Layering method of extraction kettle

Abstract

The invention provides a layering method of an extraction kettle, which comprises the following steps: stirring materials in a first extraction kettle, and forming a first water phase layer, a first impurity layer and a first oil phase layer with fuzzy layering interfaces in the first extraction kettle after the reaction of the materials is completed; discharging the first water phase layer from the bottom of the first extraction kettle, enabling the first water phase layer to flow through a first sight glass, and continuously collecting pictures in the first sight glass; the server analyzes the picture in the first sight glass through a picture prediction algorithm, and when impurities appear in the picture in the first sight glass, the picture prediction algorithm predicts the time when the first water phase layer is discharged out of the first sight glass; the server transmits the time when the first water phase layer is discharged out of the first sight glass to a central control system, the central control system automatically adjusts a valve body of the first extraction kettle, and the first water phase layer is collected. And the time when the water phase layer is completely discharged is predicted through a picture prediction algorithm, and the state of the valve body is automatically adjusted according to the prediction time, so that sectional discharging of the reaction feed liquid is realized.

Description

Technical Field

[0001] This application relates to the field of industrial automation, and particularly to a layering method for an extraction kettle. Background Art

[0002] With the continuous innovation and development of science and technology in our country, the production methods of enterprises have also changed greatly, from the traditional production and manufacturing mode that relies on manual labor in the production process to the automated production that gradually uses automated production equipment to replace manual labor. Along with the elimination of manual labor, the operating conditions of automated production have become increasingly important for production safety and enterprise efficiency.

[0003] In traditional layering operations, it usually relies on the human eye to observe the sight glass below the extraction kettle and manually adjust the opening degree of the regulating valve to achieve material separation. However, due to the similar colors of some material liquids, there is no obvious layering interface during the layering process; the experience and operation methods of each staff member are different; the reaction phenomena of different materials are diverse and not obvious. Therefore, it is easy for the human eye to make misjudgments, resulting in certain errors in controlling the layering. Such misjudgments often lead to problems such as insufficient product purity and material mixing. If completely relying on manual operation, it is difficult to ensure the consistency of the process, which has a negative impact on the quality of the produced products and even brings potential safety hazards in production. Summary of the Invention

[0004] Therefore, in order to overcome the above-mentioned shortcomings of the prior art, this application proposes a layering method for an extraction kettle, which solves the problems that some material liquids have similar colors, there is no obvious layering interface during the layering process; the experience and operation methods of each staff member are different; the reaction phenomena of different materials are diverse and not obvious, it is easy for the human eye to make misjudgments, resulting in certain errors in controlling the layering, and the obtained layer contains impurities.

[0005] This application provides a layering method for an extraction kettle, including: S1, stirring the materials in the first extraction kettle, waiting for the materials to complete the reaction, and a first aqueous phase layer, a first impurity layer, and a first oil phase layer with a blurred layering interface are formed in the first extraction kettle; S2, discharging the first aqueous phase layer from the bottom of the first extraction kettle, the first aqueous phase layer flows through a first sight glass, and pictures in the first sight glass are continuously collected; S3, the server analyzes the pictures in the first sight glass through a picture prediction algorithm, and when impurities appear in the pictures in the first sight glass, the picture prediction algorithm predicts the time when the first aqueous phase layer discharges from the first sight glass; S4, the server transmits the time when the first aqueous phase layer discharges from the first sight glass to the central control system, and the central control system automatically adjusts the valve body of the first extraction kettle to collect the first aqueous phase layer.

[0006] Specifically, the layering method of the extraction kettle further includes: S501, blowing the impurity layer and the oil phase layer in the first sight glass back into the first extraction kettle, adding clear water into the first extraction kettle, stirring the materials in the first extraction kettle, waiting for the reaction of the materials to be completed, and a second aqueous phase layer, a second impurity layer, and a second oil phase layer with a blurred layering interface are formed in the first extraction kettle; S502, discharging the second aqueous phase layer from the bottom of the first extraction kettle, the second aqueous phase layer flows through the first sight glass, and pictures in the first sight glass are continuously collected; S503, the server analyzes the pictures in the first sight glass through a picture prediction algorithm, and when impurities appear in the pictures in the first sight glass, the picture prediction algorithm predicts the time when the second aqueous phase layer discharges from the first sight glass; S504, the server transmits the time when the second aqueous phase layer discharges from the first sight glass to the central control system, and the central control system automatically adjusts the valve body of the first extraction kettle to collect the second aqueous phase layer.

[0007] Specifically, the layering method of the extraction kettle further includes: S601, collecting the first aqueous phase layer or the first aqueous phase layer and the second aqueous phase layer into the second extraction kettle, stirring the materials in the second extraction kettle, waiting for the reaction of the materials to be completed, and standing for a period of time until complete layering, a third aqueous phase layer, a third emulsion layer, and a third oil phase layer with an obvious layering interface are formed in the second extraction kettle; S602, discharging the third aqueous phase layer from the bottom of the second extraction kettle, the materials flow through the second sight glass, and pictures in the second sight glass are continuously collected; S603, the server analyzes the pictures in the second sight glass through a layering detection algorithm, and when a layering interface appears in the pictures in the second sight glass, the layering detection algorithm obtains the proportion of each layer of liquid; S604, the server transmits the proportion of each layer of liquid to the central control system, and the central control system automatically adjusts the valve body of the second extraction kettle to collect the third aqueous phase layer.

[0008] Specifically, the layering method of the extraction kettle further includes: S701, discharging the first impurity layer and the first oil phase layer in the first extraction kettle, then collecting the third aqueous phase layer into the first extraction kettle, standing the third aqueous phase layer in the first extraction kettle for a period of time, and a fourth aqueous phase layer, a fourth emulsion layer, and a fourth oil phase layer with an obvious layering interface are formed in the first extraction kettle; S702 discharging the fourth aqueous phase layer from the bottom of the first extraction kettle, the fourth aqueous phase layer flows through the first sight glass, and pictures in the first sight glass are continuously collected; S703, the server analyzes the pictures in the first sight glass through a layering detection algorithm, and when a layering interface appears in the pictures in the first sight glass, the layering detection algorithm obtains the proportion of each layer of liquid; S704, the server transmits the proportion of each layer of liquid to the central control system, and the central control system automatically adjusts the valve body of the first extraction kettle to collect the fourth aqueous phase layer.

[0009] Specifically, the central control system includes a DCS-700 system. The method uses the structured text language and function block diagram programming language in the DCS-700 system, and the input signals cover analog input, analog output, digital input, and digital output.

[0010] Specifically, the valve body includes an electric control valve and a cut-off valve.

[0011] Specifically, the outflow of materials in the first extraction kettle and the second extraction kettle is achieved by nitrogen-pressurized materials.

[0012] Specifically, the server is connected through a network. The server communicates with the RTU transmission model protocol gateway, transmits the liquid level ratio signal to the central control system. The central control system converts the preset time signal and the liquid level ratio signal into the opening signal of the electric control valve, and gradually reduces the opening of the control valve.

[0013] Specifically, the pictures in the first sight glass and the pictures in the second sight glass are collected by a high-speed camera.

[0014] Compared with the prior art, a layering method of an extraction kettle provided by the present application combines automation technology and picture prediction algorithm technology. By collecting impurities appearing in the pictures in the sight glass of the extraction kettle, the system can obtain the discharging time, thereby automatically adjusting the state of the valve body, so as to achieve segmented discharging of the reaction liquid, and obtain an impurity-free aqueous layer. This not only reduces the labor cost, reduces the losses caused by human errors, but also improves the resource utilization rate, ensures the consistency and accuracy of the production process, and reduces the errors caused by human factors. At the same time, the method has higher efficiency, lower cost, better quality control and stronger adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0016] Figure 1 is a schematic flow chart of the extraction kettle layering method provided by the present application;

[0017] Figure 2 is a schematic flow chart of the four-stage layering process provided by the present application;

[0018] Figure 3 is a schematic diagram of the extraction kettle for the four-stage layering process provided by the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0020] The following specific examples illustrate the implementation manners of the present application. Those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The present application can also be implemented or applied through other different specific implementation manners. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that, without conflict, the following embodiments and the features in the embodiments can be combined with each other. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without making creative efforts belong to the scope of protection of the present application.

[0021] It should be noted that the following description relates to various aspects of embodiments within the scope of the appended claims. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is illustrative only. Based on the present application, those skilled in the art should understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number and aspects described herein can be used to implement a device and / or practice a method. Additionally, this device and / or method can be implemented using other structures and / or functionality in addition to one or more of the aspects described herein.

[0022] It should also be noted that the drawings provided in the following embodiments only schematically illustrate the basic concept of the present application. The drawings only show the components related to the present application, rather than being drawn according to the number, shape, and size of the components in actual implementation. The type, quantity, and ratio of each component in actual implementation can be arbitrarily changed, and the component layout type may also be more complex.

[0023] In addition, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the examples can be practiced without these specific details.

[0024] After the materials react in the extraction kettle, when the layers formed after the reaction are not allowed to stand still to form different layers, the topmost layer is the oil phase, the second layer is the impurity layer, the third layer is the water phase layer, and the emulsion layer is the intermediate product of the reaction. The control of the standing time and temperature will affect the existence of the emulsion layer. When the layers formed after the reaction are not allowed to stand still, the emulsion layer does not exist, and the layer interface obtained when the layers formed after the reaction are not allowed to stand still is not clear.

[0025] Impurities are the residues after chemical reactions between drugs. The impurities exhibit different colors from the oil phase layer and the water phase layer under different lights. Due to natural light issues, the color of the impurities detected and visible to the naked eye is black. In the following steps, the visible impurities are all referred to as black substances.

[0026] The picture prediction algorithm (LSTM) is a special type of recurrent neural network (RNN) used to process time series data. By obtaining the layered video of the intelligent sight glass under the extraction kettle, especially the video of the impurity layer from empty to full in the sight glass, as the input of the LSTM algorithm, frame extraction is performed on the video to form a sequence of 10 - 20 frames. The neural network (CNN) module is used for feature extraction. The LSTM unit consists of three key gates (input gate, forget gate, output gate) and a cell state. The extracted feature sequence is input into the LSTM module through the input gate. LSTM updates the cell state and hidden state through the gating mechanism, capturing the temporal dependencies between frames. In the prediction stage, LSTM outputs a time quantity using the current updated cell state and hidden state. The system uses the time quantity output by LSTM to control the opening of the solenoid valve in advance, so that the impurity layer remains in the extraction kettle.

[0027] Drain the un - static - settled material in layers to obtain the first water phase layer. As Figure 1 shown, this embodiment provides a method for layering in an extraction kettle, including the following steps:

[0028] S1, Stir the material in the first extraction kettle, let it stand and wait for the material reaction to complete; in the first extraction kettle, a first water phase layer with a blurred layering interface, a first impurity layer, and a first oil phase layer are formed.

[0029] S2, Drain the first water phase layer from the bottom of the first extraction kettle. The first water phase layer flows through the first sight glass, and pictures in the first sight glass are continuously collected through a camera.

[0030] S3, The server analyzes the pictures in the first sight glass through the picture prediction algorithm. When black substances appear in the pictures in the first sight glass, that is, impurities that appear black under natural light, the picture prediction algorithm predicts the time when the first water phase layer completely drains out of the first sight glass and the first extraction kettle.

[0031] S4, The server transmits the time when the first water phase layer completely drains out of the first sight glass and the first extraction kettle to the central control system, and the central control system automatically adjusts the valve body of the first extraction kettle to collect the first water phase layer in the first extraction kettle.

[0032] In one embodiment, the method for layering in the extraction kettle further includes:

[0033] S501. Blow the impurity layer and the oil phase layer in the first sight glass into the first extraction kettle, add clear water to the first extraction kettle, stir the materials in the first extraction kettle, wait for the reaction of the materials to complete, and a second aqueous phase layer, a second impurity layer, and a second oil phase layer with a blurred stratification interface are formed in the first extraction kettle.

[0034] S502. Drain the second aqueous phase layer from the bottom of the first extraction kettle. The second aqueous phase layer flows through the first sight glass, and pictures in the first sight glass are continuously collected through a camera.

[0035] S503. The server analyzes the pictures in the first sight glass through a picture prediction algorithm. When impurities appear in the pictures in the first sight glass, the picture prediction algorithm predicts the time when the second aqueous phase layer drains out of the first sight glass.

[0036] S504. The server transmits the time when the second aqueous phase layer drains out of the first sight glass to the central control system, and the central control system automatically adjusts the valve body of the first extraction kettle to collect the second aqueous phase layer.

[0037] Through the layering by the picture prediction algorithm twice, there are no impurities in the obtained second aqueous phase layer.

[0038] After the reaction of the materials in the extraction kettle, when the stratified layer after the reaction is allowed to stand for a period of time until an obvious stratification is formed, the top layer is the oil phase, which contains catalyst activated carbon, butyl acetate, etc. The second layer is the impurity layer, the third layer is the emulsion layer, and the fourth layer is the aqueous phase layer. In the above method, the impurity layer and the oil layer are removed. After the materials are allowed to stand, an aqueous phase layer, an emulsion layer, and an oil phase layer will appear. The standing needs to be operated with timing and temperature control, including ensuring that the standing is carried out within the specified time range and precisely controlling the temperature during this process. The system will monitor and adjust the temperature to ensure stable conditions throughout the standing stage.

[0039] The hierarchical detection algorithm is written using the Torch programming language, the OpenCV computer vision library, and the Numpy library. Data preprocessing is carried out through the Pandas library. In the hierarchical detection algorithm, Torch can be used to train and optimize neural network models, process a large amount of image data, perform deep learning, and extract features. By using Torch, researchers can easily implement various complex neural network structures and improve the performance of the hierarchical detection algorithm. OpenCV is an open-source computer vision library that contains a large number of image processing and computer vision algorithms. In the hierarchical detection algorithm, OpenCV can be used for tasks such as image preprocessing, feature extraction, and object detection. For example, operations such as image grayscaling, edge detection, and feature point extraction can be performed using OpenCV to extract key information from the image and provide support for subsequent hierarchical detection. Numpy is a numerical computing extension library for Python that supports a large number of dimensional array and matrix operations. In addition, it provides a large number of mathematical function libraries for array operations. In the hierarchical detection algorithm, Numpy can be used to process and analyze a large amount of data, including image data and feature data.

[0040] By performing HSV transformation (color space conversion) on the collected pictures, and then locating the position of the required interface through gray histogram statistics based on gray statistics. The HS transformation facilitates color extraction and classification: in the HSV color space, the hue H represents the basic attribute of the color and can be used for color extraction and classification. By converting the image from the RGB color space to the HSV color space, it is easier to identify and extract regions of specific hues, and classify and process them. The HSV transformation is suitable for illumination invariance processing: since the value V represents the brightness of the color and is not affected by illumination changes, when processing images with unstable illumination, the HSV transformation can better retain color information and perform corresponding processing. The HSV transformation simplifies image processing algorithms: in the HSV color space, color information is separated into three components: hue, saturation, and value, which makes the algorithm simpler and more intuitive. When performing image processing, different components can be processed separately to achieve more efficient and flexible image processing.

[0041] In one embodiment, the extraction kettle layering method further includes:

[0042] S601, collecting the first aqueous phase layer or the first aqueous phase layer and the second aqueous phase layer into a second extraction kettle, stirring the materials in the second extraction kettle, waiting for the materials to complete the reaction, and standing for a period of time until complete layering occurs. A third aqueous phase layer, a third emulsion layer, and a third oil phase layer with obvious layering interfaces are formed in the second extraction kettle.

[0043] S602, drain the third aqueous phase layer from the bottom of the second extraction kettle. The material flows through the second sight glass, and pictures in the second sight glass are continuously collected by the camera;

[0044] S603, the server analyzes the pictures in the second sight glass through a layering detection algorithm. When a layering interface appears in the pictures in the second sight glass, the layering detection algorithm obtains the proportion of each layer of liquid material;

[0045] S604, the server transmits the proportion of each layer of liquid material to the central control system, and the central control system automatically adjusts the valve body of the second extraction kettle to collect the third aqueous phase layer.

[0046] In one embodiment, the extraction kettle layering method further includes:

[0047] S701, discharge the first impurity layer and the first oil phase layer in the first extraction kettle, then collect the third aqueous phase layer into the first extraction kettle, and let the third aqueous phase layer stand in the first extraction kettle for a period of time. A fourth aqueous phase layer, a fourth emulsion layer, and a fourth oil phase layer with obvious layering interfaces are formed in the first extraction kettle;

[0048] S702, drain the fourth aqueous phase layer from the bottom of the first extraction kettle. The fourth aqueous phase layer flows through the first sight glass, and pictures in the first sight glass are continuously collected;

[0049] S703, the server analyzes the pictures in the first sight glass through a layering detection algorithm. When a layering interface appears in the pictures in the first sight glass, the layering detection algorithm obtains the proportion of each layer of liquid material;

[0050] S704, the server transmits the proportion of each layer of liquid material to the central control system, and the central control system automatically adjusts the valve body of the first extraction kettle to collect the fourth aqueous phase layer.

[0051] After multiple layering through the picture prediction algorithm and the layering detection algorithm, the content of impurities and oil phase in the obtained fourth aqueous phase layer reaches the lowest, ensuring the purity of the fourth aqueous phase layer to the greatest extent.

[0052] In one embodiment, the central control system includes a DCS-700 system. The DCS-700 adopts advanced control algorithms and data processing technologies, featuring high precision, high stability, and high reliability. It can achieve efficient control and data processing. The DCS-700 has flexible configuration and configuration functions and can be customized and expanded according to actual needs to meet the control requirements of different fields and industries. The DCS-700 has a simple and easy-to-use user interface and operation method, facilitating users to monitor, operate, and manage, and reducing the difficulty of use and maintenance. The DCS-700 has high reliability and stability. By adopting redundant technology and fault diagnosis functions, it can ensure long-term fault-free operation, reducing the failure rate and maintenance cost. The method uses the structured text language and function block diagram programming language in the DCS-700 system, and the input signals cover analog input, analog output, digital input, and digital output. The entire process flow method is written using the ST language and FBD function blocks in the central control DCS-700 system. The input signals cover AI (analog input), AO (analog output), DI (digital input), and DO (digital output). For convenient management, the parameters of each instrument are imported into the tag table of the central control DCS-700 with one key in accordance with the specified format. Ensure that the system can accurately identify and process the input and output signals of each instrument, thereby achieving reliable control of the entire process flow. All parameters of the entire process flow method need to be manually input. When the central control DCS-700 system receives these parameters, it will automatically start the entire process flow.

[0053] In one embodiment, the valve body includes an electric control valve and a cut-off valve. The central control system automatically adjusts the electric control valve and the cut-off valve on the extraction kettle, thereby realizing the discharging of different products.

[0054] In one embodiment, the outflow of materials in the first extraction kettle and the second extraction kettle is achieved by nitrogen-pressurized materials. The mutual flow of materials between the first extraction kettle and the second extraction kettle is through nitrogen-pressurized materials. Then, the layering of different material layers is realized by the opening and closing of the valve body. At the same time, in order to prevent the pressure inside the extraction kettle from exceeding the safe range, a pressure interlock mechanism is adopted.

[0055] In one embodiment, the server is connected through a network. The server communicates with the Modbus (RTU transmission model) protocol gateway, transmitting preset time signals and liquid level ratio signals to the central control system. The central control system is responsible for converting the preset time signals and liquid level ratio signals into the opening signals of the valve body and gradually reducing the opening of the valve body.

[0056] In one embodiment, the pictures in the first sight glass and the pictures in the second sight glass are collected by a high-speed camera. The high-speed camera has a very high shooting frequency, which can reach thousands or even tens of thousands of frames per second. The high-speed camera will generate a large amount of images and data in a short period of time. The high-speed camera has capabilities such as high image stability, high transmission, and high anti-interference.

[0057] Embodiment 1

[0058] A method for separating layers in an extraction kettle is provided. This method undergoes four separation processes to minimize the impurities and oil phase content in the aqueous layer and ensure the purity of the aqueous layer to the greatest extent.

[0059] As Figure 2 shown, the method for separating layers in the extraction kettle includes the following steps:

[0060] S101, After adding the material to the extraction kettle and allowing it to react fully, without standing still, at this time, there is no obvious separation interface in the material in the extraction kettle, and the first separation and the second separation are started.

[0061] S102, Analyze the pictures in the sight glass collected through the picture prediction algorithm, and predict the preset time amount of the discharge of the aqueous layer starting from this time according to the detection of black substances in the pictures.

[0062] S103, Transmit the preset time amount to the central control system through the server, and the central control system automatically adjusts the opening degree of the valve body of the extraction kettle to obtain the aqueous layer.

[0063] S104, Allow the obtained aqueous layer to stand still fully. At this time, there is an obvious separation interface in the material in the extraction kettle, and the third separation and the fourth separation are started.

[0064] S105, Analyze the pictures in the sight glass collected through the separation detection algorithm. When a separation interface appears in the pictures in the sight glass, the separation detection algorithm obtains the proportion of each layer of liquid material.

[0065] S106, Transmit the proportion of each layer of liquid material to the central control system through the server, and the central control system automatically adjusts the opening degree of the valve body of the extraction kettle to obtain the aqueous layer.

[0066] As Figure 3 shown, specifically, the embodiment provides two extraction kettles, namely the first extraction kettle V1 and the second extraction kettle V2. Four separation processes are implemented in the first extraction kettle V1 and the second extraction kettle V2. The detailed description of the four separation and discharging methods includes the following steps:

[0067] The first separation

[0068] First, put the material into the first extraction kettle V1, stir and stand still the material in the first extraction kettle V1, and wait for the material reaction to complete;

[0069] Turn off the stirring in the first extraction kettle V1, open the cut-off valve XV3, close the tail gas recovery valve XV1, and at the same time open the medium-pressure nitrogen valve XV2, and wait for the pressure in the kettle to rise to 0.1 MPa.

[0070] Open the cut-off valves XV4 and XV6, and at the same time adjust the opening of the electromagnetic control valve CV1 to 50%. The reacted material is discharged from the bottom of the first extraction kettle V1, and the aqueous layer in the material flows into the first sight glass AD1 and then flows into the second extraction kettle V2 through the first sight glass AD1.

[0071] The high-speed camera continuously captures the pictures in the first sight glass AD1 and transmits them to the server once every 10 frames of the video sequence. The server analyzes the pictures through the picture prediction algorithm.

[0072] If no black substance is detected, continue to repeat the operation.

[0073] When a black substance appears in the picture, the system performs anomaly detection and triggers an alarm output. The picture prediction algorithm predicts the time when all the aqueous layer in the first sight glass AD1 is drained out of the first sight glass AD1.

[0074] The server transmits the time when the aqueous layer is drained out of the first sight glass AD1 to the central control system. The central control system converts the time into an analog quantity and controls the opening of the electromagnetic control valve CV1. The initial opening of the electromagnetic control valve CV1 is 50%, and the opening of the electromagnetic control valve CV1 is reduced proportionally according to the time until it is closed.

[0075] All the aqueous layer enters the second extraction kettle V2, and the impurity layer and the oil layer remain in the first extraction kettle V1 and the first sight glass AD1.

[0076] Second layering

[0077] Blow the material in the first sight glass AD1 back into the first extraction kettle V1, add clear water to the first extraction kettle V1, stir the material in the first extraction kettle V1, let it stand, and wait for the material reaction to be completed;

[0078] Turn off the stirring in the first extraction kettle V1, open the cut-off valve XV3, close the tail gas recovery valve XV1, and at the same time open the medium-pressure nitrogen valve XV2, and wait for the pressure in the kettle to rise to 0.1 MPa.

[0079] Open the cut-off valves XV4 and XV6, and at the same time adjust the opening of the electromagnetic control valve CV1 to 50%. The reacted material is discharged from the bottom of the first extraction kettle V1, and the aqueous layer in the material flows into the first sight glass AD1 and then flows into the second extraction kettle V2 through the first sight glass AD1.

[0080] The high-speed camera continuously captures the images in the first sight glass AD1 and transmits them to the server once every 10 frames of the video sequence. The server analyzes the images through an image prediction algorithm.

[0081] If no black substance is detected, continue to repeat the operation.

[0082] When a black substance appears in the image, the system performs anomaly detection and triggers an alarm output. The time when all the aqueous layer in the first sight glass AD1 is drained out as predicted by the image prediction algorithm.

[0083] The server transmits the time when the aqueous layer is drained out of the first sight glass AD1 to the central control system. The central control system converts the time into an analog quantity and controls the opening of the electromagnetic regulating valve CV1. The initial opening of the electromagnetic regulating valve CV1 is 50%, and the opening of the electromagnetic regulating valve CV1 is reduced proportionally according to the time until it is closed.

[0084] All the aqueous layer enters the second extraction kettle V2, then close the cut-off valve XV6, open the cut-off valves XV4 and XV7, drain the oil layer and the impurity layer into the waste collection tank, and close the cut-off valves XV4 and XV7.

[0085] The third layer separation

[0086] Stir, stand still the materials in the second extraction kettle V2, and wait for the materials to complete the reaction, and stand still for a period of time until complete layer separation.

[0087] Turn off the stirring in the second extraction kettle V2, open the cut-off valve XV10, make the materials flow into the second sight glass AD2, and the materials stand still in the second sight glass AD2 for 15 - 20 minutes.

[0088] After waiting for the materials to stand still sufficiently, the layer separation reaction is completed. At this time, the second sight glass AD2 is all filled with the lower aqueous layer.

[0089] Close the tail gas recovery valve XV8, and at the same time open the medium-pressure nitrogen valve XV9. Wait until the pressure in the kettle rises to 0.1 MPa. Start the pressure interlock, and keep the pressure in the extraction kettle around 0.1 MPa.

[0090] Open the cut-off valves XV11 and XV13, and at the same time adjust the opening of the electromagnetic regulating valve CV2 to 100%, and start discharging materials to the water layer receiving device, the first extraction kettle V1.

[0091] The high-speed camera continuously captures the images in the second sight glass AD2. Transmit them to the server once every 10 frames of the video sequence. The server analyzes the images in the second sight glass AD2 through a layer separation detection algorithm.

[0092] The layered detection algorithm performs HSV transformation on the collected images, conducts histogram equalization to increase image contrast, reduces the influence of light changes on edge detection, and then uses an edge detection algorithm to locate the position of the required interface.

[0093] If the layered interface is not detected, continue to repeat the operation to ensure continuous effective image processing before finding the layered interface.

[0094] When a layered interface appears in the image, the system performs anomaly detection and triggers an alarm output. The system will switch between global or local search and output the position of the interface.

[0095] The server transmits the signals of the proportion of each liquid to the central control DCS-700 system. The central control DCS-700 system converts the received signals into the control signal of the opening degree CV2 of the electromagnetic regulating valve.

[0096] When the water layer gradually decreases and an emulsion layer appears in the second sight glass AD2, adjust the opening degree of the electromagnetic regulating valve CV2 to 0%, and then slowly adjust it to 10%, and close it again after 7s.

[0097] Repeat the above operation until an interface of an oil layer, an emulsion layer, and an aqueous phase layer from top to bottom appears in the second sight glass AD2, and slowly reduce the opening degree of the electromagnetic regulating valve CV2 as the height of the water layer decreases.

[0098] Until the height of the aqueous phase layer approaches 0, close the electromagnetic regulating valve CV2.

[0099] All of the aqueous phase layer enters the first extraction kettle V1, close the cut-off valve XV10, open the tail gas recovery valve XV8, and at the same time close the medium-pressure nitrogen valve XV9. Open the cut-off valve XV11 and the cut-off valve XV14 to drain the oil layer and the impurity layer into the waste collection tank, and then close the cut-off valve XV11 and the cut-off valve XV14.

[0100] The fourth layering

[0101] Stir the materials in the first extraction kettle V1, let them stand, and wait for the materials to complete the reaction, and let them stand for a period of time until complete layering.

[0102] Turn off the stirring in the first extraction kettle V1, open the cut-off valve XV3 to make the materials flow into the first sight glass AD1, and the materials stand in the first sight glass AD1 for 15 - 20 minutes.

[0103] After waiting for the materials to stand fully, the layering reaction is completed. At this time, the inside of the first sight glass AD1 is all the lower aqueous phase layer.

[0104] Close the tail gas recovery valve XV1 and open the medium-pressure nitrogen valve XV2 simultaneously. Wait until the pressure in the kettle rises to 0.1 MPa. Start the pressure interlock, and keep the pressure in the first extraction kettle V1 around 0.1 MPa.

[0105] Open the cut-off valves XV4 and XV6, and adjust the opening of the electromagnetic control valve CV1 to 100% simultaneously. Start discharging to the water layer receiving device, the second extraction kettle V2.

[0106] The high-speed camera continuously captures the pictures in the first sight glass AD1. Transmit the pictures to the server once every 10 frames of the video sequence. The server analyzes the pictures in the first sight glass AD1 through the hierarchical detection algorithm.

[0107] The hierarchical detection algorithm performs HSV transformation on the captured pictures, conducts histogram equalization to increase the image contrast, reduces the influence of light changes on edge detection, and then uses the edge detection algorithm to locate the position of the required interface.

[0108] If no stratification interface is detected, continue to repeat the operation. To ensure that effective image processing can be continuously carried out before finding the stratification interface.

[0109] When a stratification interface appears in the picture, the system performs anomaly detection and triggers an alarm output. The system will switch between global or local search and output the position of the interface.

[0110] The server transmits the proportion signals of each liquid to the central control DCS-700 system. The central control DCS-700 system converts the received signals into control signals for the opening of the electromagnetic control valve CV1.

[0111] When the water layer gradually decreases and an emulsion layer appears in the first sight glass AD1, adjust the opening of the electromagnetic control valve CV1 to 0%, then slowly adjust it to 10%, and close it again after 7 s.

[0112] Repeat the above operation until an interface of the oil layer, emulsion layer, and water phase layer from top to bottom appears in the first sight glass AD1, and slowly reduce the opening of the electromagnetic control valve CV1 as the height of the water layer decreases.

[0113] Until the height of the water phase layer is close to 0, close the electromagnetic control valve CV1.

[0114] All the water phase layer enters the second extraction kettle V2. Close the cut-off valve XV6, open the cut-off valves XV4 and XV7, discharge the oil layer and impurity layer to the waste collection tank, and then close the cut-off valves XV4 and XV7.

[0115] All four stratifications are completed, and the impurities and oil phase content in the obtained water phase layer reach the lowest, ensuring the purity of the water phase layer to the greatest extent.

[0116] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A layering method for an extraction kettle, characterized in that: include: S1, stirring the material in the first extraction kettle, waiting for the material reaction to be completed, and forming a first water phase layer, a first impurity layer and a first oil phase layer with fuzzy stratification interfaces in the first extraction kettle; S2, releasing the first aqueous phase layer from the bottom of the first extraction kettle, allowing the first aqueous phase layer to flow through a first viewing mirror, and continuously collecting images in the first viewing mirror; S3, the server analyzes the image in the first viewing mirror by using an image prediction algorithm. When impurities appear in the image in the first viewing mirror, the image prediction algorithm predicts the time when the first aqueous phase is discharged from the first viewing mirror; S4, the server transmits the time when the first aqueous phase layer is discharged from the first viewing mirror to the central control system, and the central control system automatically adjusts the valve body of the first extraction kettle to collect the first aqueous phase layer.

2. The layering method of the extraction kettle according to claim 1, characterized in that: Also includes: S501, back-blowing the impurity layer and the oil phase layer in the first sight glass into the first extraction kettle, adding clean water into the first extraction kettle, stirring the materials in the first extraction kettle, waiting for the materials to react, and forming a second water phase layer, a second impurity layer, and a second oil phase layer with fuzzy stratification interfaces in the first extraction kettle; S502, releasing the second aqueous phase layer from the bottom of the first extraction kettle, allowing the second aqueous phase layer to flow through a first viewing mirror, and continuously collecting images in the first viewing mirror; S503, the server analyzes the image in the first viewing mirror by using an image prediction algorithm, and when impurities appear in the image in the first viewing mirror, the image prediction algorithm predicts the time when the second aqueous phase is discharged from the first viewing mirror; S504, the server transmits the time when the second aqueous phase layer is discharged from the first sight glass to the central control system, and the central control system automatically adjusts the valve body of the first extraction kettle to collect the second aqueous phase layer.

3. The layering method of the extraction kettle according to claim 1 or 2, characterized in that: Also includes: S601, collecting the first aqueous phase layer or the first aqueous phase layer and the second aqueous phase layer into a second extraction kettle, stirring the material in the second extraction kettle, waiting for the material reaction to be completed, and standing for a period of time until the material is completely separated, and a third aqueous phase layer, a third emulsified layer, and a third oil phase layer with obvious separation interfaces are formed in the second extraction kettle; S602, releasing the third aqueous phase from the bottom of the second extraction kettle, allowing the material to flow through a second viewing mirror, and continuously collecting images in the second viewing mirror; S603, the server analyzes the image in the second mirror by using a layered detection algorithm. When a layered interface appears in the image in the second mirror, the layered detection algorithm obtains the proportion of each layer of material and liquid; S604, the server transmits the proportion of each layer of material and liquid to the central control system, and the central control system automatically adjusts the valve body of the second extraction kettle to collect the third aqueous phase.

4. The layering method of the extraction kettle according to claim 3, characterized in that: Also includes: S701, discharging the first impurity layer and the first oil phase layer in the first extraction kettle, collecting the third water phase layer into the first extraction kettle, and leaving the third water phase layer in the first extraction kettle for a period of time, so that a fourth water phase layer, a fourth emulsified layer, and a fourth oil phase layer with obvious stratification interfaces are formed in the first extraction kettle; S702: releasing the fourth aqueous phase from the bottom of the first extraction kettle, the fourth aqueous phase flowing through the first viewing mirror, and continuously collecting images in the first viewing mirror; S703, the server analyzes the image in the first mirror by using a layered detection algorithm. When a layered interface appears in the image in the first mirror, the layered detection algorithm obtains the proportion of each layer of material and liquid; S704, the server transmits the proportion of each layer of material and liquid to the central control system, and the central control system automatically adjusts the valve body of the first extraction kettle to collect the fourth aqueous phase.

5. The layering method of the extraction kettle according to claims 1 to 4, characterized in that: The central control system includes a DCS-700 system, the method adopts a structured text language and a function block diagram programming language in the DCS-700 system, and the input signals include analog input, analog output, digital input and digital output.

6. The layering method of the extraction kettle according to claims 1 to 4, characterized in that: The valve body includes an electric regulating valve and a cut-off valve.

7. The layering method of the extraction kettle according to claims 1 to 4, characterized in that: The outflow of materials in the first extraction kettle and the second extraction kettle is achieved by punching the materials with nitrogen.

8. The layering method of the extraction kettle according to claims 1 to 4, characterized in that: The server is connected via a network, and the server communicates with the RTU transmission model protocol gateway to transmit the liquid level proportion signal to the central control system. The central control system converts the preset time signal and the liquid level proportion signal into an opening signal of the electric control valve, and gradually reduces the opening of the control valve.

9. The layering method of the extraction kettle according to claims 1 to 4, characterized in that: The image in the first mirror and the image in the second mirror are collected by a high-speed camera.

Images