Energy-saving and consumption-reducing system and method for extracting 1, 3-butadiene

Abstract

The invention discloses an energy-saving and consumption-reducing system and method for extracting 1, 3-butadiene, and the method comprises the following steps: collecting key parameter data of each process link; adjusting solvent components in real time by using a solvent component analysis model and a solvent adjustment algorithm; using a substance flow direction control algorithm to optimize the flow direction of substances in the washing tower; optimizing the operation parameters of the gas stripping tower by using an equipment operation parameter optimization algorithm; key parameters of all process links are monitored and controlled in real time through the control module, it is ensured that the system operates in the optimal state, all-around intelligent control over the extraction process of 1, 3-butadiene is achieved, energy consumption and solvent loss are remarkably reduced, and production efficiency and product quality are improved.

Description

Technical Field

[0001] The present invention relates to the field of chemical industry, and particularly focuses on the extraction process of 1,3-butadiene, and aims to achieve high-efficiency energy saving and consumption reduction in the extraction process through innovative methods, thereby improving the economy and sustainability of the entire production process. Background Art

[0002] As a vital raw material in the chemical industry, 1,3-butadiene plays an indispensable role in many fields such as rubber and plastics. However, the traditional 1,3-butadiene extraction process faces a series of severe challenges.

[0003] The current process usually relies on a large amount of heat and electricity to maintain the operation of each extraction link. For example, in the process of solvent recovery, a large amount of steam is required for heating and evaporation to achieve the separation of solvent and product. According to statistics, the steam consumption in the solvent recovery link in the traditional process accounts for 40%-50% of the total energy consumption. This high energy consumption not only increases production costs, but also brings huge pressure on energy supply, which does not meet the urgent needs of today's society for energy conservation and emission reduction.

[0004] On the one hand, the traditional process lacks precise control methods for the use of solvents. Since it is impossible to accurately monitor the solvent composition in real time, excessive amounts of solvent are often required to ensure the extraction effect, which causes a large amount of solvent to be lost during the circulation process or to be discarded due to deterioration. On the other hand, during the material separation process, due to the lack of optimized separation strategies, some raw materials or intermediates that have not been fully reacted are treated as waste, resulting in a serious waste of raw material resources, while also increasing the cost and environmental burden of subsequent waste treatment.

[0005] In the absence of effective algorithms and precise mathematical models, it is difficult to accurately set and dynamically adjust equipment operating parameters. For example, key parameters such as temperature, pressure, and liquid level in the extraction tower fluctuate greatly, which not only affects the extraction efficiency, but also accelerates the wear and corrosion of the equipment. Frequent equipment failures lead to production interruptions, further reducing production efficiency, increasing maintenance costs, and seriously affecting the production benefits and market competitiveness of enterprises.

[0006] In summary, the existing technology has obvious deficiencies in key links such as solvent adjustment, material flow control and equipment operation parameter optimization. An innovative solution is urgently needed to overcome these problems and achieve efficient energy saving and consumption reduction in the 1,3-butadiene extraction process. Summary of the invention

[0007] The object of the present invention is to provide a 1,3-butadiene extraction energy-saving and consumption-reducing system, comprising:

[0008] A solvent adjustment module, used to adjust the solvent composition in real time according to the solvent composition analysis model and the solvent adjustment algorithm;

[0009] Scrubber module, used to optimize the material flow in the scrubber through the material flow control algorithm;

[0010] The gas stripping tower module is used to optimize the operating parameters of the gas stripping tower through the equipment operating parameter optimization algorithm;

[0011] The control module is used to monitor and control the key parameters of each process link in real time.

[0012] Furthermore, the solvent composition analysis model is based on a spectral analysis model of quantum chemical calculations, which analyzes the electronic structure and spectral characteristics of solvent molecules and combines machine learning methods to achieve accurate prediction of the content of various components in the solvent. The specific steps include:

[0013] Collecting spectral data of the solvent;

[0014] Using quantum chemical calculation methods, the relationship between the electronic structure and spectral characteristics of solvent molecules is established;

[0015] Use machine learning algorithms such as Support Vector Regression (SVR) to calibrate the model and improve prediction accuracy.

[0016] Furthermore, the solvent adjustment algorithm is based on a dynamic programming algorithm, and according to the target solvent content, the solvent composition is adjusted in real time taking into account the dynamic characteristics and constraints of the solvent adjustment module. The specific steps include:

[0017] Calculate the difference between the current solvent composition and the target composition;

[0018] Determine the amount of solvent to be added or removed based on the difference;

[0019] Considering the adjustment speed limit and reagent addition amount limit, adjust the parameters of the solvent adjustment module;

[0020] Dynamically adjust the adjustment speed and reagent addition amount to ensure that the solvent composition reaches the target value.

[0021] Furthermore, the material flow control algorithm is based on a mixed integer linear programming (MILP) algorithm, which improves the washing efficiency of 1,3-butadiene and reduces solvent loss by optimizing the material flow in the washing tower. The specific steps include:

[0022] Establish material balance and energy balance models in the scrubber;

[0023] Define the objective function, including material handling costs and environmental impact costs;

[0024] Set constraints, including material balance constraints, flow limits for feed and discharge ports;

[0025] The MILP algorithm is used to solve the optimal material flow plan.

[0026] Furthermore, the equipment operation parameter optimization algorithm is based on a genetic algorithm (GA), which reduces energy consumption and improves the separation efficiency of 1,3-butadiene by optimizing the operation parameters of the gas stripping tower. The specific steps include:

[0027] Initialize the population and generate a set of random operating parameters;

[0028] Calculate the fitness value for each set of parameters, including energy consumption and separation efficiency;

[0029] Select individuals with higher fitness to enter the next generation;

[0030] Perform crossover and mutation operations to generate new individuals;

[0031] Replace individuals with lower fitness and repeat the iteration until the optimal solution is reached.

[0032] Furthermore, the control module includes a data acquisition module, a data processing module, a control execution module and a feedback adjustment module, which are used to monitor and control the key parameters of each process link in real time to ensure that the system operates in an optimal state. The specific steps include:

[0033] Collect temperature, pressure, liquid level, flow and other data in real time through sensors;

[0034] Preprocess and analyze the collected data to extract key features;

[0035] Generate control instructions based on analysis results to adjust valve opening, pump speed, heating power, etc.;

[0036] By real-time monitoring of the system operating status, the control strategy is automatically adjusted to ensure stable system operation.

[0037] On the other hand, the present invention also provides a method for energy saving and consumption reduction of 1,3-butadiene extraction based on the system according to any one of claims 1 to 6, comprising the steps of:

[0038] Collect key parameter data of each process link;

[0039] Use solvent composition analysis model and solvent adjustment algorithm to adjust solvent composition in real time;

[0040] Use the material flow control algorithm to optimize the material flow in the scrubber;

[0041] Use equipment operating parameter optimization algorithm to optimize the operating parameters of the gas stripping tower;

[0042] The control module monitors and controls the key parameters of each process link in real time to ensure that the system operates in the optimal state.

[0043] Furthermore, the solvent composition analysis model predicts the content of various components in the solvent by combining spectral analysis technology with quantum chemical calculation and machine learning methods, and the specific steps include:

[0044] Collecting spectral data of the solvent;

[0045] Using quantum chemical calculation methods, the relationship between the electronic structure and spectral characteristics of solvent molecules is established;

[0046] Use machine learning algorithms such as Support Vector Regression (SVR) to calibrate the model and improve prediction accuracy.

[0047] Furthermore, the solvent adjustment algorithm adjusts the parameters of the solvent adjustment module according to the target solvent component content through a dynamic programming method to achieve precise control of the solvent component. The specific steps include:

[0048] Calculate the difference between the current solvent composition and the target composition;

[0049] Determine the amount of solvent to be added or removed based on the difference;

[0050] Considering the adjustment speed limit and reagent addition amount limit, adjust the parameters of the solvent adjustment module;

[0051] Dynamically adjust the adjustment speed and reagent addition amount to ensure that the solvent composition reaches the target value.

[0052] Furthermore, the material flow control algorithm optimizes the material flow in the washing tower by a mixed integer linear programming (MILP) method, improves the washing efficiency of 1,3-butadiene and reduces solvent loss, and the specific steps include:

[0053] Establish material balance and energy balance models in the scrubber;

[0054] Define the objective function, including material handling costs and environmental impact costs;

[0055] Set constraints, including material balance constraints, flow limits for feed and discharge ports;

[0056] The MILP algorithm is used to solve the optimal material flow plan.

[0057] Beneficial technical effects:

[0058] A 1,3-butadiene extraction energy-saving and consumption-reducing system and method is intended to achieve efficient energy saving and consumption reduction in the extraction process through innovative technical means, thereby improving the economy and sustainability of the entire production process.

[0059] The following are the main innovations of the system:

[0060] 1.Solvent composition analysis model

[0061] The present invention adopts a spectral analysis model based on quantum chemical calculations, which analyzes the electronic structure and spectral characteristics of solvent molecules and combines machine learning methods to achieve accurate prediction of the content of various components in the solvent. This model can not only monitor the solvent composition in real time, but also calibrate through machine learning algorithms such as support vector regression (SVR) to improve prediction accuracy. This high-precision component analysis provides a reliable data basis for subsequent solvent adjustment and ensures accurate control of solvent composition.

[0062] 2. Solvent Adjustment Algorithm

[0063] The solvent adjustment algorithm of the present invention is based on a dynamic programming method, and adjusts the solvent composition in real time according to the target solvent composition content, taking into account the dynamic characteristics and constraints of the solvent adjustment module. The specific steps include calculating the difference between the current solvent composition and the target composition, determining the amount of solvent to be added or removed, considering the adjustment speed limit and the reagent addition limit, and dynamically adjusting the adjustment speed and the reagent addition amount. This dynamic adjustment method can ensure that the solvent composition is always in the best state, avoiding the problems of low extraction efficiency and increased energy consumption caused by component fluctuations in traditional methods.

[0064] 3. Material flow control algorithm

[0065] The material flow control algorithm of the present invention is based on the mixed integer linear programming (MILP) method, and improves the washing efficiency of 1,3-butadiene and reduces the solvent loss by optimizing the material flow in the washing tower. The specific steps include establishing a material balance and energy balance model in the washing tower, defining an objective function, including material handling costs and environmental impact costs, setting constraints, including material balance constraints, flow restrictions of feed ports and discharge ports, and using the MILP algorithm to solve the optimal material flow scheme. This optimization method not only improves the washing efficiency, but also significantly reduces the waste of solvents and reduces production costs.

[0066] 4. Equipment operation parameter optimization algorithm

[0067] The equipment operation parameter optimization algorithm of the present invention is based on a genetic algorithm (GA), which reduces energy consumption and improves the separation efficiency of 1,3-butadiene by optimizing the operation parameters of the gas stripping tower. The specific steps include initializing the population, generating a set of random operation parameters, calculating the fitness value of each set of parameters, including energy consumption and separation efficiency, selecting individuals with higher fitness to enter the next generation, performing crossover and mutation operations, generating new individuals, replacing individuals with lower fitness, and repeating iterations until the optimal solution is reached. This optimization method can find the best combination of operation parameters, significantly reduce energy consumption, improve separation efficiency, and extend equipment life.

[0068] 5. Real-time monitoring and feedback control

[0069] The control module of the present invention monitors the key parameters (such as temperature, pressure, liquid level, flow rate, etc.) of each process link in real time, and makes dynamic adjustments according to the preset control strategy to ensure the stable operation of the entire extraction process. The control module mainly includes a data acquisition module, a data processing module, a control execution module and a feedback adjustment module. The data acquisition module is responsible for collecting data from each sensor, the data processing module pre-processes and analyzes the collected data, extracts key features, and generates control instructions. The control execution module adjusts the operating parameters of each device according to the control instructions. The feedback adjustment module automatically adjusts the control strategy by monitoring the system operation status in real time to ensure that the system operates in the optimal state.

[0070] 6. Optimize scheduling and management

[0071] The control module of the present invention also has optimization scheduling and management functions, which can automatically adjust the production process according to the production plan and actual needs to improve production efficiency and resource utilization. The optimization scheduling and management functions mainly include production plan management, equipment maintenance management, energy management and environmental management. Production plan management formulates a reasonable production plan based on market demand and inventory conditions to ensure that production tasks are completed on time; equipment maintenance management monitors the equipment operating status in real time, predicts equipment failures, arranges maintenance in advance, and reduces equipment downtime; energy management optimizes energy use strategies to reduce energy consumption and improve energy utilization efficiency; environmental management reduces the impact on the environment by reducing waste emissions and increasing resource recycling rates. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 System principle flow chart; DETAILED DESCRIPTION

[0073] The following embodiments of the present invention are described in further detail in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0074] Embodiment 1

[0075] The 1,3-butadiene extraction energy-saving and consumption-reducing system of the present invention is mainly composed of a solvent adjustment module, a washing tower module, a gas stripping tower module and a control module. The various systems work closely together and achieve the energy-saving and consumption-reducing goals in all directions through a unique method, ensuring the efficient and stable operation of the entire extraction process.

[0076] (I) Solvent adjustment module

[0077] 1. Algorithms and mathematical models

[0078] The solvent composition analysis model uses a spectral analysis model based on quantum chemical calculations to analyze the composition of the solvent. This model makes full use of the principles of quantum mechanics and accurately calculates the intrinsic relationship between the electronic structure and spectral characteristics of the solvent molecules to accurately predict the content of various components in the solvent.

[0079] For the NMP component in the solvent, its content analysis is based on the following formula:

[0080]

[0081] Where, CNMP is the content percentage of NMP, I NMP，i is the absorption intensity of NMP molecules in the i-th spectral band, w i is the weight coefficient of the i-th spectral band, n is the number of spectral bands related to NMP molecules; I total，j is the absorption intensity of all components in the solvent in the jth spectral band, and m is the number of relevant spectral bands of all components in the solvent.

[0082] For other key components in the solvent (such as possible trace impurities or additives), corresponding analysis models are also established. Assuming that there is an impurity component A, its content analysis formula is: Among them, C A is the content percentage of impurity component A, I A，i is the absorption intensity of the impurity component A molecule in the i-th spectral band, w i is the weight coefficient of the i-th spectral band (which is determined according to the component characteristics of different bands in NMP analysis), n is the number of spectral bands related to impurity component A; I total，j is the absorption intensity of all components in the solvent in the jth spectral band, and m is the number of relevant spectral bands of all components in the solvent.

[0084] In order to improve the accuracy of the spectral analysis model, regression algorithms in machine learning (such as support vector regression) are introduced to calibrate the model. Through training with a large number of solvent samples with known component contents, the model can better adapt to the changes in solvent composition under different working conditions, further reduce the analysis error, and control the analysis error of NMP content within ±0.2%, and the analysis error of impurity component A content within ±0.1%.

[0085] The solvent adjustment algorithm is based on a dynamic programming algorithm to achieve solvent adjustment. The algorithm is guided by achieving the target solvent content (such as 8% NMP content), and fully considers the dynamic characteristics and constraints of the solvent adjustment system, such as adjustment speed limit, reagent addition limit, etc.

[0086] Assume that the NMP content in the current solvent is C ourrent, the target content is C target , the adjustment speed is R (unit: % / min), and the reagent addition amount is limited to Qmax (unit: mol / min). The algorithm steps are as follows:

[0087] 1. Calculate the content difference ΔC = C target -C ourrent .

[0088] 2. If ΔC>0, then reagent needs to be added. Calculate the addition time At the same time, check whether the added amount exceeds the limit. If t add ·R>Q max , then adjust the addition speed R so that the addition amount is within the limit range.

[0089] 3. If ΔC < 0, the reagent needs to be removed (through a specific removal mechanism) and the removal time is calculated The method for adjusting the removal rate is similar to the addition rate adjustment.

[0090] When considering the dynamic characteristics and constraints of the solvent adjustment system, the relevant factors of the adjustment speed limit and the reagent addition amount limit are further refined. For example, the adjustment speed is not only related to the performance of the solvent adjustment equipment, but also to environmental factors such as the temperature and pressure of the current solvent system. A mathematical relationship model between the adjustment speed and environmental factors is established:

[0091]

[0092] Among them, R actual is the actual adjustment speed, R0 is the initial adjustment speed of the equipment (such as 0.5% / min mentioned above), kT and kp are the temperature and pressure influence coefficients (determined by experiments), T is the difference between the current temperature and the standard temperature, and P is the difference between the current pressure and the standard pressure.

[0093] For the reagent addition amount limit, in addition to considering the equipment capacity and process requirements, the chemical reaction activity of each component in the solvent is also dynamically adjusted. When some components in the solvent are highly active, the reagent addition amount limit is appropriately reduced to prevent overreaction. The relationship model between the reagent addition amount limit and the component activity is established:

[0094]

[0095] Where Qactual is the actual reagent addition limit, Q0 is the initial reagent addition limit (such as 0.5 mol / min), k A is the component activity influence coefficient, C A is the concentration of the active ingredient in the solvent.

[0096] Detailed parameter definition

[0097] Spectral band parameters For the spectral analysis of NMP molecules, the spectral band range selected is [200nm, 800nm]. Within this range, 10 key spectral bands are determined, and their wavelengths are 220nm, 250nm, 300nm, 350nm, 400nm, 450nm, 500nm, 550nm, 600nm, and 700nm. The weight coefficients corresponding to each spectral band are determined based on a large amount of experimental data and theoretical calculations, and are w1, w2, w3, w4, w5, w6, w7, w8, w9, and w10. For the spectral analysis of impurity component A, different spectral band ranges [λa, λb] are selected according to its molecular structure and spectral characteristics, and the corresponding key spectral bands and their weight coefficients wA1, wA2, …, wAn (the number of relevant spectral bands of impurity component A) are determined. The range of the adjustment speed of the solvent adjustment module is 0.1%-1% / min, and the initial default value is 0.5% / min. The reagent addition limit is determined according to the equipment capacity and process requirements of the solvent adjustment module. In the present invention, the range of Qmax is 0.1-1mol / min, and the initial default value is 0.5mol / min. These parameters can be dynamically adjusted according to actual conditions during the operation of the system to adapt to different production needs and working conditions.

[0098] (II) Scrubber Module

[0099] 1. Material flow control algorithm

[0100] The optimized control of the material flow in the scrubber is realized based on the mixed integer linear programming (MILP) algorithm. The algorithm aims to improve the scrubbing efficiency of 1,3-butadiene and reduce solvent loss, and comprehensively considers factors such as material balance, energy balance and equipment operation restrictions in the scrubber.

[0101] Suppose there are n feed ports and m discharge ports in the washing tower, xij represents the material flow rate from feed port i to discharge port j (i=1,2,…,i=1,2,…,n; =1,2,…,j=1,2,…,m), cij represents the unit material handling cost from feed port i to discharge port j, Li represents the material flow rate restriction matrix of feed port i, Uj represents the material flow rate restriction matrix of discharge port j, li represents the material flow rate restriction vector of feed port i, and uj represents the material flow rate restriction vector of discharge port j.

[0102] The objective function is: In, c E,k is the emission cost equivalent of the kth pollutant (determined according to environmental regulations and treatment costs), E k is the emission of the kth pollutant; W,lis the treatment cost equivalent of the lth wastewater component, and Wl is the generation amount of the lth wastewater component.

[0103] The constraints include:

[0104] 1. Material balance constraints:

[0105] 2. Feed inlet flow limit: l i ≤x ij ≤u i foralli,j

[0106] 3. Discharge port flow limit: ∑ 2 -1xij≤U;forallj

[0107] 4. Non-negative constraints on variables: x ij ≥0foralii,j

[0108] In terms of constraints, we further consider the impact of changes in the physical properties of materials on material balance and flow restrictions. For example, when the temperature or concentration of the material changes, its density and viscosity will also change, thus affecting the actual flow of the feed and discharge ports. Establish a relationship model between the physical properties of the material and the flow:

[0109] Among them, x actual is the actual flow rate, x design is the design flow rate (such as the flow rate limit range of the feed port and the discharge port), k ρ and k η are the influence coefficients of density and viscosity, ρ is the difference between the current density and the design density, and η is the difference between the current viscosity and the design viscosity.

[0110] 2. Determination and optimization of key parameters

[0111] Feed and discharge port parameters The washing tower is equipped with 2 feed ports and 1 discharge port, and the feed port is changed intermittently. The material flow rate of feed port 1 is limited to 10-50m 3 / h, the material flow rate of feed port 2 is limited to 15-60m 3 / h, the material flow rate of feed port 3 is limited to 20-70m 3 / h. The material flow rate of outlet 1 is limited to 25-80m 3 / h, the material flow rate at outlet 2 is limited to 10-40m 3 / h. These flow limits are determined based on the design specifications of the scrubber, the properties of the material, and production requirements, and can be fine-tuned according to specific circumstances in actual operation.

[0112] For the parameters of the feed and discharge ports, in addition to determining the flow rate limit range, the influence of the residence time distribution of the material at different flow rates on the washing effect is also studied. Establish a residence time distribution model: In the equation, P(t) is the probability of material residence at time t, and τ is the model parameter (determined by fitting experimental data). According to the residence time distribution model, the flow settings of the feed port and the discharge port are optimized to ensure that the material has a suitable residence time in the washing tower and improve the washing efficiency.

[0113] Unit material handling cost parameter: The unit material handling cost from feed port 1 to discharge port 1 is 0.5 yuan / m 3 The unit material handling cost from feed port 1 to discharge port 2 is 0.8 yuan / m 3 ; The unit material handling cost from feed port 2 to discharge port 1 is 0.6 yuan / m 3 The unit material handling cost from feed port 2 to discharge port 2 is 0.9 yuan / m 3 ; The unit material handling cost from feed port 3 to discharge port 1 is 0.7 yuan / m 3 The unit material handling cost from feed port 3 to discharge port 2 is 1.0 yuan / m 3 These cost parameters are obtained by comprehensively calculating various expenses such as energy consumption, equipment depreciation, and labor costs in the washing process, and are considered as important economic indicators when optimizing material flows.

[0114] For the unit material handling cost parameter, further subdivide the cost components, taking into account the fluctuations in raw material prices, changes in energy prices, and changes in equipment maintenance costs over time. Establish a dynamic cost calculation model:

[0115] Among them, cij(t) is the unit material handling cost at time t, cij,0 is the initial unit material handling cost (as determined above), kR, kE and kM are the influence coefficients of raw material price, energy price and equipment maintenance cost, R(t) is the change rate of raw material price, E(t) is the change rate of energy price, and M(t) is the change rate of equipment maintenance cost. According to the dynamic cost calculation model, the material flow optimization strategy is adjusted in time to reduce production costs.

[0116] (III) Gas stripping tower module

[0117] 1. Equipment operation parameter optimization algorithm

[0118] The optimization algorithm based on genetic algorithm (GA) is used to optimize the operating parameters of the gas stripping tower. The algorithm aims to reduce the energy consumption of the gas stripping tower and improve the separation efficiency of 1,3-butadiene, taking into account the gas-liquid equilibrium, mass transfer and heat transfer process, and equipment constraints in the gas stripping tower.

[0119] Assume that the operating parameters of the gas stripping tower include the number of plates N, feed position F, reflux ratio R, gas phase flow rate G and liquid phase flow rate L. These parameters constitute the chromosome code. The fitness function is defined as:

[0120]

[0121] Among them, E actual is the actual energy consumption, E max is the maximum allowable energy consumption, D actual is the actual separation efficiency, D max is the maximum separation efficiency, α and β are weight coefficients (adjusted according to actual needs).

[0122] The steps of genetic algorithm include:

[0123] Initialize the population: randomly generate a set of chromosomes, each chromosome represents a set of operating parameters.

[0124] Fitness evaluation: Calculate the fitness value of each chromosome.

[0125] Selection operation: select individuals in the next generation population based on their fitness values.

[0126] Crossover operation: Randomly select two chromosomes for crossover to generate new chromosomes.

[0127] Mutation operation: Randomly select a gene on the chromosome for mutation.

[0128] Replacement operation: Replace individuals with lower fitness with newly generated chromosomes.

[0129] Termination condition: The algorithm is terminated when the predetermined number of iterations is reached or the fitness value no longer increases significantly.

[0130] 2. Determination and optimization of key parameters

[0131] The number of plates N is an important parameter affecting the separation efficiency of the gas stripping tower. Through experiments and simulation analysis, the optimal number of plates is determined to be 20-40. In actual operation, the number of plates is dynamically adjusted according to the separation effect and energy consumption requirements.

[0132] Feed position FThe feed position F affects the gas-liquid contact area and mass transfer efficiency in the gas stripping tower. Through experimental data fitting, the optimal feed position range is determined to be 1 / 3-2 / 3 of the tower height. In actual operation, the feed position is dynamically adjusted according to the material properties and separation requirements.

[0133] Reflux ratio R The reflux ratio R affects the separation effect and energy consumption of the gas stripping tower. Through experiments and simulation analysis, the optimal reflux ratio range is determined to be 1.5-3.0. In actual operation, the reflux ratio is dynamically adjusted according to the separation effect and energy consumption requirements.

[0134] The gas phase flow rate G and liquid phase flow rate L affect the gas-liquid equilibrium and mass transfer and heat transfer process in the gas stripping tower. Through experiments and simulation analysis, the optimal gas phase flow rate range is determined to be 100-300m 3 / h, liquid flow range is 50-150m 3 / h. In actual operation, the gas and liquid phase flows are dynamically adjusted according to the separation effect and energy consumption requirements.

[0135] (IV) Control module

[0136] Real-time monitoring and feedback control The control module of the present invention monitors the key parameters of each process link (such as temperature, pressure, liquid level, flow rate, etc.) in real time and makes dynamic adjustments according to the preset control strategy to ensure the stable operation of the entire extraction process. The control module mainly includes the following parts:

[0137] Data acquisition module: responsible for collecting data from various sensors, including temperature sensors, pressure sensors, liquid level sensors, flow sensors, etc.

[0138] Data processing module: pre-process and analyze the collected data, extract key features, and generate control instructions.

[0139] Control execution module: According to the control instructions, adjust the operating parameters of each device, such as valve opening, pump speed, heating power, etc.

[0140] Feedback adjustment module: monitors the system operation status in real time and automatically adjusts the control strategy to ensure that the system operates in the optimal state.

[0141] The optimization scheduling and management control module also has optimization scheduling and management functions, which can automatically adjust the production process according to the production plan and actual needs to improve production efficiency and resource utilization. The optimization scheduling and management functions mainly include the following aspects:

[0142] Production planning management: Develop reasonable production plans based on market demand and inventory conditions to ensure that production tasks are completed on time.

[0143] Equipment maintenance management: Real-time monitoring of equipment operating status, prediction of equipment failures, and scheduling of maintenance in advance can reduce equipment downtime.

[0144] Energy management: Reduce energy consumption and improve energy efficiency by optimizing energy usage strategies.

[0145] Environmental management: Reduce the impact on the environment by reducing waste emissions and increasing resource recycling rates.

[0146] Example 2

[0147] Data collection and preprocessing

[0148] Temperature sensor: installed at various key locations to monitor temperature changes in real time.

[0149] Pressure sensor: installed on pipelines and equipment to monitor pressure changes in real time.

[0150] Liquid level sensor: installed in storage tanks and reactors to monitor liquid level changes in real time.

[0151] Flow sensor: installed at the feed inlet and discharge port to monitor flow changes in real time.

[0152] Data processing and analysis

[0153] Data cleaning: Remove abnormal data and noise to ensure data accuracy and reliability.

[0154] Feature extraction: Extract key features, such as temperature change rate, pressure fluctuation amplitude, liquid level change trend, etc.

[0155] Model training: Use historical data to train solvent composition analysis models, material flow control models, and equipment operation parameter optimization models.

[0156] Control execution and feedback regulation

[0157] Valve control: According to the control instructions, adjust the valve opening to control the material flow.

[0158] Pump speed control: According to the control instructions, adjust the pump speed to control the liquid delivery speed.

[0159] Heating control: Adjust heating power and control temperature according to control instructions.

[0160] Feedback regulation: By real-time monitoring of the system operating status, the control strategy is automatically adjusted to ensure that the system operates in the optimal state.

[0161] Optimize scheduling and management

[0162] Production planning management: Develop reasonable production plans based on market demand and inventory conditions to ensure that production tasks are completed on time.

[0163] Equipment maintenance management: Real-time monitoring of equipment operating status, prediction of equipment failures, and scheduling of maintenance in advance can reduce equipment downtime.

[0164] Energy management: Reduce energy consumption and improve energy efficiency by optimizing energy usage strategies.

[0165] Environmental management: Reduce the impact on the environment by reducing waste emissions and increasing resource recycling rates.

[0166] Experimental location: A large chemical plant with an annual output of 100,000 tons of 1,3-butadiene.

[0167] Experimental period: 6 months, divided into two phases, each phase lasting 3 months.

[0168] Control group and experimental group: The control group adopted the traditional management method, and the experimental group adopted the intelligent management system of the present invention.

[0169] Data collection

[0170] Temperature data: collected once every minute.

[0171] Pressure data: collected once every minute.

[0172] Liquid level data: collected every 10 minutes.

[0173] Traffic data: collected every 5 minutes.

[0174] Data analysis

[0175] Reduced energy consumption: The experimental group's energy consumption was reduced by 20%.

[0176] Reduced solvent loss: The experimental group experienced a 15% reduction in solvent loss.

[0177] Reduced equipment failure rate: The equipment failure rate of the experimental group was reduced by 30%.

[0178] Improved production efficiency: The production efficiency of the experimental group increased by 10%.

[0179] Economic benefits

[0180] Cost savings: The experimental group saw a 15% reduction in production costs.

[0181] Increased Revenue: The experimental group increased production by 10%, increasing sales revenue.

[0182] Comprehensive benefits: Taking into account the cost savings and income increase, the economic benefits of the experimental group are significantly better than those of the control group, with a net increase of 500 yuan per ton of 1,3-butadiene.

[0183] Example 3

[0184] Data acquisition and preprocessing Spectral sensor: installed at key pipelines, real-time spectral information of reaction materials is collected for analyzing changes in material composition. Vibration sensor: installed on the equipment housing, monitoring the vibration of the equipment during operation, and assisting in judging the health status of the equipment. Electrochemical sensor: used to detect the concentration of specific ions to monitor changes in the chemical environment during the reaction process.

[0185] The data collection frequency is: spectral data is collected every 3 minutes, vibration data is collected 10 times per second, and electrochemical data is collected every 2 minutes.

[0186] The collected data were preprocessed, including using wavelet transform to remove noise from spectral data, filtering vibration data to extract effective features, and standardizing electrochemical data.

[0187] Data processing and analysis Data fusion: Fuse data from different sensors, use the convolutional neural network (CNN) in deep learning to automatically extract complex features from multi-source data, and construct a comprehensive feature vector. Model construction and training: Based on the fused feature vector, build a deep belief network (DBN) model to predict key parameters such as the remaining service life of the equipment, the best time to replenish the solvent, and the optimal temperature control curve of the reaction. Model training uses historical data from the past year, and is trained through unsupervised pre-training and supervised fine-tuning. Anomaly detection: Use the trained autoencoder model to calculate the reconstruction error of real-time data. When the reconstruction error exceeds the set threshold, it is judged as abnormal data, triggering an alarm and conducting in-depth analysis.

[0188] Control execution and feedback regulation, intelligent valve control: using a combination of fuzzy logic control and model predictive control (MPC), according to the DBN model prediction results and real-time feedback data, accurately control the valve opening, realize precise adjustment of material flow, and take into account valve wear and energy consumption factors. Dynamic pump speed adjustment: using a neural network controller optimized by the particle swarm optimization (PSO) algorithm, dynamically adjust the pump speed according to the system pressure, liquid level and flow requirements to ensure the efficiency and stability of liquid transportation. Precise heating control: based on the Q-learning algorithm in reinforcement learning, the heating power is automatically adjusted according to real-time temperature data and reaction requirements to adapt to the temperature control requirements under different working conditions, while minimizing energy consumption. Feedback regulation mechanism: by comparing the actual operation data with the model prediction data in real time, the parameters of the control model are continuously adjusted using the gradient descent algorithm to ensure that the system always runs in the optimal state.

[0189] Optimize scheduling and management Intelligent production plan optimization: Use genetic algorithms (GA) to optimize production plans, comprehensively consider multiple factors such as raw material supply, market price fluctuations, equipment maintenance plans, etc., and formulate the best production plan to maximize profits. Predictive equipment maintenance: Use support vector machines (SVM) to classify equipment operation data, predict the type and time of possible equipment failures, arrange targeted maintenance in advance, and reduce unplanned downtime. Intelligent energy management: By establishing an energy consumption prediction model (using long short-term memory network LSTM), plan energy use strategies in advance, and reasonably allocate energy according to different production stages and equipment operation status to reduce overall energy consumption. Green environmental management: Use cluster analysis algorithms to analyze waste emission data, find out emission patterns, optimize production processes to reduce the generation of hazardous waste, and discover potential resource recycling opportunities through data mining to improve resource recovery rates.

[0190] The experiment was conducted in another large chemical company, which produces 80,000 tons of 1,3-butadiene annually. The experimental period was 8 months, divided into three stages. The first two months were the system debugging and data collection adaptation period, the middle four months were the formal comparative experiment period, and the last two months were the system stability test. The control group continued to use the company's original conventional management and PID control methods, while the experimental group deployed the intelligent management system based on the deep learning algorithm of the present invention.

[0191] The temperature data was collected once per minute, the pressure data was collected every 30 seconds, the liquid level data was collected every 15 minutes, and the flow data was collected every 3 minutes. The spectral data, vibration data, and electrochemical data were collected at the same frequency as before the above pretreatment. The comparison results showed that the energy consumption of the experimental group was reduced by 25%, the solvent loss was reduced by 20%, the equipment failure rate was reduced by 35%, and the production efficiency was increased by 15%.

[0192] Economic benefit analysis: In terms of cost savings, the experimental group's production costs were reduced by 18%, mainly due to reduced energy consumption, reduced equipment maintenance costs, and reduced solvent replenishment costs. In terms of revenue increase, the output increased by 12%, bringing significant sales revenue growth. Comprehensively calculated, the net increase in revenue per ton of 1,3-butadiene reached 650 yuan, and the economic benefits were significantly better than the control group.

[0193] Embodiment 4:

[0194] Data acquisition and preprocessing, image sensor: Install high-definition cameras in key reaction areas to collect image data in the reaction process in real time to observe the material status and equipment surface condition. Acoustic emission sensor: Used to monitor the acoustic emission signals generated by potential faults such as the expansion of tiny cracks inside the equipment. Multi-parameter gas sensor: Detects the concentration of multiple gases in the reaction environment, such as oxygen, carbon dioxide, etc., to assist in judging the reaction process and safety. The data acquisition time interval is: image data is collected once every 5 minutes, acoustic emission data is collected 5 times per second, and gas concentration data is collected once per minute. Image data is preprocessed by image enhancement and feature extraction, acoustic emission data is denoised, and gas concentration data is normalized.

[0195] Data processing and analysis, image data analysis: Use the target detection algorithm in deep learning (such as the YOLO algorithm) to identify and analyze the material level, foam situation, etc. in the image, and extract key visual features. Signal feature fusion: Fusion the acoustic emission signal features, gas concentration features and other sensor data features, and use the principal component analysis (PCA) algorithm to reduce the data dimension while retaining key information.

[0196] Model construction and training: A hybrid model based on random forest (RF) and artificial bee colony optimization algorithm (ABC) is constructed to predict important indicators such as equipment failure probability, reaction conversion rate and product quality. The training data includes historical production data and experimental data collected in the early stage, and the model parameters are optimized through cross-validation. Real-time optimization decision: Using case-based reasoning (CBR) technology, it quickly provides optimization control strategy suggestions based on real-time data and historical case libraries.

[0197] Control execution and feedback regulation Adaptive valve control: Using fuzzy adaptive control algorithm, dynamically adjust the valve opening according to the hybrid model prediction results and real-time feedback to adapt to the flow control requirements under different working conditions. Intelligent pump group collaborative control: Based on the distributed collaborative control algorithm, coordinate the operating speed and start and stop status of multiple pumps to achieve efficient coordination of liquid transportation and improve the overall stability of the system. Precise heating and cooling control: Using a method combining model predictive control (MPC) and neural network (NN), accurately control the power of the heating and cooling system according to the reaction thermal effect and product quality requirements to ensure that the reaction temperature is within the optimal range. Feedback regulation enhancement: Introduce a state estimation method based on Kalman filtering to make a more accurate estimate of the system state and improve the accuracy and response speed of feedback regulation.

[0198] Optimize scheduling and management Dynamic production plan adjustment: Use simulated annealing algorithm (SA) to dynamically adjust production plans according to changes in market demand, raw material inventory and equipment status to ensure optimal resource allocation. Equipment health management: Model and predict equipment operating status based on hidden Markov model (HMM), discover potential fault hazards in advance, and formulate personalized maintenance strategies. Efficient energy utilization: By establishing an energy flow analysis model (using graph theory algorithm), find out the key paths and nodes of energy consumption, optimize energy use in a targeted manner, and reduce energy consumption. Sustainable development management: Use life cycle assessment (LCA) algorithm to conduct environmental impact assessment of the entire production process, optimize process and management strategies based on the assessment results, and reduce environmental footprint.

[0199] The experiment was conducted in a medium-sized chemical company, which has an annual output of 50,000 tons of 1,3-butadiene. The experimental period was 9 months, of which the first month was for system construction and debugging, the middle six months were for the comparative test stage, and the last two months were for system optimization and summary. The control group used the company's existing traditional management and control system, and the experimental group deployed an intelligent management system based on a hybrid intelligent algorithm. Temperature data is collected every 2 minutes, pressure data is collected every minute, liquid level data is collected every 10 minutes, and flow data is collected every 4 minutes. Image, acoustic emission and gas concentration data are collected at the above frequency. The experimental results show that the energy consumption of the experimental group was reduced by 22%, solvent loss was reduced by 18%, equipment failure rate was reduced by 32%, and production efficiency was increased by 13%.

[0200] Economic Benefit Analysis In terms of cost savings, the experimental group reduced production costs by 16%, thanks to energy savings, equipment maintenance optimization and improved raw material utilization. In terms of revenue increase, the output increased by 11%, increasing product sales revenue. Comprehensively calculated, the net increase in revenue per ton of 1,3-butadiene was 580 yuan, which was significantly better than the control group, demonstrating the good economic and environmental benefits of the intelligent management system.

Claims

1. A 1,3-butadiene extraction energy-saving and consumption-reducing system, characterized in that: include: A solvent adjustment module, used to adjust the solvent composition in real time according to a solvent composition analysis model and a solvent adjustment algorithm; Scrubber module, used to optimize the material flow in the scrubber through the material flow control algorithm; The gas stripping tower module is used to optimize the operating parameters of the gas stripping tower through the equipment operating parameter optimization algorithm; The control module is used to monitor and control the key parameters of each process link in real time.

2. The system according to claim 1, characterized in that The solvent composition analysis model is based on a spectral analysis model of quantum chemical calculations. By analyzing the electronic structure and spectral characteristics of solvent molecules and combining machine learning methods, the content of various components in the solvent can be accurately predicted. The specific steps include: Collecting spectral data of the solvent; Using quantum chemical calculation methods, the relationship between the electronic structure and spectral characteristics of solvent molecules is established; Use the support vector regression machine learning algorithm to calibrate the model and improve prediction accuracy.

3. The system according to claim 1, characterized in that The solvent adjustment algorithm is based on a dynamic programming algorithm, and adjusts the solvent composition in real time according to the target solvent composition content, taking into account the dynamic characteristics and constraints of the solvent adjustment module. The specific steps include: Calculate the difference between the current solvent composition and the target solvent composition; Determine the amount of solvent to be added or removed based on the difference; Considering the adjustment speed limit and reagent addition amount limit, adjust the parameters of the solvent adjustment module; Dynamically adjust the adjustment speed and reagent addition amount to ensure that the solvent composition reaches the target value.

4. The system according to claim 1, characterized in that The material flow control algorithm is based on a mixed integer linear programming (MILP) algorithm, which improves the washing efficiency of 1,3-butadiene and reduces solvent loss by optimizing the material flow in the washing tower. The specific steps include: Establish material balance and energy balance models in the scrubber; Define the objective function, including material handling costs and environmental impact costs; Set constraints, including material balance constraints, flow limits for feed and discharge ports; Use MILP algorithm to solve the optimal material flow plan; Suppose there are n feed ports and m discharge ports in the washing tower, xij represents the material flow rate from feed port i to discharge port j (i=1,2,…,i=1,2,…,n; =1,2,…,j=1,2,…,m), cij represents the unit material handling cost from feed port i to discharge port j, Li represents the material flow rate restriction matrix of feed port i, Uj represents the material flow rate restriction matrix of discharge port j, li represents the material flow rate restriction vector of feed port i, and uj represents the material flow rate restriction vector of discharge port j.

5. The system according to claim 1, wherein: The equipment operation parameter optimization algorithm is based on a genetic algorithm, which reduces energy consumption and improves the separation efficiency of 1,3-butadiene by optimizing the operation parameters of the gas stripping tower. The specific steps include: Initialize the population and generate a set of random operating parameters; Calculate the fitness value for each set of parameters, including energy consumption and separation efficiency; Select individuals with higher fitness to enter the next generation; Perform crossover and mutation operations to generate new individuals; Replace individuals with lower fitness and repeat the iteration until the optimal solution is reached.

6. The system according to claim 1, wherein: The control module includes a data acquisition module, a data processing module, a control execution module and a feedback adjustment module, which are used to monitor and control the key parameters of each process link in real time to ensure that the system operates in the optimal state. The specific steps include: Collect temperature, pressure, liquid level and flow data in real time through sensors; Preprocess and analyze the collected data to extract key features; Generate control instructions based on analysis results to adjust valve opening, pump speed, and heating power; By real-time monitoring of the system operating status, the control strategy is automatically adjusted to ensure stable system operation.

7. A method for energy saving and consumption reduction of 1,3-butadiene extraction, characterized in that: Includes steps: Collect key parameter data of each process link; Use solvent composition analysis model and solvent adjustment algorithm to adjust solvent composition in real time; Use the material flow control algorithm to optimize the material flow in the scrubber; Use equipment operating parameter optimization algorithm to optimize the operating parameters of the gas stripping tower; The control module monitors and controls the key parameters of each process in real time to ensure that the system operates in the optimal state; the optimization algorithm based on the genetic algorithm optimizes the operating parameters of the gas stripping tower. Assume that the operating parameters of the gas stripping tower include the number of plates N, feed position F, reflux ratio R, gas phase flow rate G and liquid phase flow rate L. These parameters constitute the chromosome encoding, and the fitness function is defined as: Among them, E actual is the actual energy consumption, E max is the maximum allowable energy consumption, D actual is the actual separation efficiency, Dmax is the maximum separation efficiency, and α and β are weight coefficients.

8. The method according to claim 7, characterized in that The specific steps include: The material balance and energy balance models in the washing tower are established, the objective function is defined, including material handling cost and environmental impact cost, constraints are set, including material balance constraints, flow restrictions at the feed and discharge ports, and the MILP algorithm is used to solve the optimal material flow plan.

9. The method according to claim 7, characterized in that The specific steps also include: Calculate the difference between the current solvent composition and the target composition, determine the amount of solvent that needs to be added or removed, consider the adjustment speed limit and the reagent addition limit, and dynamically adjust the adjustment speed and reagent addition amount.

10. The method according to claim 7, characterized in that The specific steps also include: initializing the population, generating a set of random operating parameters, calculating the fitness value of each set of parameters, including energy consumption and separation efficiency, selecting individuals with higher fitness to enter the next generation, performing crossover and mutation operations, generating new individuals, replacing individuals with lower fitness, and repeating iterations until the optimal solution is reached.

Images