A method and system for calculating a heating value based on energy-saving requirements

Abstract

This invention relates to the field of energy system optimization and control technology, specifically a method and system for calculating calorific value based on energy-saving requirements. The method includes: establishing a correlation model between main steam flow and electrical and thermal loads based on historical operating data; calculating the required main steam flow for the forecast period based on the electrical and thermal loads in the final load forecast; dynamically calculating the minimum calorific value based on the main steam flow, the maximum output of the pulverizing system, the enthalpy of the main steam, and the rated operating enthalpy; and constructing a multi-constraint mixed integer programming model with the objective function of minimizing fuel costs, and solving and outputting multiple fuel configuration schemes based on the minimum calorific value. This invention employs a hybrid forecasting strategy that combines LSTM deep learning with manual empirical rules, significantly reducing forecasting errors and improving load forecasting accuracy. It effectively solves the problems of high arbitrariness and low accuracy in manual forecasting, providing reliable load data for coal blending.

Description

Technical Field

[0001] This invention relates to the field of energy system optimization and control technology, and in particular to a method and system for calculating calorific value based on energy-saving requirements. Background Technology

[0002] Thermal power generation is the mainstay of my country's electricity supply, with coal costs accounting for 60%-70% of the total power generation cost. With the deepening of power market reform and the large-scale grid connection of new energy sources, thermal power units face complex operating conditions such as deep peak shaving, frequent start-ups and shutdowns, and variable coal quality. The traditional coal blending method is no longer able to meet the comprehensive requirements of modern power systems for safety, economy, and environmental protection.

[0003] Coal blending requires forecasting the unit load for the following day to determine the appropriate slag replenishment plan for each unit. Currently, most power plants still rely on operators to manually revise the D+1 day 96-point load plan based on the grid's D+1 day load plan and Ningxia's new energy load forecast, providing this information for coal blending. Then, based on the electricity load plan, the previous day's average heat supply, and the coal inventory at the coal yard, a coal supply plan is formulated for each pulverizing system. On the one hand, manual forecasting suffers from high arbitrariness, low accuracy, and difficulty in real-time tracking of new energy output fluctuations and grid peak-shaving demands. On the other hand, coal supply plans formulated under these circumstances often suffer from uneconomical coal calorific value and inability to handle high loads.

[0004] Therefore, there is an urgent need to provide a method and system for calculating calorific value based on energy-saving requirements, so as to improve the accuracy of load forecasting and the economic efficiency of coal calorific value. Summary of the Invention

[0005] In this section, as well as in the abstract and title of this application, some simplifications or omissions may be made to avoid obscuring the purpose of this section, the abstract, and the title of this application, and such simplifications or omissions shall not be used to limit the scope of the invention.

[0006] To address the shortcomings of existing technologies, one objective of this invention is to provide a method for calculating calorific value based on energy-saving requirements.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: Based on historical operating data, a correlation model between main steam flow and electrical and thermal loads is established. Based on the electrical and thermal loads in the final load forecast, the required main steam flow for the forecast period is calculated.

[0008] The minimum calorific value is dynamically calculated based on the main steam flow rate, the maximum output of the pulverizing system, the enthalpy of the main steam, and the rated enthalpy under operating conditions.

[0009] Construct a multi-constraint mixed integer programming model with the objective function of minimizing fuel cost, and solve for multiple fuel configuration schemes based on the lowest calorific value.

[0010] As a preferred embodiment of the energy-saving demand-based calorific value calculation method of the present invention, the correlation model adopts polynomial fitting; the fitting parameters are automatically updated daily based on the data of the day. The automatic update is triggered by the following conditions: daily timed triggering or when the daily load of D-1 is lower than 30% of the rated load, the model parameters of day D-2 are used, and the update of the day is skipped. Here, D represents the current day, D-1 represents the previous day, and D-2 represents the previous two days.

[0011] In a preferred embodiment of the energy-saving demand-based calorific value calculation method of the present invention, the final load forecast value is determined by comparing the manual forecast value and the model forecast value. When the deviation between the two exceeds 20%, the manual forecast value is taken as the final load forecast value; otherwise, the model forecast value is taken as the final load forecast value. The model forecast value includes, based on a Long Short-Term Memory (LSTM) neural network, using the AGC scheduling values, planned values, and time characteristics of the past N days, as well as the planned value and time characteristics of D+1, to predict the AGC scheduling value for day D+1, thus obtaining the model forecast value. Here, D represents the current day, and D+1 represents the next day.

[0012] As a preferred embodiment of the energy-saving demand-based calorific value calculation method of the present invention, the formula for calculating the minimum calorific value of the coal fed into the furnace is: in, Where F is the minimum calorific value, k is the empirical coefficient for boiler thermal efficiency, and F is the minimum calorific value. ms Main steam flow rate, M max For the pulverizing system to operate at maximum capacity, h actual For the main steam enthalpy under variable load conditions, h rated This is the enthalpy value of the main steam under rated operating conditions.

[0013] As a preferred embodiment of the calorific value calculation method based on energy-saving requirements described in this invention, the decision variables of the objective function include the coal type selection and coal feeder output of each pulverizing system.

[0014] As a preferred embodiment of the calorific value calculation method based on energy-saving requirements described in this invention, the constraints of the multi-constraint mixed integer programming model include the output limit of the pulverizing system and the energy quality limit of the furnace feed, wherein the energy quality limit of the furnace feed includes the lower limit of calorific value, the upper limit of sulfur content, the upper limit of ash content, and the lower limit of ash melting point.

[0015] As a preferred embodiment of the energy-saving demand-based calorific value calculation method of the present invention, the constraints of the multi-constraint mixed integer programming model further include the coal feeder output balance constraint: , Where, x i To provide power to the i-th coal feeder, The average output of each coal feeder is Δmax, which is the preset maximum allowable deviation.

[0016] As a preferred embodiment of the energy-saving demand-based calorific value calculation method of the present invention, the step of solving and outputting multiple fuel configuration schemes based on the minimum calorific value includes calculating the number of all possible coal type combinations based on the available coal types configured in each pulverizing system. N comb = , Where, N comb Let n be the number of fuel type combinations, and n be the number of pulverizing systems in operation. Let be the number of available coal types for the i-th pulverizing system; For each coal type combination, with the coal type selection variable fixed, a linear programming problem concerning the output of the coal feeder is solved. Combinations that do not meet the constraints are eliminated, and the combinations are arranged by cost to provide an energy allocation scheme.

[0017] As a preferred embodiment of the energy-saving demand-based calorific value calculation method of the present invention, it further includes a virtual coal type management step: receiving two original coal types and their blending ratio α input by the user. Where 0 < α < 1; automatically calculate the industrial analysis parameters of the virtual mixed coal type, including the received lower heating value, total moisture, received total sulfur, received ash, received volatile matter, received fixed carbon and ash fusion point temperature, and each parameter is calculated by weighted average according to the blending ratio; when the coal blending scheme is output, the virtual mixed coal type is decomposed and displayed as the original coal type combination form.

[0018] To address the shortcomings of existing technologies, one objective of this invention is to provide a calorific value calculation system based on energy-saving requirements, comprising: a load forecasting module that receives 96-point planned output curves of the power grid, new energy load forecasting data, and planned unit shutdown information; predicts the load curve for day D+1 based on an LSTM neural network and corrects it using manual experience rules; and allocates the load to each operating unit according to the unit shutdown status. The main steam flow calculation module stores and updates the polynomial fitting model of main steam flow with electrical load and heat load daily; it calculates the required main steam flow based on the predicted load. The dynamic calorific value calculation module stores the sliding pressure operation curve of the storage unit and the calculation formula of IAPWS-IF97. The main steam pressure is determined based on the load rate, and the corresponding enthalpy value is calculated; the minimum calorific value of the coal fed into the furnace is calculated based on the main steam flow rate, the maximum output of the pulverizing system, and the enthalpy ratio. The multi-constraint optimization solution module stores a coal type information database, the available coal type configuration in the coal bunker, the output limit of the pulverizing system, and the quality limit of the coal fed into the furnace; it constructs and solves a mixed integer programming model with the goal of minimizing coal consumption cost; and generates and sorts multiple alternative coal blending schemes. The scheme management module receives the coal blending scheme selected by the user and generates the final coal blending instruction containing the suggested silo quantity; it manages the generation, calculation and deletion of virtual mixed coal types, which are industrial analysis parameters calculated by weighting two original coal types according to the configuration ratio. The evaluation and feedback module collects actual operating data from the SIS system, calculates the deviation rate between the planned calorific value and the actual calorific value, and assigns the deviation to the responsible work team based on the time offset algorithm, generating performance evaluation numbers.

[0019] The time offset algorithm includes: setting a fixed time offset Δt, attributing the deviation data of the actual furnace feeding time to the fuel operation team before the time Δt; dividing 24 hours into four shift periods: A, B, C, and D, according to the power plant shift schedule; and calculating the average deviation rate of each shift period as the monthly performance indicator of that shift.

[0020] The beneficial effects of this invention are: 1. This invention adopts a hybrid prediction strategy that combines LSTM deep learning with human experience rules, which significantly reduces prediction errors and improves load prediction accuracy. It effectively solves the problems of high arbitrariness and low accuracy in manual prediction, and provides reliable load basis data for coal blending and combustion.

[0021] 2. This invention establishes a dynamic calculation model for the minimum calorific value of coal fed into the furnace based on the sliding pressure curve and the IAPWS-IF97 formula. It tracks the change of main steam pressure with load rate in real time, accurately calculates the steam enthalpy under variable load conditions, solves the problem of high load not being able to drive the furnace and low load calorific value waste caused by fixed calorific value standards, and realizes refined fuel management under variable pressure operation mode.

[0022] 3. This invention achieves cost savings through dynamic enthalpy correction, deeply coupling the first law of thermodynamics (IAPWS-IF97 equation of state) with the economic operation requirements of the power system, accurately calculating the steam enthalpy of variable load, reflecting the real energy demand, and avoiding waste due to excessive calorific value and load limitation due to insufficient calorific value. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 These are the preliminary prediction results of the prediction model based on the LSTM algorithm.

[0025] Figure 2 This is a prediction deviation diagram for Unit 1.

[0026] Figure 3 The load forecast curves for each unit are shown.

[0027] Figure 4 This is the sliding pressure curve of the first phase unit.

[0028] Figure 5 This is the sliding pressure curve for the second-phase unit.

[0029] Figure 6 This is the result of the program's calculation.

[0030] Figure 7 This is a comparison chart of the calorific value deviations of coal blending. Detailed Implementation

[0031] To make the objectives, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0032] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0033] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0034] Example 1

[0035] This embodiment provides a method for calculating calorific value based on energy-saving requirements. Coal is used as the fuel type in this embodiment as an example, specifically including: S1: Load Forecasting

[0036] The system acquires the planned output curve for 96 points on day D+1, renewable energy load forecast data, and planned unit shutdown information issued by the power grid. Based on a Long Short-Term Memory (LSTM) neural network, it uses the AGC scheduling values, planned values, and time characteristics of the past N days to predict the AGC scheduling value for day D+1, obtaining the model's predicted value. The model's specific parameters are updated during training, and the model is automatically updated by periodically updating the training data. D represents the current day, and D+1 represents the next day.

[0037] The model prediction is compared with the human prediction calculated based on human experience rules. When the deviation exceeds a preset threshold (such as 20%), the human prediction is used to overwrite the model prediction to obtain the final load prediction.

[0038] The rules for manual prediction are shown in the table below:

[0039] Table 1

[0040] Based on the planned shutdown status of the generating units, the final load forecast is allocated to each unit: for Phase I units, the load is evenly distributed when both units are running, and when only one unit is running, the operating unit bears the entire load but does not exceed the rated load limit; for Phase II units, the grid planned value is directly adopted.

[0041] S2: Main Steam Flow Calculation

[0042] Based on historical operating data, a correlation model between main steam flow and electrical and thermal loads is established using polynomial fitting. The correlation model automatically updates the fitting parameters daily based on the data of that day. The main steam flow required for the forecast period is calculated based on the electrical and thermal loads in the final load forecast.

[0043] The fitted parameters are automatically updated daily based on the data of that day. The automatic update is triggered by the following conditions: daily scheduled triggering or when the electrical load of day D-1 is lower than 30% of the rated load, the model parameters of day D-2 are used and the update for that day is skipped. Here, D represents the current day, D-1 represents the previous day, and D-2 represents the previous two days.

[0044] S3: Dynamic calculation of minimum calorific value of coal fed into the furnace

[0045] Based on the unit's sliding pressure operation curve, the main steam pressure under different load rates is determined, and the main steam enthalpy corresponding to the main steam pressure is calculated based on the IAPWS-IF97 formula.

[0046] Based on the main steam flow rate, maximum output of the pulverizing system, main steam enthalpy, and rated operating enthalpy, the minimum calorific value of the coal fed into the furnace is calculated using the following formula: in, Where F is the minimum calorific value, k is the empirical coefficient for boiler thermal efficiency, and F is the minimum calorific value. ms Main steam flow rate, M max For the pulverizing system to operate at maximum capacity, h actual For the main steam enthalpy under variable load conditions, h rated This is the enthalpy value of the main steam under rated operating conditions.

[0047] S4: Multi-constraint coal blending optimization solution

[0048] Obtain the available coal types in the coal bunker, the output limits of the pulverizing system, and the quality limits of the coal fed into the furnace. The quality limits include the lower limit of calorific value, the upper limit of sulfur content, the upper limit of ash content, and the lower limit of ash fusion point. A mixed-integer programming model is constructed with the objective function of minimizing coal consumption costs, and constraints including pulverizing system output limitations and coal quality limitations (specifically calorific value, sulfur content, ash content, and ash fusion point). The decision variables of the objective function include the coal type selection and feeder output for each pulverizing system. The constraints of the mixed-integer programming model also include feeder output balance constraints. , Where, x i To provide power to the i-th coal feeder, The average output of each coal feeder is Δmax, which is the preset maximum allowable deviation.

[0049] Iterate through all feasible coal type combinations and calculate the number of all possible coal type combinations based on the available coal types configured in each pulverizing system: N comb = , Where, N comb Let n be the number of fuel type combinations, and n be the number of pulverizing systems in operation. Let be the number of available coal types for the i-th pulverizing system.

[0050] Solve the linear programming problem for each combination to obtain the lowest cost solution that satisfies all constraints, and output multiple alternative solutions with the lowest cost for the user to choose from.

[0051] S5: Implementation and Evaluation Steps

[0052] Based on the user-selected coal blending scheme, the recommended silo quantity for each coal bunker is calculated, and the final coal blending instruction is generated. Actual operating data is collected, and the deviation between the actual combustion calorific value and the planned calorific value is calculated. The actual combustion calorific value is obtained by back-calculation based on the main steam flow rate and the actual coal quantity. According to the time offset algorithm, the deviation is attributed to the corresponding fuel operation team, and a performance evaluation report is generated.

[0053] The time offset algorithm includes: setting a fixed time offset Δt, attributing the deviation data of the actual furnace feeding time to the fuel operation team before the time Δt; dividing 24 hours into four shift periods: A, B, C, and D, according to the power plant shift schedule; and calculating the average deviation rate of each shift period as the monthly performance indicator of that shift.

[0054] Example 2

[0055] This is the second embodiment of the present invention. Taking a power plant located in Ningxia as an example, this embodiment specifically demonstrates a calorific value calculation system based on energy-saving requirements.

[0056] The installed capacity of this power plant is: Phase I units (Units 1 and 2, each with a rated capacity of 600MW, superheated steam outlet pressure of 17.4MPa, and temperature of 541℃); Phase II units (Units 3 and 4, each with a rated capacity of 1060MW, superheated steam outlet pressure of 25.98MPa, and temperature of 605℃).

[0057] S1: Load Forecasting

[0058] Data collection includes: time, upper limit of units 1-4 (corresponding to units #1, #2, #3, and #4 in the table below), 96-point grid plan for Phase I, new energy load, 96-point grid plan for Phase II, lower limit declared value of units 1-4, deep adjustment declared value of units 1-4, and planned shutdown information of units.

[0059] The model uses data from the past two years to train a prediction model based on the LSTM algorithm. This model uses the AGC scheduling values, planned values, and time characteristics of the past seven days, as well as the planned values ​​and time characteristics of day D+1, to predict the AGC scheduling value for day D+1. The specific parameters of the model are updated during training, and the model is automatically updated by periodically updating the training data. Based on data obtained from the power grid's 96-point technical output curves, combined with the prediction model, periodic predictions are performed to obtain preliminary prediction results, as shown below. Figure 1 As shown.

[0060] The model's prediction accuracy is no less than that of the user's manual prediction. Figure 2 The prediction bias of Unit 1 is shown.

[0061] Taking the data from the first period of 2024 as an example, the error statistics between the corrected value and the AGC scheduling value in 2024, after manual correction by the user, are as follows: Daily average error: 119.46 MW, daily average error percentage: 17.45%

[0062] Maximum daily error: 306.47 MW, maximum daily error percentage: 43.46%

[0063] Percentage of data points with a daily error exceeding 10%: 56.39%

[0064] Taking the data from December 2024 as an example (excluding data from three days of downtime), the error statistics between the corrected value and the AGC scheduling value in December 2024, after manual correction by the user, are as follows: Daily average error: 43.47MW, daily average error percentage: 5.17%

[0065] Maximum daily error: 145.50MW, maximum daily error percentage: 17.89%

[0066] The percentage of data points with a daily error exceeding 10% was 16.05%.

[0067] The load forecast values ​​are allocated to each unit, and curves are plotted, such as... Figure 3 As shown. The allocation rule is that when the deviation between the model prediction value and the user's manual prediction method exceeds 20%, the user's manual algorithm prediction is used for that point (see Table 1 for the rules of manual prediction values). Only the first phase prediction is performed; for the second phase, no model prediction is performed, and the planned values ​​of 96 points issued by the power grid are directly taken. The predicted load of units 1 and 2 is evenly distributed according to the total load of the first phase; the predicted load of units 3 and 4 is evenly distributed according to the total load of the second phase. Furthermore, a single unit cannot exceed its rated load (units 1 and 2 are rated at 600MW, units 3 and 4 are rated at 1060MW), and cannot be lower than 30% of its rated load. If a unit has a planned shutdown on the second day, that unit will not participate in the load allocation. If the load of a unit participating in the allocation is greater than its rated load, it will still be evenly distributed at that time (the same applies to the second phase).

[0068] S2: Main Steam Flow Calculation

[0069] Based on historical operating data, a correlation model between main steam flow and electrical and thermal loads is established using polynomial fitting. The correlation model automatically updates the fitting parameters daily based on the data of that day. The main steam flow required for the forecast period is calculated based on the electrical and thermal loads in the final load forecast.

[0070] By fitting historical data from Unit 1 on November 2, 2024, December 1, 2024, and December 2, 2024, the following three formulas were obtained:

[0071] November 2, 2024: Main steam flow rate = 52.6 + 3.411 × electrical load + 0.0905 × heat load, average deviation -0.02%, maximum deviation at a single point 1.04%.

[0072] December 1, 2024: Main steam flow rate = -163.5 + 4.0435 × electrical load + 0.005 × heat load, average deviation -0.03%, maximum deviation at a single point 1.85%.

[0073] December 2, 2024: Main steam flow rate = -296.2 + 4.2.05 × electrical load + 0.127 × heat load, average deviation -0.03%, maximum deviation at a single point 1.14%.

[0074] It can be observed that expressions for main steam flow rate, electrical load, and thermal load can be obtained through polynomial fitting, and the accuracy meets the coal blending requirements.

[0075] Using data from November 2, 2024 and December 1, 2024, the situation on December 2, 2024 is predicted, as shown in Table 2.

[0076] Table 2. Forecasting Across Time Periods

[0077] It can be observed that the prediction accuracy does not significantly deteriorate with each subsequent day, but the further away the date is from December 2nd, the greater the prediction deviation. Therefore, the model needs to refit the formula daily based on the current day's data to predict the next day's situation. If a fixed fitting relationship is used, it is foreseeable that the prediction deviation will increase over time. Verification has shown that when the time span reaches 3 months, the deviation reaches approximately 10%. Therefore, polynomial fitting is required, and it should be automatically updated daily. When the daily load on day D-1 is less than 30% of the rated load, the model parameters from day D-2 should continue to be used.

[0078] S3: Dynamic calculation of minimum calorific value of coal fed into the furnace

[0079] Based on boiler equipment selection and thermodynamic formulas, and according to the main steam flow rate of the rated production parameters, the total heat carried by the boiler outlet steam can be determined. Combining the boiler efficiency guarantee value and experimental corrections, the minimum calorific value of coal for each boiler can be calculated using the maximum output of the pulverizing system (the pulverizing system output is manually entered based on actual conditions). The calculation formula is as follows: in, Where F is the minimum calorific value, k is the empirical coefficient for boiler thermal efficiency, and F is the minimum calorific value. ms Main steam flow rate, M max For the pulverizing system to operate at maximum capacity, h actual For the main steam enthalpy under variable load conditions, h rated This is the enthalpy value of the main steam under rated operating conditions.

[0080] The boiler's BRL corresponds to the turbine's TRL. THA represents the turbine's rated operating condition at maximum efficiency, which is an ideal condition that can be maintained briefly during performance testing. Although the THA condition offers the highest efficiency, it is not used because it can only be maintained briefly.

[0081] Table 3 Equipment BRL Operating Parameters

[0082] Under rated load: Boiler No. 1: Minimum calorific value of coal - rated load = 3.05 × main steam flow rate / maximum coal quantity

[0083] Boiler No. 2: Minimum calorific value of coal - rated load = 2.951 × main steam flow rate / maximum coal quantity

[0084] Boiler No. 3: Minimum calorific value of coal - rated load = 3.13 × main steam flow rate / maximum coal quantity

[0085] Boiler No. 4: Minimum calorific value of coal - rated load = 3.027 × main steam flow rate / maximum coal quantity

[0086] The boiler adopts a constant sliding pressure and constant variable pressure operation mode. The rated load of the first-phase unit is 600MW, and the sliding pressure curve is as follows: Figure 4 As shown, considering the pressure loss from the boiler side to the main steam valve, it is expressed as follows: , The sliding pressure curve of the second-phase unit is as follows: Figure 5 As shown, considering the pressure loss from the boiler side to the main steam valve, it is expressed as follows: , According to the IAPWS-IF97 formula, we can know...

[0087] Under rated load of 17.4 MPa and 541℃, the enthalpy of steam for the first-phase unit is 3399.2248 kJ / kg.

[0088] At rated load of 25.98 MPa and 605℃, the enthalpy of steam for the second-phase unit is 3499.5894 kJ / kg.

[0089] The steam pressure under different loads is determined based on the sliding pressure curve. The temperature is taken as the rated temperature. The enthalpy value of the corresponding state, i.e., the enthalpy value of the main steam under varying loads, is obtained using the IAPWS-IF97 formula and denoted as h. actual .

[0090] Therefore, under arbitrary load:

[0091] Boiler No. 1: Minimum calorific value of coal - arbitrary load = 3.05 × main steam flow rate / maximum coal consumption × h actual / 3399.2248

[0092] Boiler No. 2: Minimum calorific value of coal - arbitrary load = 2.951 × main steam flow rate / maximum coal flow rate × h actual / 3399.2248

[0093] Boiler No. 3: Minimum calorific value of coal - arbitrary load = 3.13 × main steam flow rate / maximum coal consumption × h actual / 3499.5894

[0094] Boiler No. 4: Minimum calorific value of coal - arbitrary load = 3.027 × main steam flow rate / maximum coal consumption × h actual / 3499.5894

[0095] S4: Solution steps for multi-constraint coal blending optimization

[0096] Obtain the available coal types in the coal bunker, the output limits of the pulverizing system, and the quality limits of the coal entering the furnace. The quality limits include the lower limit of calorific value, the upper limit of sulfur content, the upper limit of ash content, and the lower limit of ash fusion point.

[0097] This power plant project involves external coal blending, which involves mixing two types of raw coal in different proportions to create a new type of coal (hereinafter referred to as blended coal). Industrial analysis of the blended coal is calculated based on the weighted average of the raw coal.

[0098] Table 4 Coal Type Information

[0099] Each of the four generating units is configured with its own available coal type. Table 5 shows the available coal types in the raw coal bunker of Unit 5.

[0100] The output limits of the pulverizing system are shown in the table below.

[0101] Table 6 Output Limitations of the Mechanized Powder System

[0102] A mixed-integer programming model is constructed with the objective function of minimizing coal consumption cost and constraints of pulverizing system output and coal quality limits. The decision variables of the objective function include the coal type selection and feeder output for each pulverizing system. The constraints of the mixed-integer programming model also include feeder output balance constraints. , Where, x i To provide power to the i-th coal feeder, The average output of each coal feeder is Δmax, which is the preset maximum allowable deviation.

[0103] Iterate through all feasible coal type combinations and calculate the number of all possible coal type combinations based on the available coal types configured in each pulverizing system: , Where n is the number of powder-making systems activated. Let be the number of available coal types for the i-th pulverizing system.

[0104] Solve the linear programming problem for each combination to obtain the lowest cost solution that satisfies all constraints, and output multiple alternative solutions with the lowest cost for the user to choose from.

[0105] Through planning and solving, the minimum coal cost for each pulverizing system, based on the following output, is 136,050 yuan / h. The weighted average calorific value is 21.422 MJ / kg, the weighted average sulfur content is 0.938%, the weighted average ash content is 20.28%, and the ash fusion point is 1226℃. This meets the requirements of a weighted average calorific value ≥16.5 MJ / kg, sulfur content ≤1.2%, ash content ≤30%, and ash fusion point ≥1200℃, and can satisfy the output requirements of an electrical load of 450MW and a thermal load of 380MW.

[0106] Weighted average price of standard coal (yuan / t) = Weighted average price of raw coal / Weighted average calorific value * Standard coal calorific value = Weighted average price of raw coal / Weighted average calorific value * 29.271

[0107] =[(651*x1+627*x2+627*x3+418*x4+651*x5) / (x1+x2+x3+x4+x5)] / [(24.42*x1+20.18*x2+20.18*x3+16.51*x4+24.42*x5) / (x1+x2+x3+x4+x5)]*29.271

[0108] =(651*x1+627*x2+627*x3+418*x4+651*x5) / [(24.42*x1+20.18*x2+20.18*x3+16.51*x4+24.42*x5)*29.271

[0109] The program calculation results are displayed as follows Figure 6 The second solution is as follows, and can be compared with the table below: Table 7 Comparison of the two scheme combinations

[0110] Considering that coal mills A and F use the same type of coal, and coal mills B and C use the same type of coal, if the output of coal mill B+C is 52.392 t / h, there can be multiple output schemes, which are equivalent in terms of the equations.

[0111] Furthermore, the program was used to traverse and calculate the remaining 127 combinations, and to find the top five combinations with the lowest coal-fired cost for reference.

[0112] Excessive deviation in the output of the coal feeder can easily lead to excessively high local heat load, and the output deviation of the coal feeder is generally limited.

[0113] Based on the experience-based strategy of maximizing the output of low-quality coal, given that the coal mill's output is generally average, this approach is to...

[0114] The output of pulverizing system A is 50t / h, the output of pulverizing system B is 55t / h, the output of pulverizing system C is 55t / h, the output of pulverizing system D is 60t / h, and the output of pulverizing system E is 30t / h. Then the weighted calorific value is 20.656, the weighted sulfur content is 0.9876, the weighted ash content is 22.8552, and the price is 146,130 yuan / h > 136,050 yuan / h, which increases the cost by nearly 10,000 yuan / h.

[0115] If the deviation between the output of each pulverizing system and the average total output is required to be no more than 5 t / h, then additional conditions need to be added and the solution recalculated.

[0116] |x1-(x1+x2+x3+x4+x5) / 5|<5,

[0117] |x2-(x1+x2+x3+x4+x5) / 5|<5,

[0118] |x3-(x1+x2+x3+x4+x5) / 5|<5,

[0119] |x4-(x1+x2+x3+x4+x5) / 5|<5,

[0120] |x5-(x1+x2+x3+x4+x5) / 5|<5,

[0121] A new interpretation can be obtained.

[0122] The output of pulverizing system A is 50.651 t / h.

[0123] The output of pulverizing system B is 42.067 t / h.

[0124] The output of the C-type pulverizing system is 42.067 t / h.

[0125] The output of the D pulverizing system is 52.067 t / h.

[0126] The output of the E pulverizing system is 48.48 t / h.

[0127] The weighted calorific value is 21.154, the weighted sulfur content is 0.9411, the weighted ash content is 21.149, the weighted ash melting point is 1220℃, and the price is 139,050 yuan / h > 136,050 yuan / h, which increases the cost by nearly 3,000 yuan / h.

[0128] It is evident that, compared to manual blending, the equations can be solved much faster through program calculations.

[0129] S5: Implementation and Evaluation Steps

[0130] Based on the user-selected coal blending scheme, the recommended capacity of each coal bunker is calculated, and the final coal blending instruction is generated. Actual operating data is collected, and the deviation between the actual and planned calorific values ​​is calculated. The results are as follows: Figure 6 As shown, the actual calorific value is calculated based on the main steam flow rate and the actual coal quantity. According to the time offset algorithm, the deviation is attributed to the corresponding fuel operation team, and a performance evaluation report is generated, as shown in the table below.

[0131] Table 8

[0132] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for calculating calorific value based on energy-saving requirements, characterized in that: include, Based on historical operating data, a correlation model between main steam flow and electrical load and thermal load is established. Based on the electrical load and thermal load in the final load forecast, the required main steam flow for the forecast period is calculated, and fuel output is controlled according to the results. The minimum calorific value is dynamically calculated based on the main steam flow rate, the maximum output of the pulverizing system, the enthalpy of the main steam, and the rated enthalpy under operating conditions. Construct a multi-constraint mixed integer programming model with the objective function of minimizing fuel cost, and solve for multiple fuel configuration schemes based on the lowest calorific value.

2. The calorific value calculation method based on energy-saving demand as described in claim 1, characterized in that: The correlation model uses polynomial fitting; the fitting parameters are automatically updated daily based on the data of that day. The automatic update is triggered by the following conditions: daily timed triggering or when the electrical load of day D-1 is lower than 30% of the rated load, the model parameters of day D-2 are used and the update of that day is skipped, where D represents the day.

3. The calorific value calculation method based on energy-saving demand as described in claim 2, characterized in that: The final load forecast is determined by comparing the manual forecast and the model forecast. When the deviation between the two exceeds 20%, the manual forecast is taken as the final load forecast; otherwise, the model forecast is taken as the final load forecast. The model prediction value includes the prediction of the AGC scheduling value on day D+1 based on the Long Short-Term Memory Neural Network (LSTM), using the AGC scheduling value, planned value, and time characteristics of the past few days, as well as the planned value and time characteristics of D+1.

4. The method for calculating calorific value based on energy-saving demand as described in any one of claims 1 to 3, characterized in that: The formula for calculating the minimum calorific value of the fuel fed into the furnace is as follows: in, Where F is the minimum calorific value, k is the empirical coefficient for boiler thermal efficiency, and F is the minimum calorific value. ms Main steam flow rate, M max For the pulverizing system to operate at maximum capacity, h actual For the main steam enthalpy under variable load conditions, h rated This is the enthalpy value of the main steam under rated operating conditions.

5. The calorific value calculation method based on energy-saving demand as described in claim 1, characterized in that: The decision variables of the objective function include the fuel type selection of each pulverizing system and the output of the output unit.

6. The calorific value calculation method based on energy-saving demand as described in claim 5, characterized in that: The constraints of the multi-constraint mixed integer programming model include the output limit of the pulverizing system and the energy quality limit of the furnace feed. The energy quality limit of the furnace feed includes the lower limit of calorific value, the upper limit of sulfur content, the upper limit of ash content, and the lower limit of ash melting point.

7. The calorific value calculation method based on energy-saving requirements as described in claim 5 or 6, characterized in that: The constraints of the multi-constraint mixed integer programming model also include the output balance constraint of the output unit: ∣x i − ∣≤Δmax, Where, x i For the output of the i-th power unit, Δmax represents the average output of each output unit, and Δmax is the preset maximum allowable deviation.

8. The calorific value calculation method based on energy-saving demand as described in claim 7, characterized in that: The process involves solving for and outputting multiple fuel configuration schemes based on the lowest calorific value, including: Based on the available fuel types configured for each pulverizing system, calculate the number of all possible fuel type combinations: ， Where, N comb Let n be the number of fuel type combinations, and n be the number of pulverizing systems in operation. The number of available fuel types for the i-th pulverizing system; For each fuel type combination, with the fuel type selection variable fixed, solve the linear programming problem concerning the output of the coal feeder, eliminate combinations that do not meet the constraints, arrange them by cost, and provide energy configuration schemes.

9. The calorific value calculation method based on energy-saving demand as described in claim 8, characterized in that: It also includes virtual fuel type management steps: Receive user input of two types of raw fuel and their blending ratio α and Where 0 < α < 1; automatically calculate the industrial analysis parameters of the virtual blended fuel type, including the received lower heating value, total moisture, received total sulfur, received ash, received volatile matter, received fixed carbon and ash melting point temperature, and each parameter is calculated by weighted average according to the blending ratio; when the fuel scheme is output, the virtual blended fuel is disassembled and displayed as the original combination form.

10. A calorific value calculation system based on energy-saving requirements, utilizing the method described in any one of claims 1-3, 5-6, and 8-9, characterized in that: include, The load forecasting module receives 96-point planned power output curves of the power grid, new energy load forecast data, and planned outage information of generating units. The load curve for day D+1 is predicted based on an LSTM neural network and corrected by manual experience rules; the load is allocated to each operating unit according to the unit's shutdown status. The main steam flow calculation module stores and updates the polynomial fitting model of main steam flow with electrical load and heat load daily; it calculates the required main steam flow based on the predicted load. The dynamic calorific value calculation module stores the sliding pressure operation curve of the storage unit and the calculation formula of IAPWS-IF97. Determine the main steam pressure and calculate the corresponding enthalpy value based on the load rate; The minimum calorific value of the fuel entering the furnace is calculated based on the main steam flow rate, the maximum output of the pulverizing system, and the enthalpy ratio. The multi-constraint optimization solution module stores a database of fuel type information, available fuel type configurations, pulverizing system output limits, and fuel quality limits for the furnace; it constructs and solves a mixed-integer programming model with the goal of minimizing fuel costs; and generates and sorts multiple alternative fuel configuration schemes. The scheme management module receives the configuration scheme selected by the user and generates the final fuel configuration instruction containing the suggested stockpiling amount; Manage the generation, calculation, and deletion of virtual blended fuels, wherein the virtual blended fuels are calculated by weighting two original fuels according to their configuration ratio to obtain industrial analysis parameters; The evaluation and feedback module collects actual operating data from the SIS system, calculates the deviation rate between the planned calorific value and the actual calorific value, and assigns the deviation to the responsible work team based on the time offset algorithm, generating performance evaluation data. The time offset algorithm includes: setting a fixed time offset Δt, attributing the deviation data of the actual furnace feeding time to the fuel operation team before time Δt; dividing 24 hours into multiple shift periods according to the power plant shift schedule; and calculating the average deviation rate of each shift's responsible period as the monthly performance indicator of that shift.

Images