High-efficiency washing system and intelligent control method for marine desulfurization device

Abstract

The application discloses a marine desulfurization device high-efficiency washing system and an intelligent control method, and relates to the field of intelligent control, and comprises the following steps: deploying multiple types of sensors at key nodes, using edge computing to realize data preprocessing; establishing a sulfur flux dynamic calculation model, combining the correlation between fuel sulfur content and SO2 generation amount to calculate sulfur flux in real time; predicting alkali demand based on sulfur flux data and future working conditions, and using a feedforward-feedback collaborative control strategy to optimize alkali addition; dynamically adjusting washing liquid flow, temperature and spraying angle to improve desulfurization efficiency; constructing a device safety protection mechanism, monitoring key parameters in real time and warning of abnormalities; and finally simulating system state in real time through a digital twin model, predicting potential problems and generating warnings. The application has the advantages that: by dynamically calculating sulfur flux, predicting alkali demand and optimizing washing parameters through feedforward-feedback collaborative control, combining device safety protection and digital twin warning, efficient, energy-saving and safe operation of the desulfurization system is realized.

Description

Technical Field

[0001] This invention relates to the field of intelligent control, and in particular to a high-efficiency scrubbing system and intelligent control method for marine desulfurization devices. Background Technology

[0002] Traditional desulfurization technologies, such as open-loop scrubbing towers, pose a risk of wastewater pollution, while closed-loop systems are limited by efficiency and energy consumption issues, making it difficult to balance economic efficiency and environmental friendliness. Meanwhile, the complex and variable operating conditions of ships mean that existing control systems lack dynamic adaptability, leading to large fluctuations in desulfurization efficiency and excessive chemical consumption. Against this backdrop, the development of efficient scrubbing systems and intelligent control methods has become a focus of the industry.

[0003] Current marine desulfurization control methods suffer from several shortcomings. Firstly, they rely heavily on single or limited sensors to monitor basic parameters, lacking multi-dimensional data coverage from distributed sensor networks. Furthermore, they fail to incorporate edge computing for real-time noise reduction and standardization, resulting in insufficient data accuracy and timeliness. Secondly, control logic is largely based on simple feedback control, such as adjusting slurry supply based on outlet SO2 concentration or pH value, leading to a 5-15 minute system lag. The lack of a dynamic sulfur flux model makes rapid adaptation during fuel switching difficult, potentially causing excessive emissions or alkali waste. Thirdly, parameter optimization is limited to fixed settings like scrubbing fluid flow rate and spray angle, failing to dynamically adjust based on flue gas flow, temperature, and seawater salinity. In complex marine environments, desulfurization efficiency can easily drop below 80%, making stable compliance difficult. Fourthly, safety and early warning systems lack full-condition digital twin simulation capabilities, relying primarily on passive alarms after malfunctions occur. They cannot proactively predict equipment blockages or corrosion, and equipment protection largely depends on manual inspections, resulting in delayed emergency response. Summary of the Invention

[0004] To improve existing systems and methods, this paper provides a high-efficiency scrubbing system and intelligent control method for marine desulfurization units. This method optimizes scrubbing parameters through dynamic calculation of sulfur flux, prediction of alkali demand, and feedforward-feedback collaborative control. It also combines equipment safety protection and digital twin early warning to achieve efficient, energy-saving, and safe operation of the desulfurization system.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] Intelligent control methods for marine desulfurization units include:

[0007] Based on a distributed sensor network, multiple types of sensors are deployed at key nodes of the ship desulfurization system. The data acquisition process is preprocessed through edge computing, including data noise reduction, outlier removal, and data format standardization.

[0008] A dynamic calculation model for sulfur flux is constructed based on the collected real-time data. The correlation between fuel sulfur content and SO2 generation is established through historical operating data. The sulfur flux under different fuel types is calculated by calling the model parameters corresponding to the correlation relationship, with flue gas flow rate and SO2 concentration in flue gas as input parameters.

[0009] Based on the acquired dynamic sulfur flux data and combined with the ship's future operating conditions data, an alkaline liquid demand forecasting model is constructed to predict the alkaline liquid demand in the future time period.

[0010] Feedforward control is implemented based on the alkali demand forecast, and feedback control is implemented in combination with the real-time monitoring of flue gas outlet SO2 concentration and scrubbing liquid pH value, forming a feedforward-feedback collaborative control mechanism.

[0011] Based on the collected flue gas flow rate, flue gas temperature and sulfur flux data, the flow rate, temperature and spray angle of the scrubbing liquid are dynamically optimized.

[0012] Based on the collected equipment operating parameters and washing liquid status data, an equipment safety protection mechanism is constructed, the safety threshold range of each device is set, and the operating status of key equipment in the washing system is monitored in real time.

[0013] Based on real-time and historical operating data, a digital twin model of the marine desulfurization device is constructed to simulate the operating status of the desulfurization system in real time, simulate the system response under different operating conditions, predict potential problems in the system, and generate early warning information in advance.

[0014] Preferably, the distributed sensor network-based approach deploys multiple types of sensors at key nodes of the ship's desulfurization system. The data acquisition process utilizes edge computing for data preprocessing, including data noise reduction, outlier removal, and data format standardization. Specifically, this includes:

[0015] The key nodes include flue gas inlet pipe, flue gas outlet pipe, different height areas inside the scrubbing tower, alkali storage tank, alkali conveying pipe and seawater replenishment pipe.

[0016] The various types of sensors include infrared SO2 concentration sensors, flue gas flow sensors, and flue gas temperature sensors for collecting flue gas parameters; washing liquid pH sensors, washing liquid flow sensors, washing liquid temperature sensors, and alkali liquid level sensors for collecting washing system parameters; main engine load sensors and ship speed sensors for collecting ship operating condition parameters; and seawater temperature sensors and seawater salinity sensors for collecting environmental parameters.

[0017] The collected data is denoised by edge computing devices to eliminate interference signals, outliers are removed to avoid deviations, and finally the data format is unified to achieve standardization.

[0018] Preferably, the step of constructing a dynamic sulfur flux calculation model based on collected real-time data, establishing a correlation between fuel sulfur content and SO2 generation through historical operating data, and using flue gas flow rate and SO2 concentration in flue gas as input parameters, specifically includes calculating the sulfur flux for different fuel types by calling the model parameters corresponding to the correlation relationship:

[0019] A dynamic calculation model for sulfur flux was constructed, and flue gas flow rate and flue gas SO2 concentration were extracted from multi-source data as input parameters, while main engine load and real-time fuel consumption rate were retrieved as correction parameters.

[0020] Instantaneous sulfur emissions are calculated based on flue gas flow rate and SO2 concentration, and the instantaneous sulfur emissions are dynamically corrected by combining the trend of main engine load changes, so as to eliminate the sulfur emission calculation deviation caused by the fluctuation of main engine load.

[0021] The correlation between fuel sulfur content and SO2 generation is established by calling historical operating data. If the ship switches fuel, the corresponding correlation is automatically matched, the parameters of the calculation model are updated, and the sulfur flux under different fuel types is calculated.

[0022] Preferably, the step of constructing an alkali demand forecasting model based on the acquired dynamic sulfur flux data and combined with future ship operating condition data to predict the alkali demand in the future time period specifically includes:

[0023] Based on the acquired sulfur flux data, the ship navigation system obtains route information and sea state forecast data, and combines them with main engine load planning data to generate the expected speed changes and main engine load range parameters for future time periods.

[0024] A time-series prediction algorithm was used to construct an alkaline solution demand prediction model. The model was trained using historical sulfur flux data and historical alkaline solution consumption data as training samples to obtain the time-series correlation between sulfur flux and alkaline solution consumption.

[0025] The basic demand for alkali solution is calculated based on the trained alkali solution demand prediction model for the future time period, and the basic demand is then corrected based on the alkali solution storage capacity and the alkali solution concentration decay rate.

[0026] Preferably, the feedforward control based on the alkali demand forecast, combined with feedback control based on real-time monitoring of the flue gas outlet SO2 concentration and the washing liquid pH value, forming a feedforward-feedback collaborative control mechanism, specifically includes:

[0027] Based on the alkali demand forecast, the expected alkali demand for different time periods is determined. The feedforward control adjusts the alkali delivery flow rate in advance according to the predicted demand by controlling the speed of the alkali delivery pump.

[0028] The feedback control collects the SO2 concentration at the flue gas outlet and the pH value of the washing liquid in real time, and monitors the alkali concentration through a concentration sensor in the alkali storage tank.

[0029] When the SO2 concentration at the flue gas outlet monitored in real time is higher than the standard value or the pH value of the washing liquid is lower than the lower limit of the optimal range, the alkaline solution delivery flow rate is automatically increased. When the SO2 concentration at the flue gas outlet monitored in real time is lower than the standard value and the pH value of the washing liquid is higher than the upper limit of the optimal range, the alkaline solution delivery flow rate is automatically decreased.

[0030] Preferably, the dynamic optimization of the washing liquid flow rate, temperature, and spray angle based on the collected flue gas flow rate, flue gas temperature, and sulfur flux data specifically includes:

[0031] Based on the collected flue gas flow rate, flue gas temperature, seawater salinity data, and the generated dynamic sulfur flux data, if the sulfur flux is higher than 10 kg / h or the flue gas flow rate increases, the washing liquid flow rate should be increased.

[0032] By adjusting the angle of the adjustable spray nozzles inside the scrubbing tower, the contact area between the scrubbing liquid and the flue gas can be increased, thereby improving the SO2 absorption efficiency.

[0033] When the flue gas temperature is higher than 60℃, control the heat exchange valve of the seawater supply pipeline to introduce low-temperature seawater to regulate the temperature of the washing liquid and stabilize it in the optimal absorption range of 30-45℃.

[0034] Real-time monitoring of seawater salinity; if salinity is above 35‰, increase the frequency of detergent discharge; if salinity is below 35‰, maintain the normal discharge frequency.

[0035] The used washing liquid is filtered and settled in the clarification tank at the bottom of the washing tower to remove solid impurities. After the pH value and SO3²⁻ concentration of the clarified washing liquid are tested and found to be within acceptable limits, it is recycled.

[0036] Preferably, the step of constructing a device safety protection mechanism based on collected equipment operating parameters and washing liquid status data, setting safety threshold ranges for each device, and real-time monitoring of the operating status of key equipment in the washing system specifically includes:

[0037] Based on the collected equipment operating parameters and washing liquid status data, threshold values ​​are set for key equipment parameters such as alkali transfer pump, washing liquid circulation pump, spray nozzle, and heat exchanger.

[0038] When the detected parameters exceed the threshold, the pump body vibration or current exceeds the standard, the load will be automatically reduced and an early warning will be issued. If the nozzle pressure is lower than the threshold, it is judged to be blocked and the backup nozzle group will be switched.

[0039] Monitor the pH value of the detergent in real time. When the pH value remains below 7.0, add an anti-scaling agent.

[0040] If the SO2 concentration at the flue gas outlet exceeds 1.5 times the IMO standard for 10 consecutive seconds, the pressure inside the tower exceeds the threshold, or the alkali liquid level is below the emergency lower limit, shut down the machine immediately. First, turn off the main engine fuel supply and flue gas valves, then stop the conveying system and start the emergency ventilation.

[0041] Preferably, the step of constructing a digital twin model of the marine desulfurization device based on real-time data and historical operating data, simulating the operating status of the desulfurization system in real time, simulating the system's response under different operating conditions, predicting potential system problems, and generating early warning information specifically includes:

[0042] A digital twin model is built based on real-time data and historical operational data, and the virtual model and the physical system are accurately mapped in terms of state through real-time data transmission;

[0043] The simulation results of the system response were obtained by simulating typical operating conditions such as sudden changes in main engine load, changes in fuel sulfur content, and fluctuations in ambient temperature using a digital twin model.

[0044] Based on the simulation system response results, potential problems in the system can be predicted, and early warning information can be generated in advance.

[0045] Furthermore, a high-efficiency scrubbing system for marine desulfurization devices is proposed, including:

[0046] Distributed sensor network module: Multiple types of sensors are deployed at key nodes such as flue gas inlet and outlet, scrubbing tower, and alkali pipeline to collect flue gas, scrubbing liquid, ship operating conditions and environmental parameters in real time;

[0047] Edge computing preprocessing module: Performs real-time noise reduction, outlier removal, and format standardization on raw sensor data using edge computing devices;

[0048] Sulfur flux dynamic calculation module: Based on flue gas flow rate, SO2 concentration and main engine load data, combined with the historical model of fuel sulfur content-SO2 generation, it dynamically calculates the real-time sulfur flux under different fuel types.

[0049] Alkali demand forecasting module: Based on sulfur flux data and future ship operating conditions, it outputs the alkali demand for future periods through a time series forecasting model, and makes secondary corrections based on storage capacity and concentration decay.

[0050] Feedforward-feedback coordinated control module: The flow rate of the delivery pump is adjusted based on the predicted value of the alkali solution, and the dynamic fine-tuning is performed in real time based on the SO2 concentration at the flue gas outlet and the pH value of the washing liquid.

[0051] The dynamic optimization module for washing parameters dynamically optimizes the washing liquid flow rate, spray angle, and temperature based on flue gas flow rate, temperature, and sulfur flux data to maintain the best SO2 absorption efficiency.

[0052] Equipment safety protection module: Real-time monitoring of the operating parameters of key equipment such as pump body and nozzle, as well as the status of washing liquid; Automatic load reduction / switch to backup equipment when thresholds are exceeded; Trigger shutdown protection process when abnormalities are severe.

[0053] Digital twin simulation early warning module: Constructs a digital twin model of the desulfurization system, predicts system anomalies through real-time simulation under multiple operating conditions, and generates early warning information on equipment failure or excessive emissions in advance;

[0054] Processor: The processor is used to handle the calculation process of each formula and the construction calculation process of each model.

[0055] Compared with the prior art, the advantages of the present invention are:

[0056] Multi-dimensional data is collected in real time through a distributed sensor network, and noise reduction and standardization preprocessing are performed using edge computing to provide high-quality data support for control. A dynamic sulfur flux model is constructed to adapt to different fuel types. Combined with alkali demand prediction and feedforward-feedback collaborative control, the alkali supply is adjusted in advance to avoid exceeding emission standards and to reduce material waste. The flow rate, temperature, and spray angle of the scrubbing liquid are dynamically optimized and waste liquid is recycled, which significantly improves SO2 absorption efficiency and reduces energy consumption. The equipment safety protection mechanism realizes fault early warning, backup switching, and emergency shutdown through threshold monitoring to ensure system stability. The digital twin model simulates multi-condition response, predicts problems in advance and provides early warning, and reduces downtime due to failure. The overall desulfurization process achieves a unified approach of intelligence, efficiency, safety, and energy saving. Attached Figure Description

[0057] Figure 1 This is a schematic diagram of the method proposed in this invention;

[0058] Figure 2 This is a schematic diagram of the distributed sensor network proposed in this invention;

[0059] Figure 3 This is a schematic diagram illustrating the dynamic calculation of sulfur flux proposed in this invention;

[0060] Figure 4 This is a schematic diagram illustrating the alkali demand prediction proposed in this invention.

[0061] Figure 5 This is a schematic diagram of the feedforward-feedback coordinated control proposed in this invention;

[0062] Figure 6 This is a schematic diagram illustrating the optimized washing parameters proposed in this invention;

[0063] Figure 7 This is a schematic diagram of the equipment safety protection proposed in this invention;

[0064] Figure 8 This is a schematic diagram of the digital twin simulation proposed in this invention. Detailed Implementation

[0065] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0066] The high-efficiency scrubbing system for marine desulfurization units includes:

[0067] Distributed sensor network module: Multiple types of sensors are deployed at key nodes such as flue gas inlet and outlet, scrubbing tower, and alkali pipeline to collect flue gas, scrubbing liquid, ship operating conditions and environmental parameters in real time;

[0068] Edge computing preprocessing module: Performs real-time noise reduction, outlier removal, and format standardization on raw sensor data using edge computing devices;

[0069] Sulfur flux dynamic calculation module: Based on flue gas flow rate, SO2 concentration and main engine load data, combined with the historical model of fuel sulfur content-SO2 generation, it dynamically calculates the real-time sulfur flux under different fuel types.

[0070] Alkali demand forecasting module: Based on sulfur flux data and future ship operating conditions, it outputs the alkali demand for future periods through a time series forecasting model, and makes secondary corrections based on storage capacity and concentration decay.

[0071] Feedforward-feedback coordinated control module: The flow rate of the delivery pump is adjusted based on the predicted value of the alkali solution, and the dynamic fine-tuning is performed in real time based on the SO2 concentration at the flue gas outlet and the pH value of the washing liquid.

[0072] The dynamic optimization module for washing parameters dynamically optimizes the washing liquid flow rate, spray angle, and temperature based on flue gas flow rate, temperature, and sulfur flux data to maintain the best SO2 absorption efficiency.

[0073] Equipment safety protection module: Real-time monitoring of the operating parameters of key equipment such as pump body and nozzle, as well as the status of washing liquid; Automatic load reduction / switch to backup equipment when thresholds are exceeded; Trigger shutdown protection process when abnormalities are severe.

[0074] Digital twin simulation early warning module: Constructs a digital twin model of the desulfurization system, predicts system anomalies through real-time simulation under multiple operating conditions, and generates early warning information on equipment failure or excessive emissions in advance;

[0075] Processor: The processor is used to handle the calculation process of each formula and the construction calculation process of each model.

[0076] See Figure 1 As shown, the intelligent control method for marine desulfurization devices includes:

[0077] Step 1: Based on a distributed sensor network, multiple types of sensors are deployed at key nodes of the ship's desulfurization system. The data acquisition process is preprocessed through edge computing, including data noise reduction, outlier removal, and data format standardization.

[0078] Step 2: Construct a dynamic calculation model for sulfur flux based on the collected real-time data. Establish the correlation between fuel sulfur content and SO2 generation through historical operating data. Use flue gas flow rate and SO2 concentration in flue gas as input parameters. Calculate the sulfur flux for different fuel types by calling the corresponding model parameters of the correlation relationship.

[0079] Step 3: Based on the acquired dynamic sulfur flux data and combined with the ship's future operating conditions data, construct an alkaline liquid demand forecasting model to predict the alkaline liquid demand in the future time period.

[0080] Step 4: Implement feedforward control based on the alkali demand forecast results, and implement feedback control in combination with the real-time monitoring of flue gas outlet SO2 concentration and scrubbing liquid pH value to form a feedforward-feedback collaborative control mechanism.

[0081] Step 5: Based on the collected flue gas flow rate, flue gas temperature, and sulfur flux data, dynamically optimize the flow rate, temperature, and spray angle of the scrubbing liquid.

[0082] Step Six: Based on the collected equipment operating parameters and washing liquid status data, construct an equipment safety protection mechanism, set the safety threshold range for each device, and monitor the operating status of key equipment in the washing system in real time;

[0083] Step 7: Based on real-time data and historical operating data, construct a digital twin model of the marine desulfurization device, perform real-time simulation of the desulfurization system's operating status, simulate the system's response under different operating conditions, predict potential system problems, and generate early warning information in advance.

[0084] See Figure 2 As shown, based on a distributed sensor network, multiple types of sensors are deployed at key nodes of the ship's desulfurization system. The data acquisition process utilizes edge computing for data preprocessing, including data noise reduction, outlier removal, and data format standardization. Specifically, this includes:

[0085] The key nodes include flue gas inlet pipe, flue gas outlet pipe, different height areas inside the scrubbing tower, alkali storage tank, alkali conveying pipe and seawater replenishment pipe.

[0086] The various types of sensors include infrared SO2 concentration sensors, flue gas flow sensors, and flue gas temperature sensors for collecting flue gas parameters; washing liquid pH sensors, washing liquid flow sensors, washing liquid temperature sensors, and alkali liquid level sensors for collecting washing system parameters; main engine load sensors and ship speed sensors for collecting ship operating condition parameters; and seawater temperature sensors and seawater salinity sensors for collecting environmental parameters.

[0087] The collected data is denoised by edge computing devices to eliminate interference signals, outliers are removed to avoid deviations, and finally the data format is unified to achieve standardization.

[0088] See Figure 3 As shown, a dynamic calculation model for sulfur flux is constructed based on collected real-time data. A correlation between fuel sulfur content and SO2 generation is established using historical operational data. Using flue gas flow rate and SO2 concentration in the flue gas as input parameters, the sulfur flux for different fuel types is calculated by calling the corresponding model parameters based on the correlation relationship. Specifically, this includes:

[0089] A dynamic calculation model for sulfur flux was constructed, and flue gas flow rate and flue gas SO2 concentration were extracted from multi-source data as input parameters, while main engine load and real-time fuel consumption rate were retrieved as correction parameters.

[0090] Instantaneous sulfur emissions are calculated based on flue gas flow rate and SO2 concentration, and the instantaneous sulfur emissions are dynamically corrected by combining the trend of main engine load changes, so as to eliminate the sulfur emission calculation deviation caused by the fluctuation of main engine load.

[0091] The correlation between fuel sulfur content and SO2 generation is established by calling historical operating data. If the ship switches fuel, the corresponding correlation is automatically matched, the parameters of the calculation model are updated, and the sulfur flux under different fuel types is calculated.

[0092] Specifically, based on the extracted parameters, and following the logic that flue gas flow rate reflects the amount of flue gas passing through per unit time and SO2 concentration reflects the sulfur content of flue gas per unit volume, a preliminary calculation of instantaneous sulfur emissions is performed. For example, when the real-time flue gas flow rate is 50 cubic meters per second and the SO2 concentration is 800 milligrams per cubic meter, the preliminary value of sulfur emissions per unit time in the flue gas is obtained through the correlation calculation between the two. That is, when the flue gas flow rate increases, the total amount of flue gas passing through per unit time increases. If the SO2 concentration remains unchanged, the preliminary calculated instantaneous sulfur emissions will increase accordingly. If the SO2 concentration increases, even if the flue gas flow rate is stable, the instantaneous sulfur emissions will also increase accordingly.

[0093] A correlation database of main unit load change range and sulfur emission correction coefficient is established by using historical operating data. The trend of main unit load change is monitored in real time. If the current main unit load increases compared to the previous second, the correction coefficient corresponding to the load increase range is retrieved from the correlation database to adjust the initially calculated instantaneous sulfur emission upward, eliminating the sulfur emission calculation deviation caused by the increase in main unit combustion intensity. If the main unit load decreases, the corresponding correction coefficient is retrieved to adjust downward.

[0094] When a ship switches to fuel with different sulfur contents, the calculation model parameters are adjusted based on real-time extracted fuel sulfur content data. A database of fuel sulfur content-sulfur emission correlations is established using historical data. If the online fuel sulfur content detector detects a change in sulfur content exceeding 5%, the system automatically retrieves the corresponding correlation from the database and updates the basic parameters of the sulfur flux calculation model. This parameter adjustment ensures the calculation results closely match the actual sulfur emissions from fuel combustion, avoiding calculation deviations caused by changes in fuel type. The dynamic sulfur flux calculation formula is as follows:

[0095]

[0096] in, For real-time sulfur flux, For the flow-sulfur conversion dynamic calibration coefficient, This refers to the exhaust flow rate of the main unit. For fuel sulfur content, This is the temperature influence coefficient. The exhaust temperature, This is the SO2 concentration compensation coefficient. This represents the SO2 concentration at the inlet of the scrubbing tower.

[0097] See Figure 4 As shown, based on the acquired dynamic sulfur flux data and combined with future ship operating condition data, an alkali demand forecasting model is constructed to predict the alkali demand in the future time period, specifically including:

[0098] Based on the acquired sulfur flux data, the ship navigation system obtains route information and sea state forecast data, and combines them with main engine load planning data to generate the expected speed changes and main engine load range parameters for future time periods.

[0099] A time-series prediction algorithm was used to construct an alkaline solution demand prediction model. The model was trained using historical sulfur flux data and historical alkaline solution consumption data as training samples to obtain the time-series correlation between sulfur flux and alkaline solution consumption.

[0100] The basic demand for alkali solution is calculated based on the trained alkali solution demand prediction model for the future time period, and the basic demand is then corrected based on the alkali solution storage capacity and the alkali solution concentration decay rate.

[0101] Specifically, the LSTM time-series prediction model is started and parameters are initialized. Basic model parameters trained based on historical samples are loaded, including input layer feature weights, the number of hidden layer neurons, and the output layer mapping relationship. Model parameters are adapted according to the current ship fuel type: if low-sulfur fuel with a sulfur content of 0.5% is currently used, the model calibration parameters under historical low-sulfur fuel conditions are retrieved, and the weight of sulfur flux's influence on alkali demand is reduced; if high-sulfur fuel with a sulfur content of 3.5% is used, this weight is increased.

[0102] The integrated operating condition prediction dataset is input into the adapted LSTM model. With a prediction interval of 5 minutes, the preliminary alkali demand for each period in the next 10-30 minutes is calculated. The model calculates the preliminary alkali demand for that period based on the correlation between sulfur flux and alkali consumption under similar historical operating conditions. The fluctuation range of the demand for each period is generated simultaneously to reflect the confidence level of the prediction results.

[0103] See Figure 5 As shown, a feedforward control mechanism is implemented based on the alkali demand forecast, and a feedback control mechanism is implemented in conjunction with the real-time monitoring of the flue gas outlet SO2 concentration and the scrubbing liquid pH value, forming a feedforward-feedback coordinated control mechanism. Specifically, this includes:

[0104] Based on the alkali demand forecast, the expected alkali demand for different time periods is determined. The feedforward control adjusts the alkali delivery flow rate in advance according to the predicted demand by controlling the speed of the alkali delivery pump.

[0105] The feedback control collects the SO2 concentration at the flue gas outlet and the pH value of the washing liquid in real time, and monitors the alkali concentration through a concentration sensor in the alkali storage tank.

[0106] When the SO2 concentration at the flue gas outlet monitored in real time is higher than the standard value or the pH value of the washing liquid is lower than the lower limit of the optimal range, the alkaline solution delivery flow rate is automatically increased. When the SO2 concentration at the flue gas outlet monitored in real time is lower than the standard value and the pH value of the washing liquid is higher than the upper limit of the optimal range, the alkaline solution delivery flow rate is automatically decreased.

[0107] Specifically, the deviation between the feedforward predicted value and the actual feedback value is compared every 5 minutes. If the deviation between the feedforward predicted alkali demand and the actual flow rate after feedback adjustment exceeds 7%, the cause of the deviation is analyzed. If it is due to the actual change in sulfur flux exceeding the prediction range, the sulfur flux prediction weight of the feedforward model is corrected. If it is due to the alkali concentration decay rate being faster than expected, a concentration decay correction coefficient is added to the next feedforward prediction. After calibration, the corrected parameters are synchronized to the feedforward model so that subsequent pre-adjustment is more in line with the actual operating conditions and the burden of feedback adjustment is reduced.

[0108] By combining real-time data from the concentration sensor inside the alkali storage tank, the system achieves coordinated regulation of concentration and flow rate. If the alkali concentration is detected to be below 20%, the alkali powder addition device is activated first until the concentration rises to the optimal range of 25-30%. If the feedback shows that the SO2 concentration is still above the standard, the alkali flow rate is adjusted according to the feedback logic. If the SO2 concentration is still above 5% after the concentration reaches the standard, the flow rate is increased from 20L / min to 22L / min. If the concentration is above 30%, the flow rate is appropriately reduced to avoid alkali waste, while ensuring that the pH value does not exceed 8.5, thus achieving a coordinated control logic that prioritizes concentration and adapts to flow rate.

[0109] See Figure 6 As shown, based on the collected flue gas flow rate, flue gas temperature, and sulfur flux data, the dynamic optimization of the scrubbing liquid flow rate, temperature, and spray angle specifically includes:

[0110] Based on the collected flue gas flow rate, flue gas temperature, seawater salinity data, and the generated dynamic sulfur flux data, if the sulfur flux is higher than 10 kg / h or the flue gas flow rate increases, the washing liquid flow rate should be increased.

[0111] By adjusting the angle of the adjustable spray nozzles inside the scrubbing tower, the contact area between the scrubbing liquid and the flue gas can be increased, thereby improving the SO2 absorption efficiency.

[0112] When the flue gas temperature is higher than 60℃, control the heat exchange valve of the seawater supply pipeline to introduce low-temperature seawater to regulate the temperature of the washing liquid and stabilize it in the optimal absorption range of 30-45℃.

[0113] Real-time monitoring of seawater salinity; if salinity is above 35‰, increase the frequency of detergent discharge; if salinity is below 35‰, maintain the normal discharge frequency.

[0114] The used washing liquid is filtered and settled in the clarification tank at the bottom of the washing tower to remove solid impurities. After the pH value and SO3²⁻ concentration of the clarified washing liquid are tested and found to be within acceptable limits, it is recycled.

[0115] Specifically, the used washing liquid is first sent to the bottom clarifier for sedimentation treatment, and then filtered through a quartz sand filter to remove suspended particles. After filtration, the pH value and SO3²⁻ concentration of the clarified liquid are tested: if pH ≥ 7.0 and SO3²⁻ concentration ≤ 5 g / L, the circulation pump is started to pump the clarified liquid back to the spray system; if pH < 7.0 or SO3²⁻ concentration > 5 g / L, the clarified liquid is introduced into the sewage pipe, and new alkali solution is added to the circulation system to adjust the pH value to 7.5-8.0. The recycling rate is counted daily. If it is less than 60%, the sedimentation time in the clarifier or the clogging of the filter layer is checked to ensure that the recycling efficiency meets the standard.

[0116] See Figure 7As shown, based on the collected equipment operating parameters and detergent status data, an equipment safety protection mechanism is constructed, safety threshold ranges are set for each device, and the operating status of key equipment in the washing system is monitored in real time. Specifically, this includes:

[0117] Based on the collected equipment operating parameters and washing liquid status data, threshold values ​​are set for key equipment parameters such as alkali transfer pump, washing liquid circulation pump, spray nozzle, and heat exchanger.

[0118] When the detected parameters exceed the threshold, the pump body vibration or current exceeds the standard, the load will be automatically reduced and an early warning will be issued. If the nozzle pressure is lower than the threshold, it is judged to be blocked and the backup nozzle group will be switched.

[0119] Monitor the pH value of the detergent in real time. When the pH value remains below 7.0, add an anti-scaling agent.

[0120] If the SO2 concentration at the flue gas outlet exceeds 1.5 times the IMO standard for 10 consecutive seconds, the pressure inside the tower exceeds the threshold, or the alkali liquid level is below the emergency lower limit, shut down the machine immediately. First, turn off the main engine fuel supply and flue gas valves, then stop the conveying system and start the emergency ventilation.

[0121] Specifically, differentiated protection actions are executed based on the anomaly level:

[0122] Minor anomaly (parameter exceeds threshold by 10%): Triggers audible and visual alarm, marks the location of the abnormal device and the parameter deviation value on the system interface, prompts the operator to pay attention, and does not interrupt the operation of the device;

[0123] Moderate abnormality (parameter exceeds threshold by 10%-20%): Automatic load reduction operation is executed, the speed of the alkali transfer pump is reduced from 1500r / min to 1200r / min, the spray system is switched to the standby nozzle group at the same time, the desulfurization efficiency is kept basically stable, and the warning is upgraded to orange light;

[0124] Severe anomaly (parameter exceeds threshold by 20%): Immediately stop the operation of the abnormal equipment, start the backup equipment, if there is no backup equipment, trigger a local system shutdown, the warning light turns red, and push the fault information to the crew's mobile terminal simultaneously.

[0125] See Figure 8 As shown, based on real-time and historical operating data, a digital twin model of the marine desulfurization device is constructed to simulate the operating status of the desulfurization system in real time, simulate the system's response under different operating conditions, predict potential system problems, and generate early warning information in advance. Specifically, this includes:

[0126] A digital twin model is built based on real-time data and historical operational data, and the virtual model and the physical system are accurately mapped in terms of state through real-time data transmission;

[0127] The simulation results of the system response were obtained by simulating typical operating conditions such as sudden changes in main engine load, changes in fuel sulfur content, and fluctuations in ambient temperature using a digital twin model.

[0128] Based on the simulation system response results, potential problems in the system can be predicted, and early warning information can be generated in advance.

[0129] Specifically, based on synchronized real-time data, the virtual model is driven to achieve dynamic mapping of the physical system: at the equipment level, the virtual pump adjusts its speed display according to real-time current data, and flashes a red warning when vibration exceeds the threshold; at the medium level, the liquid in the virtual scrubbing tower dynamically simulates the pH change of the scrubbing liquid, and the flue gas flow trajectory adjusts with the real-time flue gas flow rate; at the system level, a working condition-state-efficiency correlation mapping is constructed. When the load of the physical system's main unit increases from 70% to 90%, the virtual model synchronously updates the sulfur flux calculation results and dynamically displays the SO2 concentration along the scrubbing tower; during the mapping process, a virtual-real comparison is performed every 5 minutes. If the virtual value of a certain parameter deviates from the physical measured value by more than 5%, the model parameters are automatically fine-tuned to ensure the consistency between the virtual mapping and the physical system.

[0130] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0131] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0132] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart control method for marine desulfurization devices, characterized in that, include: Based on a distributed sensor network, multiple types of sensors are deployed at key nodes of the ship desulfurization system. The data acquisition process is preprocessed through edge computing, including data noise reduction, outlier removal, and data format standardization. A dynamic calculation model for sulfur flux is constructed based on the collected real-time data. The correlation between fuel sulfur content and SO2 generation is established through historical operating data. The sulfur flux under different fuel types is calculated by calling the model parameters corresponding to the correlation relationship, with flue gas flow rate and SO2 concentration in flue gas as input parameters. Based on the acquired dynamic sulfur flux data and combined with the ship's future operating conditions data, an alkaline liquid demand forecasting model is constructed to predict the alkaline liquid demand in the future time period. Feedforward control is implemented based on the alkali demand forecast, and feedback control is implemented in combination with the real-time monitoring of flue gas outlet SO2 concentration and scrubbing liquid pH value, forming a feedforward-feedback collaborative control mechanism. Based on the collected flue gas flow rate, flue gas temperature and sulfur flux data, the flow rate, temperature and spray angle of the scrubbing liquid are dynamically optimized. Based on the collected equipment operating parameters and washing liquid status data, an equipment safety protection mechanism is constructed, the safety threshold range of each device is set, and the operating status of key equipment in the washing system is monitored in real time. Based on real-time and historical operating data, a digital twin model of the marine desulfurization device is constructed to simulate the operating status of the desulfurization system in real time, simulate the system response under different operating conditions, predict potential problems in the system, and generate early warning information in advance.

2. The intelligent control method for marine desulfurization devices according to claim 1, characterized in that, The aforementioned distributed sensor network-based approach deploys multiple types of sensors at key nodes of the ship's desulfurization system. The data acquisition process utilizes edge computing for data preprocessing, including noise reduction, outlier removal, and data format standardization. Specifically, this includes: The key nodes include flue gas inlet pipe, flue gas outlet pipe, different height areas inside the scrubbing tower, alkali storage tank, alkali conveying pipe and seawater replenishment pipe. The various types of sensors include infrared SO2 concentration sensors, flue gas flow sensors, and flue gas temperature sensors for collecting flue gas parameters; washing liquid pH sensors, washing liquid flow sensors, washing liquid temperature sensors, and alkali liquid level sensors for collecting washing system parameters; main engine load sensors and ship speed sensors for collecting ship operating condition parameters; and seawater temperature sensors and seawater salinity sensors for collecting environmental parameters. The collected data is denoised by edge computing devices to eliminate interference signals, outliers are removed to avoid deviations, and finally the data format is unified to achieve standardization.

3. The intelligent control method for marine desulfurization devices according to claim 1, characterized in that, The aforementioned dynamic sulfur flux calculation model, constructed based on real-time collected data, establishes a correlation between fuel sulfur content and SO2 generation using historical operational data. Using flue gas flow rate and SO2 concentration in the flue gas as input parameters, the model calculates the sulfur flux for different fuel types by calling the corresponding parameters of the correlation relationship. Specifically, this includes: A dynamic calculation model for sulfur flux was constructed, and flue gas flow rate and flue gas SO2 concentration were extracted from multi-source data as input parameters, while main engine load and real-time fuel consumption rate were retrieved as correction parameters. Instantaneous sulfur emissions are calculated based on flue gas flow rate and SO2 concentration, and the instantaneous sulfur emissions are dynamically corrected by combining the trend of main engine load changes, so as to eliminate the sulfur emission calculation deviation caused by the fluctuation of main engine load. The correlation between fuel sulfur content and SO2 generation is established by calling historical operating data. If the ship switches fuel, the corresponding correlation is automatically matched, the parameters of the calculation model are updated, and the sulfur flux under different fuel types is calculated.

4. The intelligent control method for marine desulfurization devices according to claim 1, characterized in that, The process of constructing an alkali demand forecasting model based on the acquired dynamic sulfur flux data and combined with future ship operating condition data to predict the alkali demand in the future time period specifically includes: Based on the acquired sulfur flux data, the ship navigation system obtains route information and sea state forecast data, and combines them with main engine load planning data to generate the expected speed changes and main engine load range parameters for future time periods. A time-series prediction algorithm was used to construct an alkaline solution demand prediction model. The model was trained using historical sulfur flux data and historical alkaline solution consumption data as training samples to obtain the time-series correlation between sulfur flux and alkaline solution consumption. The basic demand for alkali solution is calculated based on the trained alkali solution demand prediction model for the future time period, and the basic demand is then corrected based on the alkali solution storage capacity and the alkali solution concentration decay rate.

5. The intelligent control method for marine desulfurization devices according to claim 1, characterized in that, The feedforward control based on the alkali demand forecast, combined with feedback control based on real-time monitoring of flue gas outlet SO2 concentration and scrubbing liquid pH, forms a feedforward-feedback collaborative control mechanism, specifically including: Based on the alkali demand forecast, the expected alkali demand for different time periods is determined. The feedforward control adjusts the alkali delivery flow rate in advance according to the predicted demand by controlling the rotation speed of the alkali delivery pump. The feedback control collects the SO2 concentration at the flue gas outlet and the pH value of the washing liquid in real time, and monitors the alkali concentration through a concentration sensor in the alkali storage tank. When the SO2 concentration at the flue gas outlet monitored in real time is higher than the standard value or the pH value of the washing liquid is lower than the lower limit of the optimal range, the alkaline solution delivery flow rate is automatically increased. When the SO2 concentration at the flue gas outlet monitored in real time is lower than the standard value and the pH value of the washing liquid is higher than the upper limit of the optimal range, the alkaline solution delivery flow rate is automatically decreased.

6. The intelligent control method for marine desulfurization devices according to claim 1, characterized in that, The dynamic optimization of the scrubbing liquid flow rate, temperature, and spray angle based on the collected flue gas flow rate, temperature, and sulfur flux data specifically includes: Based on the collected flue gas flow rate, flue gas temperature, seawater salinity data, and the generated dynamic sulfur flux data, if the sulfur flux is higher than 10 kg / h or the flue gas flow rate increases, the washing liquid flow rate should be increased. By adjusting the angle of the adjustable spray nozzles inside the scrubbing tower, the contact area between the scrubbing liquid and the flue gas can be increased, thereby improving the SO2 absorption efficiency. When the flue gas temperature is higher than 60℃, control the heat exchange valve of the seawater supply pipeline to introduce low-temperature seawater to regulate the temperature of the washing liquid and stabilize it in the optimal absorption range of 30-45℃. Real-time monitoring of seawater salinity; if salinity is above 35‰, increase the frequency of detergent discharge; if salinity is below 35‰, maintain the normal discharge frequency. The used washing liquid is filtered and settled in the clarification tank at the bottom of the washing tower to remove solid impurities. After the pH value and SO3²⁻ concentration of the clarified washing liquid are tested and found to be within acceptable limits, it is recycled.

7. The intelligent control method for marine desulfurization devices according to claim 1, characterized in that, The aforementioned mechanism for constructing a safety protection system based on collected equipment operating parameters and detergent status data, setting safety threshold ranges for each device, and real-time monitoring of the operating status of key equipment in the washing system specifically includes: Based on the collected equipment operating parameters and washing liquid status data, threshold values ​​are set for key equipment parameters such as alkali transfer pump, washing liquid circulation pump, spray nozzle, and heat exchanger. When the detected parameters exceed the threshold, the pump body vibration or current exceeds the standard, the load will be automatically reduced and an early warning will be issued. If the nozzle pressure is lower than the threshold, it is judged to be blocked and the backup nozzle group will be switched. Monitor the pH value of the detergent in real time. When the pH value remains below 7.0, add an anti-scaling agent. If the SO2 concentration at the flue gas outlet exceeds 1.5 times the IMO standard for 10 consecutive seconds, the pressure inside the tower exceeds the threshold, or the alkali liquid level is below the emergency lower limit, shut down the machine immediately. First, turn off the main engine fuel supply and flue gas valves, then stop the conveying system and start the emergency ventilation.

8. The intelligent control method for marine desulfurization devices according to claim 1, characterized in that, The process of constructing a digital twin model of the marine desulfurization device based on real-time and historical operating data, simulating the operating status of the desulfurization system in real time, simulating the system's response under different operating conditions, predicting potential system problems, and generating early warning information specifically includes: A digital twin model is built based on real-time data and historical operational data, and the virtual model and the physical system are accurately mapped in terms of state through real-time data transmission; The simulation results of the system response were obtained by simulating typical operating conditions such as sudden changes in main engine load, changes in fuel sulfur content, and fluctuations in ambient temperature using a digital twin model. Based on the simulation system response results, potential problems in the system can be predicted, and early warning information can be generated in advance.

9. A high-efficiency scrubbing system for marine desulfurization devices, used to implement the intelligent control method for marine desulfurization devices as described in any one of claims 1-8, characterized in that, include: Distributed sensor network module: Multiple types of sensors are deployed at key nodes such as flue gas inlet and outlet, scrubbing tower, and alkali pipeline to collect flue gas, scrubbing liquid, ship operating conditions and environmental parameters in real time; Edge computing preprocessing module: Performs real-time noise reduction, outlier removal, and format standardization on raw sensor data using edge computing devices; Sulfur flux dynamic calculation module: Based on flue gas flow rate, SO2 concentration and main engine load data, combined with the historical model of fuel sulfur content-SO2 generation, it dynamically calculates the real-time sulfur flux under different fuel types. Alkali demand forecasting module: Based on sulfur flux data and future ship operating conditions, it outputs the alkali demand for future periods through a time series forecasting model, and makes secondary corrections based on storage capacity and concentration decay. Feedforward-feedback coordinated control module: The flow rate of the delivery pump is adjusted based on the predicted value of the alkali solution, and the dynamic fine-tuning is performed in real time based on the SO2 concentration at the flue gas outlet and the pH value of the washing liquid. The dynamic optimization module for washing parameters dynamically optimizes the washing liquid flow rate, spray angle, and temperature based on flue gas flow rate, temperature, and sulfur flux data to maintain the best SO2 absorption efficiency. Equipment safety protection module: Real-time monitoring of the operating parameters of key equipment such as pump body and nozzle, as well as the status of washing liquid; Automatic load reduction / switch to backup equipment when thresholds are exceeded; Trigger shutdown protection process when abnormalities are severe. Digital twin simulation early warning module: Constructs a digital twin model of the desulfurization system, predicts system anomalies through real-time simulation under multiple operating conditions, and generates early warning information on equipment failure or excessive emissions in advance; Processor: The processor is used to handle the calculation process of each formula and the construction calculation process of each model.

Images