Combined optimization method and system for double-tower desulfurization slurry circulating pump based on artificial intelligence

Abstract

The invention relates to the technical field of desulfurization circulating pump frequency conversion and slurry supply cooperative control, in particular to a double-tower desulfurization slurry circulating pump combination optimization method and system based on artificial intelligence, and the method comprises the following steps: predicting boiler load values at multiple time points in the future, and defining a discrete operation situation in combination with the start-stop state and the operation frequency of a circulating pump group; and establishing a slurry amount output value corresponding to each situation, matching the slurry amount required by the boiler load at each time point in the future, and generating a future working condition demand and operation situation mapping set. According to the method, by predicting boiler loads at multiple time points in the future and combining the start-stop state and the operation frequency of the circulating pump group, discrete operation situations are dynamically defined, and the situations are matched with load requirements; and the detailed evaluation of the comprehensive cost of the migration path between the operation situations is realized by calculating the energy consumption transfer value generated by the change of the operation frequency and the start-stop wear cost of the equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of frequency conversion and slurry supply coordinated control of a desulfurization circulation pump, and in particular to a method and system for optimizing the combination of a double-tower desulfurization slurry circulation pump based on artificial intelligence. Background Art

[0002] The technical field of desulfurization circulation pump frequency conversion and slurry supply coordinated control mainly involves dynamically adjusting the operating frequency of the circulation pump and combining start-stop strategies to achieve the goals of precise control of the slurry supply in the desulfurization system and efficient energy utilization.

[0003] In practical applications, existing technologies rely on variable frequency regulation and combined start-stop strategies. This lacks the accuracy to accurately predict boiler load changes and slurry demand, and the ability to match boiler load demand and equipment frequency with start-stops at multiple future time points. Furthermore, they lack sufficient quantitative means for detailed analysis of equipment start-stop losses and energy consumption. This leads to frequent switching of operating states and increased equipment start-stop cycles, which in turn increases equipment wear and maintenance costs. Therefore, improvements are needed. Summary of the Invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and propose an artificial intelligence-based double-tower desulfurization slurry circulation pump combination optimization method and system.

[0005] In order to achieve the above object, the present invention adopts the following technical solution, which is an artificial intelligence-based dual-tower desulfurization slurry circulation pump combination optimization method, comprising the following steps:

[0006] Predict the boiler load values ​​at multiple future time points, define discrete operating states based on the start / stop status and operating frequency of the circulating pump group, establish the slurry volume output value corresponding to each state, match the slurry volume required for the boiler load at each future time point, and generate a mapping set of future operating conditions and operating states;

[0007] Based on the mapping set of future operating condition requirements and operating status, all operating statuses are extracted, and the energy consumption transfer value caused by the change in operating frequency between any two statuses is calculated. An energy consumption transfer cost table is established, and based on the energy consumption transfer cost table, the quantified wear cost of the start-stop equipment is superimposed to generate a comprehensive cost of the operating status migration path;

[0008] Based on the future working condition demand and operating situation mapping set, all operating situations that meet the slurry volume demand at the first time point are screened to obtain a feasible situation set for the first node; based on the feasible situation set for the first node, operating situations that meet the constraints are matched for subsequent time points in sequence and connections are established to establish a feasible migration network between multiple time nodes;

[0009] Based on the feasible migration network between multiple time nodes, starting from the current operating situation and tracing to the end time node, all optional complete path sequences are generated to obtain a fully quantified path candidate list. Based on the fully quantified path candidate list, the comprehensive cost of the operating situation migration path is called to calculate the total cost of each path and sort them, and the optimal scheduling instruction path for the circulating pump combination is established.

[0010] Preferably, the steps for obtaining the mapping set of future operating condition requirements and operating status are:

[0011] Based on the boiler operation records, the boiler load value, sulfur dioxide concentration at the absorption tower inlet, slurry circulation volume, liquid-gas ratio, pH value, oxygen concentration and circulation pump operation frequency at each time point in the boiler operation records are extracted to obtain the boiler load prediction input set;

[0012] Calculating boiler load forecast values ​​at each future time point based on the boiler load forecast input set;

[0013] Based on the boiler load prediction value, the start and stop status and operating frequency of the circulating pump are called to define a combined situation set, the predicted load is matched with the slurry output corresponding to each situation, and a mapping set of future working condition requirements and operating situations is generated.

[0014] Preferably, the steps for obtaining the energy consumption transfer cost table are:

[0015] Based on the future working condition demand and operating status mapping set, the operating frequency values ​​of all circulating pumps in each operating status are sequentially extracted, the frequencies of the corresponding circulating pumps under any two different operating statuses are paired, and the target operating duration of the circulating pump after the status transition is simultaneously extracted to generate an operating frequency transition and duration pairing sequence;

[0016] Calculating a total energy consumption transfer value based on the operating frequency transfer and duration pairing sequence;

[0017] Based on the total energy consumption transfer value, the energy consumption changes between each pair of operating states are registered item by item, a matrix structure is constructed according to the operating state number, the energy consumption differences of the transfer paths between all states are output in a row and column manner, and an energy consumption transfer cost table is generated.

[0018] Preferably, the steps for obtaining the comprehensive cost of the operation status transition path are:

[0019] Based on the operational status start and stop records, the operating status identifiers of each device before and after each group of operational status transitions are compared to identify devices that have changed from running to stopped or from stopped to running. Each device with a status change is recorded as a start and stop action, and the operating time of the most recent operating cycle is extracted to generate a list of started and stopped devices and their running time before the start and stop.

[0020] Calculate the cost of the operation status migration path based on the list of start-stop devices and their running time before start-stop and the energy consumption transfer cost table;

[0021] Based on the migration path cost, the migration path costs between all operating situation pairs are sequentially filled into the cost table structure matrix, and classified and arranged according to the starting situation and target situation indexes to generate the comprehensive cost of the operating situation migration path.

[0022] Preferably, the steps for obtaining the feasible situation set of the first node are:

[0023] Based on the future working condition demand and operating status mapping set, searching the operating frequency and start / stop status of all circulation pumps in each operating status in the future working condition demand and operating status mapping set one by one, calculating the slurry output flow rate corresponding to each circulation pump one by one, and summing the slurry output flow rates of all circulation pumps to form a total slurry output flow rate for each operating status;

[0024] Based on the total slurry output flow rate of each operating situation, the slurry demand value required for the boiler load corresponding to the first time point in the future operating condition demand and the operating situation mapping set is extracted. The total slurry output flow rate of each operating situation is compared with the demand value one by one. All operating situations with a total slurry output flow rate equal to or higher than the demand value are selected to form a preliminary operating situation set that meets the boiler load slurry demand at the first time point;

[0025] Based on the preliminary operating status set that meets the boiler load slurry demand at the first time point, check item by item whether the start / stop status and frequency of each circulating pump in the operating status are within the safe operating range allowed by the equipment, eliminate the operating status with start / stop status or frequency exceeding the safe operating range, and form a feasible status set for the first node.

[0026] Preferably, the steps of obtaining the feasible migration network between multiple time nodes are:

[0027] Based on the feasible situation set of the first node, the operating frequency, start / stop status, and slurry output flow rate of all circulating pumps corresponding to each operating situation in the set are extracted one by one; according to the slurry volume demand value corresponding to the boiler load demand at the subsequent second time point, all estimated operating situations that can meet the slurry volume demand at the second time point are matched to each operating situation in turn, thereby forming a feasible situation connection set from the first node to the second time point;

[0028] Based on the feasible situation connection set from the first node to the second time point, all circulating pump operating frequencies, start / stop states, and slurry output flows of all operating situations at the second time point in the connection set are sequentially retrieved and extracted; based on the slurry volume demand value corresponding to the boiler load demand at the third time point, each situation is judged item by item whether it can meet the slurry demand at the third time point under the constraints of the circulating pump frequency adjustment amplitude and the safe frequency interval; the connection situations that meet the conditions are selected to form a feasible situation connection set from the second time point to the third time point;

[0029] Based on the set of feasible situation connections from the second time point to the third time point, it progresses to all subsequent remaining time points in sequence, performs operational situation connections and constraint verification, and connects the feasible situation connections of all time points in series to form a feasible migration network between multiple time nodes.

[0030] Preferably, the steps for obtaining the full quantization path candidate list are:

[0031] Based on the feasible migration network between multiple time nodes, the current operating status in the current time node is selected as the initial departure node, and all operating status nodes connected to the next time node under the initial departure node are extracted one by one. The operating frequency, start and stop status, and slurry output flow rate of all circulating pumps contained in each status node are recorded in turn, and the connection relationship between each node is marked to form a status path set from the initial departure node to the second time node;

[0032] Based on the situation path set from the initial departure node to the second time node, all the circulation pump operating frequencies, start / stop states, and slurry output flows corresponding to each operating situation node within the second time node are retrieved in sequence, all feasible connection operating situations are matched to the third time node, and the connection relationship of each pair of nodes is recorded, and the connection is progressed one by one to the end time node to form a node connection chain set from the initial departure node to the end time node;

[0033] Based on the node connection chain set from the initial starting node to the end time node, starting from the initial starting node, all node connection chains are traversed one by one, and a continuous and complete path sequence of multiple time nodes is formed in series. The connection chains that are broken or cannot continue to the end time node are eliminated to form a fully quantified path alternative list.

[0034] Preferably, the steps for obtaining the optimal scheduling instruction path of the circulating pump combination are:

[0035] Based on the fully quantified path candidate list, each complete path sequence in the list is retrieved one by one, and the comprehensive cost of the operation situation transition path corresponding to each pair of adjacent operation situation nodes between the starting node and the end node of the path is gradually extracted. The comprehensive cost of the operation situation transition path between each pair of nodes is recorded one by one, and the comprehensive cost of the operation situation transition path of all adjacent node pairs is accumulated to form a cumulative comprehensive cost of the operation situation transition path corresponding to each path;

[0036] Based on the cumulative operation situation transition path comprehensive cost corresponding to each path, the cumulative operation situation transition path comprehensive cost of all complete path sequences is sorted from low to high to form a ranked list of cumulative path comprehensive costs;

[0037] Based on the ranked list of cumulative path comprehensive costs, a complete path sequence with the lowest comprehensive cost of the operation status migration path in the ranked list is selected, and the optimal scheduling instruction path for the circulation pump combination is formed according to the circulation pump operation frequency and start / stop status recorded at each node in the path sequence.

[0038] The present invention also provides a double-tower desulfurization slurry circulation pump combination optimization system, comprising:

[0039] Prediction module: Predicts boiler load values ​​at multiple future time points, defines discrete operating states based on the start / stop status and operating frequency of the circulating pump group, establishes slurry output values ​​corresponding to each state, matches the slurry volume required for boiler load at each future time point, and generates a mapping set between future operating condition requirements and operating states;

[0040] Cost modeling module: Based on the future operating condition demand and operating status mapping set, all operating statuses are extracted, the energy consumption transfer value caused by the change in operating frequency between any two statuses is calculated, and an energy consumption transfer cost table is established. Based on the energy consumption transfer cost table, the quantified start-stop equipment wear cost is superimposed to generate the comprehensive cost of the operating status migration path;

[0041] Network construction module: Based on the future working condition demand and operation status mapping set, all operation statuses that meet the slurry volume demand at the first time point are screened, and a feasible status set of the first node is obtained. Based on the feasible status set of the first node, operation statuses that meet the constraints are matched for subsequent time points in sequence and connections are established to establish a feasible migration network between multiple time nodes;

[0042] Path optimization module: Based on the feasible migration network between the multiple time nodes, starting from the current operation status, tracing to the end time node, generating all optional complete path sequences, and obtaining a fully quantified path candidate list. Based on the fully quantified path candidate list, calling the comprehensive cost of the operation status migration path to calculate the total cost of each path and sort them, and establish the optimal scheduling instruction path for the circulation pump combination.

[0043] Compared with the prior art, the advantages and positive effects of the present invention are:

[0044] The present invention predicts the boiler load at multiple time points in the future and combines the start-stop status and operating frequency of the circulating pump group to dynamically define discrete operating states and match each state with the load demand; then, by calculating the energy consumption transfer value generated by the change in operating frequency and the start-stop wear cost of the equipment, a detailed evaluation of the comprehensive cost of the migration path between operating states is achieved; then, the correlation between the slurry demand at each time point and the operating state combination is screened, a feasible migration network is formed and gradually optimized, thereby improving the operating accuracy, economy and stability of the desulfurization circulating pump; achieving real-time matching of the operating state with the boiler load demand, reducing the operating energy consumption and the maintenance cost caused by the start-stop wear of the equipment, improving the stability of the equipment operation, and avoiding the problem of accelerated equipment aging caused by frequent switching of operating states. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 Schematic diagram of the steps of the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0047] See also Figure 1 The present invention provides a technical solution, an artificial intelligence-based double-tower desulfurization slurry circulation pump combination optimization method, comprising the following steps:

[0048] Predict the boiler load values ​​at multiple future time points, define discrete operating states based on the start / stop status and operating frequency of the circulating pump group, establish the slurry volume output value corresponding to each state, match the slurry volume required for the boiler load at each future time point, and generate a mapping set of future operating conditions and operating states;

[0049] Based on the mapping set of future working condition requirements and operating status, all operating statuses are extracted, and the energy consumption transfer value caused by the change in operating frequency between any two statuses is calculated. An energy consumption transfer cost table is established. Based on the energy consumption transfer cost table, the quantified wear cost of the start-stop equipment is superimposed to generate the comprehensive cost of the operating status migration path;

[0050] Based on the mapping set of future working condition requirements and operating status, all operating statuses that meet the slurry volume requirements at the first time point are screened to obtain the feasible status set of the first node. Based on the feasible status set of the first node, the operating statuses that meet the constraints are matched for each subsequent time point and connected to establish a feasible migration network between multiple time nodes.

[0051] Based on the feasible migration network between multiple time nodes, starting from the current operating status and tracing to the end time node, all optional complete path sequences are generated to obtain a fully quantified path alternative list. Based on the fully quantified path alternative list, the comprehensive cost of the operating status migration path is called to calculate the total cost of each path and sort them, and the optimal scheduling instruction path for the circulating pump combination is established.

[0052] The steps to obtain the future working condition requirements and operating status mapping set are as follows:

[0053] Based on the boiler operation records, the boiler load value, sulfur dioxide concentration at the absorption tower inlet, slurry circulation volume, liquid-gas ratio, pH value, oxygen concentration and circulation pump operation frequency at each time point in the boiler operation records are extracted to obtain the boiler load prediction input set;

[0054] Based on the boiler load forecast input set, the boiler load forecast value at each future time point is calculated using the following formula:

[0055]

[0056] Among them, F t is the boiler load forecast value at the tth time point, h is the total number of data points included in the current forecast window, C c is the SO2 concentration at the absorption tower inlet at the cth time point, Q c is the slurry circulation volume at the cth time point, G c is the liquid-gas ratio at the cth time point, p c is the pH value at the cth time point, Freq c is the operating frequency of the circulating pump at the cth time point, O2 c is the oxygen concentration at the cth time point, k0, k1, k2, k3, k4, k5, k6 are fixed regression coefficients determined based on sample training;

[0057] Based on the boiler load forecast value, the start and stop status and operating frequency of the circulating pump are called to define a combined situation set. The predicted load is matched with the slurry output corresponding to each situation to generate a mapping set of future working condition requirements and operating situations.

[0058] Specifically, based on the historical operating data records of the boiler, which are usually stored in the distributed control system (DCS) or production information system (SIS) database of the power plant, the data is first aggregated and aligned at preset time intervals (for example, every minute or every five minutes) to ensure the consistency of the timestamps of various parameters. Then, for each time point, the actual load value of the boiler is accurately extracted from the record. This value is usually measured in megawatts (MW) or tons / hour (t / h evaporation), which directly reflects the output status of the boiler. At the same time, the real-time concentration value of sulfur dioxide (SO2) in the flue gas at the absorption tower inlet is extracted. This data comes from the continuous flue gas monitoring system (CEMS) and is usually measured in milligrams / normal cubic meter (mg / Nm 3 ), accurately record the initial amount of SO2 entering the desulfurization system, and then extract the total slurry circulation volume delivered by the slurry circulation pump, and obtain it through the electromagnetic flowmeter or ultrasonic flowmeter installed on the outlet pipe of the circulation pump, in cubic meters / hour (m 3 / h). The key to this parameter is to maintain sufficient liquid-solid contact. At the same time, the liquid-gas ratio (L / G) at that time is calculated or extracted. If it is directly extracted, it comes from the calculation module preset by the system. If calculation is required, it is calculated by the formula L / G = slurry circulation volume / flue gas volume. The flue gas volume is estimated based on parameters such as boiler load, fuel characteristics and excess air coefficient, or directly measured by the flue gas flow meter. The unit is liter / cubic meter (L / m 3 ) or dimensionless, then obtain the slurry pH value, which is monitored in real time by an online pH analyzer inserted into the slurry tank or pipeline, reflecting the acidity and alkalinity of the slurry, which is crucial for limestone dissolution and calcium sulfite oxidation. Then extract the oxygen concentration in the flue gas or slurry, usually referring to the oxygen content in the flue gas, measured by the oxygen analyzer in the CEMS, and the unit is percentage (%). Finally, record the operating frequency of each slurry circulation pump, obtained from the feedback signal of the variable frequency drive (VFD), in Hertz (Hz). This frequency directly affects the pump speed and slurry output. The boiler load values, SO2 concentration at the absorption tower inlet, slurry circulation volume, liquid-gas ratio, pH value, oxygen concentration and operating frequency of each circulation pump collected or calculated at the same time point are integrated to form a structured data set, in which each record represents a complete operating condition snapshot at a specific time point. This data set is the boiler load prediction input set.

[0059] formula: (k0+k1·C c +k2·Q c +k3·G c +k4·p c +k5·Freq c +k6·O2 c), the formula is beneficial in that it uses a multiple linear regression model combined with a sliding average method to predict future boiler loads. Its benefits are reflected in the following aspects: First, by incorporating multiple operating parameters (C c ,Q c ,G c ,p c ,Freq c ,O2 c ), the model can more comprehensively capture the complex factors affecting load changes, improving the accuracy and robustness of the prediction; secondly, the sliding average Processing can effectively smooth short-term data fluctuations and noise, making the prediction results more stable and avoiding prediction deviations caused by instantaneous disturbances; secondly, the structure of the linear regression model is relatively simple, with high computational efficiency, easy to understand and deploy, and convenient for engineering and technical personnel to quickly master and apply it; finally, the regression coefficients (k0 to k6) are determined through historical data sample training, which can reflect the operating characteristics of a specific unit and have a certain degree of adaptability.

[0060] Parameter description:

[0061] h represents the total number of historical data points included in the current forecast window. This is a parameter used for sliding average calculations and determines the length of historical data used for model predictions. For example, if the data collection frequency is once per minute, setting h = 10 means using the past 10 minutes of data for the average calculation. The selection of this parameter requires analyzing historical data to balance the smoothness of the forecast and the speed of response to load changes. The method for obtaining this parameter is to backtest different values ​​of h (for example, 5, 10, 15, and 20 data points) on the historical validation set, evaluate the forecast error (such as the root mean square error (RMSE), and select the h value that minimizes the RMSE. For example, a test of a power plant's historical data for one year at minute intervals found that when h = 12 (i.e., a 12-minute window), the combined performance of forecast accuracy and stability was the best, so h = 12 was set.

[0062] C c Indicates the SO2 concentration at the absorber inlet at the cth time point in the past. This parameter reflects the initial pollutant load entering the desulfurization system. This data is collected in real time by the continuous flue gas monitoring system (CEMS) installed on the flue duct at the absorber inlet. For example, at a certain historical data point c, the SO2 concentration monitored and recorded by the CEMS system is 1500 mg / Nm 3 .

[0063] Q cIndicates the slurry circulation volume at the cth time point in the past. This parameter represents the slurry flow rate used by the desulfurization system to absorb SO2 at that time. This data is obtained through real-time monitoring by an electromagnetic flowmeter or ultrasonic flowmeter installed on the outlet main pipe or each branch pipe of the slurry circulation pump. For example, at a certain historical data point c, the flowmeter reading is 3000m 3 / h.

[0064] G c It represents the liquid-gas ratio at the cth time point in the past. It reflects the amount of slurry involved in the reaction per unit flue gas volume and is a key operating parameter affecting desulfurization efficiency. This parameter is determined by the DCS system based on the real-time slurry circulation volume (Q c ) and real-time flue gas volume (V gas,c ) is calculated, and the calculation formula is: G c =Q c / V gas,c The flue gas volume can be estimated by parameters such as boiler load and fuel composition, or directly measured by a flue gas flow meter. For example, if Q c =3000m 3 / h and the flue gas volume V gas,c =300000Nm 3 / h, then G c =(3000×1000) / 300000=10L / Nm 3 .

[0065] p c The pH value of the slurry at time point c in the past. This dimensionless parameter reflects the pH of the slurry. This data is monitored in real time by an online pH meter installed in the slurry tank or circulation line. The pH value has a significant impact on the dissolution rate of limestone and the progress of the desulfurization reaction. For example, at a historical data point c, the online pH meter reading was 5.8.

[0066] Freq c Indicates the average operating frequency of the main circulation pump at the cth time point in the past or the operating frequency of the circulation pump that contributes most to the total slurry volume. The unit is Hz (Hertz). This parameter reflects the working intensity of the circulation pump and is directly related to the output capacity of the slurry. This data is obtained from the control system or feedback signal of the variable frequency drive (VFD). If there are multiple pumps running, the average value of their frequencies can be taken. For example, if two circulation pumps of the same model run at 40Hz and 45Hz respectively, then Freq c

[0067] =(40+45) / 2=42.5Hz.

[0068] O2 cRepresents the oxygen concentration in the flue gas at time point c in the past. The unit is % (volume percentage). Flue gas oxygen concentration not only affects boiler combustion conditions but also influences the oxidation of desulfurization byproducts (such as gypsum). This data is monitored in real time by the oxygen analyzer in the CEMS system. For example, at a historical data point c, the oxygen analyzer reading was 6.5%.

[0069] k0, k1, k2, k3, k4, k5, k6 are fixed regression coefficients determined based on historical operation data sample training. These coefficients are the core parameters of the model, which quantify the degree and direction of the influence of each input variable on the boiler load. The acquisition method is: collect historical operation data for at least three consecutive months (including all the above input variables and the corresponding actual boiler load) to form a training data set. Use the multivariate linear regression analysis method in statistics (such as the least squares method) to fit the training data and solve the coefficient that minimizes the sum of squares of the residuals between the predicted value and the actual value. For example, through regression analysis of historical data, we get k0 = 20.5, k1 = 0.005, k2 = 0.015, k3 = -5.2, k4 = 8.1, k5 = 1.2, k6 = -3.3. The units of these coefficients must be consistent with the units of the corresponding variables and F t to ensure dimensional consistency. For example, if F t The unit is MW, C c The unit is mg / Nm 3 , then the unit of k1 is MW / (mg / Nm 3 ).

[0070] Calculation process:

[0071] The calculation process of this formula is divided into two steps: first, calculate the weighted sum of each historical data point c, and then average these h weighted sums.

[0072] Take the prediction of boiler load F1 at the first future time point t=1 as an example.

[0073] Set the prediction window h=2.

[0074] The regression coefficients used (trained using historical data) are:

[0075] k0=20.5(MW);

[0076] k1=0.005(MW / (mg / Nm 3 ));

[0077] k2=0.015(MW / (m 3 / h));

[0078] k3=-5.2(MW / (L / Nm 3 ));

[0079] k4=8.1(MW / pH);

[0080] k5=1.2(MW / Hz);

[0081] k6 = -3.3 (MW / %);

[0082] Parameter values ​​for historical data point 1 (c=1):

[0083] C1=1550mg / Nm 3 ;

[0084] Q1=3050m 3 / h;

[0085] G1=10.5L / Nm 3 ;

[0086] p1=5.7;

[0087] Freq1=42.0Hz;

[0088] O21=6.2%;

[0089] Calculate the load contribution term L1 of historical data point 1:

[0090] L1=k0+k1·C1+k2·Q1+k3·G1+k4·p1+k5·Freq1+k6·O21;

[0091]

[0092] L1=20.5+7.75+45.75-54.6+46.17+50.4-20.46;

[0093] L1=95.51MW;

[0094] Parameter values ​​for historical data point 2 (c=2):

[0095] C2=1500mg / Nm 3 ;

[0096] Q2=3000m 3 / h;

[0097] G2=10.0L / Nm 3 ;

[0098] p2=5.8;

[0099] Freq2=42.5Hz;

[0100] O22=6.5%;

[0101] Calculate the load contribution term L2 of historical data point 2:

[0102] L2=k0+k1·C2+k2·Q2+k3·G2+k4·p2+k5·Freq2+k6·O22;

[0103]

[0104] L2=20.5+7.5+45.0-52.0+46.98+51.0-21.45;

[0105] L2=97.53MW;

[0106] Finally, calculate the boiler load forecast value F at the future time point t t (Here is F1):

[0107]

[0108] F1=96.52MW;

[0109] The results indicate that, based on the operating parameters from the last two historical time points and the pre-set regression model, the predicted boiler load at the next time point is 96.52 MW. This predicted value will serve as the basis for determining the future slurry volume required for the desulfurization system. A higher predicted value indicates that more slurry volume will be needed to meet desulfurization requirements, while a lower predicted value indicates that more slurry volume will be needed to meet desulfurization requirements.

[0110] Based on the boiler load forecast values ​​at each future time point calculated in the previous steps, for example, F t =96.52MW. First, a pre-built combination state set of circulating pump start / stop states and operating frequency definitions needs to be called. This set is generated during system initialization or regular calibration based on the physical characteristics of all slurry circulating pumps in the dual-tower desulfurization system (for example, there are three circulating pumps A, B, and C), the allowable operating frequency range (for example, the operating frequency range of a single pump is 25Hz to 50Hz), the minimum start / stop time interval, and the interlocking logic between pumps (for example, at least one pump A or B must be running before pump C starts). By filtering out combinations that do not meet the operating conditions, each combination state uniquely describes the start / stop state of all circulating pumps (0 represents stop, 1 represents running) and their respective operating frequencies (discretized within the allowable range, for example, with a step size of 1Hz). The total slurry output flow rate that the system can stably provide under each state is pre-calculated or calibrated. For example, state S1 is described as "Pump A running @ 35Hz, Pump B running @ 40Hz, Pump C stopped", and its corresponding total slurry output is 2800m 3 / h, the corresponding relationship between these “status-slurry output” is stored as a lookup table or database, and then the predicted boiler load F t Through a preset conversion model or empirical curve (for example, the relationship curve of "boiler load-required slurry volume" fitted according to historical data, which also takes into account factors such as inlet SO2 concentration, for example, when the load is 80-100MW and the SO2 concentration is 1000-1500mg / Nm 3 When the required slurry volume is Q req =0.025×F t +500, unit is m 3 / h), converted to the theoretical slurry volume required by the desulfurization system under this load. For example, if F t =96.52MW, the required slurry volume Q is calculated according to the conversion model req =0.025×96.52+500=2.413+500=2913m 3 / h, then, the calculated required slurry amount Q req (2913m 3 / h) and the slurry output of each situation in the combined situation set are compared one by one, and all the slurry outputs that can provide no less than Q are screened. req and as close as possible (e.g., no more than Q req 110% is the preset upper limit coefficient, which aims to avoid oversupply. The coefficient is set according to historical operation data analysis and energy-saving targets. For example, by analyzing the energy consumption and desulfurization effects under different redundancies, it is found that 10% redundancy is the balance point between economy and effect). A candidate situation subset is formed. This matching process is performed for each future forecast time point, and finally a mapping set of future working condition requirements and operation situations is obtained.

[0111] The steps to obtain the energy consumption transfer cost table are:

[0112] Based on the mapping set of future working condition requirements and operating status, the operating frequency values ​​of all circulating pumps in each operating status are extracted in sequence. The frequencies of the corresponding circulating pumps under any two different operating statuses are paired, and the target operating duration of the circulating pump after the status transition is simultaneously extracted to generate an operating frequency transfer and duration pairing sequence.

[0113] According to the operating frequency transfer and duration pairing sequence, the total energy consumption transfer value is calculated using the following formula:

[0114]

[0115] Among them, E i,j is the total energy consumption transfer value corresponding to the migration from the i-th operating state to the j-th operating state, s is the total number of circulating pumps in the operating state, ZCu is the power proportional coefficient of the u-th circulating pump (unit: kW / Hz 3 ),φ i,u is the operating frequency of the u-th circulating pump in operating state i (in Hz), φ j,u is the operating frequency of the u-th circulating pump in operating state j (in Hz), λ u is the duration that the u-th circulating pump continues to work in operating state j (in hours);

[0116] Based on the total energy consumption transfer value, the energy consumption changes between each pair of operating states are registered item by item, and a matrix structure is constructed according to the operating state number. The energy consumption differences of the transfer paths between all states are output in a row and column manner to generate an energy consumption transfer cost table.

[0117] Specifically, based on the future operating condition demand and operating status mapping set generated in the previous step, the mapping set records in detail the boiler load corresponding to each predicted future time point, the calculated required slurry volume, and one or more candidate operating statuses that can meet the slurry demand and their specific configurations (including the start and stop status and operating frequency of each circulating pump). First, all non-repeated operating statuses are extracted from the mapping set, and an identifier is assigned to each unique status, such as S1, S2, etc. Then, for each unique operating status, such as status Sk, its definition is parsed to obtain the specific operating frequency value φ of each circulating pump (for example, s pumps numbered u=1, 2, ..., s) under the status. k,u If a pump is in a stopped state in state Sk, its operating frequency is recorded as 0 Hz. Then, the transition between any two different operating states is systematically considered, that is, from the starting state Si to the target state Sj (where i is not equal to j). For each pair of such state transitions (Si→Sj), the operating frequency φ of each circulating pump u in the starting state Si is extracted. i,u and the operating frequency φ of the corresponding circulating pump u in the target state Sj j,u At the same time, it is necessary to determine the target operating time λ of each circulating pump u running in Sj after the system switches to the target state Sj. u , which is usually determined by the system's optimization scheduling cycle or prediction time step. For example, if the system's scheduling instructions are updated every 15 minutes, then λ u It is usually set to 0.25 hours. For the pump that stops in the target state Sj, its corresponding λ u It can be understood as 0, or it can be understood as "running" at 0Hz for 0.25 hours, but due to the subsequent calculation of energy consumption transfer, if φ j,u = 0 and φ i,u ≠0, the frequency difference still exists, so λ uThe planned duration of the situation Sj should be unified, and the starting frequency φ of each circulating pump u in each pair (starting situation Si, target situation Sj) should be i,u , target frequency φ j,u and the planned duration λ of the target situation Sj Sj (i.e. the λ of all pumps working in this situation u ) are combined into a recording unit, and finally these recording units of all pumps under all possible situation pairs are collected to generate a sequence of running frequency transfer and duration pairings.

[0118] formula: The usefulness of the formula is that it quantifies the energy consumption costs associated with the frequency adjustment (including start and stop) of the circulating pump when switching from one operating state i to another operating state j. Based on the similarity theory of fluid machinery, that is, the power consumption of the pump is approximately proportional to the cube of its speed (frequency), The nonlinear characteristics of the effect of frequency change on power are captured. The use of absolute values ​​ensures that whether the frequency increases or decreases (or starts and stops), it is regarded as a "transfer behavior" that consumes energy or generates disturbances, and is thus included in the cost. In addition, the formula introduces the independent power proportional coefficient ZC for each pump u , so that the calculation can adapt to circulation pumps of different specifications and efficiencies, enhancing the accuracy and versatility of the model. At the same time, by multiplying the continuous working time of the pump under the target situation λ u , the impact of instantaneous power changes is accumulated as energy values ​​over a period of time, making it more practical and directly reflecting the energy consumption transfer cost within a specific scheduling period.

[0119] Parameter description:

[0120] s represents the total number of circulating pumps involved in the scheduling. This is a fixed integer value determined by the actual hardware configuration of the desulfurization system. For example, if a desulfurization system is configured with three adjustable-frequency slurry circulating pumps, then s = 3.

[0121] ZC u is the power proportional coefficient of the u-th circulating pump, in kW / Hz 3 (kW per hertz cubed). This coefficient characterizes the proportional relationship between the input power of the u-th pump and the cube of its operating frequency, reflecting the energy consumption characteristics of the pump. It can be obtained by consulting the equipment manual of the pump to obtain the power-frequency characteristic curve data, or by measuring and recording the actual input power P of the pump at multiple different stable operating frequency points (for example, at least 5 points covering its common operating range, such as 30Hz, 35Hz, 40Hz, 45Hz, 50Hz, etc.) during actual operation. k and the corresponding frequency f k, and then use the least squares method to find the data points (f k ,P k ) to fit and solve P = ZC u ·f 3 +P fixed ZC u Value (where P fixed is the fixed power loss, which can be ignored if it is small or determined by fitting. For example, the power of circulating pump No. 1 is 30kW at 30Hz, 48kW at 35Hz, 72kW at 40Hz, 103kW at 45Hz, and 145kW at 50Hz. fixed =0, calculate P / f at each point 3 :30 / 30 3 ≈0.001111, 48 / 35 3 ≈0.001119, 72 / 40 3 =0.001125,103 / 45 3 ≈0.001128, 145 / 50 3 =0.00116. Taking the average value or through regression analysis, we can get ZC1=0.001128kW / Hz 3 .

[0122] φ i,u Indicates the operating frequency of the u-th circulating pump in the initial operating state i, in Hertz (Hz). This value is directly extracted from the "operating frequency transfer and duration pairing sequence" generated in the previous step for a specific state i and pump u. If pump u is in the stopped state in state i, then φ i,u = 0 Hz. For example, in situation i, the operating frequency of circulation pump No. 1 is recorded as 38 Hz.

[0123] φ j,u Indicates the operating frequency of the u-th circulating pump in the target operating state j, in Hertz (Hz). Similarly, this value is directly extracted from the "Operating Frequency Transfer and Duration Pairing Sequence" for a specific state j and pump u. If pump u is about to stop running in state j, then φ j,u =0Hz. For example, in situation j, the operating frequency of circulation pump No. 1 is planned to be adjusted to 42Hz.

[0124] λ u Indicates the expected duration of operation of the u-th circulating pump in the target operating state j, in hours (h). This duration is usually consistent with the time step of the system optimization scheduling or the time granularity of future operating condition prediction. For example, if the system makes an optimization decision every 30 minutes, then λ uTypically, this parameter is set to 0.5 hours. This parameter is also obtained from the "Operation Frequency Shift and Duration Pairing Sequence." For all pumps participating in situation j, this value is generally the same, i.e., the planned duration of situation j. For example, situation j is planned to last 0.5 hours.

[0125] Calculation process:

[0126] Assume that the system has a total of s=2 circulation pumps.

[0127] Power proportional coefficient of pump 1: ZC1 = 0.001128kW / Hz 3 .

[0128] Power proportional coefficient of pump 2: ZC2 = 0.001250kW / Hz 3 .

[0129] Frequency configuration of the initial operating state i:

[0130] Pump 1 operating frequency φ i,1 =38Hz;

[0131] Pump 2 operating frequency φ i,2 =32Hz;

[0132] Frequency configuration of target operation status j:

[0133] Pump 1 operating frequency φ j,1 =42Hz;

[0134] Pump 2 operating frequency φ j,2 =0 Hz (i.e. pump 2 is stopped in state j);

[0135] The planned duration of the target operating situation j is λ = 0.5 h, so λ1 = 0.5 h and λ2 = 0.5 h.

[0136] Calculate the energy transfer value contribution E of pump 1 i,j,1 :

[0137]

[0138] E i,j,1 =0.001128·|42 3 -38 3 |·0.5;

[0139] E i,j,1 =0.001128·|74088-54872|·0.5;

[0140] E i,j,1 =0.001128·|19216|·0.5;

[0141] E i,j,1=0.001128·19216·0.5;

[0142] E i,j,1 =21.675648·0.5;

[0143] E i,j,1 =10.837824kWh;

[0144] Calculate the energy transfer value contribution E of pump 2 i,j,2 :

[0145]

[0146] E i,j,2 =0.001250·|0 3 -32 3 |·0.5;

[0147] E i,j,2 =0.001250·|0-32768|·0.5;

[0148] E i,j,2 =0.001250·|-32768|·0.5;

[0149] E i,j,2 =0.001250·32768·0.5;

[0150] E i,j,2 =40.96·0.5;

[0151] E i,j,2 =20.48kWh;

[0152] Calculate the total energy consumption transfer value E from state i to state j i,j :

[0153] E i,j =E i,j,1 +E i,j,2 ;

[0154] E i,j =10.837824+20.48;

[0155] E i,j =31.317824kWh;

[0156] The results show that when the circulation pump combination is adjusted from operating state i (pump 1 @ 38 Hz, pump 2 @ 32 Hz) to operating state j (pump 1 @ 42 Hz, pump 2 stopped), and state j is planned to last for 0.5 hours, the total energy consumption transfer value associated with this frequency adjustment behavior is 31.317824 kWh.

[0157] Based on each pair of running situations calculated in the previous step (for example, from situation S k To Situation S l ) between the total energy consumption transfer value E k,l First, a list of all unique operating states is compiled. These states are usually identified and numbered in the previous steps. Suppose there are N unique operating states, numbered S1, S2, ..., S N Then, initialize an N×N two-dimensional matrix, the row index and column index of the matrix correspond to the numbers of these operating states, and each element M in the matrix k,l Used to store the running status S k Migrate to operating state S l The total energy consumption transfer value E k,l , then, traverse all possible ordered situation pairs (S k ,S l ), where k and l range from 1 to N. For each situation pair, find the corresponding E from the calculated result set. k,l (If k≠l), or for the diagonal elements (i.e. k=l, indicating that it is maintained from one state to itself, and no migration occurs), its energy consumption transfer value E k,k By definition, it is usually set to 0 because there is no frequency change. k,l The value is filled in the position of row k and column l of the matrix. For example, if the E from situation S1 to S2 is calculated 1,2 =31.317824kWh, then fill in this value in the first row and second column of the matrix, and complete all N 2 The final fully filled N×N matrix is ​​the energy transfer cost table.

[0158] The steps to obtain the comprehensive cost of the running situation migration path are:

[0159] Based on the operational status start and stop records, the operating status identifiers of each device before and after each group of operational status transitions are compared to identify devices that have changed from running to stopped or from stopped to running. Each device with a status change is recorded as a start and stop action, and the operating time of the most recent operating cycle is extracted to generate a list of started and stopped devices and their running time before the start and stop.

[0160] Based on the list of start-stop equipment and its running time before start-stop and the energy consumption transfer cost table, the cost of the operation status migration path is calculated. The calculation formula is:

[0161]

[0162] Among them, R i,j is the migration path cost from the i-th operating state to the j-th operating state, Ei,j is the total energy consumption transfer value corresponding to the operating status, x is the total number of devices that start and stop during the migration process, ω d is the start-stop cost coefficient of the dth start-stop device (in kWh), τ d is the continuous running time of the dth device before this migration (in hours). Indicates the equivalent wear intensity of a single start and stop of the equipment after going through this operating cycle;

[0163] Based on the migration path cost, the migration path costs between all pairs of operating situations are sequentially filled into the cost table structure matrix, and classified and arranged according to the starting situation and target situation indexes to generate the comprehensive cost of the operating situation migration path.

[0164] Specifically, based on the start and stop records of the operating status, these records usually come from the system's detailed log of each status change, which contains the running (1) or stopping (0) status identification of each circulating pump under each status. First, for the migration of each group of operating status, for example, from the starting status S i Transition to target state S j , the system will compare S one by one i and S j The operating status of each circulating pump (equipment number is d=1,2,…,s) in S i The state in is 1 (running) and in S j The state in S is 0 (stop), or i 0 (stop) in S j If the value is 1 (running), it is determined that the device d has a start-stop action, and the device number is recorded. At the same time, for each device d that is identified to have a start-stop action, it is necessary to query and extract the duration τ of the most recent continuous uninterrupted operation of the device before the start-stop action occurs from its historical operation data (usually maintained by the DCS or device management system). d , the duration is in hours. If the device is started from a stopped state, its "running time before start and stop" can be understood as 0. The device numbers of all identified start and stop actions and their corresponding τ d The values ​​are organized into a list, each element of which is a two-tuple (device number d, τ d ) to form a list of start and stop devices and their running time before start and stop.

[0165] formula: The benefit of the formula is that it takes into account two different types of costs to evaluate the total cost of operational status migration, namely the energy consumption transfer cost E caused by frequency adjustment. i,jand the equivalent wear cost caused by the start-up and shutdown operations of the equipment. This comprehensive evaluation makes the decision more comprehensive, not only pursuing the lowest short-term energy consumption, but also taking into account the long-term operating reliability and maintenance costs of the equipment. The time τ that the equipment has been running continuously before starting and stopping is introduced d As an impact factor, through the logarithmic function To quantify the wear intensity, frequent start-stop or start-stop after long-term continuous operation will cause greater impact and loss to the equipment. The wear intensity increases with τ d The increase of τ increases, but the growth rate gradually slows down, which is in line with the general law of equipment aging and fatigue accumulation, and also avoids τ d When the wear cost is too large, it will increase linearly. By setting an independent start-stop cost coefficient ω for each device d Different weights can be assigned to different equipment based on factors such as importance, purchase cost, and difficulty of maintenance, making the calculation of wear cost more targeted and reasonable.

[0166] Parameter description:

[0167] E i,j is the total energy consumption transfer value corresponding to the migration from operating state i to operating state j, in kWh. This parameter is directly obtained from the "Energy Consumption Transfer Cost Table" calculated in the previous step, according to the index of the starting state i and the target state j. For example, in the above example, the calculated E i,j =31.317824kWh.

[0168] x represents the total number of devices (circulation pumps) that started and stopped during the transition from situation i to situation j. This value is calculated from the previous step, "Generate a list of started and stopped devices and their running time before starting and stopping." For example, if a situation transition causes one pump to start and one pump to stop, then x = 2.

[0169] ω d is the start-stop cost coefficient of the dth device that has started and stopped, in kWh. This coefficient represents the cost of a single "standard" start-stop operation of the device (i.e. The equivalent energy loss or maintenance cost corresponding to a baseline value (when the equipment is used as a baseline value) is calculated. This factor is determined based on factors such as the equipment's purchase cost, expected service life, mean time between failures (MTBF), the average cost of a single repair (including spare parts and labor), and the known impact of startups and shutdowns on key equipment components (such as motor windings, bearings, and seals). This factor is obtained by first collecting historical failure data and maintenance records for a specific model of circulating pump and analyzing the correlation between the number of startups and shutdowns and failure rates and repair costs. Second, reference is made to the manufacturer's guidance on startup and shutdown limits and their impact on lifespan. Third, expert experience can be used for evaluation. For example, for a critical circulating pump valued at 100,000 yuan and expected to have a maximum startup and shutdown frequency of 10,000, if the accumulated depreciation due to startups and shutdowns is calculated at 20% of the purchase cost, the base cost of a single startup and shutdown can be estimated as 100,000 yuan x 20% / 10,000 startups = 2 yuan per startup and shutdown. Then, this monetary cost is converted into equivalent energy using the local average industrial electricity price (e.g., 0.6 yuan / kWh): 2 yuan / 0.6 yuan / kWh ≈ 3.33 kWh. This value can be used as ω d The energy consumption of pump No. 1 (critical pump) is set to ω1 = 4.0 kWh, and that of pump No. 2 (general pump) is set to ω2 = 2.5 kWh.

[0170] τ d The dth device that started or stopped the previous migration (start or stop) had its last continuous operation duration (in hours) before the migration (start or stop) occurred. This data is extracted from the "List of Start and Stop Devices and Their Pre-Start and Stop Operation Duration Pairs" generated in the previous step. For example, if Pump 1 had been running continuously for 150.5 hours before being stopped, then τ1 = 150.5h.

[0171] Indicates that the device has experienced a time duration of τ d The equivalent wear intensity of a single start-stop operation after a running cycle. This is a dimensionless amplification factor that converts the basic start-stop cost coefficient ω d Adjust according to the recent "fatigue" of the equipment. The use of natural logarithm makes the wear intensity increase with τ d , but the growth rate decreases. For example, if τ d =150.5

[0172] h, then ln(1+150.5 2 )=ln(1+22650.25)=ln(22651.25)≈10.028.

[0173] Calculation process:

[0174] Consider migrating from operating state i to operating state j.

[0175] From the “Energy Consumption Transfer Cost Table”, we can find E i,j =31.317824kWh.

[0176] In this migration, for example, x=2 devices started and stopped:

[0177] Equipment 1 (Pump 1): A stop occurs. Its start-stop cost coefficient ω1 = 4.0 kWh. It had been running continuously for τ1 = 150.5 hours before this stop.

[0178] Device 2 (Pump 2): Starts. Its start-stop cost coefficient ω2 = 2.5 kWh. For a device that switches from shutdown to operation, its "last continuous operation duration" is 0. Therefore, τ2 = 0 h.

[0179] Calculate the start-stop wear cost M1 of device 1:

[0180]

[0181] M1=4.0·ln(1+150.5 2 );

[0182] M1=4.0·ln(1+22650.25);

[0183] M1=4.0·ln(22651.25);

[0184] M1=4.0·10.02806;

[0185] M1=40.11224kWh;

[0186] Calculate the start-stop wear cost M2 of device 2:

[0187]

[0188] M2=2.5·ln(1+0 2 );

[0189] M2=2.5·ln (1);

[0190] M2=2.5·0;

[0191] M2=0kWh;

[0192] Calculate the total start-stop wear cost ∑M d :

[0193]

[0194] Calculate the total operational situation migration path cost R i,j :

[0195]

[0196] R i,j =31.317824+40.11224;

[0197] R i,j =71.430064kWh;

[0198] The results show that the comprehensive cost of migrating from operating state i to operating state j is 71.430064kWh. This value combines the energy consumption impact of frequency adjustment (31.317824kWh) and the equivalent wear cost (40.11224kWh) caused by equipment start-up and shutdown (mainly the shutdown of pump 1 after a long period of operation). The higher this comprehensive cost value, the greater the overall "cost" of the state migration. When performing path optimization, you should try to avoid choosing a path with high R i,j The migration step of the value.

[0199] Based on all the running situation pairs calculated in the previous step (for example, from situation S k To Situation S l ) between the migration path cost R k,l First, it is necessary to confirm again all the unique operating states that have been identified and numbered in the previous steps, for example, there are still N, numbered S1, S2, ..., S N Then, create a new N×N two-dimensional matrix, which is used to store these comprehensive migration path costs. Its row index and column index also correspond to the numbers of the N operating situations. Each element C in the matrix k,l Store the running status S k Migrate to operating state S l The comprehensive cost R k,l , then, traverse all possible ordered situation pairs (S k ,S l ), k, l∈[1,N], for each situation pair, the result set calculated from the previous formula (which contains all R k,l value) to find the corresponding R k,l If k = l, that is, the situation does not change, then the migration path cost R k,k Usually set to 0, since there is neither energy transfer nor start-stop wear (or E k,k = 0 and the start-stop wear item is also 0), the R k,l The value is accurately filled into the kth row and lth column of the cost table structure matrix. For example, if the comprehensive migration path cost R from situation S1 to S2 is calculated 1,2 =71.430064kWh, then fill in this value in the first row and second column of the cost table, and complete all N items one by one.2 The assignment of values ​​to each element finally forms a fully filled N×N matrix, which is the comprehensive cost table of the running state migration path. It clearly classifies and arranges the comprehensive cost of all direct migration steps according to the index of the starting state (row) and the target state (column).

[0200] The steps to obtain the feasible situation set of the first node are:

[0201] Based on the future working condition demand and operating status mapping set, the operating frequency and start / stop status of all circulating pumps in each operating status in the future working condition demand and operating status mapping set are retrieved one by one, the slurry output flow rate corresponding to each circulating pump is calculated one by one, and the slurry output flow rates of all circulating pumps are summed to form the total slurry output flow rate for each operating status;

[0202] Based on the total slurry output flow rate of each operating situation, the slurry demand value required for the boiler load corresponding to the first time point in the future operating condition demand and the operating situation mapping set is extracted. The total slurry output flow rate of each operating situation is compared with the demand value one by one. All operating situations with a total slurry output flow rate equal to or higher than the demand value are selected to form a preliminary operating situation set that meets the boiler load slurry demand at the first time point;

[0203] Based on the preliminary operating status set that meets the boiler load slurry demand at the first time point, check item by item whether the start / stop status and frequency of each circulating pump in the operating status are within the safe operating range allowed by the equipment, eliminate the operating status with start / stop status or frequency exceeding the safe operating range, and form a feasible status set for the first node.

[0204] Specifically, based on the future operating condition demand and operation status mapping set obtained in the previous step, the mapping set contains the boiler load, required slurry volume, and candidate operation statuses that meet the conditions and their detailed configurations at each predicted time point. First, each unique operation status recorded in the mapping set is traversed. For each operation status, such as status S k , read the operating frequency φ of each circulating pump defined therein (for example, there are s pumps, numbered u=1,2,…,s) k,u and start / stop status (1 for running, 0 for stopping), then, according to the characteristic curve of each circulation pump u (this curve is usually provided by the pump manufacturer or obtained through on-site calibration, which describes the output flow of the pump at different operating frequencies, for example where a u ,b u ,c u is the specific coefficient of pump u, if pump u is in state S k If the pump is in the stopped state, the slurry output flow rate is 0), calculate the pump in the current state S k Slurry output flow Q under k,u, then, the situation S k The slurry output flow Q calculated by all the circulating pumps running in the k,u Accumulate and get the running status S k The total slurry output flow that the system can provide This calculation is performed for all unique operating scenarios in the mapping set, and each operating scenario is annotated with its total slurry output flow rate.

[0205] Based on the total slurry output flow Q calculated for each operating state in the previous step total,k First, locate the relevant records of the first predicted time point (e.g., t=1) from the future working condition demand and operation status mapping, and extract the theoretical slurry demand value corresponding to the predicted boiler load at this time point from the record, which is recorded as Q demand,t＝1 , such as Q demand,t＝1

[0206] =2913m 3 / h, then all the unique operating states in the system (for example, there are N states, S1, S2, ..., S N ) The total slurry output flow Q total,k (where k = 1, ..., N) and the slurry volume demand value Q at this first time point demand,t＝1 Compare them one by one, and the screening condition is the total slurry output flow Q of the operating status total,k Must be equal to or slightly higher than the demand value Q demand,t＝1 , the specific standard can be set as Q total,k ≥Q demand,t＝1 And Q total,k ≤α·Q demand,t＝1 , where α is a coefficient slightly larger than 1, such as 1.1, representing the maximum allowable redundancy. This coefficient is set based on actual operating experience and energy-saving requirements. For example, if historical data shows that 10% flow redundancy can effectively cope with short-term fluctuations and the increase in energy consumption is within an acceptable range, then α = 1.1, and all operating states that meet this condition (for example, if the total slurry output of state S5 is 2950m 3 / h, and 2950 ≥ 2913 and 2950 ≤ 1.1 × 2913 = 3204.3, then S5 is selected) and collected to form a preliminary operating situation set that meets the boiler load slurry demand at the first time point.

[0207] Based on the preliminary operation status set that meets the boiler load slurry demand at the first time point obtained by the previous step, it is necessary to perform a safety check on each operation status in this set. For any operation status in the preliminary set, such as status S p , check the start and stop status and operating frequency φ of each circulating pump u defined internallyp,u The verification of the start-stop status is mainly to ensure that the minimum start-stop time interval requirement is met. For example, if a pump has just stopped for less than 5 minutes (the 5 minutes is the minimum stop-restart interval recommended by the equipment manufacturer, or the protective threshold set based on historical operating experience to avoid motor overheating or mechanical shock), then the situation S of starting the pump immediately is included. p It will be considered unsafe. For the verification of the operating frequency, it is necessary to ensure that φ p,u Strictly within the safe operating frequency range specified in the pump equipment manual or on-site calibration. For example, the allowable operating frequency range of a pump is 25Hz to 50Hz. If the Situation S p If the frequency of the pump is set to 20Hz or 55Hz, the situation is unsafe. In addition, it is necessary to check whether there are any interlocking conditions that are not met. For example, the start-up of some pumps depends on other specific pumps being in operation, or the total number of running pumps does not exceed the system power supply or pipe network limit. All operating situations in the preliminary operating situation set where the start-stop status of any circulating pump (such as violating the minimum start-stop interval) or the operating frequency (such as exceeding the upper and lower safety limits) does not meet the preset safety specifications are eliminated from the set. After this round of safety filtering, the retained operating situations constitute the final feasible situation set for the first node (i.e., the first prediction time point).

[0208] The steps to obtain a feasible migration network between nodes at multiple times are:

[0209] Based on the feasible situation set of the first node, the operating frequency, start / stop status, and slurry output flow rate of all circulating pumps corresponding to each operating situation in the set are extracted one by one. According to the slurry volume demand value corresponding to the boiler load demand at the second time point, each operating situation is matched with all estimated operating situations that can meet the slurry volume demand at the second time point, forming a feasible situation connection set from the first node to the second time point;

[0210] Based on the feasible situation connection set from the first node to the second time point, all the circulation pump operating frequencies, start / stop states, and slurry output flows of all the operating situations in the connection set at the second time point are retrieved and extracted in sequence. According to the slurry volume demand value corresponding to the boiler load demand at the third time point, each situation is judged item by item whether it can meet the slurry volume demand at the third time point under the constraints of the circulation pump frequency adjustment range and the safe frequency range. The connection situations that meet the conditions are selected to form a feasible situation connection set from the second time point to the third time point.

[0211] Based on the set of feasible situation connections from the second time point to the third time point, it progresses to all subsequent remaining time points in sequence, performs operational situation connections and constraint verification, and connects the feasible situation connections of all time points in series to form a feasible migration network between multiple time nodes.

[0212] Specifically, based on the feasible situation set of the first node (i.e., the first prediction time point) determined in the previous step, the system will perform a 1,k , where the first subscript represents the time node 1, and the second subscript k represents the kth feasible situation at the time node) is processed. First, for each S 1,k , extract its detailed configuration information, including the current operating frequency, start and stop status of all circulating pumps and the total slurry output flow rate under this situation. This information has been calculated or clarified in the previous step. At the same time, the slurry demand value corresponding to the boiler load predicted at the second time point (t = 2) is obtained from the future working condition demand and operating situation mapping set, which is recorded as Q demand,t＝2 , then, for the currently processed S 1,k , the system will traverse all pre-defined or dynamically generated potential target operating situations (these target situations are for the second time point and can be called estimated operating situations S 2,m ), for each estimated operating situation S 2,m , calculate or look up the table to get the total slurry output flow And determine whether the flow meets the demand at the second time point, that is, and (where α is set as before, for example 1.1), if it satisfies, then it is considered that 1,k to S 2,m is a potential valid connection, connect this pair (S 1,k →S 2,m ) is recorded, and all S in the feasible situation set of the first node are recorded. 1,k This matching process is performed, and finally all successful connection pairs are summarized to form a set of feasible situational connections from the first node to the second time point.

[0213] Based on the first node generated in the previous step, the feasible situation connection set to the second time point contains all feasible situations S from the first time point. 1,k Successfully connect to a feasible situation S at the second time point 2,m Now, the system will process all the running status S at the second time point in this connection set in turn. 2,m , for each S 2,m , extract its complete circulation pump operating frequency, start and stop status and total slurry output flow information, and at the same time, obtain the slurry volume demand value corresponding to the boiler load demand at the third time point (t=3) from the future working condition demand and operating status mapping set, recorded as Q demand,t＝3 , then, for the current S 2,mThe system needs to consider all possible next operating situations (i.e. the estimated operating situation S at the third time point). 3,n ), and for each S 3,n Make a judgment: First, check the 2,m Convert to S 3,n When the frequency adjustment range of each circulating pump is within the allowable range, for example, the upper limit of a single frequency adjustment is 5Hz (this value is set according to the equipment characteristics and process requirements. For example, to protect the inverter and motor and avoid excessive current shock, the manufacturer or operating procedures stipulate that the frequency change between adjacent scheduling cycles shall not exceed 5Hz). That is, for each pump u, Hz, then confirm S 3,n The operating frequencies of all pumps in the system are within their respective safe frequency ranges (e.g. 25Hz to 50Hz). Again, ensure that S 3,n Total slurry output flow Meeting the needs of the third time point, namely and Finally, the start-stop constraints should be checked, such as whether the minimum running or stopping time is met (for example, the pump is not allowed to stop if it has been running for less than 10 minutes. This 10 minutes is an empirical setting value to ensure that the pump is fully effective or to avoid frequent switching). Only when all these constraints are met can it be considered that the pump is started from S 2,m to S 3,n The connection is feasible, record this connection (S 2,m →S 3,n ), for all the situations S at the second time point 2,m This screening process is performed in all cases, and all connections that meet the conditions are summarized to form a set of feasible situational connections from the second time point to the third time point.

[0214] Based on the set of feasible situation connections from the second time point to the third time point constructed in the previous step, the process will continue to iterate forward, covering all subsequent predicted time points until the last time point (for example, if a total of T time points are predicted, it will advance from time point T-1 to time point T). In each iteration, for example, when building a connection from time point t to time point t+1, the system will use the known set of feasible situations at time point t (these situations are obtained by the connection screening and constraint verification in the previous t-1 step), and for each feasible situation S at time point t t,p , extract its configuration information, and then obtain the slurry volume demand Q at time point t+1 demand,t+1 , and then traverse all potential estimated operating situations S at time point t+1 t+1,q , and for each pair (S t,p →S t+1,q ) to conduct comprehensive constraint verification, including: S t+1,qDoes the total slurry output flow meet Q demand,t+1 The upper and lower limit requirements, from S t,p to S t+1,q Whether the frequency adjustment range of each pump is within the maximum allowable single-step adjustment value (for example, 5Hz), S t+1,q Whether the operating frequency of each pump is within its safe operating range, and whether all start and stop operations meet the requirements of equipment protection and process stability such as the minimum run / stop time interval, the connection (S t,p →S t+1,q ) is considered valid and is added to the set of feasible situation connections from time point t to time point t+1. Through this method of forward recursion, screening and connection of each time node, all the situation connections verified to be feasible between all time points are finally connected in series to form a network structure consisting of nodes (feasible operating situations) and directed edges (feasible migrations) covering the entire prediction time domain, that is, a feasible migration network between multiple time nodes.

[0215] The steps to obtain the full quantitative path alternative list are:

[0216] Based on the feasible migration network between multiple time nodes, the current operating status within the current time node is selected as the initial starting node. All operating status nodes connected to the next time node under the initial starting node are extracted one by one. The operating frequency, start / stop status, and slurry output flow rate of all circulating pumps contained in each status node are recorded in turn, and the connection relationship between each node is marked to form a status path set from the initial starting node to the second time node.

[0217] Based on the situation path set from the initial starting node to the second time node, all the circulation pump operating frequencies, start / stop states, and slurry output flows corresponding to each operating situation node within the second time node are retrieved in sequence, all feasible connection operating situations are matched to the third time node, and the connection relationship of each pair of nodes is recorded, and the connection is carried out one by one to the last time node, forming a node connection chain set from the initial starting node to the last time node;

[0218] Based on the set of node connection chains from the initial starting node to the end time node, starting from the initial starting node, all node connection chains are traversed one by one, and a continuous and complete path sequence of multiple time nodes is formed in series. The connection chains that are broken or cannot continue to the end time node are eliminated to form a fully quantified path alternative list.

[0219] Specifically, based on the feasible migration network between multiple time nodes constructed in the previous steps, the network stores all operating conditions (nodes) that meet the constraints at all predicted time points and all feasible migration paths (directed edges) between them in the form of a graph. First, the current actual operating circulation pump combination state (including the start and stop and frequency of each pump) is mapped to a specific operating state in the first time node (i.e., t=1) in the network. This state serves as the initial starting node for path search. Then, starting from the initial starting node, the feasible migration network is queried to find all operating state nodes directly connected to the next time node (t=2). For each such next node, the system will record in detail the operating frequency, start and stop status of all the circulation pumps it contains, and the total slurry output flow under this state. This information is part of the network node attributes. At the same time, the connection relationship from the initial starting node to this next node is clearly marked. For example, if the initial node is S 1,current , it is connected to the second time node with S 2,a , S 2,b , S 2,c , three preliminary path segments will be generated: (S 1,current →S 2,a ), (S 1,current →S 2,b ), (S 1,current →S 2,c ), and store these path segments and the node detailed information (frequency, start and stop, flow) contained in them to form a set of situation paths from the initial departure node to the second time node.

[0220] Based on the situation path set from the initial departure node to the second time node obtained in the previous step, each path segment in the set ends with a certain operation situation at the second time node. The system will expand these path segments and process the end nodes of each path segment in turn. For example, for the path segment (S 1,current →S 2,a ), the current end node is S 2,a , the system will retrieve S 2,a The corresponding operating frequency, start and stop status and slurry output flow of all circulating pumps are obtained. Then, all nodes from S are found in the feasible migration network between multiple time nodes. 2,a Start, connect to the feasible operation status node (such as S 3,x , S 3,y ), for each such feasible connection (e.g. S 2,a →S 3,x and S 2,a →S 3,y ), the system expands the original path segment to form a new longer path segment, such as (S 1,current →S2,a →S 3,x ) and (S 1,current →S 2,a →S 3,y ) and records the detailed information of these new connection relationships and newly added nodes. This process will be carried out recursively (or iteratively), that is, for all currently constructed path segments, their end nodes are processed, and the feasible situation of connecting to the next time node is found and connected until all path segments are extended to the predetermined end time node (for example, if the total prediction duration is T time units, it is extended to the Tth time node). Each step of extension is strictly based on the pre-verified connection relationship in the feasible migration network. Finally, a chain set consisting of a series of orderly connected running situation nodes that can reach the end time node step by step along the feasible migration path starting from the initial starting node is obtained. This is the node connection chain set from the initial starting node to the end time node.

[0221] Based on the node connection chain set from the initial departure node to the end time node generated in the previous step, each chain in this set represents a potential scheduling path starting from the current moment (initial departure node), spanning all future predicted time points, and ending at the end time node. However, these chains are not all complete and valid. Therefore, a final screening and sorting is required. Starting from the initial departure node, each node connection chain in the set is traversed one by one. For each chain, check whether it truly constitutes a complete path sequence that extends continuously and uninterruptedly from the initial time node to the end time node. This means that each intermediate node in the chain must have a valid connection to the next node, and The last node of the chain must be located at the predetermined end time node. If a chain is found to be broken at a certain time point in the middle (that is, the current node has no outgoing edge pointing to the next time node in the feasible migration network), or the chain fails to continue to the end time node (for example, it only extends to the second-to-last time point and then cannot continue), then such a chain will be regarded as an invalid path and will be eliminated. After this screening, all the chains retained are complete and continuous from beginning to end, and each node records in detail all the quantitative information of the operating status at that time point (such as the frequency of each pump, start and stop status, total slurry flow, etc.). These verified complete path sequences together constitute the full quantitative path candidate list.

[0222] The steps for obtaining the optimal scheduling instruction path of the circulating pump combination are:

[0223] Based on the fully quantified path candidate list, each complete path sequence in the list is retrieved one by one. The comprehensive cost of the operation situation transition path corresponding to each pair of adjacent operation situation nodes between the starting node and the end node of the path is gradually extracted. The comprehensive cost of the operation situation transition path between each pair of nodes is recorded one by one, and the comprehensive cost of the operation situation transition path of all adjacent node pairs is accumulated to form the cumulative comprehensive cost of the operation situation transition path corresponding to each path;

[0224] Based on the cumulative operation situation transition path comprehensive cost corresponding to each path, the cumulative operation situation transition path comprehensive cost of all complete path sequences is sorted from low to high to form a ranked list of cumulative path comprehensive costs;

[0225] Based on the path cumulative comprehensive cost ranking list, the complete path sequence with the lowest comprehensive cost of the operation status migration path in the ranking list is selected, and the optimal scheduling instruction path of the circulation pump combination is formed according to the circulation pump operation frequency and start-stop status recorded at each node in the path sequence.

[0226] Specifically, based on the full quantitative path alternative list generated in the previous step, the list contains multiple complete and continuous operation status path sequences starting from the current operation status to the end time node. The system will perform cost evaluation on each complete path sequence in the list. Specifically, for any path sequence, such as path P k =(S 1,a →S 2,b →S 3,c →…→S T,z ), the system will start from the starting node S of this path 1,a First, extract each pair of adjacent running status nodes in turn, for example, the first pair is (S 1,a ,S 2,b ), the second pair is (S 2,b ,S 3,c ), and so on, until the last pair (S T-1,y ,S T,z ), for each pair of such adjacent nodes (S t,curr ,S t+1,next ), the system will call the previously calculated and stored "operation situation migration path comprehensive cost table" and find the transition path from situation S t,curr Migrate to Situation S t+1,next The corresponding comprehensive cost R of the operation status migration path curr,next , the cost R curr,next The cost of energy transfer and equipment start-up and stop wear has been integrated, and the comprehensive cost R between each pair of adjacent nodes is curr,next Record them one by one, and then, this path P kThe comprehensive cost of the operation status migration path of all these adjacent node pairs is accumulated and summed up. Through this accumulation process, the complete path sequence P is obtained. k The corresponding total cumulative comprehensive cost of the operation status migration path.

[0227] Based on the cumulative comprehensive cost of the operation status transition path calculated for each complete path sequence in the fully quantified path candidate list in the previous step, for example, the total cost of path P1 is C(P1), the total cost of path P2 is C(P2), and so on. For example, if there are M such complete path sequences and their corresponding total costs in the list, the system will then sort these M paths according to their cumulative comprehensive cost of the operation status transition path. The sorting rule is from low to high, that is, the path with the smallest total cost is at the front, and the path with the largest total cost is at the back. If there are two or more paths with the same total cost, the preset secondary sorting rule (for example, giving priority to the path with the largest number of starts and stops) can be used. These secondary rules can be determined by introducing a small bias term when calculating the comprehensive cost, or by comparing these indicators when the costs are the same. For example, here, only the total cost is sorted. If the costs are the same, the order is arbitrary or in lexicographic order of the path numbers. These M (path sequence, total cost) pairs are processed by a standard sorting algorithm (such as quick sort, merge sort, etc.) to finally generate an ordered list. Each item in the list is a complete path sequence, and these path sequences are strictly arranged from smallest to largest according to their cumulative running status migration path comprehensive costs. This ordered list is the sorted list of cumulative path comprehensive costs.

[0228] Based on the path cumulative path comprehensive cost ranking list generated in the previous step, the list arranges all feasible complete scheduling paths from the current state to the future end time node according to their total cost from low to high. The system will directly select the complete path sequence at the top of this ranking list (i.e. the first one), because according to the aforementioned comprehensive cost calculation model, the cumulative operation status migration path comprehensive cost of this path is the lowest among all alternative paths, representing the optimal economic operation strategy under the current forecast and constraint conditions. Once this optimal path sequence is selected, for example, path P optimal =(S 1,opt →S 2,opt →…→S T,opt ), the system will follow the running status S of each time node (from t=1 to t=T) recorded in this path sequence t,optThe detailed configuration information of the slurry circulation pumps is extracted at each time point t, and the operating frequency and start-stop status of each slurry circulation pump are extracted. These time-series specific operation instructions (for example: when t=1, pump A runs @35Hz, pump B runs @40Hz, and pump C stops; when t=2, pump A runs @38Hz, pump B runs @42Hz, and pump C stops; ...; when t=T, ...) are organized to form a set of clear scheduling instruction sequences that can be directly issued to the desulfurization control system. This set of instruction sequences is the final optimal scheduling instruction path for the circulation pump combination.

[0229] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. The combined optimization method of double-tower desulfurization slurry circulation pump based on artificial intelligence is characterized by: The following steps are involved: Predict the boiler load values ​​at multiple future time points, define discrete operating states based on the start / stop status and operating frequency of the circulating pump group, establish the slurry volume output value corresponding to each state, match the slurry volume required for the boiler load at each future time point, and generate a mapping set of future operating conditions and operating states; Based on the mapping set of future operating condition requirements and operating status, all operating statuses are extracted, and the energy consumption transfer value caused by the change in operating frequency between any two statuses is calculated. An energy consumption transfer cost table is established, and based on the energy consumption transfer cost table, the quantified wear cost of the start-stop equipment is superimposed to generate a comprehensive cost of the operating status migration path; Based on the future working condition demand and operating situation mapping set, all operating situations that meet the slurry volume demand at the first time point are screened to obtain a feasible situation set for the first node; based on the feasible situation set for the first node, operating situations that meet the constraints are matched for subsequent time points in sequence and connections are established to establish a feasible migration network between multiple time nodes; Based on the feasible migration network between multiple time nodes, starting from the current operating situation and tracing to the end time node, all optional complete path sequences are generated to obtain a fully quantified path candidate list. Based on the fully quantified path candidate list, the comprehensive cost of the operating situation migration path is called to calculate the total cost of each path and sort them, and the optimal scheduling instruction path for the circulating pump combination is established.

2. The artificial intelligence-based double-tower desulfurization slurry circulation pump combination optimization method according to claim 1 is characterized in that: The steps for obtaining the future operating condition requirements and operating status mapping set are as follows: Based on the boiler operation records, the boiler load value, sulfur dioxide concentration at the absorption tower inlet, slurry circulation volume, liquid-gas ratio, pH value, oxygen concentration and circulation pump operation frequency at each time point in the boiler operation records are extracted to obtain the boiler load prediction input set; Calculating boiler load forecast values ​​at each future time point based on the boiler load forecast input set; Based on the boiler load prediction value, the start and stop status and operating frequency of the circulating pump are called to define a combined situation set, the predicted load is matched with the slurry output corresponding to each situation, and a mapping set of future working condition requirements and operating situations is generated.

3. The artificial intelligence-based double-tower desulfurization slurry circulation pump combination optimization method according to claim 1 is characterized in that: The steps for obtaining the energy consumption transfer cost table are as follows: Based on the future working condition demand and operating status mapping set, the operating frequency values ​​of all circulating pumps in each operating status are sequentially extracted, the frequencies of the corresponding circulating pumps under any two different operating statuses are paired, and the target operating duration of the circulating pump after the status transition is simultaneously extracted to generate an operating frequency transition and duration pairing sequence; Calculating a total energy consumption transfer value based on the operating frequency transfer and duration pairing sequence; Based on the total energy consumption transfer value, the energy consumption changes between each pair of operating states are registered item by item, a matrix structure is constructed according to the operating state number, the energy consumption differences of the transfer paths between all states are output in a row and column manner, and an energy consumption transfer cost table is generated.

4. The artificial intelligence-based double-tower desulfurization slurry circulation pump combination optimization method according to claim 1, characterized in that: The steps for obtaining the comprehensive cost of the operation status transition path are as follows: Based on the operational status start and stop records, the operating status identifiers of each device before and after each group of operational status transitions are compared to identify devices that have changed from running to stopped or from stopped to running. Each device with a status change is recorded as a start and stop action, and the operating time of the most recent operating cycle is extracted to generate a list of started and stopped devices and their running time before the start and stop. Calculate the cost of the operation status migration path based on the list of start-stop devices and their running time before start-stop and the energy consumption transfer cost table; Based on the migration path cost, the migration path costs between all operating situation pairs are sequentially filled into the cost table structure matrix, and classified and arranged according to the starting situation and target situation indexes to generate the comprehensive cost of the operating situation migration path.

5. The artificial intelligence-based double-tower desulfurization slurry circulation pump combination optimization method according to claim 1, characterized in that: The steps for obtaining the feasible situation set of the first node are: Based on the future working condition demand and operating status mapping set, searching the operating frequency and start / stop status of all circulation pumps in each operating status in the future working condition demand and operating status mapping set one by one, calculating the slurry output flow rate corresponding to each circulation pump one by one, and summing the slurry output flow rates of all circulation pumps to form a total slurry output flow rate for each operating status; Based on the total slurry output flow rate of each operating situation, the slurry demand value required for the boiler load corresponding to the first time point in the future operating condition demand and the operating situation mapping set is extracted. The total slurry output flow rate of each operating situation is compared with the demand value one by one. All operating situations with a total slurry output flow rate equal to or higher than the demand value are selected to form a preliminary operating situation set that meets the boiler load slurry demand at the first time point; Based on the preliminary operating status set that meets the boiler load slurry demand at the first time point, check item by item whether the start / stop status and frequency of each circulating pump in the operating status are within the safe operating range allowed by the equipment, eliminate the operating status with start / stop status or frequency exceeding the safe operating range, and form a feasible status set for the first node.

6. The artificial intelligence-based double-tower desulfurization slurry circulation pump combination optimization method according to claim 1, characterized in that: The steps for obtaining the feasible migration network between the multiple time nodes are: Based on the feasible situation set of the first node, the operating frequency, start / stop status, and slurry output flow rate of all circulating pumps corresponding to each operating situation in the set are extracted one by one; according to the slurry volume demand value corresponding to the boiler load demand at the subsequent second time point, all estimated operating situations that can meet the slurry volume demand at the second time point are matched to each operating situation in turn, thereby forming a feasible situation connection set from the first node to the second time point; Based on the feasible situation connection set from the first node to the second time point, all circulating pump operating frequencies, start / stop states, and slurry output flows of all operating situations at the second time point in the connection set are sequentially retrieved and extracted; based on the slurry volume demand value corresponding to the boiler load demand at the third time point, each situation is judged item by item whether it can meet the slurry demand at the third time point under the constraints of the circulating pump frequency adjustment amplitude and the safe frequency interval; the connection situations that meet the conditions are selected to form a feasible situation connection set from the second time point to the third time point; Based on the set of feasible situation connections from the second time point to the third time point, it progresses to all subsequent remaining time points in sequence, performs operational situation connections and constraint verification, and connects the feasible situation connections of all time points in series to form a feasible migration network between multiple time nodes.

7. The artificial intelligence-based double-tower desulfurization slurry circulation pump combination optimization method according to claim 1, characterized in that: The steps for obtaining the full quantization path candidate list are as follows: Based on the feasible migration network between multiple time nodes, the current operating status in the current time node is selected as the initial departure node, and all operating status nodes connected to the next time node under the initial departure node are extracted one by one. The operating frequency, start and stop status, and slurry output flow rate of all circulating pumps contained in each status node are recorded in turn, and the connection relationship between each node is marked to form a status path set from the initial departure node to the second time node; Based on the situation path set from the initial departure node to the second time node, all the circulation pump operating frequencies, start / stop states, and slurry output flows corresponding to each operating situation node within the second time node are retrieved in sequence, all feasible connection operating situations are matched to the third time node, and the connection relationship of each pair of nodes is recorded, and the connection is progressed one by one to the end time node to form a node connection chain set from the initial departure node to the end time node; Based on the node connection chain set from the initial starting node to the end time node, starting from the initial starting node, all node connection chains are traversed one by one, and a continuous and complete path sequence of multiple time nodes is formed in series. The connection chains that are broken or cannot continue to the end time node are eliminated to form a fully quantified path alternative list.

8. The artificial intelligence-based double-tower desulfurization slurry circulation pump combination optimization method according to claim 1, characterized in that: The steps for obtaining the optimal scheduling instruction path of the circulating pump combination are: Based on the fully quantified path candidate list, each complete path sequence in the list is retrieved one by one, and the comprehensive cost of the operation situation transition path corresponding to each pair of adjacent operation situation nodes between the starting node and the end node of the path is gradually extracted. The comprehensive cost of the operation situation transition path between each pair of nodes is recorded one by one, and the comprehensive cost of the operation situation transition path of all adjacent node pairs is accumulated to form a cumulative comprehensive cost of the operation situation transition path corresponding to each path; Based on the cumulative operation situation transition path comprehensive cost corresponding to each path, the cumulative operation situation transition path comprehensive cost of all complete path sequences is sorted from low to high to form a ranked list of cumulative path comprehensive costs; Based on the ranked list of cumulative path comprehensive costs, a complete path sequence with the lowest comprehensive cost of the operation status migration path in the ranked list is selected, and the optimal scheduling instruction path for the circulation pump combination is formed according to the circulation pump operation frequency and start / stop status recorded at each node in the path sequence.

9. The double-tower desulfurization slurry circulation pump combination optimization system according to any one of claims 1 to 8, characterized in that: include: Prediction module: Predicts boiler load values ​​at multiple future time points, defines discrete operating states based on the start / stop status and operating frequency of the circulating pump group, establishes slurry output values ​​corresponding to each state, matches the slurry volume required for boiler load at each future time point, and generates a mapping set between future operating condition requirements and operating states; Cost modeling module: Based on the future operating condition demand and operating status mapping set, all operating statuses are extracted, the energy consumption transfer value caused by the change in operating frequency between any two statuses is calculated, and an energy consumption transfer cost table is established. Based on the energy consumption transfer cost table, the quantified start-stop equipment wear cost is superimposed to generate the comprehensive cost of the operating status migration path; Network construction module: Based on the future working condition demand and operation status mapping set, all operation statuses that meet the slurry volume demand at the first time point are screened, and a feasible status set of the first node is obtained. Based on the feasible status set of the first node, operation statuses that meet the constraints are matched for subsequent time points in sequence and connections are established to establish a feasible migration network between multiple time nodes; Path optimization module: Based on the feasible migration network between the multiple time nodes, starting from the current operation status, tracing to the end time node, generating all optional complete path sequences, and obtaining a fully quantified path candidate list. Based on the fully quantified path candidate list, calling the comprehensive cost of the operation status migration path to calculate the total cost of each path and sort them, and establish the optimal scheduling instruction path for the circulation pump combination.

Images