A closed-loop intelligent regulation and control refined diesel hydrogenation filter parameter optimization method

Abstract

This invention relates to the fields of petroleum refining processes and industrial automatic control technology, and particularly to a closed-loop intelligent control method for optimizing the hydrofiltration parameters of refined diesel fuel. The method includes: Step 1: Real-time acquisition of sulfur content in refined diesel fuel and feedstock flow rate data, followed by dynamic weighted filtering to generate an effective sulfur concentration sequence; Step 2: Calculation of the catalyst local clogging coefficient; Step 3: Output of the catalyst real-time activity factor based on the activity decay relationship; Step 4: Dynamic calculation and prediction of adsorption capacity; Step 5: Generation of instructions to shorten the backwash cycle and reduce the filtration rate; Step 6: Generation of backwash intensity adjustment instructions proportional to the absolute value of the sulfur concentration change slope; Step 7: Execution of adjustments to the filtration rate, backwash cycle, and backwash intensity; Step 8: If the target value is not reached, return to Step 2 for catalyst activity reassessment. This method effectively improves the accuracy and stability of the data, providing reliable real-time data support for subsequent filtration parameter optimization.

Description

Technical Field

[0001] This invention relates to the fields of petroleum refining technology and industrial automatic control technology, and in particular to a closed-loop intelligent control method for optimizing the hydrofiltration parameters of refined diesel. Background Technology

[0002] In the hydrotreating process of refined diesel, the oil effluent from the hydrotreating reactor needs to be filtered to remove residual sulfides. The settings of the filtration parameters (including filtration rate, backwash cycle, and backwash intensity) directly affect product quality and equipment operating efficiency. Currently, the industry commonly adopts the following technical solutions:

[0003] Filtration parameters are set fixed based on initial design experience values, such as a uniform backwash cycle of 8 hours and a constant filtration rate at the calibrated flow rate. However, the sulfur content of refined diesel feedstock fluctuates unpredictably due to the source of crude oil (e.g., sudden changes in sulfur concentration caused by blending low-sulfur and high-sulfur crude oils), while the hydrogenation catalyst experiences activity decay over time (manifested as a decrease in sulfur adsorption capacity). Static parameters cannot adapt to these dynamic conditions, leading to filtration overload during high-sulfur periods (significantly increasing the risk of sulfur breakthrough), while low-sulfur periods result in wasted backwash hydrogen.

[0004] In addition, adjusting the filtration parameters relies on manually collecting oil samples for offline sulfur content testing, a process that takes 12-24 hours to obtain results. During this period, if the sulfur content of the feedstock suddenly increases, the filtration system will continue to produce oil with excessive sulfur content due to the lag in parameter updates, requiring rework. Furthermore, if the catalyst suddenly deactivates, excessively delayed backwashing adjustments will accelerate filter element clogging and may even lead to unplanned shutdowns.

[0005] Therefore, there is an urgent need for a closed-loop intelligent control method to optimize the hydrofiltration parameters of refined diesel fuel and solve the above problems. Summary of the Invention

[0006] To achieve the above objectives, this invention provides a closed-loop intelligent control method for optimizing the hydrofiltration parameters of refined diesel fuel, comprising:

[0007] Step 1: Real-time data on the sulfur content of refined diesel oil and the flow rate of feedstock oil are collected using an online sulfur analyzer at the outlet of the hydrotreating reactor and an inlet flow meter. The effective sulfur concentration sequence is then generated through dynamic weighted filtering.

[0008] Step 2: Based on the pressure drop gradient obtained by three sets of differential pressure transmitters at the upper, middle and lower reaches of the catalyst bed in the reactor, and combined with the hydrogen purity correction theory to clean the pressure drop, calculate the local blockage coefficient of the catalyst.

[0009] Step 3: Based on the local blockage coefficient, reaction temperature, and hydrogen partial pressure data, output the real-time catalyst activity factor through the activity decay relationship;

[0010] Step 4: Input the effective sulfur concentration sequence and real-time active factor into the sulfur adsorption capacity prediction model to dynamically calculate the predicted adsorption capacity;

[0011] Step 5: When the real-time activity factor is below the dynamic threshold and the predicted adsorption capacity decrease rate exceeds the limit, generate instructions to shorten the backwashing cycle and reduce the filtration rate.

[0012] Step 6: When the variance of the effective sulfur concentration sequence exceeds the allowable range, a backwash intensity adjustment command is generated proportionally to the absolute value of the slope of the sulfur concentration change.

[0013] Step 7: Transmit the optimization instructions to the feed pump inverter, timing controller and proportional valve of the filtration system to adjust the filtration rate, backwash cycle and rinsing intensity;

[0014] Step 8: After backwashing, collect the filter element pressure difference recovery rate. If the target value is not reached, return to step 2: to re-evaluate the catalyst activity.

[0015] Preferably, the dynamic weighted filtering in step 1 includes:

[0016] Set the length of the sliding time window, which shortens as the feed oil flow rate increases and lengthens as the flow rate decreases;

[0017] The standard deviation of sulfur content data is calculated within the window. When the standard deviation exceeds the first set range, the current sampling point is assigned a higher weight. The weight increment is proportional to the range of the standard deviation exceeding the set range.

[0018] When the standard deviation is lower than the second set range, an equal-weighted average is used for calculation;

[0019] The flow data is subjected to lag compensation processing. The compensation time is calculated based on the pipeline length and flow velocity to align the sulfur content with the flow data timestamp.

[0020] The output filtered effective sulfur concentration sequence has the same data point density as the original sampling frequency.

[0021] Preferably, the hydrogen purity correction for the theoretical clean pressure drop in step 2 includes:

[0022] During the initial commissioning phase of the reactor, the baseline pressure drop value was measured under different hydrogen purity conditions, and a purity-pressure drop compensation factor mapping table was established.

[0023] Real-time monitoring of circulating hydrogen purity; when the purity change exceeds the set range, select the two compensation factors closest to the purity from the mapping table for linear interpolation.

[0024] Multiply the interpolated compensation factor by the initial commissioning baseline pressure drop value to obtain the theoretical clean pressure drop at the current purity.

[0025] The calculation of the local blockage coefficient is as follows: take the absolute value of the pressure drop difference between the upstream and downstream sides, divide it by the theoretical clean pressure drop, and then multiply it by the catalyst bed length correction coefficient.

[0026] Preferably, the construction of the activity decay relationship in step 3 includes:

[0027] In the accelerated aging experiment of the catalyst, the activity decay rate was measured at different temperatures and hydrogen partial pressures.

[0028] The relationship between the decay rate and the reaction temperature is fitted into an exponential function, and the coefficient of the exponential term is selected according to the catalyst type.

[0029] The relationship between the decay rate and the hydrogen partial pressure is fitted into a power function form, and the power exponent is dynamically adjusted according to the rate of change of bed pressure drop.

[0030] The real-time activity factor is calculated as 1 minus the cumulative decay amount, which is obtained by integrating the decay rate over the running time, with the integration step size decreasing as the activity factor decreases.

[0031] Preferably, the generation of the dynamic threshold in step 5 includes:

[0032] Input the cumulative catalyst running time into the S-shaped curve function, and output the basic threshold.

[0033] The number of backwashes per unit time is counted. When the number exceeds the set value, the basic threshold is reduced according to the excess ratio.

[0034] The limit for the rate of decrease in adsorption capacity is set as a percentage of the historical maximum rate of decrease, and this percentage increases as the real-time activity factor increases.

[0035] Preferably, the execution of the backwash intensity adjustment command in step 6 includes a buffering mechanism:

[0036] When the absolute value of the slope of the sulfur concentration change first exceeds the critical value, only a partial intensity adjustment is applied;

[0037] If the slope continues to increase in the next cycle, then the remaining adjustment amount will be executed;

[0038] The critical value is dynamically set based on the cumulative usage time of the filter element; the longer the usage time, the lower the critical value.

[0039] The calculation method for a portion of the ratio is as follows: the ratio of the absolute value of the slope to the critical value is used as input, and the ratio is mapped to a preset ratio range through an inverted S-shaped curve.

[0040] Preferably, the target value for the differential pressure recovery rate in step 8 includes:

[0041] Record the differential pressure recovery rate after the most recent backwashes and calculate its moving average.

[0042] When the moving average rises continuously, the target value is increased to the weighted value of the moving average;

[0043] When the moving average value decreases continuously, the priority for triggering a catalyst activity reassessment is increased to the highest level;

[0044] The weighting coefficients of the weighted values ​​are adjusted inversely based on the dispersion of the recovery rate data.

[0045] Preferably, it also includes abnormal operating condition handling:

[0046] When the rate of decrease of the real-time active factor exceeds the alarm limit within multiple consecutive sampling cycles, a catalyst regeneration prompt signal is generated.

[0047] At the same time, the filtration rate is forcibly reduced to a safe lower limit, which is calculated according to a preset ratio based on the current feed oil flow rate.

[0048] The alarm limits are set in segments based on the catalyst operating time, with lower limits for longer operating times.

[0049] Preferably, the decoupling control of the filtration rate and backwashing cycle in step 7 includes:

[0050] When it is necessary to shorten the backwashing cycle, first reduce the filtration rate to the first critical value, and then perform the cycle shortening operation.

[0051] When it is necessary to extend the backwashing cycle, first increase the filtration rate to the second critical value, and then perform the cycle extension operation.

[0052] The first threshold value is calculated by taking the difference between the current filtration rate and the maximum allowable rate, and multiplying it by a coefficient related to the real-time activity factor.

[0053] The second critical value is calculated by taking the difference between the minimum allowable rate and the current filtration rate, and multiplying it by a coefficient related to the variance of sulfur concentration fluctuation.

[0054] Preferably, the dynamic updating of the sulfur adsorption capacity prediction model in step 4 includes:

[0055] When the reactor inlet temperature undergoes a step change, record the difference in sulfur adsorption before and after the temperature change.

[0056] The ratio of this difference to the temperature change amplitude is used as the temperature sensitivity coefficient;

[0057] Compare the temperature sensitivity coefficient with the historical sensitivity coefficient sequence; if the deviation exceeds the set tolerance, refit the model parameters.

[0058] The fitting method employed recursive least squares, and the forgetting factor decreased linearly with the catalyst running time.

[0059] The beneficial effects of this invention are:

[0060] 1. This invention utilizes real-time online monitoring and dynamic weighted filtering technology to collect and adjust sulfur concentration data in real time. The effective sulfur concentration sequence is dynamically optimized based on changes in flow rate, avoiding filtration system incompatibility caused by sudden changes in sulfur concentration and ensuring stable operation of the filtration system under different sulfur concentration conditions.

[0061] 2. This invention achieves real-time monitoring and adjustment of catalyst activity decay by combining the pressure drop gradient of the catalyst bed, the hydrogen purity correction theory, and the calculation of the real-time catalyst activity factor. Utilizing the activity decay relationship and the real-time activity factor, filtration parameters are dynamically calculated and adjusted in a timely manner, avoiding excessive filter clogging and unplanned downtime caused by catalyst deactivation.

[0062] 3. This invention utilizes closed-loop intelligent control to optimize and adjust the filtration system based on real-time data. Dynamic threshold generation, backwash intensity adjustment, and filtration rate adjustment commands enable the system to respond to changes in real time, avoiding the problems of excessive sulfur output and wasted backwash hydrogen caused by parameter lag.

[0063] 4. This invention enables more flexible adjustment between the filtration rate and the backwashing cycle by decoupling the backwashing intensity adjustment command and the filtration rate from the backwashing cycle, ensuring the high efficiency of the backwashing operation, while avoiding ineffective backwashing during low sulfur periods, thus improving the energy efficiency of the filtration system and reducing hydrogen waste.

[0064] 5. This invention, through an abnormal operating condition handling mechanism, automatically generates a catalyst regeneration prompt signal and forcibly adjusts the filtration rate when the real-time active factor decline rate exceeds the alarm limit, responding promptly to abnormal changes and avoiding equipment downtime and product quality problems caused by delayed processing. Attached Figure Description

[0065] To more clearly illustrate the technical solutions in this invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, those skilled in the art can obtain other drawings based on these drawings without creative effort.

[0066] Figure 1 This is a flowchart of the steps of the method of the present invention;

[0067] Figure 2 This is a flowchart of the hydrogen purity correction steps for the theoretical cleaning pressure drop in the method of the present invention;

[0068] Figure 3 This is a flowchart illustrating the buffer mechanism steps of the backwash intensity adjustment command in the method of the present invention. Detailed Implementation

[0069] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should also be noted that, to make the embodiments more comprehensive, the following embodiments are the best and preferred embodiments, and those skilled in the art can use other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0070] Please see Figures 1-3 This invention provides a closed-loop intelligent control method for optimizing the hydrofiltration parameters of refined diesel fuel. In step 1, an online sulfur analyzer is installed at the outlet of the hydrotreating reactor to measure the sulfur content of the refined diesel fuel in real time; simultaneously, an inlet flow meter is used to measure the feedstock flow rate. The data is processed using a dynamic weighted filtering algorithm to generate an effective sulfur concentration sequence. This sequence reflects real-time changes in sulfur concentration, providing a basis for subsequent adjustments to the filtration parameters. Using weighted filtering effectively reduces data noise, making the collected data more accurate and avoiding the lag inherent in traditional manual sampling.

[0071] In step 2, three sets of differential pressure transmitters are installed upstream, midstream, and downstream of the catalyst bed to measure the pressure drop gradient at different locations within the bed. Using the hydrogen purity correction theory, the pressure drop can be cleared, and the local blockage coefficient of the catalyst can be calculated. This coefficient reflects the flow state and degree of blockage in the catalyst bed, providing crucial data for catalyst activity degradation.

[0072] In step 3, the real-time activity factor of the catalyst is calculated using the activity decay relationship, combined with data on local blockage coefficient, reaction temperature, and hydrogen partial pressure. This factor dynamically reflects the working state and degree of decay of the catalyst and is a key indicator for subsequent optimization of filtration parameters.

[0073] In step 4, the real-time collected effective sulfur concentration sequence and the real-time activity factor of the catalyst are input into the sulfur adsorption capacity prediction model to dynamically calculate the sulfur adsorption capacity of the catalyst. This prediction can determine whether the catalyst can effectively adsorb sulfur and guide the adjustment of the filtration system.

[0074] In step 5, when the real-time activity factor falls below the set dynamic threshold and the rate of decrease in sulfur adsorption capacity exceeds the limit, the system generates an instruction to shorten the backwashing cycle and reduce the filtration rate. This adjustment effectively reduces the risk of premature catalyst deactivation due to overuse, while also avoiding filtration overload during high-sulfur periods.

[0075] In step 6, when the variance of the effective sulfur concentration sequence exceeds the set allowable range, the system generates a backwash intensity adjustment command based on the slope of the sulfur concentration change. This command adjusts the backwash intensity according to the magnitude of the sulfur concentration change, thereby avoiding the waste of backwash hydrogen during low-sulfur periods and ensuring the efficient operation of the filtration system under various operating conditions.

[0076] In step 7, the optimization command is transmitted to the filtration system via the feed pump frequency converter, timing controller, and proportional valve to adjust the filtration rate, backwash cycle, and backwash intensity. These parameter adjustments directly affect the operating efficiency of the filtration system, ensuring optimal filtration performance under different operating conditions.

[0077] In step 8, after backwashing, the system collects the differential pressure recovery rate of the filter element in real time. If the recovery rate does not reach the set target value, the system will automatically return to step 2 for catalyst activity reassessment. This mechanism ensures the effectiveness of the backwashing operation and makes timely adjustments in case of catalyst failure or filter element blockage, avoiding system failures and unplanned downtime.

[0078] Through the above steps, this invention can dynamically adjust filtration parameters based on real-time data, ensuring that the filtration system maintains efficient and stable operation under complex conditions such as fluctuating sulfur content and catalyst degradation. This method overcomes the shortcomings of traditional static parameter settings, which cannot adapt to dynamic conditions, effectively reducing hydrogen waste, improving filtration system efficiency, and reducing the risk of catalyst deactivation and unplanned shutdowns. Furthermore, the closed-loop intelligent control system can respond to system status changes in real time, improving the automation level of the production process, reducing the risk of human intervention and operational errors, and providing more precise technical support for the refined diesel hydrotreating process.

[0079] In one possible implementation, the length of the sliding time window determines the calculation range of the weighted filter. This window length is dynamically adjusted based on the feedstock flow rate. When the flow rate increases, the time window shortens, thereby accelerating the data response; when the flow rate decreases, the time window lengthens to smooth data fluctuations and ensure the stability of the filter. This design effectively adapts to changes in response under different flow conditions, ensuring rapid response at high flow rates and accurate data smoothing at low flow rates.

[0080] Within a sliding window, the standard deviation of the sulfur content data is calculated. The standard deviation reflects the degree of data fluctuation. When the standard deviation exceeds a preset first threshold, the current data point is considered to have significant fluctuations, potentially influenced by external interference or possessing special significance. In this case, a higher weight is assigned to the data point, with the weight increment proportional to the extent of the standard deviation exceeding the threshold. This highlights data with greater fluctuations and reduces the impact of noise on the final result. Increasing the weight when the standard deviation exceeds the limit helps in effectively responding to sudden changes.

[0081] When the standard deviation is below the second set range, it indicates that the data fluctuation is small, representing that the data tends to be stable. In this case, to avoid overemphasizing the influence of individual data points, an equal-weighted averaging method is used to process the data. This method ensures that even with small data fluctuations, erroneous filtering results will not be caused by local fluctuations in certain data points.

[0082] Considering the fluid transport characteristics in the hydrogenation reactor and piping system, flow rate data may exhibit a certain lag. Therefore, lag compensation processing is required for the flow rate data. The compensation time is calculated based on pipeline length and flow velocity to ensure that the timestamps of sulfur content data and flow rate data are aligned. This step is crucial for ensuring data synchronization and preventing data errors caused by time mismatches.

[0083] The effective sulfur concentration sequence after dynamic weighted filtering will have a data point density consistent with the original sampling frequency. After filtering, the sulfur concentration data is smoother and more reliable, removing noise interference while maintaining a high sampling rate, allowing subsequent process control to more accurately reflect the actual situation.

[0084] Through dynamic weighted filtering, the system can adjust the data processing method in real time according to changes in feedstock flow rate and sulfur content, optimizing the filtering effect. Adjusting the sliding time window based on flow rate better adapts to rapid response and accurate calculation under different operating conditions; the introduction of standard deviation assigns higher weight to data with large fluctuations, thus avoiding overreaction during periods of low fluctuation; hysteresis compensation ensures accurate alignment of data time sequence, avoiding data errors. This method effectively improves the accuracy and stability of data, providing reliable real-time data support for subsequent filtration parameter optimization, enhancing the intelligence level of the refined diesel hydrotreating process, and reducing manual intervention and operational errors.

[0085] In one possible implementation, during the initial commissioning phase of the reactor, a baseline pressure drop value needs to be measured under different hydrogen purity conditions. This process requires measurements under multiple hydrogen purity conditions to obtain baseline pressure drop data corresponding to each purity. The baseline pressure drop value serves as a reference for subsequent pressure drop correction, reflecting the normal operating state of the system under ideal conditions (i.e., when hydrogen purity is constant). By measuring the pressure drop under different purity conditions, a purity-pressure drop compensation factor mapping table can be established for subsequent real-time compensation.

[0086] As the reactor operates, the purity of the circulating hydrogen will change, potentially influenced by various factors. To address these changes, the system needs to monitor hydrogen purity in real time. When the change in hydrogen purity exceeds a preset threshold, the system selects two compensation factors from a previously established mapping table that are closest to the current purity value. To obtain a more accurate correction factor, the system performs linear interpolation on these two compensation factors, resulting in a precise compensation factor. This interpolation process ensures a smooth transition of the compensation factor, thus enabling the system to dynamically respond to changes in hydrogen purity.

[0087] By multiplying the interpolated compensation factor by the baseline pressure drop value measured during the initial commissioning phase, the theoretical clean pressure drop value at the current hydrogen purity is obtained. This theoretical pressure drop value, after considering the impact of hydrogen purity fluctuations, more accurately reflects the pressure drop changes of the reactor during actual operation. This correction step is to ensure that the pressure drop of the reactor remains consistent with the reaction efficiency at different hydrogen purities.

[0088] Localized blockage is a common phenomenon in hydrogenation reactors, characterized by increased resistance in localized areas of the catalyst bed, leading to elevated local pressure drops. To quantify this phenomenon, calculating the localized blockage coefficient is necessary. The calculation method involves taking the absolute value of the pressure drop difference between the upstream and downstream sides, dividing it by the theoretical clean pressure drop, and obtaining the ratio of the pressure drop increment to the normal pressure drop. This ratio is then adjusted using a correction factor for the catalyst bed length to arrive at the final localized blockage coefficient. This coefficient allows the system to determine whether blockage has occurred within the reactor and to make timely adjustments.

[0089] By dynamically adjusting the pressure drop based on hydrogen purity, the intelligent control level of the hydrogenation process is effectively improved, ensuring the efficient and stable operation of the refined diesel hydrogenation reactor.

[0090] In one possible implementation, to construct the activity decay relationship, data collection is first required in accelerated aging experiments of the catalyst. In these experiments, the catalyst is aged under different reaction temperatures and hydrogen partial pressures. The rate of activity decay is measured using the reaction data from these conditions. The decay rate reflects the speed at which the catalyst loses its catalytic activity over time under different environmental conditions. Through this experiment, the fundamental laws governing catalyst decay under different operating conditions can be obtained.

[0091] The measured decay rate was fitted to the relationship between the reaction temperature. Generally, temperature has a significant impact on the activity decay of the catalyst; therefore, the relationship is expressed as an exponential function. Specifically, the relationship between the decay rate (y) and temperature (T) can be expressed as:

[0092] ;

[0093] Here, A and B are preset coefficients based on the catalyst type. This fitting process helps predict the degradation behavior of the catalyst under different temperature conditions and provides theoretical support for the control of temperature changes.

[0094] Similarly, there is a relationship between the decay rate and the hydrogen partial pressure; generally, the higher the hydrogen partial pressure, the slower the catalyst decay rate. To describe this relationship, a power function is used for fitting. This relationship can be expressed as:

[0095] ;

[0096] Among them, P H This represents the partial pressure of hydrogen. C and D are coefficients related to the catalyst type and reaction conditions. The value of D is dynamically adjusted based on the rate of change of the bed pressure drop. The rate of change of the bed pressure drop reflects the changes in the fluid dynamics inside the catalyst bed and directly affects the catalyst decay rate; therefore, the power exponent D needs to be adjusted in real time.

[0097] Once the relationship between the decay rate and temperature and hydrogen partial pressure is established, the catalyst activity factor can be calculated in real time. The formula for calculating the activity factor is:

[0098] Activity factor = 1 − cumulative decay;

[0099] The cumulative decay is obtained by integrating the decay rate over time. To more accurately reflect the catalyst decay process, the integration step size decreases as the activity factor decreases. This means that when the catalyst activity decreases, the system uses shorter time intervals to integrate the decay rate, thereby improving the accuracy and real-time performance of the decay model.

[0100] The construction of the activity decay relationship and the calculation of the real-time activity factor are the core links in realizing intelligent control of the refined diesel hydrofiltration process. By accurately predicting and dynamically adjusting catalyst decay, it ensures the stability and efficiency of the reaction process, significantly improving the intelligent control level of the entire system.

[0101] In one possible implementation, the cumulative operating time of the catalyst is a crucial factor affecting its activity decay. In step 5, the cumulative operating time of the catalyst is first input into an S-shaped curve function. This curve is typically designed based on the catalyst aging process and can simulate the trend of catalyst activity changes over long-term use. An S-shaped curve generally has the following characteristics: slow initial activity decay, accelerated decay in the middle stage, and a tendency to stabilize in the later stage. Therefore, the S-shaped curve can provide the system with a basic threshold for adapting to catalyst decay, serving as an initial reference for regulation. Specifically, the S-shaped curve can be expressed as:

[0102] ;

[0103] Where k is the slope of the curve, t is the cumulative operating time of the catalyst, and t0 is the inflection point of the curve. This function allows the calculation of a basic threshold based on the cumulative operating time of the catalyst, reflecting its current operating state.

[0104] In hydrofiltration, backwashing is a crucial cleaning method used to remove accumulated solids from the filter. In step 5, the system counts the number of backwashes per unit time. When the number of backwashes exceeds a preset value, it indicates potential clogging or decreased catalyst activity in the filtration system. The system will then adjust the baseline threshold according to a set excess percentage. This adjustment mechanism aims to identify system anomalies early and prevent excessive catalyst deterioration. The excess percentage is set based on practical experience and historical system operating data; it can be a fixed percentage or dynamically adjusted according to specific circumstances. For example, if the number of backwashes exceeds a set value, the baseline threshold may be reduced by a set percentage (e.g., 10%), thereby activating more stringent operational control measures.

[0105] The rate of adsorption capacity decline is a key indicator for measuring changes in catalyst activity during the reaction process. As mentioned in step 5, the limit for the adsorption capacity decline rate is set as a percentage of the historical maximum decline rate. The historical maximum decline rate refers to the fastest recorded rate of adsorption capacity decline during catalyst use. To ensure the system can respond promptly in the early stages of catalyst deactivation, the decline rate limit is set based on historical data and increases with the real-time activity factor. This means that when the catalyst activity factor decreases, the system becomes more sensitive and employs a stricter control strategy to limit the rate of adsorption capacity decline, preventing further catalyst loss.

[0106] Specifically, the descent rate limit can be expressed as:

[0107] Limit value = historical maximum rate of decline × (1 + ΔA);

[0108] Wherein, ΔA is the change in the activity factor. As the activity factor increases, the limit will be more stringent in order to improve the reactor's sensitivity to catalyst degradation.

[0109] The method for generating dynamic thresholds achieves precise control over the hydrofiltration process of refined diesel by combining multiple factors such as catalyst running time, backwashing frequency, and adsorption capacity reduction rate. This not only improves catalyst utilization efficiency but also significantly enhances system stability and adaptability, resulting in significant economic benefits and technological advantages.

[0110] In one possible implementation, when the absolute value of the slope of the sulfur concentration change first exceeds a critical value, the system does not immediately perform a full intensity adjustment, but only a partial intensity adjustment. This design aims to prevent the system from responding too drastically to the initial fluctuation in sulfur concentration, as the initial fluctuation may not represent a long-term trend. In this case, the system first calculates a suitable adjustment ratio based on the ratio of the sulfur concentration change slope to the set critical value using an inverted S-curve mapping, and then gradually adjusts the backwash intensity. The mapping relationship of the inverted S-curve can be expressed as:

[0111] ;

[0112] in, It is the slope of the curve. It is the ratio of the slope of the sulfur concentration change to the critical value. This is the inflection point of the curve. This curve effectively maps smaller changes to smaller adjustments, and larger changes to larger adjustments.

[0113] If the slope of the sulfur concentration change continues to increase in the next cycle, the system will perform the remaining intensity adjustment again based on the new slope data. This operation indicates that the backwash intensity adjustment will only be increased when the sulfur concentration change shows a continuous trend, thus better adapting to the continuously changing working environment. In this way, the system avoids frequent or ineffective adjustments, ensuring the gradual and accurate adjustment of the backwash intensity.

[0114] In this method, the threshold value is dynamically set based on the cumulative usage time of the filter element; the longer the usage time, the lower the threshold value. This is because as the usage time of the filter element increases, its filtration performance may gradually decline, making it more susceptible to changes in sulfur concentration. Therefore, by dynamically lowering the threshold value based on the cumulative usage time, the system makes the backwashing intensity adjustment more sensitive, enabling it to respond promptly to changes in sulfur concentration even when the filter element is aging, thus avoiding a decrease in system efficiency due to filter element failure.

[0115] This backwashing intensity adjustment method based on a buffer mechanism effectively improves the system's stability, adaptability, and long-term operation capability by gradually responding to changes in sulfur concentration, dynamically adjusting the threshold, and combining it with the aging of the filter element. It has significant technical advantages for improving the overall performance of the refined diesel hydrofiltration process.

[0116] In one possible implementation, the system continuously tracks the differential pressure recovery rate after each backflushing and records the recovery rate data from the most recent cycles. Based on this data, the system calculates a moving average. The purpose of this process is to smooth out fluctuations and eliminate short-term, sudden changes, in order to more accurately reflect the overall operating trend of the system. The moving average is calculated as follows:

[0117] ;

[0118] in, It is the most recent Recovery rate after one backwash This refers to the number of backwashes considered when calculating the moving average.

[0119] When the calculated moving average shows a continuous upward trend, it indicates that the system's filtration performance is gradually improving. At this point, the system will appropriately increase the target value of the differential pressure recovery rate. The magnitude of the increase is determined by a weighting factor, which is a weighting coefficient based on the moving average. Specifically, the target value is increased by multiplying the moving average by a weighting factor. The weighting factor is adjusted based on the stability of the data, typically using the formula:

[0120] ;

[0121] in, These are weighting coefficients that are adjusted inversely based on the dispersion of the recovery rate data; if the data fluctuations are small, Larger, and vice versa.

[0122] If the moving average value decreases continuously over multiple periods, it may indicate a decline in filtration system efficiency or reduced catalyst activity. In this case, the system triggers a catalyst activity reassessment and prioritizes it to the highest level. This means the system needs to immediately evaluate the catalyst's performance to ensure it can return to optimal operating conditions. This prevents a hysteretic response to performance degradation and ensures the catalyst is evaluated and adjusted promptly.

[0123] The dynamic adjustment of the weighting coefficients is based on the dispersion of the recovery rate data. When the recovery rate data fluctuates significantly, it indicates that there is considerable uncertainty in the system's filtering process. In this case, a smaller weighting coefficient is used to limit excessive fluctuations in the target value and prevent overly drastic adjustments to the system. Conversely, when the data dispersion is low, indicating a more stable system operation, a larger weighting coefficient is used to allow the target value to respond more sensitively.

[0124] This technology, which dynamically adjusts the target value of differential pressure recovery rate, effectively improves the filtration performance, system stability, and operating efficiency in the hydrofiltration process of refined diesel through a smooth response mechanism, intelligent weighting coefficients, and timely catalyst evaluation, resulting in significant benefits.

[0125] In one possible implementation, the real-time activity factor (typically an indicator of catalyst activity) reflects the catalyst's efficiency and activity. When the rate of decrease of this factor exceeds a preset alarm limit over multiple consecutive sampling periods, the system identifies this as an abnormal operating condition. At this point, the system generates a catalyst regeneration alert signal, indicating to the operator that the catalyst activity may have decreased and regeneration is necessary as soon as possible. To achieve this, the system periodically collects the catalyst's activity factor and calculates its rate of change. If this rate exceeds a set threshold, the system triggers an alarm and executes appropriate safety procedures.

[0126] To prevent further catalyst damage due to overload, when the system detects an abnormally low real-time activity factor, in addition to generating a regeneration alert signal, it will also force the filtration rate to a safe lower limit. This lower limit is calculated based on the current feedstock flow rate according to a preset ratio. Specifically, the system monitors the feedstock flow rate in real time and calculates an appropriate lower limit based on the flow data according to a certain ratio, ensuring that the system operates within a safe range under abnormal conditions and avoiding additional burden on the system and catalyst due to excessively high filtration rates.

[0127] The alarm limit setting depends not only on the rate of decrease in the active factor but also on the catalyst's operating time. To more accurately address catalysts at different stages of use, the system sets alarm limits in segments based on the catalyst's operating time. Specifically, the longer the catalyst's operating time, the lower its tolerance for activity factor decline, and therefore the lower the alarm limit. This setting can more effectively adapt to the catalyst aging process, identify catalyst failure trends in advance, and thus take timely measures to prevent the system from continuing to operate when the catalyst fails.

[0128] This abnormal operating condition handling mechanism not only improves the system's ability to manage and protect the catalyst, but also enhances the system's response speed and handling capability to abnormal situations, effectively ensuring the safe, stable and efficient operation of the refined diesel hydrofiltration process.

[0129] In one possible implementation, when the system needs to shorten the backwash cycle, the filtration rate is first reduced to a first critical value. This critical value is calculated by multiplying the difference between the current filtration rate and the maximum allowable rate by a coefficient related to the real-time activity factor. The specific calculation formula is as follows:

[0130] First critical value = (maximum allowable rate − current filtering rate) × K1;

[0131] Here, K1 is a coefficient related to the real-time activity factor, reflecting the current activity level of the catalyst. When the real-time activity factor is low, the filtration rate needs to be reduced accordingly to avoid overloading and prevent catalyst damage or decreased filtration efficiency. By controlling the filtration rate to this critical value, the system ensures that equipment overload or excessive catalyst wear does not occur during the shortening of backwashing cycles.

[0132] When the system needs to extend the backwash cycle, the filtration rate will first be increased to a second critical value. This critical value is calculated by multiplying the difference between the minimum allowable rate and the current filtration rate by a coefficient related to the variance of sulfur concentration fluctuations. The specific calculation formula is as follows:

[0133] Second critical value = (current filtering rate − minimum allowable rate) × K2;

[0134] Here, K2 is a coefficient related to the variance of sulfur concentration fluctuations. The greater the sulfur concentration fluctuation, the smaller the adjustment range of the filtration rate, to prevent system instability caused by sulfur concentration fluctuations. By increasing the filtration rate to the second critical value, the system can effectively extend the backwashing cycle while avoiding instability in filtration performance or equipment damage due to excessively high filtration rates.

[0135] By adjusting the filtration rate first and then the backwashing cycle, this method effectively decouples the control of the filtration rate and the backwashing cycle. Specifically:

[0136] Ensure that while accelerating the backwash cycle, excessively high filtration rates do not overburden the catalyst and filtration equipment. When extending the backwash cycle, appropriately increase the filtration rate to ensure that the filtration system continues to operate efficiently during longer backwash cycles.

[0137] By decoupling the control of filtration rate and backwash cycle, this technology not only enhances the system's intelligent adjustment capability but also improves the efficiency of the backwash process, reduces equipment load, extends catalyst life, and helps achieve efficient and stable operation of the refined diesel hydrofiltration process.

[0138] In one possible implementation, when the reactor inlet temperature undergoes a step change, the system needs to record the difference in sulfur adsorption before and after the temperature change in real time. A step change refers to a sudden change in temperature, such as a sudden increase or decrease in reactor temperature due to adjustments in operating conditions or changes in the external environment. By recording the difference in sulfur adsorption before and after this change, the system can capture the impact of temperature changes on the sulfur adsorption process.

[0139] Next, the system will use the ratio between the difference in sulfur adsorption and the magnitude of the temperature change (i.e., the amount of temperature change) to calculate the temperature sensitivity coefficient. The specific formula is as follows:

[0140] ;

[0141] This temperature sensitivity coefficient reflects the degree to which temperature changes affect the amount of sulfur adsorbed. It helps the system understand the trend of sulfur adsorption capacity changes under specific temperature variations.

[0142] The calculated temperature sensitivity coefficient will be compared with the historical sensitivity coefficient sequence. If the deviation between the currently calculated sensitivity coefficient and the historical sensitivity coefficient exceeds the set tolerance value, the system will determine that the effect of temperature change on sulfur adsorption has changed significantly. At this time, the system will trigger the process of refitting the sulfur adsorption capacity prediction model.

[0143] The refitting of the prediction model employs recursive least squares. This method optimizes model parameters by continuously updating historical data to minimize prediction errors. Recursive least squares offers high computational efficiency in practical applications, enabling real-time parameter updates and thus making the prediction model more closely reflect actual operating conditions. To accommodate the catalyst aging process, the forgetting factor decreases linearly with increasing catalyst operating time. In this way, the system prioritizes recent data and gradually reduces reliance on older historical data, improving the sensitivity and accuracy of the prediction model.

[0144] By accurately calculating the temperature sensitivity coefficient, adjusting model parameters in a timely manner, and introducing recursive least squares and forgetting factors, the system's response to temperature changes and prediction accuracy can be significantly improved. This helps optimize the sulfur adsorption efficiency in the hydrofiltration process of refined diesel, extend the catalyst's lifespan, and improve the overall system's economy and stability.

[0145] The following examples will illustrate this in detail:

[0146] In this embodiment, the system is applied to a refined diesel hydrotreating reactor, which removes sulfides from the feedstock diesel to meet environmental standards. The reactor's operating conditions include:

[0147] Operating temperature: 350°C

[0148] Working pressure: 10MPa

[0149] Feed sulfur concentration: 200 ppm (parts per million)

[0150] Backwashing cycle: Perform a backwash every 100 hours, lasting 30 minutes.

[0151] Filtration rate: 50L / h

[0152] Catalyst type: NiMo / Al2O3 catalyst

[0153] Specifically, the model parameters are updated and corrected based on the recursive least squares (RLS) method.

[0154] The prediction model formula is as follows:

[0155] ;

[0156] in, This represents the predicted sulfur adsorption capacity at time t. This is the parameter vector updated at time t. The input feature vector includes temperature, feed sulfur concentration, etc.

[0157] Recursive least squares formula:

[0158] ;

[0159] in, This represents the actual sulfur adsorption capacity. The update formula for the gain matrix is:

[0160] ;

[0161] ;

[0162] in, Let covariance matrix be the variance matrix. The forgetting factor is usually set to 0.98, indicating a high degree of dependence on historical data.

[0163] Based on experimental data, the temperature sensitivity coefficient ( It can be calculated using the following formula:

[0164] ;

[0165] in, This represents the change in sulfur adsorption capacity. This represents the change in temperature. Experimental data shows that within a temperature range of 350°C to 380°C, The value is approximately 0.15.

[0166] The backwashing cycle depends on the predicted sulfur adsorption capacity. If the predicted sulfur adsorption capacity is lower than a set threshold (e.g., 80%), the backwashing cycle will be shortened to once every 80 hours. Conversely, if the sulfur adsorption capacity is higher than the threshold, the backwashing cycle can be extended to once every 120 hours.

[0167] When the predicted sulfur adsorption capacity decreases, the system will automatically reduce the filtration rate to reduce overload and prevent rapid catalyst aging. The filtration rate ranges from 30 to 60 L / h, with optimal efficiency typically achieved at 50 L / h.

[0168] The system collects data such as temperature and sulfur concentration in the reactor using sensors. In this embodiment, the temperature is 350°C and the feed sulfur concentration is 200 ppm.

[0169] The collected temperature and sulfur concentration data are input into the sulfur adsorption capacity prediction model:

[0170] x t =[350,200];

[0171] The parameters are updated using recursive least squares to predict the sulfur adsorption capacity at the current moment.

[0172] Based on temperature sensitivity coefficient =0.15, the system predicts the effect of temperature change on sulfur adsorption capacity. For example, if the temperature increases by 5°C (from 350°C to 355°C), the change in sulfur adsorption capacity is:

[0173] ;

[0174] This indicates that a 5°C increase in temperature leads to a 0.75-unit increase in sulfur adsorption capacity.

[0175] Adjust the backwash cycle and filtration rate:

[0176] Based on the sulfur adsorption capacity predicted by the model (in this embodiment, the predicted value is 90%), the system decides to extend the backwashing cycle to 120 hours while maintaining the filtration rate at 50 L / h.

[0177] Based on real-time data and model predictions, the system automatically adjusts the backwashing cycle and filtration rate to ensure stable reactor operation and maximize catalyst utilization efficiency.

[0178] Experimental comparison:

[0179] Traditional method: The traditional method operates according to a fixed backwashing cycle, without considering the influence of factors such as temperature and sulfur concentration. The backwashing cycle is usually set once every 100 hours, and the filtration rate is 50L / h.

[0180] The method of this invention: Based on real-time data such as temperature and sulfur concentration, this invention dynamically adjusts the backwashing cycle and filtration rate through a sulfur adsorption capacity prediction model.

[0181] Experimental data show that the catalyst's lifespan is 800 hours when using traditional methods; however, when using the method of this invention, with optimized backwashing cycles and filtration rates, the catalyst's lifespan is extended to 1100 hours, an improvement of 37.5%. Simultaneously, sulfur removal efficiency is increased by 10%, meeting stricter environmental standards.

[0182] Through the application of this embodiment, the sulfur adsorption capacity prediction and control system of the present invention effectively improves the reactor's operating efficiency, extends the catalyst's lifespan, and optimizes the backwashing cycle and filtration rate. Experimental data and comparative results show that the method of the present invention has significant performance advantages over traditional methods, maintaining stable reaction results under varying operating conditions, thus possessing significant technical value and market prospects in practical industrial applications.

[0183] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0184] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A closed-loop intelligent control method for optimizing hydrofiltration parameters of refined diesel fuel, characterized in that, include: Step 1: Real-time data on the sulfur content of refined diesel oil and the flow rate of feedstock oil are collected using an online sulfur analyzer at the outlet of the hydrotreating reactor and an inlet flow meter. The effective sulfur concentration sequence is then generated through dynamic weighted filtering. Step 2: Based on the pressure drop gradient obtained by three sets of differential pressure transmitters at the upper, middle and lower reaches of the catalyst bed in the reactor, and combined with the hydrogen purity correction theory to clean the pressure drop, calculate the local blockage coefficient of the catalyst. Step 3: Based on the local blockage coefficient, reaction temperature, and hydrogen partial pressure data, output the real-time catalyst activity factor through the activity decay relationship; Step 4: Input the effective sulfur concentration sequence and real-time active factor into the sulfur adsorption capacity prediction model to dynamically calculate the predicted adsorption capacity; Step 5: When the real-time activity factor is below the dynamic threshold and the predicted adsorption capacity decrease rate exceeds the limit, generate instructions to shorten the backwashing cycle and reduce the filtration rate. Step 6: When the variance of the effective sulfur concentration sequence exceeds the allowable range, a backwash intensity adjustment command is generated proportionally to the absolute value of the slope of the sulfur concentration change. Step 7: Transmit the optimization instructions to the feed pump inverter, timing controller and proportional valve of the filtration system to adjust the filtration rate, backwash cycle and rinsing intensity; Step 8: After backwashing, collect the filter element pressure difference recovery rate. If the target value is not reached, return to step 2: to re-evaluate the catalyst activity.

2. The method for optimizing refined diesel hydrotreating filtration parameters with closed-loop intelligent control according to claim 1, characterized in that, The dynamic weighted filtering in step 1 includes: Set the length of the sliding time window, which shortens as the feed oil flow rate increases and lengthens as the flow rate decreases; The standard deviation of sulfur content data is calculated within the window. When the standard deviation exceeds the first set range, the current sampling point is assigned a higher weight. The weight increment is proportional to the range of the standard deviation exceeding the set range. When the standard deviation is lower than the second set range, an equal-weighted average is used for calculation; The flow data is subjected to lag compensation processing. The compensation time is calculated based on the pipeline length and flow velocity to align the sulfur content with the flow data timestamp. The output filtered effective sulfur concentration sequence has the same data point density as the original sampling frequency.

3. The method for optimizing refined diesel hydrotreating filtration parameters with closed-loop intelligent control according to claim 1, characterized in that, The hydrogen purity correction for the theoretical clean pressure drop in step 2 includes: During the initial commissioning phase of the reactor, the baseline pressure drop value was measured under different hydrogen purity conditions, and a purity-pressure drop compensation factor mapping table was established. Real-time monitoring of circulating hydrogen purity; when the purity change exceeds the set range, select the two compensation factors closest to the purity from the mapping table for linear interpolation. Multiply the interpolated compensation factor by the initial commissioning baseline pressure drop value to obtain the theoretical clean pressure drop at the current purity. The calculation of the local blockage coefficient is as follows: take the absolute value of the pressure drop difference between the upstream and downstream sides, divide it by the theoretical clean pressure drop, and then multiply it by the catalyst bed length correction coefficient.

4. The method for optimizing refined diesel hydrotreating filtration parameters with closed-loop intelligent control according to claim 1, characterized in that, The construction of the activity decay relationship in step 3 includes: In the accelerated aging experiment of the catalyst, the activity decay rate was measured at different temperatures and hydrogen partial pressures. The relationship between the decay rate and the reaction temperature is fitted into an exponential function, and the coefficient of the exponential term is selected according to the catalyst type. The relationship between the decay rate and the hydrogen partial pressure is fitted into a power function form, and the power exponent is dynamically adjusted according to the rate of change of bed pressure drop. The real-time activity factor is calculated as 1 minus the cumulative decay amount, which is obtained by integrating the decay rate over the running time, with the integration step size decreasing as the activity factor decreases.

5. The method for optimizing refined diesel hydrotreating filtration parameters with closed-loop intelligent control according to claim 1, characterized in that, Step 5, generating the dynamic threshold, includes: Input the cumulative catalyst running time into the S-shaped curve function, and output the basic threshold. The number of backwashes per unit time is counted. When the number exceeds the set value, the basic threshold is reduced according to the excess ratio. The limit for the rate of decrease in adsorption capacity is set as a percentage of the historical maximum rate of decrease, and this percentage increases as the real-time activity factor increases.

6. The method for optimizing refined diesel hydrotreating filtration parameters with closed-loop intelligent control according to claim 1, characterized in that, The execution of the backwash intensity adjustment command in step 6 includes a buffering mechanism: When the absolute value of the slope of the sulfur concentration change first exceeds the critical value, only a partial intensity adjustment is applied; If the slope continues to increase in the next cycle, then the remaining adjustment amount will be executed; The critical value is dynamically set based on the cumulative usage time of the filter element; the longer the usage time, the lower the critical value. The calculation method for a portion of the ratio is as follows: the ratio of the absolute value of the slope to the critical value is used as input, and the ratio is mapped to a preset ratio range through an inverted S-shaped curve.

7. The method for optimizing refined diesel hydrotreating filtration parameters with closed-loop intelligent control according to claim 1, characterized in that, The target value for differential pressure recovery rate in step 8 includes: Record the differential pressure recovery rate after the most recent backwashes and calculate its moving average. When the moving average rises continuously, the target value is increased to the weighted value of the moving average; When the moving average value decreases continuously, the priority for triggering a catalyst activity reassessment is increased to the highest level; The weighting coefficients of the weighted values ​​are adjusted inversely based on the dispersion of the recovery rate data.

8. The method for optimizing refined diesel hydrotreating filtration parameters with closed-loop intelligent control according to claim 1, characterized in that, It also includes handling abnormal operating conditions: When the rate of decrease of the real-time active factor exceeds the alarm limit within multiple consecutive sampling cycles, a catalyst regeneration prompt signal is generated. At the same time, the filtration rate is forcibly reduced to a safe lower limit, which is calculated according to a preset ratio based on the current feed oil flow rate. The alarm limits are set in segments based on the catalyst operating time, with lower limits for longer operating times.

9. The method for optimizing refined diesel hydrotreating filtration parameters with closed-loop intelligent control according to claim 1, characterized in that, The decoupling control of filtration rate and backwash cycle in step 7 includes: When it is necessary to shorten the backwashing cycle, first reduce the filtration rate to the first critical value, and then perform the cycle shortening operation. When it is necessary to extend the backwashing cycle, first increase the filtration rate to the second critical value, and then perform the cycle extension operation. The first threshold value is calculated by taking the difference between the current filtration rate and the maximum allowable rate, and multiplying it by a coefficient related to the real-time activity factor. The second critical value is calculated by taking the difference between the minimum allowable rate and the current filtration rate, and multiplying it by a coefficient related to the variance of sulfur concentration fluctuation.

10. The method for optimizing refined diesel hydrotreating filtration parameters with closed-loop intelligent control according to claim 1, characterized in that, The dynamic update of the sulfur adsorption capacity prediction model in step 4 includes: When the reactor inlet temperature undergoes a step change, record the difference in sulfur adsorption before and after the temperature change. The ratio of this difference to the temperature change amplitude is used as the temperature sensitivity coefficient; Compare the temperature sensitivity coefficient with the historical sensitivity coefficient sequence; if the deviation exceeds the set tolerance, refit the model parameters. The fitting method employed recursive least squares, and the forgetting factor decreased linearly with the catalyst running time.

Images