Coal tar component separation and directional upgrading method and system

Abstract

The invention discloses a coal tar component separation and directional upgrading method and system. The method comprises the steps that raw material coal tar and all logistics detection data are collected online, and component fingerprint vectors are constructed; an emulsification state quantity is calculated according to a raw material component fingerprint vector, a pretreatment separation working condition is subjected to closed-loop adjustment, and oil-water-solid collaborative removal and stable breaking of an emulsification system are realized; inputting the fraction component fingerprint vector into a quality prediction model in fractional distillation, generating a cutting point correction amount according to a quality prediction deviation, and adjusting a side-draw temperature set value, a side-draw amount set value and a reflux ratio set value in a linkage manner to stabilize a cutting boundary; according to a fingerprint threshold set and a path selection matrix, each fraction is automatically routed to enter at least two upgrading branches for treatment, upgrading discharged materials are retested, substandard material flows trigger a backflow bypass for retreatment, the backflow material flows must be subjected to recovery cutting and redistribution nodes, and a re-routing closed loop is executed; and high-consistency coal tar fractions and controllable-mass heavy residues which can be stably supplied are output.

Description

Technical Field

[0001] This invention relates to the field of coal chemical and coking by-product deep processing and process control technology, specifically to a method and system for coal tar component separation and targeted upgrading. Background Technology

[0002] Coal tar is one of the main byproducts of the coking process. Its components include light aromatics, phenols, naphthalenes, polycyclic aromatic hydrocarbons, and asphaltenes, and it commonly contains free water, emulsified water, fine slag, and salts. It also contains sulfur-containing components, nitrogen-containing components, oxygen-containing active components, and trace metals. In industrial settings, variations in the type and blending of coal from which coal tar originates, fluctuations in coke oven temperature and condensation conditions, and differences in residence time in the primary separator and storage tanks can all lead to significant batch-to-batch fluctuations in coal tar, typically manifested as follows:

[0003] (1) Differences in water content and emulsion structure: Changes in the proportion of free water, the size of emulsion droplets, the strength of the interfacial film and the distribution of salt cause significant fluctuations in the difficulty of demulsification and the separation efficiency;

[0004] (2) Solid entrainment and salt migration: Fine slag and ash entering the subsequent tower system can easily cause heat exchanger scaling, tower tray blockage and corrosion risk, and salt migration can cause local corrosion and coking precursor accumulation.

[0005] (3) Boiling range distribution drift and fraction mixing: When fractional distillation relies on empirical cut points, the boiling range drift of the feedstock will lead to the drift of the side stream composition. Light fractions carry heavy components or medium-heavy fractions carry light components, resulting in quality dispersion and instability in downstream processing.

[0006] (4) Sudden changes in the loading and insufficiency of stability: When the impurity spectrum is dispersed and the loading changes abruptly, the adsorption bed breakthrough is advanced, the extraction partition coefficient fluctuates, or the hydrogenation stabilization load is abnormal, which can easily lead to the phenomenon of "the same fraction index fluctuating", making it difficult to guarantee the consistency of storage, transportation and subsequent deep processing.

[0007] Therefore, the single core technical problem that urgently needs to be solved in this field is: under the conditions of fluctuating coal tar feedstock and dispersed impurity spectrum, how to establish a closed-loop control system across pretreatment separation, fractional distillation, directional upgrading and reflux bypass reprocessing, so as to stabilize the cutting boundary, execute the upgrading path and force closed-loop reprocessing of substandard streams, thereby outputting target fractions with consistent quality in the long term while taking into account yield and energy consumption. Summary of the Invention

[0008] Technical Objective: To address the shortcomings of existing technologies, this invention discloses a method and system for the separation and targeted upgrading of coal tar components. By establishing a quality closed-loop control mechanism that integrates pretreatment separation, staged distillation, targeted upgrading, and reflux bypass reprocessing, the impurity content of the target fraction meets the upper limit constraint, the yield of the target fraction meets the lower limit constraint, and the energy consumption meets the upper limit constraint, thereby outputting a highly consistent coal tar fraction with stable feed and controllable quality heavy residue.

[0009] Technical solution: To achieve the above technical objectives, the present invention adopts the following technical solution:

[0010] A method for separating and directional upgrading coal tar components specifically includes the following steps:

[0011] S1. Online acquisition of detection data for raw coal tar and various materials in the separation process; construction of raw material component fingerprint vectors based on the detection data. and distillation group fingerprint vector , , ,in Corresponding to light fractions, Corresponding to the middle fraction, Corresponding heavy fraction;

[0012] S2, Fingerprint vector of raw material components Calculate the emulsion state quantity E, and use the emulsion state quantity E as the control target to adjust the pretreatment separation conditions in a closed loop so that the pretreatment output meets the preset emulsion state constraints, thereby completing the synergistic removal of oil, water and solids and the stable breakdown of the emulsion system.

[0013] S3. The pretreated output from step S2 is fed into a fractional distillation unit to obtain light fraction, middle fraction, heavy fraction and heavy residue.

[0014] S4. Divide the distillation group into fingerprint vectors. , , Input a quality prediction model, output a corresponding quality prediction index vector. , , and based on , , With the corresponding target quality index vector , , The deviation generates the cutting point correction amount The side stream extraction temperature setpoint, side stream extraction rate setpoint, and reflux ratio setpoint of the fractionation distillation unit are adjusted in a coordinated manner to achieve online stability of the fractionation boundary.

[0015] S5, Based on fingerprint threshold set Automatic routing is performed on the light, middle, and heavy fractions using the path selection matrix R, assigning the light, middle, and heavy fractions to at least two of the adsorption-upgrading, extraction-upgrading, and hydrogenation-stabilization-upgrading branches to remove sulfur-containing components, nitrogen-containing components, oxygen-containing active components, and metal impurities.

[0016] S6. Perform online retesting of the material output from each upgrading branch and construct the upgraded fingerprint vector. , , When the retest results do not meet the corresponding target quality index vector, the reflux bypass reprocessing is triggered, and the substandard material is returned to the fractional distillation unit and enters the recycling, cutting and redistribution process, or bypassed to the corresponding quality improvement branch to form a quality closed loop.

[0017] S7. The controller performs constrained rolling optimization, jointly adjusting the pretreatment separation conditions, reflux ratio setpoint, side stream extraction setpoint, and reflux bypass flow rate to simultaneously satisfy the upper limit constraint of impurity content, the lower limit constraint of target fraction yield, and the upper limit constraint of energy consumption.

[0018] In one embodiment, the raw material component fingerprint vector and the distillation group fingerprint vector , , This includes light and heavy fingerprint components, impurity spectrum fingerprint components, and stability fingerprint components; the light and heavy fingerprint components include... , , Three boiling range characteristic quantities; impurity spectrum fingerprint components including sulfur fingerprint components. Nitrogen fingerprint components Oxygen fingerprint components and the proportion of metallic fingerprints Stability fingerprint components include aggregation tendency fingerprint components. and residual emulsified fingerprint components .

[0019] In one embodiment, the emulsification state quantity E is calculated from the online water content W, the online electrical conductivity K, and the online water droplet size characteristic quantity D; the pretreatment separation conditions in step S2 with closed-loop regulation include the amount of demulsifier added, the coalescence electric field strength, and the rotation speed of the solid-liquid separation device; the preset emulsification state constraint includes the upper limit of the water content of the pretreatment output and the upper limit of the solid content of the pretreatment output.

[0020] In one embodiment, the pretreatment separation process in step S2 includes a mixing demulsification process, a three-phase separation process, and a fine solids removal process, wherein the fine solids removal is accomplished by hydrocyclone separation or filtration.

[0021] In one implementation, the cutting point correction amount in step S4 It simultaneously affects the side-stream extraction temperature setpoint, the side-stream extraction rate setpoint, and the reflux ratio setpoint, and Constrained by a preset correction range, the quality prediction model establishes a quality prediction mapping based on the distillation fraction fingerprint vector and the operational variables of fractional distillation, to output a quality prediction index vector. , , .

[0022] In one implementation, in step S5, the path selection matrix R outputs a unique quality-improvement branch routing instruction for each fraction within each control cycle, and satisfies the routing interlock condition: the routing vector for the i-th fraction... The amount satisfy and Where K is the number of quality improvement branches, and k is the index of the quality improvement branch; the available indicator for the k-th quality improvement branch. With load permit indication satisfy When any upgrading branch does not satisfy the requirements for the fraction When the distillate is in this state, the controller directs it to the reflux bypass reprocessing channel in step S6 and prevents it from bypassing the reflux bypass reprocessing channel into the product tank.

[0023] A coal tar component separation and targeted upgrading system is provided to implement the coal tar component separation and targeted upgrading method described above. The system includes an online detection device, a pretreatment separation module, a staged distillation module, a targeted upgrading module, a reflux bypass module, and a controller, wherein:

[0024] The online detection device is used to collect detection data of raw coal tar and various logistics and output it to the controller;

[0025] The pretreatment separation module is used to complete the synergistic removal of oil, water, and solids and the stable disintegration of the emulsion system under the regulation of the controller, and output the pretreated material.

[0026] The fractional distillation module is connected to the pretreatment separation module to divide the pretreatment output into light fraction, middle fraction, heavy fraction and heavy residue, and the fractional distillation module is equipped with at least two side stream sampling loops.

[0027] The targeted upgrading module is connected to the fractional distillation module and includes at least two upgrading branches from the following: adsorption upgrading branch, extraction upgrading branch, and hydrogenation stabilization upgrading branch.

[0028] The reflux bypass module is used to return substandard materials to the fractional distillation module and enter the recycling, cutting and redistribution process under the controller command, or to bypass to the corresponding quality improvement branch to form a quality closed loop;

[0029] The controller is used to construct component fingerprint vectors and calculate emulsion state quantities, generate cut point correction quantities and adjust side-line sampling setpoints and reflux ratio setpoints in conjunction, execute path selection matrix to realize quality improvement branch routing and execute route interlocking, execute retest to trigger reflux bypass reprocessing, and execute constrained rolling optimization to collaboratively meet the upper limit of impurity content, the lower limit of target fraction yield, and the upper limit of energy consumption.

[0030] In one embodiment, the online detection device includes a function for outputting three boiling range characteristics. , , Boiling range characteristic detection unit, used to output sulfur fingerprint component Compared with nitrogen fingerprint components The online component detection unit and the online physical property detection unit for outputting online water content W, online conductivity K, and online water droplet size characteristic D.

[0031] In one embodiment, the fractional distillation module is equipped with a recovery, cutting, and redistribution node. The return pipeline of the return bypass module is connected to the recycling, cutting, and redistribution node. The connection ensures that substandard return logistics must pass through the aforementioned recycling, cutting, and redistribution nodes. Then it enters the recovery, cutting and redistribution process of the fractional distillation module; the controller reconstructs the corresponding distillate fraction fingerprint vectors for the light, middle and heavy fractions output by the recovery, cutting and redistribution process and re-executes the path selection matrix calculation to output the rerouting instruction, thereby forming a rerouting closed loop.

[0032] In one embodiment, the controller includes a quality prediction unit, a cut-point correction unit, a routing decision unit, a retest determination unit, a reflux bypass reprocessing trigger unit, and a rolling optimization unit. The rolling optimization unit uses a weighted combination of quality deviation, energy consumption, and yield loss as the objective function, and uses upper limit constraints on impurity content, lower limit constraints on target fraction yield, and upper limit constraints on energy consumption as the constraint set, and outputs the linkage setting values ​​of the pretreatment separation module, the fractional distillation module, and the reflux bypass module.

[0033] Beneficial effects: The coal tar component separation and targeted upgrading method and system provided by the present invention have the following beneficial effects:

[0034] 1. This invention utilizes raw material component fingerprint vectors Closed-loop control of the emulsion state quantity E couples the key execution quantities of pretreatment separation with the constraints of the upper limits of water content and solid content in the pretreatment effluent, thereby stabilizing the pretreatment effluent boundary and reducing the disturbance of emulsion carryover and solid entrainment on distillation load and heat transfer from the source.

[0035] 2. This invention drives the cutting point correction amount by the deviation between the quality prediction index vector and the target quality index vector, and adjusts the side stream extraction temperature setting value, side stream extraction amount setting value and reflux ratio setting value in conjunction with the correction range. Under the constraint of the correction range, the cutting boundary is stabilized, and the impurity excess and quality dispersion caused by fraction mixing are suppressed.

[0036] 3. This invention ensures that the quality improvement route is executable and that substandard logistics will not bypass the reprocessing closed loop by using route uniqueness and interlocking conditions, the necessary recycling, cutting and redistribution nodes for return logistics and the rerouting closed loop, as well as the constraint-bound rolling optimization output linkage setting value. At the same time, it achieves cross-module collaborative control under three types of constraints: upper limit of impurity content, lower limit of yield and upper limit of energy consumption, thereby outputting highly consistent distillate and controllable quality heavy residuals. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0038] Figure 1 This is a block diagram of the system structure of the present invention;

[0039] Figure 2 This is a schematic diagram of the method flow of the present invention;

[0040] Figure 3 This is a functional structure diagram of the controller of the present invention;

[0041] Figure 4 A schematic diagram of the stability curves for key quality indicators;

[0042] Figure 5 This is a schematic diagram illustrating the trade-off between energy consumption and yield loss. Detailed Implementation

[0043] The present invention will now be described more clearly and completely by way of a preferred embodiment in conjunction with the accompanying drawings, but this does not limit the invention to the scope of the described embodiment.

[0044] like Figure 1The present invention provides a coal tar component separation and targeted upgrading system, comprising: an online detection device, a pretreatment separation module, a staged distillation module, a targeted upgrading module, a reflux bypass module, and a controller. The online detection device is signal-connected to the controller; the outlet of the pretreatment separation module is connected to the inlet stream of the staged distillation module; the light fraction outlet, middle fraction outlet, and heavy fraction outlet of the staged distillation module are respectively connected to the inlet valve group of the targeted upgrading module; the outlet of the targeted upgrading module is connected to the reflux bypass module; the reflux pipeline of the reflux bypass module is connected to the staged distillation module, and a recovery, cutting, and redistribution node is provided. This ensures that substandard return logistics undergoes a recycling, cutting, and redistribution process before entering a closed loop of graded distillation and rerouting.

[0045] To achieve cross-module closed-loop operation, this invention uses component fingerprint vectors as a unified state representation. The controller constructs component fingerprint vectors for the raw materials entering the system. Construct fractional fingerprint vectors for the light, middle, and heavy fractions output from fractional distillation. , , .in , , , The vector structure remains consistent, which is used for comparison, threshold determination and routing decisions between different logistics.

[0046] The online monitoring device is used to provide three types of data streams during the control cycle:

[0047] Lightness and heaviness related data stream: Output boiling range characteristic quantity , , The unit is Celsius;

[0048] Impurity spectrum correlation data stream: Output used to construct sulfur fingerprint components Nitrogen fingerprint components Oxygen fingerprint components Metal fingerprint components The detection quantity, the above fingerprint components are dimensionless quantities;

[0049] Emulsification-related data stream: Outputs online water content W (mass percentage), online conductivity K (millisiemens per centimeter), and online water droplet size characteristic D (micrometers).

[0050] The controller performs time alignment and quality verification on the above data:

[0051] Time alignment: Aligning the timestamps of different detection units to the same control cycle;

[0052] Anomaly rejection: Sampling points that are out of range, distorted, or abrupt are marked as invalid and trigger a retest request from the retest determination unit;

[0053] Data retention: when a short period of invalidity occurs, the previous valid value is retained for key variables, and a degraded state is marked inside the controller. The magnitude of cut-off point correction and the frequency of routing switching are limited to avoid instability caused by erroneous data.

[0054] The pretreatment separation module, in order of material flow direction, includes a mixing and demulsification unit, a coalescence separation unit, a three-phase separation unit, and a fine solids removal unit.

[0055] The mixing demulsification unit includes a demulsifier dispensing device and a mixer; the controller outputs a demulsifier dispensing amount set value and drives the dispensing device to reduce the interfacial film strength and weaken the emulsion stability.

[0056] The coalescence separation unit includes a coalescence device; the controller outputs a coalescence intensity setpoint and drives the coalescence device to coalesce the fine water droplets into separable water droplets.

[0057] The three-phase separation unit includes a three-phase separator; it performs preliminary separation of the oil phase, water phase, and solid phase.

[0058] Fine solids removal unit, including cyclone separator or filter; controller outputs corresponding execution quantity setpoint to reduce fine slag entrainment.

[0059] To ensure the pretreatment closed loop has a clearly calculable state quantity, this invention introduces the emulsification state quantity E as the control objective of the pretreatment closed loop. E is a dimensionless quantity used to comprehensively characterize the effects of online water content W, online conductivity K, and online water droplet size characteristic D on emulsification stability, where W is in mass percentage, K is in millisiemens per centimeter, and D is in micrometers. The controller converts W, K, and D into dimensionless quantities respectively. , , Then, calculate using the following formula:

[0060]

[0061] in , , The weights are dimensionless and satisfy the following conditions: .

[0062] The controller uses E as the closed-loop target and the upper limit of the pretreated effluent water content as the target. (mass percentage) and upper limit of solid content in pretreated output Using (mg / kg) as the operating boundary, the set values ​​for demulsifier dosing, coalescence strength, and fine solids removal are adjusted in a coordinated manner to stabilize the pretreatment discharge boundary and reduce the load fluctuation and clogging risk of the subsequent fractional distillation module.

[0063] The fractional distillation module includes a main distillation column, a reboiler section, a condenser section, a reflux section, and at least two side-stream collection loops. Each side-stream collection loop includes a side-stream collection valve, a side-stream heat exchanger, and a side-stream reflux valve.

[0064] This invention does not use a single temperature point as the basis for cutting control, but instead uses "quality prediction deviation-driven linkage correction" to stabilize the cutting boundary:

[0065] The controller is designed for light, middle, and heavy fractions respectively. , , The quality prediction unit outputs a quality prediction index vector. , , .in This is the quality prediction index vector corresponding to the i-th fraction. .

[0066] To achieve controllable objectives, a corresponding target quality index vector is set. , , The superscript * indicates the target value.

[0067] Cutting point correction unit according to and The deviation generates the cutting point correction amount ,in This is a combined correction factor that includes side-stream extraction temperature correction, side-stream extraction volume correction, and reflux ratio correction. The controller will... The mapping is used to update the side-stream extraction temperature setpoint, side-stream extraction rate setpoint, and reflux ratio setpoint, and to apply device safety boundaries and change rate boundaries to prevent tower system oscillation caused by over-adjustment.

[0068] The above design ensures the stability of the cutting boundary from two aspects:

[0069] 1) Through The prediction maps boiling range drift plus impurity spectrum drift into a controllable deviation;

[0070] 2) Through The three-variable linkage correction simultaneously constrains the cutting position and the reflux steady state, reducing the mutual cancellation caused by adjusting only one manipulation amount.

[0071] The targeted upgrading module includes at least two upgrading branches, and routing is achieved through branch switching valve assemblies. To cover differences in impurity profiles, the branch types are defined as: adsorption upgrading branch, extraction upgrading branch, and hydrogenation stabilization upgrading branch; each branch is equipped with an inlet valve, key execution quantities, and an outlet detection point for closed-loop determination by the controller.

[0072] The routing decision unit uses a fingerprint threshold set. Output routing instructions with the path selection matrix R:

[0073] fingerprint threshold set Used to The key components are mapped to the main risk types of that fraction;

[0074] The routing matrix R is used to map risk types and branch capabilities to specific branch numbers.

[0075] In one specific embodiment, two types of engineering constraints are explicitly given:

[0076] 1) Routing uniqueness: For the same fraction within the same control cycle, only one upgrading branch is allowed. Before issuing the branch valve group instruction, the controller performs a uniqueness check on the routing result of the fraction. If the check fails, the instruction is rejected and the process enters the reprocessing process.

[0077] 2) Interlocking condition: Available indication quantity of controller maintenance branch Branch load permissible indication Both values ​​are 0 or 1; when the selected branch is not available and the load is permissible, the controller will not issue the routing instruction, but will instead trigger the return bypass module and close the valve position of the fraction leading to the product tank, ensuring that substandard logistics does not bypass the closed-loop link to enter the product.

[0078] The reflux bypass module includes a reflux line, a bypass line, and a valve assembly, and a recovery, cutting, and redistribution node is installed on the reflux line. . Its function is not simply to merge the flow, but to send the reflux material into a defined entry point of the recycling, cutting and redistribution process, so that the reflux material must go through a fixed redistribution logic before entering the fractional distillation module, thus avoiding the failure of the closed loop caused by the drift of the reflux point.

[0079] When the retest judgment unit outputs a non-compliance judgment signal, the return bypass reprocessing trigger unit executes the following action chain:

[0080] 1) Close the valve position that leads the material to the product tank;

[0081] 2) Open the reflux valve to allow substandard materials to enter the reflux pipeline;

[0082] 3) Ensure the return logistics enter ;

[0083] 4) The refluxed material is fed into the recycling, cutting and redistribution process, and the redistributed light, middle and heavy fractions are output.

[0084] 5) The controller reconstructs the redistribution output logistics. , , Execute again The system receives a rerouting instruction from R, thus achieving a closed loop of backflow—recycling, cutting, redistribution, and rerouting.

[0085] When the interlock conditions allow and the branch has the ability to reprocess, the controller can also open the bypass valve to allow substandard logistics to enter the designated quality improvement branch; the issuance of this bypass command is also subject to the uniqueness of the route and the interlock conditions to avoid the bypass causing an uncontrollable cycle.

[0086] like Figure 3 As shown, the controller can be implemented collaboratively by an industrial control computer and a DCS / PLC: the industrial control computer is responsible for quality prediction, routing decisions, and rolling optimization solutions; the DCS / PLC is responsible for real-time closed-loop control of actuators such as valve positions, pumps, and coalescing devices. The controller internally includes at least the following implementation mechanisms:

[0087] 1) Data Management: Perform time alignment, quality verification, invalidation marking, and valid value preservation on online detection data;

[0088] 2) Quality prediction: based on , , Output , , and with , , Comparison leads to bias;

[0089] 3) Cutting correction: Mapping the deviation to And update the settings based on the device boundary and change rate boundary limits;

[0090] 4) Routing decision: based on It outputs routing commands to R and performs route uniqueness and interlocking;

[0091] 5) Retest Judgment: Retest the upgraded output and output a compliance judgment signal. If the retest fails or the data is invalid, output "non-compliant or unknown" judgment and trigger the safety closed loop.

[0092] 6) Rolling optimization: The weighted combination of quality deviation, energy consumption and yield loss is used as the objective, and the upper limit of impurity content, the lower limit of yield and the upper limit of energy consumption are used as the constraint set. The output is the linkage setting value of the pretreatment separation module, the fractional distillation module and the reflux bypass module. If the solution result of rolling optimization conflicts with the interlock, the interlock priority is higher than the optimization output to ensure executability and safety boundaries.

[0093] like Figure 2 As shown, this invention also provides a method for separating and directional upgrading coal tar components, which is executed cyclically by a controller within a control cycle. The control cycle is set based on the dynamic response of the device, and the controller completes data acquisition, state estimation, setpoint update, route interlocking, retest determination, and backflow bypass reprocessing closed loop in cycles. Specifically, it includes the following steps:

[0094] S1. Online acquisition of detection data for raw coal tar and various materials in the separation process; construction of raw material component fingerprint vectors based on the detection data. and distillation group fingerprint vector , , ,in Corresponding to light fractions, Corresponding to the middle fraction, Corresponding to heavy fraction.

[0095] In this embodiment, the fingerprint vector is used to characterize weight, impurity profile, and stability information in a unified manner. The controller employs a consistent fingerprint structure for each stream to ensure comparability between different streams and to facilitate threshold determination and routing decisions. The detection data includes at least: boiling range characteristics. , , (Unit: degrees Celsius); Impurity spectrum correlation detection quantities are used to construct sulfur fingerprint components. Nitrogen fingerprint components Oxygen fingerprint components Metal fingerprint component (The fingerprint components mentioned above are dimensionless); emulsification-related detection quantities include online water content W (in mass percentage), online conductivity K (in millisiemens per centimeter), and online water droplet size characteristic D (in micrometers). The controller generates the data after time alignment and range calibration of the collected data. , , , .

[0096] S2, Fingerprint vector of raw material components The emulsion state quantity E is calculated, and the pretreatment separation conditions are adjusted in a closed loop with the emulsion state quantity E as the control target so that the pretreatment output meets the preset emulsion state constraints, thereby completing the synergistic removal of oil, water and solids and the stable breakdown of the emulsion system.

[0097] In this embodiment, the emulsification state quantity E is a dimensionless quantity used to comprehensively characterize the influence of water content, salt migration characteristics, and droplet size on emulsification stability. The controller maps the online water content W, online conductivity K, and online droplet size characteristic quantity D into dimensionless quantities. , , Then, calculate the emulsified state quantity using the following formula:

[0098]

[0099] in , , The weights are dimensionless and satisfy the following conditions: .

[0100] Preset emulsification state constraints It means that among them This represents the upper limit threshold of the emulsified state quantity, which is a dimensionless quantity. The controller uses E and... The deviation is the feedback quantity, which is used to adjust the execution quantity settings of the pretreatment separation condition, including the demulsifier injection quantity setting, the coalescence strength setting, and the solid-liquid separation execution quantity setting, so that the emulsification state of the pretreatment output is controlled and the load fluctuation of subsequent staged distillation is reduced.

[0101] S3. The pretreated output from step S2 is fed into a fractional distillation unit to obtain light fraction, middle fraction, heavy fraction and heavy residue.

[0102] In this embodiment, the light fraction, middle fraction, and heavy fraction correspond to the components subsequently constructed. , , Heavy residues are output as heavy end product streams and included in energy consumption and yield accounting.

[0103] S4. Divide the distillation group into fingerprint vectors. , , Input a quality prediction model, output a corresponding quality prediction index vector. , , and based on , , With the corresponding target quality index vector , , The deviation generates the cutting point correction amount The side stream extraction temperature setpoint, side stream extraction rate setpoint, and reflux ratio setpoint of the fractionation distillation unit are adjusted in a coordinated manner to achieve online stability of the fractionation boundary.

[0104] In this embodiment, Let be the quality prediction index vector for the i-th fraction. For the corresponding target quality index vector, the superscript * indicates the target value. Cut-off point correction amount. The linkage correction amount includes at least the update components of the side-stream extraction temperature setpoint, the side-stream extraction rate setpoint, and the reflux ratio setpoint. The controller synchronously updates the above setpoints based on the quality deviation and applies boundary constraints and change rate constraints to the update amplitude to ensure stable operation of the distillation unit and suppress boundary oscillations caused by over-adjustment.

[0105] S5, Based on fingerprint threshold set Automatic routing is performed on the light, middle, and heavy fractions using the path selection matrix R, assigning them to at least two of the following upgrading branches: adsorption upgrading branch, extraction upgrading branch, and hydrogenation stabilization upgrading branch, to remove sulfur-containing components, nitrogen-containing components, oxygen-containing active components, and metal impurities.

[0106] In this embodiment, the fingerprint threshold set Used for , , The key fingerprint components are used for risk classification. The path selection matrix R is used to map the risk classification to the quality improvement branch number and output the branch valve group command. The adsorption quality improvement branch is used to reduce the load of polar impurities and metal-related impurities; the extraction quality improvement branch is used to reduce the load of oxygen-containing active impurities; and the hydrogenation stabilization quality improvement branch is used to reduce sulfur-containing and nitrogen-containing active components and improve storage and transportation stability. When issuing routing commands, the controller performs route executability verification and uses the verification results for subsequent retesting and reprocessing triggering.

[0107] The controller outputs a unique branch routing instruction for each fraction and issues the route when the branch availability indicator and the load allowance indicator meet; otherwise, it prohibits entry into the product path and switches to the reflux bypass for reprocessing.

[0108] S6. Perform online retesting of the material output from each upgrading branch and construct the upgraded fingerprint vector. , , When the retest results do not meet the corresponding target quality index vector, the reflux bypass reprocessing is triggered, and the substandard material is returned to the fractional distillation unit and enters the recycling, cutting and redistribution process, or bypassed to the corresponding quality improvement branch to form a quality closed loop.

[0109] In this embodiment, Let be the fingerprint vector of the i-th fraction after upgrading, with the superscript ' indicating the upgraded state. The retest determination uses and A consistent pass / fail assessment signal is output. If the product fails to meet the standard, the controller closes the product destination valve and triggers the backflow bypass reprocessing path.

[0110] Reflux reprocessing path: Substandard streams are refluxed to the recycling, cutting and redistribution process of the fractional distillation unit. The recycling, cutting and redistribution process re-outputs light, middle and heavy fractions and re-enters the cutting correction and routing decision in steps S4 and S5.

[0111] Bypass reprocessing path: Substandard logistics are bypassed to the corresponding quality improvement branch for reprocessing, and after retesting, it is decided whether to enter the product path or the return reprocessing path.

[0112] The controller ensures that the return logistics must pass through the recycling, cutting, and redistribution nodes. It then enters the recycling, cutting, and redistribution process, and is rebuilt after the redistribution output. , , Perform rerouting to form a closed loop of backflow—reclaiming, cutting, redistribution—rerouting.

[0113] S7. The controller performs constrained rolling optimization, jointly adjusting the pretreatment separation conditions, reflux ratio setpoint, side stream extraction setpoint, and reflux bypass flow rate to simultaneously satisfy the upper limit constraint of impurity content, the lower limit constraint of target fraction yield, and the upper limit constraint of energy consumption.

[0114] In this implementation, rolling optimization updates on a control cycle basis. Within each rolling window, it calculates the linkage setpoints and sends them to the pretreatment separation unit, the fractional distillation unit, and the reflux bypass reprocessing path. The constraint set includes: the impurity content of each target fraction does not exceed the upper limit threshold set; the yield of each target fraction is not lower than the lower limit threshold set; and the energy consumption per unit product does not exceed the upper limit threshold set. The controller combines the rolling optimization output with the unit boundary constraints and routing interlock constraints, and then sends it for execution to ensure that the setpoints are executable and the operating boundaries are controlled.

[0115] like Figure 5As shown, under the constraint-bound rolling optimization simulation, the relative value of energy consumption and yield loss are taken as the two core indicators for optimization trade-offs: when the operating point violates the constraints corresponding to the upper limit of energy consumption or the lower limit of yield, its solution falls into the infeasible region; within the feasible region that simultaneously satisfies the upper limit constraint of impurity content, the lower limit constraint of target fraction yield, and the upper limit constraint of energy consumption, the rolling optimization unit can form a set of non-dominant Pareto front operating points, and select a better operating point based on the weighted combination objective function of quality deviation, energy consumption, and yield loss, thereby realizing the linkage setpoint output of the pretreatment separation module, the fractional distillation module, and the reflux bypass module, so that the quality indicators, energy consumption, and yield are optimized synergistically within the constraint set.

[0116] like Figure 4 As shown, under the simulation conditions of disturbance, the impurity content is used as the key quality indicator for comparison: Under the fixed cutting strategy, disturbances in raw materials and operating conditions will cause the impurity content to rise significantly and fluctuate considerably, and it is easier to approach or exceed the upper limit of impurity content; When the closed-loop control strategy of this invention is adopted, the controller adjusts the pretreatment separation condition, the fractional distillation cutting boundary and the reflux bypass reprocessing strategy in a coordinated manner based on the fingerprint vector and the quality prediction results, so that the fluctuation amplitude of impurity content is significantly converged and more stably maintained within the target quality range, thereby improving the consistency of the target fraction quality and reducing the transmission of disturbances to the product side.

[0117] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for separating and directional upgrading coal tar components, characterized in that, Specifically, the following steps are included: S1. Online acquisition of detection data for raw coal tar and various materials in the separation process; construction of raw material component fingerprint vectors based on the detection data. and distillation group fingerprint vector , , ,in Corresponding to light fractions, Corresponding to the middle fraction, Corresponding heavy fraction; S2, Fingerprint vector of raw material components Calculate the emulsion state quantity E, and use the emulsion state quantity E as the control target to adjust the pretreatment separation conditions in a closed loop so that the pretreatment output meets the preset emulsion state constraints, thereby completing the synergistic removal of oil, water and solids and the stable breakdown of the emulsion system. S3. The pretreated output from step S2 is fed into a fractional distillation unit to obtain light fraction, middle fraction, heavy fraction and heavy residue. S4. Divide the distillation group into fingerprint vectors. , , Input a quality prediction model, output a corresponding quality prediction index vector. , , and based on , , With the corresponding target quality index vector , , The deviation generates the cutting point correction amount The side stream extraction temperature setpoint, side stream extraction rate setpoint, and reflux ratio setpoint of the fractionation distillation unit are adjusted in a coordinated manner to achieve online stability of the fractionation boundary. S5, Based on fingerprint threshold set Automatic routing is performed on the light, middle, and heavy fractions using the path selection matrix R, assigning the light, middle, and heavy fractions to at least two of the adsorption-upgrading, extraction-upgrading, and hydrogenation-stabilization-upgrading branches to remove sulfur-containing components, nitrogen-containing components, oxygen-containing active components, and metal impurities. S6. Perform online retesting of the material output from each upgrading branch and construct the upgraded fingerprint vector. , , When the retest results do not meet the corresponding target quality index vector, the reflux bypass reprocessing is triggered, and the substandard material is returned to the fractional distillation unit and enters the recycling, cutting and redistribution process, or bypassed to the corresponding quality improvement branch to form a quality closed loop. S7. The controller performs constrained rolling optimization, jointly adjusting the pretreatment separation conditions, reflux ratio setpoint, side stream extraction setpoint, and reflux bypass flow rate to simultaneously satisfy the upper limit constraint of impurity content, the lower limit constraint of target fraction yield, and the upper limit constraint of energy consumption.

2. The method for separating and directional upgrading coal tar components according to claim 1, characterized in that, The raw material component fingerprint vector and the distillation group fingerprint vector , , Including light and heavy fingerprint components, impurity spectrum fingerprint components, and stability fingerprint components; Light and heavy fingerprint components include , , Three boiling range characteristic quantities; impurity spectrum fingerprint components including sulfur fingerprint components. Nitrogen fingerprint components Oxygen fingerprint components and the proportion of metallic fingerprints ; Stable fingerprint components include fingerprint components with aggregation tendency. and residual emulsified fingerprint components .

3. The method for separating and directional upgrading coal tar components according to claim 1, characterized in that, The emulsification state quantity E is calculated from the online water content W, online conductivity K, and online water droplet size characteristic quantity D; the pretreatment separation conditions in step S2 with closed-loop regulation include the amount of demulsifier added, the coalescence electric field strength, and the rotation speed of the solid-liquid separation device; the preset emulsification state constraint includes the upper limit of the water content of the pretreatment output and the upper limit of the solid content of the pretreatment output.

4. The method for separating and directional upgrading coal tar components according to claim 1, characterized in that, The pretreatment separation process in step S2 includes a mixing and demulsification process, a three-phase separation process, and a fine solids removal process, wherein the fine solids removal is accomplished by hydrocyclone separation or filtration.

5. The method for separating and directional upgrading coal tar components according to claim 1, characterized in that, Cutting point correction amount in step S4 It simultaneously affects the side-stream extraction temperature setpoint, the side-stream extraction rate setpoint, and the reflux ratio setpoint, and Constrained by a preset correction range, the quality prediction model establishes a quality prediction mapping based on the distillation fraction fingerprint vector and the operational variables of fractional distillation, to output a quality prediction index vector. , , .

6. The method for separating and directional upgrading coal tar components according to claim 1, characterized in that, In step S5, the path selection matrix R outputs a unique quality-improvement branch routing instruction for each fraction within each control cycle, and satisfies the routing interlock condition: the routing vector for the i-th fraction... The amount satisfy and Where K is the number of quality improvement branches, and k is the index of the quality improvement branch; the available indicator for the k-th quality improvement branch. With load permit indication satisfy ; When any of the upgrading branches does not meet the requirements for the fraction When the distillate is in this state, the controller directs it to the reflux bypass reprocessing channel in step S6 and prevents it from bypassing the reflux bypass reprocessing channel into the product tank.

7. A coal tar component separation and targeted upgrading system, characterized in that, A method for separating and directionally upgrading coal tar components as described in any one of claims 1-6, comprising an online detection device, a pretreatment separation module, a staged distillation module, a directional upgrading module, a reflux bypass module, and a controller, wherein: The online detection device is used to collect detection data of raw coal tar and various logistics and output it to the controller; The pretreatment separation module is used to complete the synergistic removal of oil, water, and solids and the stable disintegration of the emulsion system under the regulation of the controller, and output the pretreated material. The fractional distillation module is connected to the pretreatment separation module to divide the pretreatment output into light fraction, middle fraction, heavy fraction and heavy residue, and the fractional distillation module is equipped with at least two side-stream sampling loops. The targeted upgrading module is connected to the fractional distillation module and includes at least two upgrading branches from the following: adsorption upgrading branch, extraction upgrading branch, and hydrogenation stabilization upgrading branch. The reflux bypass module is used to return substandard materials to the fractional distillation module and enter the recycling, cutting and redistribution process under the controller command, or to bypass to the corresponding quality improvement branch to form a quality closed loop; The controller is used to construct component fingerprint vectors and calculate emulsion state quantities, generate cut point correction quantities and adjust side-line sampling setpoints and reflux ratio setpoints in conjunction, execute path selection matrix to realize quality improvement branch routing and execute route interlocking, execute retest to trigger reflux bypass reprocessing, and execute constrained rolling optimization to collaboratively meet the upper limit of impurity content, the lower limit of target fraction yield, and the upper limit of energy consumption.

8. The coal tar component separation and targeted upgrading system according to claim 7, characterized in that, The online detection device includes components for outputting three boiling range characteristic quantities. , , Boiling range characteristic detection unit, used to output sulfur fingerprint component Compared with nitrogen fingerprint components The online component detection unit and the online physical property detection unit for outputting online water content W, online conductivity K, and online water droplet size characteristic D.

9. A coal tar component separation and targeted upgrading system according to claim 7, characterized in that, The fractional distillation module is equipped with a recycling, cutting, and redistribution node. The return pipeline of the return bypass module is connected to the recycling, cutting, and redistribution node. The connection ensures that substandard return logistics must pass through the aforementioned recycling, cutting, and redistribution nodes. It then enters the recovery, cutting, and redistribution process of the fractional distillation module; The controller reconstructs the corresponding distillate fraction fingerprint vectors for the light, middle, and heavy fractions output from the recycling, cutting, and redistribution process, and re-executes the path selection matrix calculation to output a rerouting instruction, thereby forming a rerouting closed loop.

10. A coal tar component separation and targeted upgrading system according to claim 7, characterized in that, The controller includes a quality prediction unit, a cut-point correction unit, a routing decision unit, a retest judgment unit, a reflux bypass reprocessing trigger unit, and a rolling optimization unit. The rolling optimization unit uses a weighted combination of quality deviation, energy consumption, and yield loss as the objective function, and uses upper limit constraints on impurity content, lower limit constraints on target fraction yield, and upper limit constraints on energy consumption as the constraint set, and outputs the linkage setting values ​​of the pretreatment separation module, the fractional distillation module, and the reflux bypass module.

Images