A fast and slow bi-level strategy optimization method for virtual power plant considering carbon price signal and network constraints

Abstract

This invention relates to a fast-slow two-layer strategy optimization method for virtual power plants that considers electricity price, carbon price signals, and network constraints. The method includes the following steps: Based on the virtual power plant's carbon market compliance cycle, including free carbon allowances, historical carryover carbon allowances, initial carbon inventory, market carbon price, and predicted conventional adjustable unit power generation during this cycle, an equivalent carbon price signal for intraday scheduling is calculated. Based on the equivalent carbon price signal, electricity price signal, operating parameters of various resources within the virtual power plant, and network constraint parameters, an optimization model is constructed with the goal of minimizing overall operating costs and constrained by resource operating limitations and power flow transmission limitations. This model is then used to solve for the time-series scheduling instructions for each resource's output or power consumption plan. This invention unifies the three major elements of electricity, carbon, and grid within an optimization framework, outputting instructions that conform to market principles, meet carbon emission reduction requirements, and comply with grid physical security rules, achieving synergistic optimization of economy, low carbon emissions, and safety.

Description

Technical Field

[0001] This invention relates to the interdisciplinary field of virtual power plant optimal scheduling and electricity market operation, and in particular to a fast-slow two-layer strategy optimization method for virtual power plants that considers electricity price and carbon price signals as well as network constraints. Background Technology

[0002] With the deepening of the energy transition, various types of resources, such as distributed photovoltaics, energy storage devices, adjustable conventional and adjustable generating units, and flexible loads, are widely connected on the distribution and user sides. Virtual Power Plants (VPPs), as an important technological form for aggregating heterogeneous distributed resources, can achieve unified modeling, coordinated control, and market trading of geographically dispersed and characteristically diverse resources. This enhances the capacity for renewable energy absorption, strengthens grid regulation flexibility, and promotes coordinated and optimized operation of power generation, grid, load, and storage.

[0003] Against the backdrop of the ongoing development of my country's electricity market, the scope of continuous operation of provincial spot markets is expanding, and the integration of medium- and long-term, spot, and ancillary service markets is accelerating. Virtual power plants have become an important emerging market participant. Simultaneously, the national carbon emissions trading market is becoming increasingly sophisticated, with the power generation industry being incorporated into carbon market management. Carbon costs are gradually becoming a key variable influencing energy system operation decisions. In this context, the optimal dispatch of virtual power plants needs to simultaneously respond to electricity market price signals and carbon market constraints to achieve a balance between economic efficiency and low carbon emissions.

[0004] However, existing virtual power plant dispatching methods still have significant shortcomings: First, most methods only consider electricity price signals and fail to effectively incorporate carbon price signals into the dispatching cost system, making it difficult to balance operational economics and carbon emission reduction targets; second, they do not adequately consider the synergistic relationships among various resources such as photovoltaics, energy storage, conventional adjustable units, flexible loads, and grid power purchase and sale, and lack full exploration of the overall optimization capabilities of virtual power plants as aggregators; in addition, existing methods often lack a unified description and coordination method for the operating mechanism under the dual background of "electricity market price signals + carbon market compliance constraints," making it difficult to adapt to the development needs of my country's current market environment.

[0005] Therefore, it is necessary to propose a virtual power plant optimization method for the coordinated environment of the electricity-carbon market. This method can introduce network physical constraints under the combined effect of electricity price and carbon price signals, and adopt a fast and slow two-layer coordination mechanism adapted to the slow variable characteristics of the carbon market, thereby effectively improving the low-carbon nature, economy and safety of virtual power plant operation.

[0006] It should be noted that the information disclosed in the background section above is only for understanding the background of this application, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0007] The technical problem this application aims to solve is: "the multi-objective collaborative optimization problem of virtual power plants in a dual electricity-carbon market environment, which makes it difficult to simultaneously consider economic efficiency, low carbon emissions, and grid security."

[0008] The technical solution adopted in this application to solve the above-mentioned technical problems is as follows.

[0009] This application provides a fast-slow two-level strategy optimization method for virtual power plants that considers electricity price and carbon price signals as well as network constraints, including the following steps: Slow-stage carbon market coordination steps: Based on the virtual power plant's carbon market compliance cycle: free carbon allowances, historical carryover carbon allowances, initial carbon inventory, market carbon price, and the predicted power generation of conventional adjustable units during the cycle, an equivalent carbon price signal for intraday dispatch is calculated and generated. The fast-level power market optimization scheduling steps are as follows: Based on the equivalent carbon price signal, the electricity price signal of the power market, the operating parameters of each resource within the virtual power plant, and the network constraint parameters of the distribution network, an optimization model is constructed; the optimization model aims to minimize the overall operating cost of the virtual power plant, and is constrained by the operating limitations of each resource within the virtual power plant and the power flow transmission limitations of the distribution network; the optimization model is solved to obtain the output or consumption plans of each resource within the virtual power plant in multiple scheduling periods, which serve as time-series scheduling instructions; The virtual power plant's internal resources include distributed renewable energy units, conventional adjustable units, energy storage devices, flexible adjustable loads, and main grid trading interfaces; the comprehensive operating cost includes the carbon emission cost calculated from the equivalent carbon price signal and the carbon emissions of conventional adjustable units.

[0010] In some embodiments, the slow-layer carbon market coordination steps specifically include: obtaining free carbon allowances, historical carryover carbon allowances, initial carbon inventory, market carbon price, and predicted conventional adjustable unit power generation; and calculating the expected total carbon emissions for the period based on the emission factors of conventional adjustable units and the predicted conventional adjustable unit power generation. The total available carbon allowances for the calculation period are the sum of free carbon allowances, historical carryover carbon allowances, and initial carbon inventory. The emission gap between the expected total carbon emissions for the period and the total available carbon allowances for the period is calculated; based on the emission gap and the market carbon price, an equivalent carbon price signal is generated through a risk adjustment function.

[0011] In some embodiments, an equivalent carbon price signal is generated through a risk adjustment function based on the emission gap and the market carbon price. Specifically, if the emission gap is less than or equal to zero, the equivalent carbon price signal is the market carbon price or a benchmark carbon price determined based on the market carbon price; if the emission gap is greater than zero, the gap ratio is calculated, and the equivalent carbon price signal is calculated through a risk adjustment function based on the market carbon price and the gap ratio. The equivalent carbon price signal increases as the gap ratio increases.

[0012] In some embodiments, the risk adjustment function includes a linear function, a piecewise linear function, a quadratic function, or an exponential function.

[0013] In some embodiments, the constraints of the optimization model include at least one of the following: renewable energy output constraints; conventional adjustable unit output constraints, start-up and shutdown constraints, and ramping constraints; energy storage power constraints and state of charge constraints; flexible load adjustment constraints; power balance constraints; main grid trading constraints; and network power flow constraints based on power transmission distribution factor or linear distribution power flow model.

[0014] In some embodiments, the electricity price signal includes at least one of the following: medium- and long-term contract electricity price, day-ahead spot electricity price, intraday spot electricity price, retail electricity price, grid connection settlement price, agent purchase electricity price, or green electricity transaction settlement price.

[0015] In some embodiments, the overall operating cost also includes at least one of the following costs: conventional adjustable unit fuel cost, conventional adjustable unit start-up and shutdown cost, energy storage degradation cost, main grid power purchase cost, flexible load adjustment cost, load loss penalty cost, and curtailment penalty cost; the overall operating cost is the sum of all cost items minus the main grid power sales revenue and local retail revenue, in order to minimize the overall operating cost in the optimization model.

[0016] In some embodiments, the optimization model is solved by using a mixed-integer linear programming algorithm.

[0017] In some embodiments, when using a mixed-integer linear programming algorithm to solve the problem, if the global optimum cannot be reached within a preset time, a feasible solution that satisfies all constraints is used as the timing scheduling instruction.

[0018] In some embodiments, a virtual power plant fast-slow two-layer strategy optimization system considering electricity price and carbon price signals and network constraints is also provided, including: The slow-layer carbon market coordination module is configured to: calculate and generate an equivalent carbon price signal for intraday dispatch based on the virtual power plant's free carbon allowances, historical carryover carbon allowances, initial carbon inventory, market carbon price, and the predicted power generation of conventional adjustable units during the carbon market compliance period. The fast-layer optimization scheduling solution module is configured to: construct an optimization model based on the equivalent carbon price signal, the electricity price signal of the electricity market, the operating parameters of each resource within the virtual power plant, and the network constraint parameters of the distribution network; the optimization model aims to minimize the overall operating cost of the virtual power plant, and is constrained by the operating limitations of each resource within the virtual power plant and the power flow transmission limitations of the distribution network; solve the optimization model to obtain the output or consumption plans of each resource within the virtual power plant in multiple scheduling periods, which serve as time-series scheduling instructions; The virtual power plant's internal resources include distributed renewable energy units, conventional adjustable units, energy storage devices, flexible adjustable loads, and main grid trading interfaces; the comprehensive operating cost includes the carbon emission cost calculated from the equivalent carbon price signal and the carbon emissions of conventional adjustable units.

[0019] The present invention has the following beneficial effects: This invention, by employing the technical feature of a "slow-layer carbon market coordination step," decouples the long-cycle, low-frequency carbon market quota management problem from the short-cycle, high-frequency power dispatch problem. It quantifies macro-level carbon compliance pressure into a specific, operable daily cost signal—the "equivalent carbon price signal"—thus effectively introducing the long-term constraint of carbon into the intraday optimization decision-making process, laying the foundation for collaborative optimization. Furthermore, by employing the technical feature of a "fast-layer power market optimization dispatch step," it constructs a multi-variable, multi-constraint mathematical optimization model, providing a unified decision-making framework capable of simultaneously addressing various challenges. This invention addresses various information and constraints from the electricity market, carbon market, and physical systems. Specifically, by receiving and processing "equivalent carbon price signals," it internalizes external environmental costs. This allows the objective function of the "optimization model" to simultaneously reflect economic and environmental costs, thereby enabling the solved "time-series scheduling instructions" to automatically reduce the use of high-carbon-emission resources such as "conventional adjustable generating units" while pursuing economic efficiency, directly promoting low-carbon operation. Furthermore, by receiving and processing "electricity market price signals," the optimization model responds to price fluctuations in the electricity market. This ensures the economic viability of "virtual power plant" operation, enabling it to capture market opportunities and reduce electricity costs by optimizing the purchase and sale of electricity through the "main grid trading interface" and the scheduling of internal resources. By receiving and processing the operating parameters of various resources within the virtual power plant, it accurately models the physical operating limits of different resources, ensuring that the generated "time-series scheduling instructions" for "distributed renewable energy units," "conventional adjustable units," "energy storage devices," and "flexible adjustable loads" remain within their safe operating range, thus guaranteeing equipment safety and power supply reliability. Furthermore, by receiving and processing the network parameters of the distribution network... The technical feature of "constraint parameters" incorporates the physical transmission capacity limitations of the power grid into the decision-making process, ensuring that the optimized dispatch scheme will not lead to line overload or voltage exceedance. This fundamentally solves the technical problem that the dispatch scheme is not feasible in the physical power grid, greatly enhancing the practicality and safety of the scheme. By incorporating "carbon emission costs" into "comprehensive operating costs" and minimizing them as the goal, the technical feature unifies the quantitative evaluation standards of economic efficiency and low carbon emissions. This allows the solution process of the "optimization model" to automatically find the optimal balance between economic costs and environmental costs, thereby maximizing the overall benefits of the "virtual power plant".

[0020] In summary, through the synergistic effect of all the above-mentioned technical features, the present invention successfully unifies the three major elements of electricity, carbon, and grid within an optimization framework, outputting "time-sequenced dispatch instructions" that conform to market economic laws, meet carbon emission reduction requirements, and strictly comply with grid physical security rules. Ultimately, this enables virtual power plants to achieve multi-objective synergistic optimization of economy, low carbon emissions, and grid security in a dual electricity-carbon market environment.

[0021] Other beneficial effects of the present invention will be further described below. Attached Figure Description

[0022] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a block diagram illustrating the synergistic optimization of fast and slow dual-layer electro-carbon coatings in an embodiment. Figure 2 For example, a comparison of annual scenarios: revenue structure, annual carbon emissions, and power generation diagrams of high-carbon / low-carbon units; Figure 3 This is a schematic diagram illustrating monthly carbon quota trading and inventory in an example. Figure 4 This is a schematic diagram showing the daily output comparison of representative equipment in an example.

[0023] Figure 5 This is a schematic diagram of the representative daily line load rate for an example. Figure 6 This is a schematic diagram illustrating price signals and transaction responses for an example. Detailed Implementation

[0024] The embodiments of the present invention will be described in detail below. It should be emphasized that the following description is merely exemplary and not intended to limit the scope and application of the present invention.

[0025] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of embodiments of the present invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0026] This invention recognizes that existing technologies for virtual power plants in a dual electricity-carbon market environment struggle to simultaneously achieve multi-objective coordinated optimization that balances economic efficiency, low carbon emissions, and grid security. This is because existing technologies fail to acknowledge the fundamental time-scale mismatch between the slow-cycle nature of carbon market compliance mechanisms and the fast-cycle demands of electricity market dispatch decisions, as well as the inherent mechanistic obstacle of directly quantifying carbon constraints and integrating them into real-time economic dispatch. Therefore, this invention, by discovering and based on the mechanism that "slow-cycle carbon quota constraints can be transformed into fast-cycle dispatch costs through equivalent carbon price signals," proposes a core scheme for constructing a fast-slow dual-layer coordinated optimization framework to solve the specific technical problem of multi-objective coordination among electricity, carbon, and the grid.

[0027] In some embodiments, the present invention provides a fast-slow two-layer strategy optimization method for virtual power plants that considers electricity price and carbon price signals and network constraints, in order to solve the following problems existing in the prior art: 1. Electricity prices and carbon prices are difficult to coordinate in decision-making regarding virtual power plant dispatching; 2. Insufficient synergy among various aggregated resources such as photovoltaics, energy storage, and market-based power purchase and sale, failing to fully realize the overall optimization advantages of virtual power plants; 3. It is difficult to balance power supply reliability, market benefits, and renewable energy consumption under low-carbon constraints.

[0028] In some embodiments, the overall conceptual framework of the method of the present invention is as follows: The slow layer is the carbon market calculation layer. Based on the virtual power plant's carbon market compliance cycle: free carbon allowances, historical carryover carbon allowances, initial carbon inventory, market carbon price, and the predicted power generation of conventional adjustable units during the cycle, it calculates and generates an equivalent carbon price signal for intraday dispatch, providing a trading planning reference for the subsequent power market dispatch layer. The fast layer is the electricity market dispatch layer, used to process a given equivalent carbon price signal. Under actual or predicted electricity price signals, and in conjunction with network constraints, the virtual power plant's distributed renewable energy units (such as photovoltaic units, wind turbine units, etc.), conventional adjustable units, energy storage devices, flexible adjustable loads, and power exchanged with the main grid through the main grid trading interface (main grid interaction power) are jointly optimized.

[0029] By using the above two-level coordination method, the slow variables of the carbon market and the fast variables of power dispatch are decoupled, and the compliance pressure and low-carbon constraints are mapped into part of the operating cost of conventional adjustable units through the equivalent carbon price, thereby realizing the low-carbon economic operation of the virtual power plant.

[0030] In some embodiments, the method of the present invention is applicable to the aggregation and control of virtual power plants, and includes the following resources: 1. Distributed photovoltaic power; 2. Conventional adjustable units: including but not limited to conventional adjustable units of coal, gas, and oil, or other units with controllable output and carbon emission attributes; 3. Energy storage devices; 4. Flexible and adjustable loads: including loads that can be reduced, loads that can be transferred, and loads that can be interrupted; 5. Mainnet transaction interface; 6. Constraints on power distribution network and related lines.

[0031] The integrated energy system of the park and the distribution network-level resource aggregation platform can both be used as specific application scenarios for this invention.

[0032] In some embodiments, the virtual power plant system involved in the method of the present invention, through its slow-layer carbon market coordination module and fast-layer optimization scheduling solution module, calls or integrates the following functional units to work collaboratively: The distributed renewable energy generation unit is used to collect renewable energy available output predictions, node access locations and output constraints, and provide renewable energy output upper limit data to the fast-layer optimization scheduling solution module; The conventional adjustable unit is used to collect the unit's upper and lower output limits, ramp rate, start-up and shutdown status, fuel cost, start-up and shutdown cost and carbon emission factor, and execute the unit's output and start-up and shutdown strategies based on the output results of the fast layer optimization scheduling solution module. The energy storage unit is used to collect parameters such as energy storage capacity, upper limit of charging and discharging power, charging and discharging efficiency, initial state of charge and degradation cost, and execute charging and discharging plans based on the output results of the fast layer optimization scheduling solution module; The flexible and adjustable load management unit is used to collect user adjustable capacity, adjustment time window, adjustment cost and contract constraints, and provide adjustable load boundary data to the fast layer optimization scheduling solution module; The main grid trading interface unit is used to receive information on electricity purchase price, electricity sales price, trading power limit and grid connection point, and output the main grid purchase power and electricity sales power. The network constraint modeling unit is used to receive node topology, line capacity, node load distribution and equipment access location, and calculate line power flow constraints for use by the fast layer optimization scheduling solution module; The slow-layer carbon market coordination module receives free carbon allowances, historical carryover carbon allowances, initial carbon inventory, market carbon prices, compliance cycle parameters, and predicted generation of conventional adjustable units within the cycle. This module first calculates the total expected carbon emissions for the cycle (i.e., predicted emissions) based on the emission factors of conventional adjustable units and the predicted generation of conventional adjustable units. Then, it calculates the emission gap and equivalent carbon price signal. The emission factor of conventional adjustable units is a known constant or a parameter that can be obtained by querying the unit model and fuel type, representing the carbon emissions generated per unit of electricity generated.

[0033] The fast-layer optimization scheduling solution module is used to establish objective functions and constraints based on data from various functional units, solve for and output time-series scheduling instructions for each resource.

[0034] In some embodiments, the slow-layer carbon market coordination method is further described. The slow-layer carbon market coordination module operates on a cyclical scale. The slow-layer operating cycle can be selected according to the carbon market rules and the internal management needs of the virtual power plant. It is not limited to a single cycle. The annual compliance cycle is used as the final balance cycle of carbon quotas, and the monthly or quarterly cycle is used as the rolling update cycle. The equivalent carbon price is updated according to the latest emission forecasts, carbon inventory and carbon price levels, and then transmitted to the fast-layer scheduling model.

[0035] Its main steps include: Step 1 Obtain free carbon allowances for the period, historical carryover carbon allowances, initial carbon inventory, market carbon price, and forecasted generation from conventional adjustable units within the period.

[0036] Step 2 The total expected carbon emissions for the period are calculated based on the emission factors of conventional adjustable generating units and the predicted power generation.

[0037] Step 3 Based on the projected emissions, free carbon allowances, historical carryover allowances, initial carbon inventory, and available carbon market prices during the compliance period, calculate the available allowances and expected emission gap for the period, and compare the total expected carbon emissions of the period with the total available carbon allowances of the period to determine the emission gap.

[0038] Step 4: Construct the equivalent carbon price based on the emission gap. If the expected emissions do not exceed the available allowances, the equivalent carbon price can be the reference carbon price, the internal carbon price, or the benchmark carbon price after considering trading frictions. If the expected emissions exceed the available allowances, the emission gap ratio is further calculated, and a carbon price increase term is generated through a preset risk adjustment function. The risk adjustment function can be a linear function, a piecewise linear function, a quadratic function, or an exponential function; a strict one-to-one correspondence between the gap ratio and the carbon price increase ratio is not required.

[0039] Step 5 The generated equivalent carbon price is passed to the fast-layer optimization scheduling model as the carbon emission cost coefficient for conventional adjustable units.

[0040] By adopting the above approach, we can avoid explicitly modeling the carbon market as a frequent trading process on a 15-minute or hourly basis, which is more in line with the operational characteristics of my country's carbon market, which is mainly based on quota management, verification, and compliance settlement.

[0041] The following are some variable calculations for the slow-layer carbon market: Periodic predicted emissions:

[0042] Total available carbon allowances for the cycle:

[0043] Emissions gap:

[0044] when At that time, the equivalent daily carbon price is:

[0045] when First, calculate the gap ratio:

[0046] Recalculate the equivalent daily carbon price:

[0047] Carbon price series for different time periods:

[0048] The equivalent carbon price is not the actual transaction price in the carbon market, but rather a shadow carbon cost signal used for intraday dispatch within virtual power plants. It is composed of the market carbon price, transaction friction costs, compliance risk premium, capital occupation costs, and emission gap pressure, and is subject to upper and lower limits to avoid the carbon price signal being too weak or too strong.

[0049] The risk adjustment function of the equivalent carbon price does not require a strict one-to-one correspondence between the emission deficit ratio and the carbon price increase ratio. Instead, it reflects the increasing trend of carbon quota procurement costs, compliance risks, capital occupation costs, and transaction friction costs faced by virtual power plants as the emission deficit widens. To avoid insufficient incentives for low-carbon dispatch due to an excessively low equivalent carbon price, or distortion of fast-layer optimization results due to an excessively high equivalent carbon price, lower and upper limits can be set for the equivalent carbon price.

[0050] In some embodiments, the fast-level power market optimization dispatch method is further described. The fast-layer optimization scheduling module solves the problem according to intraday time periods. It is preferred to use 15-minute discrete time periods for modeling, but 30-minute or hourly models can also be used.

[0051] Fast layer optimization variables should include at least: 1. Actual output of each distributed renewable energy unit; 2. Curtailment power of each distributed renewable energy unit; 3. Standard adjustable unit output; 4. The start-up and shutdown status of the unit is normally adjustable; 5. Energy storage charging power; 6. Energy storage discharge power; 7. Energy storage state of charge; 8. Power purchased from the main grid; 9. Main grid power sales capacity; 10. Flexible load adjustment amount; 11. Load loss.

[0052] The following are some variables in the fast-track electricity market: Actual photovoltaic output:

[0053] Curtailment power:

[0054] Energy storage charging power:

[0055] Energy storage discharge power:

[0056] Energy storage state of charge:

[0057] Energy storage charge and discharge state variables:

[0058] Output of conventional adjustable units:

[0059] Standard adjustable unit operating status:

[0060] Conventional adjustable unit start-up variables:

[0061] Main network power purchase capacity:

[0062] Main grid power sales:

[0063] Mutual exclusion state variables for electricity purchase and sale:

[0064] Flexible load reduction:

[0065] Load loss:

[0066] In some embodiments, the composition of the electricity price signal is further described. Electricity price signals include, but are not limited to, one or more of the following: 1. Medium- and long-term contract electricity prices; 2. Spot electricity price as of today or within the day; 3. Retail electricity price; 4. Settle electricity price upon connection to the grid; 5. Electricity purchase price on behalf of the customer; 6. Settlement price for green electricity transactions; 7. Other market-based electricity purchase and sale price signals.

[0067] Among these, medium- and long-term electricity contracts are typically signed between electricity consumers and power generators within the power system, with the duration of the contract determined by its content. The concept of "medium- and long-term" generally refers to contracts that are longer than the current day's contract, ranging from monthly to annual terms.

[0068] In one implementation, the fast-layer model simultaneously considers the purchase price of electricity, the sales price of electricity, and the local load retail electricity price to characterize the revenue sources and transaction costs of the virtual power plant as an aggregated operating entity.

[0069] In some embodiments, the fast layer constraint conditions are further described. The fast-layer optimization scheduling model should include at least the following constraints: (1) Renewable energy output constraint: The actual output of each renewable energy unit shall not exceed the available output for the corresponding time period, and the sum of the abandoned power and the actual output shall be equal to the available output.

[0070] Upper and lower limits of photovoltaic power output:

[0071] Abandoned Light and Upper and Lower Realms:

[0072] Decomposition of available photovoltaic power:

[0073] (2) Output constraints of conventional adjustable units The output of conventional adjustable units is limited by minimum and maximum output.

[0074] Output lower bound constraint:

[0075] Output upper bound constraint:

[0076] (3) Start-up and ramping constraints of conventional adjustable units The start-up and shutdown process of conventional adjustable units meets the start-up and shutdown logic constraints, and the output changes in adjacent time periods meet the ramping constraints.

[0077] Initial startup logic constraints:

[0078] Startup logic constraints for subsequent periods:

[0079] Nonnegativity constraint:

[0080] Uphill constraints:

[0081] Downhill constraints:

[0082] (4) Energy storage power and state of charge constraints: The energy storage charging and discharging power is limited by the rated power, and the energy storage state is updated according to the time period evolution equation and meets the upper and lower boundary constraints. The energy storage state at the beginning and end of the day is set to be equal or approximately equal to ensure the closed-loop operation within the day.

[0083] Initial charge state constraints:

[0084] SOC upper and lower bound constraints:

[0085] Charging power constraints:

[0086] Discharge power constraint:

[0087] SOC state transition equation:

[0088] Beginning-of-Day and End-of-Day SOC Consistency Constraints:

[0089] (5) Flexible load constraints: The flexible load adjustment amount shall not exceed the set upper limit and shall correspond to certain compensation costs, production disturbance costs, or comfort loss costs. The cost of flexible adjustable load can be determined according to user contracts or demand response compensation rules. In one embodiment, the flexible load is manifested as a load that can be reduced, and a linear cost function can be used, that is, the flexible load cost is equal to the product of the unit adjustment compensation price and the adjusted electricity amount; the flexible load cost scheme can also be set as a piecewise linear cost function to reflect the increase in user comfort loss or production disturbance cost when the adjustment depth increases; or a quadratic cost function can be used to smoothly characterize the increase in marginal compensation cost brought about by the increase in adjustment amount.

[0090] Flexible load reduction limit:

[0091] Non-negativity constraint for load loss:

[0092] (6) Power balance constraints: For each scheduling period, the following power balance relationship is satisfied: Local renewable energy output + conventional adjustable unit output + energy storage discharge power - energy storage charging power + main grid power purchase power - main grid power sales power + load loss = original load demand - flexible load adjustment.

[0093]

[0094] (7) Main grid transaction constraints: During the same period, the main grid purchase power and the main grid sales power are not simultaneously positive to avoid meaningless simultaneous buying and selling of electricity during the same period.

[0095] Power purchase constraints:

[0096] Electricity sales capacity constraints:

[0097] (8) Network power flow constraints are established based on node injected power and distribution network topology to ensure that the power of each line does not exceed its capacity limit. DC power flow, power transfer distribution factor (PTDF, used to represent the impact of node injected power changes on the active power flow of each line under given network topology and reference node conditions), linear DistFlow (linear DistFlow is an approximate power flow model of the distribution network obtained by linearizing the distribution network power flow equations; it is usually applicable to radial or weak-loop distribution systems and can describe the approximate relationship between branch power, voltage drop, and node injection. It is used for simplified network modeling with small voltage amplitude changes and active power flow constraints; linear DistFlow is more suitable for modeling distribution network scenarios considering node voltage and feeder power constraints.) or other linearized power flow models are used for expression.

[0098] Node load sharing factor:

[0099] The load allocation factor for each node can be determined based on the node's historical electricity consumption, user-registered capacity, transformer capacity, feeder capacity, or node-level load forecast data. When the load of each node can be directly measured, the load allocation factor is calculated from the actual measured or forecast data; when only the aggregated total load is available, the aggregated load can be allocated to each node according to the historical electricity consumption ratio or capacity ratio, ensuring that the sum of the allocation factors for each node is 1.

[0100] Node-injected power vector:

[0101] Line power flow calculation:

[0102] Power flow constraints on the line:

[0103] Line transmission limit relaxation:

[0104] In some embodiments, the construction of the objective function is further explained. The fast-layer optimization objective of this invention is to minimize the combination of the overall operating cost of the virtual power plant and maximize the operating revenue, which is equivalent to minimizing "total cost minus total revenue".

[0105] The total cost includes at least: 1. Fuel cost of conventional adjustable generating units; 2. Start-up and shutdown costs of conventional adjustable generating units; 3. Energy storage degradation cost (expressed using a charge / discharge throughput model, i.e., the energy storage degradation cost is proportional to the sum of the charging and discharging amounts in each time period); 4. Cost of purchasing electricity from the main grid; 5. Costs of flexible load adjustment; 6. Load loss penalty cost; 7. Costs of power curtailment penalties; 8. Carbon emission costs.

[0106] Among them, the carbon emission cost is obtained by multiplying the equivalent carbon price of slow-layer transmission by the emissions of conventional adjustable units, which is used to suppress disorderly output of high-carbon units.

[0107] The fast-layer optimization scheduling model can be constructed as a linear programming, mixed-integer linear programming, or mixed-integer quadratic programming model based on the cost function form. In one embodiment, to improve solution efficiency, fuel cost, flexible load cost, and energy storage degradation cost are expressed in linear or piecewise linear form, and mixed-integer linear programming is used for solution. For larger-scale systems, a solution time cap, an optimal gap threshold, and a rolling optimization window can be set, and a feasible solution that satisfies the constraints can be used as the scheduling execution scheme if the global optimum is not reached.

[0108] Total revenue includes at least: 1. Revenue from local electricity sales; 2. Revenue from selling surplus electricity to the grid; 3. Other market-based benefits.

[0109] Local load electricity sales revenue is calculated based on actual service load and retail electricity price. Virtual power plant operators, as electricity sales companies, park energy service providers, or load aggregators, sell electricity to users within their coverage area and collect electricity fees. Local users are located outside the economic settlement boundary of the virtual power plant operator, and their electricity fees belong to the external operating income obtained by the virtual power plant operator from contract users.

[0110] Revenue from selling surplus electricity to the grid is calculated based on the electricity sold and the grid settlement price.

[0111] Overall objective function:

[0112] Equivalent to:

[0113] Organized by time period:

[0114] Fuel costs for thermal power plants:

[0115] Startup costs:

[0116] Energy storage degradation costs:

[0117] Cost of abandonment penalty:

[0118] Loss of load penalty cost:

[0119] Electricity purchase cost:

[0120] Flexible load costs are calculated using a linear cost function.

[0121] Carbon emission costs:

[0122] Revenue from electricity sales:

[0123] Actual service load:

[0124] Local retail revenue:

[0125] In some embodiments, the solution process is further described. The method of the present invention includes at least the following steps: Step A: Input the virtual power plant resource parameters, load curve, renewable energy forecast curve, electricity price curve, network parameters, carbon quota parameters, and reference carbon price.

[0126] Step B calculates the cyclical emission gap in the slow layer and generates the intraday equivalent carbon price accordingly.

[0127] Step C: Input the equivalent carbon price into the fast layer optimization scheduling model.

[0128] Step D: Establish the objective function and various operational constraints in the fast layer, and perform joint optimization of resources for each time period.

[0129] Step E outputs the timing scheduling instructions and market operation results for each resource of the virtual power plant.

[0130] Steps A to E above constitute the basic framework of the fast and slow two-layer optimization solution process of this invention, as follows: Figure 1 As shown in the figure. Steps B and C demonstrate the coupling mechanism between the slow-layer carbon market coordination module and the fast-layer optimization scheduling module; after receiving the equivalent carbon price signal, step D constructs the objective function and applies various operational constraints, including network constraints, and finally outputs the optimization result that meets the requirements of electricity-carbon-grid coordination.

[0131] In some embodiments, the output results are further explained. The output should include at least: 1. Output and curtailment of renewable energy; 2. The unit's output and start / stop status are normally adjustable; 3. Energy storage charging and discharging power and state of charge; 4. Flexible load adjustment amount; 5. Main grid power purchase and power sales; 6. Load service rate; 7. Renewable energy consumption rate; 8. Electricity volume purchased and sold on a time-of-use basis; 9. Power flow and key congested lines for each route; 10. Total carbon emissions and carbon costs; 11. Net operating income of virtual power plants.

[0132] This invention is not limited to a single equivalent carbon price calculation method. The risk adjustment function in the equivalent carbon price can be a linear function, a piecewise linear function, a quadratic function, an exponential function, or a lookup function based on historical data; the fast layer optimization algorithm can be centralized optimization, rolling time-domain optimization, decomposition coordination optimization, or heuristic repair optimization; the network constraints can be DC power flow, PTDF, linear DistFlow, or other linearized power flow models.

[0133] The present invention will be further described below with reference to an embodiment, but the scope of protection of the present invention is not limited thereto.

[0134] System Structure The system employs a 30-node network as the distribution network backbone upon which the virtual power plant relies. The system includes: 1. The distributed photovoltaic unit has a total of 6 photovoltaic access points, which are connected to different nodes to simulate new energy resources with multiple decentralized access points.

[0135] 2. Two conventional adjustable units are set up, with their access nodes being node 6 and node 24, respectively, to provide adjustable support when the load is high or when renewable energy is insufficient.

[0136] 3. The energy storage device is a centralized energy storage device used to perform peak shaving and valley filling, time-series arbitrage and short-term power support.

[0137] 4. Flexible loads can be set up with one or more types of flexible loads, which can be reduced or adjusted under given cost constraints.

[0138] 5. The main network transaction interface main network access point is set at node 1, which is used to purchase electricity when the system is short of power and sell electricity when the system has surplus power.

[0139] 6. The carbon market coordination module generates an intraday equivalent carbon price based on the periodic free allowances, carbon inventory, reference carbon price, and expected emissions.

[0140] Scheduling input Input typical daily load curves, electricity purchase price curves, electricity sales price curves, local retail electricity price curves, available output curves for distributed photovoltaic power, network parameters, and carbon market parameters.

[0141] Among them, the fast-layer scheduling adopts 15-minute discrete modeling to more accurately depict the operation strategy of the virtual power plant at different times of the day.

[0142] Solution process First, the emission gap is calculated in the slow layer based on the quota and emission forecast results within the compliance period. Second, an equivalent carbon price is generated based on the emission gap. Then, the equivalent carbon price is passed to the fast layer optimization model. Finally, the optimal operating strategy for a typical day is obtained by comprehensively considering electricity price, carbon price, power balance constraints, energy storage constraints, conventional adjustable unit constraints, main grid trading constraints, and network flow constraints in the fast layer model.

[0143] Solution results During a typical day, the system meets the aggregated load demand of the virtual power plant through distributed photovoltaics, conventional adjustable generating units, energy storage, and main grid power purchase. When local resources are abundant and line constraints are met, electricity can be sold to external users through the main grid interface. When transmission pressure occurs on some lines, the optimization model maintains the system's power flow feasibility by adjusting the unit output, energy storage charging and discharging, main grid power purchase and sales, and flexible load regulation.

[0144] The results show that the method of the present invention can balance economy, low carbon emissions and power supply reliability while taking into account network constraints, and improve the capacity for renewable energy absorption and the operating income of aggregate entities.

[0145] Technical Mechanism Explanation In this invention, the equivalent carbon price output by the slow layer is introduced into the objective function of the fast layer through the carbon emission cost term. This enables conventional adjustable generating units to automatically reduce their high-emission operation tendency when carbon costs increase, thereby improving the priority of renewable energy consumption. Simultaneously, the fast-layer model coordinates grid power purchase and sales, energy storage charging and discharging, flexible load regulation, and the operation of conventional adjustable generating units under node network constraints, ensuring that the scheduling results meet both economic optimization requirements and physical feasibility requirements.

[0146] Example execution results To verify the feasibility of the fast-slow two-layer strategy optimization method of this invention, a campus-level virtual power plant example system was constructed for computational testing. This example system includes distributed photovoltaic power, conventional adjustable generating units, energy storage devices, flexible adjustable loads, a main grid trading interface, and simplified campus feeder constraints. The computational test uses a 15-minute scheduling period, with a total of 96 scheduling periods per day. The slow layer coordinates carbon assets based on monthly carbon quota status and annual compliance requirements, while the fast layer performs day-ahead optimization scheduling based on electricity price signals, equivalent carbon price signals, and equipment operating constraints.

[0147] This example sets up two comparison scenarios: Scenario 1 is a traditional electricity trading scenario. In this scenario, virtual power plants mainly optimize resource scheduling based on electricity price signals, without feeding back carbon price signals and quota inventory pressure to the intraday scheduling process, nor do they proactively optimize carbon quota assets.

[0148] Scenario 2 is an electricity-carbon coupled quota trading scenario. In this scenario, the slow-layer carbon market coordination module generates an equivalent carbon price signal based on the monthly carbon price, free quotas, predicted emissions, and carbon inventory status, and transmits it to the fast-layer scheduling model. The fast layer, while satisfying power balance, equipment operation, and line capacity constraints, jointly optimizes photovoltaic output, conventional adjustable unit output, energy storage charging and discharging, flexible load regulation, and grid power purchase and sale.

[0149] The annual comprehensive benefit comparison results of the calculation example are as follows: Figure 2 As shown in the figure, compared with the traditional electricity trading scenario, the annual total revenue of the electricity-carbon coupled quota trading scenario is slightly higher, while the annual total carbon emissions are significantly reduced. In terms of power generation structure, the power generation of high-carbon units is significantly reduced in this scenario, while the proportion of low-carbon units and renewable energy consumption is correspondingly increased, verifying the effectiveness of this method in the synergistic optimization of the low-carbon economy. Figure 3 The study further demonstrates the monthly carbon quota trading and inventory changes in the electricity-carbon coupling scenario. The results show that this method, through flexible quota trading and inventory management, not only reduces compliance risks but also brings additional carbon asset trading revenue.

[0150] Figure 4 The output curves of various resources are shown on two sets of comparative scenarios for typical days (regular trading days in July). It can be seen that... Figure 4 In the electricity-carbon coupling quota trading scenario below, conventional adjustable units (especially high-carbon units) actively reduce their output during peak photovoltaic power generation periods during the daytime, energy storage devices participate more actively in peak shaving and valley filling, and grid power purchase and sale behavior becomes more intensive. Figure 5 The diagram shows the load factor heatmap of each feeder line in the virtual power plant operating on the distribution network during the aforementioned representative days. As can be seen from the diagram, in both scenario one and scenario two, the load factor of each line is controlled within the safe constraint range (not exceeding 100%), indicating that the network power flow constraints constructed by this method effectively ensure the safety of system operation. Figure 6 The study demonstrates the temporal coupling relationship between the electricity price signal, the carbon price signal, and the virtual power plant trading response. It shows that both the net output of energy storage and the net power purchased by the main grid can accurately follow the changes in the purchase and sale of electricity prices to optimize arbitrage, achieving the economically optimal response under the constraint of the carbon price (74.0 yuan / tCO2).

[0151] In summary, compared with traditional electricity trading scenarios, the electricity-carbon coupled quota trading scenario, while satisfying power balance, equipment operation constraints, and line capacity constraints, can guide virtual power plants to reduce the output of high-carbon units and improve the utilization of low-carbon resources, and obtain carbon asset revenue through monthly quota trading and inventory management. The annual and representative day results together demonstrate that this method can simultaneously improve the operating revenue of the aggregation entity, reduce annual carbon emissions, and keep the load rate of each feeder within the capacity constraint range, verifying the feasibility and effectiveness of the fast-slow two-layer strategy optimization method.

[0152] The background section of this invention may include background information about the problems or environment in which the invention is being developed, and is not necessarily a description of prior art. Therefore, the content included in the background section does not constitute an admission of prior art by the applicant.

[0153] The above description provides a further detailed explanation of the present invention in conjunction with specific / preferred embodiments, and it should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various substitutions or modifications can be made to these described embodiments without departing from the concept of the present invention, and all such substitutions or modifications should be considered within the scope of protection of the present invention. In the description of this specification, the reference to terms such as "an embodiment," "some embodiments," "preferred embodiment," "example," "specific example," or "some examples," etc., indicates that the specific features, structures, materials, or characteristics described in connection with that embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any suitable manner in one or more embodiments or examples. Without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification and the features of different embodiments or examples. Although the embodiments of the present invention and their advantages have been described in detail, it should be understood that various changes, substitutions, and modifications can be made herein without departing from the scope of protection of the patent application.

Claims

1. A virtual power plant fast-slow two-level strategy optimization method considering electricity price and carbon price signals and network constraints, characterized in that, Includes the following steps: Slow-stage carbon market coordination steps: Based on the virtual power plant's carbon market compliance cycle: free carbon allowances, historical carryover carbon allowances, initial carbon inventory, market carbon price, and the predicted power generation of conventional adjustable units during the cycle, an equivalent carbon price signal for intraday dispatch is calculated and generated. Fast-level power market optimization scheduling steps: Based on the equivalent carbon price signal, the electricity price signal of the power market, the operating parameters of each resource within the virtual power plant, and the network constraint parameters of the distribution network, an optimization model is constructed; the optimization model aims to minimize the overall operating cost of the virtual power plant, and is constrained by the operating limitations of each resource within the virtual power plant and the power flow transmission limitations of the distribution network. Solving the optimization model yields the output or consumption plans of each resource in the virtual power plant during multiple scheduling periods, which serve as time-series scheduling instructions. The virtual power plant's internal resources include distributed renewable energy units, conventional adjustable units, energy storage devices, flexible adjustable loads, and a main grid trading interface; the comprehensive operating cost includes the carbon emission cost calculated from the equivalent carbon price signal and the carbon emissions of the conventional adjustable units.

2. The virtual power plant fast and slow two-layer strategy optimization method considering electricity price and carbon price signals and network constraints as described in claim 1, characterized in that, The slow-layer carbon market coordination steps specifically include: obtaining the free carbon allowances, historical carryover carbon allowances, initial carbon inventory, market carbon price, and predicted conventional adjustable unit power generation; and calculating the expected total carbon emissions for the period based on the emission factors of the conventional adjustable units and the predicted conventional adjustable unit power generation. The total available carbon allowances for the calculation period are the sum of the free carbon allowances, historical carryover carbon allowances, and the initial carbon inventory. Calculate the emission gap between the expected total carbon emissions for the period and the total available carbon allowances for the period; generate the equivalent carbon price signal using a risk adjustment function based on the emission gap and the market carbon price.

3. The virtual power plant fast and slow two-layer strategy optimization method considering electricity price and carbon price signals and network constraints as described in claim 2, characterized in that, Based on the emission gap and the market carbon price, the equivalent carbon price signal is generated through a risk adjustment function. Specifically, if the emission gap is less than or equal to zero, the equivalent carbon price signal is the market carbon price or a benchmark carbon price determined based on the market carbon price; if the emission gap is greater than zero, the gap ratio is calculated, and based on the market carbon price and the gap ratio, the equivalent carbon price signal is calculated through the risk adjustment function, wherein the equivalent carbon price signal increases as the gap ratio increases.

4. The virtual power plant fast and slow two-layer strategy optimization method considering electricity price and carbon price signals and network constraints as described in claim 3, is characterized in that... The risk adjustment function includes a linear function, a piecewise linear function, a quadratic function, or an exponential function.

5. The virtual power plant fast and slow two-layer strategy optimization method considering electricity price and carbon price signals and network constraints as described in claim 1, characterized in that, The constraints of the optimization model include at least one of the following: renewable energy output constraints; conventional adjustable unit output constraints, start-up and shutdown constraints, and ramping constraints; energy storage power constraints and state of charge constraints; flexible load adjustment constraints; power balance constraints; main grid trading constraints; and network power flow constraints based on power transmission distribution factor or linear distribution power flow model.

6. The virtual power plant fast-slow two-layer strategy optimization method considering electricity price and carbon price signals and network constraints as described in claim 1, characterized in that, The electricity price signal includes at least one of the following: medium- and long-term contract electricity price, day-ahead spot electricity price, intraday spot electricity price, retail electricity price, grid connection settlement electricity price, agent purchase electricity price, or green electricity transaction settlement price.

7. The virtual power plant fast and slow two-layer strategy optimization method considering electricity price and carbon price signals and network constraints as described in claim 1, characterized in that, The comprehensive operating cost also includes at least one of the following costs: fuel cost of conventional adjustable generating units, start-up and shutdown cost of conventional adjustable generating units, energy storage degradation cost, main grid power purchase cost, flexible load adjustment cost, load loss penalty cost, and curtailment penalty cost; the comprehensive operating cost is the sum of each cost item minus the main grid power sales revenue and local retail revenue, in order to minimize the comprehensive operating cost in the optimization model.

8. The virtual power plant fast and slow two-layer strategy optimization method considering electricity price and carbon price signals and network constraints as described in claim 1, characterized in that, The optimization model is solved using a mixed-integer linear programming algorithm.

9. The virtual power plant fast and slow two-layer strategy optimization method considering electricity price and carbon price signals and network constraints as described in claim 8, characterized in that, When using the mixed-integer linear programming algorithm to solve the problem, if the global optimum cannot be reached within a preset time, then a feasible solution that satisfies all constraints is used as the timing scheduling instruction.

10. A virtual power plant fast-slow two-level strategy optimization system considering electricity price and carbon price signals and network constraints, characterized in that, include: The slow-layer carbon market coordination module is configured to: calculate and generate an equivalent carbon price signal for intraday dispatch based on the virtual power plant's free carbon allowances, historical carryover carbon allowances, initial carbon inventory, market carbon price, and the predicted power generation of conventional adjustable units during the carbon market compliance period. The fast-layer optimization scheduling solution module is configured to: construct an optimization model based on the equivalent carbon price signal, the electricity price signal of the electricity market, the operating parameters of each resource inside the virtual power plant, and the network constraint parameters of the distribution network; the optimization model aims to minimize the overall operating cost of the virtual power plant, and uses the operating constraints of each resource inside the virtual power plant and the power flow transmission constraints of the distribution network as constraints. Solving the optimization model yields the output or consumption plans of each resource in the virtual power plant during multiple scheduling periods, which serve as time-series scheduling instructions. The virtual power plant's internal resources include distributed renewable energy units, conventional adjustable units, energy storage devices, flexible adjustable loads, and a main grid trading interface; the comprehensive operating cost includes the carbon emission cost calculated from the equivalent carbon price signal and the carbon emissions of the conventional adjustable units.

Images