A method of avoiding coke oven gas incorporation into a TSA adsorption system

Abstract

This invention discloses a method to prevent coke oven gas from being incorporated into the TSA adsorption system. It relates to the field of coke oven gas TSA coarse and fine desorption purification process control, primarily addressing the process pain point of significant fluctuations in raw gas flow, pressure, and impurity concentration caused by coke oven commissioning, maintenance, and hot standby switching, resulting in impurity penetration into the TSA adsorption tower and exceeding purified gas standards. The method first involves hardware pipeline modifications, upgrading the original 422 regeneration adsorption process to a 4-3-1 process, significantly improving the flexibility of regeneration gas regulation to adapt to frequent coke oven switching conditions. Then, a digital twin and a first-principles dynamic mechanism model of the adsorption tower are built to achieve advance prediction of operating conditions, disturbance simulation, risk classification and control, PLC pre-adjustment, model closed-loop self-calibration, and quantitative evaluation of defense effects. This constructs a full-process proactive defense closed-loop control system, forming a complete automated closed loop of "prediction-simulation-control-verification-iteration," continuously ensuring the long-term stable and compliant operation of the TSA system.

Description

Technical Field

[0001] This invention relates to the field of TSA coarse and fine removal purification process control technology for coke oven gas, and particularly to a method for preventing coke oven gas from being incorporated into the TSA adsorption system. Background Technology

[0002] In coking production, multiple coke ovens will be put into operation, under maintenance or in hot standby in turn according to the production schedule. When coke ovens are connected to / out of the pipeline network, the instantaneous flow rate of raw coal gas, the pressure of the gas collecting pipe and the concentration of coal gas impurities will be subject to violent disturbances and fluctuations.

[0003] Traditional furnace gas purification TSA adsorption section generally adopts the 422 regeneration process, that is, two towers adsorption and two towers regeneration under the four-tower configuration. One regeneration tower is heated and the other is cold blown. The heating regeneration gas source relies solely on the cold blown gas recovery and reuse. The regeneration gas pipeline control valve configuration is simple, and the regeneration condition adjustment flexibility is insufficient, making it difficult to adapt to large gas impact conditions.

[0004] Furthermore, current TSA system control is mostly a reactive adjustment mode, which can only make delayed adjustments after the impurities at the adsorption tower outlet exceed the standard. At the same time, no disturbance risk classification and judgment system has been established. All disturbances are only adjusted using a uniform adjustment strategy. Over-adjustment under low-impact conditions leads to energy waste, while insufficient adjustment under high-impact conditions causes the purified gas to exceed the standard.

[0005] To address the aforementioned technical shortcomings, a solution is proposed. Summary of the Invention

[0006] The purpose of this invention is to provide a method to avoid coke oven gas from being incorporated into the TSA adsorption system, which aims to offset the impact of gas flow, pressure, and impurity concentration caused by the commissioning or shutdown of coke ovens in advance, prevent impurities from penetrating the adsorption tower instantaneously, and stabilize the purified gas indicators.

[0007] The objective of this invention can be achieved through the following technical solution: a method for preventing coke oven gas from being incorporated into a TSA adsorption system, comprising the following steps:

[0008] Step 1: Optimize the regeneration adsorption process using 422 to a 4-3-1 process;

[0009] Step 2: Read and analyze the coke oven pushing plan table to generate a "4-3-1 operating condition prediction timetable" for the next 24 hours, and mark "pre-action coke oven";

[0010] Step 3: Build a digital twin model of the coke oven pipeline network, collect real-time data from the site as the simulation boundary, simulate the entire process of furnace body switching, output continuous time-series simulation curves, extract key feature values, and combine with the impurity precipitation dynamic analysis model to output the instantaneous peak concentration of impurities, cumulative precipitation amount and impact curve.

[0011] Step 4: Based on the TSA first-principles dynamic mechanism model, import the gas flow temperature-pressure-flow dynamic curve, impurity concentration impact curve, and read the tower internal state simulation. Calculate the minimum safety margin Tamin, flow impact intensity Lflow, and impurity concentration impact intensity Imp. Determine the safety, controllable, severe, and critical levels under multiple conditions, and match and output the corresponding level control strategy.

[0012] Step 5: Convert the hierarchical control strategy into PLC timing instructions and pre-set them. After the working condition ends, compare the deviation between the measured and simulated curves, calibrate multiple model parameters for over-limit calibration, calculate the purification qualification rate and impact attenuation rate, obtain the active defense evaluation coefficient by weighting, and issue stabilization or alarm instructions according to the threshold.

[0013] Preferably, the 4-3-1 process includes:

[0014] The regeneration gas is equipped with calibrated programmable valves A1 and A2 as the main switch for coarse and fine regeneration gas, and programmable valves A3 and A4 as shut-off valves during hot blowing regeneration.

[0015] The upstream of the calibrated programmable valves A1 and A2 are respectively added to the designated heater inlets E1 and E2, branch pipes PG1 and PG2;

[0016] Branch lines PG3 and PG4 are added to the outlets of heaters E1 and E2, and are respectively connected to the downstream of programmable valves A3 and A4 on the bottom horizontal pipe of the valves with designated serial numbers.

[0017] Preferably, the 24-hour "4-3-1 operating condition prediction timetable" includes the coke oven number, switching type (exit / integration), and planned time.

[0018] When the system time reaches T_adv minutes before any operating condition switching time, the "active defense process" is automatically triggered, and the coke oven is marked as a "pre-action coke oven", and the trigger signal and pre-action coke oven ID are output.

[0019] Preferably, the analysis process in step three is as follows:

[0020] Based on the CAD drawing of 4-3-1, a digital twin sub-model of coke oven-pipeline network is constructed. The initial boundary conditions of the digital twin sub-model of coke oven-pipeline network are set to simulate the complete 4-3-1 switching process and output continuous time-series simulation curves: instantaneous flow rate of main pipe F(t), pipeline pressure P(t), and gas temperature T(t).

[0021] Calculate key characteristic values: maximum instantaneous flow rate change (ΔF / Δt), maximum flow / pressure overshoot, and fluctuation duration;

[0022] The dynamic analysis model for impurity precipitation, which is pre-trained offline using historical data, outputs the instantaneous peak concentration of each impurity, the cumulative precipitation of impurities over the entire cycle, and generates a simulated impurity concentration impact curve that changes over time.

[0023] Preferably, the analysis process of the hierarchical control strategy is as follows:

[0024] The simulation was performed based on the first-principles dynamic mechanism model of the TSA adsorption tower. During the simulation, the concentration of each key impurity Ciout(t) in the gas phase at the outlet of the adsorption tower was continuously monitored, where i represents the key impurity and i is a natural number greater than zero.

[0025] The Ciout(t) is compared with the preset purified gas qualification threshold in real time. The precise moment when Ciout(t) ≥ the preset purified gas qualification threshold for the first time is recorded. The earliest time when penetration occurs among all key impurities is taken as the predicted impurity penetration time Tbt in this evaluation.

[0026] Obtain the remaining adsorption time Tsxc of each online adsorption tower and calculate the safety margin curve as a function of time: Tmar(t) = remaining breakthrough time (t) - Tsxc, where the remaining breakthrough time (t) is: if the impact terminates at time t, the dynamic prediction value of the remaining breakthrough time calculated from time t. The minimum value of this curve is found during the entire fluctuation period, which is the minimum safety margin Tamin.

[0027] Calculate the flow impact intensity Lflow, the impurity concentration impact intensity Imp, and the comprehensive stress index Syl. Combined with the set logical judgment conditions, output the safety level / controllable level / severe level / critical level.

[0028] Preferably, based on a pre-built level-control strategy library matching, the hierarchical control strategy corresponding to the current level is output;

[0029] The graded control strategy includes adsorption timing adjustment commands, micro-purification module switching commands, regeneration gas flow rate adjustment values, pressure equalization duration parameters, and action execution time nodes.

[0030] Preferably, the analysis process for the stabilization or alarm command is as follows:

[0031] Based on the hierarchical control strategy corresponding to the current level, it is broken down into a sequence of digital and analog timing action instructions that can be recognized by the PLC, and the execution pre-time of each instruction is marked.

[0032] Before the planned switchover time for the coke oven 4-3-1 operating condition arrives, the entire instruction sequence is sent to the TSA local control unit PLC. The PLC executes the instructions sequentially according to the time nodes. After the 4-3-1 operating condition switchover event is completed, curves of the actual flow rate, temperature, pressure, and online impurity concentration data are plotted.

[0033] The difference between the actual impact curve on site and the simulated dynamic curve in the digital twin simulation is calculated step by step to obtain the global average deviation and peak deviation. If the global average deviation or peak deviation exceeds the system's preset deviation threshold, a calibration command is triggered.

[0034] The cumulative time during which all indicators of the purified gas are qualified within the evaluation window is calculated. The evaluation window represents the time from the start of the operating condition switch to the end of the system's determination that the fluctuation has ended and the TSA has returned to steady state.

[0035] The overall pass rate of purified gas is calculated based on the cumulative duration / evaluation window duration. The impact attenuation rate is then calculated as follows: Impact attenuation rate = (1 - actual maximum fluctuation amplitude / assumed maximum fluctuation amplitude without defense) × 100%.

[0036] The active defense evaluation coefficient is calculated based on the comprehensive pass rate of purified gas × preset weight coefficient 1 + impact attenuation rate × preset weight coefficient 2. Combined with the threshold of the active defense evaluation coefficient, a set command or alarm command is output.

[0037] The beneficial effects of this invention are as follows:

[0038] (1) The present invention transforms the traditional 422 adsorption regeneration process into a 4-3-1 process, adds and reuses special programmable valves and optimizes pipelines and safety valve pressure tapping pipelines, reconstructs independent regeneration gas circuits for coarse and fine desorption units, flexibly adapts to the gas flow and composition disturbances caused by the start-up and shutdown of coke ovens, and greatly improves the freedom of regeneration gas supply regulation, so as to solve the problem of poor adaptability of the original regeneration mode to cope with large airflow impacts.

[0039] (2) This invention analyzes the coke oven pushing plan in advance to generate a 24-hour operating timetable, triggers the defense process in advance before the oven body is switched, abandons the traditional passive adjustment mode after the fact, and completes the disturbance simulation based on the coke oven-pipeline digital twin to accurately predict the impact curves of flow, pressure and impurity concentration, lock the disturbance intensity and duration in advance, reserve sufficient operation window for pre-control, and at the same time, graded and matched differentiated control strategies to accurately balance purification stability and energy consumption, and compare the deviation between the on-site measured curve and the simulation curve after the operating condition is completed, realize closed-loop self-calibration iteration, and the simulation model continues to be highly accurate. Attached Figure Description

[0040] The invention will now be further described with reference to the accompanying drawings;

[0041] Figure 1This is a schematic diagram illustrating the method of the present invention;

[0042] Figure 2 This is a schematic diagram of the hierarchical control decision analysis of the present invention. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments;

[0045] Example 1: Please refer to Figures 1 to 2 As shown, the present invention is a method for preventing coke oven gas from being incorporated into a TSA adsorption system, comprising the following steps:

[0046] Step 1: Optimize the regeneration adsorption process using 422 to a 4-3-1 process;

[0047] Among them, the regeneration adsorption of 422 is as follows: two towers in the four towers are adsorbed and two towers are regenerated. Among the two towers of regeneration, one tower is cold-blown and the other tower is heated. The regeneration gas of the heated regeneration tower comes from the reuse of the cold-blown gas.

[0048] 4-3-1 Process: The regenerated gas is equipped with calibrated programmable valves A1 and A2 as the main switch for coarse and fine regenerated gas.

[0049] For example, programmable valve A1: programmable valve 22XV-110, 22: section code, representing the furnace gas TSA coarse desorption adsorption unit, XV instrument tag number: two-position pneumatic programmable shut-off valve (fully open / fully closed only, DCS time-sequenced automatic control, TSA adsorption regeneration process dedicated valve), numeric suffix: pipeline function sequence number, that is, the main control programmable shut-off valve of the regeneration gas main pipe of the coarse desorption unit of section 22 (regeneration gas main programmable valve).

[0050] Programmable valve A2:22XV-210;

[0051] Add programmable valves A3 and A4 as shut-off valves during hot blowing regeneration;

[0052] For example, the programmable valve A3:22XV-108 (reused), a programmable shut-off valve for the hot blowing regeneration circuit of the coarse stripping unit in section 22 (the original hot blowing regeneration programmable valve).

[0053] Programmable valve A4:22XV-212, a dedicated programmable shut-off valve for the hot blowing regeneration circuit of the fine descaling unit in section 22 (corresponding one-to-one with the function of 22XV-108).

[0054] The upstream of the calibrated programmable valves A1 and A2 are respectively added to the designated heater inlets E1 and E2, branch pipes PG1 and PG2;

[0055] For example, heater E1: E-2201, heater E2: E-2203;

[0056] Branch pipe PG1: Branch pipe PG-22109-400-B01B;

[0057] Branch pipe PG2: PG-22209-350-B01B;

[0058] Branch lines PG3 and PG4 are added to the outlets of heaters E1 and E2, and they are respectively connected to the downstream of programmable valves A3 and A4 on the bottom horizontal pipe of valve No. 3.

[0059] For example: Branch line PG3: PG-10-400-D01B, Branch line PG4: G-22210-350-D01B

[0060] The root pipes of the calibrated safety valves P1 and P2 are moved upstream of the calibrated programmable valves A1 and A2 as protection for the regeneration gas system.

[0061] For example: safety valve P1 is calibrated as 22PSV-104, and safety valve P2 is calibrated as 22PSV-204;

[0062] Step Two: Read and analyze the coke oven pushing plan table to generate a "4-3-1 operating condition prediction timetable" for the next 24 hours, and mark "pre-action coke ovens". Specifically, this includes:

[0063] S1: Read and parse the coke oven pushing plan in real time, identify the "planned maintenance / hot standby start time" and "planned commissioning time" of the target coke oven, obtain countdown information, and generate a "4-3-1 operating condition prediction timetable" for the next 24 hours, including information such as coke oven number, switching type (exit / integration), and planned time.

[0064] S2: When the system time reaches T_adv minutes (e.g., 30 minutes) before any operating condition switching time, the "active defense process" is automatically triggered, and the coke oven is marked as a "pre-action coke oven", and information such as the trigger signal and the pre-action coke oven ID is output;

[0065] Step 3: Build a digital twin model of the coke oven pipeline network, collect real-time data from the site as the simulation boundary, simulate the entire process of furnace body switching, output continuous time-series simulation curves, and extract key feature values. Combined with the impurity precipitation dynamic analysis model, output the instantaneous peak concentration of impurities, cumulative precipitation amount, and impact curves. Specifically, this includes:

[0066] Based on the CAD drawing of 4-3-1, a digital twin sub-model of coke oven-pipeline network is constructed. The DCS real-time database is called up to collect the current output status of the pre-action coke oven, the pressure of the gas collecting pipe, the flow rate / temperature of the raw coal gas outlet, and the current pressure, flow rate and resistance coefficient of the corresponding main pipe and branch pipe of the pipeline network, which are used as the initial boundary conditions for the simulation of the digital twin sub-model of coke oven-pipeline network.

[0067] Dynamic disturbance simulation: Simulate the entire 4-3-1 switching process of the pre-action coke oven, and output continuous time-series simulation curves: instantaneous flow rate F(t) of the main pipe, pipeline pressure P(t) and gas temperature T(t), and store millisecond-level simulation time-series data, where t is the sampling timestamp and t>0;

[0068] Calculate the key characteristic values ​​of the instantaneous flow rate F(t), pipeline pressure P(t), and gas temperature T(t) in the simulated main pipe: maximum instantaneous flow rate change rate (ΔF / Δt), maximum flow / pressure overshoot, and fluctuation duration, etc.

[0069] The time-series simulation curve simulated in step S3 is used as input and fed into the impurity precipitation dynamic analysis model that has been pre-trained offline based on historical data. The model outputs the instantaneous peak concentration of each impurity, the cumulative precipitation of impurities over the entire cycle, and generates a simulated impurity concentration impact curve that changes over time.

[0070] Example 2: Step 4: Based on the TSA first-principles dynamic mechanism model, import the gas flow temperature-pressure-flow dynamic curves and impurity concentration impact curves, and read the tower internal state simulation. Calculate the minimum safety margin (Tamin), flow impact intensity (Lflow), and impurity concentration impact intensity (Imp). Determine the safety, controllable, severe, and critical levels under multiple conditions, and match the corresponding graded control strategies. Specifically, this includes:

[0071] The first-principles dynamic mechanism model of the TSA adsorption tower (which essentially describes the coupled process of "adsorption-heat transfer-mass transfer-flow" of gas in a porous adsorbent bed) is used to fully couple the multi-component adsorption competition effect, gas-solid mass transfer process, and gas-solid heat transfer process in the bed, providing a basis for high-precision dynamic simulation.

[0072] The simulation is based on the first-principles dynamic mechanism model of the TSA adsorption tower, with the following simulation boundary and initial condition configurations:

[0073] Time-varying boundary conditions: Import dynamic curves of airflow temperature, pressure and flow, and impact curves of impurity concentration;

[0074] Initial internal state: Read the current adsorption time, bed temperature, adsorbent loading rate, and remaining adsorption capacity in each adsorption tower of TSA as the initial state of the simulation;

[0075] During the simulation, the concentration of each key impurity Ciout(t) in the gas phase at the outlet of the adsorption tower is continuously monitored, where i represents the key impurity and i is a natural number greater than zero.

[0076] The Ciout(t) is compared with the preset qualified threshold for purified gas in real time, and the precise moment when Ciout(t) ≥ the preset qualified threshold for purified gas is recorded.

[0077] The earliest time of penetration among all critical impurities is taken as the predicted impurity penetration time Tbt in this assessment.

[0078] Read the real-time operating data of the TSA system to obtain the remaining adsorption time Tsxc of each online adsorption tower. As the simulation progresses, the remaining adsorption time Tsxc of the current adsorption tower is known and continuously decreases. Calculate the safety margin curve that changes with time: Tmar(t) = remaining penetration time (t) - Tsxc, where the remaining penetration time (t) is: if the impact terminates at time t, the dynamic prediction value of the remaining penetration time calculated from time t.

[0079] The minimum value of the curve during the entire period of fluctuation is the minimum safety margin Tamin.

[0080] Calculate the flow impact intensity Lflow: simulate the maximum instantaneous flow rate change (ΔF / Δt) / the maximum flow rate change that the equipment can withstand at the factory;

[0081] Calculate the impurity concentration impact intensity Imp: instantaneous peak impurity concentration / process design baseline impurity concentration;

[0082] Comprehensive Stress Index Syl: Flow impact intensity Lflow × Impurity concentration impact intensity Imp;

[0083] Set logical judgment conditions:

[0084] If all the following conditions are met simultaneously: minimum safety margin Tamin > preset safety margin threshold, flow impact intensity Lflow < preset flow low impact intensity upper limit coefficient, impurity concentration impact intensity Imp < preset impurity low impact intensity upper limit coefficient, and comprehensive stress index Syl < low comprehensive stress threshold, then it is judged as a safe level.

[0085] If the safety level is not met, but all conditions are met simultaneously: minimum safety margin Tamin > 0 (no negative safety margin, no instantaneous penetration will occur throughout the process), flow impact intensity Lflow ≤ preset upper limit coefficient of impact intensity in the controllable flow range, impurity concentration impact intensity Imp ≤ preset upper limit coefficient of impact intensity in the controllable impurity range, and comprehensive stress index Syl < medium comprehensive stress threshold, then it is judged as a controllable level.

[0086] If the safety and controllability level is not met, and any of the following conditions are met, and there is no critical qualitative risk signal: -Tneg≤Minimum safety marginTamin<0 (minimum margin is slightly negative, and there is only a short-term instantaneous penetration risk), or medium comprehensive stress threshold≤comprehensive stress indexSyl<high comprehensive stress threshold, then it is judged as a severe level;

[0087] Wherein, Tneg represents the critical threshold of negative margin. Critical qualitative risk signal: Real-time reading of the online monitoring signals of the front-end gas-liquid separator and electrostatic tar collector to determine whether the raw gas carries liquid tar mist or a large amount of dust qualitative risk signal. If so, a critical qualitative risk signal is generated; if not, no signal is generated, i.e., there is no critical qualitative risk signal.

[0088] If any of the following conditions are met, the system is directly classified as critical: minimum safety margin Tamin < -Tneg (minimum safety margin has a large negative offset, and there is a persistent risk of penetration in the disturbance peak range), comprehensive stress index Syl ≥ high comprehensive stress threshold, or receiving qualitative risk signals carrying liquid tar mist and a large amount of dust.

[0089] Based on the output safety level / controllable level / severity level / critical level, and based on the pre-built level-control strategy library, the corresponding hierarchical control strategy is output, including information such as adsorption timing adjustment command, micro-purification module switching command, regeneration gas flow rate adjustment value, pressure equalization duration parameter, and action execution time node.

[0090] For example: the risk level is safe / controllable: generate a conventional mild feedforward scheme, only fine-tune the TSA adsorption tower switching sequence and slightly adjust the regeneration gas volume, without any additional hardware switching actions;

[0091] The risk level is severe / critical, and two core defense strategies are implemented simultaneously:

[0092] Strategy A – Adsorption Timing Reconstruction Strategy: Before the arrival of this disturbance, adjust the TSA running sequence and add a short-term micro-pressure equalization step before the adsorption tower to be put into operation officially enters the adsorption process; extract the high-pressure clean purification gas in the current adsorption end-stage tower and pre-pressurize the standby tower to dilute the inlet impurity concentration and reduce the risk of penetration.

[0093] Strategy B – Diversion Assisted Purification Strategy: Issue a start command to put into operation the hot standby micro fast-response purification modules arranged in parallel at the front end of the TSA; divert part of the fluctuating airflow during the 4-3-1 operating condition switching to the micro modules for pre-adsorption and removal of impurities, thereby reducing the impurity load on the main adsorption tower.

[0094] Step 5: Convert the hierarchical control strategy into PLC timing instructions and pre-set them. After the operating conditions are completed, compare the deviations between the measured and simulated curves. Perform over-limit calibration on multiple model parameters, calculate the purification pass rate and impact attenuation rate, and obtain the weighted active defense evaluation coefficient. Issue stabilization or alarm instructions based on the threshold values. Specifically, this includes:

[0095] Based on the hierarchical control strategy corresponding to the current level, it is broken down into PLC-recognizable digital and analog timing action instruction sequences, and the execution lead time of each instruction is marked.

[0096] Before the planned switching time of coke oven 4-3-1 operating condition arrives, the entire set of instructions is sent to the TSA local control unit PLC. The PLC executes pre-pressure equalization, micro-purification start-up, timing offset and other pre-adjustment operations in sequence according to the time nodes. The system state optimization is completed before the impact airflow reaches the adsorption tower to achieve active defense against disturbances.

[0097] After the 4-3-1 working condition switching event is completed, collect real-time actual flow rate, temperature, pressure, and online impurity concentration data at the TSA inlet, and plot the curves of the actual flow rate, temperature, pressure, and online impurity concentration data at the site.

[0098] The time-step difference between the actual impact curve on site and the simulated dynamic curve in the digital twin simulation is calculated to obtain the global average deviation and peak deviation.

[0099] If the global average deviation or peak deviation exceeds the system's preset deviation threshold, a calibration command is triggered. In response to the calibration command, the system performs the preset warning operation corresponding to the calibration command, such as adjusting the core parameters of at least one of the following models: coke oven-pipeline sub-model, impurity precipitation dynamic analysis model, and TSA first-principles adsorption model, including resistance, mass transfer coefficient, and adsorption equilibrium constant.

[0100] At the same time, the timestamps of all practical data and strategy execution records are precisely aligned with the start time of the working condition switching event to eliminate clock deviations between systems;

[0101] Calculate the cumulative time during which all indicators of the purified gas meet the standards within the evaluation window;

[0102] The evaluation window represents the period from the start of the operating condition switch until the system determines that the fluctuation has ended and the TSA has returned to steady state;

[0103] The overall pass rate of purified gas is calculated based on the cumulative duration / evaluation window duration.

[0104] Using the actual inlet fluctuation as input, run a virtual simulation assuming no defensive action to obtain the assumed outlet fluctuation curve. Calculate the maximum fluctuation amplitude (or standard deviation) of the actual outlet curve and the assumed no-defense curve respectively, and calculate the impact attenuation rate: Impact attenuation rate = (1 - actual maximum fluctuation amplitude / assumed no-defense maximum fluctuation amplitude) × 100%;

[0105] The active defense evaluation coefficient is calculated based on the comprehensive pass rate of purified gas × preset weight coefficient 1 + impact attenuation rate × preset weight coefficient 2. Combined with the active defense evaluation coefficient threshold, if the active defense evaluation coefficient ≥ the preset active defense evaluation coefficient threshold, a stabilization command is generated; if the active defense evaluation coefficient < the preset active defense evaluation coefficient threshold, an alarm command is generated. The preset early warning operation corresponding to the stabilization command or alarm command is immediately executed to adjust the entire active defense scheme and ensure the reliability of subsequent active defense.

[0106] In summary, this solution first involves hardware modification, optimizing the original 422 regeneration adsorption process into a 4-3-1 process. This includes adding and reusing multiple programmable valves, modifying pipelines and safety valve pressure tapping lines, and reconstructing the regeneration gas supply loop. The system analyzes the coke oven pushing schedule in real time to generate a 24-hour operating timetable. Active defense is triggered in advance before switching, marking coke ovens that are about to take action. A digital twin model of the coke oven and pipeline network is built, and field parameters are collected to complete disturbance simulation. Flow and pressure fluctuation characteristics are extracted, and impurity impact curves are output based on the impurity precipitation dynamic analysis model. Safety is calculated using the TSA adsorption first-mechanism model simulation. The system combines margin, flow rate, impurity impact intensity, and comprehensive stress index with tar dust signals to classify the system into four levels: safe, controllable, severe, and critical. It then matches these levels with tiered control strategies such as timing fine-tuning, backup tower pre-pressurization, and micro-purification diversion. The control strategies are translated into PLC timing instructions that are issued in advance to pre-adjust and resist airflow disturbances. After the operating conditions are completed, the deviation between the measured and simulated curves is compared. The system performs out-of-limit calibration of multiple model core parameters, synchronizes data timestamps to eliminate clock errors, calculates the purification pass rate and impact attenuation rate, and calculates the defense evaluation coefficient using weighted averages. Based on the threshold, it outputs a stable or alarm command and iteratively optimizes the entire active defense system.

[0107] The threshold is set for result comparison and analysis to determine whether it is good or bad. The value of the threshold is determined by a combination of large-scale model analysis of the sample data and human experience, and can also be adjusted appropriately based on seasonal or common-sense influencing factors. The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for preventing coke oven gas from being incorporated into a TSA adsorption system, characterized in that, Includes the following steps: Step 1: Optimize the regeneration adsorption process using 422 to a 4-3-1 process; Step 2: Read and analyze the coke oven pushing plan table to generate a "4-3-1 operating condition prediction timetable" for the next 24 hours, and mark "pre-action coke oven"; Step 3: Build a digital twin model of the coke oven pipeline network, collect real-time data from the site as the simulation boundary, simulate the entire process of furnace body switching, output continuous time-series simulation curves, extract key feature values, and combine with the impurity precipitation dynamic analysis model to output the instantaneous peak concentration of impurities, cumulative precipitation amount and impact curve. Step 4: Based on the TSA first-principles dynamic mechanism model, import the gas flow temperature-pressure-flow dynamic curve, impurity concentration impact curve, and read the tower internal state simulation. Calculate the minimum safety margin Tamin, flow impact intensity Lflow, and impurity concentration impact intensity Imp. Determine the safety, controllable, severe, and critical levels under multiple conditions, and match and output the corresponding level control strategy. Step 5: Convert the hierarchical control strategy into PLC timing instructions and pre-set them. After the working condition ends, compare the deviation between the measured and simulated curves, calibrate multiple model parameters for over-limit calibration, calculate the purification qualification rate and impact attenuation rate, obtain the active defense evaluation coefficient by weighting, and issue stabilization or alarm instructions according to the threshold.

2. The method for preventing coke oven gas from being incorporated into the TSA adsorption system according to claim 1, characterized in that, The 4-3-1 process includes: The regeneration gas is equipped with calibrated programmable valves A1 and A2 as the main switch for coarse and fine regeneration gas, and programmable valves A3 and A4 as shut-off valves during hot blowing regeneration. The upstream of the calibrated programmable valves A1 and A2 are respectively added to the designated heater inlets E1 and E2, branch pipes PG1 and PG2; Branch lines PG3 and PG4 are added to the outlets of heaters E1 and E2, and are respectively connected to the downstream of programmable valves A3 and A4 on the bottom horizontal pipe of the valves with designated serial numbers.

3. The method for preventing coke oven gas from being incorporated into the TSA adsorption system according to claim 1, characterized in that, The 24-hour "4-3-1 Operating Condition Prediction Timetable" includes the coke oven number and switching type; When the system time reaches T_adv minutes before any operating condition switching time, the "active defense process" is automatically triggered, and the coke oven is marked as a "pre-action coke oven", and the trigger signal and the pre-action coke oven ID are output.

4. The method for preventing coke oven gas from being incorporated into the TSA adsorption system according to claim 1, characterized in that, The analysis process in step three is as follows: Based on the CAD drawing of 4-3-1, a digital twin sub-model of coke oven-pipeline network is constructed. The initial boundary conditions of the digital twin sub-model of coke oven-pipeline network are set to simulate the complete 4-3-1 switching process and output continuous time-series simulation curves: instantaneous flow rate of main pipe F(t), pipeline pressure P(t), and gas temperature T(t). Calculate key characteristic values: maximum instantaneous flow rate change rate (ΔF / Δt), maximum flow / pressure overshoot, and fluctuation duration. Combined with a dynamic analysis model for impurity precipitation that has been pre-trained offline using historical data, output the peak instantaneous concentration of each impurity, the cumulative precipitation of impurities over the entire cycle, and generate a simulated impurity concentration impact curve that changes over time.

5. The method for preventing coke oven gas from being incorporated into the TSA adsorption system according to claim 1, characterized in that, The analysis process of the hierarchical control strategy is as follows: The simulation was performed based on the first-principles dynamic mechanism model of the TSA adsorption tower. During the simulation, the concentration of each key impurity Ciout(t) in the gas phase at the outlet of the adsorption tower was continuously monitored, where i represents the key impurity and i is a natural number greater than zero. The Ciout(t) is compared with the preset purified gas qualification threshold in real time. The precise moment when Ciout(t) ≥ the preset purified gas qualification threshold for the first time is recorded. The earliest time when penetration occurs among all key impurities is taken as the predicted impurity penetration time Tbt in this evaluation. Obtain the remaining adsorption time Tsxc of each online adsorption tower and calculate the safety margin curve as a function of time: Tmar(t) = remaining breakthrough time (t) - Tsxc, where the remaining breakthrough time (t) is: if the impact terminates at time t, the dynamic prediction value of the remaining breakthrough time calculated from time t. The minimum value of this curve is found during the entire fluctuation period, which is the minimum safety margin Tamin. Calculate the flow impact intensity Lflow, the impurity concentration impact intensity Imp, and the comprehensive stress index Syl. Combined with the set logical judgment conditions, output the safety level / controllable level / severe level / critical level.

6. The method for preventing coke oven gas from being incorporated into the TSA adsorption system according to claim 5, characterized in that, Based on a pre-built level-control strategy library, the hierarchical control strategy corresponding to the current level is output. The graded control strategy includes adsorption timing adjustment commands, micro-purification module switching commands, regeneration gas flow rate adjustment values, pressure equalization duration parameters, and action execution time nodes.

7. The method for preventing coke oven gas from being incorporated into the TSA adsorption system according to claim 6, characterized in that, The analysis process for the stabilization or alarm commands is as follows: Based on the hierarchical control strategy corresponding to the current level, it is broken down into a sequence of digital and analog timing action instructions that can be recognized by the PLC, and the execution pre-time of each instruction is marked. Before the planned switchover time for the coke oven 4-3-1 operating condition arrives, the entire instruction sequence is sent to the TSA local control unit PLC. The PLC executes the instructions sequentially according to the time nodes. After the 4-3-1 operating condition switchover event is completed, curves of the actual flow rate, temperature, pressure, and online impurity concentration data are plotted. The difference between the actual impact curve on site and the simulated dynamic curve in the digital twin simulation is calculated step by step to obtain the global average deviation and peak deviation. If the global average deviation or peak deviation exceeds the system's preset deviation threshold, a calibration command is triggered. The cumulative time during which all indicators of the purified gas are qualified within the evaluation window is calculated. The evaluation window represents the time from the start of the operating condition switch to the end of the system's determination that the fluctuation has ended and the TSA has returned to steady state. The overall pass rate of purified gas is calculated based on the cumulative duration / evaluation window duration. The impact attenuation rate is then calculated as follows: Impact attenuation rate = (1 - actual maximum fluctuation amplitude / assumed maximum fluctuation amplitude without defense) × 100%. The active defense evaluation coefficient is calculated based on the comprehensive pass rate of purified gas × preset weight coefficient 1 + impact attenuation rate × preset weight coefficient 2. Combined with the threshold of the active defense evaluation coefficient, a set command or alarm command is output.

Images