Intelligent control method for adding amount of activated carbon in glycerol decolorization section

Abstract

The present application belongs to the technical field of chemical process control, and relates to an intelligent control method for the addition amount of activated carbon in a glycerol decolorization section, comprising the following steps: collecting the crude glycerol inlet chroma, the outlet chroma after decolorization, the crude glycerol feed flow, the inlet and outlet pressure difference of the plate and frame filter press and the actual activated carbon addition mass flow of the glycerol decolorization section in real time through an industrial control system, constructing a first-in-first-out data queue by using a historical database, and performing time sequence alignment on the crude glycerol inlet chroma, the actual activated carbon addition mass flow and the outlet chroma after decolorization according to the lag time of the decolorization kettle; based on the historical data and the current data after time sequence alignment; through a feedforward-feedback cascade control architecture, while eliminating the lagging influence, the differential suppression logic is used to realize the active defensive protection of the filter press.

Description

Technical Field

[0001] This invention belongs to the field of chemical process control technology and relates to an intelligent control method for the amount of activated carbon added in the glycerol decolorization process. It is applied to the glycerol decolorization process to solve the problems of large lag, fluctuation in activated carbon efficiency, and easy clogging of filter press, so as to accurately control the amount, ensure quality, reduce costs, and protect equipment. Background Technology

[0002] Glycerin decolorization is a crucial step in the crude glycerin refining production line. Its core process involves adding powdered activated carbon in a specific ratio to a decolorization vessel containing crude glycerin, utilizing the physical adsorption properties of the activated carbon to remove pigments and colloidal impurities from the glycerin. In this stage, the precision of the activated carbon dosage directly affects product quality and production costs. Insufficient dosage will result in the glycerin's color failing to meet national standards, leading to product downgrading; excessive dosage not only increases the cost of raw materials but also causes the filter cake layer in the subsequent plate and frame filter press to thicken too quickly, shortening the filtration cycle.

[0003] Patent application CN105037093A discloses a method for decolorizing glycerol, which uses static adsorption decolorization with powdered activated carbon. Activated carbon is added at a fixed ratio of 0.3–0.5% of the crude glycerol weight, and the mixture is heated, stirred, and then filtered to obtain decolorized glycerol. However, this method relies on a fixed ratio and manual experience, lacks online detection and hysteresis compensation, and cannot adapt to fluctuations in activated carbon efficiency, easily leading to dosage imbalances and the risk of filter press clogging. Patent application CN120590239A discloses a deep purification process for high-purity glycerol. This process includes a decolorization unit, using preset parameters to quantitatively add activated carbon, combined with filtration to achieve decolorization. It focuses on optimizing the overall purification process. However, this method uses fixed-value control of the dosage, lacks real-time adsorption efficiency inversion and time-series alignment, does not consider the large hysteresis characteristics of the decolorization vessel, and lacks a safety protection mechanism for the filter press. Patent application CN103965019A discloses a purification process for glycerol. This method uses distillation as the core purification method, supplemented by activated carbon adsorption for decolorization. The addition of activated carbon is manually controlled, focusing on improving the purity and yield of glycerol. However, the decolorization step is an auxiliary process, and the addition is not intelligently controlled. It does not solve the problems of large time delay and adsorption efficiency fluctuation, and it lacks pressure differential prediction and active protection logic for the filter press.

[0004] In existing technologies, the addition of activated carbon mainly relies on manual experience or simple PID feedback control. However, because decolorization reactors typically have large volumes, the residence time of materials within the reactor can be as long as 30 to 45 minutes. This results in a significant pure lag between the outlet color detection value and the inlet addition operation. Conventional PID algorithms struggle to handle such large lags, easily causing system oscillations, i.e., fluctuating between insufficient addition leading to excessive color and excessive addition resulting in waste. Furthermore, the adsorption efficiency of activated carbon is significantly affected by the batch of raw materials, the degree of moisture absorption, and the temperature of glycerol. Existing control methods usually assume that the adsorption efficiency is constant, and cannot automatically adjust the addition ratio when the quality of activated carbon fluctuates.

[0005] More seriously, the existing control logic lacks effective consideration of equipment safety boundaries. As a downstream device in the decolorization process, the plate and frame filter press experiences an exponential increase in operating resistance as the filter cake thickness increases. In the later stages of filter press operation, if the control system continuously increases the activated carbon dosage to forcibly reduce color, it can cause a sharp rise in the filter press's feed pressure, easily leading to safety accidents such as pressure buildup, filter cloth rupture, and even frame deformation. Currently, manual intervention by operators monitoring the pressure gauge is crucial, which carries the risk of delayed response and human error. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the prior art and solve the technical problems of control oscillation caused by large time delay, inability to adapt to the fluctuation of activated carbon performance, and easy clogging of filter press due to lack of equipment safety constraints. The invention provides an intelligent control method for the amount of activated carbon added in the glycerol decolorization process.

[0007] To achieve the above-mentioned objective, this invention provides an intelligent control method for the amount of activated carbon added in a glycerol decolorization process, comprising the following steps: S1, through the industrial control system, collects in real time the crude glycerol inlet color, the decolorized outlet color, the crude glycerol feed flow rate, the inlet and outlet pressure difference of the plate and frame filter press, and the actual mass flow rate of activated carbon added in the glycerol decolorization section. It constructs a first-in-first-out data queue using the historical database and aligns the crude glycerol inlet color, the actual mass flow rate of activated carbon added, and the decolorized outlet color according to the lag time of the decolorization kettle. S2, based on historical data and current data after time-series alignment, utilizes the logarithmic decay law of liquid phase adsorption kinetics to reverse calculate the actual decolorization capacity of unit mass activated carbon under current working conditions, and obtains the unit adsorption efficiency index. S3. Based on the current measured differential pressure and differential pressure change rate of the plate and frame filter press, a safety inhibition factor is constructed. The safety inhibition factor identifies the clogging trend of the filter press by introducing the differential pressure. When the differential pressure rises rapidly, the value of the safety inhibition factor is reduced. S4 calculates the activated carbon dosage setpoint based on the current crude glycerol feed flow rate, the target color required by the process, the unit adsorption efficiency index after smoothing treatment, and the equipment safety inhibition factor, and sends the activated carbon dosage setpoint as a control command to the feeding equipment for execution.

[0008] This invention solves the problem of severe material lag caused by large-volume reactors in the glycerol decolorization process by introducing a time-series alignment technology based on the decolorization reactor volume and real-time flow rate. Compared with traditional manual experience adjustment or conventional PID control, it can ensure that the collected crude glycerol inlet information and the lag output outlet information are logically precisely matched, eliminating frequent oscillations in the control system when dealing with objects with large time delays, and achieving zero-delay response in application scenarios with fluctuations in feed flow rate and concentration, greatly improving the automation and stability level of the decolorization process.

[0009] The formula for calculating the unit adsorption efficiency index described in this invention is as follows:

[0010] In the formula, for The unit adsorption efficiency index obtained by time-inversion; The historical entry chromaticity prior to the lag time; The current export color value; The density of activated carbon in the decolorization kettle before the lag time; This is a numerical stability constant used to prevent the denominator from being zero.

[0011] This invention utilizes liquid-phase adsorption kinetics to perform online inversion of the decolorization capacity of activated carbon, transforming the adsorption performance, which is difficult to measure directly, into a quantifiable efficiency index. Compared with traditional control modes that rely on fixed-ratio addition or simply on outlet feedback, this invention can keenly capture the efficiency decay of activated carbon caused by batch differences, moisture, or changes in ambient temperature. This gives the system the ability to adapt to fluctuations in raw material quality, ensuring that the color of the outlet glycerol remains consistently within the standard even in complex chemical production environments.

[0012] The calculation formula for the equipment safety inhibition factor described in this invention is as follows:

[0013] In the formula, As a safety inhibitory factor for equipment; This is the current measured pressure difference; This is the maximum allowable operating pressure differential of the filter press. This represents the rate of change of pressure difference; It is a static pressure-weighted index; These are the dynamic rate weighting coefficients; when the calculated values ​​are... If the value is less than 0, assign it the value 0.

[0014] This invention, by constructing a safety suppression factor that includes a differential pressure term, endows the control system with the function of trend prediction. Compared with existing defensive measures that only alarm and shut down after overpressure, it can identify the rate trend of filter cake thickening inside the filter press. In the early stage of blockage when the differential pressure rises rapidly, it smoothly reduces the dosage through nonlinear suppression logic, realizing active defensive protection for the plate and frame filter press. This effectively avoids safety accidents such as pressure buildup, filter cloth rupture, and frame deformation, and extends the filtration cycle.

[0015] The calculation formula for the activated carbon dosage setting value described in this invention is as follows:

[0016] In the formula, Set the value for activated carbon dosage; This represents the current measured feed flow rate; This is the currently measured inlet chromaticity; The target colorimetric value required by the process; The unit adsorption efficiency index after smoothing treatment; It is a device safety inhibition factor.

[0017] This invention forms a precise dosage control architecture by decoupling adsorption efficiency feedback, material load feedforward, and equipment safety constraints through multivariate calculations. Compared with feedback adjustment of a single indicator, this invention can dynamically correct the dosage command in real time according to the equipment safety status while ensuring the process target color. In application scenarios that guarantee the quality of glycerol refining, it maximizes the optimization of activated carbon raw material consumption and achieves the optimal balance between production efficiency and economic cost.

[0018] The method for calculating the lag time described in this invention is as follows: divide the effective volume of the decolorizing reactor by the current crude glycerol feed flow rate; the specific operation of time alignment is as follows: extract the historical inlet color and historical activated carbon addition mass flow rate at the current moment minus the lag time from the first-in-first-out data queue, and combine them with the decolorized outlet color at the current moment to form a calculation group.

[0019] This invention uses dynamic flow velocity correlation to calculate lag time, enabling the system to automatically adjust with the production rhythm. Compared with the lag compensation method using a fixed time constant, it can scale the time alignment window in real time according to the increase or decrease of production load, ensuring real-time synchronization between the historical data queue and the current production status, and improving the accuracy of online model calculation.

[0020] The method for calculating the density of activated carbon in the decolorization kettle of the present invention is as follows: divide the actual mass flow rate of activated carbon added before the lag time by the crude glycerol feed flow rate before the lag time.

[0021] The differential pressure change rate described in this invention is obtained by the differential method. The calculation formula is the measured differential pressure at the current moment minus the measured differential pressure at the previous sampling moment, divided by the sampling time interval.

[0022] The step of sending the activated carbon dosage setpoint as a control command to the feeding device for execution, as described in this invention, further includes: Based on the flow rate and frequency calibration curve of the feeding equipment, the calculated activated carbon dosage setpoint is converted into the operating frequency command of the frequency converter; The flow rate and frequency calibration curves have a linear relationship. Specifically, the frequency command equals the product of the calibration slope and the activated carbon dosage setting plus the intercept, to ensure that the control signal matches the physical actuator.

[0023] The method described in this invention further includes the following anomaly protection steps: The value of the equipment safety inhibition factor is monitored in real time. When the equipment safety inhibition factor is lower than the preset safety threshold for a continuous preset time, the plate and frame filter press is determined to be in a state of serious blockage risk. The system automatically outputs an interlock shutdown signal, stopping the operation of the feeding equipment and triggering an audible and visual alarm to prompt the operator to clean or replace the filter cloth.

[0024] The hysteresis time of the decolorization kettle described in this invention The calculation formula is: ,in, The effective volume of the decolorization kettle, This represents the crude glycerol feed flow rate.

[0025] Compared with existing technologies, the beneficial effects of this invention are as follows: It constructs a control architecture based on time-series alignment performance inversion and differential safety constraints. Its core lies in breakthroughs in two dimensions: First, in the time dimension, it utilizes first-in-first-out queue alignment technology to solve the problem of large-delay input-output pairing that is common in continuous flow chemical processes, making it possible to calculate unmeasurable process parameters online; Second, in the safety dimension, it proposes a nonlinear differential pressure suppression algorithm, upgrading the traditional overpressure alarm to trend suppression, which can smoothly intervene in the early stage of deterioration of the filter press operating conditions, solving the technical contradiction of the disconnect between process quality closed loop and equipment safety operation management in traditional control, and significantly improving the robustness of the system. Attached Figure Description

[0026] Figure 1 This is a schematic flowchart illustrating the intelligent control method for the amount of activated carbon added in the glycerol decolorization process according to the present invention.

[0027] Figure 2 This is a schematic diagram showing the effect comparison of the input-output characteristic distribution of the decolorization section involved in this invention.

[0028] Figure 3 This is a schematic diagram illustrating the effect of the safety suppression characteristic curve of the filter press involved in this invention. Detailed Implementation

[0029] The technical solution of the present invention will now be clearly and completely described in conjunction with the embodiments and accompanying drawings.

[0030] Example 1: The intelligent control method for the amount of activated carbon added in the glycerol decolorization section provided in this embodiment is as follows: Figure 1 As shown, its main process includes the following steps: S1 collects multiple parameters, establishes a first-in-first-out queue, and aligns them according to the time lag: The industrial control system collects real-time data on the crude glycerol inlet color, the decolorized outlet color, the crude glycerol feed flow rate, the inlet and outlet pressure difference of the plate and frame filter press, and the actual mass flow rate of activated carbon added in the glycerol decolorization section. A first-in-first-out (FIFO) data queue is constructed using a historical database, and the crude glycerol inlet color, the actual mass flow rate of activated carbon added, and the decolorized outlet color are time-series aligned according to the lag time of the decolorization kettle.

[0031] In this step, the data acquisition frequency is set to 1Hz. For ease of understanding, the following actual operating parameters are used for illustration: the effective volume of the decolorization kettle. Current moment Collected crude glycerol feed flow rate .

[0032] At this point, the system calculates the current lag time. :

[0033] This means that the glycerin flowing out of the decolorizing kettle at the current moment actually entered the decolorizing kettle 30 minutes ago. Therefore, the current inlet data cannot be used directly to analyze the current outlet results.

[0034] The system maintains a first-in-first-out (FIFO) data queue in memory with a length of at least one hour; the system calculates... min, backtracking queue, extract Historical data at the min time: Historical entry chromaticity Hazen; Historical activated carbon dosage Historical feed flow rate The measured export color at the current moment. Hazen.

[0035] Through dynamic lag calculation and data extraction based on volume and flow, the system successfully constructed a strictly corresponding set of input and output samples. This eliminated the 30-minute time lag misalignment, laying a data foundation for the subsequent accurate calculation of adsorption efficiency.

[0036] S2, align the data according to the time sequence, and back-calculate the activated carbon decolorization efficiency index according to the adsorption law: Based on historical data and current data after time-series alignment, the actual decolorization capacity of a unit mass of activated carbon under current operating conditions is calculated in reverse using the logarithmic decay law of liquid phase adsorption kinetics, thus obtaining the unit adsorption efficiency index.

[0037] This step aims to quantify the actual adsorption efficiency of activated carbon under current operating conditions, and the calculation formula strictly follows:

[0038] In the formula, for The unit adsorption efficiency index obtained by time-inversion; The historical entry chromaticity prior to the lag time; The current export color value; The density of activated carbon in the decolorization kettle before the lag time; This is a numerical stability constant used to prevent the denominator from being zero.

[0039] This formula is based on the logarithmic decay law of liquid-phase adsorption kinetics. Its core is to inversely derive the actual decolorization capacity per unit mass of activated carbon using time-aligned input-output data, thus achieving a quantitative characterization of activated carbon adsorption efficiency. Its logical principle revolves around the essential characteristics of the adsorption process and the practicality of engineering calculations. The formula uses the logarithmic difference in chroma as the numerator because the chroma removal effect of activated carbon on pigments in glycerol during liquid-phase adsorption exhibits a logarithmic decay relationship with the adsorbent dosage. The natural logarithmic difference between the inlet and outlet chroma accurately reflects the actual decolorization requirements and adsorption effect, avoiding the problem of linear calculations failing to match the adsorption kinetic characteristics. The denominator is selected as the activated carbon density in the decolorization vessel at the lag time, and a numerical stability constant is introduced. This is because the decolorization efficiency of activated carbon is determined by the amount of activated carbon added per unit volume of glycerol, and the activated carbon density at the lag time can accurately correspond to the adsorbent dosage conditions for the current outlet color. The addition of [a specific parameter] is to avoid engineering calculation anomalies such as a denominator of zero when the density of activated carbon in the reactor approaches zero, thus ensuring the robustness of the formula. Overall, the formula transforms the intangible activated carbon adsorption efficiency into a quantifiable index that can be calculated online by using the ratio of the quantified adsorption effect value to the adsorbent dosage density. Furthermore, by combining historical data aligned with current measured data, the impact of large time lag in the decolorization vessel on the calculation is eliminated, which can truly reflect the actual decolorization capacity of activated carbon under the current working conditions and provide a core basis for the accurate calculation of subsequent dosage.

[0040] First, calculate the density of activated carbon in the reactor at the lag time. :

[0041] Next, substitute the chromaticity data to calculate. Take the numerical stability constant Molecular part: ; Denominator: Final calculation:

[0042] Calculation results This indicates that under the current operating conditions, adding 1 kg of activated carbon to 1 cubic meter of glycerol can reduce the logarithm of color by 1.294.

[0043] If the activated carbon becomes damp at this time, its adsorption capacity will decrease. Under the same dosage, the outlet color may become 30 Hazen, at which point the molecules will change. Calculated Will be reduced to The system keenly detected the deterioration of activated carbon through this decrease in value.

[0044] S3, based on the pressure difference and rate of change, a safety factor is constructed to predict blockage and adjust the factor value accordingly: Based on the current measured differential pressure and differential pressure change rate of the plate and frame filter press, a safety inhibition factor is constructed. The safety inhibition factor identifies the clogging trend of the filter press by introducing a differential term of the differential pressure, and reduces the value of the safety inhibition factor when the differential pressure rises rapidly.

[0045] This step introduces equipment protection logic and sets the maximum permissible pressure differential of the filter press. MPa, static pressure weighted index Dynamic rate weighting coefficient .

[0046] The calculation formula is:

[0047] In the formula, As a safety inhibitory factor for equipment; This is the current measured pressure difference; This is the maximum allowable operating pressure differential of the filter press. The rate of change of pressure difference; when the calculated If the value is less than 0, assign it the value 0.

[0048] This formula is built around the safety protection requirements of plate and frame filter presses. Its core is to construct a nonlinear equipment safety inhibition factor by integrating static pressure differential state and dynamic pressure differential change trend. This allows for the prediction and quantitative constraint of filter press clogging risk, and its logical principle takes into account both the actual operating conditions of the equipment and the practicality of engineering control. The formula uses 1.0 as the baseline value because equipment safety inhibition needs to be within the range of 0-1. 1 represents no safety risk and no need to inhibit dosage, while 0 represents high risk and complete restriction of dosage, conforming to the numerical logic of engineering control. The subtrahend term is obtained by multiplying the static pressure term and the dynamic rate term, where... To characterize the static pressure difference, the measured pressure difference is compared with the maximum allowable pressure difference, and a static pressure weighted index A is introduced to amplify the influence of the overpressure trend and accurately reflect the current static load state of the filter press. To characterize dynamic trends, an exponential function amplifies the impact of differential pressure change rate. The dynamic rate weighting coefficient B can be adjusted to better reflect the sensitivity to clogging trends based on equipment characteristics, enabling early identification of filter cake thickening and rapid differential pressure increases in the filter press. Multiplying two terms couples the static and dynamic risks; the larger the differential pressure and the faster the rate of increase, the larger the subtractive term and the smaller the inhibition factor, thus smoothly reducing activated carbon dosage through numerical changes. The formula also stipulates that a negative calculated value is assigned a value of zero to avoid ineffective inhibition values, ensuring the robustness of engineering calculations and upgrading equipment protection from passive overpressure alarms to active trend suppression.

[0049] Scenario Demonstration: Assuming the filter press is in the later stages of operation, the current measured pressure difference is... Although the pressure was below the 0.6MPa alarm threshold, it was already increasing due to filter cake blockage. The rate of increase in MPa / min is rapid, that is .

[0050] The first step is to calculate the static pressure ratio to the power of A: The second step is to calculate the exponential part of the dynamic differential term: The third step is to calculate the overall inhibition factor: .

[0051] Results Analysis: Although the measured pressure difference of 0.5 MPa did not reach the threshold and seemed to be within a safe range, the calculated pressure difference was too rapid. Only This means the system will forcibly reduce the injection amount by about 90% to prevent the pressure from suddenly reaching its peak.

[0052] If the pressure is stable, the exponent term is 1. The degree of inhibition is relatively mild.

[0053] S4 calculates the dosage based on multiple parameters and issues instructions to the feeding equipment for execution: Based on the current crude glycerol feed flow rate, the target color required by the process, the unit adsorption efficiency index after smoothing treatment, and the equipment safety inhibition factor, the activated carbon dosage setpoint is calculated, and the activated carbon dosage setpoint is sent as a control command to the feeding equipment for execution.

[0054] This step calculates the final command based on all parameters; current measured feed flow rate; Current entry chromaticity Hazen; Target chromaticity Hazen; Smoothed Performance Index Safety factor .

[0055] Calculation formula:

[0056] In the formula, Set the value for activated carbon dosage; This represents the current measured feed flow rate; This is the currently measured inlet chromaticity; The target colorimetric value required by the process; The unit adsorption efficiency index after smoothing treatment; It is a device safety inhibition factor.

[0057] This formula is built around the precise calculation of activated carbon dosage in the glycerol decolorization section. It integrates three core dimensions: material load feedforward, adsorption efficiency feedback, and equipment safety constraints. Through multivariate decoupling calculations, it achieves dynamic and intelligent control of the dosage. The formula is based on the current crude glycerol feed flow rate. The base multiplier is directly linked to the production load, ensuring that the dosage is adjusted in real time according to the feed flow rate, achieving a zero-delay feedforward response to changes in material load. The numerator uses the natural logarithmic difference between the current inlet color and the target process color, continuing the logarithmic decay law of liquid-phase adsorption kinetics, accurately quantifying the adsorption requirements needed to achieve the target decolorization effect, and maintaining consistency with the calculation logic of the adsorption efficiency index to ensure data correlation. The denominator is the smoothed unit adsorption efficiency index. Based on the real-time inversion of the actual decolorization capacity of activated carbon, the theoretical adsorption requirements are corrected through feedback to avoid dosing deviations caused by fluctuations in activated carbon efficiency. Finally, an equipment safety inhibition factor is introduced. By implementing product constraints, the risk of equipment blockage is linked to the dosage calculation. When the pressure difference is abnormal, the dosage is automatically reduced by decreasing the factor value. This ensures that the bottom line of equipment safety is maintained while meeting the process color requirements, achieving the optimal balance between production quality, raw material cost and equipment safety.

[0058] The first step is to calculate the required logarithmic decrease in decolorization: The second step is to calculate the theoretical activated carbon density requirement based on performance: The third step is to calculate the basic mass flow rate based on the flow rate: Fourth step, apply safety restraint: Ultimately, the system will The set values ​​are sent to the feeder. For project implementation, the system adjusts the feeder according to the calibration curve. Convert to inverter frequency Hz, driving the motor to rotate.

[0059] In addition, regarding the attached Figure 2 and attached Figure 3 Effect description: like Figure 2 As shown, the X-axis represents the inlet chromaticity and the Y-axis represents the dosage. The scattered distribution of data points in the prior art is chaotic, indicating disordered control. In contrast, the data points of this invention are closely distributed around a logarithmic curve, proving that the algorithm can accurately provide the optimal dosage under different inlet chromaticities.

[0060] like Figure 3 As shown, the X-axis represents the comprehensive risk index, and the Y-axis represents... The curve exhibits a clear non-linear steep drop characteristic: in the low-risk area, Maintaining the level around 1.0 will not affect production; however, if the risk index rises and enters the high-risk zone, The sharp drop visually demonstrates the algorithm's proactive risk avoidance capabilities.

Claims

1. A method for intelligent control of activated carbon addition in a glycerol decolorization process, characterized in that, Includes the following steps: The industrial control system collects in real time the crude glycerol inlet color, the decolorized outlet color, the crude glycerol feed flow rate, the inlet and outlet pressure difference of the plate and frame filter press, and the actual mass flow rate of activated carbon added in the glycerol decolorization section. It constructs a first-in-first-out data queue using the historical database and aligns the crude glycerol inlet color, the actual mass flow rate of activated carbon added, and the decolorized outlet color according to the lag time of the decolorization kettle. Based on time-aligned historical and current data, and utilizing the logarithmic decay law of liquid-phase adsorption kinetics, the actual decolorization capacity of a unit mass of activated carbon under current operating conditions is calculated in reverse, yielding the unit adsorption efficiency index; the unit adsorption efficiency index is: for The unit adsorption efficiency index obtained by time-inversion; The historical entry color before the lag time; The current export color value; The density of activated carbon in the decolorization kettle before the lag time; This is a numerical stability constant used to prevent the denominator from being zero; Based on the current measured differential pressure and differential pressure change rate of the plate and frame filter press, a safety inhibition factor is constructed. This factor identifies the clogging trend of the filter press by introducing a differential term in the differential pressure, and reduces its value when the differential pressure rises rapidly. The safety inhibition factor is: As a safety inhibitory factor for equipment; This is the current measured pressure difference; This is the maximum allowable operating pressure differential of the filter press. This represents the rate of change of pressure difference; It is a static pressure-weighted index; These are the dynamic rate weighting coefficients; when the calculated values ​​are... When the value is less than 0, assign it the value 0; Based on the current crude glycerol feed flow rate, the target color required by the process, the unit adsorption efficiency index after smoothing treatment, and the equipment safety inhibition factor, the activated carbon dosage set value is calculated, and the activated carbon dosage set value is sent to the feeding equipment as a control command for execution. The activated carbon dosage setting is: Set the value for activated carbon dosage; This represents the current measured feed flow rate; This is the currently measured inlet chromaticity; The target colorimetric value required by the process; The unit adsorption efficiency index after smoothing treatment; It is a device safety inhibition factor.

2. The intelligent control method for the amount of activated carbon added in the glycerol decolorization section according to claim 1, characterized in that, The lag time is calculated by dividing the effective volume of the decolorizing reactor by the current crude glycerol feed flow rate. The specific operation of time alignment is as follows: extract the historical inlet color and historical activated carbon addition mass flow rate at the current moment minus the lag time from the first-in-first-out data queue, and combine them with the current decolorized outlet color to form a calculation group.

3. The intelligent control method for the amount of activated carbon added in the glycerol decolorization section according to claim 1, characterized in that, The method for calculating the density of activated carbon in the decolorization kettle is as follows: divide the actual mass flow rate of activated carbon added before the lag time by the crude glycerol feed flow rate before the lag time.

4. The intelligent control method for the amount of activated carbon added in the glycerol decolorization section according to claim 1, characterized in that, The rate of change of pressure difference is obtained by the differential method. The calculation formula is the measured pressure difference at the current moment minus the measured pressure difference at the previous sampling moment, divided by the sampling time interval.

5. The intelligent control method for the amount of activated carbon added in the glycerol decolorization section according to claim 1, characterized in that, The steps of sending the activated carbon dosage setpoint as a control command to the feeding equipment for execution also include: Based on the flow rate and frequency calibration curve of the feeding equipment, the calculated activated carbon dosage setpoint is converted into the operating frequency command of the frequency converter; The flow rate and frequency calibration curves have a linear relationship. Specifically, the frequency command equals the product of the calibration slope and the activated carbon dosage setting plus the intercept, to ensure that the control signal matches the physical actuator.

6. The intelligent control method for the amount of activated carbon added in the glycerol decolorization section according to claim 1, characterized in that, The method also includes the following exception protection steps: The value of the equipment safety inhibition factor is monitored in real time. When the equipment safety inhibition factor is lower than the preset safety threshold for a continuous preset time, the plate and frame filter press is determined to be in a state of serious blockage risk. The system automatically outputs an interlock shutdown signal, stopping the operation of the feeding equipment and triggering an audible and visual alarm to prompt the operator to clean or replace the filter cloth.

7. The intelligent control method for the amount of activated carbon added in the glycerol decolorization section according to claim 1, characterized in that, hysteresis time of decolorization kettle The calculation formula is: ,in, The effective volume of the decolorization kettle, This represents the crude glycerol feed flow rate.

Images