Source-load coordinated scheduling method based on wind power fluctuation characteristics and electrical smelting furnace modeling

Abstract

The invention provides a source-load coordinated scheduling method based on wind power fluctuation characteristics and electrical smelting furnace modeling, and relates to the technical field of power system scheduling, the source-load coordinated scheduling method based on wind power fluctuation characteristics and electrical smelting furnace modeling comprises the following steps: obtaining a technological process of an electrical smelting furnace, analyzing the operation characteristics of the electrical smelting furnace for magnesia, and constructing a mathematical model of the electrical smelting furnace for magnesia according to an analysis result; according to the electrical smelting magnesia furnace mathematical model, a double-layer optimization model is constructed, the double-layer optimization model is solved through an iterative approximation method, and a final decision variable is obtained; and based on the obtained final decision variable, carrying out validity verification through a preset case library. The invention provides an optimized source-load coordinated dispatching double-layer optimization model by combining the volatility of wind power and the established load model of the electric smelting furnace for magnesia, and aims to improve the wind power consumption capability, minimize the operation cost of a power system and verify the effectiveness and rationality of the provided model in a specified scheme example.

Description

Technical Field

[0001] The present invention relates to the technical field of power system dispatching, and in particular to a source-load coordinated dispatching method based on wind power fluctuation characteristics and electric molten magnesium furnace modeling. Background Art

[0002] With the rapid development of society, carbon emissions in countries around the world continue to rise, and reducing carbon emissions has become a focus of common concern in the international community. In this context, the problem of insufficient absorption of renewable energy and excessive carbon emissions in integrated energy systems has become increasingly prominent, becoming an important challenge that needs to be solved urgently. In order to effectively meet this challenge, we need to deeply explore the scheduling potential in each link of the coupling of multiple energy sources such as electricity, heat, and gas, and formulate scientific and reasonable scheduling plans through innovative technical means and management strategies. This can not only promote the full utilization of renewable energy with environmental protection effects and improve energy utilization efficiency, but also promote the low-carbon operation of integrated energy systems and contribute to global sustainable development. Therefore, research on the problem of renewable energy absorption and carbon emissions in integrated energy systems has become a hot topic and focus in the current energy field.

[0003] With the rapid development of renewable energy, especially wind power, the grid-connected capacity of wind power has been increasing. The randomness, volatility and uncertainty of wind power output have increased the peak load regulation pressure of thermal power units, bringing challenges to the control and operation costs of the power system.

[0004] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0005] In view of this, the present invention provides a source-load coordinated scheduling method based on wind power fluctuation characteristics and electric molten magnesium furnace modeling to solve the above-mentioned problem of increased peak-shaving pressure on thermal power units caused by the randomness, volatility and uncertainty of wind power output.

[0006] In order to solve the above problems, the specific technical solutions adopted by the present invention are as follows:

[0007] The source-load coordinated scheduling method based on wind power fluctuation characteristics and electric molten magnesium furnace modeling includes the following steps:

[0008] S1. Obtain the process flow of the fused magnesium furnace, analyze the operating characteristics of the fused magnesium furnace and construct a mathematical model of the fused magnesium furnace according to the analysis results;

[0009] S2. According to the mathematical model of the fused magnesium furnace, a double-layer optimization model is constructed, and the double-layer optimization model is solved by an iterative approximation method to obtain the final decision variables;

[0010] S3. Based on the final decision variables obtained, the validity is verified through the preset case library.

[0011] Preferably, obtaining the process flow of the fused magnesium furnace, analyzing the operating characteristics of the fused magnesium furnace and constructing a mathematical model of the fused magnesium furnace according to the analysis results comprises the following steps:

[0012] S11. According to the process flow of the fused magnesium furnace, the production stages of the fused magnesium furnace are divided, and the production stages include a start-up stage, a smelting stage and a shutdown stage;

[0013] S12. Analyze the current and power characteristics of the fused magnesium furnace at each production stage;

[0014] S13. Based on the analysis results, a mathematical model of the fused magnesium furnace at each production stage of the fused magnesium furnace is constructed;

[0015] S14. Based on the process flow and operating characteristics of the electric molten magnesium furnace, production constraints are established during the production stage of the electric molten magnesium furnace.

[0016] Preferably, the mathematical model of the fused magnesium furnace includes a current model and a power model, wherein:

[0017] The formulas of the current model and power model in the startup phase are:

[0018] I 启动 (t) = λt (t≤t1);

[0019]

[0020]

[0021] In the formula, I 启动 (t) represents the current value in the time period t during the start-up phase of the fused magnesium furnace; λ represents the current increase coefficient; P 启动 (t) represents the active power consumed in the time period t during the startup phase of the fused magnesium furnace; U represents the voltage value of the smelting process during the startup phase of the fused magnesium furnace; It indicates the power factor of the power supply system of the electric fused magnesium furnace; δ indicates the power increase coefficient;

[0022] The formulas of the current model and power model in the smelting stage are:

[0023] I 熔炼 (t) = I n (t1≤t≤t2);

[0024] P 熔炼 (t) = P N (t1≤t≤t2);

[0025] In the formula, I 熔炼(t) represents the current value in the smelting stage of the fused magnesium furnace in the time period t; I n Indicates the rated current of n phases in the working cycle; P 熔炼 (t) represents the power value of the smelting stage of the fused magnesium furnace in the time period t; P n Indicates the rated power of the electric fused magnesium furnace;

[0026] The formulas of the current model and power model in the shutdown stage are:

[0027] I 停炉 (t) = I n -(t-t2) (t≥t2);

[0028] P 停炉 (t) = P N -δ(t-t2) (t≥t2);

[0029] In the formula, I 停炉 (t) represents the current value of the fused magnesium furnace during the shutdown phase in time period t; I n Indicates the rated current of n phases in the working cycle; P 停炉 (t) represents the power value of the smelting stage of the fused magnesium furnace in the time period t; P n It indicates the rated power of the electric fused magnesium furnace; δ indicates the power increase coefficient.

[0030] Preferably, the production constraints include capacity adjustment constraints, power ramp constraints, adjustable state constraints and power balance constraints, wherein:

[0031] The formula for capacity adjustment constraint is:

[0032]

[0033]

[0034] P Mg，max ≤P(t)≤P Mg，min (t1≤t≤t2);

[0035] Where, P(t) represents the power value of the fused magnesium furnace in the production stage during time period t; P pre N represents the active power consumed in time period t during the fused magnesium production stage; Mg (t) represents the number of furnaces that can be adjusted in time period t; Pup and P down represents the adjustment capacity of each fused magnesium furnace; St1 and St2 represent the upper and lower power adjustment variables of the fused magnesium furnace in time period t; S t 1=1 means that the power is adjusted upward in time period t, S t 2 = 0 means that the power is adjusted downward in time period t; P mg,max and P mg,minThey respectively represent the upper and lower limits of active power during normal production of the fused magnesium furnace;

[0036] The formula for the power ramp constraint is:

[0037]

[0038] In the formula, and Respectively represent the power values ​​in the time periods ti and ti-1; RDg and RUg represent the maximum power values ​​that the fused magnesium furnace can reduce or increase in unit time;

[0039] The formula for the adjustable state constraint is:

[0040] P F =P pre +μ F P adj-st ;

[0041] Where P F P represents the active power actually consumed in the fused magnesium production stage after adjustment in time period t; pre represents the active power consumed in time period t during the fused magnesium production stage; μ F Represents a state variable, with a value of 0 or 1; P adj-st Indicates adjustable quantity;

[0042] The formula for the power balance constraint is:

[0043]

[0044] Where, T Mg Represents the operation cycle of the electric fused magnesium furnace production stage; P H，k It indicates the power of the fused magnesium furnace during different operation stages of the production process.

[0045] Preferably, according to the mathematical model of the fused magnesium furnace, a double-layer optimization model is constructed, and the double-layer optimization model is solved by an iterative approximation method to obtain the final decision variables, including the following steps:

[0046] S21. Based on the fluctuation of wind power and the sampling differences of different wind power time series, the upper optimization model is constructed with the maximum wind power consumption as the objective function and the power balance of the power system and the power regulation of the output electric molten magnesium furnace smelting load of wind turbines and thermal power units as constraints;

[0047] S22. Taking the minimum total operating cost of the power system as the objective function and the power balance of thermal power units as the constraint condition, the lower optimization model is constructed;

[0048] Among them, the total operating cost of the power system is the sum of the operating cost of thermal power units and the penalty cost of wind power abandonment;

[0049] S23, constructing a double-layer optimization model based on the upper-layer optimization model and the lower-layer optimization model;

[0050] S24. Use the iterative approximation method to solve the optimal solution of the two-layer optimization model and obtain the final decision variables.

[0051] Preferably, the expression of the upper optimization model is:

[0052]

[0053] Where P w-act represents the total output power of the wind farm during the dispatching period; ΔT1 represents the number of time periods involved in the power system dispatching; represents the output power of the i-th wind farm in time period t; N wind represents the number of wind farms incorporated into the power system; T represents the dispatching period.

[0054] Preferably, the expression of the lower layer optimization model is:

[0055] min F=M G +M W .

[0056]

[0057]

[0058] Where F represents the total operating cost of the power system; M G Represents the operating cost of thermal power units; M W Said wind power cuts penalty costs; A binary variable representing the on or off status of the jth thermal power unit; represents the power generation of the jth thermal power unit in period t; a j 、b j and c j They represent the maximum output capacity, minimum output capacity and operating cost parameters of the j-th thermal power unit respectively; η represents the unit reduction penalty cost of the i-th wind farm in the power system; ΔT2 represents the total number of time periods for wind power reduction in the power system; N G Indicates the number of thermal power units involved in power generation; T indicates the scheduling period; represents the original active power forecast output of the i-th wind farm during period t; represents the output power of the i-th wind farm in time period t; N wind Represents the number of wind farms incorporated into the power system.

[0059] Preferably, the expression of the two-layer optimization model is:

[0060]

[0061] Wherein, G(x, y) represents the objective function of the upper optimization model; O(x, z) represents the objective function of the lower optimization model; x represents the optimization variable; y and z both represent decision variables; i represents the i-th equality constraint in the two-level optimization model; j represents the j-th inequality constraint in the two-level optimization model; m represents the number of equality constraints in the upper optimization model; n represents the number of inequality constraints in the upper optimization model; u represents the number of equality constraints in the lower optimization model; v represents the number of inequality constraints in the lower optimization model; h(x, y) represents the equality constraint in the upper optimization model; k(x, y) represents the inequality constraint in the upper optimization model; r(x, y) represents the equality constraint in the lower optimization model; s(x, y) represents the inequality constraint in the lower optimization model.

[0062] Preferably, solving the optimal solution of the two-layer optimization model by an iterative approximation method to obtain the final decision variables includes the following steps:

[0063] S241, randomly selecting an initial solution, and using the initial solution as the value of the upper decision variable in the two-level optimization model;

[0064] S242, fixing the value of the upper-level decision variable, and solving the lower-level optimization problem in the two-level optimization model by using a genetic algorithm with changed parameters to obtain the optimal value of the lower-level decision variable;

[0065] S243, using the optimal lower-level decision variable value to update the upper-level decision variable to optimize the upper-level objective function;

[0066] S244, determine whether the update of the upper-level decision variables meets the predetermined convergence conditions. If not, return to step S242 to continue iteration. If so, stop iteration and output the upper-level decision variable value as the optimal solution of the two-level optimization model.

[0067] Preferably, solving the lower-level optimization problem in the two-level optimization model by using a genetic algorithm with changed parameters to obtain the optimal lower-level decision variable value comprises the following steps:

[0068] S2421, a random number generator initializes a population of a genetic algorithm, and an individual of each population represents a state of a power system;

[0069] S2422. Calculate the fitness value of each population individual according to the objective function and constraint conditions of the lower optimization model;

[0070] S2423, through the selection operation, select the individual with the highest fitness from the current population as the parent of the next generation population;

[0071] S2424, performing crossover and mutation operations on the selected parent individuals to obtain a new population;

[0072] S2425. Repeat steps S2422 to S2424 until the preset maximum number of iterations is reached, and output the solution corresponding to the optimal population individual as the optimal solution to the lower-level optimization problem.

[0073] The beneficial effects of the present invention are:

[0074] The present invention optimizes the fused magnesium furnace model by analyzing the load operation characteristics of the fused magnesium furnace, and integrates it with the thermal power unit as a controllable load resource to promote wind power consumption. Subsequently, combining the volatility of wind power and the established fused magnesium furnace load model, an optimized source-load coordinated scheduling two-layer optimization model is proposed, aiming to improve the wind power consumption capacity while minimizing the operating cost of the power system. The effectiveness and rationality of the proposed model are then verified in a specified scenario example. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0076] Figure 1 The process of the source-load coordinated scheduling method based on wind power fluctuation characteristics and electric magnesium furnace modeling according to an embodiment of the present invention is Figure 2

[0077] Figure 2 The power curve of the production process of the molten magnesium furnace in the source-load coordinated scheduling method based on wind power fluctuation characteristics and molten magnesium furnace modeling according to an embodiment of the present invention is Figure 2

[0078] Figure 3 This is an error area diagram of wind power output in different time series in a source-load coordinated scheduling method based on wind power fluctuation characteristics and electric magnesium furnace modeling according to an embodiment of the present invention

[0079] Figure 4 The source-load coordination double-layer optimization model process in the source-load coordination scheduling method based on wind power fluctuation characteristics and electric magnesium furnace modeling according to an embodiment of the present invention Figure 2

[0080] Figure 5 Comparison between conventional load forecast value and improved wind power output forecast value in source-load coordinated dispatching method based on wind power fluctuation characteristics and molten magnesium furnace modeling according to an embodiment of the present invention Figure 2

[0081] Figure 6 The output of thermal power units and the energy consumption of high-energy-consuming loads of molten magnesium in the source-load coordinated scheduling method based on wind power fluctuation characteristics and molten magnesium furnace modeling according to an embodiment of the present invention (Scheme 1) Figure 2

[0082] Figure 7 The output of thermal power units and the energy consumption of high-energy-consuming loads of molten magnesium in the source-load coordinated scheduling method based on wind power fluctuation characteristics and molten magnesium furnace modeling according to an embodiment of the present invention (Scheme 2) Figure 2

[0083] Figure 8 The output of four thermal power units in the source-load coordinated scheduling method based on wind power fluctuation characteristics and electric magnesium furnace modeling according to an embodiment of the present invention (Scheme 1) Figure 2

[0084] Fig. 9 The output of four thermal power units in the source-load coordinated scheduling method based on wind power fluctuation characteristics and electric magnesium furnace modeling according to an embodiment of the present invention (Scheme 2) Figure 2

[0085] Fig.10 The present invention is based on the wind power fluctuation characteristics and the molten magnesium furnace modeling in the source-load coordinated scheduling method according to the embodiment of the present invention, and the wind power consumption comparison without considering the wind power fluctuation under different schemes Figure 2

[0086] Fig.11 The present invention is based on the wind power fluctuation characteristics and the electric magnesium furnace modeling in the source-load coordinated scheduling method according to the embodiment of the present invention. The wind power consumption comparison considering wind power fluctuation under different schemes Figure 2

[0087] Fig.12 Comparison of wind power abandonment under different schemes in the source-load coordinated scheduling method based on wind power fluctuation characteristics and electric magnesium furnace modeling according to an embodiment of the present invention Figure 2

[0088] Fig.13 The comparison of the operating costs of thermal power units under different schemes in the source-load coordinated scheduling method based on wind power fluctuation characteristics and electric magnesium furnace modeling according to an embodiment of the present invention is Figure 2

[0089] Fig.14 It is a comparison chart of the total operating cost of the power system under different schemes in the source-load coordinated scheduling method based on wind power fluctuation characteristics and electric magnesium furnace modeling according to an embodiment of the present invention. DETAILED DESCRIPTION

[0090] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0091] According to an embodiment of the present invention, a source-load coordinated scheduling method based on wind power fluctuation characteristics and electric molten magnesium furnace modeling is provided.

[0092] The present invention will now be further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to an embodiment of the present invention, a source-load coordinated scheduling method based on wind power fluctuation characteristics and electric molten magnesium furnace modeling includes the following steps:

[0093] S1. Obtain the process flow of the fused magnesium furnace, analyze the operating characteristics of the fused magnesium furnace and construct a mathematical model of the fused magnesium furnace according to the analysis results;

[0094] It should be noted that the fused magnesium furnace has the characteristics of large capacity, strong regulation ability, fast response speed, high degree of automation, etc. Its operating characteristics and regulation ability have a significant impact on the power system's consumption of wind power and the control system's operating cost. In order to effectively utilize large-capacity fused magnesium loads in power grid peak regulation, it is necessary to establish an accurate mathematical model. The fused magnesium furnace model takes into account the voltage, current, power and other parameters during the operation of the fused magnesium furnace, which plays a key role in the subsequent control and regulation of the fused magnesium load. The different production stages of the fused magnesium furnace are analyzed, and the process is divided into three stages: furnace startup, smelting, and furnace shutdown. Figure 2 shown, and Figure 2 Area is a code, representing the energy required for upward or downward adjustment.

[0095] As a preferred embodiment, obtaining the process flow of the fused magnesium furnace, analyzing the operating characteristics of the fused magnesium furnace and constructing a mathematical model of the fused magnesium furnace according to the analysis results includes the following steps:

[0096] S11. According to the process flow of the fused magnesium furnace, the production stages of the fused magnesium furnace are divided, and the production stages include a start-up stage, a smelting stage and a shutdown stage;

[0097] It should be noted that the startup phase (t0-t1 period) mainly includes the preparation of power supply, electrodes and raw materials.

[0098] Melting stage (t1-t2 period), during this stage, the current and power parameters are monitored, and the position and output of the electrode and power supply are adjusted to ensure the stability of the arc and the uniformity of the melt in the furnace.

[0099] Shutdown phase (t2-t N During the shutdown phase, the current and temperature are monitored and gradually reduced to cool the fused magnesium furnace.

[0100] S12. Analyze the current and power characteristics of the fused magnesium furnace at each production stage;

[0101] S13. Based on the analysis results, a mathematical model of the fused magnesium furnace is constructed for each production stage of the fused magnesium furnace; the mathematical model of the fused magnesium furnace includes a current model and a power model, wherein:

[0102] The formulas of the current model and power model in the startup phase are:

[0103] I 启动 (t) = λt (t≤t1);

[0104]

[0105]

[0106] In the formula, I 启动 (t) represents the current value in the time period t during the start-up phase of the fused magnesium furnace; λ represents the current increase coefficient; P 启动 (t) represents the active power consumed in the time period t during the startup phase of the fused magnesium furnace; U represents the voltage value of the smelting process during the startup phase of the fused magnesium furnace; It indicates the power factor of the power supply system of the electric fused magnesium furnace; δ indicates the power increase coefficient;

[0107] The formulas of the current model and power model in the smelting stage are:

[0108] I 熔炼 (t) = I n (t1≤t≤t2);

[0109] P 熔炼 (t) = P N (t1≤t≤t2);

[0110] In the formula, I 熔炼 (t) represents the current value in the smelting stage of the fused magnesium furnace in the time period t; I n Indicates the rated current of n phases in the working cycle; P 熔炼 (t) represents the power value of the smelting stage of the fused magnesium furnace in the time period t; P N Indicates the rated power of the electric fused magnesium furnace;

[0111] The formulas of the current model and power model in the shutdown stage are:

[0112] I 停炉 (t) = I n -(t-t2) (t≥t2);

[0113] P 停炉 (t) = P N -δ(t-t2) (t≥t2);

[0114] In the formula, I 停炉 (t) represents the current value of the fused magnesium furnace during the shutdown phase in time period t; I n Indicates the rated current of n phases in the working cycle; P 停炉 (t) represents the power value of the smelting stage of the fused magnesium furnace in the time period t; P n It indicates the rated power of the electric fused magnesium furnace; δ indicates the power increase coefficient.

[0115] S 14. Based on the process flow and operating characteristics of the electric fused magnesium furnace, establish production constraints in the production stage of the electric fused magnesium furnace.

[0116] It should be noted that in the model of the electric magnesium furnace, the present invention adds additional constraints to the traditional basic electric magnesium furnace model. These constraints specifically address the operational flexibility required for effective demand response strategies in the face of variable wind power generation, and are able to explore the nuances of the load regulation capabilities of the electric magnesium furnace in the broader context of low-carbon power system dispatch.

[0117] Specifically, production constraints include capacity adjustment constraints, power ramp constraints, adjustable state constraints, and power balance constraints, among which:

[0118] The formula for capacity adjustment constraint is:

[0119]

[0120]

[0121] P Mg，max ≤P(t)≤P Mg，min (t1≤t≤t2);

[0122] Where, P(t) represents the power value of the fused magnesium furnace in the production stage during time period t; P pre N represents the active power consumed in time period t during the fused magnesium production stage; Mg (t) represents the number of furnaces that can be adjusted in time period t; P up and P down Indicates the adjustment capacity of each fused magnesium furnace; S t 1 and S t2 represents the upper and lower power adjustment variable of the fused magnesium furnace in time period t; S t 1=1 means that the power is adjusted upward in time period t, S t 2 = 0 means that the power is adjusted downward in time period t; P mg,max and P mg,min They respectively represent the upper and lower limits of active power during the normal production of fused magnesium;

[0123] Specifically, in order to enhance the stability and accuracy of the electrode current, the power level should not increase or decrease too quickly, and a "ramp constraint" should be used. The formula for the power ramp constraint is:

[0124]

[0125] In the formula, and Respectively expressed in t i and t i-1 Power value of the time period; RD g and RU g They respectively represent the maximum power value that the electric magnesium furnace can reduce or increase in unit time; that is, the limit value of the ramp constraint.

[0126] The formula for the adjustable state constraint is:

[0127] P F =P pre +μ F P adj-st ;

[0128] Where P F P represents the active power actually consumed in the fused magnesium production stage after adjustment in time period t; pre represents the active power consumed in time period t during the fused magnesium production stage; μ F Represents a state variable, with a value of 0 or 1; P adj-st Indicates adjustable quantity;

[0129] Specifically, in a single production cycle, the total energy consumption of the equipment is fixed and will not fluctuate significantly with changes in output. Although the energy consumption of the equipment can be optimized in each cycle through certain adjustments, this adjustment space is limited and cannot change the basic level of equipment energy consumption; the formula for the power balance constraint is:

[0130]

[0131] Where, T Mg Represents the operation cycle of the electric fused magnesium furnace production stage; P H，k It indicates the power of the fused magnesium furnace during different operation stages of the production process.

[0132] S2. According to the mathematical model of the fused magnesium furnace, a double-layer optimization model is constructed, and the double-layer optimization model is solved by an iterative approximation method to obtain the final decision variables;

[0133] It should be noted that the two-level optimization model, also known as the two-level mixed integer linear programming (BMILP) model, is a two-stage computational framework that effectively decomposes the complex decision-making process into two interrelated hierarchical layers. Each layer is designed with its own unique objective function and a set of constraints. The process first inputs the raw data into the upper layer, which makes strategic decisions based on the overall goals and parameters. Then, the results of the upper layer calculations are cascaded down as input constraints to the lower layer, which performs more detailed operational optimization and fine-tunes decisions within the scope established by the upper layer, running in tandem to solve multi-faceted optimization problems.

[0134] As a preferred implementation, a double-layer optimization model is constructed according to the mathematical model of the fused magnesium furnace, and the double-layer optimization model is solved by an iterative approximation method to obtain the final decision variables, including the following steps:

[0135] S21. Based on the fluctuation of wind power and the sampling differences of different wind power time series, the upper optimization model is constructed with the maximum wind power consumption as the objective function and the power balance of the power system and the power regulation of the output electric molten magnesium furnace smelting load of wind turbines and thermal power units as constraints;

[0136] It should be noted that the upper-level optimization model: establishes a wind power consumption optimization model that takes into account the characterization of wind power volatility. The volatility of wind power significantly affects the operational safety and economy of the power grid, increases regulatory pressure and operating costs, and reduces wind energy utilization and its benefits. Therefore, characterizing and evaluating the volatility of wind power is crucial to enhancing wind power consumption capacity. The wind power consumption optimization model aims to maximize the use of wind energy while meeting the requirements of power grid operation, taking into account various factors such as conventional power sources, conventional loads, electric magnesium furnaces and other loads, and the impact of wind power forecast errors.

[0137] Specifically, the expression of the upper optimization model is:

[0138]

[0139] Where P W-act represents the total output power of the wind farm during the dispatching period; ΔT1 represents the number of time periods involved in the power system dispatching; represents the output power of the i-th wind farm in time period t; N wind represents the number of wind farms incorporated into the power system; T represents the dispatching period.

[0140] In addition, the constraints of the upper optimization model mainly consider the power system constraints, the output constraints of wind turbines / thermal power units, and the power regulation constraints of large-capacity molten magnesium smelting loads.

[0141] Power system constraints include power balance constraints and spinning reserve constraints.

[0142] The expression of the equilibrium constraint is:

[0143]

[0144] In the formula, represents the output power of the i-th wind farm in time period t; It means that the thermal power unit outputs power in time period t; represents the predicted active power of conventional load in time period t; Represents the active power of each fused magnesium furnace in time period t; represents the regulation power of each fused magnesium furnace in the same period; T represents the scheduling cycle, that is, the number of time periods in a day; k represents the time period in which the fused magnesium furnace participates in the regulation; N Mg Indicates the number of fused magnesium furnaces involved in the regulation.

[0145] The expression of spinning reserve constraint is:

[0146]

[0147]

[0148] In the formula, Respectively represent the upper and lower limits of the thermal power unit output; Indicates the output of thermal power unit in time period t; They represent the positive and negative spinning reserves required to adjust the load forecast error during time period t respectively;

[0149] For wind turbine output constraints, from the perspective of wind power time series, sampling at different time intervals within the sequence will produce differences between the sequences. Especially when the original wind power series exhibits high volatility, a larger difference means a greater divergence between the two sequences. On the contrary, a smaller difference indicates a closer consistency between the two sequences. Suppose two time series: the first is the reference time series and the second is the target time series. The reference time series represents the most original continuous time series, but due to the limitations of the accuracy of the measurement system, its time interval can be adjusted according to the required accuracy. The difference area between the reference time series and the target time series is called the "waveform difference", such as Figure 3 The shaded area in and its sub-figures are shown. In order to eliminate the influence of installed capacity differences and time length on the calculation, the series are uniformly dimensionless before quantization.

[0150]

[0151] S = ∫ N P' ΔW (x)dx;

[0152]

[0153] Where P' ΔW Indicates the unit value of the fluctuation difference; P ΔW represents the wind power fluctuation difference between the reference time series and the target time series; P N represents the rated power of the electric fused magnesium furnace; S represents the sum of all differences in the degree of fluctuation between different time series; τ w The final expression for the fluctuation difference is used as an indicator to characterize the volatility of the wind power series;

[0154]

[0155]

[0156] In the formula, represents the predicted active power output of the ith wind farm in time period T without considering the change of time series prediction error, represents the original active power forecast output of the i-th wind farm in time period t; κ represents the parameters of different time series size states; P W-N Represents the installed capacity of a single wind farm.

[0157] The operation constraints of thermal power units include upper and lower limits of output power and ramp constraints of thermal power units;

[0158] The expressions for the upper and lower limits of output power are:

[0159]

[0160] The expression of the thermal power unit climbing constraint is:

[0161]

[0162]

[0163] In the formula, It represents the active output power of the thermal power unit in the time period t; It represents the active output power of the thermal power unit in the time period t-1; Respectively represent the lower and upper limits of the active output power of the thermal power unit; P Therup , P Therdown They represent the lifting and lowering limits of thermal power units respectively;

[0164] The power regulation constraints of the fused magnesium load include power regulation upper and lower limit constraints, regulation times constraints, fused magnesium load power ramp constraints and regulation energy balance constraints;

[0165] Among them, the expressions of upper and lower limits of power regulation are:

[0166]

[0167] Where P Hmax and P Hmin They respectively represent the upper and lower limits of the regulating power of the electric fused magnesium furnace; represents the regulating power of each fused magnesium furnace in the same period; k represents the period of time when the fused magnesium furnace participates in the regulation; N Mg Indicates the number of fused magnesium furnaces involved in the regulation.

[0168] The expression of the adjustment number constraint is:

[0169]

[0170] In the formula, S H t and S H t-1 represents a 0, 1 variable, indicating the regulation state of the fused magnesium furnace in the time period t and t-1, where S H t =0 means not participating in regulation, S H t =1 means participating in regulation, and M means the maximum number of adjustments during the production cycle of the electric fused magnesium furnace.

[0171] The expression of the power ramp constraint of fused magnesium load is:

[0172]

[0173]

[0174] In the formula, It represents the active output of the conventional generator set during period t; The unit representing the total active power output of a conventional generator set during the period t-1; P H，up and P H,down They respectively represent the upper and lower limits of the ramp rate of the electric magnesium furnace.

[0175] The expression for adjusting the energy balance constraint is:

[0176]

[0177] Where, T Mg Indicates the number of time periods in an operating cycle of the fused magnesium furnace; Indicates the regulation power of each electric magnesium furnace during the same period.

[0178] S22. Taking the minimum total operating cost of the power system as the objective function and the power balance of thermal power units as the constraint condition, the lower optimization model is constructed;

[0179] It should be noted that for the lower optimization model, the objective function of the lower model is to minimize the total operating cost of the power system, which is defined as the sum of the operating cost of the thermal power units and the penalty cost of wind power abandonment. This process aims to optimize the overall cost-effectiveness of the power system operation. The intermediary variable between the two layers is the total output power of the thermal power units. Based on the optimal dispatch solution of the upper optimization model and acknowledging the changes in the parameters of the thermal power units in the lower model, the output power distribution between the thermal power units is optimized to achieve the lowest operating cost of these units. By taking into account the previously calculated wind power abandonment cost, the total operating cost of the power system is further refined.

[0180] The expression of the lower optimization model is:

[0181] min F=M G +M W .

[0182]

[0183]

[0184] Where F represents the total operating cost of the power system; M G Represents the operating cost of thermal power units; M W Said wind power cuts penalty costs; A binary variable representing the on or off status of the jth thermal power unit; represents the power generation of the jth thermal power unit in period t; a j 、b j and c j They represent the maximum output capacity, minimum output capacity and operating cost parameters of the j-th thermal power unit respectively; η represents the unit reduction penalty cost of the i-th wind farm in the power system; ΔT2 represents the total number of time periods for wind power reduction in the power system; N G Indicates the number of thermal power units involved in power generation; T indicates the scheduling period; represents the original active power forecast output of the i-th wind farm during period t; represents the output power of the i-th wind farm in time period t; N wind Represents the number of wind farms incorporated into the power system.

[0185] Among them, the total operating cost of the power system is the sum of the operating cost of thermal power units and the penalty cost of wind power abandonment;

[0186] In addition, the expression of the power balance constraint of the thermal power unit is:

[0187]

[0188] Where N G Indicates the number of thermal power units, It represents the power generation of the jth thermal power unit in the t period, and T represents a scheduling cycle, that is, the number of time periods in a day.

[0189] S23, constructing a double-layer optimization model based on the upper-layer optimization model and the lower-layer optimization model;

[0190] It should be noted that the two-level optimization model belongs to the category of nonlinear mixed integer programming problems, and its computational complexity is relatively high. In order to effectively solve this problem, a hybrid algorithm based on genetic algorithm is adopted. The mathematical formula of the two-level optimization model is provided below. The method flow is as follows: Figure 4 shown.

[0191] Specifically, the expression of the two-level optimization model is:

[0192]

[0193] Wherein, G(x, y) represents the objective function of the upper optimization model; O(x, z) represents the objective function of the lower optimization model; x represents the optimization variable; y and z both represent decision variables; i represents the i-th equality constraint in the optimization model; .j represents the j-th inequality constraint in the optimization model; m represents the number of equality constraints in the upper optimization model; n represents the number of inequality constraints in the upper optimization model; u represents the number of equality constraints in the lower optimization model; v represents the number of inequality constraints in the lower optimization model; h(x, y) represents the equality constraint in the upper optimization model; k(x, y) represents the inequality constraint in the upper optimization model; r(x, y) represents the equality constraint in the lower optimization model; s(x, y) represents the inequality constraint in the lower optimization model.

[0194] In the optimization framework proposed by the present invention, the upper objective function is represented by G(x, y), and the objective function of the lower problem is represented by O(x, z), where the variable x encapsulates the optimization variables, including the total output of the thermal power plant, the specified power at the wind power intersection, and the power of the management molten magnesium smelting process. The variable y represents the operating state, such as the switching state of the thermal power plant and the wind farm. In contrast, z represents lower-level decision variables, which are determined by the results of the higher-level decisions represented by x.

[0195] In this structured model, higher-level constraints are expressed as the equation h(x, y) = 0 and the inequality k(x, y) ≤ 0, which stipulate the conditions for thermal power output and wind power integration. In the lower-level model, the equation r(x, z) = 0 and the inequality s(x, z) ≤ 0 are used to constrain the power balance of the system, including the operation constraints of thermal power and wind turbines, as well as the constraints on the energy storage system.

[0196] Considering the complexity of this two-level scheduling model, it is not feasible to solve it directly. Therefore, this model requires an iterative approximation method with a suitable optimization algorithm to converge to the optimal solution.

[0197] S24. Use the iterative approximation method to solve the optimal solution of the two-layer optimization model and obtain the final decision variables.

[0198] As a preferred implementation, using an iterative approximation method to solve the optimal solution of the two-level optimization model and obtain the final decision variables includes the following steps:

[0199] S241, randomly selecting an initial solution, and using the initial solution as the value of the upper decision variable in the two-level optimization model;

[0200] S242, fixing the value of the upper-level decision variable, and solving the lower-level optimization problem in the two-level optimization model by using a genetic algorithm with changed parameters to obtain the optimal value of the lower-level decision variable;

[0201] Step S242 is to solve the nonlinear mixed integer programming problem. Since the genetic algorithm has a strong exploration ability across complex solution spaces and adaptability to nonlinear and constraint-rich environments, it is very suitable for dual-objective optimization problems. The genetic algorithm with changed parameters is used to improve the algorithm. In order to reduce the probability of the algorithm falling into the local optimum, parameters such as crossover probability and mutation probability are continuously changed during the solution process, that is, the search space of the algorithm is increased and the diversity within its search range is improved.

[0202] As a preferred implementation, solving the lower-level optimization problem in the two-level optimization model by using a genetic algorithm with changed parameters to obtain the optimal lower-level decision variable value includes the following steps:

[0203] S2421, a random number generator initializes a population of a genetic algorithm, and an individual of each population represents a state of a power system;

[0204] Specifically, a random number generator initializes the population of the genetic algorithm, and each individual of the population represents the state of the power system and also includes a code. The total output power of the thermal power unit and the adjustment power of the molten magnesium smelting process are used as individuals in the genetic algorithm, and they are represented as chromosomes using binary coding.

[0205] S2422. Calculate the fitness value of each population individual according to the objective function and constraint conditions of the lower optimization model;

[0206] S2423, through the selection operation, select the individual with the highest fitness from the current population as the parent of the next generation population;

[0207] S2424, performing crossover and mutation operations on the selected parent individuals to obtain a new population;

[0208] S2425. Repeat steps S2422 to S2424 until the preset maximum number of iterations is reached, and output the solution corresponding to the optimal population individual as the optimal solution to the lower-level optimization problem.

[0209] S243. Use the optimal value of the lower-level decision variable to update the upper-level decision variable to optimize the upper-level objective function; (the specific process is to solve the output of each thermal power unit under the current conditions through the lower layer, update the original output data of the thermal power unit of the upper-level optimization model in the next iteration, and then perform a new round of iteration until the result meets the convergence condition and the iteration is terminated).

[0210] S244, determine whether the update of the upper-level decision variables meets the predetermined convergence conditions. If not, return to step S242 to continue iteration. If so, stop iteration and output the upper-level decision variable value as the optimal solution of the two-level optimization model.

[0211] S3. Based on the final decision variables obtained, the validity is verified through the preset case library.

[0212] Specifically, the rationality and effectiveness of the proposed source-load coordination scheduling model are verified through specific cases, including:

[0213] (1) Verification of the effectiveness of wind power volatility indicators;

[0214] The variability of wind power output affects the accuracy of forecasts based on different time series. When relying on data from a specific time series, larger variability leads to less precise forecasts, while smaller variability leads to more accurate forecasts. Therefore, forecast inaccuracy can indirectly indicate the volatility of wind power. In order to evaluate τ wWhether the index can effectively measure the volatility of wind power, the present invention adopts two prediction models, the ARMA time series model and the back propagation neural network. The ARMA model is a linear prediction method that predicts future wind power generation by analyzing the time series of historical data. It aims to identify and utilize linear patterns and trends in the data to provide enhanced predictions for future wind power output. In contrast, the back propagation neural network is a nonlinear prediction method based on artificial intelligence, which solves complex data patterns by developing and training neural networks, especially nonlinear and irregular changes commonly found in wind power data. Learning from historical data patterns allows back propagation neural networks to provide more accurate wind power generation forecasts. The application of these models in wind power generation forecasting depends largely on the characteristics of the data and the specific forecasting needs. These techniques are applied to the short-term power output forecast of a single wind farm.

[0215] The forecast horizon was set to one day (96 time periods), and the root mean square error (RMSE) was used to evaluate the forecast accuracy.

[0216] (2) Optimization scheduling model verification;

[0217] In order to fully verify the effectiveness and rationality of the proposed indicators and models, the present invention conducted a series of simulation analyses based on the following conditions and parameter configurations: a total of 4 thermal power units were involved, with a total installed capacity of 1500MW; the maximum output capacity of each unit (P max )、Minimum output capacity(P min ) and operating cost parameters (denoted as a, b, c) are shown in Table 1.

[0218] Table 1 Maximum output capacity, minimum output capacity and operating cost parameters of each unit

[0219]

[0220]

[0221] At the same time, the total installed capacity of the wind farm is set to 350MW, which ensures the unit reduction penalty cost. The total power consumption of the high-energy load of the molten magnesium furnace is set to 230MW, and it can produce continuously for more than 24 hours. Its upward and downward regulation capabilities are 20% and 15% of the rated power, respectively. Taking into account the complexity of the internal reaction mechanism of the molten magnesium furnace during the startup and shutdown stages, and its regulation capabilities during these stages, the present invention intentionally excludes the consideration of these stages when determining the regulation interval. This is done to improve the accuracy and practicality of the simulation analysis and to provide a solid theoretical and practical basis for evaluating the effectiveness of the proposed model.

[0222] Figure 5It shows the relationship between conventional load and wind power output, reflecting the anti-peak regulation characteristics of wind power, which not only increases the peak load burden of thermal power units, but also reduces the efficiency of wind energy utilization.

[0223] For in-depth analysis and comparison, two different scenarios are set: Scheme 1, ignoring the optimization modeling of fused magnesium load regulation and not considering the operating scenario of such load regulation; Scheme 2, adopting the source-load coordinated optimization scheduling method, including the optimization modeling of fused magnesium load regulation. Both schemes will be evaluated under a unified simulation setting to ensure a balanced comparison. This comparative analysis aims to reveal the potential benefits of regulating fused magnesium load in improving wind power utilization and reducing peak load pressure of thermal power units, thereby providing a more accurate and effective strategy for the optimal scheduling of wind power.

[0224] As shown in Table 2, although the three calculation methods produce different results when predicting a single wind farm power series, the root mean square error (RMSE) values ​​obtained by the three methods show a consistent decreasing trend. The prediction accuracy increases steadily from low to high. This trend is consistent with τ w Therefore, using τ w Quantifying the volatility of wind power time series is a scientific and reasonable method. Therefore, by analyzing the wind power output series at different time scales, we can obtain the volatility of wind power time series based on the calculated τ. w The forecast error calculated for the time scale required for wind power output.

[0225] Table 2 Root mean square error of different wind power series predictions

[0226] Type / Method ARMA BP Neural Network <![CDATA[Volatility index τ w > Wind farm 0.02544 0.02315 0.02108

[0227] According to the above case scheme, the energy consumption results of the thermal power unit and the high energy load of the fused magnesium under different schemes of Scheme 1 and Scheme 2 are calculated, such as Figure 6 and Figure 7 As shown. Figure 6 and Figure 7 It can be clearly seen that the two curves represent the total capacity and energy consumption of the four thermal power units and the high-load energy consumption of fused magnesium in the next 24 hours, respectively, and contain a total of 96 time intervals. It can be clearly seen from the figure that in the case of Scheme 1, the high-load energy consumption of fused magnesium remains unchanged, while the total capacity of the four thermal power units is low in the first 24 hours, corresponding to the trough period in the 96 time intervals. In contrast, the power generation in the 28th-44th and 68th-88th periods increased significantly, reaching the daily peak, with a peak-to-valley difference of 799MW. In Scheme 2, the high-load energy load of fused magnesium participates in the flexible regulation of the power system. In the 1st-28th and 48th-64th periods, there was a significant increase in electricity, which enhanced the absorption capacity of wind power. At the same time, the total capacity of the four thermal power units in Scheme 2 is reduced compared with the peak-to-valley difference of Scheme 1, falling to 649MW, and the output is more stable.

[0228] from Figure 8 and Fig. 9 The operating status of the four thermal power units (units 1-4) can be seen in the figure. A macro analysis of these two figures shows that the output of the units in Scheme 2 is more stable than that in Scheme 1 within the 24-hour scheduling cycle.

[0229] Fig.10 and Fig.11 The wind power absorption capacity of Schemes 1 and 2 with and without considering the changes in renewable energy is compared respectively. Fig.12 The wind curtailment in Scheme 1 and Scheme 2 was compared, taking into account the volatility of renewable energy. It is obvious that compared with wind curtailment, Scheme 2 brings more additional wind energy absorption. The highest levels were reached in the 10-25, 45-70 and 90-96 time periods, accounting for about 90% of the total wind curtailment in the entire scheduling cycle, and the maximum wind curtailment power in a single time period reached 96MW·h. After the implementation of Scheme 2, the maximum wind curtailment power in a single time interval dropped to 36MW·h. Therefore, the coordinated scheduling method involving high-energy loads of fused magnesium can significantly reduce wind power curtailment as a whole. Comparison Fig.10 and Fig.11 It can be seen that when considering the wind power fluctuation error, such as around the time intervals of 45, 65, and 83-90, Scheme 2 has completely absorbed the wind power without considering the fluctuation error. However, when considering the fluctuation error, there is still some remaining wind power that has not been completely absorbed. Therefore, in the analysis considering the wind power fluctuation characterization, a more accurate absorption of wind power in the power grid and a more accurate calculation of the wind abandonment cost can be provided.

[0230] Fig.13 The individual cost distribution of the four thermal power units under Scheme 1 and Scheme 2 is given. Among them, the costs of Units 1, 2, and 3 are lower under Scheme 2, but the cost of Unit 4 is lower under Scheme 1, indicating that overall, the cost of Scheme 2 is lower. Fig.14 The total cost of power system operation under the two schemes is shown. Scheme 2 considers the participation of fused magnesium load in the source load coordinated dispatch method, which effectively reduces the power system operation cost.

[0231] Through simulation analysis, it is verified that the use of source-load coordinated scheduling upper and lower level models can enhance the flexibility of power system control, effectively reduce wind power abandonment (or increase wind power consumption rate), and reduce power system operating costs.

[0232] In summary, with the help of the above-mentioned technical scheme of the present invention, the present invention optimizes the electric molten magnesium furnace model by analyzing the load operation characteristics of the electric molten magnesium furnace, and integrates it with the thermal power unit as a controllable load resource to promote wind power consumption. Subsequently, combined with the volatility of wind power and the established electric molten magnesium furnace load model, an optimized source-load coordinated scheduling two-layer optimization model is proposed, aiming to improve the wind power consumption capacity while minimizing the operating cost of the power system, and then the effectiveness and rationality of the proposed model are verified in the specified scheme example.

[0233] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, optical storage, etc.) containing computer-usable program code.

[0234] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A source-load coordinated scheduling method based on wind power fluctuation characteristics and molten magnesium furnace modeling, characterized in that: The source-load coordinated scheduling method based on wind power fluctuation characteristics and molten magnesium furnace modeling includes the following steps: S 1. Obtain the process flow of the fused magnesium furnace, analyze the operating characteristics of the fused magnesium furnace and construct a mathematical model of the fused magnesium furnace based on the analysis results; S2. According to the mathematical model of the fused magnesium furnace, a double-layer optimization model is constructed, and the double-layer optimization model is solved by an iterative approximation method to obtain the final decision variables; S3. Based on the final decision variables obtained, the validity is verified through the preset case library.

2. The source-load coordinated scheduling method based on wind power fluctuation characteristics and molten magnesium furnace modeling according to claim 1 is characterized in that: The process of obtaining the process flow of the fused magnesium furnace, analyzing the operation characteristics of the fused magnesium furnace and constructing a mathematical model of the fused magnesium furnace according to the analysis results comprises the following steps: S 11. According to the process flow of the fused magnesium furnace, the production stages of the fused magnesium furnace are divided, and the production stages include a startup stage, a smelting stage and a shutdown stage; S 12. Analyze the current and power characteristics of the fused magnesium furnace at each production stage; S 13. Based on the analysis results, a mathematical model of the fused magnesium furnace at each production stage of the fused magnesium furnace is constructed; S 14. Based on the process flow and operation characteristics of the fused magnesium furnace, production constraints are established in the production stage of the fused magnesium furnace.

3. The source-load coordinated scheduling method based on wind power fluctuation characteristics and molten magnesium furnace modeling according to claim 2 is characterized in that: The mathematical model of the fused magnesium furnace includes a current model and a power model, wherein: The formulas of the current model and power model in the startup phase are: I 启动 (t)=λt(t≤t1); In the formula, I 启动 (t) represents the current value in the time period t during the start-up phase of the fused magnesium furnace; λ represents the current increase coefficient; P 启动 (t) represents the active power consumed in the time period t during the startup phase of the fused magnesium furnace; U represents the voltage value of the smelting process during the startup phase of the fused magnesium furnace; It indicates the power factor of the power supply system of the electric fused magnesium furnace; δ indicates the power increase coefficient; The formulas of the current model and power model in the smelting stage are: I 熔炼 (t)=I n (t1≤t≤t2); P 熔炼 (t)=P N (t1≤t≤t2); In the formula, I 熔炼 (t) represents the current value in the smelting stage of the fused magnesium furnace in the time period t; I n Indicates the rated current of n phases in the working cycle; P 熔炼 (t) represents the power value of the smelting stage of the fused magnesium furnace in the time period t; P n Indicates the rated power of the electric fused magnesium furnace; The formulas of the current model and power model in the shutdown stage are: I 停炉 (t)=I n -(t-t2)(t≥t2); P 停炉 (t)=P N -δ(t-t2)(t≥t2); In the formula, I 停炉 (t) represents the current value of the fused magnesium furnace during the shutdown phase in time period t; I n Indicates the rated current of n phases in the working cycle; P 停炉 (t) represents the power value of the smelting stage of the fused magnesium furnace in the time period t; P n It indicates the rated power of the electric fused magnesium furnace; δ indicates the power increase coefficient.

4. The source-load coordinated scheduling method based on wind power fluctuation characteristics and molten magnesium furnace modeling according to claim 2 is characterized in that: The production constraints include capacity adjustment constraints, power ramp constraints, adjustable state constraints and power balance constraints, where: The formula for the capacity adjustment constraint is: P Mg，max ≤P(t)≤P Mg，min (t1≤t≤t2); Where, P(t) represents the power value of the fused magnesium furnace in the production stage during time period t; P pre N represents the active power consumed in time period t during the fused magnesium production stage; Mg (t) represents the number of furnaces that can be adjusted in time period t; P up and P down Indicates the adjustment capacity of each fused magnesium furnace; S t 1 and S t 2 represents the upper and lower power adjustment variable of the fused magnesium furnace in time period t; S t 1=1 means that the power is adjusted upward in time period t, S t 2 = 0 means that the power is adjusted downward in time period t; P mg,max and P mg,min They respectively represent the upper and lower limits of active power during the normal production of fused magnesium; The formula for the power ramp constraint is: In the formula, and Respectively represent the power values ​​in the time periods ti and ti-1; RDg and RUg represent the maximum power values ​​that the fused magnesium furnace can reduce or increase in unit time; The formula for the adjustable state constraint is: P F =P pre +μ F P adj-st ; Where P F P represents the active power actually consumed in the fused magnesium production stage after adjustment in time period t; pre represents the active power consumed in time period t during the fused magnesium production stage; μ F Represents a state variable, with a value of 0 or 1; P adj-xt Indicates adjustable quantity; The formula for the power balance constraint is: Where, T Mg Represents the operation cycle of the electric fused magnesium furnace production stage; P H，k It indicates the power of the fused magnesium furnace during different operation stages of the production process.

5. The source-load coordinated scheduling method based on wind power fluctuation characteristics and molten magnesium furnace modeling according to claim 1 is characterized in that: The method of constructing a double-layer optimization model based on the mathematical model of the fused magnesium furnace and solving the double-layer optimization model by an iterative approximation method to obtain the final decision variables includes the following steps: S21. Based on the fluctuation of wind power and the sampling differences of different wind power time series, the upper optimization model is constructed with the maximum wind power consumption as the objective function and the power balance of the power system and the power regulation of the output electric molten magnesium furnace smelting load of wind turbines and thermal power units as constraints; S22. Taking the minimum total operating cost of the power system as the objective function and the power balance of thermal power units as the constraint condition, the lower optimization model is constructed; The total operating cost of the power system is the sum of the operating cost of the thermal power units and the penalty cost for wind power abandonment; S23, constructing a double-layer optimization model based on the upper-layer optimization model and the lower-layer optimization model; S24. Use the iterative approximation method to solve the optimal solution of the two-layer optimization model and obtain the final decision variables.

6. The source-load coordinated scheduling method based on wind power fluctuation characteristics and molten magnesium furnace modeling according to claim 5 is characterized in that: The expression of the upper optimization model is: Where P w-act represents the total output power of the wind farm during the dispatching period; ΔT1 represents the number of time periods involved in the power system dispatching; represents the output power of the i-th wind farm in time period t; N wind represents the number of wind farms incorporated into the power system; T represents the dispatching period.

7. The source-load coordinated scheduling method based on wind power fluctuation characteristics and fused magnesium furnace modeling according to claim 5 is characterized in that: The expression of the lower layer optimization model is: minF=M G +M W . Where F represents the total operating cost of the power system; M G Represents the operating cost of thermal power units; M W Said wind power cuts penalty costs; A binary variable representing the on or off status of the jth thermal power unit; represents the power generation of the jth thermal power unit in period t; a j , b j and c j They represent the maximum output capacity, minimum output capacity and operating cost parameters of the j-th thermal power unit respectively; η represents the unit curtailment penalty cost of the i-th wind farm in the power system; ΔT2 represents the total number of time periods for wind power curtailment in the power system; N G Indicates the number of thermal power units involved in power generation; T indicates the scheduling period; represents the original active power forecast output of the i-th wind farm during period t; represents the output power of the i-th wind farm in time period t; N wind Represents the number of wind farms incorporated into the power system.

8. The source-load coordinated scheduling method based on wind power fluctuation characteristics and fused magnesium furnace modeling according to claim 7 is characterized in that: The expression of the two-layer optimization model is: Wherein, G(x, y) represents the objective function of the upper optimization model; O(x, z) represents the objective function of the lower optimization model; x represents the optimization variable; .y and z both represent decision variables; i represents the i-th equality constraint in the optimization model; j represents the j-th inequality constraint in the optimization model; m represents the number of equality constraints in the upper optimization model; n represents the number of inequality constraints in the upper optimization model; u represents the number of equality constraints in the lower optimization model; v represents the number of inequality constraints in the lower optimization model; h(x, y) represents the equality constraint in the upper optimization model; k(x, y) represents the inequality constraint in the upper optimization model; r(x, y) represents the equality constraint in the lower optimization model; s(x, y) represents the inequality constraint in the lower optimization model.

9. The source-load coordinated scheduling method based on wind power fluctuation characteristics and fused magnesium furnace modeling according to claim 5 is characterized in that: The method of solving the optimal solution of the two-layer optimization model by using the iterative approximation method to obtain the final decision variables includes the following steps: S241, randomly selecting an initial solution, and using the initial solution as the value of the upper decision variable in the two-level optimization model; S242, fixing the value of the upper-level decision variable, and solving the lower-level optimization problem in the two-level optimization model by using a genetic algorithm with changed parameters to obtain the optimal value of the lower-level decision variable; S243, using the optimal lower-level decision variable value to update the upper-level decision variable to optimize the upper-level objective function; S244, determine whether the update of the upper-level decision variables meets the predetermined convergence conditions. If not, return to step S242 to continue iteration. If so, stop iteration and output the upper-level decision variable value as the optimal solution of the two-level optimization model.

10. The source-load coordinated scheduling method based on wind power fluctuation characteristics and fused magnesium furnace modeling according to claim 9 is characterized in that: The method of solving the lower-level optimization problem in the two-level optimization model by using a genetic algorithm with changed parameters to obtain the optimal lower-level decision variable value comprises the following steps: S2421, a random number generator initializes a population of a genetic algorithm, and an individual of each population represents a state of a power system; S2422. Calculate the fitness value of each population individual according to the objective function and constraint conditions of the lower optimization model; S2423, through the selection operation, select the individual with the highest fitness from the current population as the parent of the next generation population; S2424, performing crossover and mutation operations on the selected parent individuals to obtain a new population; S2425. Repeat steps S2422 to S2424 until the preset maximum number of iterations is reached, and output the solution corresponding to the optimal population individual as the optimal solution to the lower-level optimization problem.

Images