A method and system for operation optimization based on a desulfurization system

Abstract

The present application provides a kind of method and system based on the operation optimization of desulfurization system, it is related to desulfurization equipment operation control technical field, the present application obtains the health coefficient of pump by calculating the real-time stability and performance loss degree of liquid circulating pump, oxygen pump, and judges health degree, the blockage coefficient of spray system is calculated, the health degree of pump is set by setting blockage threshold value, the equipment failure rate of desulfurization system is calculated and judged, according to the cost ratio of desulfurization system and equipment failure rate, historical operation parameter data is screened to form training data set, desulfurization system operation optimization model is constructed by long short-term memory neural network, and the operation optimization model is trained, then the real-time detection parameters are collected as the input of model, the optimized result is output by optimization model, the operation optimization model is optimized by sparrow search algorithm, and according to loss function as fitness evaluation index.

Description

Technical Field

[0001] This invention relates to the field of desulfurization equipment operation control technology, specifically to a method and system for optimizing the operation of a desulfurization system. Background Technology

[0002] For decades, coal has been a major pillar of industrialization and energy development in many countries, but it also faces serious environmental problems. Sulfur dioxide produced during coal combustion not only severely impacts the environment but also poses a threat to human health. Traditional methods of flue gas desulfurization rely too heavily on manual control, often resulting in substandard emissions and excessive costs. Therefore, optimizing and upgrading desulfurization systems has become an urgent issue to address.

[0003] In the prior art, CN115355163A discloses an optimization and management method for a desulfurization system. This method establishes a data-fusion desulfurization efficiency model based on the system's operational data. This model is used in actual production to optimize the desulfurization system in real time. While ensuring desulfurization efficiency, it outputs the optimal slurry circulation pump combination for on-site operators to adjust. This reduces the total current of the slurry circulation pumps, thereby lowering the overall power consumption of the desulfurization system. An edge management device tracks the start-stop intervals and cumulative operating time of the slurry circulation pumps, while also monitoring their health status to ensure long-term continuous normal operation. This solves the problem of relatively crude experience-based adjustment methods leading to high power consumption in desulfurization. However, this solution only considers the energy consumption of the slurry circulation pumps, neglecting the cost of desulfurizing agents and system downtime due to equipment failure. Furthermore, it lacks optimization of the overall parameters of the desulfurization system.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for optimizing the operation of a desulfurization system, in order to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A method and system for optimizing the operation of a desulfurization system, comprising the following steps:

[0008] Step 1: Obtain historical data on current, vibration, temperature, and noise of the pumps in the desulfurization system to be optimized. The pumps include slurry circulation pumps and oxygen pumps. By calculating the stability and performance loss of the pumps, obtain the pump health coefficient. Determine the health level of the pumps based on the health coefficient and arrange maintenance plans based on the overall health level.

[0009] Step 2: Obtain the pipeline pressure, slurry flow rate, and slurry level difference in the slurry tank at the same power of the slurry circulation pump from historical data. Calculate the blockage coefficient of the spray system, which includes pipelines and nozzles. By setting the blockage threshold and the health status of the pump, calculate and determine the equipment failure rate of the desulfurization system.

[0010] Step 3: Obtain the operating parameter data of the desulfurization system from the historical data, calculate the cost ratio of each data set by the cumulative sulfur dioxide amount and operating cost, and filter the operating parameters according to the equipment failure rate. Mark the priority of the pump history on the filtered data to form a sample dataset. Then, preprocess the sample dataset to form a training dataset.

[0011] Step 4: Construct an optimization model for the desulfurization system operation using a long short-term memory neural network. Use the detection parameters in the training dataset as the input to the optimization model and the execution parameters in the training dataset as the output to train the model. Then, collect real-time detection parameters as the input to the model and output the optimized results through the optimization model.

[0012] Step 5: Randomly generate a group of sparrow individuals using the sparrow search algorithm. Each individual represents a set of hyperparameter configurations for running the optimization model. Explore a better hyperparameter space by adjusting the position of the individuals. Train and run the optimization model using the hyperparameter configurations of each individual. Optimize the model using the loss function as a fitness evaluation index.

[0013] Furthermore, the specific steps for obtaining the pump's health coefficient by calculating the pump's stability and performance loss are as follows:

[0014] Historical data on current, vibration, temperature, and noise were analyzed according to specific time periods. Statistical analysis was conducted on historical data for each day at different time intervals. Once the data is collected, the fluctuations in the pump's current, vibration, temperature, and noise are calculated. The pump's stability is then calculated based on the weights of these fluctuations.

[0015] The formulas for the fluctuation of the pump's current, vibration, temperature, and noise are as follows;

[0016] ;

[0017] ;

[0018] ;

[0019] ;

[0020] in, Due to the fluctuation of current, For the wave nature of vibration, For temperature fluctuations, For the fluctuation of noise, In order to be in Current at any moment In order to be in Vibration at any moment In order to be in Temperature at any moment In order to be in Momentary noise, The time of data collection Time period Number of data collection times within;

[0021] The formula for calculating the stability of a pump is:

[0022] ;

[0023] in, For the stability of the pump, The weights are assigned to the fluctuations of current, vibration, temperature, and noise, respectively.

[0024] Furthermore, the formula for calculating the degree of performance loss of the pump is as follows;

[0025] ;

[0026] in, The degree of performance loss of the pump, Time period The stability of the pump, Time period The stability of the pump To ensure the stability of the pump during the initial operation of the desulfurization system.

[0027] Furthermore, the formula for calculating the health coefficient of the pump is as follows:

[0028] ;

[0029] in, For the pump's health coefficient, The degree of performance loss of the pump, Time period The stability of the pump, Time period The stability of the pump, The degree of performance loss of the pump;

[0030] The specific method for judging the pump's health level based on the health coefficient and arranging a maintenance plan based on the overall health level is as follows:

[0031] when When the pump is in good condition, its priority is 1, and no maintenance is required.

[0032] when At this time, the pump's status is judged as normal, with a priority of 2, and no maintenance is required;

[0033] when When the pump is deemed unhealthy, with a priority of 3, it requires maintenance.

[0034] in, These are the criteria for judging health and unhealth. .

[0035] Furthermore, the calculation steps for the clogging coefficient of the spray system are as follows;

[0036] The formula for calculating the pipeline blockage coefficient is as follows:

[0037] ;

[0038] in, The blockage coefficient of the pipeline. This represents the current slurry flow rate in the pipeline. This represents the slurry flow rate at the moment the pipeline maintenance was completed. This represents the current pipeline pressure. The pipeline pressure at the moment the maintenance was completed;

[0039] The formula for calculating the clogging coefficient of the nozzle is:

[0040] ;

[0041] in, The nozzle clogging coefficient is... The difference in slurry tank level before and after the slurry pump circulation pump starts, assuming the nozzle is not clogged. This is the time it takes for the liquid level difference backflow to disappear when the nozzle is not clogged. This represents the difference in slurry tank level before and after the slurry pump circulation pump starts. This represents the time it takes for the backflow of the current liquid level difference to disappear.

[0042] The formula for calculating the clogging coefficient of the spray system is as follows:

[0043] ;

[0044] in, The clogging coefficient of the spray system;

[0045] when In this case, determine if the sprinkler system is clogged;

[0046] The formula for calculating the equipment failure rate of the desulfurization system is as follows:

[0047] ;

[0048] in, The failure rate of the desulfurization system, The number of sprinkler systems identified as clogged based on the clogging coefficient. This represents the total number of sprinkler systems. The number of pumps deemed unhealthy based on a health coefficient. This represents the total number of pumps.

[0049] Furthermore, the operating parameters include detection parameters and execution parameters. The detection parameters include flue gas intake volume, flue gas temperature, sulfur dioxide concentration, slurry tank level, limestone concentration, pH value, gypsum concentration, number of circulating pumps, number of oxygen pumps, real-time power consumption unit price, real-time limestone cost, and cumulative desulfurization time of the system. The execution parameters include the start / stop status of the circulating pumps and the start / stop status of the oxygen pumps.

[0050] The formula for calculating the cost ratio of each set of data based on the cumulative amount of sulfur dioxide and operating costs is as follows:

[0051] ;

[0052] in, For cost ratio, The unit price is the real-time power consumption. For the real-time cost of limestone, This refers to the flue gas intake volume. This refers to the concentration of sulfur dioxide.

[0053] The specific method for filtering operating parameters and marking the priority of pump history in the filtered data to form a sample dataset is as follows:

[0054] Historical data is divided into days, with each group consisting of ten days' worth of data. The data is then sorted based on cost comparison, from largest to smallest, and also based on equipment failure rate, from largest to smallest. The top two-tenths of the data in each group are then removed. The filtered data are then timestamped, and the circulating pumps and oxygen pumps in the data are numbered and prioritized.

[0055] Furthermore, the desulfurization system operation optimization model includes an input gate, a forget gate, and an output gate;

[0056] The equation for the forgetting gate is:

[0057] ;

[0058] in, The current open / closed state of the forget gate. This represents the sigmoid function. For the weight of the forget gate, Forget gate bias parameters, This is the hidden state from the previous moment. Input switch for the current data;

[0059] Determine how many cell states from the previous time step need to be retained in the current time step;

[0060] The equation for the input gate is:

[0061] ;

[0062] in, To input the current open / closed state of the gate, This represents the sigmoid function. For the input gate weights, For the input gate bias parameters, This is the output hidden state from the previous time step. This refers to the current detection parameter data;

[0063] ;

[0064] in, Candidate cell state To represent the hyperbolic tangent function, For candidate cell weights, These are the candidate cell state bias parameters. This is the hidden state from the previous moment. Input switch for the current data;

[0065] Determine how much of the network's input data at the current moment needs to be saved to the cell state;

[0066] The updated equation is:

[0067] ;

[0068] in, Current cell state This represents the cell state at the previous moment;

[0069] The equation for the output gate is:

[0070] ;

[0071] ;

[0072] in, Output the current open / closed state of the gate. This represents the sigmoid function. For the output gate weights, These are the output gate bias parameters. This is the hidden state from the previous moment. The current data input switch, This contains the execution parameter data for the current moment.

[0073] Control how much of the current cell state needs to be output to the current output value;

[0074] The mathematical expressions for the sigmoid function and the hyperbolic tangent function are:

[0075] ;

[0076] ;

[0077] in, For the sigmoid function, For the input of the sigmoid function, It is the hyperbolic tangent function. For function input.

[0078] Furthermore, the process of the sparrow search algorithm may include discoverer and follower;

[0079] The formula for updating the discoverer's location is:

[0080] ;

[0081] in, For the updated number The middle generation Only sparrows in the first The position of the sparrow The types and dimensions of weights in the optimization model during hyperparameter configuration. The weight dimensions in the hyperparameter configuration are used to optimize the model. For random numbers that follow a normal distribution, A uniformly random number in [0,1]. It is the warning threshold, and its value ranges from [0.5, 1].

[0082] Follower position update formula;

[0083] ;

[0084] in, Let t be the position of the best-fit sparrow in the population in dimension d. To generate a random number from -1 to 1;

[0085] The formula for calculating the loss function is as follows;

[0086] ;

[0087] in, For the sample size, For the true value, These are predicted values.

[0088] The present invention also provides a system for optimizing the operation of a desulfurization system, wherein the optimized system is used to execute the above-described method for optimizing the operation of a desulfurization system, comprising:

[0089] The data acquisition model is used to obtain historical data on current, vibration, temperature and noise of slurry circulation pump and oxygen pump, pipeline pressure, slurry flow rate and slurry level difference in slurry tank when the equipment is started and stopped, and operating parameter data of desulfurization system.

[0090] The equipment analysis module is used to calculate the blockage coefficient and equipment failure rate of the spraying system based on the pipeline pressure, slurry flow rate and slurry level difference in the slurry tank when the equipment starts and stops, under the same power of the slurry circulation pump. It also calculates the cost ratio of each set of data by the cumulative amount of sulfur dioxide and the operating cost.

[0091] The data filtering module is used to filter operating parameters based on cost ratio and equipment failure rate, mark the priority of pump history on the filtered data to form a sample dataset, and then preprocess the sample dataset to form a training dataset.

[0092] The optimization system module constructs an optimization model for the desulfurization system operation through a long short-term memory neural network. The detection parameters in the training dataset are used as the input to the optimization model, and the execution parameters in the training dataset are used as the output to train the model. Then, by collecting real-time detection parameters as the input to the model, the optimization model outputs the optimization result.

[0093] The system parameter optimization module is used to randomly generate a group of sparrow individuals using a sparrow search algorithm. Each individual represents a set of hyperparameter configurations for running the optimization model. By adjusting the position of the individuals, a better hyperparameter space is explored. The optimization model is trained using the hyperparameter configuration of each individual and optimized using a loss function as a fitness evaluation index.

[0094] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention obtains the health coefficient of the pump by calculating the real-time stability and performance loss of the liquid circulation pump and oxygen pump, and judges the health level. By calculating the blockage coefficient of the spray system and setting the blockage threshold and the health level of the pump, the equipment failure rate of the desulfurization system is calculated and judged. Based on the cost ratio of the desulfurization system and the equipment failure rate, historical operating parameter data is filtered to form a training dataset. A desulfurization system operation optimization model is constructed through a long short-term memory neural network and trained. Then, real-time detection parameters are collected as input to the model, and the optimization results are output through the optimization model. The operation optimization model is optimized by using a sparrow search algorithm and using a loss function as a fitness evaluation index.

[0095] This invention uses historical operating parameter data to form a training dataset based on the cost ratio of the desulfurization system and the equipment failure rate. It fully considers the energy consumption of the slurry circulation pump and oxygen pump, the cost of the desulfurizing agent, and the factors of equipment failure. By optimizing the operation of the desulfurization system using the filtered data, it effectively eliminates situations where the desulfurization effect is good but will damage the equipment and cause additional energy consumption and excessive reagent costs. It optimizes the desulfurization system more comprehensively and meticulously from multiple perspectives. By constructing a desulfurization system operation optimization model through a long short-term memory neural network and optimizing the optimization model, the overall parameters of the desulfurization system can be optimized more accurately and quickly. Attached Figure Description

[0096] Figure 1 This is a schematic diagram of the overall method flow of the present invention;

[0097] Figure 2 This is a schematic diagram of the overall system structure of the present invention. Detailed Implementation

[0098] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0099] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0100] Example:

[0101] Please see Figure 1 The present invention provides a technical solution:

[0102] A method for optimizing the operation of a desulfurization system, comprising the following steps:

[0103] Step 1: Obtain historical data on current, vibration, temperature, and noise of the pumps in the desulfurization system to be optimized. The pumps include slurry circulation pumps and oxygen pumps. By calculating the stability and performance loss of the pumps, obtain the pump health coefficient. Determine the health level of the pumps based on the health coefficient and arrange maintenance plans based on the overall health level.

[0104] The specific steps for obtaining the pump's health coefficient by calculating the pump's stability and performance loss are as follows:

[0105] Historical data on current, vibration, temperature, and noise were analyzed according to specific time periods. Statistical analysis was conducted on historical data for each day at different time intervals. Once a data acquisition is performed, the fluctuations in the pump's current, vibration, temperature, and noise are calculated. The pump's stability is then calculated based on the weights assigned to the fluctuations in current, vibration, temperature, and noise.

[0106] Slurry circulation pumps and oxygen pumps primarily consist of an impeller driven by a motor. Therefore, the health of these pumps can be assessed by monitoring current, vibration, temperature, and noise. The magnitude of the current is directly related to the pump's load. Under normal circumstances, the current should remain within the design range. Abnormal fluctuations in current may indicate changes in the pump's load, such as blockage, wear, or motor problems. For example, an increase in current usually means increased pump resistance, possibly due to blockage or mechanical failure. Vibration levels reflect the pump's mechanical stability. During normal operation, vibration should be kept at a low level. Abnormal vibration is usually related to mechanical failures, such as imbalance, looseness, or bearing damage. Increased temperature usually indicates increased friction, poor lubrication, or overload, which may lead to component damage. Sustained high temperatures can provide early warning of potential faults. Pumps generate a certain amount of noise during normal operation; abnormal noise changes can indicate internal faults. For example, increased noise may indicate cavitation, impeller damage, or other mechanical problems. By comprehensively analyzing the fluctuations of these parameters, the operating status and health of slurry circulation pumps and oxygen pumps can be fully identified.

[0107] In this embodiment, the formulas for the fluctuation of the pump's current, vibration, temperature, and noise are as follows;

[0108] ;

[0109] ;

[0110] ;

[0111] ;

[0112] in, Due to the fluctuation of current, For the wave nature of vibration, For temperature fluctuations, For the fluctuation of noise, In order to be in Current at any moment In order to be in Vibration at any moment In order to be in Temperature at any moment In order to be in Momentary noise, The time of data collection Time period Number of times data was collected within the scope.

[0113] The time period is one day, and the fluctuations of pump current, vibration, temperature and noise are calculated within one day.

[0114] The formula for calculating the stability of a pump is:

[0115] ;

[0116] in, For the stability of the pump, The weights are assigned to the fluctuations of current, vibration, temperature, and noise, respectively.

[0117] Pump stability reflects its stable state. Higher real-time stability indicates a greater probability of pump failure. Real-time stability indicators can help identify potential failure risks. By monitoring changes in real-time stability, problems can be detected promptly.

[0118] In this embodiment, the formula for calculating the degree of performance loss of the pump is as follows;

[0119] ;

[0120] in, The degree of performance loss of the pump, Time period The stability of the pump, Time period The stability of the pump To ensure the stability of the pump during the initial operation of the desulfurization system.

[0121] The stability of the pump during the initial operation of the desulfurization system is most representative, reflecting the pump's optimal stability. The statistical time length for pump stability is... This can be done over one or more days, reducing the impact of unforeseen factors on the stability assessment of the pump.

[0122] The degree of performance loss is used to quantify the difference between the overall performance of the pump under its current state and its overall performance under its rated state. The greater the degree of performance loss, the lower the pump's current operating efficiency.

[0123] In this embodiment, the formula for calculating the health coefficient of the pump is:

[0124] ;

[0125] in, For the pump's health coefficient, The degree of performance loss of the pump, Time period The stability of the pump, Time period The stability of the pump, The degree of performance loss of the pump;

[0126] The specific method for judging the pump's health level based on the health coefficient and arranging a maintenance plan based on the overall health level is as follows:

[0127] when When the pump is in good condition, its priority is 1, and no maintenance is required.

[0128] when At this time, the pump's status is judged as normal, with a priority of 2, and no maintenance is required;

[0129] when When the pump is deemed unhealthy, with a priority of 3, it requires maintenance.

[0130] in, These are the criteria for judging health and unhealth. .

[0131] Step 2: Obtain historical data on the pipeline pressure, slurry flow rate, and slurry level difference in the slurry tank at the same power for the slurry circulation pump. Calculate the clogging coefficient of the spray system, which includes pipelines and nozzles. By setting clogging thresholds and pump health status, calculate and determine the equipment failure rate of the desulfurization system.

[0132] In desulfurization systems, pipeline pressure is a key parameter in fluid dynamics. During normal operation of the spray system, the pressure within the pipeline should be maintained within a certain range. If blockage occurs, the pressure will be affected, typically manifesting as an increased pressure drop. Monitoring pressure changes can determine if there is an increase in flow resistance. Flow rate is an important indicator of liquid transport capacity. In spray systems, the stability and changes in flow rate reflect system efficiency and blockage status. When the flow rate is significantly lower than expected, it may indicate a blockage, obstructing liquid flow. The slurry level difference in the slurry tank also provides important information during equipment start-up and shutdown. Level changes reflect the fluid's inflow and outflow. Abnormal level changes, such as a failure to quickly drop the level upon shutdown, may indicate a blockage, preventing free liquid flow.

[0133] In this embodiment, the calculation steps for the clogging coefficient of the spray system are as follows;

[0134] The formula for calculating the pipeline blockage coefficient is as follows:

[0135] ;

[0136] in, The blockage coefficient of the pipeline. This represents the current slurry flow rate in the pipeline. This represents the slurry flow rate at the moment the pipeline maintenance was completed. This represents the current pipeline pressure. This refers to the pipeline pressure at the moment the maintenance was completed.

[0137] The formula for calculating the clogging coefficient of the nozzle is:

[0138] ;

[0139] in, The nozzle clogging coefficient is... The difference in slurry tank level before and after the slurry pump circulation pump starts, assuming the nozzle is not clogged. This is the time it takes for the liquid level difference backflow to disappear when the nozzle is not clogged. This represents the difference in slurry tank level before and after the slurry pump circulation pump starts. This represents the time it takes for the backflow of the liquid level difference to disappear at the current moment.

[0140] The formula for calculating the clogging coefficient of the spray system is as follows:

[0141] ;

[0142] in, The clogging coefficient of the spray system;

[0143] when When this happens, it is determined that the sprinkler system is blocked.

[0144] The clogging factor is a comprehensive indicator. Generally, an increase in the clogging factor means that flow is obstructed, and the fluid cannot pass through the system smoothly at the designed flow rate. This factor can be used to link pressure loss, flow rate, and liquid level changes through a formula, forming a quantitative standard to facilitate monitoring and judgment of clogging conditions.

[0145] The formula for calculating the equipment failure rate of the desulfurization system is as follows:

[0146] ;

[0147] in, The failure rate of the desulfurization system, The number of sprinkler systems identified as clogged based on the clogging coefficient. This represents the total number of sprinkler systems. The number of pumps deemed unhealthy based on a health coefficient. This represents the total number of pumps.

[0148] Equipment failure rate indicates the proportion of equipment that fails in the desulfurization system. Equipment failure rate means that the desulfurization system cannot be used normally or that the desulfurization effect is seriously affected.

[0149] Step 3: Obtain the operating parameter data of the desulfurization system from historical data, calculate the cost ratio of each data set by the cumulative sulfur dioxide amount and operating cost, and filter the operating parameters according to the equipment failure rate. Mark the priority of the pump history on the filtered data to form a sample dataset. Then, preprocess the sample dataset to form a training dataset.

[0150] In this embodiment, the operating parameters include detection parameters and execution parameters. The detection parameters include flue gas intake volume, flue gas temperature, sulfur dioxide concentration, slurry tank level, limestone concentration, pH value, gypsum concentration, number of circulating pumps, number of oxygen pumps, real-time power consumption unit price, real-time limestone cost, and cumulative desulfurization time of the system. The execution parameters include the start / stop status of the circulating pumps and the start / stop status of the oxygen pumps.

[0151] The flue gas intake directly affects the scale and efficiency of the desulfurization reaction. Excessive intake may result in insufficient reaction of the limestone desulfurizing agent, reducing desulfurization efficiency. Flue gas temperature affects the rate and efficiency of the desulfurization reaction. Higher temperatures generally accelerate the reaction, but excessively high temperatures may damage equipment or cause incomplete reaction. Sulfur dioxide concentration affects the consumption of the desulfurizing agent and the reaction rate; excessively high concentrations will accelerate agent depletion. A slurry tank level that is too low may cause the pumps to run dry or result in insufficient desulfurizing agent concentration; a level that is too high may affect reaction efficiency or cause slurry overflow. Limestone concentration directly affects the efficiency of the desulfurization reaction. Insufficient concentration will lead to incomplete sulfur dioxide removal. pH value affects the desulfurization reaction. The optimal pH range is typically between 5.5 and 7.5. Both excessively high and low pH levels will affect reaction efficiency and gypsum formation. Excessively high gypsum concentration may cause precipitation or scaling, affecting system operation. The number of circulating pumps affects the slurry circulation efficiency. Insufficient pumps may lead to poor slurry flow, affecting desulfurization performance; excessive pumps will result in resource waste. The number of oxygen pumps affects the oxygen supply in the reaction, influencing the limestone oxidation reaction and sulfur dioxide removal efficiency. Power consumption affects operating costs, thus impacting overall economics. High power consumption may lead to increased operating expenses. The cost of limestone directly affects the economics of the desulfurization system. Increased costs may lead to decreased operating profits. Cumulative desulfurization time is related to equipment operating efficiency and wear levels. Long-term operation may lead to increased equipment failures.

[0152] The filtered data is labeled with the historical priority of the pumps to form a sample dataset. The sample dataset is then preprocessed to form a training dataset. Each pump is labeled with a priority for training. This helps to select pumps based on their health status and prevents unexpected shutdowns during desulfurization due to pump health issues, which could lead to serious consequences such as non-compliance with sulfur dioxide emission standards and forced production stoppages. The preprocessing uses normalization to scale each data value to a range between 0 and 1, ensuring that all input data have the same value range. This avoids some features having unreasonable impacts on the model training process due to large values. Data normalization can improve training efficiency and model accuracy.

[0153] In this embodiment, the formula for calculating the cost ratio of each set of data based on the cumulative amount of sulfur dioxide and operating costs is as follows:

[0154] ;

[0155] in, For cost ratio, The unit price is the real-time power consumption. For the real-time cost of limestone, This refers to the flue gas intake volume. This refers to the concentration of sulfur dioxide.

[0156] The specific method for filtering operating parameters and marking the priority of pump history in the filtered data to form a sample dataset is as follows:

[0157] Historical data is divided into days, with each group consisting of ten days' worth of data. The data is then sorted based on cost comparison, from largest to smallest, and also based on equipment failure rate, from largest to smallest. The top two-tenths of the data in each group are then removed. The filtered data are then timestamped, and the circulating pumps and oxygen pumps in the data are numbered and prioritized.

[0158] The cost ratio reflects the economic cost per unit of sulfur dioxide removal. By selecting operating parameter combinations with low cost ratios, operating costs can be effectively reduced, improving economic efficiency. Equipment failure rate is a key indicator for assessing system health. A high failure rate can lead to frequent downtime and maintenance, increasing operating costs. By analyzing equipment failure rates, parameters that perform stably during operation can be selected, reducing the risk of failures and improving the overall reliability of the system. Combining cost ratio and failure rate in the selection process allows for the evaluation of operating parameters from multiple perspectives, ensuring that the system is not only economically sound but also operates stably.

[0159] Step 4: Construct an optimization model for the desulfurization system operation using a long short-term memory neural network. Use the detection parameters in the training dataset as the input to the optimization model and the execution parameters in the training dataset as the output to train the model. Then, collect real-time detection parameters as the input to the model and output the optimized results through the optimization model.

[0160] Long Short-Term Memory (LSTM) neural networks can effectively capture long-term dependencies in time-series data. This is particularly important for optimizing operating parameters in desulfurization systems, as these data are typically time-dependent. LSTM neural networks can learn the patterns of data change across different time periods, thus more accurately predicting the operating parameters of the desulfurization system. Complex nonlinear relationships often exist between operating parameters and system performance in desulfurization systems. LSTM neural networks can effectively model these nonlinear relationships, thereby improving the accuracy and effectiveness of optimization.

[0161] In this embodiment, the desulfurization system operation optimization model includes an input gate, a forget gate, and an output gate;

[0162] The input gate controls the input information flowing into the current cell state. The input gate `it` receives the current input and the hidden state from the previous time step, and uses the sigmoid function to determine how much input information to use, producing a value in the range [0,1].

[0163] The equation for the input gate is:

[0164] ;

[0165] in, To input the current open / closed state of the gate, This represents the sigmoid function. For the input gate weights, For the input gate bias parameters, This is the output hidden state from the previous time step. This refers to the current detection parameter data;

[0166] ;

[0167] in, Candidate cell state To represent the hyperbolic tangent function, For candidate cell weights, These are the candidate cell state bias parameters. This is the hidden state from the previous moment. Input switch for the current data;

[0168] Determine how much of the network's input data at the current moment needs to be saved to the cell state;

[0169] The updated equation is:

[0170] ;

[0171] in, Current cell state This represents the cell state at the previous moment;

[0172] The forget gate controls the flow of historical information into the current cell state. For neurons in the hidden layer of a long short-term memory neural network, some historical information needs to be forgotten before computation. The forget gate receives the hidden state of the previous time step and the input data of the current time step as input switches, and is mapped to the range [0,1] by the sigmoid function.

[0173] The equation for the forgetting gate is:

[0174] ;

[0175] in, The current open / closed state of the forget gate. This represents the sigmoid function. For the weight of the forget gate, Forget gate bias parameters, This is the hidden state from the previous moment. Input switch for the current data;

[0176] Determine how many cell states from the previous time step need to be retained in the current time step;

[0177] The output gate controls the influence of the current time-in memory unit on the current time-in execution parameter data, i.e., which part will be output at the current time. The output gate receives the hidden state of the previous time step and the current time-in data input switch, which is mapped to the range [0,1] by the sigmoid function;

[0178] The equation for the output gate is:

[0179] ;

[0180] ;

[0181] in, Output the current open / closed state of the gate. This represents the sigmoid function. For the output gate weights, These are the output gate bias parameters. This is the hidden state from the previous moment. The current data input switch, This contains the execution parameter data for the current moment.

[0182] Control how much of the current cell state needs to be output to the current output value;

[0183] The mathematical expressions for the sigmoid function and the hyperbolic tangent function are:

[0184] ;

[0185] ;

[0186] in, For the sigmoid function, For the input of the sigmoid function, It is the hyperbolic tangent function. For function input.

[0187] Step 5: Randomly generate a group of sparrow individuals using the sparrow search algorithm. Each individual represents a set of hyperparameter configurations for running the optimization model. Explore a better hyperparameter space by adjusting the position of the individuals. Train and run the optimization model using the hyperparameter configurations of each individual. Optimize the model using the loss function as a fitness evaluation index.

[0188] The Sparrow Search algorithm is a swarm intelligence optimization algorithm that effectively explores the parameter space and avoids getting trapped in local optima. This is crucial for hyperparameter tuning of Long Short-Term Memory (LSTM) neural network models, as appropriate hyperparameter settings can significantly improve model performance. By simulating the foraging behavior of sparrows, the Sparrow Search algorithm accelerates the search process and improves convergence speed. This allows LSM neural networks to find the optimal parameter combination more quickly during training, thus shortening training time. The Sparrow Search algorithm exhibits good adaptability in dynamic environments. For LSM neural network models that require real-time adjustments, it can dynamically optimize hyperparameters based on the latest data, improving the model's flexibility and stability.

[0189] In this embodiment, the process of the sparrow search algorithm may include discoverer and follower;

[0190] The formula for updating the discoverer's location is:

[0191] ;

[0192] in, For the updated number The middle generation Only sparrows in the first The position of the sparrow The types and dimensions of weights in the optimization model during hyperparameter configuration. The weight dimensions in the hyperparameter configuration are used to optimize the model. For random numbers that follow a normal distribution, A uniformly random number in [0,1]. It is the warning threshold, and its value range is [0.5, 1].

[0193] Follower position update formula;

[0194] ;

[0195] in, Let t be the position of the best-fit sparrow in the population in dimension d. To generate a random number between -1 and 1.

[0196] The performance metrics (such as accuracy, loss function, etc.) of each LSTM model are calculated using the validation set and used as fitness values.

[0197] The formula for calculating the loss function is as follows;

[0198] ;

[0199] in, For the sample size, For the true value, These are predicted values.

[0200] Please see Figure 2 The present invention also provides a system for optimizing the operation of a desulfurization system, the system being used to execute the above-described method for optimizing the operation of a desulfurization system, comprising:

[0201] The data acquisition model is used to obtain historical data on current, vibration, temperature and noise of slurry circulation pump and oxygen pump, pipeline pressure, slurry flow rate and slurry level difference in slurry tank when the equipment is started and stopped, and operating parameter data of desulfurization system.

[0202] The equipment analysis module is used to calculate the blockage coefficient and equipment failure rate of the spraying system based on the pipeline pressure, slurry flow rate and slurry level difference in the slurry tank when the equipment starts and stops, under the same power of the slurry circulation pump. It also calculates the cost ratio of each set of data by the cumulative amount of sulfur dioxide and the operating cost.

[0203] The data filtering module is used to filter operating parameters based on cost ratio and equipment failure rate, mark the priority of pump history on the filtered data to form a sample dataset, and then preprocess the sample dataset to form a training dataset.

[0204] The optimization system module constructs an optimization model for the desulfurization system operation through a long short-term memory neural network. The detection parameters in the training dataset are used as the input to the optimization model, and the execution parameters in the training dataset are used as the output to train the model. Then, by collecting real-time detection parameters as the input to the model, the optimization model outputs the optimization result.

[0205] The system parameter optimization module is used to randomly generate a group of sparrow individuals using a sparrow search algorithm. Each individual represents a set of hyperparameter configurations for running the optimization model. By adjusting the position of the individuals, a better hyperparameter space is explored. The optimization model is trained using the hyperparameter configuration of each individual and optimized using a loss function as a fitness evaluation index.

[0206] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0207] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0208] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0209] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for optimizing the operation of a desulfurization system, characterized in that, The specific steps include: Step 1: Obtain historical data on current, vibration, temperature, and noise of the pumps in the desulfurization system to be optimized. The pumps include slurry circulation pumps and oxygen pumps. By calculating the stability and performance loss of the pumps, obtain the pump health coefficient. Determine the health level of the pumps based on the health coefficient and arrange maintenance plans based on the overall health level. Step 2: Obtain the pipeline pressure, slurry flow rate, and slurry level difference in the slurry tank at the same power of the slurry circulation pump from historical data. Calculate the blockage coefficient of the spray system, which includes pipelines and nozzles. By setting the blockage threshold and the health status of the pump, calculate and determine the equipment failure rate of the desulfurization system. Step 3: Obtain the operating parameter data of the desulfurization system from the historical data, calculate the cost ratio of each data set by the cumulative sulfur dioxide amount and operating cost, and filter the operating parameters according to the equipment failure rate. Mark the priority of the pump history on the filtered data to form a sample dataset. Then, preprocess the sample dataset to form a training dataset. Step 4: Construct an optimization model for the desulfurization system operation using a long short-term memory neural network. Use the detection parameters in the training dataset as the input to the optimization model and the execution parameters in the training dataset as the output to train the model. Then, collect real-time detection parameters as the input to the model and output the optimized results through the optimization model. Step 5: Randomly generate a group of sparrow individuals using the sparrow search algorithm. Each individual represents a set of hyperparameter configurations for running the optimization model. Explore a better hyperparameter space by adjusting the position of the individuals. Train and run the optimization model using the hyperparameter configuration of each individual. Optimize the model by using the loss function as a fitness evaluation index. The operating parameters include detection parameters and execution parameters. The detection parameters include flue gas intake volume, flue gas temperature, sulfur dioxide concentration, slurry tank level, limestone concentration, pH value, gypsum concentration, number of circulating pumps, number of oxygen pumps, real-time power consumption unit price, real-time limestone cost, and cumulative desulfurization time of the system. The execution parameters include the start / stop status of the circulating pumps and the start / stop status of the oxygen pumps. The formula for calculating the cost ratio of each set of data based on the cumulative amount of sulfur dioxide and operating costs is as follows: ； in, For cost ratio, The unit price is the real-time power consumption. For the real-time cost of limestone, This refers to the flue gas intake volume. This refers to the concentration of sulfur dioxide. The specific method for filtering operating parameters and marking the priority of pump history in the filtered data to form a sample dataset is as follows: Historical data is divided into days, with each group consisting of ten days' worth of data. The data is then sorted based on cost comparison, from largest to smallest, and also based on equipment failure rate, from largest to smallest. The top two-tenths of the data in each group are then removed. The filtered data are then timestamped, and the circulating pumps and oxygen pumps in the data are numbered and prioritized.

2. The method for optimizing the operation of a desulfurization system according to claim 1, characterized in that: The specific steps for obtaining the pump's health coefficient by calculating the pump's stability and performance loss are as follows: Historical data on current, vibration, temperature, and noise were analyzed according to specific time periods. Statistical analysis was conducted on historical data for each day at different time intervals. Once the data is collected, the fluctuations in the pump's current, vibration, temperature, and noise are calculated. The pump's stability is then calculated based on the weights of these fluctuations. The formulas for the fluctuation of the pump's current, vibration, temperature, and noise are as follows; ； ； ； ； in, Due to the fluctuation of current, For the wave nature of vibration, For temperature fluctuations, For the fluctuation of noise, In order to be in Current at any moment In order to be in Vibration at any moment In order to be in Temperature at any moment In order to be in Momentary noise, The time of data collection Time period Number of data collection times within; The formula for calculating the stability of a pump is: ； in, For the stability of the pump, The weights are assigned to the fluctuations of current, vibration, temperature, and noise, respectively.

3. The method for optimizing the operation of a desulfurization system according to claim 2, characterized in that: The formula for calculating the degree of performance loss of the pump is as follows; ； in, The degree of performance loss of the pump, Time period The stability of the pump, Time period The stability of the pump To ensure the stability of the pump during the initial operation of the desulfurization system.

4. The method for optimizing the operation of a desulfurization system according to claim 3, characterized in that: The formula for calculating the health coefficient of the pump is: ； in, For the pump's health coefficient, The degree of performance loss of the pump, Time period The stability of the pump, Time period The stability of the pump, The degree of performance loss of the pump; The specific method for judging the pump's health level based on the health coefficient and arranging a maintenance plan based on the overall health level is as follows: when When the pump is in good condition, its priority is 1, and no maintenance is required. when At this time, the pump's status is judged as normal, with a priority of 2, and no maintenance is required; when If the pump is deemed unhealthy at this time, with a priority of 3, it requires maintenance. in, These are the criteria for judging health and unhealth. .

5. The method for optimizing the operation of a desulfurization system according to claim 4, characterized in that: The steps for calculating the clogging coefficient of the spray system are as follows; The formula for calculating the pipeline blockage coefficient is as follows: ； in, The blockage coefficient of the pipeline. This represents the current slurry flow rate in the pipeline. This represents the slurry flow rate at the moment the pipeline maintenance was completed. This represents the current pipeline pressure. The pipeline pressure at the moment the maintenance was completed; The formula for calculating the clogging coefficient of the nozzle is: ； in, The nozzle clogging coefficient is... The difference in slurry tank level before and after the slurry pump circulation pump starts, assuming the nozzle is not clogged. This is the time it takes for the liquid level difference backflow to disappear when the nozzle is not clogged. This represents the difference in slurry tank level before and after the slurry pump circulation pump starts. This represents the time it takes for the backflow of the current liquid level difference to disappear. The formula for calculating the clogging coefficient of the spray system is as follows: ； in, The clogging coefficient of the spray system; when In this case, determine if the sprinkler system is clogged; The formula for calculating the equipment failure rate of the desulfurization system is as follows: ； in, The failure rate of the desulfurization system, The number of sprinkler systems identified as clogged based on the clogging coefficient. This represents the total number of sprinkler systems. The number of pumps deemed unhealthy based on a health coefficient. This represents the total number of pumps.

6. The method for optimizing the operation of a desulfurization system according to claim 1, characterized in that: The desulfurization system operation optimization model includes an input gate, a forget gate, and an output gate; The equation for the forgetting gate is: ； in, The current open / closed state of the forget gate. This represents the sigmoid function. For the weight of the forget gate, Forget gate bias parameters, This is the hidden state from the previous moment. Input switch for the current data; Determine how many cell states from the previous time step need to be retained in the current time step; The equation for the input gate is: ； in, To input the current open / closed state of the door. This represents the sigmoid function. For the input gate weights, For the input gate bias parameters, This is the output hidden state from the previous time step. This refers to the current detection parameter data; ； in, Candidate cell state To represent the hyperbolic tangent function, For candidate cell weights, These are the candidate cell state bias parameters. This is the hidden state from the previous moment. Input switch for the current data; Determine how much of the network's input data at the current moment needs to be saved to the cell state; The updated equation is: ； in, Current cell state This represents the cell state at the previous moment; The equation for the output gate is: ； ； in, To output the current open / closed state of the gate. This represents the sigmoid function. For the output gate weights, These are the output gate bias parameters. This is the hidden state from the previous moment. The current data input switch, This contains the execution parameter data for the current moment. Control how much of the current cell state needs to be output to the current output value; The mathematical expressions for the sigmoid function and the hyperbolic tangent function are: ； ； in, For the sigmoid function, For the input of the sigmoid function, It is the hyperbolic tangent function. For function input.

7. The method for optimizing the operation of a desulfurization system according to claim 1, characterized in that: The process of the sparrow search algorithm includes discoverer and follower; The formula for updating the discoverer's location is: ； in, For the updated number The middle generation Only sparrows in the first The position of the sparrow The types and dimensions of weights in the optimization model during hyperparameter configuration. The weight dimensions in the hyperparameter configuration are used to optimize the model. For random numbers that follow a normal distribution, A uniformly random number in [0,1]. It is the warning threshold, and its value ranges from [0.5, 1]. Follower position update formula; ； in, Let t be the position of the best-fit sparrow in the population in dimension d. To generate a random number from -1 to 1; The formula for calculating the loss function is as follows; ； in, For the sample size, For the true value, These are predicted values.

8. A system based on the operation optimization of a desulfurization system, characterized in that: The optimized system is used to execute the operation optimization method based on the desulfurization system according to any one of claims 1-7, including: The data acquisition model is used to obtain historical data on current, vibration, temperature and noise of slurry circulation pump and oxygen pump, pipeline pressure, slurry flow rate and slurry level difference in slurry tank when the equipment is started and stopped, and operating parameter data of desulfurization system. The equipment analysis module is used to calculate the blockage coefficient and equipment failure rate of the spraying system based on the pipeline pressure, slurry flow rate and the liquid level difference of the slurry in the slurry tank when the equipment starts and stops, under the same power of the slurry circulation pump. The cost ratio of each set of data is calculated by the cumulative amount of sulfur dioxide and the operating cost. The data filtering module filters operating parameters based on cost ratio and equipment failure rate, marks the filtered data according to the priority of pump history, forming a sample dataset, and then preprocesses the sample dataset to form a training dataset. The optimization system module constructs an optimization model for the desulfurization system operation through a long short-term memory neural network. It uses the detection parameters in the training dataset as the input to the optimization model and the execution parameters in the training dataset as the output to train the model. Then, it collects real-time detection parameters as the input to the model and outputs the optimization results through the optimization model. The system parameter optimization module is used to randomly generate a group of sparrow individuals using a sparrow search algorithm. Each individual represents a set of hyperparameter configurations for running the optimization model. By adjusting the position of the individuals, a better hyperparameter space is explored. The optimization model is trained using the hyperparameter configuration of each individual and optimized using a loss function as a fitness evaluation index.

Images