Low-carbon scheduling optimization method and system based on two-stage robustness and interval carbon constraint

Abstract

The invention discloses a low-carbon scheduling optimization method and system based on two-stage robustness and interval carbon constraint, and the method comprises the steps: constructing a two-stage robust optimization model, constructing an uncertain set of wind power and photovoltaic output, and considering a predicted value and a fluctuation state of the uncertain set to describe the uncertainty of wind and light output; performing interval uncertainty modeling on emission factors of the carbon emission energy equipment, and introducing predicted values, variation coefficients and fluctuation values to represent interval characteristics of the emission factors; designing a robust carbon emission constraint, and ensuring that the carbon emission of the system does not exceed the upper limit under the most unfavorable condition through interval upper bound modeling; and solving the two-stage robust optimization model by adopting a column constraint generation algorithm, and decomposing the problem into a main problem and a sub-problem to obtain the most economical regulation and control scheme and the worst scene under the day-ahead scheduling scheme. And dynamic cooperation and optimization balance among low carbon, robustness and economy in the system are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of energy scheduling, in particular to a low-carbon scheduling optimization method and system based on two-stage robustness and interval carbon constraints. BACKGROUND

[0002] Integrated energy systems (IEM) significantly improve energy utilization efficiency and promote renewable energy consumption by integrating multiple energy forms and optimizing energy conversion and storage. This system not only alleviates energy supply and demand pressure, but also promotes environmental improvement, which is highly consistent with the green and low-carbon energy transformation strategy advocated by the country. However, in actual operation, due to factors such as fuel quality, operating state, measurement error, etc., the emission factor (i.e., the amount of carbon emissions per unit of power generation or heating) of different energy devices (such as gas turbine units, combined heat and power units, etc.) presents obvious interval uncertainty. If this uncertainty is ignored in scheduling optimization, it may lead to emission overruns, plan deviations or loss of control. SUMMARY

[0003] The present application aims to overcome the shortcomings of the prior art and provide a low-carbon scheduling optimization method based on two-stage robustness and interval carbon constraints, comprising the following steps:

[0004] Step one, construct a two-stage robust optimization model, the model is min-max-min structure, with the minimum scheduling cost as the target, the outer layer is the day-ahead economic low-carbon scheduling problem, the optimization variable is the day-ahead scheduling scheme, the inner layer is the control sub-problem caused by source and load uncertainty, the optimization variable is the control scheme and the worst-case scenario;

[0005] Step two, construct the uncertainty set of wind power and photovoltaic output, considering the predicted value and fluctuation state to describe the uncertainty of wind and light output;

[0006] Step three, model the interval uncertainty of the emission factor of carbon emission energy devices, introduce the predicted value, change coefficient and fluctuation value to represent the interval characteristics of the emission factor;

[0007] Step four, design a robust carbon emission constraint, model the upper bound of the interval to ensure that the carbon emissions of the system do not exceed the upper limit in the most adverse situation;

[0008] Step five, use the column constraint generation algorithm to solve the two-stage robust optimization model, decompose the problem into a main problem and a sub-problem to obtain the most economic control scheme and the worst-case scenario under the day-ahead scheduling scheme.

[0009] Further, the construction formula of the uncertainty set is:

[0010]

[0011] Wherein, in the formula: and respectively represent the predicted value of wind power and photovoltaic at t period; τ WT,t and τ PV,t respectively represent the fluctuation state of wind power and photovoltaic at t period, taking 0 or 1 as value; l WT,t and l PV,t respectively represent the fluctuation value of wind power and photovoltaic at t period; Ω WT and Ω PV represent uncertainty, representing the number of periods in which wind and light output fluctuation occurs in the period.

[0012] Further, the emission factor modeling formula is:

[0013]

[0014] In the formula: ε i represents a random variable of unknown distribution; represents a predicted value; n represents a variation coefficient; ε i represents a fluctuation value.

[0015] The two-stage robust and interval carbon constraint based low-carbon scheduling optimization method according to claim 1, wherein the robust carbon emission constraint is:

[0016] The carbon emission calculation model is:

[0017]

[0018] In the formula: represents the total carbon emission amount in the system; P i,t represents the carbon emission amount of a certain carbon emission device in t period;

[0019] The robust emission constraint is:

[0020]

[0021] In the formula: represents the maximum value of carbon emission amount in the system.

[0022] Further, the objective function includes:

[0023] The two-stage robust optimization model is:

[0024]

[0025] In the formula: k1 and k2 represent weight coefficients, used to balance the economic, low-carbon and robustness targets; represents the additional control cost considering uncertainty;

[0026] The objective function is:

[0027]

[0028] In the formula: represents the operation cost of each unit in the system at the t period; represents the grid interaction cost, which represents the purchase of electricity from the grid when it is greater than 0, otherwise represents the sale of electricity to the grid; represents the carbon emission cost in the system;

[0029] The carbon emission cost is:

[0030]

[0031] In the formula: represents the carbon emission cost in the system; represents the unit carbon emission price; represents the total carbon emission in the system.

[0032] A low-carbon scheduling optimization system based on two-stage robust optimization and interval carbon constraints, the low-carbon scheduling optimization method based on two-stage robust optimization and interval carbon constraints comprises a model construction module, an uncertainty processing module, a constraint design module, a solution output module and a data processing module;

[0033] The model construction module, the uncertainty processing module, the constraint design module and the solution output module are connected with the data processing module respectively.

[0034] The model construction module is used for constructing a two-stage robust optimization model, determining a min-max-min structure and an optimization target.

[0035] The uncertainty processing module is used for constructing a wind and light output uncertainty set and an emission factor interval model.

[0036] The constraint design module is used for designing a robust carbon emission constraint and a target function.

[0037] The solution output module is used for iteratively solving the model by using a column constraint generation algorithm and outputting an optimal scheduling scheme.

[0038] The low-carbon scheduling optimization method based on two-stage robust optimization and interval carbon constraints has the beneficial effects that: two-stage robust optimization is used to realize safe and flexible real-time scheduling, thereby improving the robustness and economy of the system as a whole, interval robust constraints are embedded to effectively deal with carbon emission and uncertainty, and dynamic coordination and optimal balance among low carbon, robustness and economy in the system are realized. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 It is a flowchart of the low-carbon scheduling optimization method based on two-stage robust optimization and interval carbon constraints.

[0040] Figure 2 Flowchart for solving the main problem and the sub-problem. DETAILED DESCRIPTION

[0041] The technical solutions of the present application are described in further detail below in combination with the drawings, but the protection scope of the present application is not limited to the following description.

[0042] The features and performances of the present application are described in further detail below in combination with the examples.

[0043] As shown in the figure, the low-carbon scheduling optimization method based on two-stage robustness and interval carbon constraints includes the following steps: Figure 1 Step one, construct a two-stage robust optimization model, which is a min-max-min structure, with the goal of minimizing scheduling cost, the outer layer being a day-ahead economic low-carbon scheduling problem, the optimization variable being the day-ahead scheduling scheme, and the inner layer being a control sub-problem caused by source and load uncertainty, the optimization variable being the control scheme and the worst-case scenario;

[0044] Step two, construct an uncertain set of wind power and photovoltaic output, considering the predicted value and fluctuation state, to describe the uncertainty of wind and light output;

[0045] Step three, model the interval uncertainty of the emission factor of carbon-emitting energy equipment, introduce the predicted value, the change coefficient and the fluctuation value to represent the interval characteristics of the emission factor;

[0046] Step four, design a robust carbon emission constraint, and ensure that the carbon emission amount does not exceed the upper limit in the most adverse situation through interval upper bound modeling;

[0047] Step five, solve the two-stage robust optimization model by using the column constraint generation algorithm, decompose the problem into a main problem and a sub-problem, and obtain the most economic control scheme and the worst-case scenario under the day-ahead scheduling scheme.

[0048] The construction formula of the uncertain set is:

[0049]

[0050]

[0051] Wherein, in the formula: and respectively represent the predicted value of wind power and photovoltaic at t period; τ WT,t and τ PV,t respectively represent the fluctuation state of wind power and photovoltaic at t period, taking the value of 0 or 1; l WT,t and l PV,t respectively represent the fluctuation value of wind power and photovoltaic at t period; Ω WT and Ω PV represent the uncertainty degree, representing the number of periods in which wind and light output fluctuation occurs in a period.​

[0052] The emission factor modeling formula is:

[0053]

[0054] In the formula, ε i represents a random variable of an unknown distribution; represents a predicted value; n represents a variation coefficient; and ε i represents a fluctuation value.

[0055] The low-carbon scheduling optimization method based on two-stage robustness and interval carbon constraints according to claim 1, wherein the robust carbon emission constraint is:

[0056] The carbon emission calculation model is:

[0057]

[0058] In the formula, P represents the total carbon emission amount in the system; and P i,t represents the carbon emission amount of a certain carbon emission device within a t period;

[0059] The robust emission constraint is:

[0060]

[0061] In the formula, P represents the maximum value of the carbon emission amount in the system.

[0062] The objective function comprises:

[0063] The two-stage robust optimization model is:

[0064]

[0065] In the formula, κ1 and k2 represent weight coefficients for balancing the economic, low-carbon, and robustness targets; represents an additional control cost caused by considering uncertainty;

[0066] The objective function is:

[0067]

[0068] In the formula, C represents the operation cost of each unit in the system within a t period; represents a grid interaction cost, which represents power purchase from the grid when it is greater than 0, and represents power sale to the grid otherwise; represents the carbon emission cost in the system;

[0069] The carbon emission cost is:

[0070]

[0071] wherein: represents the carbon emission cost within the system; represents the unit carbon emission price; represents the total amount of carbon emission within the system.

[0072] A low-carbon scheduling optimization system based on two-stage robust optimization and interval carbon constraints, the low-carbon scheduling optimization method based on two-stage robust optimization and interval carbon constraints comprises a model construction module, an uncertainty processing module, a constraint design module, a solution output module and a data processing module;

[0073] The model construction module, the uncertainty processing module, the constraint design module and the solution output module are connected with the data processing module respectively.

[0074] The model construction module is used to construct a two-stage robust optimization model, determine a min-max-min structure and an optimization objective.

[0075] The uncertainty processing module is used to construct a wind and light output uncertainty set and an emission factor interval model.

[0076] The constraint design module is used to design a robust carbon emission constraint and a target function.

[0077] The solution output module is used to iteratively solve the model by using a column constraint generation algorithm and output an optimal scheduling scheme.

[0078] Specifically, the two-stage robust method is used to process the uncertainty of wind power and photovoltaic power, the model is a min-max-min structure, and the optimization objective is the minimum scheduling cost in the worst case. The system is modeled as a two-stage optimization model, wherein the first stage is a decision-making strategy layer (operation plan, emission control), and the second stage is an adjusting strategy layer (re-scheduling in the face of source and load fluctuations). Meanwhile, interval uncertainty modeling (i.e. "interval robustness") is embedded in the decision-making process of the first stage to process the interval uncertainty and upper limit control requirement of the carbon emission factor.

[0079] Uncertainty set construction

[0080]

[0081] wherein: and respectively represent the predicted values of wind power and photovoltaic power at t period; τ WT,t and τ PV,t respectively represent the fluctuation states of wind power and photovoltaic power at t period, and take values of 0 or 1; l WT,t and l PV,trespectively represent the fluctuation values of wind power and photovoltaic power in t period; Ω WT and Ω PV represent the uncertainty, representing the number of periods in which wind and light output fluctuations occur in the period.

[0082] Emission factor modeling

[0083] For each type of carbon emission energy equipment i, the carbon emission factor corresponding to the unit output is:

[0084]

[0085] In the formula: ε i represents a random variable with unknown distribution; represents the predicted value; n represents the variation coefficient; ε i represents the fluctuation value.

[0086] Robust carbon emission constraint design

[0087] The total carbon emission calculation model of the system is:

[0088]

[0089] In the formula: represents the total carbon emission in the system; P i,t represents the carbon emission of a certain carbon emission device in t period.

[0090] In order to ensure that it is still not over-standard in the worst case, the upper bound of the interval is modeled, and the following robust emission constraint is constructed:

[0091]

[0092] In the formula: represents the maximum value of the carbon emission in the system.

[0093] 3. Model and solution

[0094] Based on the above uncertain problem, the following two-stage robust optimization model is established to minimize the scheduling cost:

[0095]

[0096] In the formula: k1 and k2 represent weight coefficients, used to balance the economic, low-carbon and robustness targets

[0097] represents the additional regulation cost considering uncertainty.

[0098] The outer layer of the model is the day-ahead economic low-carbon dispatching problem, and the optimization variable x is the day-ahead dispatching scheme, i.e., to find the most economic and low-carbon day-ahead dispatching scheme of IES under the worst scenario; the inner layer is the regulation sub-problem caused by the uncertainty of source and load, and the optimization variable is the regulation scheme y and the worst scenario u, i.e., to find the most economic regulation scheme and the worst scenario under the day-ahead dispatching scheme x.

[0099] Objective function

[0100] The objective function in a single integrated energy system includes the operation cost of units in the system, the interaction cost with the power grid, and the carbon trading cost.

[0101]

[0102] In the formula: represents the operation cost of each unit in the system in the t period; represents the grid interaction cost, which represents the purchase of electricity from the grid when it is greater than 0, otherwise represents the sale of electricity to the grid; represents the carbon emission cost in the system.

[0103] The carbon emission cost in the integrated energy system can be represented as:

[0104]

[0105] In the formula: represents the carbon emission cost in the system; represents the unit carbon emission price; represents the total carbon emission in the system.

[0106] Solution method:

[0107] In order to facilitate the solution, the above model can be transformed into:

[0108]

[0109] In the formula: u represents the predicted value of the load power; c, b represent the coefficient matrix corresponding to the objective function; D, M, F, G, Q represent the coefficient matrix under the corresponding coefficient; m and h represent the operating parameters of the system; x and y represent the optimization variables.

[0110] The column constraint generation algorithm is used for solution, and the original problem is decomposed into the main problem (outer min) and the sub-problem (inner max-min).

[0111] The main problem is:

[0112]

[0113] In the formula: n represents the number of iterations; y ndenotes the solution of the sub-problem after n iterations; denotes the wind power and photovoltaic output value in the worst scenario after n iterations.

[0114] The sub-problem is:

[0115]

[0116] After conversion of the sub-problem by using strong duality theory, the following form is obtained:

[0117]

[0118] In the formula: γ, π, ν are dual variables corresponding to constraints.

[0119] By using Big-M method to linearize the above formula, the following form is obtained:

[0120]

[0121] In the formula: u p T = [u pv p (t), u wt p (t)] denotes the point prediction value of photovoltaic output and wind power output; lu = [d pv (t), d wt (t)] denotes the point prediction value deviation of photovoltaic output and wind power output; J' = [J' pv (t), J' wt (t)] denotes the added continuous auxiliary variable; N denotes the upper boundary of the dual variable.

[0122] As shown in the above, the flow of solving the main problem and the sub-problem is as follows: Figure 2

[0123] (1) Given the initial uncertain situation u1, the initial upper limit U = +∞, the initial lower limit L = -∞, the convergence threshold σ = 0.001, and the initial iteration number n = 1.

[0124] (2) Solve the main problem to obtain the optimal solution (x n , y l ), that is, the minimum operation cost after optimization, and the system state in each period. At the same time, the main problem objective function value is taken as the new lower limit L.

[0125] (3) Substitute the decision variable x n obtained by the main problem into the sub-problem to continue solving, and obtain the objective function value h n (x n ) of the sub-problem and the worst scenario u n+1 ​and update the upper bound U = min{h n (x n ), U}.

[0126] (4) If U - L < σ, output the optimal solution of the master problem; otherwise, continue to increase the control variable y n+1 in the master problem, increase the column constraints, and let n = n + 1, jump to step 2 to continue iteration.

Claims

1. A low-carbon scheduling optimization method based on two-stage robustness and interval carbon constraints, characterized in that, Includes the following steps: Step 1: Construct a two-stage robust optimization model. The model has a min-max-min structure and aims to minimize scheduling costs. The outer layer is the day-ahead economic low-carbon scheduling problem, with the day-ahead scheduling scheme as the optimization variable. The inner layer is the regulation sub-problem caused by source-load uncertainty, with the regulation scheme and worst-case scenario as the optimization variables. Step 2: Construct an uncertainty set for wind and solar power output, considering its predicted values ​​and fluctuations, to describe the uncertainty of wind and solar power output; Step 3: Model the range uncertainty of the emission factors of carbon emission energy equipment, and introduce predicted values, variation coefficients and fluctuation values ​​to characterize the range characteristics of the emission factors. Step 4: Design robust carbon emission constraints, and ensure that the carbon emissions of the system do not exceed the upper limit under the worst-case scenario by modeling the upper bound of the interval. Step 5: Use the column constraint generation algorithm to solve the two-stage robust optimization model, decompose the problem into the main problem and sub-problems, and obtain the most economical control scheme and the worst-case scenario under the day-ahead scheduling scheme.

2. The low-carbon scheduling optimization method based on two-stage robustness and interval carbon constraints according to claim 1, characterized in that, The formula for constructing the uncertain set is: Wherein: and τ represents the predicted values ​​of wind power and solar power respectively during time period t; WT,t and τ PV,t These represent the fluctuation states of wind power and photovoltaic power during time period t, respectively, with values ​​of 0 or 1; WT,t and l PV,t Ω represents the fluctuation values ​​of wind power and photovoltaic power respectively during time period t; WT and Ω PV This indicates uncertainty and represents the number of periods during which wind and solar power output fluctuates within the cycle.

3. The low-carbon scheduling optimization method based on two-stage robustness and interval carbon constraints according to claim 1, characterized in that, The emission factor modeling formula is as follows: Where: ε i Represents a random variable with an unknown distribution; Indicates the predicted value; n represents the coefficient of variation; ε i This indicates the fluctuation value. The low-carbon scheduling optimization method based on two-stage robustness and interval carbon constraints according to claim 1 is characterized in that the robust carbon emission constraint is: The carbon emission calculation model is as follows: In the formula: P represents the total carbon emissions in the system. i,t This represents the carbon emissions of a specific carbon-emitting device during time period t. Robust emission constraints: In the formula: This indicates the maximum carbon emissions within the system.

4. The low-carbon scheduling optimization method based on two-stage robustness and interval carbon constraints according to claim 1, characterized in that, The objective function includes: Two-stage robust optimization model: In the formula: k1 and k2 represent weighting coefficients, used to balance the objectives of economy, low carbon emissions and robustness; This indicates that additional regulatory costs caused by uncertainty have been taken into consideration; The objective function is: In the formula: This represents the operating cost of each unit in the system during time period t; This represents the grid interaction cost; when it is greater than 0, it indicates purchasing electricity from the grid; otherwise, it indicates selling electricity to the grid. Indicates the cost of carbon emissions within the system; The carbon emission cost is: In the formula: This indicates the carbon emission cost within the system; Indicates the price per unit of carbon emissions; This indicates the total carbon emissions within the system.

5. A low-carbon scheduling optimization system based on two-stage robust optimization and interval carbon constraints, characterized in that, The low-carbon scheduling optimization method based on two-stage robustness and interval carbon constraints as described in any one of claims 1-4 includes a model building module, an uncertainty handling module, a constraint design module, a solution output module, and a data processing module. The model building module, uncertainty handling module, constraint design module, and solution output module are respectively connected to the data processing module; The aforementioned model building module is used to construct a two-stage robust optimization model and determine the min-max-min structure and optimization objective. The uncertainty handling module is used to construct the uncertainty set of wind and solar power output and the range model of emission factors; The constraint design module is used to design robust carbon emission constraints and objective functions. The solution output module is used to iteratively solve the model using a column constraint generation algorithm and output the optimal scheduling scheme.

Images