Recovery treatment method and system based on chemical product production waste gas

Abstract

The invention provides a recovery treatment method and system based on chemical product production waste gas, and relates to the technical field of waste gas treatment. Time sequence identification is carried out, a time sequence exhaust gas parameter table is determined, and exhaust gas time sequence nodes and exhaust gas types, exhaust gas concentrations and exhaust gas contents corresponding to the nodes are stored in the time sequence exhaust gas parameter table according to the time sequence; performing stage division, and outputting a multi-stage waste gas flow; the multi-branch waste gas recovery system is connected with a multi-branch waste gas recovery device, comprises a plurality of recovery branches, and inputs a multi-stage waste gas flow into a multi-branch mapping encoder to obtain a recovery mapping code; and when chemical products are produced, valves of the multiple recovery branches are controlled to be started or closed according to the recovery mapping coding time sequence. The technical problems that in the prior art, most chemical waste gas is subjected to unified recovery and centralized treatment, equipment cannot adjust the working mode according to the characteristics of different waste gas during treatment, the waste gas treatment efficiency is low, and the equipment burden is too heavy are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of waste gas treatment, and particularly to a method and system for recycling and treating waste gas generated in the production of chemical products. Background Art

[0002] During the production process of chemical products, the generation of waste gas is an inevitable by-product. Different chemical reactions produce different types of waste gas, which may have adverse effects on the environment, human health, and the equipment itself. Therefore, timely and effective waste gas recycling and treatment have become an important link in the production process. However, the existing chemical waste gas recycling methods usually gather the waste gas together for unified treatment, which leads to interference between different waste gas components and may affect the recycling efficiency. For example, when the waste gas contains different gases, some gases may be over-recycled while the recycling efficiency of other gases is low, resulting in waste of resources. Moreover, the production process of chemical products is usually carried out in stages, and the generation of waste gas also occurs in stages. The waste gas generated in different stages may have different characteristics such as composition, concentration, and flow rate. The existing waste gas recycling usually does not optimize the treatment method according to the different stages of waste gas, resulting in a mismatch in timing during the treatment process, which in turn leads to unnecessary energy waste and a decrease in treatment efficiency. Summary of the Invention

[0003] The present application provides a method and system for recycling and treating waste gas generated in the production of chemical products, aiming to solve the technical problems that in the existing chemical production process, most chemical waste gases are recycled and centrally treated in a unified manner, and the equipment cannot adjust its working mode according to the characteristics of different waste gases, resulting in low waste gas treatment efficiency and excessive equipment burden.

[0004] In the first aspect disclosed in the present application, a method for recycling and treating waste gas generated in the production of chemical products is provided. The method includes: collecting waste gas process record data during the production of chemical products; performing time series identification on the waste gas process record data to determine a time series waste gas parameter table, where the time series waste gas parameter table stores waste gas time series nodes and the corresponding waste gas types, waste gas concentrations, and waste gas contents in chronological order; performing stage division according to the time series waste gas parameter table to output a multi-stage waste gas process; connecting a multi-branch waste gas recycling device, where the multi-branch waste gas recycling device includes multiple recycling branches, inputting the multi-stage waste gas process into a multi-branch mapping encoder, establishing a mapping code between the multi-stage waste gas process and the multiple recycling branches to obtain a recycling mapping code; when producing the chemical product, the multi-branch waste gas recycling device controls the valves of the multiple recycling branches to start or close according to the recycling mapping code in chronological order.

[0005] The second aspect disclosed in this application provides a recovery and treatment system for waste gas from chemical product production. The system is used for the above-mentioned recovery and treatment method of waste gas from chemical product production, and the system includes: a data acquisition module for acquiring waste gas process record data during the production of chemical products; a timing identification module for performing timing identification on the waste gas process record data to determine a timing waste gas parameter table, which stores waste gas timing nodes and the corresponding waste gas types, waste gas concentrations, and waste gas contents at each node in chronological order; a stage division module for performing stage division according to the timing waste gas parameter table and outputting a multi-stage waste gas process; a mapping coding module for connecting a multi-branch waste gas recovery device, where the multi-branch waste gas recovery device includes multiple recovery branches, inputting the multi-stage waste gas process into a multi-branch mapping encoder, establishing a mapping coding between the multi-stage waste gas process and the multiple recovery branches, and obtaining a recovery mapping coding; a valve control module for, when producing the chemical product, the multi-branch waste gas recovery device controlling the valves of the multiple recovery branches to start or close according to the recovery mapping coding in chronological order.

[0006] One or more technical solutions provided in this application have at least the following beneficial effects: By collecting real-time data on waste gas emissions during chemical production, accurate waste gas generation conditions can be obtained, ensuring that waste gas recovery and treatment can be controlled based on accurate and complete data, and enhancing the reliability of subsequent steps; through timing identification, waste gas process data is arranged and organized in chronological order into a timing waste gas parameter table, enabling each waste gas node to be clearly marked on the time axis and ensuring that the generation laws and characteristics of waste gas can be clearly and accurately reflected; by performing stage division according to the timing waste gas parameter table, the waste gas process is divided into multiple different stages, allowing more targeted treatment methods to be adopted in each stage, avoiding the limitations of a single treatment method. The multi-stage waste gas process enables the system to flexibly respond to changes in different waste gas types during the production process and adapt to the diverse needs in chemical product production; by connecting a multi-branch waste gas recovery device and using a multi-branch mapping encoder to map the waste gas process and recovery branches, the appropriate recovery branch can be automatically selected for waste gas recovery according to the requirements of waste gas in different stages, avoiding resource waste or improper configuration, thereby improving the recovery efficiency and reducing unnecessary energy consumption; by controlling the start and stop of the valves of the recovery branches according to the recovery mapping coding in chronological order, precise scheduling of the waste gas treatment process is achieved, and according to the recovery mapping coding, the waste gas recovery device can automatically adjust the working states of multiple recovery branches, reducing the need for manual intervention and making the waste gas recovery process more automated and intelligent.

[0007] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of this application more obvious and understandable, the following specific embodiments of this application are specifically given. Description of the Drawings

[0008] Figure 1 It is a schematic flow chart of the recovery and treatment method for waste gas generated in the production of chemical products provided by an embodiment of this application.

[0009] Figure 2 It is the change of the exemplary waste gas concentration with time sequence nodes in the recovery and treatment method for waste gas generated in the production of chemical products provided by an embodiment of this application.

[0010] Figure 3 It is a schematic structural diagram of the recovery and treatment system for waste gas generated in the production of chemical products provided by an embodiment of this application.

[0011] Description of the reference numerals: Data acquisition module 10, time sequence identification module 20, stage division module 30, mapping coding module 40, valve control module 50. Detailed Description of the Embodiments

[0012] By providing a recovery and treatment method and system for waste gas generated in the production of chemical products, an embodiment of this application solves the technical problems in the prior art that in the chemical production process, most chemical waste gases are recovered and treated in a unified and centralized manner, and the equipment cannot adjust the working mode according to the characteristics of different waste gases during treatment, resulting in low waste gas treatment efficiency and excessive equipment burden.

[0013] After introducing the basic principle of this application, the following will specifically introduce various non-limiting embodiments of this application in conjunction with the drawings of the specification. It should be understood that the specific embodiments described herein are only used to explain this application and are not used to limit this application.

[0014] Embodiment 1, as Figure 1 shown, an embodiment of this application provides a recovery and treatment method for waste gas generated in the production of chemical products, and the method includes: Collect the waste gas process record data during the production of chemical products.

[0015] During the chemical production process, different parameters of the waste gas are continuously monitored through multiple waste gas sensors or monitoring devices, including waste gas types (such as organic gases, inorganic gases, etc.), concentration, flow rate, etc. These monitoring devices are usually installed at various key nodes of the waste gas emission source or pipeline to continuously record the waste gas emission situation. The collected waste gas process record data can provide detailed information about the generation and emission of the waste gas.

[0016] Perform temporal identification on the recorded data of the exhaust gas process to determine a temporal exhaust gas parameter table, which stores exhaust gas temporal nodes and the corresponding exhaust gas types, exhaust gas concentrations, and exhaust gas contents for each node in temporal order.

[0017] Perform temporal identification on the recorded data of the exhaust gas process. The purpose of temporal identification is to find the temporal pattern of exhaust gas generation. The exhaust gas emissions in chemical production are not continuously and evenly released, but are associated with different stages of the production process. For example, a certain type of exhaust gas is emitted at the start of the reaction, and after a certain period of time, another type of exhaust gas is generated, and then it enters different stages of exhaust gas components and concentrations. Use data mining techniques, such as time series analysis, clustering analysis, etc., to identify the time series information in the data and distinguish each temporal node, that is, the key time points of exhaust gas generation.

[0018] Based on the temporal nodes, construct a temporal exhaust gas parameter table. This table records each temporal node in chronological order and assigns the corresponding exhaust gas type, concentration, and content to each node. For example, assume that a certain type of exhaust gas is generated within 10 minutes after the start of the reaction. Then record the temporal node of 10 minutes and clearly label the type, concentration, and content of the exhaust gas on this node. An exemplary temporal exhaust gas parameter table is shown in Table 1. As Figure 2 shown, it is the change of the exemplary exhaust gas concentration with the temporal node.

[0019] Table 1: Temporal Exhaust Gas Parameter Table Perform stage division according to the temporal exhaust gas parameter table and output a multi-stage exhaust gas process.

[0020] Arrange the various nodes in the temporal exhaust gas parameter table in chronological order and construct a feature vector for each temporal node. The feature vector includes the concentration, flow rate, gas type of the exhaust gas, and the change trend of these parameters over time. For example, at a certain moment, the exhaust gas concentration is a certain value, and over time, the concentration may change, forming a concentration change rate. Incorporate this information into the feature vector.

[0021] To reasonably divide different stages of the exhaust gas process, use the sliding window technique to calculate the similarity between temporal feature vectors. By calculating the similarity of the feature vectors between adjacent time nodes, determine the time points with significant changes in similarity as candidate variant temporal nodes. These points are important moments of change in the exhaust gas generation process and usually indicate the start of a new stage.

[0022] The K - value clustering algorithm is used to cluster the candidate mutation timing nodes to determine the boundaries of different stages. The K - value clustering algorithm will divide the candidate nodes into different stages based on a cost function, such as minimizing the within - group distance and maximizing the between - group distance. The optimal K - value solution finally output is the optimal stage division result. According to the result of K - value clustering, the timing exhaust gas parameter table is divided into multiple stages, and a multi - stage exhaust gas process is output. Each stage corresponds to certain exhaust gas characteristics and treatment requirements. These stage divisions provide a basis for the subsequent control of the exhaust gas recovery branch, so that different exhaust gas flows and compositions can be guided to appropriate recovery devices for treatment.

[0023] Connect the multi - branch exhaust gas recovery device. The multi - branch exhaust gas recovery device includes multiple recovery branches. Input the multi - stage exhaust gas process into the multi - branch mapping encoder to establish the mapping coding between the multi - stage exhaust gas process and the multiple recovery branches, and obtain the recovery mapping coding.

[0024] Connect the multi - branch exhaust gas recovery device. The multi - branch exhaust gas recovery device contains multiple recovery branches, and each recovery branch has different treatment capabilities. For example, some recovery branches are suitable for treating a single exhaust gas component, while others are suitable for treating mixed gases. Each recovery branch may also have different treatment efficiencies, exhaust gas flow treatment capabilities, exhaust gas concentration adaptabilities, etc.

[0025] Input the multi - stage exhaust gas process into the multi - branch mapping encoder. The function of this encoder is to select a suitable recovery branch for each stage according to the characteristics of the exhaust gas in each stage. That is to say, the encoder will map the exhaust gas process of each stage to one or more recovery branches through rule matching.

[0026] In the multi - branch mapping encoder, the relationship between the exhaust gas process of each stage and the multiple recovery branches is represented by the recovery mapping coding. The recovery mapping coding contains multiple pieces of information, such as the type and concentration of the exhaust gas in each stage and the corresponding recovery branches. This coding provides a clear control diagram, telling the system which recovery branches should be started in each timing stage and how to regulate the valves of these branches.

[0027] When producing the chemical product, the multi - branch exhaust gas recovery device controls the valves of the multiple recovery branches to start or close according to the recovery mapping coding in time sequence.

[0028] During the production process of chemical products, the types, concentrations, and emission amounts of waste gases will vary with different stages of the production process. Therefore, it is necessary to precisely control the opening and closing of the recovery branches according to the characteristics of the waste gases at each stage. In this step, the recovery mapping code provides an indication of the recovery branches required for each waste gas stage, and the recovery device adjusts the valves of the recovery branches in real time according to this code. For example, in a certain stage, a single recovery branch needs to be opened to treat a specific type of waste gas, while in another stage, multiple recovery branches need to be opened to treat mixed waste gases. The control system accurately controls which recovery branches start or close at what time according to the timing information in the recovery mapping code. For example, after a certain waste gas stage starts, the valves of the recovery branches will be opened at a specified time point to start recovering the waste gas generated in this stage; at the same time, the opening and closing states of the recovery branches are dynamically adjusted. If the waste gas volume or composition changes, the valves will also be adjusted according to the new recovery mapping code.

[0029] This control method is implemented using an automated system (such as a PLC system, a DCS system, etc.). The system reads the recovery mapping code in real time and adjusts the states of the recovery branches stage by stage according to the timing nodes of the waste gas process to ensure the efficiency and accuracy of waste gas recovery and treatment.

[0030] Furthermore, the multiple recovery branches at least include one single-component recovery branch and at least one mixed-component recovery branch; the single-component recovery branch is used for recovering and treating a single waste gas type, and the mixed-component recovery branch is used for recovering and treating a mixed waste gas type.

[0031] During the waste gas recovery process, according to the types and compositions of the waste gases, the recovery branches are designed as single-component recovery branches dedicated to treating a single waste gas component and mixed-component recovery branches dedicated to treating waste gases containing multiple components.

[0032] Among them, the single-component recovery branch is specifically used to treat a certain specific type of waste gas. For example, the gas generated during a certain chemical reaction process, if its composition is fixed and single, can be recovered through the single-component recovery branch. This branch is targeted and can efficiently treat and recover a single waste gas component, such as carbon dioxide, nitrogen, ammonia, etc.

[0033] Different from the single-component recovery branch, the mixed-component recovery branch is used to process mixed waste gas, that is, waste gas containing multiple different components. Such mixed waste gas is generated in many chemical production processes and contains different harmful gases. The mixed-component recovery branch has relatively complex processing capabilities to adapt to the recovery and treatment of various waste gas components. In actual operation, the selection of the waste gas recovery branch depends on the type and composition of the waste gas and the changes in the waste gas during the production process. For example, at a certain stage, if the waste gas contains only one type of gas (such as ammonia), the single-component recovery branch is activated, while when the waste gas is a mixed gas, the mixed-component recovery branch needs to be activated.

[0034] The setting of these recovery branches ensures the flexibility and efficiency of the waste gas recovery process, and the appropriate recovery branch can be dynamically selected for treatment according to the actual composition of the waste gas and the recovery requirements.

[0035] Furthermore, the multi-branch mapping encoder includes a preset rule library, in which rule tags of the multiple recovery branches are stored. The rule tags include the single / mixed-component processing capabilities of each recovery branch, the types of waste gas suitable for processing, and the branch processing efficiency. When the multi-branch mapping encoder receives the multi-stage waste gas process, the sequential waste gas parameter table of each stage of the waste gas process is matched with the rule tags of each recovery branch in the preset rule library to obtain the recovery branches that pass the matching, and an encoding is established between each stage of the waste gas process and the recovery branches that pass the matching, and the recovery mapping encoding is output.

[0036] The multi-branch mapping encoder includes a preset rule library. Among them, the multi-branch mapping encoder needs to select the most suitable recovery branch to process the waste gas according to the sequential waste gas parameter table in different stages. The preset rule library stores rule tags of multiple recovery branches, which describe the types of waste gas that the recovery branch can adapt to, the processing capabilities, and the recovery efficiency. Among them, the single / mixed-component processing capabilities characterize whether each recovery branch can process a single waste gas component (such as nitrogen, carbon dioxide, etc.) or mixed waste gas (waste gas containing multiple components); the types of waste gas suitable for processing characterize the types of waste gas that each recovery branch is suitable for processing. For example, some recovery branches are suitable for organic gases, while others are suitable for inorganic gases or specific gases; the branch processing efficiency characterizes the processing efficiency of each recovery branch under different waste gas concentrations, flow rates, etc., which determines the effect of the branch when recovering waste gas. The recovery branch with higher processing efficiency can recover the waste gas more quickly. These rule tags provide detailed references for the selection of the recovery branch, and the encoder can perform reasonable matching of the recovery branch based on these rule tags.

[0037] When the multi-branch mapping encoder receives the exhaust gas flow at any stage, it first analyzes the sequential exhaust gas parameter table of this exhaust gas flow. The sequential exhaust gas parameter table contains information such as the exhaust gas type, concentration, and flow rate at this stage. For example, the sequential exhaust gas parameter table contains: at the first stage, the exhaust gas is mainly nitrogen, with a concentration of 100 ppm and a low flow rate; at the second stage, the exhaust gas is a mixed exhaust gas, with a concentration of 500 ppm and a high flow rate.

[0038] The encoder matches these exhaust gas parameters with the recovery branch rule labels in the preset rule library. For example, if the exhaust gas at the first stage is nitrogen, the encoder searches for all recovery branches suitable for treating nitrogen and determines which branches are most suitable for treating this exhaust gas. If the exhaust gas at the second stage is a mixed exhaust gas, the encoder selects the recovery branches suitable for treating the mixed gas and further filters the suitable branches according to the flow rate, concentration, etc. of the exhaust gas.

[0039] According to the result of the label matching, a recovery mapping code is established for each exhaust gas stage. This code represents the exact mapping relationship between the exhaust gas flow and the recovery branches, and it can be determined which branches need to be started at what time point through this code. For example, the recovery mapping code is: at the first stage, use the single-component recovery branch A to treat nitrogen; at the second stage, use the multi-component recovery branch B to treat the mixed exhaust gas.

[0040] Furthermore, the stage division is performed according to the sequential exhaust gas parameter table, and a multi-stage exhaust gas flow is output. The method includes: According to the sequential node arrangement of the sequential exhaust gas parameter table, a sequential feature vector sequence of each sequential node is constructed. The sequential feature vector sequence includes the concentrations of each gas corresponding to the time node and the change rate of the concentration with time; a set of sequential feature vector sequences of each sequential node is obtained; the vector similarity change of the set of sequential feature vector sequences is calculated using a sliding window, and candidate mutation sequential nodes are output. The candidate mutation sequential nodes are output as a set of candidate division points; the set of candidate division points is used as the candidate stage boundary and K-value clustering is performed using a cost function to output the optimal solution of K value, where the K value is the number of stages for the stage division; the sequential exhaust gas parameter table is divided into stages according to the optimal solution of K value, and a multi-stage exhaust gas flow is output.

[0041] In the sequential exhaust gas parameter table, each sequential node contains the specific data of the exhaust gas, including the exhaust gas type, concentration, and time, etc. The sequential nodes are arranged in chronological order, and can be time nodes in units of minutes, seconds, etc. For example, assume that a certain sequential node represents the 10th minute of the production process, and the exhaust gas concentration is 50 ppm, and then list the subsequent sequential nodes in chronological order.

[0042] The feature vector of each time series node includes not only the waste gas concentration, but also the change rate of the concentration with respect to time. The change rate can reflect the speed and trend of the concentration change, and can be obtained by calculating the concentration difference and time difference between adjacent time series nodes. It is used to judge the turning points and changes in the waste gas stage. Each time series node will be represented as a feature vector containing the concentration and the change rate. For example, if the concentration at a certain moment is 50 ppm and the change rate is 2 ppm / min, then the feature vector of this time series node is (50, 2).

[0043] Organize the feature vectors of each obtained time series node into a set of feature vector sequences, and this set represents the overall trend of the waste gas concentration and its change over time.

[0044] The sliding window is a time series analysis method. By defining a window in the set of time series feature vector sequences, calculate the similarity between the feature vectors within the window, and slide the window over the entire sequence to obtain the similarity measure at each position. For example, assume the window size is 3 time series nodes, then the first window contains the first 3 nodes, the second window contains nodes 2 to 4, and so on.

[0045] Within each window, calculate the similarity between the feature vector sequences, and use measurement methods such as cosine similarity, Euclidean distance, or Pearson correlation coefficient to measure the similarity of the feature vectors within the window. If the similarity within the window changes significantly, that is, the vectors within the window change greatly, it indicates that the waste gas process has changed, which may be a sudden change in the waste gas composition, concentration, or flow rate.

[0046] By calculating the similarity change, identify those time series nodes with significant similarity changes, and take these nodes as candidate mutation time series nodes. For example, if the similarity change between two adjacent time series nodes is greater than the set threshold, then this node represents a change in the waste gas process and becomes a candidate mutation time series node. Combine all the identified candidate mutation time series nodes into a set of candidate partition points. These points represent the possible stage change points of the waste gas process, providing possible split points for subsequent stage division and helping to determine different stages of waste gas treatment.

[0047] The set of candidate partition points represents the time nodes at which the waste gas process may change, and is the potential stage boundary in the waste gas process. The cost function is used to measure the quality of the stage division. Its goal is to find the best division by minimizing the within-group similarity and maximizing the between-group difference, that is, to group the time series nodes with similar waste gas characteristics into the same stage and increase the difference between the stages as much as possible. The cost function is based on the following: within-group similarity, that is, the similarity of the feature vectors of the time series nodes within the same stage. If the within-group similarity is too high, the cost is lower; between-group difference, that is, the difference in waste gas characteristics between different stages. If the between-group difference is large, the cost is lower.

[0048] The K - value clustering algorithm is used to cluster the time - series nodes in the candidate partition point set. The purpose is to divide these nodes into K stages. K is the number of clusters, and the K - value is determined by optimizing the cost function. The clustering algorithm adjusts the K - value to select the most appropriate number of stages, making the exhaust gas characteristics within each stage as similar as possible, while maximizing the differences between stages. The K - value is determined based on the optimal value of the cost function. Specifically, when minimizing the cost function, the obtained K - value is the number of the best stage partitions. This K - value defines the start and end time - series nodes for each stage.

[0049] According to the number of stages determined in the K - value optimal solution and the candidate partition point set, the time - series nodes of the exhaust gas process are divided into several stages. Each stage corresponds to a continuous time - series segment, and the exhaust gas characteristics are relatively consistent within this segment. For example, if the K - value is 3 and the candidate partition points are nodes 5, 10, and 15, then the exhaust gas process will be divided into 3 stages. The first stage is from the first node to node 5, the second stage is from node 5 to node 10, and the third stage is from node 10 to node 15. After completing the stage division, a multi - stage exhaust gas process is obtained, providing necessary information for subsequent exhaust gas recovery branch control and stage - by - stage processing.

[0050] Furthermore, a sliding window is used to calculate the vector similarity change of the time - series feature vector sequence set and output candidate mutated time - series nodes. The method includes: Among them, the vector similarity change includes calculating the cosine similarity change rate of the time - series feature vector sequence set; performing a significance test on the cosine similarity change rate, and outputting the time - series nodes with a cosine similarity change rate greater than the preset cosine similarity change threshold as candidate mutated time - series nodes.

[0051] The similarity change between different time nodes is measured by calculating the cosine similarity change rate of the time - series feature vector sequence set. Cosine similarity is a method for measuring the similarity between two vectors. It evaluates their similarity by calculating the cosine value of the angle between two vectors. During the processing of time - series exhaust gas, each time - series node has a feature vector containing concentration, change rate, etc. Cosine similarity is used to calculate the similarity between adjacent time - series nodes.

[0052] The change rate is the change amount of the cosine similarity between adjacent time - series nodes. The cosine similarity change rate reflects the change trend of exhaust gas characteristics over time. If the similarity change rate between adjacent time - series nodes is large, it means that the exhaust gas characteristics between these two nodes are quite different, which may indicate that an important change has occurred in the exhaust gas production process. For example, an increase in the reaction time leads to a change in the exhaust gas composition.

[0053] Significance tests can use statistical methods, such as t-tests, z-tests, etc., to verify whether there are statistically significant differences in the change rate. Another simple method is to set a preset threshold for the change in cosine similarity. The threshold is set according to the actual situation and is used to filter out those time series nodes with insignificant changes. If the similarity change between time series nodes exceeds this threshold, then it is considered that this node represents a significant change in the exhaust gas process, and this time series node is output as a candidate variant time series node. The candidate variant time series node characterizes the potential stage turning points in the exhaust gas process, that is, it represents the transition of the exhaust gas from one stage to another. For example, assume that between the 10th time series node and the 11th time series node in the exhaust gas process, the change rate of the cosine similarity exceeds the preset threshold, then the 11th time series node will be marked as a candidate variant time series node and used as a reference point for stage division.

[0054] Furthermore, record the start and end time points corresponding to the multi-stage exhaust gas process. The multi-branch exhaust gas recovery device controls the valves of the multiple recovery branches to start or close according to the recovery mapping coding time series and the start and end time points.

[0055] Record the start and end time points corresponding to the multi-stage exhaust gas process. The start and end time points represent the start and end moments of each stage in the exhaust gas process. These time points are determined through clustering analysis or time series feature vector changes in the previous steps. Usually, the start time and end time of each stage are clearly recorded. For example, assume that after stage division, the exhaust gas process is divided into 3 stages: the first stage starts from time point t 1 and ends at t 2 ; the second stage starts from time point t 2 and ends at t 3 ; the third stage starts from time point t 3 and ends at t 4 . The start and end time points of each stage are the key time information for adjusting the valves of the recovery branches during the exhaust gas recovery process.

[0056] The recovery mapping coding is a time series control coding that indicates which recovery branches should be used for exhaust gas recovery in each stage. The recovery mapping coding contains detailed information about the recovery branches that need to be started in each stage, including the type of recovery branch, working time, etc. For example, in the first stage (from t 1 to t 2 ), the mapping coding indicates that the single-component recovery branch A needs to be started to process nitrogen; in the second stage (from t 2 to t 3 ), it indicates that the multi-component recovery branch B needs to be started to process multiple exhaust gas components.

[0057] Based on the recycling mapping coding timing and the start and end time points of each stage, the waste gas recycling device realizes the dynamic treatment of waste gas through valve control. At the start time of each stage, the corresponding recycling branch is opened according to the recycling mapping coding, and the corresponding branch is closed at the end time. For example, when the waste gas process enters the first stage (from t 1 to t 2 ), the single-component recycling branch A is opened according to the indication of the recycling mapping coding, and this branch is closed at the end of this stage (at time t 2 ). If other branches need to be opened in the second stage, the corresponding recycling branch will be opened at time t 2 through the indication of the recycling mapping coding.

[0058] To achieve this process, an automated control system is used to execute the control instructions for the recycling branch valves. The control system reads the recycling mapping coding and the stage time points, and sends on-off signals to the recycling device at appropriate moments to ensure the high efficiency and accuracy of waste gas recycling throughout the production process.

[0059] Furthermore, the multi-branch waste gas recycling device controls the valves of the multiple recycling branches to start or close according to the recycling mapping coding timing and the start and end time points. The method further includes: Identifying the valve closing timing corresponding to the recycling branch in the starting state; collecting the gas flow rate and pressure feedback information corresponding to the recycling branch in the starting state; predicting the predicted gas flow rate and predicted pressure feedback information at the closing timing based on the gas flow rate and the pressure feedback information; judging whether each recycling branch meets the closing response condition based on the predicted gas flow rate and predicted pressure feedback information. If not, dynamically correcting the closing timing of the valve corresponding to the recycling branch.

[0060] In the waste gas recycling system, multiple recycling branches work in parallel, and it is monitored in real time which recycling branches are in the enabled state, that is, the branches currently recycling waste gas. Specifically, it is judged which valves of the recycling branches have been opened and are in the normal working state through the feedback signals of sensors or the control system, and these branches are marked as the starting state. Identifying the valve closing timing corresponding to the recycling branch in the starting state, and each valve of the recycling branch has a closing timing, that is, a predetermined time when it should be closed at a certain moment.

[0061] Flow meters and pressure sensors are installed at the inlet and outlet of each recycling branch to collect the real-time gas flow rate and pressure data of each recycling branch. These data can reflect the actual situation in the waste gas recycling process and provide a basis for predicting the valve closing timing.

[0062] Based on the gas flow rate and pressure feedback information, mathematical models, physical models or machine learning algorithms are used to predict the changes in waste gas flow rate and pressure under the valve closing timing. These models are based on historical data, production process parameters and the dynamic characteristics of the waste gas recovery system, and can predict the flow situation of waste gas after the valve is closed. The goal of the prediction is to determine whether abnormal fluctuations will occur in the waste gas flow rate and pressure at a specific moment when the valve is closed. Through prediction, the changes in waste gas flow rate and pressure when the valve is closed are obtained. For example, assume that at a certain moment, closing the valve may cause a sudden increase or decrease in the waste gas flow rate, and this change can be foreseen through predictive analysis.

[0063] Some closing response conditions are preset, such as the maximum fluctuation range of the flow rate and pressure. These conditions are formulated based on the design and operation requirements of the waste gas recovery equipment to ensure that the equipment will not cause a reduction in treatment efficiency or equipment damage due to unstable waste gas flow when the valve is closed. If the predicted flow rate or pressure change of a certain recovery branch exceeds the closing response conditions, for example, the flow rate is too high or the pressure is too low, then the closing timing of the valve corresponding to this recovery branch is dynamically corrected. The dynamic correction includes delaying the closing time of the valve or closing the valve in advance to ensure the smoothness of waste gas flow. For example, if it is predicted that the flow rate will be too fast after closing the valve, the closing time of the valve can be extended to gradually reduce the waste gas flow rate and prevent sudden pressure changes.

[0064] Furthermore, using a cost function to perform K - value clustering on the set of candidate partition points as stage boundary candidates, including clustering constraint conditions; the clustering constraint condition is that the K value is less than the total number of branches of the recovery branches in the multi - branch waste gas recovery device.

[0065] Adding the clustering constraint condition to the K - value clustering process ensures that the number of stage divisions is less than the total number of recovery branches of the waste gas recovery device. Among them, the total number of recovery branches is a fixed resource parameter determined by the design of the waste gas recovery device. For example, the recovery device has 10 recovery branches, and these branches can process waste gas at different stages. In the clustering process, set the constraint condition: the K value is less than the total number of recovery branches, which means that the number of stages of the waste gas process must be less than the number of recovery branches. This can ensure that each stage has enough recovery branches for processing and avoid having no corresponding recovery branches in some stages. For example, if the recovery device has 10 recovery branches, then the K value (i.e., the number of waste gas stages) must be less than 10, so as to ensure that each stage can be assigned corresponding recovery branches for processing.

[0066] Furthermore, establish the encoding between the waste gas process of each stage and the recovery branches that pass the matching. The method includes: If there are multiple returned recovery branches that pass the matching, obtain the recovery effect indicators corresponding to the recovery branches that pass the matching; determine the first recovery branch from the recovery branches that pass the matching according to the recovery effect indicators, and establish the encoding between the exhaust gas flow in each stage and the first recovery branch.

[0067] When the exhaust gas flow in a certain stage matches multiple recovery branches, evaluate the effects of each recovery branch to determine which recovery branch is most suitable for treating the exhaust gas in this stage. For example, if the exhaust gas contains multiple components, there may be multiple recovery branches that can treat these components, but the efficiency, processing capacity, etc. of each branch will be different. Each recovery branch has a set of recovery effect indicators, which describe the recovery efficiency, processing capacity, exhaust gas component adaptability, and processing time of this branch under specific conditions. These effect indicators help the system evaluate the performance of each recovery branch and thus make a choice.

[0068] By comparing the recovery effect indicators of multiple recovery branches, select the recovery branch with the best effect as the first recovery branch. For example, if the recovery branch A has a higher recovery efficiency and stronger processing capacity, then select branch A as the first recovery branch in this stage. After selecting the first recovery branch, establish the encoding between the exhaust gas flow in each stage and the first recovery branch. This mapping encoding contains information such as the time node of the exhaust gas stage, the exhaust gas type, and the identifier of the recovery branch. Through this encoding, the exhaust gas recovery process can be accurately controlled in subsequent operations to ensure that the exhaust gas in each stage is treated by a suitable recovery branch.

[0069] In summary, the recovery and treatment method for exhaust gas based on chemical product production provided by the embodiments of the present application has the following technical effects: By collecting real-time data on waste gas emissions during the chemical production process, accurate information on waste gas generation can be obtained, ensuring that waste gas recovery and treatment can be controlled based on accurate and complete data, enhancing the reliability of subsequent steps; through time series identification, the waste gas process data is arranged in chronological order and organized into a time series waste gas parameter table, enabling each waste gas node to be clearly calibrated on the time axis, ensuring that the generation patterns and characteristics of waste gas can be clearly and accurately reflected; by dividing stages according to the time series waste gas parameter table, the waste gas process is divided into multiple different stages, allowing more targeted treatment methods to be adopted in each stage, avoiding the limitations of a single treatment method. The multi-stage waste gas process enables the system to flexibly respond to changes in different types of waste gas during the production process and adapt to the diverse needs in chemical product production; by connecting a multi-branch waste gas recovery device and using a multi-branch mapping encoder to map the waste gas process and recovery branches, the appropriate recovery branch can be automatically selected for waste gas recovery according to the requirements of waste gas in different stages, avoiding resource waste or improper configuration, thereby improving the recovery efficiency and reducing unnecessary energy consumption; by controlling the opening and closing of the valves of the recovery branches through the recovery mapping code time series, precise scheduling of the waste gas treatment process is achieved, and according to the recovery mapping code, the waste gas recovery device can automatically adjust the working states of multiple recovery branches, reducing the need for manual intervention and making the waste gas recovery process more automated and intelligent.

[0070] Embodiment 2, based on the same inventive concept as the waste gas recovery and treatment method for chemical product production in the foregoing embodiment, as Figure 3 shown, the embodiment of the present application provides a waste gas recovery and treatment system for chemical product production, and the system includes: A data acquisition module 10 for collecting waste gas process record data in the production of chemical products.

[0071] A time series identification module 20 for performing time series identification on the waste gas process record data to determine a time series waste gas parameter table, and the time series waste gas parameter table stores waste gas time series nodes and the corresponding waste gas types, waste gas concentrations, and waste gas contents at each node in chronological order.

[0072] A stage division module 30 for dividing stages according to the time series waste gas parameter table and outputting a multi-stage waste gas process.

[0073] A mapping coding module 40 for connecting a multi-branch waste gas recovery device, where the multi-branch waste gas recovery device includes multiple recovery branches, inputting the multi-stage waste gas process into a multi-branch mapping encoder, establishing a mapping code between the multi-stage waste gas process and the multiple recovery branches, and obtaining a recovery mapping code.

[0074] The valve control module 50 is used to control the valves of the multiple recovery branches to start or close according to the recovery mapping coding time sequence when producing the chemical product by the multi-branch waste gas recovery device.

[0075] Furthermore, the multiple recovery branches at least include one single-component recovery branch and at least one mixed-component recovery branch; the single-component recovery branch is used for recovering and processing a single waste gas type, and the mixed-component recovery branch is used for recovering and processing a mixed waste gas type.

[0076] Furthermore, the multi-branch mapping encoder includes a preset rule library, and the preset rule library stores rule tags of the multiple recovery branches. The rule tags include the single / mixed component processing capabilities, the waste gas types suitable for processing, and the branch processing efficiency of each recovery branch; when the multi-branch mapping encoder receives the multi-stage waste gas process, it matches the time-sequence waste gas parameter table of each stage of the waste gas process with the rule tags of each recovery branch in the preset rule library, obtains the recovery branches that pass the matching, establishes the coding between each stage of the waste gas process and the recovery branches that pass the matching, and outputs the recovery mapping coding.

[0077] Furthermore, the stage division module 30 is used to perform the following operation steps: According to the time-sequence node arrangement of the time-sequence waste gas parameter table, construct a time-sequence feature vector sequence for each time-sequence node. The time-sequence feature vector sequence includes the gas concentrations corresponding to the time nodes and the change rate of the concentration with time; obtain a set of time-sequence feature vector sequences for each time-sequence node; use a sliding window to calculate the vector similarity change of the set of time-sequence feature vector sequences, output candidate mutation time-sequence nodes, and output the candidate mutation time-sequence nodes as a set of candidate division points; use a cost function to perform K-value clustering on the set of candidate division points as the stage boundary candidates, and output the optimal K-value solution, where the K-value is the number of stages for division; perform stage division on the time-sequence waste gas parameter table according to the optimal K-value solution, and output the multi-stage waste gas process.

[0078] Furthermore, the stage division module 30 is used to perform the following operation steps: Among them, the vector similarity change includes calculating the cosine similarity change rate of the set of time-sequence feature vector sequences; performing a significance test on the cosine similarity change rate, and outputting the time-sequence nodes greater than the preset cosine similarity change threshold as candidate mutation time-sequence nodes.

[0079] Furthermore, record the start and end time points corresponding to the multi-stage waste gas process, and the multi-branch waste gas recovery device controls the valves of the multiple recovery branches to start or close according to the recovery mapping coding time sequence and the start and end time points.

[0080] Furthermore, the stage division module 30 is used to perform the following operation steps: Identify the valve closing timings corresponding to the recovery branches in the startup state; collect the gas flow rate and pressure feedback information corresponding to the recovery branches in the startup state; predict the predicted gas flow rate and predicted pressure feedback information at the closing timings based on the gas flow rate and the pressure feedback information; determine whether each recovery branch meets the closing response condition based on the predicted gas flow rate and predicted pressure feedback information, and if not, dynamically correct the closing timings of the valves corresponding to the recovery branches.

[0081] Furthermore, using the candidate division point set as the stage boundary candidates, perform K-means clustering using a cost function, including clustering constraint conditions; the clustering constraint condition is that the value of K is less than the total number of branches of the recovery branches in the multi-branch waste gas recovery device.

[0082] Furthermore, the mapping and encoding module 40 is used to perform the following operation steps: If multiple matched recovery branches are returned, obtain the recovery effect indicators corresponding to the matched recovery branches; determine the first recovery branch from the matched recovery branches according to the recovery effect indicators, and establish the encoding between the waste gas flow of each stage and the first recovery branch.

[0083] Through the foregoing detailed description of the recovery and treatment method for waste gas from chemical product production in this specification, those skilled in the art can clearly know the recovery and treatment system for waste gas from chemical product production in this embodiment. Since it corresponds to the method disclosed in the embodiment, it is described relatively simply. For related parts, reference can be made to the description in the method part.

[0084] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for recycling waste gas from chemical product production, characterized in that: The method comprises: Collecting waste gas process record data of chemical products production; Performing time sequence identification on the exhaust gas process record data to determine a time sequence exhaust gas parameter table, wherein the time sequence exhaust gas parameter table stores exhaust gas time sequence nodes and exhaust gas types, exhaust gas concentrations, and exhaust gas contents corresponding to each node in a time sequence order; Divide the exhaust gas flow into multiple stages according to the exhaust gas parameter table; Connecting a multi-branch exhaust gas recovery device, the multi-branch exhaust gas recovery device includes a plurality of recovery branches, inputting the multi-stage exhaust gas process into a multi-branch mapping encoder, establishing a mapping code between the multi-stage exhaust gas process and the plurality of recovery branches, and obtaining a recovery mapping code; When the chemical product is produced, the multi-branch waste gas recovery device controls the valves of the multiple recovery branches to start or close according to the recovery mapping code timing.

2. The method according to claim 1, characterized in that The multiple recovery branches include at least one single-component recovery branch and at least one mixed-component recovery branch; The single-component recovery branch is used for recovering and processing a single type of exhaust gas, and the mixed-component recovery branch is used for recovering and processing mixed types of exhaust gas.

3. The method according to claim 1, characterized in that The multi-branch mapping encoder includes a preset rule base, in which the rule labels of the multiple recovery branches are stored, and the rule labels include the single / mixed component processing capacity of each recovery branch, the exhaust gas type adapted for processing, and the branch processing efficiency; When the multi-branch mapping encoder receives the multi-stage exhaust gas process, it matches the timing exhaust gas parameter table of each stage of the exhaust gas process with the rule label of each recovery branch in the preset rule library, obtains the recovery branch that passes the match, establishes the encoding between each stage of the exhaust gas process and the recovery branch that passes the match, and outputs the recovery mapping code.

4. The method according to claim 1, characterized in that The method includes: dividing the exhaust gas flow into stages according to the time sequence exhaust gas parameter table and outputting a multi-stage exhaust gas flow. According to the arrangement of the time series nodes of the time series exhaust gas parameter table, a time series feature vector sequence of each time series node is constructed, wherein the time series feature vector sequence includes the concentration of each gas corresponding to the time node and the change rate of the concentration with time; Obtain a set of time series feature vector sequences of each time series node; Using a sliding window to calculate the vector similarity change of the time series feature vector sequence set, outputting candidate variation time series nodes, and outputting the candidate variation time series nodes as a candidate partition point set; The candidate partition point set is used as the stage boundary candidate to perform K value clustering using the cost function, and the K value optimal solution is output, where the K value is the number of stage divisions; The time sequence exhaust gas parameter table is divided into stages according to the K value optimal solution, and a multi-stage exhaust gas process is output.

5. The method according to claim 4, characterized in that A sliding window is used to calculate the vector similarity change of the time series feature vector sequence set, and a candidate variation time series node is output. The method include: The vector similarity change includes calculating the cosine similarity change rate of the time series feature vector sequence set; A significance test is performed on the cosine similarity change rate, and the time series nodes with a cosine similarity change rate greater than a preset cosine similarity change threshold are output as candidate variant time series nodes.

6. The method according to claim 4, characterized in that The start and end time points corresponding to the multi-stage exhaust gas process are recorded, and the multi-branch exhaust gas recovery device controls the valves of the multiple recovery branches to start or close according to the recovery mapping coding timing and the start and end time points.

7. The method according to claim 6, characterized in that The multi-branch exhaust gas recovery device controls the valves of the multiple recovery branches to start or close according to the recovery mapping coding timing and the start and end time points, and the method also includes: Identify the valve closing sequence corresponding to the recovery branch in the start-up state; Collect the gas flow rate and pressure feedback information corresponding to the recovery branch in the startup state; Predicting predicted gas flow rate and predicted pressure feedback information at the closing time sequence according to the gas flow rate and the pressure feedback information; Based on the predicted gas flow rate and predicted pressure feedback information, it is determined whether each recovery branch meets the closing response condition. If not, the closing timing of the corresponding valve of the recovery branch is dynamically corrected.

8. The method according to claim 4, characterized in that The candidate partition point set is used as a stage boundary candidate, and K-value clustering is performed using a cost function, including clustering constraints; The clustering constraint condition is that the K value is less than the total number of recovery branches in the multi-branch exhaust gas recovery device.

9. The method according to claim 3, characterized in that Establish the coding between each stage of the exhaust gas process and the matching recovery branch, including: If multiple recycling branches that pass the match are returned, obtain the recycling effect indicators corresponding to the recycling branches that pass the match; The first recovery branch is determined from the recovery branches that have been matched according to the recovery effect index, and a code is established between the exhaust gas process at each stage and the first recovery branch.

10. A recycling and treatment system for waste gas produced by chemical products, characterized in that: The system is used to implement the method for recovering and treating waste gas produced by chemical products according to any one of claims 1 to 9, comprising: Data acquisition module, used to collect waste gas process record data of chemical product production; A time sequence identification module, used to perform time sequence identification on the exhaust process record data and determine a time sequence exhaust gas parameter table, wherein the time sequence exhaust gas parameter table stores exhaust gas time sequence nodes and exhaust gas types, exhaust gas concentrations and exhaust gas contents corresponding to each node in a time sequence order; A stage division module, used for performing stage division according to the time sequence exhaust gas parameter table and outputting a multi-stage exhaust gas process; A mapping coding module, used to connect a multi-branch exhaust gas recovery device, the multi-branch exhaust gas recovery device includes a plurality of recovery branches, input the multi-stage exhaust gas process into a multi-branch mapping encoder, establish a mapping code between the multi-stage exhaust gas process and the plurality of recovery branches, and obtain a recovery mapping code; The valve control module is used to control the valves of the multiple recovery branches to start or close according to the recovery mapping code timing when the chemical product is produced.

Images