Power system joint planning method and readable storage medium

Abstract

This application provides a joint planning method for power systems, including establishing a shared energy storage joint planning model based on the power system's topology, and using a two-level optimization method to solve the shared energy storage joint planning model to obtain the energy storage system's operation plan. The optimization objective is to optimize the overall economic performance of the power system at the upper level, and the operational performance of the energy storage system in a typical scenario is used as the lower level optimization objective. The shared energy storage joint planning model uses a fuzzy chance-constrained programming method to describe the source-side model of wind turbine output, the load-side model describing the impact of price incentives on user demand load response, and the storage-side model considering the lifespan of the energy storage power station. This application considers the randomness and uncertainty of the power system's source-side and load-side operation, thereby achieving power system dispatch balance, improving energy utilization efficiency, and reducing overall system cost.

Description

Technical Field

[0001] This invention relates to the field of power planning technology for power systems, and specifically to a joint planning method for power systems and a readable storage medium. Background Technology

[0002] As my country's energy system gradually transitions towards cleaner and renewable energy, the proportion of clean electricity in social energy consumption is continuously increasing. A new power system primarily powered by clean energy sources such as wind and solar power will gradually replace the high-carbon-emission power system dominated by traditional fossil fuel power generation. However, the output of new energy sources is characterized by randomness and volatility, which significantly impacts the safe and stable operation of the power grid, mainly including issues such as wind and solar curtailment, peak shaving and frequency regulation, and stability.

[0003] Traditional power grid and energy storage configuration and operation modes do not take into account the flexibility of user-side response, resulting in insufficient utilization of energy storage and user-side resources. Consequently, wind power cannot be effectively absorbed, and the phenomenon of "wind curtailment and solar curtailment" occurs frequently. The lack of coordination among the various power grid modules of "source-grid-load-storage" leads to resource waste and increased equipment investment. Summary of the Invention

[0004] To address the issues of insufficient response to user-side changes and lack of coordination among power system modules in existing power systems, this invention proposes a power system joint planning method and a readable storage medium.

[0005] This invention employs the following power system joint planning method, comprising the following steps:

[0006] S100. Establish a joint planning model for shared energy storage based on the topology of the power system.

[0007] S200. Taking the optimal comprehensive economic performance of the power system as the upper-level optimization objective and the operational performance of the energy storage system in the power system under typical scenarios as the lower-level optimization objective, the two-level optimization method is used to solve the shared energy storage joint planning model to obtain the operation plan of the energy storage system.

[0008] The shared energy storage joint planning model uses grid information flow and energy flow to connect the following sub-models:

[0009] ①: A source-end model for describing the power output of wind turbines based on fuzzy chance-constrained programming;

[0010] ②: Describe the load-side model that shows how price incentives affect user demand load response;

[0011] ③: Storage-end model considering the lifespan of the energy storage power station.

[0012] Specifically, step S100, establishing a joint planning model for shared energy storage based on the topology of the power system, includes:

[0013] S110. Establish a source-end model to indirectly describe the uncertainty of wind turbine output through the output prediction error of the wind turbine: including:

[0014] S111. Using fuzzy theory to predict the power output error ε wind Modeling:

[0015]

[0016] In the formula p' wind p wind These represent the predicted and actual wind power output, respectively. Prediction error has two forms: if the actual output is higher than the predicted output, it is a positive error; if the actual output is lower than the predicted output, it is a negative error.

[0017] S113. Calculate the membership degree μ of the prediction error using the Cauchy distribution. wind :

[0018]

[0019] In the formula, E + wind E - wind σ represents the statistical average of the positive and negative errors, respectively; σ is the weight.

[0020] S115. Establish a reliability measure C based on the prediction error ξ. r (ξ≤ε wind The source-end model of wind turbine output is represented by )

[0021]

[0022] Specifically, the step S100 of establishing a joint planning model for shared energy storage based on the topology of the power system includes:

[0023] S120. Establish the load-end model. The steps include...

[0024] S121. Establish a demand response model based on incentive prices. If the upper bound of the demand response coefficient for load participation in demand-side response is ρ. up The lower bound is ρ down All of them are linearly related to the incentive price x, as expressed below:

[0025]

[0026]

[0027] Then the demand response ρ0∈[ρ] under the incentive price x down ,ρ up ],

[0028] S122. Calculate the total response load S participating in the demand response under the incentive price:

[0029]

[0030] Here, let event j be the user participation request response, and the number of participating users be N. j The incentive price is x j ,ρ(x j ) represents the response coefficient for user participation in demand response, and its value range is ρ(x) j )∈[ρ down , ρ up ], S j The total load demand of a specific user;

[0031] S123. Calculate the total incentive cost C at incentive price x. s :

[0032]

[0033] S125. Establish a model for the impact of incentive prices on load participation in demand response. Load participation in demand response takes three forms: transfer, reduction, and interruption, corresponding to the transferable load model, the reductionable load model, and the interruptible load model, respectively. Divide a day into N load regulation periods T, then:

[0034] Load reduction model:

[0035]

[0036] In the formula, ρ LAr，t S is the coefficient by which the load can be reduced at time t; t P represents the user load capacity at time t; LAr，t The load reduction power at time t; W LAr，T This refers to the amount of load reduction that can be achieved during period T. t LAr and These are the start and end times of the load reduction, respectively; W LAr，T and These represent the lower and upper limits of the capacity that can be reduced during period T, respectively.

[0037] Transferable load model:

[0038]

[0039] In the formula, ρ LAs，tP is the load transfer factor at time t; LAs，t Let t be the load transferable power. t LAs and W represents the start and end times of the load transfer. LAs，T This represents the transferable load during period T. and W LAs，T These represent the upper and lower limits of the load transferable capacity during time period T;

[0040] Interruptible load model:

[0041]

[0042] P LAt，t ρ is the interruptible power of the load at time t; LAt，t Let be the interruptibility coefficient of the load at time t; t LAt and W represents the start and end times of the load interruption. LAt，T The interruptible amount of load during period T; and W LAt，T These represent the upper and lower limits of the interruptible load interruption capacity, respectively.

[0043] 7. The power system joint planning method as described in claim 3, characterized in that the step S100 of establishing a shared energy storage joint planning model based on the power system topology includes:

[0044] S130. Establish a storage-end model to describe the impact of the charging and discharging process on battery life. After n discharge cycles, the actual battery capacity is expressed as:

[0045]

[0046] in,,

[0047] Where i is the count of discharges, Γ A It is a dimensionless coefficient, Γ R Rated lifespan of energy storage system, d i eff The loss during the i-th discharge process:

[0048]

[0049] In the formula, D i P represents the depth of discharge during a non-rated operation of an energy storage system; i ES，dis Indicates D i The corresponding discharge power, I RP represents the discharge current of the energy storage battery under rated conditions. R This represents the power of the energy storage battery under rated conditions, where a, b, and c are all influence coefficients.

[0050] Preferably, step S200, which uses the optimal overall economic performance of the power system as the upper-level optimization objective and the operational performance of the energy storage system in a typical scenario as the lower-level optimization objective, employs a two-level optimization method to solve the shared energy storage joint planning model to obtain the daily operation plan of the energy storage system, includes:

[0051] S210. The objective function for determining the upper-level optimization objective based on comprehensive economic optimization is:

[0052]

[0053] In the formula, F upper C is the objective function of the upper-level decision-making model. d inv、 C d FOM These represent the average daily investment cost and fixed maintenance cost of the energy storage system, respectively; ε sc Let be the probability of scenario sc occurring; The variable maintenance cost of an energy storage system under a typical operating scenario (SC); The operating cost of the power grid under typical operating scenario sc; To delay the returns on investment in power grid equipment upgrades;; C s Cost of responding to demand;

[0054] S220. Determine the objective function for optimizing the lower-level objective, including:

[0055] Optimization objective 1 is to minimize the variance of the net load of the power grid, expressed as:

[0056]

[0057] Optimization objective 2 is to minimize the square of the change in net power grid load, expressed as:

[0058]

[0059] Optimization objective 3 is to maximize the power support level that the energy storage system can provide during grid faults. The objective function is:

[0060]

[0061] In the formula, This serves as backup power for the energy storage system.

[0062] Preferably, step S200, which uses the optimal overall economic performance of the power system as the upper-level optimization objective and the operational performance of the energy storage system in a typical scenario as the lower-level optimization objective, employs a two-level optimization method to solve the shared energy storage joint planning model to obtain the daily operation plan of the energy storage system, includes:

[0063] S230. Using a multi-group hierarchical chaotic differential particle swarm optimization as the basic architecture, solve the upper-level decision model and solve for the optimal comprehensive operating cost.

[0064] S240. Use the fuzzy satisfaction maximization method to optimize the three objectives at the lower level; obtain the linear membership degree of the multi-objectives, and use the reduced half-trapezoidal membership function curve to obtain the linear membership degree of the three objective functions of the lower-level decision model respectively, so as to evaluate the satisfaction of the corresponding objective functions, calculate the maximum fuzzy satisfaction degree, and use it as the objective function of the lower-level decision model to obtain the optimal operation plan of the energy storage system under different typical operation scenarios.

[0065] This application also provides a readable storage medium for the above-described method, storing a computer program that, when executed by a processor, causes the processor to perform the steps of the above-described power system joint planning method. This invention considers the randomness and uncertainty of the operation of the power system's source and load ends, employs a fuzzy chance-constrained programming method to describe the uncertainty of wind power output, considers the impact of price incentives on the load end, and simultaneously considers the lifespan and cost of energy storage, thereby achieving joint planning of the power system's "source-grid-load-storage". By using a two-layer solution algorithm on the joint planning model, the system's economics and the energy storage system's operation plan are solved step-by-step, reducing algorithm complexity and improving the algorithm's practicality. This achieves power system scheduling balance, reduces resource losses, improves energy utilization efficiency, and reduces overall system cost. Attached Figure Description

[0066] Figure 1 This is an overall flowchart of the technical solution of this application;

[0067] Figure 2 This is a structural diagram of the power system of this application;

[0068] Figure 3 This is a typical daily load curve for a power system in summer;

[0069] Figure 4 It is the relationship curve between the price incentive curve and the demand impact coefficient;

[0070] Figure 5 This is a schematic diagram illustrating the composition of the demand-response curve;

[0071] Figure 6 The solution process for the upper-level decision-making model;

[0072] Figure 7 The solution process for the lower-level decision-making model. Detailed Implementation

[0073] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0074] Figure 1 This is the overall flowchart of the present application. The power system joint planning method of the present application includes the following steps:

[0075] S100. Establish a joint planning model for shared energy storage based on the topology of the power system; this joint planning model includes:

[0076] ①: A source-end model for describing the power output of wind turbines based on fuzzy chance-constrained programming;

[0077] ②: Describe the load-side model that shows how price incentives affect user demand load response;

[0078] ③: Storage-end model considering the lifespan of the energy storage power station.

[0079] S200. Taking the optimal comprehensive economic performance of the power system as the upper-level optimization objective and the operational performance of the energy storage system in the power system under typical scenarios as the lower-level optimization objective, the two-level optimization method is used to solve the shared energy storage joint planning model to obtain the operation plan of the energy storage system.

[0080] Figure 2This is a schematic diagram of a power system. In a real power system network, the source, load, and storage ends transmit energy and information through the power grid. The addition of storage ends enables shared energy storage within the grid. The source end, or power generation side, includes traditional thermal and hydropower units, as well as emerging renewable energy sources such as photovoltaic power plants and wind turbines. The load end, or power consumption side, is characterized by the dispersed nature of users and significant differences in their production and lifestyles. Therefore, the electricity demand of a single user exhibits randomness within a certain range, but statistically, the influence of specific factors on the overall direction of user demand is deterministic. The storage end, or energy storage side, is necessary because the power generation capacity of wind turbines and photovoltaic power plants is significantly affected by the environment, impacting the energy balance of the power grid. During peak power generation periods, the energy storage side absorbs excess electricity from the grid; during off-peak periods, it releases electricity to maintain grid balance. The grid end, or power grid, completes energy transfer between the source, load, and storage ends through transmission and transformation equipment. A dispatch center is also set up within the power grid to send control information to the source and storage ends, regulating the operation of each component. In other words, the joint planning model in this application connects the aforementioned source, load, and storage models through the energy and information flow at the grid end.

[0081] The dispatch center is primarily responsible for information flow interaction, controlling the information exchange among the four components of "source-grid-load-storage." By scheduling energy flow through information flow, it achieves real-time energy interaction and balance, thereby improving energy utilization efficiency. Furthermore, based on energy flow conditions, the dispatch center implements appropriate price incentives to encourage user participation in demand response within a certain range. For example, during peak load periods, price subsidies can be used to encourage users to reduce electricity consumption.

[0082] Renewable energy units such as wind turbines are given priority for grid connection. During peak wind power generation periods, if wind power cannot be fully connected to the grid, the surplus power is transmitted to energy storage power stations at low cost for backup.

[0083] The power grid interacts with renewable energy generators, energy storage, and conventional loads in real time through the dispatch center, facilitating information and energy flow. During peak load periods, priority is given to absorbing renewable energy into the grid, purchasing and storing electrical energy from energy storage stations, and encouraging users to actively participate in demand response, with multiple measures taken to reduce grid peak load. During off-peak periods, surplus power generation is sold to energy storage stations at low prices, eliminating the need for renewable energy grid connection and user participation in demand response.

[0084] Energy storage power stations store low-cost electricity from the grid and renewable energy units during off-peak periods and supply electricity to the grid during peak periods. The "low-load storage, high-load" arbitrage model, ignoring other factors, is theoretically illogical. Energy storage assists the grid in peak shaving, enhances system stability, and improves energy efficiency. Simultaneously, energy storage can act as a backup power source to supply users when the grid fails. When the grid cannot absorb users' distributed photovoltaic power generation, energy storage can also absorb surplus power generated by users.

[0085] Conventional load refers to the electricity load of general users. During peak electricity consumption periods, users can actively participate in demand response within a certain range and obtain certain revenue by taking into account the price incentives issued by the dispatch center and their own circumstances. At the same time, users can also choose to generate electricity through distributed generation devices and sell it to the grid or energy storage power stations to obtain certain revenue.

[0086] For example, taking a 110kV line as the research object, assuming the configuration of one 1MW energy storage power station and several distributed energy storage systems, the analysis is conducted using a typical summer day, divided into 24 time periods per hour. The benchmark electricity price for a typical summer day is the 10kV industrial and commercial time-of-use electricity price for a certain region during summer (July to September), which is 1.15 yuan / kWh during peak, flat, and low periods, respectively. The incentive price adopts an incentive scheme with a subsidy price of 0.3 yuan / kWh during peak periods, 0.2 yuan / kWh during flat periods, and 0.1 yuan / kWh during low periods. The load curve for a typical summer day is as follows: Figure 3 As shown.

[0087] The steps for establishing a joint planning model for shared energy storage in S100 of this application include the following steps.

[0088] S110. Establish the source-end model.

[0089] As mentioned earlier, the power generation capacity of renewable energy generating units in the power grid is affected by the environment and fluctuates significantly. To characterize the volatility and randomness of renewable energy sources, this technical solution uses a planning method based on fuzzy chance constraints to model the source end.

[0090] Taking wind turbines as an example, wind turbine output generally refers to its power generation. The randomness of wind turbine output can be transformed into the randomness of its output prediction error. That is, assuming the predicted value of wind turbine output is certain, then the randomness of wind turbine output and the randomness of output prediction error (i.e., the error between the predicted value and the actual output) are equivalent. Therefore, we can start from the perspective of output prediction error, use fuzzy theory to model the prediction error, and indirectly describe the uncertainty of wind turbine output through the uncertainty of wind turbine output prediction error.

[0091] Fuzzy chance-constrained programming can be expressed as:

[0092] M[g(x,ξ)≤0]≥α,

[0093] In the formula: α is the confidence level; g is the set of constraints; ξ is the fuzzy variable; x is the decision variable; M is the measure function of event A, where M(A) = C r (ξ∈A)

[0094]

[0095] In the formula, μ is the membership function of the fuzzy variable ξ, and sup indicates taking the upper bound.

[0096] Therefore, the steps for modeling wind turbines using the fuzzy chance constraint planning method in the source-end model based on S110 are as follows:

[0097] S111, Prediction error ε of wind turbine output wind Modeling includes:

[0098]

[0099] In the formula p' wind p wind These represent the predicted and actual wind power output, respectively. Prediction error has two forms: if the actual output is higher than the predicted output, it is a positive error; if the actual output is lower than the predicted output, it is a negative error.

[0100] S113. Calculate the membership degree μ of the prediction error using the Cauchy distribution. wind :

[0101]

[0102] In the formula, E + wind E - wind σ represents the statistical average of the positive and negative errors, respectively; σ is the weight.

[0103] S115. Establish a source-end model for wind turbine output, expressed as the reliability measure Cr (ξ≤εwind) of the prediction error ξ:

[0104]

[0105] The constraints of this model are:

[0106] C r (ξ≤ε wind )≥α.

[0107] S120. Establish the load-end model.

[0108] Since different user needs have different levels of demand rigidity, user needs can change within a certain range when incentivized by price. However, users respond differently to price incentives under different levels of demand rigidity. Therefore, it is necessary to classify and establish price incentive models.

[0109] S121. Establish a demand response model based on incentive prices. For example... Figure 3 As shown, the price incentive curve for the load-side user is illustrated, determined by the upper bound of the response coefficient ρ. up With response coefficient lower bound ρ down The triangular region enclosed by the incentive price, its boundary, and the line segment CE constitute the user's response space. When the incentive price is 0, the user has a certain response space AD, but it is highly uncertain, manifested by the maximum length of AD. As the incentive price increases, the demand response coefficient increases, and users tend to reduce load usage, reducing the randomness of load increases and the fluctuation range. When the incentive price reaches a critical point B, the user can guarantee not to increase energy load. Point B is determined by the maximum cost of load reduction / transfer / interruption, and the incentive price at the maximum of these three is X0 at point B. When the incentive price reaches a saturation point C, the user's demand response coefficient is at its maximum, and the fluctuation range is approximately negligible, representing the maximum effective incentive intensity of the model. Point C is determined by the cost at which the response coefficient is at its maximum.

[0110] The upper bound ρ of the response coefficient in the figure up With response coefficient lower bound ρ down for:

[0111]

[0112]

[0113] S122. Calculate the total response load S participating in the demand response under the incentive price:

[0114]

[0115] Here, let event j be the user participation request response, and the number of participating users be N. j The incentive price is x j ,ρ(x j ) represents the response coefficient for user participation in demand response, and its value range is ρ(x) j )∈[(ρ down , ρ up ], S j This represents the total load demand of a specific user.

[0116] S123. Calculate the total incentive cost C at the incentive price. s :

[0117]

[0118] S125. Establish a model of the impact of incentive prices on load participation in demand response. Load participation in demand response takes three forms: transfer, reduction, and interruption, corresponding to the transferable load model, the reductionable load model, and the interruptible load model, respectively. Dividing a day into N load regulation periods T, then:

[0119] Load reduction model:

[0120]

[0121] In the formula, ρ LAr，t Let be the load reduction factor at time t, and the curve of the reduction factor is as follows: Figure 5 As shown in (a); S t P represents the user load capacity at time t; LAr，t The load reduction power at time t; W LAr，T This refers to the amount of load reduction that can be achieved during period T. t LAr and These are the start and end times of the load reduction, respectively; W LAr，T and These represent the lower and upper limits of the capacity that can be reduced during period T, respectively.

[0122] Transferable load model:

[0123]

[0124] In the formula, ρ LAs，t The curves for the transferability coefficient and the reduction coefficient of the load at time t are shown below. Figure 5 (b) shows; P LAs，t Let t be the load transferable power. t LAs and W represents the start and end times of the load transfer. LAs，T This represents the transferable load during period T. and W LAs，T These represent the upper and lower limits of the load transferable capacity during time period T; the load transfer time should not be less than the shortest duration T. min .

[0125] Interruptible load model:

[0126]

[0127] Then the incentive price x j The total demand response coefficient at the load end is a linear combination of the response coefficients under three conditions:

[0128] ρ(x j )=λ·ρL Ar，t +μ·ρ LAs，t +ν·ρ LAt，t ,(λ+μ+ν=1).

[0129] Where λ, μ, and ν are undetermined coefficients, determined by statistical analysis of users' electricity consumption habits, and P LAt，t ρ is the interruptible power of the load at time t; LAt，t Let be the interruptibility factor of the load at time t. The curve of the interruptibility factor is shown below. Figure 5 As shown in (c); t LAt and The start and end times of the load interruption; W LAt，T The interruptible amount of load during period T; and W LAt，T These represent the upper and lower limits of the interruptible load interruption capacity; the load transfer time should not be less than the shortest duration T. min .

[0130] Figure 4 The synthesized demand response curve is a combination of demand response curves for reduceable / transferable / interruptible loads. Figure 5 (a) It is evident that the upper limit of cost for demand response to load reduction is low, the demand response coefficient without incentive prices is also lower than the composite curve, the maximum value of the demand response coefficient is low, and the cost at the maximum response coefficient is also low. From Figure 5 (b) It is evident that the demand response coefficient of transferable loads without price incentives is slightly lower than that of the synthesized curve. The maximum demand response cost of transferable loads is the same as that of the synthesized curve, while the maximum demand response coefficient of transferable loads is slightly lower than that of the synthesized curve. The cost corresponding to the upper limit of the response coefficient is higher than that of the synthesized curve. From Figure 5 (c) It can be seen that the demand response coefficient of interruptible loads without price incentives is the same as the synthesized curve. The maximum demand response cost of interruptible loads is lower than that of the synthesized curve, the maximum demand response coefficient is higher than that of the synthesized curve, and the cost corresponding to the upper limit of the response coefficient is lower than that of the synthesized curve.

[0131] S130. Establish the storage-side model.

[0132] In an energy storage system that uses chemical batteries for energy storage, the real-time charge / discharge capacity of the energy storage battery during the charging and discharging process is:

[0133]

[0134] The constraints for the charging and discharging process are:

[0135]

[0136] Let Δt = 1, then the charging and discharging power of the energy storage system at time t-1. Therefore, the charging and discharging power of the energy storage system during a certain period is numerically equal to the charging and discharging capacity. This represents the energy storage device's charge level at time t. This indicates that the energy storage is in a discharging state at time t-1, and vice versa; η is the charge / discharge efficiency of the energy storage battery. Constraints are applied to the charge / discharge of the energy storage system at the current time t, where S... socmax and S socmin Indicates the upper and lower limits of the energy storage state of charge; P b.max -P represents the maximum discharge power of the stored energy. b.max This represents the maximum charging power for energy storage.

[0137] The number of cycles during operation significantly impacts battery lifespan, particularly the lifespan of energy storage systems. Considering the charging and discharging behavior of energy storage devices directly reduces battery life. Under rated operating conditions, the total discharge capacity over the entire lifespan of an energy storage system is defined as the total effective discharge capacity G. R (Ah), there is

[0138] G R =L R D R C R .

[0139] Among them, L R D represents the cycle life (cycles) of energy storage under rated conditions; R Indicates the depth of discharge under rated conditions; C R This indicates the rated capacity of the energy storage (Ah).

[0140] In actual power grid operation, the lifespan degradation of energy storage batteries operating under off-rated conditions is closely related to their depth of discharge and discharge rate. This can be determined by calculating the discharge capacity d of the energy storage battery under off-rated conditions. a , will d a Equivalent to the discharge capacity d of the energy storage battery under rated conditions eff And the d of each discharge process within the power grid operating cycle eff The values ​​are superimposed to determine whether the total effective discharge capacity G has been reached. R When equal to G R When this time, it indicates that the energy storage battery is no longer usable.

[0141] The cycle life LA of an energy storage battery during operation is related to the actual depth of discharge D of the energy storage battery:

[0142] L A =a×D -b ×e -cD ,

[0143] In the formula, a, b, and c are the influence coefficients, and a>0, b>0, and c>0. It can be seen that the cycle life LA is negatively correlated with the depth of discharge D.

[0144] On the other hand, the discharge capacity d of the energy storage battery under non-rated conditions a and the discharge capacity d of the energy storage battery under rated conditions eff The relationship is:

[0145]

[0146] Since the power and current of the energy storage system exhibit a consistent trend when the power grid operates in grid-connected mode, there is a relationship between the discharge rate and the degree of energy storage lifespan loss.

[0147]

[0148] Among them, I R P represents the discharge current of the energy storage battery under rated conditions. R Indicates the power of the energy storage battery under rated conditions; I ES This indicates the discharge current of the energy storage battery under non-rated conditions; PE S,dis Indicates with I ES The corresponding discharge power.

[0149] Integrating the above relationships, we can obtain the battery capacity loss model:

[0150]

[0151] In the formula, i represents the number of discharges, and D... i This indicates the depth of discharge during a specific energy storage operation under non-rated conditions. and They represent D respectively i The corresponding discharge power and discharge quantity.

[0152] Based on this, the battery's capacity can be obtained. The discharge process of the battery under non-rated conditions is transformed into a discharge process under rated conditions. Each discharge action causes battery loss. Assuming n discharge actions occur within the operating cycle, the actual battery capacity is:

[0153]

[0154] Where i is the count of discharges, Γ A It is a dimensionless coefficient, Γ R Rated lifespan of the energy storage system.

[0155] It should be noted that the modeling processes of S110, S120, and S130 in this application are modeling different subsystems, and therefore there is no requirement for the order of the steps.

[0156] As mentioned above, a dispatch center is also set up in the power grid to send control information to the source and storage ends, thereby controlling the operation of each part. That is, in the joint planning model of this application, the energy flow and information flow at the grid end (including the dispatch center) connect the aforementioned source-end model, load-end model, and storage-end model. Therefore, the grid connection relationship determines a set of boundary conditions between the source end, storage end, and load end. These boundary conditions and other constraints are as follows:

[0157] Power balance constraints of the power grid:

[0158]

[0159]

[0160] Grid node voltage constraints:

[0161]

[0162] Power transmission constraints of power grid branches:

[0163]

[0164] Timing constraints of power recursion relationship in energy storage system:

[0165]

[0166]

[0167] Daily charge-discharge cycle constraints of energy storage systems:

[0168]

[0169] Maximum charging (discharging) power constraints for energy storage systems:

[0170]

[0171] Energy storage system charge / discharge depth constraints:

[0172]

[0173] Energy storage system charging (discharging) power constraints during peak and off-peak hours:

[0174]

[0175]

[0176] User demand response constraints:

[0177] User demand response constraints include

[0178] Response capacity constraints

[0179]

[0180] Response coefficient constraints

[0181]

[0182] And reliability constraints.

[0183] 0≤ρ≤α.

[0184] For the established shared energy storage joint planning model, the grid's dispatching capacity at the load end is limited. Load demand is determined by users and can be influenced by incentive prices, but cannot be directly controlled by the grid. Regarding source output, we always aim to maximize output under permissible conditions; therefore, dispatching the source end in grid dispatching is simple and easy. The key to grid dispatching planning lies in the energy storage system's operation plan. This application employs a two-level optimization method to solve for the energy storage system's operation plan. The specific optimization objectives are: the upper-level optimization objective is to achieve the optimal overall economic performance of the power system, and the lower-level optimization objective is to assess the operational performance of the energy storage system within the power system under typical scenarios. The steps include:

[0185] S210. The objective function for determining the upper-level optimization objective based on comprehensive economic optimization is expressed as:

[0186]

[0187] In the formula, F upper C is the objective function of the upper-level decision-making model. d inv、 C d FOM These represent the average daily investment cost and fixed maintenance cost of the energy storage system, respectively; ε sc Let be the probability of scenario sc occurring; The variable maintenance cost of an energy storage system under a typical operating scenario (SC); The operating cost of the power grid under typical operating scenario sc; To delay the returns on investment in power grid equipment upgrades;; C s Cost of responding to demand;

[0188] S220. Determine the objective function for optimizing the lower-level objective.

[0189] The objective function of the lower-level decision-making is mainly based on the role of energy storage systems in peak shaving and valley filling and smoothing power fluctuations during the normal operation of new energy generating units; and in the event of a system failure, it aims to improve the power support provided by the energy storage system to the grid, thereby summarizing and formulating a fuzzy multi-objective optimization operation strategy. Therefore, it involves multiple optimization objectives.

[0190] Optimization Goal 1:

[0191] Peak shaving and valley filling can delay grid upgrades and renovations. Energy storage systems can achieve load peak shaving and valley filling by effectively charging and discharging between load peaks and valleys. On the one hand, peak-valley price differences can be used to reduce operating costs; on the other hand, they can delay grid upgrades and renovations due to load growth to a certain extent. In this embodiment, a day is divided into 24 time periods. Of course, dividing a day into other appropriate segments is also acceptable in other implementations. Based on this, one of the optimization objectives in the lower-level optimization algorithm is to minimize the net load variance of the power grid, expressed as:

[0192]

[0193] In the formula For typical power grid load scenarios, This represents the average power grid load under typical scenarios.

[0194] Optimization Goal 2:

[0195] New energy consumption and fluctuation mitigation. Rational scheduling of energy storage system charging and discharging power can mitigate the combined effects of renewable energy generation system output power and load fluctuations to a certain extent, minimizing the fluctuation level of power exchange between the distribution system and the upstream grid. Therefore, minimizing the square of the net load change of the grid is one of the optimization objectives, expressed as:

[0196]

[0197] In the formula The load at time t is a typical scenario. This represents the average load at the previous moment in a typical scenario.

[0198] Optimization Goal 3:

[0199] Improving reliability and energy storage backup capacity. The introduction of energy storage systems can provide power support for some critical loads in the grid during grid failures, improving the grid's power supply reliability and participating in fault recovery to a certain extent. To represent the power support level that the energy storage system can provide during grid failures, the backup support capacity of the energy storage system is used to reflect the maximum power support provided by the energy storage system to the grid under the operating constraints at time t. The objective function is:

[0200]

[0201] In the formula, This serves as backup power for the energy storage system.

[0202] S230. For the upper-level optimization objective of optimizing the overall economic efficiency of the power system, a multi-group hierarchical chaotic differential particle swarm optimization (MPS) algorithm is used as the basic framework to solve for the optimal overall operating cost. Generally speaking, when using the multi-group hierarchical chaotic differential particle swarm optimization algorithm to solve for the upper-level optimal overall operating cost, the input parameters include the average daily investment cost of the energy storage system, the daily operation and maintenance cost of the energy storage system, the daily operating cost of the system under typical scenarios, and relevant parameters related to delaying grid transformation benefits. Figure 6 The chaotic differential particle swarm optimization algorithm obtains the maximum and minimum values ​​of each component of the objective function, thereby calculating the optimal overall operating cost.

[0203] S240. Use the fuzzy satisfaction maximization method to optimize the three objectives at the lower level.

[0204] When two or more objective functions are considered simultaneously, a multi-objective optimization problem is formed. In the lower-level decision model, the three optimization objectives also have contradictory relationships: when the energy storage system participates in peak shaving and valley filling to smooth out fluctuations, it inevitably reduces the system's performance, leading to a decrease in its backup support capacity. Therefore, it is necessary to effectively address the three lower-level objectives to obtain a more balanced overall optimal solution.

[0205] In response, this application uses the fuzzy satisfaction maximization method based on fuzzy mathematics theory to process the three optimization objectives of the lower-level decision model, and uses the fuzzy satisfaction maximization as the optimization objective of the lower-level decision model to find the multi-objective linear membership degree. The linear membership degree of the three objective functions of the lower-level decision model is obtained by using the descending half-trapezoidal membership function curve to evaluate the satisfaction of the corresponding objective functions, and the maximum fuzzy satisfaction is calculated as the objective function of the lower-level decision model.

[0206] Based on this, a fuzzy multi-objective chaotic particle swarm optimization algorithm is used to obtain the optimal operating plan for the energy storage system under different typical operating scenarios. For example... Figure 7 As shown, the system inputs basic data of new energy generating units, power grid, energy storage, and load, as well as candidate solutions obtained from the upper-level optimization algorithm. The system operation data is solved by particle swarm optimization, and the population fitness value is obtained by power flow calculation. The population is updated considering indicators such as new energy consumption, load peak shaving and valley filling, and power grid reserve support capacity to solve for the optimal operation plan.

[0207] Considering that the chaotic differential particle swarm optimization algorithm used to solve the upper-level optimization objective and the fuzzy multi-objective chaotic particle swarm optimization algorithm used to solve the lower-level optimization objective are commonly used algorithms in the field of multi-objective optimization, and the algorithm steps are well known, this application will not provide a detailed description. However, it should be considered that those skilled in the art can optionally apply the above algorithms to the solution of the technical solution of this application based on the content disclosed in this specification.

[0208] The traditional energy storage configuration method is compared with the shared energy storage configuration method of the present invention. The three indicators of comprehensive benefits, renewable energy consumption and energy utilization rate are compared and the results are shown in Table 1.

[0209]

[0210] Table 1: Comparison of Traditional Energy Storage Configuration Methods and Shared Energy Storage Configuration Methods

[0211] Shared energy storage requires additional consideration of demand response subsidies, dispatch costs, and transformation and upgrading costs, resulting in higher construction costs than traditional energy storage. However, its overall benefits, renewable energy absorption capacity, and energy utilization efficiency are all higher than traditional energy storage. It can effectively mitigate the impact of renewable energy grid connection, improve grid stability, enhance the grid's ability to absorb renewable energy, and effectively improve energy utilization efficiency, with significant results.

[0212] In summary, this invention considers the uncertainty of power output from the source to the load. Taking the power output fluctuation of wind turbines as an example, it uses a fuzzy chance-constrained programming method to describe wind power output. It explores the potential for proactive user response, establishing a demand response model that allows for the transfer, reduction, or interruption of loads. Through price incentives, it controls user participation in demand response within a certain range, assisting the grid in peak shaving. It also considers factors such as the charging, discharging, and lifespan loss of energy storage power stations to establish a mathematical model. Based on the above models, it establishes and solves a four-in-one shared energy storage joint planning model integrating the source, grid, load, and storage systems. This aims to enhance the grid's renewable energy absorption capacity, reduce wind curtailment, and improve energy utilization efficiency.

[0213] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0214] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0215] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0216] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0217] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A joint planning method for power systems, characterized in that, Includes the following steps: S100. Establish a joint planning model for shared energy storage based on the topology of the power system. The steps in S100, establishing a joint planning model for shared energy storage based on the topology of the power system, include: S110. Establish a source-end model to indirectly describe the uncertainty of wind turbine output through the output prediction error of the wind turbine: including: S111. Using fuzzy theory to predict the power output error ε wind Modeling: ， In the formula p ' wind p wind These represent the predicted and actual wind power output, respectively. Prediction error has two forms: if the actual output is higher than the predicted output, it is a positive error; if the actual output is lower than the predicted output, it is a negative error. S113. Calculate the membership degree μ of the prediction error using the Cauchy distribution. wind : ， In the formula, E + wind E - wind σ represents the statistical average of the positive and negative errors, respectively; σ is the weight. S114. Establish a reliability measure C based on the prediction error ξ. r (ξ≤ε wind The source-end model of wind turbine output is represented by ) ； The steps of S100 in establishing a joint planning model for shared energy storage based on the topology of the power system include: S120. Establish the load-end model. The steps include: S121. Establish a demand response model based on incentive prices. If the upper bound of the demand response coefficient for load participation in demand-side response is ρ. up The lower bound is ρ down All of them are linearly related to the incentive price x, as expressed below: ， ； Then the demand response ρ0∈[ρ] under the incentive price x down ,ρ up ], S122. Calculate the total response load S participating in the demand response under the incentive price: ， Let event j be the user participation request response, and let N be the number of participating users. j The incentive price is x j , The response coefficient for user participation in demand response, its value range is: S j The total load demand of a specific user; S123. Calculate the total incentive cost C at incentive price x. s : ； S124. Establish a model for the impact of incentive prices on load participation in demand response. Load participation in demand response takes three forms: transfer, reduction, and interruption, corresponding to the transferable load model, the reductionable load model, and the interruptible load model, respectively. Divide a day into N load regulation periods T, then: Load reduction model: ， In the formula, ρ LAr，t S is the coefficient by which the load can be reduced at time t; t P represents the user load capacity at time t; LAr，t The load reduction power at time t; W LAr，T This refers to the amount of load reduction that can be achieved during period T. and These are the start and end times of the load reduction, respectively; and These represent the lower and upper limits of the capacity that can be reduced during period T, respectively. Transferable load model: ， In the formula, ρ LAs，t P is the load transfer factor at time t; LAs，t Let t be the load transferable power. and W represents the start and end times of the load transfer. LAs，T This represents the transferable load during period T. and These are the upper and lower limits of the load transferable capacity during time period T; Interruptible load model: ， P LAt，t ρ is the interruptible power of the load at time t; LAt，t Let be the interruptibility coefficient of the load at time t; and W represents the start and end times of the load interruption. LAt，T The interruptible amount of load during period T; and These are the upper and lower limits of the interruptible load interruption capacity, respectively. The steps of S100 in establishing a joint planning model for shared energy storage based on the topology of the power system include: S130. Establish a storage-end model to describe the impact of the charging and discharging process on battery life. After n discharge cycles, the actual battery capacity is expressed as: ， Where i is the count of the number of discharges, It is a dimensionless coefficient. The rated lifespan of an energy storage system The loss during the i-th discharge process: ， In the formula, D i This indicates the depth of discharge during a specific energy storage operation under non-rated conditions. Indicates D i The corresponding discharge power, I R P represents the discharge current of the energy storage battery under rated conditions. R This represents the power of the energy storage battery under rated conditions, where a, b, and c are all influence coefficients. S200. Taking the optimal comprehensive economic performance of the power system as the upper-level optimization objective and the operational performance of the energy storage system in the power system under typical scenarios as the lower-level optimization objective, the two-level optimization method is used to solve the shared energy storage joint planning model to obtain the operation plan of the energy storage system. The shared energy storage joint planning model uses grid information flow and energy flow to connect the following sub-models: A source-end model for wind turbine output is described based on fuzzy chance-constrained programming. A load-side model describing how price incentives affect user demand load response; A storage-end model considering the lifespan of an energy storage power station; The steps of S200, which take the optimal comprehensive economic performance of the power system as the upper-level optimization objective and the operational performance of the energy storage system in a typical scenario as the lower-level optimization objective, to solve the shared energy storage joint planning model using a two-level optimization method to obtain the daily operation plan of the energy storage system, include: S210. The objective function for determining the upper-level optimization objective based on comprehensive economic optimization is: ， In the formula, F upper C is the objective function of the upper-level decision-making model. d inv、 C d FOM These represent the average daily investment cost and fixed maintenance cost of the energy storage system, respectively; ε sc Let be the probability of scenario sc occurring; The variable maintenance cost of an energy storage system under a typical operating scenario (SC); The operating cost of the power grid under typical operating scenario sc; To delay the returns on investment in power grid equipment upgrades; Cs represents demand response costs; S220. Determine the objective function for optimizing the lower-level objective, including: Optimization objective 1: Minimize the net load variance of the power grid; Optimization objective 2: Minimize the square of the change in net power grid load; Optimization objective 3: to maximize the power support level that the energy storage system can provide during grid faults; The steps of S200, which take the optimal comprehensive economic performance of the power system as the upper-level optimization objective and the operational performance of the energy storage system in a typical scenario as the lower-level optimization objective, to solve the shared energy storage joint planning model using a two-level optimization method to obtain the daily operation plan of the energy storage system, include: S230. Using a multi-group hierarchical chaotic differential particle swarm optimization as the basic architecture, solve the upper-level decision model and solve for the optimal comprehensive operating cost. S240. Use the fuzzy satisfaction maximization method to optimize the three objectives at the lower level; obtain the linear membership degree of the multi-objectives, and use the reduced half-trapezoidal membership function curve to obtain the linear membership degree of the three objective functions of the lower-level decision model respectively, so as to evaluate the satisfaction of the corresponding objective functions, calculate the maximum fuzzy satisfaction degree, and use it as the objective function of the lower-level decision model to obtain the optimal operation plan of the energy storage system under different typical operation scenarios.

2. A computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the power system joint planning method as described in claim 1.

Images