An optimized control method and system based on salt refining
By constructing a trend model of ion concentration change for feedforward prediction and feedback correction, the problems of delayed addition of purifying agent and difficulty in the disposal of calcium-magnesium mud in the salt refining process were solved. This enabled precise addition of purifying agent and resource utilization of calcium-magnesium mud, and improved the system's adaptability and production stability.
Patent Information
- Application Number
- CN202611399956.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-10
- Publication Date
- 2026-11-17
AI Technical Summary
The existing salt refining process suffers from problems such as delayed addition of purifying agents, difficulty in disposing of calcium and magnesium mud, and a lack of adaptive capability in the control system.
By constructing a trend model of ion concentration change for feedforward prediction, the dosage of purifying agent can be calculated in advance. Combined with feedback correction and parameter self-correction, the precise addition of purifying agent and the resource utilization of calcium-magnesium mud can be achieved.
It achieves precise dosing of purifying agents, reduces agent consumption, reduces solid waste emissions, and enables adaptive control of the system, thereby improving production stability and resource utilization efficiency.
Smart Images

Figure CN123113547A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of salt refining technology, and in particular to an optimized control method and system based on salt refining. Background Technology
[0002] Salt refining is a crucial step in vacuum salt production. The process involves pumping raw brine from a brine well into a purification reactor, adding sodium carbonate and sodium hydroxide, causing calcium and magnesium ions in the raw brine to precipitate as calcium carbonate and magnesium hydroxide, respectively. After sedimentation and filtration, high-purity refined brine is obtained, which is then sent to an evaporator for evaporation and crystallization to produce the finished salt. This process directly affects the purity of the finished salt and the production cost.
[0003] Currently, the refining process in most salt-producing enterprises in China is still primarily manual. Operators periodically sample and analyze the calcium and magnesium content, then manually adjust the frequency of the purifying agent dosing pump based on the analysis results. The time from sampling to obtaining results is typically over 40 minutes, while the calcium and magnesium content of the raw brine fluctuates significantly with the depth of brine extraction and the grade of the raw salt. This method of measuring first and then adjusting is always a step too slow. To ensure the quality of the refined brine, operators often add too much purifying agent, resulting in wasted chemicals.
[0004] A few companies have begun using automated control systems, which involve installing online calcium and magnesium analyzers at the outlet of the purification reactor. The measured residual values are compared with set values, and the dosage of the purifying agent is automatically adjusted via a PID controller. However, this feedback control method still only addresses problems after they occur; adjustments are only initiated when the outlet concentration deviates, and the adjustment action always lags behind changes in the raw brine quality. Furthermore, once the PID parameters of the feedback control system are set, they remain unchanged for a long time. Over time, changes such as pipe scaling and valve wear occur, and the characteristics of the raw brine also slowly change, causing the control effect of the original parameters to gradually deteriorate, requiring manual recalibration.
[0005] On the other hand, the calcium-magnesium sludge produced during the purification process also presents difficulties in disposal. In the past, calcium-magnesium sludge was mostly transported off-site for landfill or stockpiled, but with the advent of environmental protection policies, many areas have banned direct landfilling, forcing companies to dry it and use it as a raw material for building materials or discharge it into salt fields. The former is costly, and the latter is greatly affected by weather, neither of which are ideal disposal methods. Summary of the Invention
[0006] Based on this, the purpose of this invention is to propose an optimized control method and system for salt refining, which aims to solve the shortcomings of existing salt refining processes, such as delayed addition of purifying agents, difficulty in disposing of calcium and magnesium mud, and lack of adaptive capability in the control system.
[0007] According to the present invention, an optimized control method based on salt refining is proposed, the method comprising:
[0008] Historical water quality data of the original brine in a continuous time series is obtained. Based on the historical water quality data, an ion concentration change trend model is constructed. The concentrations of calcium and magnesium ions in a lag period after the current moment are predicted by the ion concentration change trend model. Based on the concentration predictions, the feedforward dosage of the purifying agent is calculated and output in advance.
[0009] The actual residual concentrations of calcium and magnesium ions in the refined brine at the outlet of the purification unit are detected. The deviation between the actual residual concentration and the target residual concentration is calculated and set as the residual concentration deviation value. Based on the residual concentration deviation value, the feedforward dosage is corrected online to generate the corrected actual dosage of the purification agent.
[0010] Based on the actual dosage of the corrected purifying agent, the expected amount of calcium magnesium mud to be generated is estimated in real time through material balance. Based on the expected amount of generation, the operating parameters of the carbonization reactor are dynamically adjusted so that the output of carbonization liquid tracks the expected amount of generation.
[0011] The carbonized liquid is reinjected into the original brine inlet, the purification unit inlet, or the evaporation and crystallization unit in a certain proportion;
[0012] The cumulative integral of the deviation value is used as a feedback signal to periodically self-correct the parameters of the ion concentration change trend model.
[0013] Furthermore, the step of constructing an ion concentration change trend model based on the historical water quality data, and predicting the concentrations of calcium and magnesium ions within a lag period after the current moment using the ion concentration change trend model, includes:
[0014] Collect calcium ion concentration sequences of the original halide over multiple past sampling periods. Magnesium ion concentration sequence The trend and seasonal components of both are calculated using first-order exponential smoothing, with the following formulas: , , in, and As the smoothing coefficient, when for When t is the actual detected value of calcium ion concentration, Let be the smoothed value of the calcium ion concentration at time t. Let be the trend increment of calcium ion concentration at time t, when For When t is the actual detected value of magnesium ion concentration, Let be the smoothed value of the magnesium ion concentration at time t. The trend increment of magnesium ion concentration at time t;
[0015] The predicted concentrations of calcium and magnesium ions are calculated based on the smoothing value and trend increment, using the following formula: , among which, when Let be the smoothed value of the calcium ion concentration at time t. When the trend increment of calcium ion concentration at time t is given, The predicted concentration of calcium ions is given by [formula] within one lag period after the current time. ,when Let be the smoothed value of the magnesium ion concentration at time t. When the trend increment of magnesium ion concentration at time t is... The predicted concentration of magnesium ions is given by [formula] within a lag period following the current time. , The hysteresis period is the total residence time from when the raw brine enters the purification unit to when the solid-liquid separation is completed and the refined brine is output.
[0016] Furthermore, the step of calculating and outputting the feedforward dosage of the purifying agent in advance based on the predicted concentration includes:
[0017] The feedforward dosage of the purifying agent is calculated based on the predicted concentration value and the stoichiometric ratio, using the following formula: ,in, This refers to the feedforward dosage of the purifying agent. This represents the predicted concentration of calcium ions over a lag period following the current moment. This represents the predicted concentration of magnesium ions over a lag period following the current moment. and These are the purifier consumption coefficients for calcium ions and magnesium ions, respectively.
[0018] Furthermore, the step of correcting the feedforward dosage online based on the residual concentration deviation value to generate the corrected actual dosage of the purifying agent includes:
[0019] The deviation between the actual residual concentration and the target residual concentration is calculated using the following formula: ,in, Let be the residual concentration deviation value at time t. For the target residual concentration, This refers to the actual residual concentration measured at the export site.
[0020] Based on the deviation value, the correction compensation amount is calculated using the proportional-integral-derivative control law, as follows: , in, To correct the compensation amount, , , These are the proportional, integral, and differential coefficients, respectively. The residual ion concentration deviation at time t-1 The cumulative interval from time 0 to the current time t is the sum of all residual concentration deviation values e(t) from all past times. The integral term is... The cumulative interval is activated only when the residual concentration deviation value e(t) is continuously greater than the allowable dead zone δ. When e(t) ≤ δ, the integral term remains unchanged to avoid integral saturation.
[0021] The actual dosage of the purified agent after correction is calculated based on the correction compensation amount, using the following formula: , in, This is the corrected actual dosage of the purification agent. This refers to the feedforward injection amount.
[0022] Furthermore, the step of estimating the expected amount of calcium-magnesium sludge generated in real time through material balance based on the corrected actual dosage of the purifying agent includes:
[0023] The formula for estimating the expected amount of calcium-magnesium mud is as follows: , in, This represents the expected amount of calcium magnesium sludge to be generated. This is the corrected actual dosage of the purification agent. This represents the total calcium and magnesium concentration of the original brine. This refers to the historical calcium and magnesium precipitation rate. This represents the yield coefficient of calcium magnesium mud.
[0024] Furthermore, the step of dynamically adjusting the operating parameters of the carbonization reactor according to the expected production amount, so that the carbonization liquid output tracks the expected production amount, includes: , in, for Intake flow rate, For carbonization conversion rate, for density, These are the stoichiometric coefficients for the carbonization reaction;
[0025] The amount of carbonized liquid produced Intake flow The relationship between the reaction pressure and the dynamic response characteristics of the carbonization reactor is determined by adjusting the reaction pressure. Intake flow To ensure that the output of carbonized liquid is tracked in relation to the expected production of calcium magnesium mud. .
[0026] Furthermore, the step of proportionally reinjecting the carbonized liquid into the original brine inlet, the purification unit inlet, or the evaporation and crystallization unit includes:
[0027] When the concentration of calcium ions in the original brine is lower than the first set threshold, the proportion of the carbonized liquid reinjected into the original brine inlet is increased to replenish the calcium ion concentration to the first target range.
[0028] When the concentration of magnesium ions in the original halide is lower than the second set threshold, the proportion of the carbonized liquid reinjected into the original halide inlet is increased to replenish the magnesium ion concentration to the second target range.
[0029] Furthermore, the step of proportionally reinjecting the carbonized liquid into the original brine inlet, the purification unit inlet, or the evaporation and crystallization unit also includes:
[0030] A preset allocation ratio is set in advance to reinject the remaining carbonized liquid after calcium and magnesium supplementation back to the inlet of the evaporation crystallization unit and the purification unit according to the allocation ratio;
[0031] The preset distribution ratio is fine-tuned based on the supersaturation of the brine in the evaporator to obtain the target distribution ratio, specifically:
[0032] When the supersaturation of the brine in the evaporator is higher than the preset saturation limit, the proportion of the remaining carbonized liquid after calcium and magnesium supplementation reinjected into the evaporation and crystallization unit is increased based on the preset distribution ratio to obtain the target distribution ratio.
[0033] When the real-time supersaturation of the brine in the evaporator is lower than the preset lower limit of saturation, the proportion of the carbonized liquid reinjected into the evaporation and crystallization unit is reduced based on the first preset proportion to obtain the target proportion.
[0034] The remaining carbonized liquid after calcium and magnesium supplementation is reinjected into the inlet of the evaporation and crystallization unit and the purification unit according to the target allocation ratio.
[0035] Furthermore, the step of using the cumulative integral of the deviation value as a feedback signal to periodically self-correct the parameters of the ion concentration change trend model includes:
[0036] After each N batches run or after each time T, the average residual concentration deviation and average residual concentration during the period are statistically analyzed.
[0037] If the average residual concentration deviation value is greater than the preset upper threshold, it is determined that the trend model prediction value is systematically too high or too low, triggering automatic adjustment of the model parameters. The adjustment formula is: ,in, The adjusted smoothing coefficient. The smoothing coefficient before adjustment. To adjust the step size factor, This represents the average residual concentration deviation value.
[0038] Simultaneously, the purifier consumption coefficients for calcium ions and magnesium ions are adjusted for synchronous compensation: , ,in, , These are the adjusted purifier consumption coefficients for calcium ions and magnesium ions, respectively. , These are the purifier consumption coefficients for calcium ions and magnesium ions before adjustment. For the target residual concentration, This represents the average residual concentration.
[0039] The corrected parameters are used for feedforward prediction calculations in the next cycle.
[0040] This invention also proposes an optimization and control system based on salt refining, used to implement the above-mentioned optimization and control method based on salt refining, the system comprising:
[0041] Feedforward calculation module: used to acquire historical water quality data of raw brine in continuous time series, construct ion concentration change trend model based on the historical water quality data, and predict the concentration values of calcium ions and magnesium ions in a lag period after the current moment through the ion concentration change trend model, and calculate and output the feedforward dosage of purification agent in advance based on the concentration prediction value.
[0042] Feedback correction module: used to detect the actual residual concentration of calcium and magnesium ions in the refined brine at the outlet of the purification unit, calculate the deviation value between the actual residual concentration and the target residual concentration, set it as the residual concentration deviation value, and correct the feedforward dosage online based on the residual concentration deviation value to generate the corrected actual dosage of the purification agent;
[0043] Carbonization control module: used to estimate the expected amount of calcium magnesium mud to be generated in real time through material balance based on the actual dosage of the corrected purifying agent, and dynamically adjust the operating parameters of the carbonization reactor according to the expected amount of generated so that the output of carbonized liquid tracks the expected amount of generated.
[0044] Carbonized liquid reinjection module: used to reinject the carbonized liquid into the original brine inlet, purification unit inlet, or evaporation crystallization unit in proportion;
[0045] Parameter self-correction module: Used to periodically self-correct the parameters of the ion concentration change trend model by taking the cumulative integral of the deviation value as a feedback signal.
[0046] Compared with the prior art, the present invention has the following advantages:
[0047] This invention overcomes the lag defect of traditional feedback control by predicting the changing trend of calcium and magnesium concentration in the raw brine one lag period in advance and adjusting the dosage of purifying agent accordingly. An adjustable dead zone is set in the feedback correction stage, automatically freezing the integral term when the outlet concentration fluctuates within a small range near the target value. This avoids frequent regulator actions and reciprocating wear of the actuator caused by detection noise, thus improving system stability. By establishing a quantitative relationship between the purifying agent dosage, calcium and magnesium sludge generation, and carbonization gas demand, the output of the carbonization unit actively follows the sludge discharge rhythm of the main process, solving the matching problem between the carbonization unit and the refining process, and realizing online resource utilization of solid waste. Utilizing the feedback deviation information accumulated during operation, the smoothing coefficient and stoichiometric coefficient of the feedforward model are periodically corrected, enabling the system to adaptively track the slow changes in raw brine characteristics during long-term operation. This avoids the problem of decreased accuracy and the need for manual readjustment after a period of operation for fixed-parameter systems.
[0048] This invention achieves precise dosing of purifying agents, resource utilization of calcium-magnesium mud, and adaptive regulation of the system through synergistic control of feedforward prediction, feedback correction, parameter self-correction, and carbonization linkage, thereby reducing agent consumption and solid waste emissions.
[0049] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by means of embodiments of the invention. Attached Figure Description
[0050] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:
[0051] Figure 1 This is an overall flowchart of the optimized control method based on salt refining according to Embodiment 1 of the present invention;
[0052] Figure 2 This is a system block diagram of the optimized control system based on salt refining according to Embodiment 2 of the present invention. Detailed Implementation
[0053] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.
[0055] Example 1
[0056] Before explaining the specific control logic of this invention, a brief description of the basic process of salt refining is provided below:
[0057] The raw brine from the brine well (typically containing approximately 280-310 g / L of NaCl, along with certain amounts of calcium, magnesium, and sulfate ions) is piped into a purification reactor. Sodium carbonate and sodium hydroxide solutions are added to the reactor. The sodium carbonate reacts with calcium ions to form calcium carbonate precipitate, while the sodium hydroxide reacts with magnesium ions to form magnesium hydroxide precipitate. The resulting suspension enters a settling zone or clarification tank, where the precipitate (i.e., calcium-magnesium mud) settles to the bottom and is periodically discharged, yielding the clear liquid above, which is the refined brine. The refined brine is then further filtered through a sand filter or precision filter to remove suspended solids before being sent to an evaporator for evaporation and crystallization, yielding the finished salt product.
[0058] Throughout the process, the quality of the raw brine is not constant. It is affected by many factors, such as the depth of brine extraction, differences in the quality of the original salt deposits, switching and adjusting different brine wells, or rainwater infiltration. These factors cause the concentration of calcium and magnesium ions in the raw brine to fluctuate within a wide range, directly impacting the dosage of the purifying agent. Higher calcium and magnesium content requires more agent, while lower content requires less. However, due to the significant lag between sampling, testing, obtaining results, and adjusting the dosage, timely manual intervention is difficult. This is the core technical problem this invention aims to solve. Based on the aforementioned process flow, the optimized control method of this invention achieves precise adjustment of the purifying agent dosage through coordinated control of feedforward prediction and feedback correction.
[0059] Please see Figure 1 This invention proposes an optimized control method based on salt refining, comprising steps S101 to S105:
[0060] S101, acquire historical water quality data of the original brine in a continuous time series, construct an ion concentration change trend model based on the historical water quality data, and predict the concentration values of calcium ions and magnesium ions in a lag period after the current moment through the ion concentration change trend model, and calculate and output the feedforward dosage of the purifying agent in advance based on the concentration prediction value.
[0061] Further optionally, the step of constructing an ion concentration change trend model based on the historical water quality data, and predicting the concentrations of calcium and magnesium ions within a lag period after the current moment using the ion concentration change trend model, includes:
[0062] Collect calcium ion concentration sequences of the original halide over several past sampling periods. Magnesium ion concentration sequence The trend and seasonal components of both are calculated using first-order exponential smoothing, with the following formulas: , , in, and As the smoothing coefficient, when for When t is the actual detected value of calcium ion concentration, Let be the smoothed value of the calcium ion concentration at time t. Let be the trend increment of calcium ion concentration at time t, when For When t is the actual detected value of magnesium ion concentration, Let be the smoothed value of the magnesium ion concentration at time t. The trend increment of magnesium ion concentration at time t;
[0063] The predicted concentrations of calcium and magnesium ions are calculated based on the smoothing value and trend increment, using the following formula: , among which, when Let be the smoothed value of the calcium ion concentration at time t. When the trend increment of calcium ion concentration at time t is given, The predicted concentration of calcium ions is given by [formula] within one lag period after the current time. ,when Let be the smoothed value of the magnesium ion concentration at time t. When the trend increment of magnesium ion concentration at time t is... The predicted concentration of magnesium ions is given by [formula] within a lag period following the current time. , The hysteresis period is the total residence time from when the raw brine enters the purification unit to when the solid-liquid separation is completed and the refined brine is output.
[0064] Further optionally, the step of calculating and outputting the feedforward dosage of the purifying agent in advance based on the predicted concentration value includes:
[0065] The feedforward dosage of the purifying agent is calculated based on the predicted concentration value and the stoichiometric ratio, using the following formula: ,in, This refers to the feedforward dosage of the purifying agent. This represents the predicted concentration of calcium ions over a lag period following the current moment. This represents the predicted concentration of magnesium ions over a lag period following the current moment. and These are the purifier consumption coefficients for calcium ions and magnesium ions, respectively.
[0066] It should be noted that after the system starts, the online detector collects data on the calcium and magnesium ion concentrations of the original brine every 2 minutes and stores it in the controller's data buffer. The buffer retains the past 200 sets of data (approximately 400 minutes).
[0067] The feedforward calculation module performs a calculation every 2 minutes. Let the current time be t. First, it reads the most recent 20 sets of data (i.e., data from the past 40 minutes) from the cache and calculates the smoothing value of calcium using a recursive formula. and trend value : , , in, , The initial value can be determined through offline optimization. The method is as follows: take a segment of historical production data, and... , The grid search within the range minimizes the root mean square error of the prediction. In this embodiment, an offline tuning was performed before field commissioning to determine the optimal combination. The value is 0.42. The value is 0.18.
[0068] Lag period This can be obtained through on-site measurement. The specific procedure is as follows: A certain amount of tracer, such as concentrated NaCl solution, is instantaneously added to the feed inlet of the purification reactor, and the timing of the addition is recorded. Meanwhile, at the overflow outlet in the settlement zone, a conductivity meter was used for continuous monitoring to record the moment when the conductivity began to rise. ,but This embodiment was measured. minute.
[0069] Predicted calcium ion concentration values according to Calculation. Similarly, the predicted concentration of magnesium ions can be calculated. Feedforward dosage. In this embodiment According to chemical reaction formula The calculation yields 2.65, meaning that 2.65 kg of sodium carbonate is required to remove 1 kg of calcium ions. according to The calculation yields 3.20, meaning that 3.20 kg of sodium hydroxide is required to remove 1 kg of magnesium ions.
[0070] The feedforward calculation module sends the feedforward dosage value to the variable frequency metering pump of the dosing device as the main set value, and at the same time sends the feedforward dosage value to the feedback correction module as the reference value.
[0071] It should also be noted that this invention uses the first-order exponential smoothing method for feedforward prediction in salt refining, combined with a lag time. It enables advanced control with simple calculations, requiring no complex modeling or identification; it can be implemented with a regular PLC, making it suitable for the existing automation hardware conditions of salt production enterprises.
[0072] S102, detect the actual residual concentration of calcium and magnesium ions in the refined brine at the outlet of the purification unit, calculate the deviation value between the actual residual concentration and the target residual concentration, set it as the residual concentration deviation value, and correct the feedforward dosage online based on the residual concentration deviation value to generate the corrected actual dosage of the purification agent.
[0073] Further optionally, the step of correcting the feedforward dosage online based on the residual concentration deviation value to generate the corrected actual dosage of the purifying agent includes:
[0074] The deviation between the actual residual concentration and the target residual concentration is calculated using the following formula: ,in, Let be the residual concentration deviation value at time t. For the target residual concentration, This refers to the actual residual concentration measured at the export site.
[0075] Based on the deviation value, the correction compensation amount is calculated using the proportional-integral-derivative control law, as follows: , in, To correct the compensation amount, , , These are the proportional, integral, and differential coefficients, respectively. The residual ion concentration deviation at time t-1 The cumulative interval from time 0 to the current time t is the sum of all residual concentration deviation values e(t) from all past times. The integral term is... The cumulative interval is activated only when the residual concentration deviation value e(t) is continuously greater than the allowable dead zone δ. When e(t) ≤ δ, the integral term remains unchanged to avoid integral saturation.
[0076] The actual dosage of the purified agent after correction is calculated based on the correction compensation amount, using the following formula: , in, This is the corrected actual dosage of the purification agent. This refers to the feedforward injection amount.
[0077] It should be noted that the online detector measures the residual calcium and magnesium concentrations of the refined brine every 2 minutes. After reading this data, the feedback correction module calculates the deviation values for the calcium and magnesium channels respectively, using the following formula: Taking calcium channels as an example, the deviation value of calcium channels is: ,in, The measured residual concentration at the calcium channel outlet is used as the reference. In this embodiment, the target residual concentration of calcium ions is set at 20 mg / L.
[0078] Correction compensation amount of calcium channels , The accumulation method for the integral term is specially handled: a dead zone is set. mg / L, which is 5% of the target value, the program checks the current calcium channel deviation value every 2 minutes. The absolute value of. If If the integral terms are added normally, then the integral terms are added normally; if If the value is not accumulated, the integral term will remain unchanged from the previous value. This freezing mechanism is very useful in actual operation. Because the online analyzer itself has a measurement noise of ±0.5 mg / L, if a dead zone is not set, the integral term will fluctuate back and forth around the target point, causing frequent changes in the feed pump frequency.
[0079] , , The initial values of the three coefficients can be tuned using the critical proportional gain method. Specifically: before system commissioning, disable the integral and derivative actions, gradually increase the proportional gain until the output exhibits constant-amplitude oscillations, record the oscillation period and critical gain, and then calculate using the Ziegler-Nichols empirical formula. The tuning result in this embodiment is as follows: , , .
[0080] Magnesium channels were treated in exactly the same way as calcium channels, with a target residual concentration set at 10 mg / L and a dead zone. mg / L.
[0081] Total correction compensation ,in, This is the correction compensation amount for calcium channels. This is the correction compensation amount for the magnesium channel, and the actual dosage. The controller sends the actual dosage value to the variable frequency metering pump of the dosing device every 2 minutes.
[0082] The coordination between feedforward and feedback is as follows: when the raw brine concentration shows an upward trend, the feedforward signal begins to increase the feed rate 32 minutes in advance, while the feedback signal remains basically unchanged until the outlet concentration shows a significant deviation; when the raw brine concentration peak reaches the outlet of the purification reactor after 32 minutes, the feedback signal intervenes in time for fine-tuning. This combination of feedforward pre-adjustment and feedback fine-tuning ensures both rapid response and steady-state accuracy.
[0083] S103, based on the actual dosage of the corrected purifying agent, the expected amount of calcium magnesium mud to be generated is estimated in real time through material balance. Based on the expected amount of generation, the operating parameters of the carbonization reactor are dynamically adjusted so that the output of carbonization liquid tracks the expected amount of generation.
[0084] Optionally, the step of estimating the expected amount of calcium-magnesium sludge generated in real time through material balance based on the corrected actual dosage of the purifying agent includes:
[0085] The formula for estimating the expected amount of calcium-magnesium mud is as follows: , in, This represents the expected amount of calcium magnesium sludge to be generated. This is the corrected actual dosage of the purification agent. This represents the total calcium and magnesium concentration of the original brine. This refers to the historical calcium and magnesium precipitation rate. This represents the yield coefficient of calcium magnesium mud.
[0086] Further optionally, the step of dynamically adjusting the operating parameters of the carbonization reactor according to the expected production amount, so that the carbonization liquid output tracks the expected production amount, includes: , in, for Intake flow rate, For carbonization conversion rate, for density, These are the stoichiometric coefficients for the carbonization reaction;
[0087] The amount of carbonized liquid produced Intake flow The relationship between the reaction pressure and the dynamic response characteristics of the carbonization reactor is determined by adjusting the reaction pressure. Intake flow To ensure that the output of carbonized liquid is tracked in relation to the expected production of calcium magnesium mud. .
[0088] It should be noted that the carbonization control module reads the actual dosage from the feedback correction module. The value is read from the online detector, and the total calcium and magnesium concentration of the original brine is read simultaneously. In this embodiment, the total calcium and magnesium concentration is 850 mg / L as an example. Estimate the expected amount of calcium-magnesium mud to be generated: Wherein, 1.95 is the calcium-magnesium mud yield coefficient, which, according to actual measurements at our plant, means that 1.95 kg of dry-based calcium-magnesium mud is produced per kg of purifying agent, and 0.9 is the calcium-magnesium precipitation rate, meaning that 90% of the calcium and magnesium ions in the original brine are precipitated out.
[0089] Then calculate the required... Intake airflow: Where 1.05 is the stoichiometric coefficient for the carbonization reaction, meaning that 1.05 kg of calcium magnesium mud is required for complete carbonization of every kilogram of calcium magnesium mud. 0.80 represents the carbonization conversion rate, determined through small-scale testing, indicating that in actual production, 80% of the calcium-magnesium mud is converted into soluble bicarbonate. 1.98 represents the carbonization conversion rate under the reaction conditions. Density, in kg / m³.
[0090] The carbonization control module will Intake flow The value is sent to The mass flow controller's setting terminal automatically adjusts the valve opening to ensure the actual intake air flow tracks the set value. This method ensures that the carbonized liquid production rate always matches the calcium-magnesium sludge production rate.
[0091] The operating conditions for the carbonization reaction are: reactor temperature 45-55℃, pressure 0.15-0.30 MPa, stirring speed 30-50 rpm, single-batch reaction time 60-90 minutes, and calcium-magnesium mud slurry solid content approximately 20%. Under these conditions, the main components of the carbonization liquid are... Approximately 8~12g / L Approximately 5~8g / L, pH value 7.5~8.5.
[0092] S104, the carbonized liquid is reinjected into the original brine inlet, the purification unit inlet, or the evaporation and crystallization unit in proportion.
[0093] Further optionally, the step of proportionally reinjecting the carbonized liquid into the original brine inlet, the purification unit inlet, or the evaporation and crystallization unit includes:
[0094] When the concentration of calcium ions in the original brine is lower than the first set threshold, the proportion of the carbonized liquid reinjected into the original brine inlet is increased to replenish the calcium ion concentration to the first target range.
[0095] When the concentration of magnesium ions in the original halide is lower than the second set threshold, the proportion of the carbonized liquid reinjected into the original halide inlet is increased to replenish the magnesium ion concentration to the second target range.
[0096] Further optionally, the step of proportionally reinjecting the carbonized liquid into the original brine inlet, the purification unit inlet, or the evaporation and crystallization unit also includes:
[0097] A preset allocation ratio is set in advance to reinject the remaining carbonized liquid after calcium and magnesium supplementation back to the inlet of the evaporation crystallization unit and the purification unit according to the allocation ratio;
[0098] The preset distribution ratio is fine-tuned based on the supersaturation of the brine in the evaporator to obtain the target distribution ratio, specifically:
[0099] When the supersaturation of the brine in the evaporator is higher than the preset saturation limit, the proportion of the remaining carbonized liquid after calcium and magnesium supplementation reinjected into the evaporation and crystallization unit is increased based on the preset distribution ratio to obtain the target distribution ratio.
[0100] When the real-time supersaturation of the brine in the evaporator is lower than the preset lower limit of saturation, the proportion of the carbonized liquid reinjected into the evaporation and crystallization unit is reduced based on the first preset proportion to obtain the target proportion.
[0101] The remaining carbonized liquid after calcium and magnesium supplementation is reinjected into the inlet of the evaporation and crystallization unit and the purification unit according to the target allocation ratio.
[0102] Preferably, the first set threshold is 50 mg / L, the first target range is 80~120 mg / L, the second set threshold is 30 mg / L, and the second target range is 50~80 mg / L.
[0103] It should be noted that the carbonized solution is distributed to the three reinjection channels according to the following rules. The controller reads the original brine calcium and magnesium concentration measured by the online detector and performs conditional judgment:
[0104] If the original brine calcium ion concentration is below 50 mg / L, the carbonized liquid is reinjected into the original brine feed pipe. The goal of calcium supplementation is to increase the original brine calcium ion concentration to 80-120 mg / L. In this embodiment, when the original brine calcium ion concentration is 42 mg / L, it is calculated that it needs to be supplemented to 100 mg / L, and the required reinjection flow rate is approximately 3.5% of the original brine flow rate.
[0105] If the original magnesium halide concentration is below 30 mg / L, magnesium should be reinjected, with a target range of 50-80 mg / L. If both calcium and magnesium are low, the reinjection volume should be determined based on the weighted average of the two requirements, but the total hardness should not exceed the upper limit.
[0106] Preferably, the remaining carbonized liquid after calcium and magnesium supplementation is distributed as follows: 60%~80% is fed into the feed inlet of the purification reactor and enters the reaction zone along with the original brine; 20%~40% is fed into the evaporation and crystallization unit. The proportion reinjected into the evaporation and crystallization unit is also finely adjusted according to the supersaturation of the brine in the evaporation tank. When the brine density reaches 1.20~1.22 g / cm³, which corresponds to a supersaturation of 1.10~1.15, the reinjection ratio is appropriately increased. This utilizes the bicarbonate ions in the carbonized liquid to inhibit the excessively rapid growth of the sodium chloride crystal (100) face, which helps to obtain crystals with better sphericity.
[0107] Understandably, the bicarbonate and magnesium bicarbonate ions in the carbonized liquid reinjected into the evaporation crystallization unit are used to suppress the growth rate of the (100) crystal facet of sodium chloride crystals and promote the exposure of the (111) crystal facet, which can guide the sodium chloride crystals to grow towards a high sphericity direction.
[0108] S105, the cumulative integral of the deviation value is used as a feedback signal to periodically self-correct the parameters of the ion concentration change trend model.
[0109] Further optionally, the step of using the cumulative integral of the deviation value as a feedback signal to periodically self-correct the parameters of the ion concentration change trend model includes:
[0110] After each N batches run or after each time T, the average residual concentration deviation and average residual concentration during the period are statistically analyzed.
[0111] If the average residual concentration deviation value is greater than the preset upper threshold, it is determined that the trend model prediction value is systematically too high or too low, triggering automatic adjustment of the model parameters. The adjustment formula is: ,in, The adjusted smoothing coefficient. The smoothing coefficient before adjustment. To adjust the step size factor, This represents the average residual concentration deviation value.
[0112] Simultaneously, the purifier consumption coefficients for calcium ions and magnesium ions are adjusted for synchronous compensation: , ,in, , These are the adjusted purifier consumption coefficients for calcium ions and magnesium ions, respectively. , These are the purifier consumption coefficients for calcium ions and magnesium ions before adjustment. For the target residual concentration, This represents the average residual concentration.
[0113] The corrected parameters are applied to the feedforward prediction calculation for the next cycle.
[0114] This embodiment describes the working process of the parameter self-correction module as follows.
[0115] After the system had been running continuously for 15 days following its commissioning, the average calcium and magnesium concentration of the original brine gradually increased from 500-700 mg / L to 700-900 mg / L (based on...) because the brine well from which the original brine was sourced changed its extraction layer during this period. (Calculation). The feedforward calculation module uses... and , The values were still the initial values set before commissioning, but gradually the predicted values were lower than expected, as evidenced by the measured values of residual calcium concentration at the outlet consistently being about 3 to 5 mg / L higher than the target value.
[0116] The parameter self-correction module calculates the average deviation for the day every 24 hours. The calculation is for the 5th day. mg / L, mg / L, weighted average The concentration of mg / L exceeds the preset upper threshold of 2.5 mg / L. Calculated at 12.5% of the target value of 20 mg / L, this can be adjusted based on operational experience. Module trigger parameter correction: The calculated result is too large. It should be noted that the correction formula used in this embodiment needs to have a limited correction range in practical applications to avoid system oscillation caused by excessive adjustment at once. The maximum single correction range set in the actual program is ±30%, therefore... Actual take .
[0117] , The maximum correction range per instance is also limited to ±20%, and the actual value used in this case is... .
[0118] And so on, Adjust synchronously in the same way.
[0119] The revised parameters were put into use starting from day 6. The average deviation calculated on day 7 decreased to mg / L, mg / L, weighted average mg / L, falling back below the threshold. Thereafter, until day 15, the system did not trigger parameter correction, indicating that the corrected model had adapted to the new brine characteristics.
[0120] Since the source of raw brine for salt production enterprises is often not constant, factors such as changes in brine extraction depth, brine blending from different salt fields, and freshwater infiltration during the rainy season can all cause a slow drift in the quality of the raw brine. Through a parameter self-correction mechanism, the system can automatically track this drift, eliminating the need for process technicians to readjust the settings periodically.
[0121] It should be noted that the self-correction mechanism is limited to the parameters (smoothing coefficients) of the feedforward model. and stoichiometric coefficients , ), PID parameters of the feedback correction stage ( , , The parameters remain unchanged. This is because the feedforward model handles predictable, slowly changing disturbances, such as long-term drift in the quality of the raw brine, while the feedback correction handles unpredictable, rapidly changing disturbances, such as instantaneous fluctuations in the dosing pump and detection noise. This clear division of labor avoids system instability that might result from adjusting all parameters simultaneously.
[0122] Example 2
[0123] Please see Figure 2 This invention proposes an optimized control system based on salt refining, the system comprising:
[0124] Feedforward calculation module: used to acquire historical water quality data of raw brine in continuous time series, construct ion concentration change trend model based on the historical water quality data, and predict the concentration values of calcium ions and magnesium ions in a lag period after the current moment through the ion concentration change trend model, and calculate and output the feedforward dosage of purification agent in advance based on the concentration prediction value.
[0125] Feedback correction module: used to detect the actual residual concentration of calcium and magnesium ions in the refined brine at the outlet of the purification unit, calculate the deviation value between the actual residual concentration and the target residual concentration, set it as the residual concentration deviation value, and correct the feedforward dosage online based on the residual concentration deviation value to generate the corrected actual dosage of the purification agent;
[0126] Carbonization control module: used to estimate the expected amount of calcium magnesium mud to be generated in real time through material balance based on the actual dosage of the corrected purifying agent, and dynamically adjust the operating parameters of the carbonization reactor according to the expected amount of generated so that the output of carbonized liquid tracks the expected amount of generated.
[0127] Carbonized liquid reinjection module: used to reinject the carbonized liquid into the original brine inlet, purification unit inlet, or evaporation crystallization unit in proportion;
[0128] Parameter self-correction module: Used to periodically self-correct the parameters of the ion concentration change trend model by taking the cumulative integral of the deviation value as a feedback signal.
[0129] Further optionally, the feedforward calculation module is also used for:
[0130] Collect calcium ion concentration sequences of the original halide over multiple past sampling periods. Magnesium ion concentration sequence The trend and seasonal components of both are calculated using first-order exponential smoothing, with the following formulas: , , in, and As the smoothing coefficient, when for When t is the actual detected value of calcium ion concentration, Let be the smoothed value of the calcium ion concentration at time t. Let be the trend increment of calcium ion concentration at time t, when For When t is the actual detected value of magnesium ion concentration, Let be the smoothed value of the magnesium ion concentration at time t. The trend increment of magnesium ion concentration at time t;
[0131] The predicted concentrations of calcium and magnesium ions are calculated based on the smoothing value and trend increment, using the following formula: , among which, when Let be the smoothed value of the calcium ion concentration at time t. When the trend increment of calcium ion concentration at time t is given, The predicted concentration of calcium ions is given by [formula] within one lag period after the current time. ,when Let be the smoothed value of the magnesium ion concentration at time t. When the trend increment of magnesium ion concentration at time t is... The predicted concentration of magnesium ions is given by [formula] within a lag period following the current time. , The hysteresis period is the total residence time from when the raw brine enters the purification unit to when the solid-liquid separation is completed and the refined brine is output.
[0132] Further optionally, the feedforward calculation module is also used for:
[0133] The feedforward dosage of the purifying agent is calculated based on the predicted concentration value and the stoichiometric ratio, using the following formula: ,in, This refers to the feedforward dosage of the purifying agent. This represents the predicted concentration of calcium ions over a lag period following the current moment. This represents the predicted concentration of magnesium ions over a lag period following the current moment. and These are the purifier consumption coefficients for calcium ions and magnesium ions, respectively.
[0134] Further optionally, the feedback correction module is also used for:
[0135] The deviation between the actual residual concentration and the target residual concentration is calculated using the following formula: ,in, Let be the residual concentration deviation value at time t. For the target residual concentration, This refers to the actual residual concentration measured at the export site.
[0136] Based on the deviation value, the correction compensation amount is calculated using the proportional-integral-derivative control law, as follows: , in, To correct the compensation amount, , , These are the proportional, integral, and differential coefficients, respectively. The residual ion concentration deviation at time t-1 The cumulative interval from time 0 to the current time t is the sum of all residual concentration deviation values e(t) from all past times. The integral term is... The cumulative interval is activated only when the residual concentration deviation value e(t) is continuously greater than the allowable dead zone δ. When e(t) ≤ δ, the integral term remains unchanged to avoid integral saturation.
[0137] The actual dosage of the purified agent after correction is calculated based on the correction compensation amount, using the following formula: , in, This is the corrected actual dosage of the purification agent. This refers to the feedforward injection amount.
[0138] Further optionally, the carbonization control module is also used for:
[0139] The formula for estimating the expected amount of calcium-magnesium mud is as follows: , in, This represents the expected amount of calcium magnesium sludge to be generated. This is the corrected actual dosage of the purification agent. This represents the total calcium and magnesium concentration of the original brine. This refers to the historical calcium and magnesium precipitation rate. This represents the yield coefficient of calcium magnesium mud.
[0140] Further optionally, the carbonization control module is also used for: , in, for Intake flow rate, For carbonization conversion rate, for density, These are the stoichiometric coefficients for the carbonization reaction;
[0141] The amount of carbonized liquid produced Intake flow The relationship between the reaction pressure and the dynamic response characteristics of the carbonization reactor is determined by adjusting the reaction pressure. Intake flow To ensure that the output of carbonized liquid is tracked in relation to the expected production of calcium magnesium mud. .
[0142] Further optionally, the carbonized liquid reinjection module is also used for:
[0143] When the concentration of calcium ions in the original brine is lower than the first set threshold, the proportion of the carbonized liquid reinjected into the original brine inlet is increased to replenish the calcium ion concentration to the first target range.
[0144] When the concentration of magnesium ions in the original halide is lower than the second set threshold, the proportion of the carbonized liquid reinjected into the original halide inlet is increased to replenish the magnesium ion concentration to the second target range.
[0145] Further optionally, the carbonized liquid reinjection module is also used for:
[0146] A preset allocation ratio is set in advance to reinject the remaining carbonized liquid after calcium and magnesium supplementation back to the inlet of the evaporation crystallization unit and the purification unit according to the allocation ratio;
[0147] The preset distribution ratio is fine-tuned based on the supersaturation of the brine in the evaporator to obtain the target distribution ratio, specifically:
[0148] When the supersaturation of the brine in the evaporator is higher than the preset saturation limit, the proportion of the remaining carbonized liquid after calcium and magnesium supplementation reinjected into the evaporation and crystallization unit is increased based on the preset distribution ratio to obtain the target distribution ratio.
[0149] When the real-time supersaturation of the brine in the evaporator is lower than the preset lower limit of saturation, the proportion of the carbonized liquid reinjected into the evaporation and crystallization unit is reduced based on the first preset proportion to obtain the target proportion.
[0150] The remaining carbonized liquid after calcium and magnesium supplementation is reinjected into the inlet of the evaporation and crystallization unit and the purification unit according to the target allocation ratio.
[0151] Further optionally, the parameter self-correction module is also used for:
[0152] After each N batches run or after each time T, the average residual concentration deviation and average residual concentration during the period are statistically analyzed.
[0153] If the average residual concentration deviation value is greater than the preset upper threshold, it is determined that the trend model prediction value is systematically too high or too low, triggering automatic adjustment of the model parameters. The adjustment formula is: ,in, The adjusted smoothing coefficient. The smoothing coefficient before adjustment. To adjust the step size factor, This represents the average residual concentration deviation value.
[0154] Simultaneously, the purifier consumption coefficients for calcium ions and magnesium ions are adjusted for synchronous compensation: , ,in, , These are the adjusted purifier consumption coefficients for calcium ions and magnesium ions, respectively. , These are the purifier consumption coefficients for calcium ions and magnesium ions before adjustment. For the target residual concentration, This represents the average residual concentration.
[0155] The corrected parameters are used for feedforward prediction calculations in the next cycle.
[0156] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.
Claims
1. An optimized control method based on salt refining, characterized in that, The method includes: Historical water quality data of the original brine in a continuous time series is obtained. Based on the historical water quality data, an ion concentration change trend model is constructed. The concentrations of calcium and magnesium ions in a lag period after the current moment are predicted by the ion concentration change trend model. Based on the concentration predictions, the feedforward dosage of the purifying agent is calculated and output in advance. The actual residual concentrations of calcium and magnesium ions in the refined brine at the outlet of the purification unit are detected. The deviation between the actual residual concentration and the target residual concentration is calculated and set as the residual concentration deviation value. Based on the residual concentration deviation value, the feedforward dosage is corrected online to generate the corrected actual dosage of the purification agent. Based on the actual dosage of the corrected purifying agent, the expected amount of calcium magnesium mud to be generated is estimated in real time through material balance. Based on the expected amount of generation, the operating parameters of the carbonization reactor are dynamically adjusted so that the output of carbonization liquid tracks the expected amount of generation. The carbonized liquid is reinjected into the original brine inlet, the purification unit inlet, or the evaporation and crystallization unit in a certain proportion; The cumulative integral of the deviation value is used as a feedback signal to periodically self-correct the parameters of the ion concentration change trend model.
2. The optimized control method based on salt refining according to claim 1, characterized in that, The step of constructing an ion concentration change trend model based on the historical water quality data, and predicting the concentrations of calcium and magnesium ions within a lag period after the current moment using the ion concentration change trend model, includes: Collect calcium ion concentration sequences of the original halide over multiple past sampling periods. Magnesium ion concentration sequence The trend and seasonal components of both are calculated using first-order exponential smoothing, with the following formulas: , , in, and As the smoothing coefficient, when for When t is the actual detected value of calcium ion concentration, Let be the smoothed value of the calcium ion concentration at time t. Let be the trend increment of calcium ion concentration at time t, when For When t is the actual detected value of magnesium ion concentration, Let be the smoothed value of the magnesium ion concentration at time t. The trend increment of magnesium ion concentration at time t; The predicted concentrations of calcium and magnesium ions are calculated based on the smoothing value and trend increment, using the following formula: , among which, when Let be the smoothed value of the calcium ion concentration at time t. When the trend increment of calcium ion concentration at time t is given, The predicted concentration of calcium ions is given by [formula] within one lag period after the current time. ,when Let be the smoothed value of the magnesium ion concentration at time t. When the trend increment of magnesium ion concentration at time t is... The predicted concentration of magnesium ions is given by [formula] within a lag period following the current time. , The hysteresis period is the total residence time from when the raw brine enters the purification unit to when the solid-liquid separation is completed and the refined brine is output.
3. The optimized control method based on salt refining according to claim 2, characterized in that, The step of calculating and outputting the feedforward dosage of the purifying agent in advance based on the predicted concentration includes: The feedforward dosage of the purifying agent is calculated based on the predicted concentration value and the stoichiometric ratio, using the following formula: ,in, This refers to the feedforward dosage of the purifying agent. This represents the predicted concentration of calcium ions over a lag period following the current moment. This represents the predicted concentration of magnesium ions over a lag period following the current moment. and These are the purifier consumption coefficients for calcium ions and magnesium ions, respectively.
4. The optimized control method based on salt refining according to claim 3, characterized in that, The step of correcting the feedforward dosage online based on the residual concentration deviation value to generate the corrected actual dosage of the purifying agent includes: The deviation between the actual residual concentration and the target residual concentration is calculated using the following formula: ,in, Let be the residual concentration deviation value at time t. For the target residual concentration, This refers to the actual residual concentration measured at the export site. Based on the deviation value, the correction compensation amount is calculated using the proportional-integral-derivative control law, as follows: , in, To correct the compensation amount, , , These are the proportional, integral, and differential coefficients, respectively. The deviation of residual ion concentration at time t-1. The cumulative interval from time 0 to the current time t is the sum of all residual concentration deviation values e(t) from all past times. The integral term is... The cumulative interval is activated only when the residual concentration deviation value e(t) is continuously greater than the allowable dead zone δ. When e(t) ≤ δ, the integral term remains unchanged to avoid integral saturation. The actual dosage of the purified agent after correction is calculated based on the correction compensation amount, using the following formula: , in, This is the corrected actual dosage of the purification agent. This refers to the feedforward injection amount.
5. The optimized control method based on salt refining according to claim 1, characterized in that, The step of estimating the expected amount of calcium-magnesium sludge generated in real time through material balance based on the corrected actual dosage of the purifying agent includes: The formula for estimating the expected amount of calcium-magnesium mud is as follows: , in, This represents the expected amount of calcium magnesium sludge to be produced. This is the corrected actual dosage of the purification agent. This represents the total calcium and magnesium concentration of the original brine. This refers to the historical calcium and magnesium precipitation rate. This represents the yield coefficient of calcium magnesium mud.
6. The optimized control method based on salt refining according to claim 5, characterized in that, The step of dynamically adjusting the operating parameters of the carbonization reactor according to the expected production amount, so that the carbonization liquid output tracks the expected production amount, includes: , in, for Intake flow rate, For carbonization conversion rate, for density, These are the stoichiometric coefficients for the carbonization reaction; The amount of carbonized liquid produced Intake flow The relationship between the reaction pressure and the dynamic response characteristics of the carbonization reactor is determined by adjusting the reaction pressure. Intake flow To ensure that the output of carbonized liquid is tracked in relation to the expected production of calcium magnesium mud. .
7. The optimized control method based on salt refining according to claim 1, characterized in that, The step of proportionally reinjecting the carbonized liquid into the original brine inlet, the purification unit inlet, or the evaporation and crystallization unit includes: When the concentration of calcium ions in the original brine is lower than the first set threshold, the proportion of the carbonized liquid reinjected into the original brine inlet is increased to replenish the calcium ion concentration to the first target range. When the original halide magnesium ion concentration is lower than the second set threshold, the proportion of the carbonized liquid reinjected into the original halide inlet is increased to replenish the magnesium ion concentration to the second target range.
8. The optimized control method based on salt refining as described in claim 7, characterized in that, The step of proportionally reinjecting the carbonized liquid into the original brine inlet, the purification unit inlet, or the evaporation and crystallization unit further includes: A preset allocation ratio is set in advance to reinject the remaining carbonized liquid after calcium and magnesium supplementation back to the inlet of the evaporation crystallization unit and the purification unit according to the allocation ratio; The preset distribution ratio is fine-tuned based on the supersaturation of the brine in the evaporator to obtain the target distribution ratio, specifically: When the supersaturation of the brine in the evaporator is higher than the preset saturation limit, the proportion of the remaining carbonized liquid after calcium and magnesium supplementation reinjected into the evaporation and crystallization unit is increased based on the preset distribution ratio to obtain the target distribution ratio. When the real-time supersaturation of the brine in the evaporator is lower than the preset lower limit of saturation, the proportion of the carbonized liquid reinjected into the evaporation and crystallization unit is reduced based on the first preset proportion to obtain the target proportion. The remaining carbonized liquid after calcium and magnesium supplementation is reinjected into the inlet of the evaporation and crystallization unit and the purification unit according to the target allocation ratio.
9. The optimized control method based on salt refining according to claim 1, characterized in that, The step of using the cumulative integral of the deviation value as a feedback signal to periodically self-correct the parameters of the ion concentration change trend model includes: After each N batches run or after each time T, the average residual concentration deviation and average residual concentration during the period are statistically analyzed. If the average residual concentration deviation value is greater than the preset upper threshold, it is determined that the trend model prediction value is systematically too high or too low, triggering automatic adjustment of the model parameters. The adjustment formula is: ,in, The adjusted smoothing coefficient. The smoothing coefficient before adjustment. To adjust the step size factor, This represents the average residual concentration deviation value. Simultaneously, the purifier consumption coefficients for calcium ions and magnesium ions are adjusted for synchronous compensation: , ,in, , These are the adjusted purifier consumption coefficients for calcium ions and magnesium ions, respectively. , These are the purifier consumption coefficients for calcium ions and magnesium ions before adjustment. For the target residual concentration, This represents the average residual concentration. The corrected parameters are used for feedforward prediction calculations in the next cycle.
10. An optimization and control system based on salt refining, used to implement the optimization and control method based on salt refining as described in any one of claims 1 to 9, characterized in that, The system includes: Feedforward calculation module: used to acquire historical water quality data of raw brine in continuous time series, construct ion concentration change trend model based on the historical water quality data, and predict the concentration values of calcium ions and magnesium ions in a lag period after the current moment through the ion concentration change trend model, and calculate and output the feedforward dosage of purification agent in advance based on the concentration prediction value. Feedback correction module: used to detect the actual residual concentration of calcium and magnesium ions in the refined brine at the outlet of the purification unit, calculate the deviation value between the actual residual concentration and the target residual concentration, set it as the residual concentration deviation value, and correct the feedforward dosage online based on the residual concentration deviation value to generate the corrected actual dosage of the purification agent; Carbonization control module: used to estimate the expected amount of calcium magnesium mud to be generated in real time through material balance based on the actual dosage of the corrected purifying agent, and dynamically adjust the operating parameters of the carbonization reactor according to the expected amount of generated so that the output of carbonized liquid tracks the expected amount of generated. Carbonized liquid reinjection module: used to reinject the carbonized liquid into the original brine inlet, purification unit inlet, or evaporation crystallization unit in proportion; Parameter self-correction module: Used to periodically self-correct the parameters of the ion concentration change trend model by taking the cumulative integral of the deviation value as a feedback signal.
