Data processing method for chemical process

Abstract

The invention relates to the technical field of wastewater treatment, and discloses a data processing method for a chemical process, which comprises the following steps: S1, separately and independently monitoring wastewater generation points in a chemical production line and acquiring corresponding parameters, S2, regularly acquiring feeding amount, equipment operation state and wastewater indexes to acquire time sequence data to prepare a training set and a test set, s3, the waste water treatment equipment predicts waste water indexes in future time according to the time sequence prediction model, and the waste water conveying flow regulation and control equipment is used; through a time sequence prediction mode, the equipment parameter change in a short period in the future can be obtained, and according to the relevance of variables, the future waste liquid index change is obtained according to the relevance between the waste liquid parameters and the equipment parameters while the equipment parameter change is obtained, so that future reference is provided for subsequent waste liquid treatment, and after equipment regulation and control are completed, the waste liquid index change is obtained. The predicted actual waste liquid reaches the waste liquid treatment equipment through a pipeline, and the environment in the treatment equipment is coupled with the changing waste water.

Description

Technical Field

[0001] This invention relates to the field of wastewater treatment technology, specifically a data processing method for chemical processes. Background Technology

[0002] In chemical production, raw materials are processed along the production line, and different processes generate different types of waste liquid. When the waste liquid is collected, it forms a mixed waste liquid, which is then treated uniformly by a wastewater treatment system. In practical applications, the flow of materials on the production line is sequential, so the parameters of the waste liquid flowing in the mixed waste liquid channel will change over time.

[0003] The waste liquid generated in the initial process first enters the integrated waste liquid channel. When the initial process transitions to subsequent processes, the waste liquid generated in the subsequent processes merges with the waste liquid from the initial process and is transported together. At this time, the parameters of the waste liquid in the waste liquid channel will change. If the waste liquid is to be treated comprehensively, the waste liquid treatment mechanism needs to be dynamically controlled. However, the wastewater treatment mechanism has a time frame for dynamic changes. If the properties of the waste liquid are collected and analyzed through a detection terminal, and the waste liquid treatment mechanism is adjusted until the adjustment is completed, the integrated waste liquid has already entered the waste liquid treatment mechanism, resulting in an inaccurate correspondence between the waste liquid treatment adjustment and the waste liquid changes. For example, the adjustment cycle for pH value includes addition, mixing, and complete dispersion; temperature adjustment requires adjusting the heating temperature until the liquid temperature is close to the heating temperature.

[0004] Time series forecasting is applied in chemical production processes to predict short-term parameters based on past production process parameters. It is often used to monitor the operating status of equipment during production and to control equipment in a timely manner when potential dangers are predicted. However, time series forecasting has not been applied to non-direct production processes, such as the issue of waste liquid variables generated in direct production processes. Summary of the Invention

[0005] The purpose of this invention is to provide a data processing method for chemical processes. By using time-series prediction, changes in equipment parameters in the near future can be obtained. Based on the correlation of variables, changes in waste liquid can be obtained simultaneously with changes in equipment parameters. By analyzing the correlation between waste liquid parameters and equipment parameters, future changes in waste liquid indicators can be obtained, thus providing a future reference for subsequent waste liquid treatment. After equipment regulation is completed, the predicted actual waste liquid is delivered to the waste liquid treatment equipment through pipelines, and the internal environment of the treatment equipment is coupled with the changing wastewater, thereby solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A data processing method for a chemical process includes: S1: Separately and independently monitor the wastewater generation points in the chemical production line and obtain corresponding parameters. The corresponding parameters include feed amount, equipment operating status, wastewater index, and the pipeline for transporting wastewater to the wastewater treatment equipment. Wastewater index includes wastewater flow rate, wastewater flow velocity, and wastewater component content. The wastewater pipelines discharged from several equipment are connected in series to the main pipeline, and the wastewater is uniformly discharged to the wastewater treatment equipment from the main pipeline. S2: Time-series data on feed rate, equipment operating status, and wastewater indicators are collected periodically to prepare training and testing sets for training the time-series prediction model. The future time predicted by the time-series prediction model is greater than the ratio of pipe flow rate to wastewater flow rate. S3: Wastewater treatment equipment uses a time-series prediction model to predict wastewater indicators and control wastewater flow rates in the future. The control content of the equipment includes feed rate, feed material, feed rate, mixing rate, and heating temperature. The control time of the equipment shall not exceed the sum of the ratio of the pipe flow rate to the wastewater flow rate and the future time predicted by the time-series prediction model.

[0007] As a further aspect of the present invention: S3A: The wastewater treatment equipment predicts wastewater indicators and wastewater transport flow rate in the future based on the time series prediction model. The control content of the equipment includes equipment preheating time, treatment tank water quality adjustment time, and equipment pre-start adjustment time. The control time of the equipment is not greater than the sum of the ratio of the pipe flow rate to the wastewater flow rate and the future time predicted by the time series prediction model.

[0008] As a further aspect of the present invention: in S2: the future time predicted by the time-series prediction model is greater than the ratio of the pipe flow rate to the wastewater flow rate, and is not less than the longest time among the equipment preheating time, the treatment tank water quality adjustment time, and the equipment pre-start adjustment time.

[0009] As a further aspect of the present invention: an external discharge pipe is installed on the wastewater conveying branch of the main pipeline connected to the equipment. If any parameter in the predicted data, such as the amount of feed, the equipment operating status, the wastewater index, or the wastewater conveyance exceeds the set threshold, the external discharge pipe is opened to prevent the waste liquid from entering the wastewater in the main pipeline.

[0010] As a further aspect of the present invention: in S3, the feeding speed and mixing speed are both controlled by the motor speed.

[0011] As a further aspect of the present invention: the training set in step S1 includes wastewater indicators generated under various processing environments and processing parameters of the equipment with the same feed amount, as well as wastewater indicators under different feed amounts. The processing environment and processing parameters include the equipment operating temperature, the feed amount and feed rate of gas, liquid and solid, and the equipment operating pressure.

[0012] As a further aspect of the present invention: a heating element is provided on the outer wall of the wastewater discharge main pipeline, and the heating element is a resistance wire, a resistance block, or a heat exchange fluid channel.

[0013] As a further aspect of the present invention: in step S2, the time-series prediction model is composed of a CNN-LSTM framework, and the corresponding parameters are normalized before being input into the model, and the training data is scaled to 0-1.

[0014] Compared with the prior art, the beneficial effects of the present invention are: By using time-series forecasting, we can know the changes in equipment parameters in the near future. Based on the correlation of variables, we can obtain the changes in waste liquid while obtaining the changes in equipment parameters. By analyzing the correlation between waste liquid parameters and equipment parameters, we can obtain the future changes in waste liquid indicators, thus providing a future reference for subsequent waste liquid treatment. After the equipment is regulated, the predicted actual waste liquid will reach the waste liquid treatment equipment through pipelines, and the internal environment of the treatment equipment will be coupled with the changing wastewater. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A flowchart illustrating a data processing method for a chemical process; Detailed Implementation

[0017] Please see Figure 1 Example 1: This embodiment includes the following steps: S1: Separately and independently monitor the wastewater generation points in the chemical production line and obtain corresponding parameters. The corresponding parameters include feed amount, equipment operating status, wastewater indicators, and the pipeline route for transporting wastewater to the wastewater treatment equipment. Wastewater indicators include wastewater flow rate, wastewater velocity, and wastewater component content. The wastewater pipelines discharged from several equipment are connected in series to the main pipeline, and then uniformly discharged to the wastewater treatment equipment by the main pipeline.

[0018] The chemical production line is broken down into processes or equipment that can generate wastewater, and their processing parameters are monitored and recorded synchronously.

[0019] In step S1, the training set includes wastewater indicators generated under various processing environments and processing parameters of the equipment with the same feed amount, as well as wastewater indicators under different feed amounts. The processing environment and processing parameters include equipment operating temperature, feed amount and feed rate of gas, liquid and solid, and equipment operating pressure.

[0020] The feed rate is reflected in the amount of raw materials fed in and the amount fed in each step. The amount of raw materials fed in indicates the baseline quantity of materials, while the amount fed in each step reflects the amount of auxiliary materials used. Both the amount of raw materials fed in and the amount of auxiliary materials fed in will affect the substance content and total amount of the final wastewater. The equipment operating status can also reflect changes in wastewater indicators, such as temperature, reaction, and equipment operating pressure. Therefore, the equipment operating status is also collected synchronously.

[0021] Secondly, it is necessary to record the pipeline route for transporting wastewater to the wastewater treatment equipment, as the pipeline route indicates the time of waste liquid release at each node.

[0022] S2: Time-series data on feed rate, equipment operating status, and wastewater indicators are collected periodically to prepare training and testing sets for training the time-series prediction model. The future time predicted by the time-series prediction model is greater than the ratio of pipe flow rate to wastewater flow rate.

[0023] The time series prediction model consists of a CNN-LSTM framework, and the corresponding parameters are normalized before being input into the model, with the training data scaled to 0-1.

[0024] Dataset partitioning involves dividing the dataset into training and testing sets. The training set is used to train the model, while the testing set is used to evaluate the model's performance and generalization ability. Using early warning indicators as the raw input data for the CNN-LSTM deep learning time-series prediction model, preprocessing the input data is necessary during deep learning to reduce noise and improve model performance and accuracy. This includes data cleaning such as deduplication, handling missing and outlier values, and error handling. Missing and outlier values ​​can affect model performance, so appropriate methods are needed. Common methods for handling missing values ​​include interpolation, deleting missing values, or replacing them with the mean or median. Outliers can be handled manually or using statistical methods. Alternatively, feature selection, feature scaling, or data transformation can be used. Considering that the data input used in this study consists of process parameters with many monitoring variables and large differences in numerical ranges, further steps are taken to enhance the model's generalization ability, accelerate its convergence, and ensure efficient training of the prediction model. To assess the effectiveness of the model's predictions, the study employed a data normalization preprocessing method, ultimately scaling the training data to 0-1.

[0025] Time-series forecasting can reveal short-term changes in equipment parameters. Based on the correlation between variables, changes in wastewater volume can be obtained alongside these changes. For example, low heating temperatures and incomplete reactions may lead to an increase in the content of a certain substance in the wastewater. Conversely, high heating temperatures and intense intermolecular motion may cause the loss of a certain substance through other means, thus reducing its content. An increase in the amount of liquid added may also lead to an increase in the volume of wastewater. Therefore, while obtaining changes in equipment parameters through time-series forecasting, the correlation between wastewater parameters and equipment parameters can be used to predict future changes in wastewater indicators, providing a future reference for subsequent wastewater treatment.

[0026] S3: Wastewater treatment equipment uses a time-series prediction model to predict wastewater indicators and control wastewater flow rates in the future. The control content of the equipment includes feed rate, feed material, feed rate, mixing rate, and heating temperature. The control time of the equipment shall not exceed the sum of the ratio of the pipe flow rate to the wastewater flow rate and the future time predicted by the time-series prediction model.

[0027] It is worth noting that time-series algorithms have high accuracy in short-term predictions. In order to meet the interval problem between waste liquid treatment and waste liquid generation, the future prediction time of the time-series algorithm is set to be greater than the ratio of the pipe flow rate to the wastewater flow rate. It is only necessary to ensure that the prediction time corresponds to the time when the wastewater arrives at the waste liquid treatment equipment.

[0028] For example, if the time series of parameters for process A predicts the temperature increase within the next t time period, the changes in waste liquid parameters can be known based on the correlation between waste liquid and process equipment parameters. When the prediction result is obtained, there is still residual wastewater in the pipe. At this time, the wastewater treatment equipment immediately adjusts the wastewater treatment parameters. The wastewater treatment parameters have an adjustment cycle. After the wastewater treatment parameters are adjusted, since the equipment adjustment time is no more than twice the ratio of the pipe flow rate to the wastewater flow rate, the predicted actual waste liquid reaches the waste liquid treatment equipment through the pipeline after the equipment adjustment is completed.

[0029] Secondly, the future time predicted by the time-series prediction model is greater than the ratio of pipe length to wastewater flow rate. This is to increase the margin of error. Assuming the predicted future time is 20 seconds, and the pipe length is the length of the wastewater flowing from the generation end to the treatment end, combined with the wastewater flow rate, the time it takes for the pipe to enter the treatment unit is 10 seconds. The time-series prediction value needs to be calculated based on the interrelationships of variables to obtain the wastewater index variables, which creates a time margin. After the calculation, assuming there are still 10 seconds left until the predicted time, since wastewater changes linearly under normal circumstances, although 10 seconds have passed, the changed wastewater has just entered the treatment unit. The treatment equipment, which has already begun to adjust, gradually changes the treatment environment, causing the gradually changing wastewater to enter the equipment. The equipment adjustment is coupled with the change in wastewater parameters. Since the equipment adjustment time is no greater than the sum of the ratio of pipe length to wastewater flow rate and the future time predicted by the time-series prediction model, the equipment adjusts to meet the standard of the time-series prediction value within 30 seconds, ensuring that the predicted wastewater enters the treatment equipment after 30 seconds. The environment inside the treatment equipment is coupled with the changing wastewater.

[0030] Additional explanation: Initially, the main pipeline only contains wastewater from the first process to generate wastewater. As time progresses, the second, third, and subsequent processes also generate wastewater. Furthermore, the pipe length from the second process entering the wastewater treatment equipment through the main pipeline is reduced, consequently reducing the length of wastewater from subsequent processes. To simplify the calculations, the pipe length from multiple processes entering the wastewater treatment equipment through the main pipeline can be kept consistent. If the pipe lengths are inconsistent, the convergence time of the multiple wastewater streams needs to be calculated. Alternatively, a staggered prediction method can be used.

[0031] For example, assuming the convergence time of the waste liquid from the first process and the second process is 5 seconds, the timing prediction for the first process is 20 seconds, and the timing prediction for the second process is 15 seconds, so that the convergence state of the waste liquid from the first process and the second process is matched, and so on.

[0032] Example 2: This example includes the following steps: S1: Separately and independently monitor the wastewater generation points in the chemical production line and obtain corresponding parameters. The corresponding parameters include feed amount, equipment operating status, wastewater index, and the pipeline for transporting wastewater to the wastewater treatment equipment. Wastewater index includes wastewater flow rate, wastewater flow velocity, and wastewater component content. The wastewater pipelines discharged from several equipment are connected in series to the main pipeline, and the wastewater is uniformly discharged to the wastewater treatment equipment from the main pipeline. S2: Time-series data on feed rate, equipment operating status, and wastewater indicators are collected periodically to prepare training and testing sets for training the time-series prediction model. The future time predicted by the time-series prediction model is greater than the ratio of pipe flow rate to wastewater flow rate. S3A: Wastewater treatment equipment predicts wastewater indicators and wastewater flow rate in the future based on a time-series prediction model. The equipment's control content includes equipment preheating time, treatment tank water quality adjustment time, and equipment pre-start adjustment time.

[0033] The difference between this embodiment and Embodiment 1 is that the control content of the equipment includes the equipment preheating time, the water quality adjustment time of the treatment tank, and the equipment pre-start adjustment time. This method is more suitable for time-series prediction because the equipment heating adjustment, water quality adjustment, and pre-start adjustment all require time to be reserved in advance.

[0034] When applied to Example 2, in S2: the future time predicted by the time-series prediction model is greater than the ratio of pipe flow rate to wastewater flow rate, and is not less than the longest time among equipment preheating time, treatment tank water quality adjustment time, and equipment pre-start adjustment time.

[0035] Since the time required for equipment heating adjustment, water quality adjustment, and pre-start adjustment is different, it is necessary to use the longest time as the standard to obtain the future time predicted by the time series prediction model, so as to ensure that the wastewater entering the wastewater treatment equipment after the prediction is in the optimal treatment environment.

[0036] An external discharge pipe is installed on the wastewater conveying branch of the main pipeline connected to the equipment. If any parameter in the predicted data, such as the amount of feed, the equipment operating status, the wastewater index, or any parameter in the wastewater conveyance exceeds the set threshold, the external discharge pipe will be opened to prevent the waste liquid from entering the wastewater in the main pipeline.

[0037] To prevent nonlinear or uncontrollable changes in equipment that could lead to unpredictable wastewater content, the external discharge pipe is opened for temporary wastewater storage. This process assesses the risks and prevents wastewater from entering the wastewater treatment equipment, which could cause problems that make the entire treatment environment uncontrollable.

[0038] In S3, both the feeding speed and the mixing speed are controlled by the motor speed. This motor control method is faster, more convenient, and has a quicker response.

[0039] Heating elements are installed on the outer wall of the main wastewater discharge pipeline. The heating elements are resistance wires, resistance blocks, or heat exchange fluid channels.

[0040] This method preheats the wastewater, preventing the ambient temperature from dropping as the wastewater enters the wastewater treatment equipment.

[0041] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A data processing method for a chemical process, characterized in that: include: S1: Separately and independently monitor the wastewater generation points in the chemical production line and obtain corresponding parameters. The corresponding parameters include feed amount, equipment operating status, wastewater index, and the pipeline for transporting wastewater to the wastewater treatment equipment. Wastewater index includes wastewater flow rate, wastewater flow velocity, and wastewater component content. The wastewater pipelines discharged from several equipment are connected in series to the main pipeline, and the wastewater is uniformly discharged to the wastewater treatment equipment from the main pipeline. S2: Time-series data on feed rate, equipment operating status, and wastewater indicators are collected periodically to prepare training and testing sets for training the time-series prediction model. The future time predicted by the time-series prediction model is greater than the ratio of pipe flow rate to wastewater flow rate. S3: Wastewater treatment equipment uses a time-series prediction model to predict wastewater indicators and control wastewater flow rates in the future. The control content of the equipment includes feed rate, feed material, feed rate, mixing rate, and heating temperature. The control time of the equipment shall not exceed the sum of the ratio of the pipe flow rate to the wastewater flow rate and the future time predicted by the time-series prediction model.

2. The data processing method for a chemical process according to claim 1, characterized in that: S3A: Wastewater treatment equipment uses a time-series prediction model to predict wastewater indicators and control wastewater flow rates in the future. The control measures include equipment preheating time, treatment tank water quality adjustment time, and equipment pre-start adjustment time. The control time of the equipment is no greater than the sum of the ratio of the pipe flow rate to the wastewater flow rate and the future time predicted by the time-series prediction model.

3. The data processing method for a chemical process according to claim 2, characterized in that: In S2: The future time predicted by the time series prediction model is greater than the ratio of pipe flow rate to wastewater flow rate, and is not less than the longest time among equipment preheating time, treatment tank water quality adjustment time, and equipment pre-start adjustment time.

4. The data processing method for a chemical process according to claim 1, characterized in that: An external discharge pipe is installed on the wastewater conveying branch of the main pipeline connected to the equipment. If any parameter in the predicted data, such as the amount of feed, the equipment operating status, the wastewater index, or the wastewater conveyance exceeds the set threshold, the external discharge pipe will be opened to prevent the waste liquid from entering the wastewater in the main pipeline.

5. The data processing method for a chemical process according to claim 1, characterized in that: In S3, both the feeding speed and the mixing speed are controlled by the motor speed.

6. The data processing method for a chemical process according to claim 1, characterized in that: In step S1, the training set includes wastewater indicators generated under various processing environments and processing parameters of the equipment with the same feed amount, as well as wastewater indicators under different feed amounts. The processing environment and processing parameters include equipment operating temperature, feed amount and feed rate of gas, liquid and solid, and equipment operating pressure.

7. The data processing method for a chemical process according to claim 1, characterized in that: Heating elements are installed on the outer wall of the main wastewater discharge pipeline. The heating elements are resistance wires, resistance blocks, or heat exchange fluid channels.

8. The data processing method for a chemical process according to claim 1, characterized in that: In step S2, the time series prediction model is composed of a CNN-LSTM framework, and the corresponding parameters are normalized before being input into the model, and the training data is scaled to 0-1.

Images