Waste organic solution purification and regeneration method for regenerating isopropanol
Through phased distillation and distillation operations, combined with hydrogen peroxide and activated carbon treatment, combined with LSTM and deep reinforcement learning algorithm to optimize the distillation process, the problem of complex process and difficult to improve purity in isopropanol purification is solved, and high efficiency and low The purity of isopropanol is increased.
Patent Information
- Application Number
- CN202410990067.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-22
- Publication Date
- 2024-11-12
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing isopropanol purification methods have problems such as complex processes, low selectivity, serious equipment corrosion, large amounts of steam and energy consumption, and are difficult to achieve extremely high purity, and may cause loss of isopropanol solvent.
Stage distillation and distillation operations are adopted, combined with hydrogen peroxide solution and activated carbon particles for pretreatment, followed by vacuum suction filtration, centrifugation and ultrasonic extraction technology, and finally optimize the distillation process through the LSTM model and deep reinforcement learning algorithm to control the heating power and condenser temperature to increase the purity of isopropanol.
It significantly improves the concentration and purity of isopropanol, reduces environmental hazards, improves resource utilization, reduces energy consumption and production costs, and enhances system stability and market competitiveness.
Smart Images

Figure CN118930407A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of separation and purification of chemical products, in particular to a method for purifying and regenerating a waste organic solution for regenerating isopropanol. Background Art
[0002] Isopropyl alcohol is an important organic solvent with a wide range of uses. It is mainly used as a solvent for paints, inks, pesticides, synthetic resins, electronic appliances, as well as an extraction solvent, coating solvent, dewaxing solvent and shellac solvent. It is also a diluent for automotive polishing paints and plastic color printing inks, and can be used as an antifreeze agent. Isopropyl alcohol is also an important chemical raw material, used in a variety of synthetic, pharmaceutical and fragrance preparation processes, such as the preparation of acetone, methyl isobutyl ketone, isopropylamine and various esters. Isopropyl alcohol exists in the air in a vapor state. Its vapor is heavier than air and can cause slight irritation to the eyes, nose and throat, and can be absorbed by the human body through the skin.
[0003] As the global environmental protection voice is getting louder and louder, the consumption of chlorofluorocarbons is restricted. As a substitute for chlorofluorocarbons and other electronic industrial cleaning agents, isopropyl alcohol has become one of the fastest growing areas in the world. The main production methods of isopropyl alcohol are propylene indirect hydration, propylene direct hydration and acetone hydrogenation.
[0004] The indirect hydration method of sulfuric acid is one of the indirect hydration methods. The indirect hydration method of sulfuric acid is to dissolve propylene in a sulfuric acid aqueous solution, hydrate to synthesize isopropyl acid sulfate and diisopropyl sulfate, and the esters obtained after hydration are hydrolyzed (treated with water vapor) to obtain crude isopropyl alcohol, and then refined to obtain an isopropyl alcohol product with a mass fraction of 99%, and hydrolyzed to obtain dilute sulfuric acid, which is concentrated and removed solids or colloids and then recycled. The advantage of the indirect hydration method of sulfuric acid is that the purity of the raw material propylene is not high, the propylene conversion rate can reach more than 90%, and the concentration of the obtained crude isopropyl alcohol is as high as about 60%, so the refining cost can be reduced. The disadvantages are complicated process, low selectivity, and both the hydrolysis of esters and the regeneration of sulfuric acid require a large amount of steam, and the equipment is severely corroded. In addition, there is a problem of treatment of waste sulfuric acid. The existing purification method consumes a large amount of energy, and although purification can remove most of the impurities, it is difficult to achieve extremely high purity, and in the process of purification, the loss of isopropyl alcohol solvent may also be caused. For this reason, we propose a waste organic solution purification and regeneration method for regenerating isopropyl alcohol. Summary of the invention
[0005] The object of the present invention is to provide a waste organic solution purification and regeneration method for regenerating isopropyl alcohol.
[0006] In order to solve the problems raised in the above background technology, the present invention provides the following technical solutions: a method for purifying and regenerating a waste organic solution for regenerating isopropanol, wherein the isopropanol waste organic solution is a residual solution generated after using isopropanol, and the residual solution contains other impurities and reaction by-products, and a chemical reagent is used to treat substances in the isopropanol waste organic solution. The specific operating steps of the method for purifying and regenerating a waste organic solution for regenerating isopropanol are as follows:
[0007] Step 1: placing the isopropanol waste organic solution and distilled water in a mixing barrel at a mass ratio of 1:1, adding hydrogen peroxide solution, heating and stirring, standing and cooling to room temperature to obtain an isopropanol premix, adding activated carbon particles to the isopropanol premix, stirring for 5 minutes and standing, removing the supernatant after standing, and transferring the remaining solid into an incinerator for differentiation treatment;
[0008] Step 2: transfer the supernatant obtained in step 1 into a beaker, use a vacuum filter to vacuum filter the supernatant, use a filter membrane to filter the supernatant, the filtrate obtained is isopropanol filtrate, collect the filter membrane and filter residue and then perform heat treatment;
[0009] Step 3: transfer the isopropanol filtrate into a centrifuge tube, the volume of which does not exceed two-thirds of the volume of the centrifuge tube, place the centrifuge tube in a centrifuge for centrifugation to separate the isopropanol separation liquid and excess solid matter, and collect the solid matter into an evaporating dish for high-temperature heat treatment;
[0010] Step 4: transferring the separately collected isopropanol separated liquid into a distiller for distillation, heating and evaporating the isopropanol separated liquid, collecting the extract from the condenser into a beaker, collecting the isopropanol distillate, and then introducing the isopropanol distillate into a distillation tower, heating and distilling the isopropanol distillate, cooling the tower top vapor through a condenser to convert it into liquid, collecting different fractions according to the boiling point difference of the components, and obtaining a high-purity isopropanol distillate;
[0011] Step 5: The collected distillation filtrate is transferred into a conical flask, and an ultrasonic extraction machine is used to perform ultrasonic extraction on the isopropanol distillate. After extraction, a clean isopropanol distillate is obtained, and the purity is tested to ensure that the required purity is achieved.
[0012] As a further solution of the present invention: the isopropanol waste organic solution is a hazardous waste, the isopropanol concentration in the isopropanol waste organic solution is 30.0%-50.0%, and the isopropanol concentration after purification is 99.3%-99.7%. When handling it, it is necessary to pay attention to comply with safety operation regulations.
[0013] As a further solution of the present invention: in the step 1, hydrogen peroxide can be used to disinfect the waste organic solution of isopropyl alcohol, activated carbon particles are used to adsorb and precipitate impurities in the isopropyl alcohol premix, and then the precipitate is dried and collected.
[0014] As a further solution of the present invention: in the step 2, the pore size of the filter membrane used for vacuum filtration is 1 μm to 5 μm, and after the filtration is completed, the organic matter in the remaining solid matter is put into an incinerator for thermal decomposition treatment.
[0015] As a further solution of the present invention: in the step three, the temperature of the isopropanol separation liquid when placed in the centrifuge tube and the centrifuge is controlled at 25°C, the centrifugal speed is 2800r / min, the centrifugal time is 6min to 10min, and after the centrifugation is completed, the centrifuge tube is taken out after the centrifuge stops naturally to separate the isopropanol separation liquid and excess solid matter.
[0016] As a further solution of the present invention: in the step 4, the distillation operation is carried out twice, and after the isopropanol separation liquid is distilled, it is transferred to a distillation tower for multiple distillations.
[0017] As a further solution of the present invention: in the step five, a heating device is used to heat the inside of the tower to vaporize the isopropanol distillate, and the vaporized steam rises to the top of the distillation tower to collect high-purity distillate.
[0018] As a further solution of the present invention: the extractant in the ultrasonic extraction step is ethylene glycol, and the extracted substances include heavy metal ions and organic solutions in the isopropanol distillate.
[0019] As a further solution of the present invention: after the clean isopropanol distillate is obtained, it is placed in a light-proof container and sealed for storage at a storage temperature of 25°C to 30°C.
[0020] As a further solution of the present invention: In step 4, a high-purity isopropanol distillation process control method is added, comprising the following steps:
[0021] Step 1, data acquisition module:
[0022] Temperature sensor: installed in different sections of distillation towers and rectification towers (top, middle, bottom) to monitor the temperature of each section;
[0023] Variable T i is the temperature of the i-th position
[0024]
[0025] T means to calculate the average temperature of each location, and n is the number of sensors;
[0026] Pressure sensor: installed at the top and bottom of distillation towers and rectification towers to monitor the pressure inside the towers;
[0027] Variable P i is the pressure at the ith position
[0028]
[0029] P means to calculate the average value of pressure at each position;
[0030] Liquid level sensor: installed in the tower kettle and the middle section to monitor the liquid level;
[0031] Variable L i is the liquid level at the ith position
[0032]
[0033] It means to calculate the average value of the liquid level at each position;
[0034] Flow sensor: installed at the feed port and discharge port to monitor the flow of feed and discharge;
[0035] Variable F in is the feed flow rate, F out is the discharge flow rate;
[0036] ΔF=F in -F out
[0037] ΔF is the flow difference between the feed and the discharge, ensuring the material balance of the system;
[0038] Component concentration sensor: installed at the gas outlet at the top of the tower and the liquid outlet at the bottom of the tower to monitor the concentration of isopropanol;
[0039] Variable C top is the concentration of the top component, C bottom is the concentration of the bottom component, C avg The average concentration
[0040]
[0041] It means calculating the average component concentration at the top and bottom of the tower to evaluate the separation effect of distillation and rectification;
[0042] All the above sensors are connected to the central control system through the data acquisition module (DCS system);
[0043] Step 2, data preprocessing:
[0044] First, the data collected by the sensors are denoised separately;
[0045]
[0046] in is the denoised data at time t, is the predicted value at time t, z t is the observed value at time t, K t Kalman gain at time t;
[0047] Then perform maximum and minimum normalization on the denoised data:
[0048]
[0049] Among them, x′ is the normalized data, x is the original input data, and x max is the maximum value of the data, x min is the minimum value of the data;
[0050] Then, data fusion is performed on the data collected by different sensors. The steps are as follows:
[0051]
[0052] x f is the fused data, w i is the weight of the i-th sensor, x i is the data of the i-th sensor, n is the number of sensors;
[0053] Step 3, data prediction model:
[0054] The present invention uses the long short-term memory network (LSTM) model to perform time series prediction on the preprocessed data. Specifically, we use the preprocessed temperature, pressure, liquid level, flow rate and component concentration data to build the LSTM model and predict the changes in key parameters;
[0055] The first is the input layer, where the input data includes temperature, pressure, liquid level, flow rate, and component concentration.
[0056] X t =[T t , P t , L t , F t , C t ]
[0057] X t To group variables, T t is the temperature at time t, P t is the pressure at time t, L t is the liquid level at time t, F t The flow rate at time t, C t The concentration of the component at time t;
[0058] Next is the setting of the hidden layer, which contains multiple LSTM units. Each unit includes an input gate, a forget gate, and an output gate to capture long-term dependencies in the time series:
[0059] Input gate: i t =σ(W i ·[h t-1 ,x t ]+b i )
[0060] Forget gate: f t =σ(W f ·[h t-1 ,x t ]+b f )
[0061] Output gate: o t =σ(W o ·[h t-1 , x t ]+b o )
[0062] Cell status update:
[0063] Cell Status:
[0064] Hidden state: h t =o t tanh(C t )
[0065] Among them, i t is the input gate activation value, σ is the unactivated function, W i is the input gate weight matrix, t is the time step, h t-1 , is the hidden state at time t-1, b i is the bias vector of the input gate, f t is the forget gate activation value, W f is the weight matrix of the forget gate, b f is the bias vector of the forget gate, o t is the output gate activation value, W o is the weight matrix of the output gate, b o is the bias vector of the output gate, is the cell state update, W C is the weight matrix updated for the cell state, b C is the bias vector for updating the cell state, C t is the cell state at the current moment;
[0066] Finally, the predicted output of the output layer is
[0067] Temperature forecast:
[0068] Prediction of stress:
[0069] Liquid level prediction:
[0070] Flow control:
[0071] Control of component concentration:
[0072] in, is the predicted temperature value at the future time, is the weight matrix for temperature prediction, is the bias vector for temperature prediction, Predicted stress values for future moments is the weight matrix for pressure prediction, is the bias vector for pressure prediction, is the predicted value of the liquid level at the future moment, is the weight matrix for liquid level prediction, is the bias vector for liquid level prediction, Traffic forecast value at future time, is the weight matrix for traffic prediction, is the bias vector for flow prediction, is the predicted value of the component concentration at the future time, is the weight matrix for component concentration prediction, is the bias vector for component concentration prediction, and t is the time step;
[0073] Accurate prediction of the above parameters is crucial for optimizing the distillation and rectification processes to ensure high purity of isopropanol;
[0074] Step 4, control decision module:
[0075] Combined with the prediction results of the LSTM model, the Deep Reinforcement Learning (DRL) algorithm is used to adjust the operating parameters in the purification and regeneration method of the regenerated isopropanol waste organic solution in real time:
[0076] We chose the Deep Q Network (DQN) as the DRL algorithm. DQN evaluates the value of each state-action pair by learning the Q-value function. We use the deep learning framework TensorFlow to implement it.
[0077] The state space includes the system state at the current moment and the parameters at the future moment predicted by the LSTM model, and the action space includes the adjustable operating parameters in the control system, including:
[0078] Heating power A hp , condenser temperature A ct , Tower top pressure A tp ;
[0079] The reward function is defined based on the operation effect (isopropyl alcohol purity, energy consumption),
[0080]
[0081] r t is the reward function, α and β are weight coefficients, is the predicted component concentration at the future time, E t Energy consumption for heating and condenser;
[0082] Finally, according to the operation decision generated by DQN, the heating power, condenser temperature and tower top pressure are adjusted, and according to the feedback data, the DQN model parameters are dynamically adjusted to ensure the stability and optimal operation of the system;
[0083] Step 5: Execute the module
[0084] Based on the prediction results of the LSTM model and the optimization decision of the DRL algorithm, specific control instructions are generated. The control instructions include heating power, condenser temperature and valve opening. The control instructions are transmitted to relevant actuators through the industrial control network (PLC);
[0085] Heater Control Operation:
[0086]
[0087] Condenser control operation:
[0088]
[0089] Valve control operation:
[0090]
[0091] Among them, A hp is the heating power control instruction, k p is the proportional coefficient of the PID controller, T set is the set temperature, T actual is the actual temperature, k i is the integral coefficient of the PID controller, k d is the differential coefficient of the PID controller, A ct is the condenser temperature control instruction, Cset To set the condensation temperature, C actual is the actual condensation temperature, A vp is the valve opening control instruction, F set To set the flow rate, F actual is the actual flow rate;
[0092] Next, rely on sensors to monitor the operating status of the heater, condenser and valve in real time, obtain sensor feedback data, monitor system parameters, and if it is found that the parameters deviate from the set value, promptly issue an alarm and make adjustments.
[0093] By adopting the above technical solution, compared with the prior art, the beneficial effects of the present invention are:
[0094] 1. The present invention can improve the concentration of purified isopropanol through staged and refined distillation operations, and can separate harmful substances in the isopropanol waste organic solution and reasonably dispose of them by pre-treating the isopropanol waste organic solution, thereby reducing potential harm to the environment. The isopropanol waste organic solution is purified through specific steps, so that the purified isopropanol can be reused, thereby greatly improving the utilization rate of resources;
[0095] 2. The present invention uses hydrogen peroxide solution and activated carbon particles to adsorb the isopropanol waste organic solution, which can effectively remove impurities and pollutants therein, ensuring that the isopropanol finally obtained can have a higher concentration and can meet high-standard application requirements. By distilling and rectifying the isopropanol waste organic solution and performing ultrasonic extraction on it after rectification, the purity of the purified isopropanol is greatly improved, which can enable it to be better reused.
[0096] 3. In the purification and regeneration method of the waste organic solution of regenerated isopropanol, a high-purity isopropanol distillation process control algorithm is adopted, which significantly improves the system performance and efficiency. Through the accurate prediction of key parameters by the LSTM model and the dynamic optimization and adjustment of the operating parameters by the deep reinforcement learning (DRL) algorithm, the purification effect of isopropanol in the distillation and rectification process is guaranteed, and the purity and yield of the final product are improved. At the same time, the introduction of the high-purity isopropanol distillation process control algorithm can effectively reduce the system energy consumption, energy waste and production costs. Real-time data monitoring and feedback mechanism enhance the stability and adaptability of the system, enabling the system to respond quickly to changes and maintain stable operation. Automated control and optimization reduce manual operations, improve production efficiency, and adapt to different production conditions. In addition, precise control and optimization of the purification process also reduce the generation of waste, which helps to achieve green production and environmental protection. The comprehensive application of the high-purity isopropanol distillation process control algorithm has significantly improved the purity, output, energy consumption, stability and efficiency of this patented technology, enhancing the technical level and market competitiveness. BRIEF DESCRIPTION OF THE DRAWINGS
[0097] Figure 1 This is a flow chart for purifying isopropanol waste organic solution in an embodiment of the present invention.
[0098] Figure 2 1 is a flow chart of the control algorithm of the high-purity isopropanol distillation process in an embodiment of the present invention. DETAILED DESCRIPTION
[0099] The specific embodiments of the present invention are further described below in conjunction with the accompanying drawings. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.
[0100] As attached Figure 1 As shown, the present invention provides a method for purifying and regenerating a waste organic solution for regenerating isopropanol. The isopropanol waste organic solution is a residual solution generated after using isopropanol. The residual solution contains other impurities and reaction by-products. Chemical reagents are used to treat substances in the isopropanol waste organic solution. The specific operating steps of the method for purifying and regenerating the isopropanol waste organic solution are as follows:
[0101] Step 1: placing the isopropanol waste organic solution and distilled water in a mixing barrel at a mass ratio of 1:1, adding hydrogen peroxide solution, heating and stirring, standing and cooling to room temperature to obtain an isopropanol premix, adding activated carbon particles to the isopropanol premix, stirring for 5 minutes and standing, removing the supernatant after standing, and transferring the remaining solid into an incinerator for differentiation treatment;
[0102] Step 2: transfer the supernatant obtained in step 1 into a beaker, use a vacuum filter to vacuum filter the supernatant, use a filter membrane to filter the supernatant, the filtrate obtained is isopropanol filtrate, collect the filter membrane and filter residue and then perform heat treatment;
[0103] Step 3: transfer the isopropanol filtrate into a centrifuge tube, the volume of which does not exceed two-thirds of the volume of the centrifuge tube, place the centrifuge tube in a centrifuge for centrifugation to separate the isopropanol separation liquid and excess solid matter, and collect the solid matter into an evaporating dish for high-temperature heat treatment;
[0104] Step 4: transferring the separately collected isopropanol separated liquid into a distiller for distillation, heating and evaporating the isopropanol separated liquid, collecting the extract from the condenser into a beaker, collecting the isopropanol distillate, and then introducing the isopropanol distillate into a distillation tower, heating and distilling the isopropanol distillate, cooling the tower top vapor through a condenser to convert it into liquid, collecting different fractions according to the boiling point difference of the components, and obtaining a high-purity isopropanol distillate;
[0105] Step 5: The collected distillation filtrate is transferred into a conical flask, and an ultrasonic extraction machine is used to perform ultrasonic extraction on the isopropanol distillate. After extraction, a clean isopropanol distillate is obtained, and the purity is tested to ensure that the required purity is achieved.
[0106] In one embodiment of the present invention, the isopropanol waste organic solution is a hazardous waste, the isopropanol concentration in the isopropanol waste organic solution is 30.0%-50.0%, and the isopropanol concentration after purification is 99.3%-99.7%. When handling it, it is necessary to pay attention to comply with safety operation regulations.
[0107] In one embodiment of the present invention: in step 1, hydrogen peroxide is used to disinfect the waste organic solution of isopropyl alcohol, activated carbon particles are used to adsorb and precipitate impurities in the isopropyl alcohol premix, and the precipitate is dried and collected.
[0108] In one embodiment of the present invention: in step 2, the pore size of the filter membrane used for vacuum filtration is 1 μm to 5 μm, and after the filtration is completed, the organic matter in the remaining solid matter is put into an incinerator for thermal decomposition treatment.
[0109] In one embodiment of the present invention: in step three, the isopropanol separation liquid is placed in a centrifuge tube and the temperature when placed in a centrifuge is controlled at 25°C, the centrifugal speed is 2800r / min, the centrifugal time is 6min to 10min, and after the centrifugation is completed, the centrifuge tube is taken out after the centrifuge stops naturally to separate the isopropanol separation liquid and excess solid matter.
[0110] In one embodiment of the present invention: in step 4, the distillation operation is carried out twice, and after the isopropanol separation liquid is distilled, it is transferred to a distillation tower for multiple distillations.
[0111] In one embodiment of the present invention: in step five, a heating device is used to heat the inside of the tower to vaporize the isopropanol distillate, and the vaporized steam rises to the top of the distillation tower to collect high-purity distillate.
[0112] In one embodiment of the present invention, the extractant in the ultrasonic extraction step is ethylene glycol, and the extracting substances include heavy metal ions and organic solutions in the isopropanol distillate.
[0113] In one embodiment of the present invention: after the clean isopropanol distillate is obtained, it is placed in a light-proof container and sealed for storage at a storage temperature of 25°C to 30°C.
[0114] Example
[0115] First, a waste organic solution containing isopropanol and distilled water are placed in a stirring barrel at a mass ratio of 1:1, and a hydrogen peroxide solution is added to heat and stir. After stirring, the mixture is allowed to stand and cool to 25°C to obtain an isopropanol premixed solution. Activated carbon particles are then added thereto, and the mixture is allowed to stand after being fully stirred for 5 minutes. After standing, the supernatant is removed, and the remaining solid is transferred to an incinerator for differentiation treatment. In this step, the hydrogen peroxide solution can disinfect the waste organic solution containing isopropanol, and the activated carbon particles are used to adsorb precipitated impurities. The precipitate is dried and collected for treatment. The obtained supernatant is then transferred to a beaker, and a vacuum filter is used for vacuum filtration. The pore size of the filter membrane is 20 μm. After filtration, the filtrate, i.e., the isopropanol filtrate, is collected, and the filter membrane and filter residue are collected and heat-treated. The organic matter in the remaining solid matter is placed in an incinerator for thermal decomposition. The isopropanol filtrate is transferred to a centrifuge tube with a volume not exceeding two-thirds of the volume of the centrifuge tube. The centrifuge is placed at a temperature of 25°C and centrifuged at 2800 r / min centrifugal speed for 8 minutes. After the centrifugation is completed, it is allowed to stop naturally, and the isopropanol separation liquid and excess solid matter are separated. The solid matter is collected in an evaporating dish for high-temperature heat treatment. The isopropanol separation liquid collected separately is then transferred to a distiller for distillation. The distillation operation is performed twice. The isopropanol separation liquid is heated and evaporated. The extract is collected from the condenser into a beaker to obtain isopropanol distillate, which is then introduced into a distillation tower. It is heated by a heating device and then distilled. The top steam is cooled by a condenser and becomes a liquid. Different fractions are collected according to the boiling point difference of the components to obtain a high-purity isopropanol distillate. Finally, the collected distillation filtrate is transferred to a conical flask, and an ultrasonic extractor is used to perform ultrasonic extraction on the isopropanol distillate with ethylene glycol as an extractant, including extraction of heavy metal ions and organic solutions. After extraction, a clean isopropanol distillate is obtained, and a purity test is performed. The obtained clean isopropanol distillate is placed in a light-proof container and sealed for storage at a storage temperature of 25°C to 30°C.
[0116] Next, in step 4, the high-purity isopropanol distillation process control method is set as follows:
[0117] Step 1, data acquisition module:
[0118] Temperature sensor: installed in different sections of distillation towers and rectification towers (top, middle, bottom) to monitor the temperature of each section;
[0119] Variable T i is the temperature of the i-th position
[0120]
[0121] T means to calculate the average temperature of each location, and n is the number of sensors;
[0122] Pressure sensor: installed at the top and bottom of distillation towers and rectification towers to monitor the pressure inside the towers;
[0123] Variable P i is the pressure at the ith position
[0124]
[0125] P means to calculate the average value of pressure at each position;
[0126] Liquid level sensor: installed in the tower kettle and the middle section to monitor the liquid level;
[0127] Variable L i is the liquid level at the ith position
[0128]
[0129] It means to calculate the average value of the liquid level at each position;
[0130] Flow sensor: installed at the feed port and discharge port to monitor the flow of feed and discharge;
[0131] Variable F in is the feed flow rate, F out is the discharge flow rate;
[0132] ΔF=F in -F out
[0133] ΔF is the flow difference between the feed and the discharge, ensuring the material balance of the system;
[0134] Component concentration sensor: installed at the gas outlet at the top of the tower and the liquid outlet at the bottom of the tower to monitor the concentration of isopropanol;
[0135] Variable C top is the concentration of the top component, C bottom is the concentration of the bottom component, C avg The average concentration
[0136]
[0137] It means calculating the average component concentration at the top and bottom of the tower to evaluate the separation effect of distillation and rectification;
[0138] All the above sensors are connected to the central control system through the data acquisition module (DCS system);
[0139] Step 2, data preprocessing:
[0140] First, the data collected by the sensors are denoised separately;
[0141]
[0142] in is the denoised data at time t, is the predicted value at time t, z t is the observed value at time t, K t Kalman gain at time t;
[0143] Then perform maximum and minimum normalization on the denoised data:
[0144]
[0145] Among them, x′ is the normalized data, x is the original input data, and x max is the maximum value of the data, x min is the minimum value of the data;
[0146] Then, data fusion is performed on the data collected by different sensors. The steps are as follows:
[0147]
[0148] x f is the fused data, w i is the weight of the i-th sensor, x i is the data of the i-th sensor, n is the number of sensors;
[0149] Step 3, data prediction model:
[0150] The present invention uses the long short-term memory network (LSTM) model to perform time series prediction on the preprocessed data. Specifically, we use the preprocessed temperature, pressure, liquid level, flow rate and component concentration data to build the LSTM model and predict the changes in key parameters;
[0151] The first is the input layer, where the input data includes temperature, pressure, liquid level, flow rate, and component concentration.
[0152] X t =[T t , P t , L t , F t , C t ]
[0153] X t To group variables, T t is the temperature at time t, P t is the pressure at time t, L t is the liquid level at time t, F t The flow rate at time t, C t The concentration of the component at time t;
[0154] Next is the setting of the hidden layer, which contains multiple LSTM units. Each unit includes an input gate, a forget gate, and an output gate to capture long-term dependencies in the time series:
[0155] Input gate: i t =σ(W i ·[h t-1 , x t ]+b i )
[0156] Forget gate: f t =σ(W f ·[h t-1 , x t ]+b f ]
[0157] Output gate: o t =σ(W o ·[h t-1 ,x t ]+b o )
[0158] Cell status update:
[0159] Cell Status:
[0160] Hidden state: h t =o t tanh(C t )
[0161] Among them, i t is the input gate activation value, σ is the unactivated function, W i is the input gate weight matrix, t is the time step, h t-1 , is the hidden state at time t-1, b i is the bias vector of the input gate, f t is the forget gate activation value, W f is the weight matrix of the forget gate, b f is the bias vector of the forget gate, o t is the output gate activation value, W o is the weight matrix of the output gate, b o is the bias vector of the output gate, is the cell state update, W C is the weight matrix updated for the cell state, b C is the bias vector for updating the cell state, C t is the cell state at the current moment;
[0162] Finally, the predicted output of the output layer is
[0163] Temperature forecast:
[0164] Prediction of stress:
[0165] Liquid level prediction:
[0166] Flow control:
[0167] Control of component concentration:
[0168] in, is the predicted temperature value at the future time, is the weight matrix for temperature prediction, is the bias vector for temperature prediction, Predicted stress values for future moments is the weight matrix for pressure prediction, is the bias vector for pressure prediction, is the predicted value of the liquid level at the future moment, is the weight matrix for liquid level prediction, is the bias vector for liquid level prediction, Traffic forecast value at future time, is the weight matrix for traffic prediction, is the bias vector for flow prediction, is the predicted value of the component concentration at the future time, is the weight matrix for component concentration prediction, is the bias vector for component concentration prediction, and t is the time step;
[0169] Accurate prediction of the above parameters is crucial for optimizing the distillation and rectification processes to ensure high purity of isopropanol;
[0170] Step 4, control decision module:
[0171] Combined with the prediction results of the LSTM model, the Deep Reinforcement Learning (DRL) algorithm is used to adjust the operating parameters in the purification and regeneration method of the regenerated isopropanol waste organic solution in real time:
[0172] We chose the Deep Q Network (DQN) as the DRL algorithm. DQN evaluates the value of each state-action pair by learning the Q-value function. We use the deep learning framework TensorFlow to implement it.
[0173] The state space includes the system state at the current moment and the parameters at the future moment predicted by the LSTM model, and the action space includes the adjustable operating parameters in the control system, including:
[0174] Heating power A hp , condenser temperature A ct , Tower top pressure A tp ;
[0175] The reward function is defined based on the operation effect (isopropyl alcohol purity, energy consumption),
[0176]
[0177] r t is the reward function, α and β are weight coefficients, is the predicted component concentration at the future time, E t Energy consumption for heating and condenser;
[0178] Finally, according to the operation decision generated by DQN, the heating power, condenser temperature and tower top pressure are adjusted, and according to the feedback data, the DQN model parameters are dynamically adjusted to ensure the stability and optimal operation of the system;
[0179] Step 5: Execute the module
[0180] Based on the prediction results of the LSTM model and the optimization decision of the DRL algorithm, specific control instructions are generated. The control instructions include heating power, condenser temperature and valve opening. The control instructions are transmitted to relevant actuators through the industrial control network (PLC);
[0181] Heater Control Operation:
[0182]
[0183] Condenser control operation:
[0184]
[0185] Valve control operation:
[0186]
[0187] Among them, A hp is the heating power control instruction, k p is the proportional coefficient of the PID controller, T set is the set temperature, T actual is the actual temperature, k i is the integral coefficient of the PID controller, k d is the differential coefficient of the PID controller, A ct is the condenser temperature control instruction, Cset To set the condensation temperature, C actual is the actual condensation temperature, A vp is the valve opening control instruction, F set To set the flow rate, F actual is the actual flow rate;
[0188] Next, rely on sensors to monitor the operating status of the heater, condenser and valve in real time, obtain sensor feedback data, monitor system parameters, and if it is found that the parameters deviate from the set value, promptly issue an alarm and make adjustments.
[0189] The numerical examples are as follows:
[0190] Set temperature: 78 degrees (boiling point of isopropyl alcohol)
[0191] Set pressure: 1atm
[0192] Set condensation temperature: 20 degrees
[0193] Initial heating power: 1000W
[0194] Initial condenser temperature: 20 degrees
[0195] Initial valve opening: 50%
[0196] Initial data collected by sensors in real time:
[0197] Current temperature: 78.5 degrees
[0198] Current pressure: 1atm
[0199] Current condensation temperature: 21 degrees
[0200] Current component concentration: 85%
[0201] Use the LSTM model to predict key parameters for the next hour:
[0202] Forecasted temperature: 78.5 degrees
[0203] Predicted pressure: 1atm
[0204] Predicted condensation temperature: 19.8 degrees
[0205] Predicted component concentration: 90%
[0206] A deep reinforcement learning algorithm is used to optimize the operating parameters with the goal of maximizing component concentrations and minimizing energy consumption:
[0207] Optimized heating power: 1200W
[0208] Optimized condenser temperature: 19 degrees
[0209] Optimized valve opening: 55%
[0210] Pass the optimized control instructions to the relevant actuators:
[0211] k p is 0.1, k i is 0.05, k d 0.01
[0212] Monitor the adjusted parameters through sensors to ensure control effect:
[0213] Actual temperature: 78 degrees
[0214] Actual pressure: 1atm
[0215] Actual condensation temperature: 19 degrees
[0216] Actual component concentration: 95%
[0217] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. For those skilled in the art, various changes, modifications, substitutions and variations of these embodiments are made without departing from the principles and spirit of the present invention, and still fall within the scope of protection of the present invention.
[0218] In the description of the specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.
[0219] The above contents are merely examples and explanations of the present invention. Various modifications or additions to the specific embodiments described or replacements in similar ways by technicians in the technical field shall fall within the protection scope of the present invention as long as they do not deviate from the invention or exceed the scope defined by the claims.
[0220] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.
Claims
1. A method for purifying and regenerating a waste organic solution for regenerating isopropanol, characterized in that: The isopropanol waste organic solution is the residual solution produced after using isopropanol, and the residual solution contains other impurities and reaction by-products. The substances in the isopropanol waste organic solution are treated with chemical reagents. The specific operation steps of the waste organic solution purification and regeneration method for regenerating isopropanol are as follows: Step 1: put the waste organic solution containing isopropanol and distilled water in a mixing barrel at a mass ratio of 1:1, add hydrogen peroxide solution, heat and stir, let it stand and cool to room temperature after stirring to obtain an isopropanol premix, then add activated carbon particles to the isopropanol premix, stir it thoroughly for 5 minutes and let it stand, remove the supernatant after standing, and transfer the remaining solid into an incinerator for differentiation treatment; Step 2: transfer the supernatant obtained in step 1 into a beaker, use a vacuum filter to vacuum filter the supernatant, use a filter membrane to filter the supernatant, the filtrate obtained is isopropanol filtrate, collect the filter membrane and filter residue and then perform heat treatment; Step 3: transfer the isopropanol filtrate into a centrifuge tube, the volume of which does not exceed two-thirds of the volume of the centrifuge tube, place the centrifuge tube in a centrifuge for centrifugation to separate the isopropanol separation liquid and excess solid matter, and collect the solid matter into an evaporating dish for high-temperature heat treatment; Step 4: transferring the separately collected isopropanol separated liquid into a distiller for distillation, heating and evaporating the isopropanol separated liquid, collecting the extract from the condenser into a beaker, collecting the isopropanol distillate, and then introducing the isopropanol distillate into a distillation tower, heating and distilling the isopropanol distillate, cooling the tower top vapor through a condenser to convert it into liquid, collecting different fractions according to the boiling point difference of the components, and obtaining a high-purity isopropanol distillate; Step 5: The collected distillation filtrate is transferred into a conical flask, and an ultrasonic extraction machine is used to perform ultrasonic extraction on the isopropanol distillate. After extraction, a clean isopropanol distillate is obtained, and the purity is tested to ensure that the required purity is achieved.
2. The method for purifying and regenerating a waste organic solution for regenerating isopropanol according to claim 1, characterized in that: The isopropanol waste organic solution is a hazardous waste. The isopropanol concentration in the isopropanol waste organic solution is 30.0%-50.0%, and the isopropanol concentration after purification is 99.3%-99.7%. When handling it, it is necessary to pay attention to complying with safety operation regulations.
3. The method for purifying and regenerating a waste organic solution for regenerating isopropanol according to claim 1, characterized in that: In the step 1, hydrogen peroxide is used to disinfect the waste organic solution of isopropyl alcohol, activated carbon particles are used to adsorb and precipitate impurities in the isopropyl alcohol premix, and the precipitate is dried and collected.
4. The method for purifying and regenerating a waste organic solution for regenerating isopropanol according to claim 1, characterized in that: In the step 2, the pore size of the filter membrane used for vacuum filtration is 1 μm to 5 μm. After the filtration is completed, the organic matter in the remaining solid matter is put into an incinerator for thermal decomposition treatment.
5. The method for purifying and regenerating a waste organic solution for regenerating isopropanol according to claim 1, characterized in that: In the step three, the isopropanol separation liquid is placed in a centrifuge tube and the temperature is controlled at 25° C. when placed in a centrifuge, the centrifugal speed is 2800 r / min, and the centrifugal time is 6 min to 10 min. After the centrifugation is completed, the centrifuge tube is taken out after the centrifuge stops naturally to separate the isopropanol separation liquid and excess solid matter.
6. The method for purifying and regenerating a waste organic solution for regenerating isopropanol according to claim 1, characterized in that: In the step 4, the distillation operation is carried out twice. After the isopropanol separation liquid is distilled, it is transferred to a distillation tower for multiple distillations.
7. The method for purifying and regenerating a waste organic solution for regenerating isopropanol according to claim 1, characterized in that: In the step 5, a heating device is used to heat the inside of the tower to vaporize the isopropanol distillate, and the vaporized steam rises to the top of the distillation tower to collect high-purity distillate.
8. The method for purifying and regenerating a waste organic solution for regenerating isopropanol according to claim 1, characterized in that: The extractant in the ultrasonic extraction step is ethylene glycol, and the extracting substances include heavy metal ions and organic solutions in the isopropanol distillate.
9. The method for purifying and regenerating a waste organic solution for regenerating isopropanol according to claim 1, characterized in that: After the clean isopropanol distillate is obtained, it is placed in a light-proof container and sealed for storage at a storage temperature of 25°C to 30°C.
10. The method for purifying and regenerating a waste organic solution for regenerating isopropanol according to claim 1, characterized in that: In step 4, a high-purity isopropanol distillation process control method is added, and the specific process is as follows: Step 1, data acquisition module: Temperature sensors: installed in different sections of distillation towers and rectification towers: top, middle and bottom of towers, to monitor the temperature of each section; Variable T i is the temperature of the i-th position T means to calculate the average temperature of each location, and n is the number of sensors; Pressure sensor: installed at the top and bottom of distillation towers and rectification towers to monitor the pressure inside the towers; Variable P i is the pressure at the ith position P means to calculate the average value of pressure at each position; Liquid level sensor: installed in the tower kettle and the middle section to monitor the liquid level; Variable L i is the liquid level at the ith position It means to calculate the average value of the liquid level at each position; Flow sensor: installed at the feed port and discharge port to monitor the flow of feed and discharge; Variable F in is the feed flow rate, F out is the discharge flow rate; ΔF=F in -F out △F is the flow difference between feed and discharge, ensuring the material balance of the system; Component concentration sensor: installed at the gas outlet at the top of the tower and the liquid outlet at the bottom of the tower to monitor the concentration of isopropanol; Variable C top is the concentration of the top component, C bottom is the concentration of the bottom component, C avg The average concentration It means calculating the average component concentration at the top and bottom of the tower to evaluate the separation effect of distillation and rectification; All the above sensors are connected to the central control system through the data acquisition module; Step 2, data preprocessing: First, the data collected by the sensors are denoised separately; in is the denoised data at time t, is the predicted value at time t, z t is the observed value at time t, K t Kalman gain at time t; Then perform maximum and minimum normalization on the denoised data: Among them, x′ is the normalized data, x is the original input data, and x max is the maximum value of the data, x min is the minimum value of the data; Then, data fusion is performed on the data collected by different sensors. The steps are as follows: x f is the fused data, w i is the weight of the i-th sensor, x i is the data of the i-th sensor, n is the number of sensors; Step 3, data prediction model: Use the long short-term memory network (LSTM) model to perform time series prediction on the preprocessed data. Specifically, use the preprocessed temperature, pressure, liquid level, flow rate, and component concentration data to build the LSTM model and predict the changes in key parameters. The first is the input layer, where the input data includes temperature, pressure, liquid level, flow rate, and component concentration. X t =[T t ,P t ,L t ,F t ,C t ] Among them, X t To group variables, T t is the temperature at time t, P t is the pressure at time t, L t is the liquid level at time t, F t The flow rate at time t, C t The concentration of the component at time t; Next is the setting of the hidden layer, which contains multiple LSTM units. Each unit includes an input gate, a forget gate, and an output gate to capture long-term dependencies in the time series: Input gate: i t =σ(W i ·[h t-1 , x t ]+b i ) Forget gate: f t =σ(W f ·[h t-1 , x t ]+b f ) Output gate: o t =σ(W o ·[h t-1 , x t ]+b o ) Cell status update: Cell Status: Hidden state: h t =o t tanh(C t ) Among them, i t is the input gate activation value, σ is the unactivated function, W i is the input gate weight matrix, t is the time step, h t-1 , is the hidden state at time t-1, b i is the bias vector of the input gate, f t is the forget gate activation value, W f is the weight matrix of the forget gate, b f is the bias vector of the forget gate, o t is the output gate activation value, W o is the weight matrix of the output gate, b o is the bias vector of the output gate, is the cell state update, W C is the weight matrix updated for the cell state, b C is the bias vector for updating the cell state, C t is the cell state at the current moment; Finally, the predicted output of the output layer is Temperature forecast: Prediction of stress: Liquid level prediction: Flow control: Control of component concentration: in, is the predicted temperature value at the future time, is the weight matrix for temperature prediction, is the bias vector for temperature prediction, Predicted stress values for future moments is the weight matrix for pressure prediction, is the bias vector for pressure prediction, is the predicted value of the liquid level at the future moment, is the weight matrix for liquid level prediction, is the bias vector for liquid level prediction, Traffic forecast value at future time, is the weight matrix for traffic prediction, is the bias vector for flow prediction, is the predicted value of the component concentration at the future time, is the weight matrix for component concentration prediction, is the bias vector for component concentration prediction, and t is the time step; Accurate prediction of the above parameters is crucial for optimizing the distillation and rectification processes to ensure high purity of isopropanol; Step 4, control decision module: Combined with the prediction results of the LSTM model, the deep reinforcement learning DRL algorithm is used to adjust the operating parameters in the purification and regeneration method of the regenerated isopropanol waste organic solution in real time: We chose the Deep Q Network (DQN) as the DRL algorithm. DQN evaluates the value of each state-action pair by learning the Q-value function. This algorithm is implemented using the deep learning framework TensorFlow. The state space includes the system state at the current moment and the parameters at the future moment predicted by the LSTM model, and the action space includes the adjustable operating parameters in the control system, including: Heating power A hp , condenser temperature A ct , Tower top pressure A tp ; The reward function is defined based on the effect of the operation. Among them, r t is the reward function, α and β are weight coefficients, is the predicted component concentration at the future time, E t Energy consumption for heating and condenser; Finally, according to the operation decision generated by DQN, the heating power, condenser temperature and tower top pressure are adjusted, and according to the feedback data, the DQN model parameters are dynamically adjusted to ensure the stability and optimal operation of the system; Step 5: Execute the module Based on the prediction results of the LSTM model and the optimization decision of the DRL algorithm, specific control instructions are generated. The control instructions include heating power, condenser temperature and valve opening. The control instructions are transmitted to the relevant actuators through the industrial control network PLC. Heater Control Operation: Condenser control operation: Valve control operation: Among them, A hp is the heating power control instruction, k p is the proportional coefficient of the PID controller, T set is the set temperature, T actual is the actual temperature, k i is the integral coefficient of the PID controller, k d is the differential coefficient of the PID controller, A ct is the condenser temperature control instruction, C set To set the condensation temperature, C actual is the actual condensation temperature, A vp is the valve opening control instruction, F set To set the flow rate, F actual is the actual flow rate; Next, rely on sensors to monitor the operating status of the heater, condenser and valve in real time, obtain sensor feedback data, monitor system parameters, and if it is found that the parameters deviate from the set value, promptly issue an alarm and make adjustments.
Citation Information
Patent Citations
Isopropanol purifying method
CN106278818A
Method for separating and purifying high-purity isopropanol
CN107253901A
Separation method and separation equipment for acetone-containing isopropanol waste liquid
CN111978153A
Purification method of isopropanol
CN112142563A
Method for separating isopropanol and water by extractive distillation
CN113968773A
