A double-layer optimization method for energy storage participating in low-carbon flexible peak regulation

Abstract

A two-layer optimization method for energy storage to participate in low-carbon flexible peak shaving is proposed. S1. Improve the carbon emission metering model of thermal power units. S2. Propose a system risk cost function for energy storage flexible peak shaving. S3. Construct a two-layer optimization model for energy storage to participate in low-carbon flexible peak shaving: The upper-layer model aims to minimize system carbon emissions, load peak-valley difference rate, and energy storage operating costs, determining the energy storage charging and discharging power at each time point. The lower-layer model aims to minimize load shedding risk costs, wind curtailment risk costs, and thermal power unit regulation risk costs, using the energy storage charging and discharging power obtained from the upper layer as constraints to determine the energy storage reserve capacity at each time point. S4. Optimize control. This invention, considering the system carbon emissions and load peak-valley difference rate in the upper-layer energy storage power optimization model and the system operating risk-constraining lower-layer energy storage reserve capacity optimization model, provides a reference for the low-carbon and flexible operation of the system.

Description

Technical Field

[0001] This invention belongs to the field of low-carbon flexible peak shaving in power systems, specifically relating to a two-layer optimization method for energy storage to participate in low-carbon flexible peak shaving. Background Technology

[0002] The increasing proportion of renewable energy sources such as wind and solar power in the power system has significantly increased the difficulty of peak shaving due to the randomness, volatility, and anti-peak-shaving characteristics of their output. Traditional hydropower units are limited by seasonality, water inflow, and reservoir regulation capacity, making it difficult to meet peak-shaving demands. If thermal power units participate in peak shaving, frequent adjustments or start-ups and shutdowns will increase coal consumption and equipment maintenance costs, thereby increasing carbon emissions. Therefore, new system equipment or operating modes are needed to achieve low-carbon and flexible peak shaving.

[0003] Energy storage devices possess bidirectional response capabilities and are easy to control, providing an effective method for low-carbon and flexible peak shaving in power systems. The combined peak shaving of fused magnesium load and energy storage can effectively reduce the net load peak-to-valley difference and the total system operating cost. Day-ahead and intraday optimized scheduling applicable to renewable energy and adiabatic compressed air energy storage can be achieved through robust online operation schemes to calibrate the energy storage linear model. Optimizing the operation of wind-storage systems can address the problem of poor wind power fluctuation smoothing in these systems. The aforementioned models link energy storage operation with renewable energy output or electricity load regulation for joint scheduling; however, energy storage can actually participate in power system operation as an independent operating entity, such as grid-side energy storage.

[0004] Currently, many countries and regions have successively constructed grid-side energy storage systems. In countries such as the United States, Australia, and the United Kingdom, grid-side energy storage is mainly used to participate in the frequency regulation market. There are also many demonstration projects in China, such as the energy storage power station demonstration project in Jinjiang, Fujian, which provides peak-shaving and frequency regulation services to local substations; the energy storage demonstration project in the Guangdong power grid, which mainly addresses power constraints caused by power construction limitations in some areas; and Hunan, which has abundant hydropower and good frequency regulation capabilities, but suffers from a large peak-valley load difference, where the energy storage power station in Changsha adopts a two-charge-two-discharge operation mode daily to meet the peak-shaving needs of Changsha's electricity load during the "midday peak" and "evening peak," with relatively fixed charging and discharging times. However, the application of grid-side energy storage in these regions is limited to solving problems existing in their respective regional power grids, and its full potential has not been realized.

[0005] Most technologies for independently operated energy storage focus on capacity configuration and economic analysis. For example, by analyzing the differences between above-ground and underground geographical locations for energy storage configuration, a two-layer collaborative optimization method for underground energy storage capacity configuration is proposed. If energy storage participates in smoothing load fluctuations during non-peak-shaving periods, this operating mode can improve the economics of energy storage operation and shorten the investment payback period. By proposing benefit and cost indicators for combining energy storage with deep peak-shaving thermal power units to participate in system flexibility peak-shaving, both the overall system benefits and wind curtailment rate are considered. However, few technologies emphasize the role of independently operated energy storage in reducing wind power forecasting error risk and system carbon emissions when participating in power system peak-shaving. System carbon emission measurement needs a more precise description, and measuring the contribution of a particular component to system carbon emissions is beneficial for the subsequent operation of the carbon market and carbon trading. When energy storage is an independently operated entity, its contribution to system carbon emissions needs to be measured.

[0006] Therefore, energy storage is considered an independent operating entity, and its charging and discharging power is included in the carbon emission metering of thermal power units. The value of the system carbon emission intensity reduced by energy storage is related to the carbon emission intensity of thermal power units. Furthermore, if energy storage is equipped with a certain reserve capacity during operation, it can reduce the system operation risk caused by the uncertainty of wind power. Under the condition of reducing the economics of energy storage to a certain extent, it can increase the system reserve capacity and improve the economics of power system operation. Summary of the Invention

[0007] To address the problems existing in the prior art, this invention proposes a two-layer optimization method for energy storage to participate in low-carbon flexible peak shaving, which can effectively reduce the total operating cost of the power system, while giving full play to the peak shaving role of energy storage in the system, and reducing the risk of system carbon emissions and wind power prediction errors.

[0008] This invention provides a two-layer optimization method for energy storage to participate in low-carbon flexible peak shaving, characterized by the following steps:

[0009] S1. Improve the carbon emission measurement model of thermal power units, quantify the contribution of energy storage to emission reduction in the system, describe in detail the carbon emissions generated by the output of thermal power units, and propose a carbon emission measurement model that considers the charging and discharging power of energy storage.

[0010] S2. A system risk cost function for flexible peak shaving of energy storage is proposed, including the wind power operation risk cost function and the thermal power unit regulation risk cost function;

[0011] S3. Construct a two-layer optimization model for energy storage to participate in low-carbon flexible peak shaving:

[0012] The upper-level model aims to minimize system carbon emissions, load peak-valley difference rate, and energy storage operating costs, and determines the energy storage charging and discharging power at each moment.

[0013] The lower-level model aims to minimize the risks and costs of load shedding, wind curtailment, and thermal power unit regulation, using the energy storage charging and discharging power obtained from the upper-level model as constraints to determine the energy storage reserve capacity at each time point. Based on the wind and solar power output forecast time series and load forecast curves, day-ahead dispatching plans are optimized.

[0014] S4. Optimized Control:

[0015] The model reduces system carbon emissions while lowering the load peak-valley difference, and measures the contribution of energy storage to emission reduction; by setting energy storage reserve capacity, it improves the system's peak-shaving flexibility and reduces system operation risks.

[0016] As a further improvement of the present invention, step S1 of improving the carbon emission measurement model of thermal power units includes:

[0017] The carbon emission intensity of coal-fired power generation in thermal power units is negatively correlated with the unit load; that is, the higher the unit load, the lower the carbon emission intensity of coal-fired power generation. The linear relationship between the output of thermal power units and the carbon emission intensity of coal-fired power generation is as follows:

[0018]

[0019] In the formula: Let g be the carbon emission intensity of coal-fired power generation unit g at time t; k≥0, b≥0 are the parameters of the linear function; P represents the output of thermal power unit g at time t; g Let g be the installed capacity of thermal power unit;

[0020] The carbon emissions of coal-fired power generation unit g Calculate according to the above formula;

[0021]

[0022] In the formula: Δt is the time interval;

[0023] Incorporating the reduced system carbon emissions from energy storage charging and discharging power, the improved carbon emission measurement model for thermal power units is obtained as follows:

[0024]

[0025] In the formula: T is the number of time periods; N G γ represents the number of thermal power units; t Defined as the system carbon emission intensity reduced by the charging and discharging power of energy storage units, it is related to the current carbon emission intensity of thermal power units; N B This refers to the number of energy storage units. and , respectively, represent the charging and discharging power of energy storage i at time t.

[0026] As a further improvement to the present invention, step S1 of improving the carbon emission measurement model of thermal power units also includes:

[0027] The charging and discharging power model of energy storage changes the output of thermal power units, which in turn changes the carbon emission intensity per unit of thermal power unit output, thereby changing the overall carbon emissions of the system.

[0028] (1)γ t Upper limit of values

[0029] Because energy storage charging and discharging involves certain power losses, the carbon emissions deducted in the improved carbon emission metering model for thermal power units should be less than the carbon emissions reduced by energy storage from coal-fired power generation. Therefore, in the improved carbon emission metering model for thermal power units, γ t The value should be less than the carbon emission intensity of coal-fired power generation units, while also ensuring that the energy storage absorbs wind power. If the energy storage stores thermal power, it will lead to F c The first and second terms increase simultaneously, and the coefficient of the first term is greater than that of the second term, resulting in F c Increase;

[0030]

[0031] (2)γ t Lower bound of the value

[0032] Since the active power output of other units in the system during energy storage discharge is less than that of the units without energy storage, the system's carbon emissions will not increase during energy storage discharge. t The value of should be greater than 0, that is:

[0033] γ t ≥0.

[0034] As a further improvement of the present invention, in step S3, the two-layer optimization model first assigns initial values ​​to the variables of the upper-layer model, and the lower-layer model optimizes its objective function based on this. The result is returned to the upper-layer optimization objective. The two layers alternately iterate, and finally obtain the result of global interest equilibrium. The upper-layer problem is solved by the immune genetic algorithm, and the lower-layer problem is solved by calling the CPLEX solver in MATLAB and the point estimation method.

[0035] Compared with the prior art, the advantages of the present invention are as follows:

[0036] 1. The improved carbon emission metering proposed in this embodiment first improves the accuracy of carbon emission metering by incorporating the negative correlation between the carbon emission intensity of coal-fired power generation and the unit load into the carbon emission metering model. Furthermore, it includes the energy storage charging and discharging power in the system carbon emission metering, and the reduction in system carbon emission intensity per unit of energy storage charging and discharging power is related to the current carbon emission intensity of the thermal power unit.

[0037] 2. This embodiment describes the system risk cost of energy storage participating in flexible peak shaving through wind power prediction error distribution, and provides a wind power operation risk cost function including energy storage reserve capacity. When energy storage has a certain reserve capacity, it can significantly reduce system operation risk, that is, reduce the risk of load shedding and wind curtailment, reduce the adjustment risk of thermal power units, and make system peak shaving more flexible.

[0038] 3. In the two-layer optimization model of this embodiment, the upper-layer model focuses on describing the ability of energy storage to reduce carbon emissions and improve peak shaving, while the lower-layer model focuses on the peak shaving economy and flexibility of energy storage. The decision variables of the upper and lower layers, namely charging and discharging power and energy storage reserve capacity, are mutually constrained, providing an operating mode for energy storage that simultaneously considers the system's low carbon emissions and flexible peak shaving. Attached Figure Description

[0039] Figure 1 This is a probability distribution diagram of wind power output prediction error in this embodiment;

[0040] Figure 2 This is a flowchart of the iterative solution process for the model in this embodiment;

[0041] Figure 3 The results are the optimized energy storage discharge depths for different scenarios and time periods described in the specific application embodiments.

[0042] Figure 4 The results are the energy storage optimization results for each time period in Scenario 1 of the specific application embodiment;

[0043] Figure 5 The results are the energy storage optimization results for each time period in Scenario 3 of the specific application embodiment;

[0044] Figure 6 In specific application embodiments, different γ t The curve of energy storage output change when the value is reached;

[0045] Figure 7 In specific application embodiments, different γ t The output curve of the thermal power unit at the specified value. Detailed Implementation

[0046] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0047] like Figure 1 As shown in the figure, a two-layer optimization method for energy storage to participate in low-carbon flexible peak shaving in this embodiment includes the following steps:

[0048] S1. Improve the carbon emission measurement model of thermal power units, describe in detail the carbon emissions generated by the output of thermal power units, and propose a carbon emission measurement model that takes into account the charging and discharging power of energy storage.

[0049] S2. A system risk cost function for flexible peak shaving of energy storage is proposed, including the wind power operation risk cost function and the thermal power unit regulation risk cost function;

[0050] S3. Construct a two-layer optimization model for energy storage to participate in low-carbon flexible peak shaving: The upper-layer model aims to minimize system carbon emissions, load peak-valley difference rate, and energy storage operating costs, determining the energy storage charging and discharging power at each time point. The lower-layer model aims to minimize load shedding risk costs, wind curtailment risk costs, and thermal power unit regulation risk costs, using the energy storage charging and discharging power obtained from the upper layer as constraints to determine the energy storage reserve capacity at each time point. Based on the wind and solar power output forecast time series and load forecast curves, day-ahead dispatching plans are optimized.

[0051] S4. Optimized Control: The model can reduce system carbon emissions while lowering the load peak-valley difference, and measure the contribution of energy storage to emission reduction. By setting energy storage reserve capacity, the system's peak-shaving flexibility is improved, and system operation risks are reduced. The upper-level energy storage power optimization model, which considers system carbon emissions and load peak-valley difference rate, and the lower-level energy storage reserve capacity optimization model, which considers system operation risks, are mutually constrained, providing a reference for the low-carbon and flexible operation of the system.

[0052] Probability density function of wind power output prediction error It follows a pattern with a mean of 0 and a variance of σ. 2 normal distribution [23,24] ,like Figure 1 As shown, Let represent the upper and lower bounds of the wind power prediction error at time t when the confidence level is β. If the system has adjustable resources, namely thermal power units and energy storage devices, the adjustable capacity is the range of wind power prediction error when the confidence level is β. and When the wind power prediction error at time t is less than At time t, the system's adjustable resources are insufficient, requiring the shedding of some loads; when the wind power prediction error at time t is higher than... At that time, measures such as wind curtailment need to be taken. Figure 1 The probability weighted value of the shaded area can represent the load shedding and wind curtailment risk costs when the wind power is running at time t.

[0053] The two-level optimization model first assigns initial values ​​to the variables of the upper-level model, and then optimizes its objective function based on these initial values. The result is returned to the upper-level optimization objective. The two levels iterate alternately until a global equilibrium result is obtained. An immune genetic algorithm is used to solve the upper-level problem, and the CPLEX solver in MATLAB and the point estimation method are used together to solve the lower-level problem.

[0054] Once the output of conventional power sources and the output of energy storage are determined, the backup capacity of energy storage is optimized with the objective of minimizing system risk and cost. Therefore, a two-level optimization model is used to describe the energy storage optimization decision problem, with the convergence condition being:

[0055]

[0056] In the formula, Let ε represent the reserve capacity of energy storage i at time t during the k-th iteration, where ε is a sufficiently small positive number. The iterative solution process is as follows: Figure 2 As shown.

[0057] In step S1 of this embodiment, the carbon emission measurement model of thermal power units is improved, the carbon emissions generated by the output of thermal power units are characterized in detail, and a carbon emission measurement model that takes into account the charging and discharging power of energy storage is proposed.

[0058] The carbon emission intensity of coal-fired power generation is negatively correlated with the unit load; that is, the higher the unit load, the lower the carbon emission intensity of coal-fired power generation. The linear functional relationship between the output of thermal power units and the carbon emission intensity of coal-fired power generation is as follows:

[0059]

[0060] In the formula: Let g be the carbon emission intensity of coal-fired power generation unit g at time t; k≥0, b≥0 are the parameters of the linear function; P represents the output of thermal power unit g at time t; g Let g be the installed capacity of the thermal power unit.

[0061] The carbon emissions of coal-fired power generation unit g It can be calculated using the following formula.

[0062]

[0063] In the formula: Δt is the time interval.

[0064] Because energy storage incurs energy losses during charging and discharging, system carbon emissions are minimized when wind and thermal power units can meet load demands. However, during certain periods, significant load reductions and limitations in the ramp-up capabilities of thermal power units can lead to substantial wind curtailment. Energy storage can reduce system carbon emissions by incorporating wind power during charging or reducing thermal power unit output during discharging. Furthermore, energy storage devices can reduce carbon emissions by optimizing charging and discharging processes to increase the output of low-coal-consumption units and decrease the output of high-coal-consumption units.

[0065] By incorporating the reduced system carbon emissions from energy storage charging and discharging power, the improved carbon emission measurement model for thermal power units can be obtained as follows:

[0066]

[0067] In the formula: T is the number of time periods; N G γ represents the number of thermal power units; t Defined as the system carbon emission intensity reduced by the charging and discharging power of energy storage units, it is related to the current carbon emission intensity of thermal power units; N B This refers to the number of energy storage units. and , respectively, represent the charging and discharging power of energy storage i at time t.

[0068] The charging and discharging power of energy storage changes the output of thermal power units, which in turn changes the carbon emission intensity per unit of output of thermal power units, thereby changing the overall carbon emissions of the system.

[0069] 1.γ t Upper limit of values

[0070] Since energy storage charging and discharging involves certain power losses, the carbon emissions deducted in the above formula should be less than the carbon emissions reduced by energy storage from coal-fired power generation. Therefore, in the above formula, γ t The value should be less than the carbon emission intensity of coal-fired power generation units. This also ensures that the energy storage absorbs wind power; if the energy storage stores thermal power, it will lead to F... c The first and second terms increase simultaneously, and the coefficient of the first term is greater than that of the second term, resulting in F c Increase.

[0071]

[0072] 2.γ t Lower bound of the value

[0073] Since the active power output of other units in the system during energy storage discharge is less than that of the units without energy storage, the system's carbon emissions will not increase during energy storage discharge. t The value of should be greater than 0, that is:

[0074] γ t ≥0

[0075] In step S2 of this embodiment, a system risk cost function for flexible peak shaving of energy storage is proposed, including a wind power operation risk cost function and a thermal power unit regulation risk cost function:

[0076] Because wind power output is random and fluctuating, thermal power units and energy storage devices need to adjust their power output to maintain a balance between power generation and consumption in the system. The capacity of the system's adjustable resources determines the maximum range of wind power that the system can accept. When wind power output exceeds the maximum range that the system can accept, it will bring operational risks to the system and cause economic losses.

[0077] Probability density function of wind power output prediction error It follows a pattern with a mean of 0 and a variance of σ. 2 The normal distribution, such as Figure 1 As shown, Let represent the upper and lower bounds of the wind power prediction error at time t when the confidence level is β. If the system has adjustable resources, namely thermal power units and energy storage devices, the adjustable capacity is the range of wind power prediction error when the confidence level is β. and When the wind power prediction error at time t is less than At time t, the system's adjustable resources are insufficient, requiring the shedding of some loads; when the wind power prediction error at time t is higher than... At that time, measures such as wind curtailment need to be taken. Figure 1 The probability weighted value of the shaded area can represent the load shedding and wind curtailment risk costs when the wind power is running at time t.

[0078] The wind power operation risk cost of the system can then be expressed as:

[0079]

[0080] Where: N W Indicates the number of wind farms; and These represent the maximum and minimum power generation output of the wind farm w, respectively; μ W μ D These represent the cost coefficients for wind curtailment and load shedding, respectively. Let w be the wind power output of the wind farm at time t.

[0081] Adding energy storage to the system increases the capacity of the system's adjustable resources, significantly reducing wind power operation risk costs. Maintaining the state of charge (SOC) of the energy storage within the desired range helps address potential wind power operation risks. The wind power operation risk cost at time t, including the adjustable capacity of energy storage, is:

[0082]

[0083] In the formula: This represents the charging and discharging reserve capacity of energy storage i at time t.

[0084] 2. Risk-cost function for thermal power unit regulation

[0085] The power imbalance caused by the uncertainty of wind power output can be balanced not only by energy storage but may also require regulation by thermal power units. The fuel costs and regulation costs arising from the uncertainty of wind power output, i.e., the regulation risk costs of thermal power units, are...

[22] :

[0086]

[0087] In the formula: and These are the maximum upward ramp power, maximum downward ramp power, upward ramp power, and downward ramp power of thermal power unit g at time t; C g (·) represents the power generation cost function of thermal power unit g at time t. a g b g c g d is the power generation cost coefficient for thermal power unit g; g is the unit power regulation cost coefficient for thermal power unit g.

[0088] In step S3 of this embodiment, a two-layer optimization model for energy storage to participate in low-carbon flexible peak shaving is constructed: the upper-layer model aims to minimize system carbon emissions, load peak-valley difference rate, and energy storage operating costs, determining the energy storage charging and discharging power at each time point. The lower-layer model aims to minimize load shedding risk costs, wind curtailment risk costs, and thermal power unit regulation risk costs, using the energy storage charging and discharging power obtained from the upper layer as constraints to determine the energy storage reserve capacity at each time point. Day-ahead scheduling is performed based on the wind and solar power output forecast time series and load forecast curves.

[0089] The upper-level objective function aims to minimize system carbon emissions, load peak-valley difference rate, and energy storage operating costs, and determines the energy storage charging and discharging power at each moment.

[0090] As a further improvement of the present invention, the energy storage participation low-carbon flexible peak shaving dual-layer optimization model is shown in the following equation:

[0091] 1. Upper-level objective function

[0092]

[0093] In the formula: k1, k2, k3 are weighting coefficients, with values ​​of 100, 1, and 1 respectively; β, F b F c These are the load peak-valley difference rate, the energy storage operation cost function, and the carbon emission measurement function of thermal power units, respectively.

[0094] Energy storage discharges during peak load periods and charges during off-peak periods, thus altering the equivalent load of the system. Let the power resulting from the sum of the load power and the energy storage power be:

[0095]

[0096] Where: N L The number of nodes; Let be the load power of node k at time t.

[0097] The load peak-valley difference rate of the equivalent load is:

[0098]

[0099] The energy storage operating cost function is:

[0100]

[0101] In the formula: c b,i Let be the unit power regulation cost coefficient of energy storage i.

[0102] 2. Lower-level objective function

[0103] Optimize the backup capacity of energy storage devices at all times in the system, and adjust the ramp-up output of thermal power units to minimize the system's operating risk and cost, including wind power operating risk and cost and thermal power unit adjustment risk and cost.

[0104]

[0105] As a further improvement of the present invention, the dual-layer optimization model for energy storage participation in low-carbon flexible peak shaving established in step S3 is further provided with upper-layer optimization model constraints and lower-layer optimization model constraints.

[0106] 1. Constraints of the upper-level optimization model

[0107] 1) System power balance constraints:

[0108]

[0109] 2) Wind power output constraints:

[0110]

[0111] 3) Output constraints of thermal power units:

[0112]

[0113] In the formula: and These represent the minimum and maximum output of the thermal power unit, respectively.

[0114] 4) Gradient constraints for thermal power units:

[0115]

[0116] In the formula: r u,g r d,g Let g be the maximum uphill / downhill climbing rate of the thermal power unit.

[0117] 5) Start-up and shutdown constraints for thermal power units:

[0118]

[0119] In the formula: ug,j This is a Boolean variable representing the start-up / shutdown state of thermal power unit g at time j, where 0 indicates shutdown and 1 indicates startup; T g,on T g,off These represent the maximum continuous start-up time and the maximum continuous shutdown time of thermal power unit g, respectively.

[0120] 6) Energy storage charging and discharging power constraints:

[0121]

[0122] In the formula: P i c,max and P i d,max These are the maximum charging and discharging power of energy storage i, respectively.

[0123] 7) Energy storage state of charge constraints:

[0124]

[0125] In the formula: S i,t S represents the SOC value of energy storage i during time period t; i,max and S i,min These represent the upper and lower limits of the energy storage SOC, respectively; σ i Represents the self-discharge rate of energy storage; η c,i and η d,i These represent the energy storage charging and discharging efficiency, respectively.

[0126] 2. Constraints of the lower-level optimization model

[0127] Provided that the constraints of the upper-level optimization model are satisfied, the variables to be solved in the lower-level model must satisfy the following constraints. After the first iteration, when solving the upper-level model, it is necessary to satisfy the energy storage reserve capacity constraint and the state of charge constraint containing the energy storage reserve capacity. That is, the energy storage reserve capacity of the lower-level model will limit the upper and lower limits of the state of charge of the energy storage in the upper-level model.

[0128] 1) Adjustable power constraints for thermal power units:

[0129] The maximum adjustable power of a thermal power unit g at time t cannot exceed its maximum ramp rate, and its output cannot exceed its maximum and minimum generating power.

[0130]

[0131] 2) Energy storage backup capacity constraints:

[0132]

[0133]

[0134] In the formula: Vi max Let i be the capacity of energy storage.

[0135] 3) State of charge constraints including energy storage reserve capacity:

[0136] To ensure that energy storage can provide a certain reserve to cope with risks at time t, a certain amount of charging and discharging space needs to be reserved. Therefore, the following state-of-charge constraint with energy storage reserve capacity is proposed:

[0137]

[0138] In step S4 of this embodiment, the model can reduce system carbon emissions while lowering the load peak-valley difference, and measure the contribution of energy storage to emission reduction. By setting the energy storage reserve capacity, the system's peak-shaving flexibility is improved, and system operation risks are reduced. The upper-level energy storage power optimization model, which considers the system carbon emissions and the load peak-valley difference rate, and the lower-level energy storage reserve capacity optimization model, which considers the system operation risks, mutually constrain each other, providing a reference for the low-carbon and flexible operation of the system.

[0139] To verify the effectiveness of this invention, an IEEE 30-node, 6-unit system was selected for verification. The system comprises six identical 100MW coal-fired power units, one 200MW wind farm, and one 100MW energy storage system. The maximum and minimum grid-connected wind power outputs are 200MW and 100MW, respectively. Detailed parameters of the power units and energy storage are shown in Tables 1 and 2. The carbon emission intensity curves of a coal-fired power unit in Shandong under different load rates were used to obtain a linear function of the power unit output and the carbon emission intensity of coal-fired power generation through linear regression.

[0140]

[0141] Table 1 Operating parameters of coal-fired power units

[0142]

[0143] Table 2 Operating parameters of the energy storage system

[0144]

[0145] The remaining simulation parameters are as follows: the wind power prediction error follows a normal distribution with a mean of 0 and a variance of 64, with a confidence interval of [-16, 16]; the wind power operation risk parameter μ W =μ D =100; Thermal power unit start-up and shutdown parameters T g,on =T g,off =4; Energy storage charge / discharge efficiency parameter σ = 0.9, η c,i =η d,i =0.9, initial soc =0.5; Δt =1.

[0146] Three operating scenarios were set up for comparative analysis. Under the same system parameters, the optimization results are shown in Table 3. Scenario 1 is the model in this paper; Scenario 2 is the lower-level model where energy storage does not participate in reserve capacity optimization; Scenario 3 is where energy storage participates in reserve capacity optimization, but carbon emission measurement in the upper-level optimization does not include energy storage power.

[0147] Table 3 Comparison of optimization results in different scenarios

[0148]

[0149] As shown in Table 3, the system carbon emissions in scenarios 1 and 2 are slightly lower than those in scenario 3. The difference in system carbon emissions between scenarios 1 and 3 represents the contribution of energy storage charging and discharging power to the system's carbon emissions. This is because carbon emission measurement in scenario 3 does not include energy storage power. Including energy storage power in the model's carbon emission measurement describes the actual carbon emissions of the system. Unlike scenario 3, the model proposed in this paper, namely scenario 1, considers system carbon emissions, which necessitates maximizing energy storage power. This reduces reserve capacity and increases wind power operation risk costs. Larger energy storage power also reduces the need for thermal power unit regulation, thus reducing thermal power unit regulation risk costs.

[0150] In Table 3, the system carbon emissions in Scenario 2 are lower than those in Scenario 1 because the wind power output is more uncertain, and the reserved energy storage backup capacity will lead to the energy storage not being fully utilized, resulting in redundancy. Scenario 2, because it does not have reserved energy storage backup capacity, has a larger energy storage charging and discharging power, the highest energy storage operating cost, the smallest peak-valley difference rate, and the greatest system operation risk cost.

[0151] Figure 3 In scenarios 1 and 3, the depth of charge and discharge is less than that in scenario 2 during energy storage operation because energy storage needs to reserve a certain amount of backup capacity, resulting in a more conservative operation.

[0152] The optimization results of energy storage in scenarios 1 and 3 at different time periods are as follows: Figure 4 and Figure 5 As shown. Positive and negative values ​​for reserve capacity represent upper and lower reserves. Comparing scenarios 1 and 3 with scenario 2, it's clear that scenario 2 has a higher risk of system load shedding and wind curtailment. This is because energy storage lacks reserve capacity, and the output of thermal power units is limited by ramp-up restrictions or economic factors, failing to provide sufficient reserves. In scenarios 1 and 3, energy storage participates in peak-shaving reserves, providing some backup support and reducing system operational risks. For example, in... Figure 4 and Figure 5 During periods of low load and high wind power output, energy storage provided upward reserve during periods 1, 2, and 3, and downward reserve during periods of high load and low wind power output, specifically periods 12, 13, and 14.

[0153] In the carbon emission measurement model proposed in this paper, γ t The range of values ​​is When γ t When = 0, it corresponds to simulation scenario 3 in this paper. This is simulation scenario 1 in this paper. If γ t Pick When r > 1, the model has no solution. r takes values ​​of 0.1, 0.2, ..., 0.9. Figure 6 and Figure 7 Different γ values ​​are given t The energy storage output and the thermal power unit output at t=13 are compared. At this time, the load is relatively large, the wind power output is relatively small, and the energy storage is in a discharging state.

[0154] Depend on Figure 6 and Figure 7 It can be seen that as the value of r increases, the energy storage discharge power increases linearly, while the output of the thermal power unit decreases linearly, and the increase in energy storage discharge power is equal to the decrease in the output of the thermal power unit. This is because the larger r is, the greater the carbon emission intensity of the system with reduced energy storage discharge power. To make the objective function F... c When the minimum value is reached, the load is constant, the energy storage discharge power will increase, while the output of the thermal power unit will decrease.

[0155] Similarly, during energy storage charging, the larger the value of r, the greater the energy storage charging power, and the higher the output of the thermal power unit will be. However, increasing the output of the thermal power unit will reduce its carbon emission intensity, therefore the objective function F... c It is still possible to reach the optimal state, and as r increases, F c The value of γ will decrease. When regulating energy storage, this can be achieved by increasing the value of r, i.e., γ. t The value can increase the weight of its carbon emission reduction capacity, making energy storage dispatch strategies more aggressive.

[0156] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other way. Any modifications or equivalent changes made based on the technical essence of the present invention shall still fall within the scope of protection claimed by the present invention.

Claims

1. A two-layer optimization method for energy storage to participate in low-carbon flexible peak shaving, characterized in that, The steps include: S1. Improve the carbon emission measurement model of thermal power units, quantify the contribution of energy storage to emission reduction in the system, characterize the carbon emissions generated by the output of thermal power units in detail, and propose a carbon emission measurement model that considers the charging and discharging power of energy storage. S2. Propose a system risk cost function for flexible peak shaving of energy storage, including a wind power operation risk cost function and a thermal power unit regulation risk cost function; The step S1 of improving the carbon emission measurement model of thermal power units includes: The carbon emission intensity of coal-fired power generation in thermal power units is negatively correlated with the unit load; that is, the higher the unit load, the lower the carbon emission intensity of coal-fired power generation. The linear relationship between the output of thermal power units and the carbon emission intensity of coal-fired power generation is as follows: ； In the formula: For thermal power units exist Carbon emission intensity of coal-fired power generation at any given time; The parameters are linear function parameters; For thermal power units exist Efforts made at all times; For thermal power units The installed capacity; thermal power units Carbon emissions from coal-fired power generation Calculate according to the above formula; ； In the formula: For time intervals; Incorporating the reduced system carbon emissions from energy storage charging and discharging power, the improved carbon emission measurement model for thermal power units is obtained as follows: ； In the formula: Number of time periods; This represents the number of thermal power units. Defined as the system carbon emission intensity reduced by the charging and discharging power of energy storage units, it is related to the carbon emission intensity of thermal power units at the current moment. This refers to the number of energy storage units. and Energy storage exist The charging and discharging power at any given moment; The charging and discharging power model of energy storage changes the output of thermal power units, which in turn changes the carbon emission intensity per unit of thermal power unit output, thereby changing the overall carbon emissions of the system. (1) The upper limit of the possible values; Because energy storage charging and discharging involves certain power losses, the carbon emissions deducted in the improved carbon emission metering model for thermal power units should be less than the carbon emissions reduced by energy storage from coal-fired power generation. Therefore, in the improved carbon emission metering model for thermal power units... The value should be less than the carbon emission intensity of coal-fired power generation units, while also ensuring that the energy storage absorbs wind power. If the energy storage stores thermal power, it will lead to... The first and second terms increase simultaneously, and the coefficient of the first term is greater than that of the second term, leading to Increase; ； (2) The lower limit of the value; Since the active power output of other units in the system during energy storage discharge is less than that of the units without energy storage, the system's carbon emissions will not increase during energy storage discharge. The value of should be greater than 0, that is: ； S3. Construct a two-layer optimization model for energy storage to participate in low-carbon flexible peak shaving: The upper-level model aims to minimize system carbon emissions, load peak-valley difference rate, and energy storage operating costs, and determines the energy storage charging and discharging power at each moment. The lower-level model aims to minimize the risk costs of load shedding, wind curtailment, and thermal power unit regulation. It uses the energy storage charging and discharging power obtained from the upper level as a constraint to determine the energy storage reserve capacity at each time point. Based on the wind and solar power output forecast time series and load forecast curve, it optimizes the day-ahead dispatch plan. S4. Optimized Control: The model reduces system carbon emissions while lowering the load peak-valley difference, and measures the contribution of energy storage to emission reduction; by setting energy storage reserve capacity, it improves the system's peak-shaving flexibility and reduces system operation risks.

2. The two-layer optimization method for energy storage participating in low-carbon flexible peak shaving according to claim 1, characterized in that, In step S3, the two-layer optimization model first assigns initial values ​​to the variables of the upper-layer model, and then optimizes the objective function of the lower-layer model based on these values. The result is returned to the upper-layer optimization objective. The two layers iterate alternately until the result of global interest equilibrium is obtained. The upper-layer problem is solved by using an immune genetic algorithm, and the lower-layer problem is solved by calling the CPLEX solver in MATLAB and the point estimation method.

Images