Control optimization method and device for extracting electronic-grade chemicals from wash oil and electronic equipment

Abstract

The invention relates to a control optimization method and device for extracting electronic-grade chemicals from wash oil and electronic equipment. The control optimization method comprises the following steps: acquiring operation parameters of separation and extraction equipment, target component indexes of fed raw materials corresponding to the operation parameters and target component indexes of products; performing feature selection on the operation parameters, the target component index of the fed raw material and the target component index of the product to obtain feature data; constructing a prediction model based on the feature data, and then training and evaluating the prediction model by using historical data to obtain a prediction model; and controlling the separation and extraction equipment to separate and extract the electronic-grade chemical products according to parameters obtained by the prediction model. According to the control optimization method provided by the invention, the process optimization method for extracting the electronic-grade chemicals from the wash oil based on intelligent control realizes accurate monitoring and control of the wash oil extraction process, and can improve the consistency of the product quality, the production efficiency and the traceability.

Description

Technical Field

[0001] The present invention relates to the field of chemical extraction technology, and specifically to a control optimization method, device and electronic equipment for extracting electronic-grade chemicals from wash oil, and more particularly to a control optimization method, control optimization device and electronic equipment for separating and extracting electronic-grade chemicals from coal tar wash oil fractions. Background Art

[0002] At present, coal tar washing oil fraction is an important chemical raw material, from which various electronic-grade chemicals such as β-methylnaphthalene, acenaphthene, fluorene, and dibenzofuran can be extracted.

[0003] For example, CN101643380A discloses a process for producing industrial fluorene from coal tar washing oil, which uses heavy washing oil obtained by removing components such as naphthalene, methylnaphthalene, dimethylnaphthalene, and acenaphthene from coal tar washing oil as raw material to produce industrial fluorene. The process is characterized in that the heavy washing oil is used as raw material to extract the fluorene fraction by continuous distillation, and the fraction is subjected to solvent crystallization or cooling crystallization, solvent recovery, recrystallization, and pressing separation steps (or washing, crystallization, and solvent recovery steps) to produce qualified industrial fluorene and obtain fluorene-rich residual oil, which is then added back to the raw material of the continuous distillation step.

[0004] CN103772132A discloses a method for extracting biphenyl from coal tar washing oil, which comprises the following production steps: (1) using a packed tower to distill the coal tar washing oil to obtain a crude biphenyl fraction with a boiling point of 208-212°C; (2) mixing the crude biphenyl fraction with an organic solvent, cooling and crystallizing, and then filtering to obtain a biphenyl product.

[0005] The current control methods in the process of deep processing of washing oil to extract electronic-grade chemicals mainly include manual operation and conventional automatic control.

[0006] Among them, manual operation relies on the experience and feeling of the operator, which is easily affected by subjective factors and cannot achieve precise control. Especially in the process of extracting electronic-grade chemicals from washing oil, the quality requirements are extremely high, and a slight misoperation may lead to fluctuations and instability in product quality, thus failing to meet the strict requirements of electronic-grade chemicals. Conventional automated control methods use preset fixed parameters for control, lacking the ability to monitor and flexibly adjust real-time data. This control method cannot cope with complex production environments and process changes, and often cannot adjust parameters in time to adapt to different batches of raw materials and production conditions. This may lead to instability and fluctuations in product quality, affecting production efficiency and cost control.

[0007] In summary, the instability and fluctuation of product quality when utilizing the coal tar wash oil fraction currently exist, which affects production efficiency and cost control. Summary of the invention

[0008] In view of the problems existing in the prior art, the purpose of the present invention is to provide a control optimization method, device and electronic equipment for extracting electronic-grade chemicals from wash oil, so as to solve the problem that when electronic-grade products are extracted from coal tar wash oil fractions, product quality is unstable and fluctuating, which affects production efficiency and cost control.

[0009] To achieve this object, the present invention adopts the following technical solutions:

[0010] In a first aspect, the present invention provides a control optimization method for extracting electronic-grade chemicals from wash oil, the control optimization method comprising:

[0011] S1. Obtaining the operating parameters of the separation and extraction equipment and the target component index of the raw material and the target component index of the product when the operating parameters correspond to the operating parameters;

[0012] S2. Perform feature selection on the operation parameters, target component indexes of the input raw materials, and target component indexes of the products to obtain feature data;

[0013] S3, building a prediction model based on the feature data, and then using historical data to train and evaluate the prediction model to obtain a prediction model;

[0014] S4. Based on the parameters obtained from the prediction model, the separation and extraction equipment is controlled to separate and extract electronic-grade chemical products.

[0015] The control optimization method provided by the present invention is a process optimization method for extracting electronic-grade chemicals from wash oil based on intelligent control, which realizes accurate monitoring, control and dynamic optimization of the wash oil extraction process and real-time adjustment of operating parameters, thereby improving product quality consistency, production efficiency and traceability.

[0016] As a preferred technical solution of the present invention, the separation and extraction equipment includes crude distillation equipment, rectification equipment and high-purity separation equipment.

[0017] Preferably, the operating parameters include temperature, pressure, reflux ratio and flow rate.

[0018] As a preferred technical solution of the present invention, the feature selection method includes correlation analysis or feature importance evaluation.

[0019] As a preferred technical solution of the present invention, if the correlation coefficient in the correlation analysis is greater than 0.7, the corresponding operating parameters and the index of the target component in the raw material and the index of the target component in the product when the corresponding operating parameters are input are output as characteristic data.

[0020] Preferably, in the feature importance evaluation, if the information gain is greater than 0.05 or the Gini index is less than 0.2, the corresponding operating parameters and the index of the target component in the raw material and the index of the target component in the product when the corresponding operating parameters are input are output as feature data.

[0021] As a preferred technical solution of the present invention, the construction of the prediction model includes: using feature data to obtain the prediction model with the help of a support vector machine model.

[0022] Preferably, the kernel function used to construct the prediction model includes a nonlinear basis kernel function.

[0023] As a preferred technical solution of the present invention, the historical data includes time series data of temperature, pressure, reflux ratio, feed position, flow rate and product concentration.

[0024] As a preferred technical solution of the present invention, the use of historical data to train and evaluate the prediction model includes: using historical data to perform training and evaluation using a cross-validation method.

[0025] Preferably, if the mean square error is less than 0.01 and the coefficient of determination is greater than 0.9 in the training and evaluation of the prediction model using historical data, the prediction model is obtained, otherwise the training and evaluation is continued.

[0026] As a preferred technical solution of the present invention, the prediction model is periodically trained and evaluated using current data during production to optimize the prediction model, or the prediction model is trained and evaluated when the mass percentage of the target component in the product fluctuates by more than 2% to optimize the prediction model.

[0027] In a second aspect, the present invention provides a control optimization device for extracting electronic-grade chemical products from washing oil, the control optimization device comprising:

[0028] A data acquisition module, used to acquire the operating parameters of the separation and extraction equipment and the target component index of the raw materials fed into the corresponding operating parameters and the target component index of the product;

[0029] The feature selection module performs feature selection on the operation parameters, the target component index of the input raw materials and the target component index of the product to obtain feature data;

[0030] Construct a training and evaluation module to construct a prediction model using feature data, and then use historical data to train and evaluate the prediction model to obtain a prediction model;

[0031] The application module is used to control the separation and extraction equipment to separate and extract electronic-grade chemical products based on the parameters obtained from the prediction model.

[0032] In a third aspect, the present invention provides an electronic device, the electronic device comprising:

[0033] at least one processor; and a memory communicatively coupled to the at least one processor;

[0034] Wherein, the memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the control optimization method for extracting electronic-grade chemical products from wash oil as described in the first aspect.

[0035] Compared with the prior art solutions, the present invention has the following beneficial effects:

[0036] (1) Improve product quality consistency: Through real-time monitoring and model-based intelligent adjustment, the impact of manual intervention on product quality is reduced, ensuring the stability of electronic-grade chemicals.

[0037] (2) Improve production efficiency: By dynamically adjusting operating parameters, resource waste caused by over-adjustment or lack of adjustment is reduced, the production process is optimized, and overall production efficiency is improved.

[0038] (3) Enhanced adaptive capabilities: Using artificial intelligence technologies such as reinforcement learning, the control system can automatically adjust the control strategy according to different batches of raw materials and environmental changes to adapt to complex and changing production conditions.

[0039] (4) Improved traceability: Intelligent control methods not only ensure the stability of product quality, but also provide detailed analysis reports for the production process through data recording and model analysis, supporting the traceability and optimization of the production process.

[0040] (5) Through the optimization method of the present invention, the process control of extracting electronic-grade chemicals from coal tar washing oil can be transformed from traditional preset fixed parameter control to a dynamic control method based on real-time data feedback and intelligent optimization, which greatly improves the stability of product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a flow chart of a control optimization method for extracting electronic-grade chemicals from wash oil provided by an embodiment of the present invention;

[0042] Figure 2 It is a schematic diagram of a control optimization device for extracting electronic-grade chemical products from wash oil provided by the present invention;

[0043] Figure 3 It is a schematic diagram of an electronic device provided by the present invention.

[0044] In the figure: 10 - electronic device, 11 - processor, 12 - ROM, 13 - RAM, 14 - bus, 15 - I / O interface, 16 - input unit, 17 - output unit, 18 - storage unit, 19 - communication unit.

[0045] The present invention is further described in detail below. However, the following examples are only simplified examples of the present invention and do not represent or limit the scope of protection of the present invention. The scope of protection of the present invention shall be subject to the claims. DETAILED DESCRIPTION

[0046] To better illustrate the present invention and facilitate understanding of the technical solution of the present invention, typical but non-limiting embodiments of the present invention are as follows:

[0047] This embodiment provides a control optimization method for extracting electronic grade chemical products from washing oil, such as Figure 1 As shown, the control optimization method includes:

[0048] S1. Obtaining the operating parameters of the separation and extraction equipment and the target component index of the raw material and the target component index of the product when the operating parameters correspond to the operating parameters;

[0049] S2. Perform feature selection on the operation parameters, target component indexes of the input raw materials, and target component indexes of the products to obtain feature data;

[0050] S3, building a prediction model based on the feature data, and then using historical data to train and evaluate the prediction model to obtain a prediction model;

[0051] S4. Based on the parameters obtained from the prediction model, the separation and extraction equipment is controlled to separate and extract electronic-grade chemical products.

[0052] Specifically, the separation and extraction equipment includes crude distillation equipment, rectification equipment and high-purity separation equipment.

[0053] Specifically, the operating parameters include temperature, pressure, reflux ratio and flow rate.

[0054] Exemplarily, the operating parameters are obtained by configuring sensors corresponding to the crude distillation equipment, the rectification equipment and the high-purity separation equipment, such as temperature sensors, pressure sensors and other equipment for detection, analysis and calculation.

[0055] Exemplarily, the high-purity separation equipment includes one or a combination of at least two of an adsorption separation equipment, a crystallization equipment, and an extraction equipment.

[0056] The key parameters such as temperature, pressure, flow rate and product content at the feed inlet, top outlet and bottom outlet of each intermediate crude product distillation tower and product distillation tower (including β-methylnaphthalene distillation tower, α-methylnaphthalene distillation tower, acenaphthene distillation tower, dibenzofuran distillation tower and fluorene distillation tower) in the wash oil separation process are collected.

[0057] Specifically, the feature selection includes correlation analysis or feature importance evaluation.

[0058] The correlation analysis is performed using the Pearson correlation coefficient, and the calculation formula is as follows:

[0059]

[0060] In the formula, r is the correlation coefficient, X i is the initial value of the operating parameters of the separation and extraction equipment corresponding to product i, Y i is the initial value of the target component index in the corresponding product i, is the average value of the initial values ​​of the corresponding operating parameters of all products, It is the average value of the initial values ​​of the target component indicators of all products.

[0061] If the correlation coefficient in the correlation analysis is greater than 0.7, the corresponding operating parameters and the index of the target component in the input raw material and the index of the target component in the product when the corresponding operating parameters are input are output as characteristic data.

[0062] The feature importance evaluation is calculated and evaluated by means of a CART decision tree model.

[0063] Among them, if the information gain is greater than 0.05 or the Gini index is less than 0.2 in the feature importance evaluation, the corresponding operating parameters and the index of the target component in the raw material and the index of the target component in the product when the corresponding operating parameters are input are output as feature data.

[0064] In the present invention, optionally, feature data obtained through feature importance evaluation are weighted averaged to screen out feature data that ultimately explain the target indicator.

[0065] Specifically, weighted averaging is to assign different weights to each feature based on its contribution to the prediction of the target variable, and obtain the final feature data through weighted averaging. The weighted averaging process is as follows:

[0066] Step 1: Evaluate the importance of each feature (such as using a decision tree algorithm to calculate information gain or Gini index);

[0067] Step 2: According to the evaluation results of each feature, weights are assigned. The weights can be assigned according to the feature importance score. For example, a feature with an information gain or a lower Gini index will receive a corresponding weight.

[0068] Step 3: Calculate the average weighted value. The formula is as follows:

[0069]

[0070] In the formula, X 加权 is the feature data after weighted average, X i is the i-th feature, f i The i-th weight;

[0071] Step 4: Filter out the feature data in the weighting tool as the feature set, which will be used for subsequent models and predictions.

[0072] Specifically, the construction of the prediction model includes: using feature data to perform modeling simulation with the help of a support vector machine model to obtain a prediction model.

[0073] Wherein, the kernel function used to construct the prediction model includes a nonlinear basis kernel function.

[0074] The exemplary construction process is as follows:

[0075] Step 1: Data Preparation

[0076] A data set for training is obtained based on the feature data. Assume that the feature data X has been screened out through correlation analysis and feature importance evaluation, and contains key operating parameters (such as temperature, pressure, reflux ratio, flow rate, etc.) related to the target indicator (such as product concentration or other quality indicators);

[0077] (1) Input data

[0078] ① Characteristic data: X = (X1, X2, X3, ..., X i ), where X1, X2, X3, ..., Xi are the characteristic data corresponding to product i;

[0079] ② Target index Y: i.e., target concentration or other target quality parameters of product i, preferably, the target product mass percentage, impurity type and mass percentage in the present invention;

[0080] (2) Divide the training set and test set:

[0081] Divide the data set into training set and test set, generally in the ratio of 80% training set and 20% test set;

[0082] Step 2: Build a Support Vector Machine Model (SVM)

[0083] The basic idea of ​​the support vector machine model (SVM) is to map data to a higher-dimensional feature space through a kernel function, and find a hyperplane in the space that can maximize the interval to achieve the regression goal;

[0084] (1) Select kernel function:

[0085] Available kernel functions: linear kernel function, nonlinear basis kernel function (RBF), or generative kernel function. The present invention selects nonlinear basis kernel function (RBF);

[0086] (2) SVM regression model training:

[0087] The core of training the SVM regression model using the data set is to map the data to a high-dimensional space through a kernel function, and find the best hyperplane in the space to minimize the deviation of the data points. Specifically, the radial basis RBF kernel function is used, and the training process of the model is as follows:

[0088]

[0089] In the formula, K(X, X i ) is the kernel function, α i is the Lagrange multiplier; b is the bias term, X is the current input feature data, representing the independent variable of the model, which is the operating parameters extracted during the oil wash separation process, such as temperature, pressure, reflux ratio and flow rate, X i is the characteristic value of each sample in the training data, γ is the parameter of the kernel function, which controls the similarity between data points;

[0090] The feature data X and target variable Y are input into the SVM regression model, and simulation calculation is performed to obtain the prediction model.

[0091] Specifically, the historical data includes time series data of temperature, pressure, reflux ratio, feed position, flow rate and product concentration.

[0092] The historical data should cover a certain time range to cover the changes under different operating conditions. An exemplary time range can be selected to cover a complete production cycle; the historical data should include a wide range of fluctuations in operating parameters to ensure that the model adapts to changes under different production conditions; the historical data should include production data with a production month of ≥ 6 to avoid overfitting and ensure the robustness of the model.

[0093] Specifically, the use of historical data to train and evaluate the prediction model includes: using the historical data to perform training and evaluation using a cross-validation method.

[0094] Exemplarily, the training process is as follows:

[0095] (1) Divide the data into k subsets; usually k = 5 or k = 10, that is, divide the data into k subsets, take one subset as the validation set in turn, and use the remaining data as the training set, and repeat the training and validation process;

[0096] (2) Use k-1 subsets for model training and the remaining 1 subset for validation;

[0097] (3) Calculate the mean square error (MSE) and determination coefficient (R 2 );

[0098] (4) Repeat k times and finally calculate the MSE and R of all k times2 The average value of .

[0099] Wherein, if the mean square error is less than 0.01 and the determination coefficient is greater than 0.9 in the training and evaluation of the prediction model using historical data, the prediction model is obtained, otherwise the training and evaluation is continued.

[0100] When the model meets the requirements, it can accurately predict the relationship between the operating parameters, the concentration of the target component in the input raw materials, and the concentration of the target component in the product. For example, when the concentration of the target component in the input raw materials and the concentration of the target component in the product are known, the model can be used to predict the corresponding operating parameters and then guide the extraction equipment to produce.

[0101] The qualified prediction model obtained through training will be used to monitor and control the extraction process in real time. By inputting the operating parameters and product concentration data collected in real time, the prediction model calculates the required optimal operating parameters (such as temperature, pressure, reflux ratio, flow rate, etc.). The system will adjust the equipment parameters according to these prediction results to ensure the stability and consistency of product quality. The real-time decision-making capability is to combine historical data with real-time feedback. The system can dynamically make decisions and adjust operating parameters based on the established prediction model and optimization algorithm to ensure that the production process always runs in the optimal state. This capability enables the system to cope with different production environments, process changes and raw material fluctuations, significantly improves the stability of the production process, reduces quality fluctuations, and ensures that the product always meets the predetermined quality standards.

[0102] Furthermore, in the control optimization method provided by the present invention, an early warning feedback process can also be set up. For example, by real-time monitoring of the changing trends of key parameters and comparing them with set thresholds, the system can promptly issue early warning signals to indicate potential problems or abnormal conditions, thereby reducing risks in the production process and prompting operators to take timely measures to ensure the stability and safety of the wash oil extraction process.

[0103] In the present invention, the design and selection of threshold values ​​are key factors to ensure the safety and stability of the wash oil extraction process. By setting threshold values ​​based on historical data, process requirements and equipment parameters, and taking into account the needs of dynamic adjustment, potential risks can be discovered in time and corresponding measures can be taken to ensure the stability and safety of the production process. The setting of threshold values ​​requires the collection of historical data, including long-term monitoring data of key parameters (such as temperature, pressure, reflux ratio, etc.). By analyzing these data, the normal fluctuation range of each parameter can be determined, and based on this range, the upper and lower threshold values ​​are set. There is no limitation in the present invention, and it can be selected specifically according to the conventional requirements in this field.

[0104] Furthermore, the threshold is not fixed, but adjusted according to the dynamic changes in the production process. For example, different seasons may cause changes in production conditions (such as raw material temperature, humidity, etc.), which in turn affects the threshold of the parameter. The threshold can be adjusted in different seasons or production batches based on historical data.

[0105] For example, if the production process changes (such as changes in raw material composition, equipment upgrades, etc.), the threshold may need to be reset. Dynamic adjustments can be made through real-time data monitoring and early warning system feedback.

[0106] Taking the reflux ratio threshold as an example, the following is explained: If the reflux ratio is set to 1.5, the reflux ratio fluctuation is usually 1.5±0.1. The thresholds are set as follows: Upper threshold: 1.6. If the reflux ratio exceeds 1.6, it may lead to reduced reaction efficiency or energy waste. Lower threshold: 1.4. If the reflux ratio is lower than 1.4, it may lead to incomplete extraction and affect product quality.

[0107] Furthermore, the data of key parameters recorded and stored in the present invention can not only meet the needs of product quality traceability and analysis and be prepared for future analysis and review, but also provide a detailed understanding of the entire separation and extraction process and provide support for quality management and improvement. At the same time, traceability records also help meet regulatory requirements and compliance with standards.

[0108] Furthermore, when production is carried out after the qualified model is determined, the model that meets the requirements is trained and evaluated periodically based on the current data or when the mass percentage of the target component in the product fluctuates by more than 2%, so as to improve the accuracy of the prediction model and optimize the model to ensure the accuracy and stability of the prediction model. By periodically training and evaluating the model, it is ensured that the prediction model can always adapt to changes in production conditions and maintain high prediction accuracy of the operation process when facing different raw materials or external factors. This optimization process is adaptive and can be continuously adjusted according to real-time data so that the production process is always in the best control state, thereby continuously optimizing operating parameters and improving production efficiency and product quality.

[0109] In the present invention, once the model construction evaluation is qualified, the changing trends of key parameters (mainly including: temperature, pressure, flow rate and product content at the feed inlet, tower top discharge port and tower bottom discharge port, etc.) can be predicted, factors affecting product quality can be identified, and support can be provided for subsequent control strategies.

[0110] For example, the specific control process is usually implemented through an industrial computer or a central control system (CCS). The intelligent control system collects real-time data from sensors and then uses an adaptive control algorithm to automatically adjust the parameters of related equipment based on these data to ensure that the oil washing extraction process can operate within the target range. These parameters include temperature, pressure, liquid level, flow rate, and chemical concentration.

[0111] The control system makes real-time decisions based on previously established models and algorithms to respond to different production environments and process changes. In this way, the system can maintain the consistency and stability of the wash oil extraction process, reduce product quality fluctuations, and ensure that production meets the predetermined quality standards.

[0112] Industrial computers or central control systems play a key role in this process, being able to process large amounts of real-time data and execute corresponding control commands to achieve the required parameter adjustments, which can greatly improve production efficiency, reduce the need for human intervention, and ensure the stability and consistency of the oil washing extraction process. This is very important for process optimization and quality control in industrial production.

[0113] Furthermore, the present invention provides a control optimization device for extracting electronic grade chemical products from washing oil, such as Figure 2 As shown, the control optimization device includes:

[0114] A data acquisition module, used to acquire the operating parameters of the separation and extraction equipment and the target component index of the raw materials fed into the corresponding operating parameters and the target component index of the product;

[0115] The feature selection module performs feature selection on the operation parameters, the target component index of the input raw materials and the target component index of the product to obtain feature data;

[0116] Construct a training and evaluation module to construct a prediction model using feature data, and then use historical data to train and evaluate the prediction model to obtain a prediction model;

[0117] The application module is used to control the separation and extraction equipment to separate and extract electronic-grade chemical products based on the parameters obtained from the prediction model.

[0118] Further, Figure 31 is a schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0119] like Figure 3 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0120] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0121] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a control optimization method for extracting electronic-grade chemical products from wash oil.

[0122] In some embodiments, the control optimization method for extracting electronic-grade chemical products from washed oil can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the control optimization method for extracting electronic-grade chemical products from washed oil described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to execute the control optimization method for extracting electronic-grade chemical products from washed oil by any other appropriate means (e.g., by means of firmware).

[0123] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0124] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0125] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0126] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0127] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0128] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0129] The actual application process is as follows:

[0130] Example 1

[0131] The following takes 1000 tons / hour of washing oil raw material processing capacity as an example to illustrate that in the process of washing oil deep processing to extract electronics and chemicals, the raw material feed rate is 1000 tons / hour. The mass percentage of each component in the washing oil is shown in Table 1. The intelligent control steps for washing oil extraction of electronic grade chemicals are as follows:

[0132] S1, sensor data collection stage

[0133] The sensor equipment is used to collect the temperature, pressure, flow rate and mass percentage parameters of the feed inlet, tower top and tower bottom of each intermediate process crude product distillation tower and product distillation tower (including β-methylnaphthalene distillation tower, α-methylnaphthalene distillation tower, acenaphthene distillation tower, dibenzofuran distillation tower and fluorene distillation tower) in the wash oil separation process at high frequency, and transmit the data to the subsequent processing and analysis system;

[0134] S2. Data processing and analysis stage

[0135] Through algorithms and data processing technology, the collected data is processed and analyzed to obtain data information that is valuable for improving the process, mainly including:

[0136] Perform feature selection on the operating parameters, the target component index of the input raw materials, and the target component index of the product to obtain feature data, and use the CART decision tree model to perform calculation and evaluation. If the information gain is greater than 0.05 or the Gini index is less than 0.2, the corresponding operating parameters and the index of the target component in the input raw materials and the target component index in the product when the corresponding operating parameters are used as feature data are output;

[0137] To build the model:

[0138] Step 1: Data Preparation

[0139] A data set for training is obtained based on the feature data. The feature data X has been screened out through correlation analysis and feature importance evaluation, and includes key operating parameters (such as temperature, pressure, reflux ratio, flow rate, etc.) related to the target indicator (such as product concentration or other quality indicators);

[0140] (1) Input data

[0141] ① Characteristic data: X = (X1, X2, X3, ..., X i ), where X1, X2, X3, ..., X i is the characteristic data corresponding to product i;

[0142] ② Target index Y: i.e., target concentration or other target quality parameters of product i, preferably, the target product mass percentage, impurity type and its mass percentage;

[0143] (2) Divide the training set and test set:

[0144] Divide the data set into training set and test set, generally in the ratio of 80% training set and 20% test set;

[0145] Step 2: Build a Support Vector Machine Model (SVM)

[0146] The basic idea of ​​the support vector machine model (SVM) is to map data to a higher-dimensional feature space through a kernel function, and find a hyperplane in the space that can maximize the interval to achieve the regression goal;

[0147] (1) Select kernel function:

[0148] The present invention selects a nonlinear basis kernel function (RBF);

[0149] (2) SVM regression model training:

[0150] Use the data set to train the SVM regression model, specifically using the RBF kernel function. The model training process is as follows:

[0151]

[0152] In the formula, K(X, X i ) is the kernel function, α i is the Lagrange multiplier; b is the bias term; X is the current input feature data, representing the independent variable of the model, which is the operating parameters extracted during the oil wash separation process, such as temperature, pressure, reflux ratio and flow rate, X i is the characteristic value of each sample in the training data, γ is the parameter of the kernel function, which controls the similarity between data points;

[0153] The feature data X and target variable Y are input into the SVM regression model, and simulation calculation is performed to obtain the prediction model.

[0154] The cross-validation method is used to conduct training evaluation with the help of historical data. The training process is as follows:

[0155] (1) Divide the data into k subsets; usually k = 5 or k = 10, and in this embodiment k = 10 is selected, that is, divide the data into k subsets, take one subset as the validation set in turn, and take the remaining data as the training set, and repeat the training and validation process;

[0156] (2) Use k-1 subsets for model training and the remaining 1 subset for validation;

[0157] (3) Calculate the mean square error (MSE) and determination coefficient (R 2 );

[0158] (4) Repeat k times and finally calculate the MSE and R of all k times 2 The average value of .

[0159] Wherein, if the mean square error is less than 0.01 and the determination coefficient is greater than 0.9 in the training and evaluation of the prediction model using historical data, the prediction model is obtained, otherwise the training and evaluation is continued.

[0160] S3, Intelligent Control Strategy Stage

[0161] Automatically adjust the operating parameters according to the operating parameters obtained from the prediction model to achieve precise control. Based on the control algorithm and optimization model, the intelligent control system dynamically adjusts the operating parameters according to the real-time data and prediction model to achieve the best state of the process of washing oil separation and extraction of high value-added products. The use of adaptive and intelligent control strategies can ensure the consistency and stability of product quality.

[0162] Example 2

[0163] On the basis of Example 1, a real-time monitoring and early warning stage is added to ensure the stability and safety of the wash oil extraction process. By real-time monitoring of the changing trends of key parameters (temperature, pressure, reflux ratio) and comparing them with the set thresholds, the system can issue early warning signals in a timely manner to indicate potential problems or abnormal conditions, thereby reducing the risks in the production process and prompting operators to take timely measures to ensure the stability and safety of the wash oil extraction process.

[0164] The threshold values ​​in this embodiment are set as follows:

[0165] (1) The reflux ratio threshold is set as follows:

[0166] The reflux ratio fluctuates to 1.5±0.8, and the thresholds are set as follows:

[0167] During the process, the upper limit threshold of the reflux ratio of the crude fractionation tower is 1.8, and the lower limit threshold is 1.3; the upper limit threshold of the reflux ratio of the β-methylnaphthalene distillation tower is 2.0, and the lower limit threshold is 1.4; the upper limit threshold of the reflux ratio of the acenaphthene distillation tower is 2.2, and the lower limit threshold is 1.5; the upper limit threshold of the reflux ratio of the dibenzofuran distillation tower is 2.3, and the lower limit threshold is 1.6; the upper limit threshold of the reflux ratio of the fluorene distillation tower is 2.5, and the lower limit threshold is 1.7.

[0168] (2) The pressure threshold is set as follows:

[0169] The pressure fluctuation range of the tower bottom is usually between 5 bar ± 1.0 bar, and the pressure fluctuation range of the tower top is usually between 4 bar ± 1.0 bar, so the threshold values ​​are set as follows:

[0170] The upper threshold of the bottom pressure of the crude fraction distillation tower is 5.5 bar, and the lower threshold is 4.5 bar; the upper threshold of the bottom pressure of the β-methylnaphthalene distillation tower is 5.4 bar, and the lower threshold is 4.6 bar; the upper threshold of the bottom pressure of the acenaphthene distillation tower is 5.3 bar, and the lower threshold is 4.7 bar; the upper threshold of the bottom pressure of the dibenzofuran distillation tower is 5.6 bar, and the lower threshold is 4.4 bar; the upper threshold of the bottom pressure of the fluorene distillation tower is 5.5 bar, and the lower threshold is 4.5 bar;

[0171] During the process, the upper threshold value of the top pressure of the crude fractionation tower is 4.5 bar, and the lower threshold value is 3.5 bar; the upper threshold value of the top pressure of the β-methylnaphthalene distillation tower is 4.3 bar, and the lower threshold value is 3.7 bar; the upper threshold value of the top pressure of the acenaphthene distillation tower is 4.2 bar, and the lower threshold value is 3.8 bar; the upper threshold value of the top pressure of the dibenzofuran distillation tower is 4.4 bar, and the lower threshold value is 3.6 bar; the upper threshold value of the top pressure of the fluorene distillation tower is 4.4 bar, and the lower threshold value is 3.6 bar.

[0172] (3) Temperature threshold

[0173] The temperature fluctuation range of the tower bottom is usually 210℃±10℃, and the temperature fluctuation range of the tower top is usually 180℃±10℃, so the threshold values ​​are set as follows:

[0174] The upper limit threshold of the bottom temperature of the crude fraction distillation tower in the process is 205℃, and the lower limit threshold is 195℃; the upper limit threshold of the bottom temperature of the β-methylnaphthalene distillation tower is 210℃, and the lower limit threshold is 200℃; the upper limit threshold of the bottom temperature of the acenaphthene distillation tower is 210℃, and the lower limit threshold is 200℃; the upper limit threshold of the bottom temperature of the dibenzofuran distillation tower is 215℃, and the lower limit threshold is 205℃; the upper limit threshold of the bottom temperature of the fluorene distillation tower is 215℃, and the lower limit threshold is 205℃;

[0175] During the process, the upper limit threshold of the top temperature of the crude fractionation tower is 180°C, and the lower limit threshold is 170°C; the upper limit threshold of the top temperature of the β-methylnaphthalene distillation tower is 185°C, and the lower limit threshold is 175°C; the upper limit threshold of the top temperature of the acenaphthene distillation tower is 185°C, and the lower limit threshold is 175°C; the upper limit threshold of the top temperature of the benzofuran distillation tower is 190°C, and the lower limit threshold is 180°C; the upper limit threshold of the top temperature of the fluorene distillation tower is 190°C, and the lower limit threshold is 180°C.

[0176] Example 3

[0177] Based on Example 2, data recording and traceability stages are added to meet the needs of product quality traceability and analysis for future analysis and review. These data can provide a detailed understanding of the entire separation and extraction process and provide support for quality management and improvement. At the same time, traceability records also help meet regulatory requirements and compliance with standards.

[0178] Comparative Example 1

[0179] Conventional automatic control mode is adopted, and preset fixed parameters are used for control, as follows:

[0180] Conventional automation control methods usually use preset fixed parameters to control the operation of equipment. The specific control process includes:

[0181] (1) Parameter setting

[0182] Under conventional automated control, all operating parameters (such as temperature, pressure, reflux ratio, and flow rate) are pre-set fixed values ​​and will not be dynamically adjusted in real time based on data from the production process;

[0183] (2) Control process

[0184] Set target temperature: Set a fixed temperature value of 300°C, and the temperature will remain at this set value during the entire production process;

[0185] Set target pressure: Set a pressure value of 5 bar, and the device will always maintain this fixed pressure value;

[0186] Set flow rate: Set the flow rate to a fixed flow rate of 50L / min;

[0187] Set reflux ratio: The reflux ratio is set to a fixed value of 1.5, which will not be adjusted during the whole process;

[0188] (3) Control method

[0189] Traditional PID control: The system uses a PID controller to adjust these fixed parameters. The PID controller calculates the deviation between the current operating parameters and the target parameters to adjust the equipment operating status, but these adjustments depend on fixed set values.

[0190] Lack of real-time optimization: In this mode, the equipment cannot automatically adjust parameters based on data feedback during the production process, and has no adaptive ability to changes in the production environment;

[0191] Fixed parameter example:

[0192] The fixed values ​​of temperature, pressure, and flow rate set during the wash oil extraction process were 300 °C, 5 bar, 50 L / min, and the reflux ratio was 1.5;

[0193] In this conventional control mode, the system will always maintain these fixed parameters regardless of changes in feedstock quality or process conditions.

[0194] The mass percentages of the target products in Examples 1-3 and Comparative Example 1 of the present invention are shown in Table 1.

[0195] Table 1

[0196] Product Name β-Methylnaphthalene / % α-Methylnaphthalene / % Acenaphthene / % Fluorene / % Dibenzofuran / % other / % Washing oil raw materials 15 5 20 30 15 15 Example 1 99.95 99.82 99.93 99.97 95.1 —— Example 2 99.90 99.31 99.90 99.93 96.0 —— Example 3 99.97 99.72 99.96 99.92 95.6 —— Comparative Example 1 95％ 90％ 96.5％ 97.8％ 95.3％ ——

[0197] In case 1 of product quality improvement, the intelligent control method was used to increase the product yield from 80% under the traditional method to more than 85%, while the product quality indicators were stabilized within the specified range. This means that more high-value-added chemicals can be obtained from the oil washing extraction process, improving the quality consistency of the products and enhancing their competitiveness.

[0198] In terms of improving production efficiency, both Case 1 and Case 2 have achieved significant results. In Case 1, the intelligent control method achieved improved production efficiency by optimizing operating parameters, which was specifically manifested in reducing energy consumption and raw material losses and improving resource utilization. In Case 2, the intelligent control method improved production efficiency by more than 10% compared with traditional methods, reducing waste and energy consumption in the production process.

[0199] For case 3 of enhanced traceability, the intelligent control system realizes data recording and storage of key parameters. This data recording can meet the needs of product quality traceability, while improving the manageability and quality assurance capabilities of the production process. Operators can trace the production process and key parameter changes of each batch of products, discover potential problems in time, and take corresponding measures to adjust and improve.

[0200] It is stated that the present invention illustrates the detailed structural features of the present invention through the above embodiments, but the present invention is not limited to the above detailed structural features, that is, it does not mean that the present invention must rely on the above detailed structural features to be implemented. Those skilled in the art should understand that any improvement of the present invention, equivalent replacement of the components selected by the present invention, addition of auxiliary components, selection of specific methods, etc., all fall within the protection scope and disclosure scope of the present invention.

[0201] The preferred embodiments of the present invention are described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, a variety of simple modifications can be made to the technical solution of the present invention, and these simple modifications all belong to the protection scope of the present invention.

[0202] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the present invention will not further describe various possible combinations.

[0203] In addition, various embodiments of the present invention may be arbitrarily combined, and as long as they do not violate the concept of the present invention, they should also be regarded as the contents disclosed by the present invention.

Claims

1. A control optimization method for extracting electronic-grade chemicals from wash oil, characterized in that: The control optimization method comprises: S1. Obtaining the operating parameters of the separation and extraction equipment and the target component index of the raw material and the target component index of the product when the operating parameters correspond to the operating parameters; S2. Perform feature selection on the operation parameters, target component indexes of the input raw materials, and target component indexes of the products to obtain feature data; S3, building a prediction model based on the feature data, and then using historical data to train and evaluate the prediction model to obtain a prediction model; S4. Based on the parameters obtained from the prediction model, the separation and extraction equipment is controlled to separate and extract electronic-grade chemical products.

2. The optimization control method according to claim 1, characterized in that: The separation and extraction equipment includes crude distillation equipment, rectification equipment and high-purity separation equipment; Preferably, the operating parameters include temperature, pressure, reflux ratio and flow rate.

3. The optimization control method according to claim 1 or 2, characterized in that: The feature selection method includes correlation analysis or feature importance evaluation.

4. The optimization control method according to claim 3, characterized in that: If the correlation coefficient in the correlation analysis is greater than 0.7, the corresponding operating parameters and the target component index in the raw material and the target component index in the product when the corresponding operating parameters are input are output as characteristic data; Preferably, in the feature importance evaluation, if the information gain is greater than 0.05 or the Gini index is less than 0.2, the corresponding operating parameters and the index of the target component in the raw material and the index of the target component in the product when the corresponding operating parameters are input are output as feature data.

5. The optimization control method according to any one of claims 1 to 4, characterized in that: The construction of the prediction model includes: using feature data to perform modeling simulation with the support vector machine model to obtain the prediction model; Preferably, the kernel function used to construct the prediction model includes a nonlinear basis kernel function.

6. The optimization control method according to any one of claims 1 to 5, characterized in that: The historical data includes time series data of temperature, pressure, reflux ratio, feed position, flow rate and product concentration.

7. The optimization control method according to any one of claims 1 to 6, characterized in that: The use of historical data to train and evaluate the prediction model includes: using the historical data to perform training and evaluation using a cross-validation method; Preferably, if the mean square error is less than 0.01 and the coefficient of determination is greater than 0.9 in the training and evaluation of the prediction model using historical data, the prediction model is obtained, otherwise the training and evaluation is continued.

8. The optimization control method according to any one of claims 1 to 7, characterized in that: The prediction model is periodically trained and evaluated using current data during production to optimize the prediction model, or when the mass percentage of the target component in the product fluctuates by more than 2%, the prediction model is trained and evaluated to optimize the prediction model.

9. A control optimization device for extracting electronic grade chemical products from washing oil, characterized in that: The control optimization device comprises: A data acquisition module, used to acquire the operating parameters of the separation and extraction equipment and the target component index of the raw materials fed into the corresponding operating parameters and the target component index of the product; The feature selection module performs feature selection on the operation parameters, the target component index of the input raw materials and the target component index of the product to obtain feature data; Construct a training and evaluation module to construct a prediction model using feature data, and then use historical data to train and evaluate the prediction model to obtain a prediction model; The application module is used to control the separation and extraction equipment to separate and extract electronic-grade chemical products based on the parameters obtained from the prediction model.

10. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively coupled to the at least one processor; Wherein, the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the control optimization method for extracting electronic-grade chemical products from wash oil according to any one of claims 1-8.

Images