Source-load dispatching method considering electric smelting magnesium load regulation and wind farm power distribution

Abstract

The application discloses a source-load scheduling method considering electric smelting magnesium load regulation and wind power distribution, introduces electric smelting magnesium high-load energy load with fast regulation characteristics on the load side, participates in power grid optimal scheduling together with thermal power plants, takes the maximum system wind power consumption and the minimum sum of average variances of power fluctuations at all wind power plant busbars as objective functions, considers relevant constraints, and solves by using a genetic algorithm; the system wind power curtailment, wind power fluctuation and other aspects under different cases are compared and analyzed, and the effectiveness of the low-carbon optimal scheduling model considering electric smelting magnesium load regulation and wind power distribution is proved, the system new energy consumption can be improved, and the new energy power fluctuation can be reduced. Finally, the actual load curve of each wind turbine after smoothing is obtained. The method can reduce the wind power curtailment, realize the wind power consumption, and smooth the output curve of each wind power plant; the two objective functions echo each other, so that the calculation of each unit is more reasonable.

Description

Technical Field

[0001] This invention belongs to the field of source-cooperative optimization scheduling technology, specifically involving a source-load scheduling method that considers the regulation of fused magnesium load and the power allocation of wind farms. Background Technology

[0002] With the rapid increase in wind power installed capacity, large-scale wind power grid connection has become an inevitable trend. As of the end of November 2022, my country's grid-connected wind power installed capacity reached 350 million kilowatts. The national annual wind power generation reached 614.5 billion kilowatt-hours, and the growth rate of wind power installed capacity reached 20.9%. However, the volatility, randomness, and anti-peak-shaving characteristics of wind power output have led to a growing gap between the system's load power and wind power output as the grid-connected wind power capacity increases, resulting in problems such as voltage quality degradation and frequent frequency fluctuations.

[0003] Traditional dispatching methods rely on frequent start-ups and shutdowns of thermal power plants to accommodate fluctuations in wind power output. However, when wind power output fluctuates significantly, system stability is achieved at the cost of wind curtailment, resulting in substantial waste of wind power resources. When wind power output is limited, dispatching needs to consider the smoothing effect of multiple wind farms overlapping, optimizing the total wind power allocation. Therefore, developing an optimal power dispatching scheme that absorbs wind power and optimizes the total wind power allocation to achieve a smoothing effect is a better choice. Summary of the Invention

[0004] The purpose of this invention is to provide a source-load scheduling method that takes into account the load regulation of fused magnesium and the power distribution of wind farms, which can reduce wind curtailment and smooth the power fluctuations of each wind farm.

[0005] The technical solution adopted in this invention is a source-load scheduling method that takes into account the load regulation of fused magnesium and the power distribution of wind farms, and is implemented according to the following steps:

[0006] Step 1: Obtain the predicted total wind power and fused magnesium load for the typical load day of the target wind farm;

[0007] Step 2: With the goal of minimizing the system's wind curtailment, establish an upper-level model. Input the predicted total wind power and the predicted fused magnesium load into the upper-level model to obtain the actual total wind power and the actual fused magnesium load.

[0008] Step 3: With the goal of minimizing the sum of the average variances of power fluctuations at all wind farm busbars throughout the day, establish a lower-level model, input the actual total wind power value into the lower-level model for allocation, and obtain the actual power output curve of each wind farm.

[0009] Step 4: Perform wind power dispatch according to the actual fused magnesium load value and the actual output curve of each wind farm.

[0010] The invention is further characterized by:

[0011] Step 2, the upper-level model includes the upper-level objective function and upper-level constraints, where the upper-level objective function is:

[0012]

[0013] In the formula: T is the number of time periods in the scheduling cycle; N W P represents the number of wind farms. t Out P represents the total active power output of all wind farm busbars during time period t; t G P represents the total active power output of a thermal power plant during time period t. t L N represents the active power of the system's normal load during time period t; Mg P represents the number of fused magnesium furnaces. t Mg，k Let ΔP be the active power of the k-th electric fused magnesium furnace during time period t. t Mg，k The adjustable power of a single electric fused magnesium furnace during time period t;

[0014] The upper-level constraints include system power balance constraints, thermal power plant operation constraints, upper and lower limits of wind power output constraints, and upper and lower limits of power regulation for high-energy-consuming loads of fused magnesium.

[0015] The upper-level constraints are as follows:

[0016] 1) System power balance constraints:

[0017]

[0018] 2) Upper and lower limits of wind power output constraints:

[0019]

[0020] In the formula: P t Out-old The total active power output of the wind farm at the busbar during time period t is predicted.

[0021] 3) Operating constraints of thermal power plants:

[0022] ① Output power upper and lower limit constraints:

[0023]

[0024] ② Climbing speed constraint:

[0025]

[0026]

[0027] In the formula: P t-1G The total active power output of all thermal power plants during time period t-1; P G，up and P G，down These represent the total uphill and downhill ramp rates of all thermal power plants, respectively.

[0028] 4) Upper and lower limits of power regulation for high-energy-consuming electrofused magnesium:

[0029]

[0030] Where: ΔP t Mgmax and ΔP t Mgmin These are the upper and lower limits of the power regulation for high-energy-consuming electrofused magnesium.

[0031] The lower-level model includes a lower-level objective function and lower-level constraint functions, where the lower-level objective function is:

[0032]

[0033] Among them, P t W，j Let P be the output of the j-th wind farm at time t. t Battery，cj P represents the charging power of the energy storage device in the j-th wind farm at time t. t Battery，dj P represents the discharge power of the energy storage device at the j-th wind farm station at time t. Out，jav Let N be the average power fluctuation at the j-th wind farm bus within one period. W The number of wind farms;

[0034] The lower-level constraint functions include power balance constraints for wind farms, balance equation constraints for energy storage device charge and discharge quantities, upper and lower limit constraints for energy storage device charge and discharge power, and upper and lower limit constraints for energy storage device state of charge.

[0035] The method for calculating the average power fluctuation at the j-th wind farm bus within a cycle is as follows:

[0036]

[0037] In the formula: R W This represents the sum of the average variances of power fluctuations at all wind farm busbars throughout the day.

[0038] The lower-level constraint function is as follows:

[0039] 1) Power balance constraints of wind farms:

[0040]

[0041] 2) Constraints of the energy storage device's charge and discharge balance equation:

[0042]

[0043] In the formula: T is the total number of time periods in a day, and its value is set to 96;

[0044] 3) Upper and lower limits of charging and discharging power constraints for energy storage devices:

[0045] ① Charging power upper and lower limit constraints:

[0046]

[0047] ② Upper and lower limits of discharge power constraints;

[0048]

[0049] In the formula: P Batterycjmax P represents the upper limit of the charging output of the energy storage device in the j-th wind farm. Batterydjmax This represents the upper limit of the discharge output of the energy storage device in the j-th wind farm station;

[0050] 4) Upper and lower limits of the state of charge of energy storage devices:

[0051]

[0052] In the formula: Let SOC be the state of charge (SOC) of the energy storage device at time t in the j-th wind farm. max The upper limit of the state of charge (SOC) of an energy storage device. min This is the limit of the state of charge of the energy storage device.

[0053] The beneficial effects of this invention are:

[0054] This invention relates to a source-load dispatching method that considers load regulation of fused magnesium and power distribution in wind farms. This method can improve the absorption of new energy sources in the system and reduce power fluctuations of new energy sources. The dispatching method of this invention is of great significance for developing green electricity and implementing a sustainable development strategy. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the principle of the electric fused magnesium furnace (FMF) of this invention;

[0056] Figure 2 This is a schematic diagram illustrating the principle of high-energy-consuming loads of fused magnesium participating in wind power consumption according to the present invention;

[0057] Figure 3 This is a schematic diagram of the source-load coordination system structure of the present invention, which takes into account the load regulation of fused magnesium and the power distribution of wind farms.

[0058] Figure 4This is a schematic diagram of the source-load scheduling method of the present invention, which takes into account the load regulation of fused magnesium and the power distribution of wind farms;

[0059] Figure 5 This is a graph showing the conventional load forecast power and wind power forecast power curves of a verification example of the present invention;

[0060] Figure 6 This is a power curve diagram of fused magnesium high-energy-load load after regulation in a verification example of the present invention;

[0061] Figure 7 This is a comparison chart of wind power output under two schemes in the verification example of this invention;

[0062] Figure 8 These are wind curtailment diagrams under two schemes in the verification examples of this invention.

[0063] Figure 9 This is a battery energy storage charging and discharging power diagram of a verification example of the present invention;

[0064] Figure 10 These are the power curves of two wind farms under two different schemes for allocating total wind power. Detailed Implementation

[0065] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0066] This invention models the high-energy-consuming load of fused magnesium based on its load characteristics, and analyzes the principle of wind power absorption by fused magnesium.

[0067] Due to the unique characteristics of magnesite, such as low grade, complex mineral composition, and large compositional variations, the production of high-purity fused magnesia primarily utilizes a domestically unique three-phase AC electric fused magnesia furnace (FMF). The FMF controls the current value and arc intensity of the three-phase electrodes by appropriately adjusting the distance between the three-phase electrodes and the molten pool surface. Magnesite raw material, with MgCO3 as its main component, is melted by the heat released by the electric arc in the FMF, and then solidifies into the final product, MgO. A schematic diagram of the FMF working principle is shown below. Figure 1 1 refers to the step-down transformer, 2 to the sensor, 3 to the voltage transformer, 4 to the traction motor, 5 to the liftable graphite electrode, 6 to the electric arc, 7 to the molten pool, 8 to the turntable, and 9 to the packing device.

[0068] The production process begins with magnesite raw material being fed into the furnace via a filling device and conveyor belt. Simultaneously, power is supplied to the fused magnesia furnace through a step-down transformer, sensors, and voltage transformers. Then, the furnace's control system adjusts the speed and direction of the traction motor to raise and lower the electrodes, creating an electric arc between the electrodes and the raw material. The raw material melts by absorbing the heat released by the arc, forming a molten pool below the electrodes. The smelting process ends when the upper surface of the molten pool reaches the furnace opening. Finally, a rotary table removes the furnace from the smelting station for cooling and processing, yielding the fused magnesia product.

[0069] The smelting process in an electric fused magnesium furnace consumes a large amount of active power, making it a high-energy-consuming load. The active power consumed by the furnace can be controlled by adjusting its operating current, meaning the load of the fused magnesium furnace is adjustable. In summary, the electric fused magnesium furnace possesses characteristics such as power adjustability and speed of operation.

[0070] The active power of high-energy-consuming loads in fused magnesium is:

[0071]

[0072] In the formula: P t Mg，k U represents the operating power of the k-th fused magnesium furnace at time t; t Mg，k and I t Mg，k Let be the current value of the smelting voltage of the kth electric fused magnesium furnace at time t; This refers to the power factor of the power supply system. During the smelting process of fused magnesium furnaces, to avoid furnace blowout accidents due to excessive power, the active power of each fused magnesium furnace must not be too high. Simultaneously, to ensure the quality of the produced fused magnesium products, the active power of each fused magnesium furnace must not be too low. The constraints are as follows:

[0073]

[0074]

[0075] In the formula: P t Mg base N represents the normal operating power of a single fused magnesium furnace at time t. Mg N represents the total number of fused magnesium furnaces operating within the fused magnesium enterprise. Mgadj P represents the total number of adjustable fused magnesium furnaces within the fused magnesium enterprise. Mg up and P Mg down This indicates the margin for adjusting the power of a single electric fused magnesia furnace upwards and downwards; S t Mg，kS represents the state variable representing the upward adjustment of the power of the k-th electric molten magnesium furnace at time t. t Mg，k =1 indicates that the active power of the k-th electric molten magnesium furnace is increased at time t, S t Mg，k =0 indicates that the active power of the k-th electric fused magnesium furnace is reduced at time t; P Mg，kmax and P Mg，kmin Let be the maximum and minimum active power of the k-th electric fused magnesium furnace.

[0076] Analysis of the principle of high-energy-consuming loads of fused magnesium participating in wind power consumption:

[0077] Because wind power output is random and fluctuating, the constant changes in wind power output increase the regulation pressure on thermal power plants. When the regulation of thermal power plants cannot meet the regulation needs of wind power and load, the power supply and demand of the system can be balanced by scheduling high-energy-consuming loads of fused magnesium, thereby improving the wind power absorption capacity and reducing wind curtailment. Figure 2 A schematic diagram illustrating the principle of high-energy-consuming loads of fused magnesium participating in wind power consumption. Figure 2 In the middle, P G,max and P G,min These represent the maximum and minimum output of a thermal power plant, respectively. The adjustable range of thermal power plant output lies between the two black lines. From the perspective of grid operation, whether wind power output can be fully absorbed by the grid depends on whether the difference between the load and the minimum output of thermal power is sufficient, i.e., the grid connection capacity for wind power generation (the capacity to absorb wind power). Figure 2 The portion between the green curve and the black line below it. Adding the minimum output of thermal power to the output of wind power yields the minimum adjustable output of thermal power including wind power, such as... Figure 2 As shown by the yellow curve in the middle.

[0078] Depend on Figure 2 It is evident that the off-peak load period is the time of day with the lowest wind power absorption capacity, and also the main period for wind power curtailment (i.e., periods t1-t2 and t3-t4). When wind power output exceeds the regulation capacity of thermal power plants, if the grid accepts all of this wind power, the thermal power plants will be forced to make deep adjustments, or even shut down. To ensure the safety and economy of system operation, wind curtailment occurs. Taking the wind curtailment process during period t1-t2 as an example, the amount of wind power curtailed is:

[0079]

[0080] In the formula: ΔT is the duration of time interval t. L Power consumed by normal load; P W For wind power output. Taking the wind curtailment process during the period t1 to t2 as an example, after the high-energy-consuming load of fused magnesium electrolytic capacitors participates in regulation, the regulation power of the high-energy-consuming load of fused magnesium electrolytic capacitors during the period t1 to t2 is ΔP. Mg At this time, the amount of air curtailed is:

[0081]

[0082] Since wind power output is at its maximum during the time periods t1~t2, t3~t4, and t6~t7, and the minimum thermal power output curve including wind power is higher than the load power curve, it is necessary to adjust the active power of the high-energy-consuming load of fused magnesium electrolytic capacitors upward, i.e., ΔP. Mg >0, by Figure 1 It can be seen that after the high-energy-consuming load of fused magnesium participates in the regulation, the total system load (i.e., the purple curve in the figure) increases, and the wind curtailment area shrinks from Region II to Region I. Therefore, the wind power curtailment is less than the wind power curtailment under the traditional dispatch mode, i.e., E ’ abon <E abon Taking the wind power absorption during the period t1 to t2 as an example, the increased wind power absorption capacity of the system during this period is:

[0083]

[0084] In the formula: ΔT is the duration of time interval t. L Power consumed by normal load; P W It provides power for wind power.

[0085] Therefore, by adjusting the high-energy-consuming load of fused magnesium to absorb obstructed wind power, the area of ​​obstructed wind power is reduced, effectively improving the system's ability to absorb obstructed wind power. Similarly, during the periods t2-t3 and t4-t5, wind power output is at a relatively low value, and the minimum thermal power output curve including wind power is lower than the load power curve. Therefore, it is not necessary to adjust the size of the high-energy-consuming load of fused magnesium to absorb wind power. However, during the period t5-t6, wind power output is at its minimum. The fluctuation of wind power output can be balanced by adjusting the active power of the high-energy-consuming load of fused magnesium downward, thus alleviating the regulation pressure on thermal power plants.

[0086] This invention relates to a source-load scheduling method that considers load regulation of fused magnesium and power distribution in wind farms, such as... Figure 4 As shown, please follow these steps:

[0087] Obtain the predicted total wind power and fused magnesium load values ​​for typical load days of the target wind farm; with the goal of minimizing wind curtailment, establish an upper-level model, and input the predicted total wind power and fused magnesium load values ​​into the upper-level model to obtain the actual total wind power and actual fused magnesium load values.

[0088] The upper-level model includes an upper-level objective function and upper-level constraints, where the upper-level objective function is:

[0089]

[0090] In the formula: T is the number of time periods in the scheduling cycle; N WP represents the number of wind farms. t Out P represents the total active power output of all wind farm busbars during time period t; t G P represents the total active power output of a thermal power plant during time period t. t L N represents the active power of the system's normal load during time period t; Mg P represents the number of fused magnesium furnaces. t Mg，k Let ΔP be the active power of the k-th electric fused magnesium furnace during time period t. t Mg，k The adjustable power of a single electric fused magnesium furnace during time period t;

[0091] The upper-level constraints include system power balance constraints, thermal power plant operation constraints, upper and lower limits of wind power output constraints, and upper and lower limits of power regulation for high-energy-consuming loads of fused magnesium.

[0092] The upper-level constraints are as follows:

[0093] 1) System power balance constraints:

[0094]

[0095] 2) Upper and lower limits of wind power output constraints:

[0096]

[0097] In the formula: P t Out-old The total active power output of the wind farm at the busbar during time period t is predicted.

[0098] 3) Operating constraints of thermal power plants:

[0099] ① Output power upper and lower limit constraints:

[0100]

[0101] ② Climbing speed constraint:

[0102]

[0103]

[0104] In the formula: P t-1 G The total active power output of all thermal power plants during time period t-1; P G，up and P G，down These represent the total uphill and downhill ramp rates of all thermal power plants, respectively.

[0105] 4) Upper and lower limits of power regulation for high-energy-consuming electrofused magnesium:

[0106]

[0107] Where: ΔP t Mgmax and ΔP t Mgmin These are the upper and lower limits of the power regulation for high-energy-consuming electrofused magnesium.

[0108] With the goal of minimizing the sum of the average variances of power fluctuations at all wind farm busbars throughout the day, a lower-level model is established. The actual total wind power value is input into the lower-level model for allocation, resulting in the actual power output curve of each wind farm.

[0109] The lower-level model includes a lower-level objective function and lower-level constraint functions, where the lower-level objective function is:

[0110]

[0111] Among them, P t W，j Let P be the output of the j-th wind farm at time t. t Battery，cj P represents the charging power of the energy storage device in the j-th wind farm at time t. t Battery，dj P represents the discharge power of the energy storage device at the j-th wind farm station at time t. Out，j av Let N be the average power fluctuation at the j-th wind farm bus within one period. W Let be the number of wind farms; the average power fluctuation at the busbar of the j-th wind farm within one period is calculated as follows:

[0112]

[0113] In the formula: R W This represents the sum of the average variances of power fluctuations at all wind farm busbars throughout the day.

[0114] The lower-level constraint functions include power balance constraints for wind farms, balance constraints for energy storage device charge / discharge, upper and lower limits for energy storage device charge / discharge power, and upper and lower limits for energy storage device state of charge. Specifically:

[0115] 1) Power balance constraints of wind farms:

[0116]

[0117] 2) Constraints of the energy storage device's charge and discharge balance equation:

[0118]

[0119] In the formula: T is the total number of time periods in a day, and its value is set to 96;

[0120] 3) Upper and lower limits of charging and discharging power constraints for energy storage devices:

[0121] ③ Charging power upper and lower limit constraints:

[0122]

[0123] ④ Upper and lower limits of discharge power constraints;

[0124]

[0125] In the formula: P Batterycjmax P represents the upper limit of the charging output of the energy storage device in the j-th wind farm. Batterydjmax This represents the upper limit of the discharge output of the energy storage device in the j-th wind farm station;

[0126] 4) Upper and lower limits of the state of charge of energy storage devices:

[0127]

[0128] In the formula: Let SOC be the state of charge (SOC) of the energy storage device at time t in the j-th wind farm. max The upper limit of the state of charge (SOC) of an energy storage device. min This is the limit of the state of charge of the energy storage device.

[0129] The relationship between the state of charge (SBC) of an energy storage device and its output power is shown in the following formula. The change in electricity during the charging process of the energy storage device is as follows:

[0130]

[0131] The change in battery charge during discharge is as follows:

[0132]

[0133] In the formula, SOC t+1 j Let η be the state of charge of the energy storage device at time t+1 in the j-th wind farm. c and η d These represent the energy storage charging efficiency and discharging efficiency, respectively, with Δt being the charging / discharging time interval. Δt is 15 min, and its value is taken as 0.25 h. C Battery，j Let be the capacity of the energy storage device in the j-th wind farm.

[0134] This invention employs a two-level optimization model, a system optimization model with a hierarchical structure. The lower-level optimization optimizes its objective function based on the given solution from the upper-level decision and feeds the optimization result back to the upper level. The upper-level optimization adjusts its decision variables based on the lower-level optimization result, making a decision that balances the overall interests. The system structure diagram of this invention, which considers source-load coordination for fused magnesium load regulation and wind farm power allocation, is shown below. Figure 3 As shown.

[0135] The system mainly includes multiple thermal power plants, multiple wind farms (containing battery energy storage devices), fused magnesium loads (containing multiple fused magnesium furnaces), and conventional loads. During the transmission of electricity from wind farms to the system, large-scale grid connection of wind power leads to severe wind curtailment. This invention addresses this by introducing high-energy-consuming fused magnesium loads in conjunction with thermal power to regulate wind power absorption, and by establishing an upper-level model to minimize wind curtailment. Furthermore, the battery energy storage devices absorb or release electrical energy according to changes in wind power output, achieving spatiotemporal translation of electrical energy. By smoothing fluctuations in wind power output, the total wind power output is allocated to meet the requirements for wind power stabilization. This leads to the establishment of a lower-level model that minimizes output fluctuations for each wind farm.

[0136] The active power output of wind power at the bus in the system is:

[0137]

[0138] In the formula, P t W，j It is the output power of the j-th wind farm at time t; P is the output power of the battery energy storage device in the j-th wind farm at time t; if it is greater than 0, it means the battery energy storage device is releasing electrical energy; if it is less than 0, it means the battery energy storage device is absorbing electrical energy. t Out，j This represents the output power at the j-th wind farm bus at time t.

[0139] Considering that the optimal scheduling of each wind farm in this invention is based on the total planned wind power output obtained from the upper-level model, the output of each wind farm and the energy storage regulation power within the farm are then calculated. The variance of power fluctuations at the wind farm bus is calculated based on the total wind power output obtained from the upper-level model. Therefore, the upper and lower levels are mutually constrained in the calculation process, and repeated iterations are required to obtain the optimal solution. Thus, this invention employs a two-level optimization model to describe the source-load coordination optimization problem for large-scale wind power integration.

[0140] Wind power dispatch is carried out according to the actual fused magnesium load value and the actual output curve of each wind farm.

[0141] The two-layer model optimization solution method used in this invention:

[0142] The source-load optimization scheduling model established in this invention, which considers the load regulation of fused magnesium and the power allocation of wind farms, is a bilayer mixed integer linear programming (BMILP) problem. BMILP has a two-layer progressive structure, with different objective functions and constraints in each layer, and the upper and lower planning models mutually constrain each other. The mathematical model of the bilayer optimization is shown in equation (24). It is solved using MATLAB, and its standard form is:

[0143]

[0144] Where F(x,y) is the objective function of the upper-level optimization problem; L(x,t) is the objective function of the lower-level optimization problem; x is the variable to be optimized, including the total output of the thermal power plant, the total planned power at the wind power bus, and the high-energy-load regulation power of fused magnesium; y is the start-up and shutdown state of the thermal power plant and the wind farm; t is the lower-level variable; x is the upper-level variable; x θ These are local optimal solution variables obtained from the upper-level model, which need to be passed to the lower-level model for further solution. In this invention, they refer to the total power output of the wind farm. g(x,y)=0 and h(x,y)≤0 are the equality and inequality constraints of the upper-level model; u(x,t)=0 and r(x,t)≤0 are the equality and inequality constraints of the lower-level model. In this invention, the upper-level equality constraints include system power balance constraints and thermal power plant power balance constraints; the upper-level inequality constraints include wind power output constraints and high-energy-consuming load regulation power constraints of fused magnesium; the lower-level equality constraints include wind farm power balance constraints and energy storage device charge / discharge balance equality constraints; and the lower-level inequality constraints include energy storage operation constraints.

[0145] The source-load optimization scheduling model proposed in this invention, which considers fused magnesium load regulation and wind farm power allocation, has an optimal solution. However, since the bi-level programming problem is difficult to solve directly, it is necessary to use a suitable optimization algorithm for iterative approximation to find the optimal solution.

[0146] To verify the effectiveness of the source-load optimization scheduling method for wind power absorption that takes into account the load regulation of fused magnesium proposed in this invention, two schemes with and without fused magnesium regulation are set up for comparative analysis.

[0147] (1) Scheme 1: A two-layer optimization model considering the regulation of high energy loads of thermal power plants and fused magnesium. Based on the wind power forecast data after the smoothing of the battery energy storage device, while fully considering the operating constraints of fused magnesium, the fused magnesium load actively participates in the system's wind power consumption and carbon emission reduction.

[0148] (2) Scheme 2: A two-layer optimization model that relies solely on thermal power plants for regulation. The up and down regulation power of the high-energy-consuming load of fused magnesium is set to 0, and other conditions are consistent with the model proposed in Scheme 1.

[0149] To verify the effectiveness of the proposed lower-level scheduling model for allocating total wind power to achieve smooth wind power distribution in each wind farm, two schemes were compared and analyzed: allocation based on capacity ratio and allocation based on the solution results of the lower-level model.

[0150] Solution steps:

[0151] 1) Input the raw data, including the conventional load forecast power and wind power forecast power curves, such as... Figure 5 Input fused magnesium data. The rated power for high-energy-consuming fused magnesium is 200MW. Based on a field survey of a fused magnesium enterprise in Liaoning Province, high-energy-consuming fused magnesium production typically operates 24 hours a day without interruption. The power can be increased by 20% of the rated power, i.e., 40MW. The power can be decreased by 10% of the rated power, i.e., 20MW. The dispatch cycle is 24 hours. Figure 5 The wind power prediction curve shows significant fluctuations in wind power output, ranging from 90 to 240 MW. Peak output occurs between periods 1 and 19, coinciding with a low point in the conventional load power prediction curve. Furthermore, wind power output is at its lowest point during periods 30-42 and 69-84, when conventional load power is at its peak. This indicates a clear anti-peak characteristic of wind power output, increasing the pressure on thermal power plants and leading to severe wind curtailment. The proposed source-load coordinated two-layer optimization scheduling model, considering load-side fused magnesium load regulation, solves a two-layer mixed integer linear programming problem, with each layer employing a genetic algorithm. In the upper-layer scheduling model, the genetic algorithm has a maximum iteration count of 100, a population size of 500, a mutation probability of 0.8, and a crossover probability of 0.3. Compared to the upper-level scheduling model, the lower-level scheduling model has a significantly reduced number of variables. Therefore, the maximum number of iterations in the lower-level scheduling model is set to 50, the population size to 100, the mutation probability to 0.7, and the crossover probability to 0.3. The model iteratively seeks the optimal solution through a genetic algorithm, achieving comprehensive optimization of the power output of each wind farm, the total power output of the thermal power plant, and the power regulation power of the high-energy-consuming load of fused magnesium. The power curve after the high-energy-consuming load of fused magnesium participates in the regulation is shown below. Figure 6 As shown.

[0152] 2) In Scheme 1, when wind power output is high and the system's conventional load is at its lowest (periods 1-19), thermal power plants operate near their minimum output. To maximize wind power absorption and avoid insufficient downward regulation capacity or even shutdowns at thermal power plants, all high-energy-consuming loads of fused magnesium are increased in power. When wind power output is low and the system's conventional load is at its peak (periods 30-34), thermal power plants increase their output. To avoid wind curtailment and maintain the safe and stable operation of the system, all high-energy-consuming loads of fused magnesium are still increased in power. When wind power output increases and the system's conventional load power decreases (periods 72-89), thermal power plants decrease their output, and all high-energy-consuming loads of fused magnesium are increased in power. When wind power output is at its lowest and the system's conventional load is at its peak (periods 34-42), thermal power plants operate near their maximum output. To reduce the regulation pressure on thermal power plants, all high-energy-consuming loads of fused magnesium are decreased in power. In Scheme 2, due to the lack of high-energy-consuming loads from fused magnesium oxide (FMC) to participate in regulation, the fluctuations in wind power output are entirely borne by thermal power plants. This significantly increases the regulation pressure on thermal power plants, which are not operating at their minimum output levels. This not only increases the system's operating costs but also leads to a severe deficiency in the regulation capacity of thermal power plants, ultimately forcing wind curtailment. The wind power output and curtailment amounts for the two schemes are as follows: Figure 7 and Figure 8 As shown. Figure 7 This demonstrates the planned wind power output under two different scenarios, from... Figure 8 As can be seen from the data, when using Scheme 1, there are 2 periods when the planned wind power output is restricted, the maximum restricted power is 9.66MW, the wind curtailment rate is 0.11%, and the total restricted electricity is 64.48MW·h. When using Scheme 2, the planned wind power output is restricted in 91 out of 96 periods per day, the maximum restricted power is 177.16MW, the wind curtailment rate is 31.90%, and the total restricted electricity is 19193.51MW·h. Figure 8 The study demonstrates the wind curtailment situation under two schemes. Under Scheme 1, due to the participation of fused magnesium load in regulation, the system's wind curtailment rate is 0.11%, a reduction of 31.79% compared to Scheme 2, and the amount of wind curtailed is reduced by 99.67%. In summary, by adopting the source-load coordinated two-layer optimization scheduling model of this invention, which considers load-side fused magnesium load regulation, the wind power absorption level is significantly improved, and the amount of wind curtailed is effectively reduced.

[0153] 3) The charging and discharging power of the battery energy storage device in the wind farm obtained from the lower-level model is as follows: Figure 9 As shown. From Figure 9 As can be seen, battery energy storage has a longer charging / discharging duration and higher charging / discharging power. The operating conditions of the two wind farms obtained from the two optimization schemes are as follows: Figure 10 As shown. Combined with Figure 9 and Figure 10Analysis shows that, for example, during the rapid increase in wind power between periods 44 and 48 in Schemes 1 and 2, battery energy storage will operate in a charging state during this period, effectively mitigating the rise in wind power. Conversely, during periods 12 and 15 in Schemes 1 and 2, wind power decreases significantly, and battery energy storage will operate in a discharging state during this period to prevent excessively low grid-connected wind power. Furthermore, the charging / discharging duration of the battery energy storage device is relatively short, allowing for frequent charging / discharging throughout the entire dispatch cycle to smooth out high-frequency random fluctuations in wind power.

[0154] Through the above-described method, this invention presents a source-load scheduling method that considers fused magnesium load regulation and wind farm power allocation. First, it introduces a high-energy-consuming fused magnesium load with rapid adjustment characteristics on the load side, participating in grid optimization scheduling alongside thermal power plants. The objective function is to maximize wind power absorption and minimize the sum of the average variances of power fluctuations at all wind farm busbars. Considering relevant constraints, a genetic algorithm is used to solve the problem. Finally, comparative analyses of system wind curtailment and wind power fluctuations under different cases demonstrate the effectiveness of the proposed low-carbon optimization scheduling model considering fused magnesium load regulation and wind farm power allocation. This model can improve the system's renewable energy absorption and reduce renewable energy power fluctuations. Ultimately, the smoothed load curves of each wind turbine unit are obtained. This method can both reduce wind curtailment and achieve wind power absorption, and smooth the output curve of each wind farm. The two objective functions complement each other, making the calculations of each unit more reasonable. This is of great significance for developing green electricity and implementing a sustainable development strategy.

Claims

1. A source-load scheduling method considering fused magnesium load regulation and wind farm power distribution, characterized in that, The specific steps are as follows: Step 1: Obtain the predicted total wind power and fused magnesium load for the typical load day of the target wind farm; Step 2: With the goal of minimizing the system's wind curtailment, establish an upper-level model. Input the predicted total wind power and the predicted fused magnesium load into the upper-level model to obtain the actual total wind power and the actual fused magnesium load. The upper-level model mentioned in step 2 includes an upper-level objective function and upper-level constraints, wherein the upper-level objective function is: （1） In the formula: T The number of time periods in the scheduling cycle; N W The number of wind farms; P t Out For all wind farm busbars at t There is always effort and contribution during each period; P t G For thermal power plants t There is always effort and contribution during each period; P t L For the system's normal load at t Active power during a given time period; N Mg For the number of fused magnesium furnaces, P t Mg，k For the first k Taiwan Electric Magnesium Melting Furnace t Active power during a time period P t Mg，k For a single electric fused magnesia furnace t Adjustment power during different time periods; The upper-level constraints include system power balance constraints, thermal power plant operation constraints, upper and lower limits of wind power output constraints, and upper and lower limits of power regulation for high-energy-consuming loads of fused magnesium. Step 3: With the goal of minimizing the sum of the average variances of power fluctuations at all wind farm busbars throughout the day, establish a lower-level model, input the actual total wind power value into the lower-level model for allocation, and obtain the actual power output curve of each wind farm. The lower-level model includes a lower-level objective function and lower-level constraint functions, wherein the lower-level objective function is: （8） in, P t W，j for t Time of the first j The power output of each wind farm P t Battery，cj for t Time of the first j The charging power of the energy storage device in each wind farm station; P t Battery，dj for t Time of the first j Discharge power of energy storage devices within a wind farm station; P Out，j av For the first time in a period j Average power fluctuation at the busbar of each wind farm N W The number of wind farms; The lower-level constraint functions include power balance constraints for wind farms, balance equation constraints for energy storage device charge and discharge quantities, upper and lower limit constraints for energy storage device charge and discharge power, and upper and lower limit constraints for energy storage device state of charge. Step 4: Perform wind power dispatch according to the actual fused magnesium load value and the actual output curve of each wind farm.

2. The source-load scheduling method considering fused magnesium load regulation and wind farm power allocation according to claim 1, characterized in that, The specific upper-level constraints are as follows: 1) System power balance constraints: (2) 2) Upper and lower limits of wind power output constraints: (3) In the formula: P t Out-old For wind farms in t Total active power output at the busbar during the time period; 3) Operating constraints of thermal power plants: ① Output power upper and lower limit constraints: (4) ② Climbing speed constraint: (5) (6) In the formula: P t-1 G For all thermal power plants t- Total effort output in time period 1; P G，up and P G，down These represent the total uphill and downhill ramp rates of all thermal power plants, respectively. 4) Upper and lower limits of power regulation for high-energy-consuming electrofused magnesium: (7) In the formula: P t Mgmax and P t Mgmin These are the upper and lower limits of the power regulation for high-energy-consuming electrofused magnesium.

3. The source-load scheduling method considering fused magnesium load regulation and wind farm power allocation according to claim 1, characterized in that, Within a period of time j The method for calculating the average power fluctuation at the busbar of a wind farm is as follows: （9） In the formula: R W This represents the sum of the average variances of power fluctuations at all wind farm busbars throughout the day.

4. The source-load scheduling method considering fused magnesium load regulation and wind farm power allocation according to claim 1, characterized in that, The lower-level constraint function is specifically: 1) Power balance constraints of wind farms: (10) 2) Constraints of the energy storage device's charge and discharge balance equation: (11) In the formula: T The total number of time periods within a day is set to 96. 3) Upper and lower limits of charging and discharging power constraints for energy storage devices: ① Charging power upper and lower limit constraints: (12) ② Upper and lower limits of discharge power constraints; (13) In the formula: P Batterycjmax For the first j The maximum charging output of energy storage devices within a wind farm site P Batterydjmax For the first j The upper limit of discharge output of energy storage devices in a wind farm station; 4) Upper and lower limits of the state of charge of energy storage devices: (14) In the formula: SOC t j For the first j Energy storage device in a wind farm t State of charge at time t, SOC max This represents the upper limit of the state of charge of the energy storage device. SOC min This is the limit of the state of charge of the energy storage device.

Images