fluctuation at the time node, is the average value of the fluctuation degree of boiler wastewater equipment at multiple historical set time nodes, is the standard deviation of the fluctuation degree of boiler wastewater equipment at multiple historical set time nodes, Set the number of time points for history.

[0015] This technical solution constructs a loss function centered on the deviation between the degree of fluctuation and the historical average, and normalizes this with the standard deviation of historical fluctuations. This allows the model to focus more on states that exhibit significant anomalies relative to historical behavior during training, thereby increasing its sensitivity to unstable operating conditions. This structural logic strengthens the model's ability to identify abnormalities in equipment operating fluctuations, effectively avoiding misjudgments caused by large or small historical fluctuations, and improving the robustness and generalization of anomaly detection.

[0016] Furthermore, when hour, ;when hour, .

[0017] The technical solution avoids the error offset problem in the positive and negative directions, ensures that the loss function gives equal attention to all abnormal fluctuations during the training process, improves the comprehensive perception ability of the model to fluctuation trends, accurately identifies abnormal states whether they are too high or too low, and enhances the adaptability and robustness of the model to complex dynamic behaviors of the boiler wastewater system.

[0018] Further, the machine learning model is a long short-term memory neural network model.

[0019] Further, the operating state is divided into normal and abnormal.

[0020] Further, the method further comprises training the machine learning model, specifically: inputting a training set into a pre-constructed machine learning model for training, calculating the loss between the predicted value and the label during the training process; adjusting the model parameters using the gradient descent method to minimize the prediction error; iteratively adjusting the parameters of the machine learning model until the loss is less than a set value or a set number of training times is reached, and finally obtaining the trained machine learning model.

[0021] In a second aspect, the present application provides an intelligent control system for boiler wastewater treatment, comprising a memory and a processor, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned intelligent control method for boiler wastewater treatment is realized.

[0022] The present application has the following advantages: The present application combines equipment abnormal amplitude and response lag characteristics to construct a dynamic fluctuation index, and uses it as the core input of the machine learning model, and highlights the deviation from the historical normal fluctuation range during model training, so that the system can accurately identify the operating abnormality and automatically trigger the start-stop control in real time, thereby ensuring high sensitivity to sudden abnormalities while suppressing false positives caused by normal fluctuations, effectively improving the stability, response speed and robustness of the boiler wastewater treatment process. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 is a flow chart of an intelligent control method for boiler wastewater treatment according to an embodiment of the present application; Figure 2 is a structural block diagram of an intelligent control system for boiler wastewater treatment according to an embodiment of the present application. DETAILED DESCRIPTION

[0024] An embodiment of an intelligent control method for boiler wastewater treatment.

[0025] like Figure 1 As shown in FIG, a flow chart of an intelligent control method for boiler wastewater treatment according to an embodiment of the present invention includes the following steps: S1: Obtain the fluctuation degree of boiler wastewater equipment at the current time node.

[0026] In one embodiment, the degree of fluctuation for: , For boiler wastewater equipment The degree of abnormality at the time node, For boiler wastewater equipment Dynamic lag time of time nodes, The natural constant An exponential function with base .

[0027] By combining the severity of a device's anomaly with its dynamic lag time and using an exponential function to attenuate the anomaly, the impact of the anomaly gradually decreases as the lag time increases, reflecting the timeliness and actual dynamic characteristics of the system's response. This design not only emphasizes the importance of immediate anomalies but also effectively suppresses misjudgments caused by lag time, improving the accuracy of anomaly detection and the responsiveness of control strategies.

[0028] In one embodiment, the abnormality level for: , For boiler wastewater equipment The inlet and outlet pressure difference at the time node, It is the reference pressure difference of boiler wastewater equipment under normal operating conditions.

[0029] By comparing the current inlet and outlet pressure differential of the device with a reference pressure differential under normal operating conditions and calculating the relative difference between the two, the degree of anomaly is quantified, accurately reflecting deviations from the device's operating status. This construction logic, based on direct measurement of physical quantities, can sensitively detect problems such as blockages or flow anomalies caused by pressure differential changes, improving the accuracy and real-time nature of anomaly detection and providing a reliable foundation for subsequent intelligent control.

[0030] In one embodiment, the dynamic lag time for: , is the lag time of the boiler wastewater equipment in standby state, is the lag time of the boiler wastewater equipment under full load condition, For boiler wastewater equipment Wastewater flow at the time node, The maximum wastewater flow rate carried by the boiler wastewater equipment, To set the time window size, is the normalization function.

[0031] By normalizing the wastewater flow rate of the equipment within the set time window and its maximum carrying capacity, the lag time indicator is dynamically adjusted, so that the equipment exhibits a longer response lag when the load is low, and the lag time is shortened when the load is close to full load. This truly reflects the impact of the equipment operating status on the response speed and improves the adaptability and response efficiency of the control strategy under different load conditions.

[0032] S2: Input the degree of fluctuation into the improved machine learning model and input the operating status of the boiler wastewater equipment at the current time node.

[0033] In one embodiment, the improved machine learning model includes a loss function, wherein the loss function for: , For boiler wastewater equipment The degree of fluctuation at the time node, is the average value of the fluctuation degree of boiler wastewater equipment at multiple historical set time nodes, is the standard deviation of the fluctuation degree of boiler wastewater equipment at multiple historical set time nodes, Set the number of time points for history.

[0034] By designing a normalized square loss function based on the deviation of the fluctuation degree from its historical mean, the model pays more attention to abnormal conditions that deviate significantly from the historical normal fluctuation range during training, thereby improving the sensitivity and recognition accuracy of abnormal fluctuations of the equipment, enhancing the robustness and generalization ability of the model, and promoting the precise monitoring and intelligent control of the operating status of boiler wastewater equipment.

[0035] when hour, ;when hour, By taking the absolute value of the difference between the degree of volatility and the historical mean, we ensure that whether the volatility is above or below the historical average, it can be fairly quantified and included in the loss calculation, thus avoiding calculation errors caused by deviation signs, improving the comprehensive recognition ability of abnormal fluctuations, and enhancing the model's sensitivity and stability to various abnormal conditions.

[0036] The machine learning model is a long short-term memory neural network model, and the machine learning model is trained using the long short-term memory neural network model. Specifically, a training set is input into a pre-built machine learning model for training, and the training set includes multi-dimensional features such as the degree of fluctuation, abnormal indicators, and lag time of historical collection; during the training process, the loss between the output prediction value and the label is calculated, and the label is set manually; the model parameters are adjusted using the gradient descent method to minimize the prediction error; the parameters of the machine learning model are iteratively adjusted until the loss is less than the set value or the set number of training times is reached, and finally a trained machine learning model is obtained.

[0037] S3: Control the start and stop of the boiler wastewater equipment based on the operating status.

[0038] In one embodiment, operating states are categorized as normal and abnormal. Normal states indicate stable operation that meets operating requirements and requires no intervention. Abnormal states, on the other hand, indicate significant deviations from normal operating parameters, posing potential faults or system risks. Based on these determinations, the system automatically triggers start / stop control commands, promptly shutting down, performing maintenance, or adjusting operating parameters. This prevents the spread of faults and excessive system emissions, ensuring the safety and stability of the boiler wastewater treatment system. This control logic, centered on model judgment results and combined with the dynamic characteristics of the equipment, enables precise decision-making and rapid response to start / stop operations.

[0039] The solution of this invention constructs a fluctuation index that integrates the degree of abnormality and dynamic hysteresis characteristics, introduces an improved machine learning model to intelligently judge the operating status of the equipment, and based on this, achieves dynamic control of the start and stop of boiler wastewater equipment. Furthermore, by designing a loss function associated with historical fluctuation characteristics, the model focuses on identifying abnormal fluctuations, thereby improving the model's learning accuracy and generalization capabilities. This solution comprehensively considers the system's operational dynamics, real-time performance, and historical behavior patterns, and possesses strong robustness and adaptability. It can effectively realize intelligent monitoring and precise control of the boiler wastewater treatment process, reduce the risk of failure, and improve operational efficiency and system safety.

[0040] An embodiment of an intelligent control system for boiler wastewater treatment: like Figure 2 As shown in FIG, a structural block diagram of an intelligent control system for boiler wastewater treatment according to an embodiment of the present invention includes a processor and a memory.

[0041] The present invention also provides an intelligent control system for boiler wastewater treatment. Figure 2As shown, the system includes a processor and a memory, wherein the memory stores computer program instructions. When the computer program instructions are executed by the processor, an intelligent control method for boiler wastewater treatment according to the present invention is implemented.

[0042] The intelligent control system for boiler wastewater treatment also includes other components well known to those skilled in the art, such as a communication interface. The configuration and functions of these components are known in the art and will not be described in detail here.

[0043] In the present invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, the computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store the required information and can be accessed by an application, module, or both. Any such computer storage medium can be part of, accessible to, or connectable to a device. Any application or module described in the present invention can be implemented by computer-readable / executable instructions stored or otherwise retained by such a computer-readable medium.

[0044] In the description of this specification, "multiple" and "several" mean at least two, such as two, three or more, etc., unless otherwise clearly defined.

[0045] While several embodiments of the present invention have been shown and described herein, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Numerous modifications, variations, and alternatives will occur to those skilled in the art without departing from the concept and spirit of the present invention. It should be understood that various alternatives to the embodiments of the present invention described herein may be employed in practicing the present invention.

Claims

1. An intelligent control method for boiler wastewater treatment, characterized in that: include: Obtaining the degree of fluctuation of the boiler wastewater equipment at the current time node, wherein the degree of fluctuation is positively correlated with the abnormality degree of the boiler wastewater equipment at the current time node and is negatively correlated with the dynamic lag time of the boiler wastewater equipment at the current time node; Inputting the fluctuation degree into the improved machine learning model, inputting the operating status of the boiler wastewater equipment at the current time node, and controlling the start and stop of the boiler wastewater equipment based on the operating status; Among them, the improved machine learning model includes a loss function, which is positively correlated with the difference between the fluctuation degree of the current time node and the average of the fluctuation degrees of multiple set time nodes in history, and is negatively correlated with the standard deviation of the fluctuation degrees of multiple set time nodes in history.

2. The intelligent control method for boiler wastewater treatment according to claim 1, characterized in that: The degree of fluctuation for: , For boiler wastewater equipment The degree of abnormality at the time node, For boiler wastewater equipment Dynamic lag time of time nodes, The natural constant An exponential function with base .

3. The intelligent control method for boiler wastewater treatment according to claim 1, characterized in that: The degree of abnormality for: , For boiler wastewater equipment The inlet and outlet pressure difference at the time node, It is the reference pressure difference of boiler wastewater equipment under normal operating conditions.

4. The intelligent control method for boiler wastewater treatment according to claim 1, characterized in that: The dynamic lag time for: , is the lag time of the boiler wastewater equipment in standby state, is the lag time of the boiler wastewater equipment under full load condition, For boiler wastewater equipment Wastewater flow at a given time point, The maximum wastewater flow rate carried by the boiler wastewater equipment, To set the time window size, is the normalization function.

5. The intelligent control method for boiler wastewater treatment according to claim 1, characterized in that: The loss function for: , For boiler wastewater equipment The degree of fluctuation at the time node, is the average value of the fluctuation degree of boiler wastewater equipment at multiple historical set time nodes, is the standard deviation of the fluctuation degree of boiler wastewater equipment at multiple historical set time nodes, Set the number of time points for history.

6. The intelligent control method for boiler wastewater treatment according to claim 5, characterized in that: when hour, ;when hour, .

7. The intelligent control method for boiler wastewater treatment according to claim 1, characterized in that: The machine learning model is a long short-term memory neural network model.

8. The intelligent control method for boiler wastewater treatment according to claim 1, characterized in that: The operating status is divided into normal and abnormal.

9. The intelligent control method for boiler wastewater treatment according to claim 1, characterized in that: It also includes training the machine learning model, specifically: The training set is input into the pre-built machine learning model for training. During the training process, the loss between the output prediction value and the label is calculated; the model parameters are adjusted using the gradient descent method to minimize the prediction error; the parameters of the machine learning model are iteratively adjusted until the loss is less than the set value or the set number of training times is reached, and finally a trained machine learning model is obtained.

10. An intelligent control system for boiler wastewater treatment, characterized in that: The invention comprises a memory and a processor, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an intelligent control method for boiler wastewater treatment according to any one of claims 1 to 9 is implemented.

Images