A joint clearing calculation method, device, equipment and storage medium

Abstract

The application relates to a combined clearing calculation method, device and equipment and a storage medium. The method comprises the following steps: acquiring day-ahead unit parameters; inputting the day-ahead unit parameters into a pre-established day-ahead unit combination model, combining a first dynamic carbon emission factor sequence calculated through a factor calculation model to perform optimization calculation, and obtaining a day-ahead combined clearing result; taking a unit start-stop state in the day-ahead combined clearing result as a constraint condition, combining a second dynamic carbon emission factor sequence calculated through the factor calculation model, inputting acquired intraday unit parameters into a pre-established intraday constraint scheduling model to perform calculation, and obtaining an intraday combined clearing result. The application can improve the accuracy of the clearing calculation result.

Description

Technical Field

[0001] This application relates to the field of low-carbon operation and carbon emission accounting technology of power systems, and in particular to a joint clearing calculation method, apparatus, equipment and storage medium. Background Technology

[0002] With the advancement of new power system construction, power dispatching and trading mechanisms are becoming increasingly complex. In this process, the spot market for electricity and ancillary service markets (such as frequency regulation and peak shaving) jointly participate in unit output decisions. This leads to a more dynamic and coupled nature in unit operating status and marginal emission characteristics.

[0003] In related technologies, joint clearing calculation models for electricity and ancillary services typically employ static or average unit carbon emission parameters, or ignore the impact of carbon emission costs on clearing decisions. Joint clearing calculation refers to the decision-making process of determining the traded electricity volume, reserved capacity, and market prices for ancillary services such as frequency regulation. This results in joint clearing models failing to accurately reflect the dynamic changes in carbon emission intensity of units under different treatment levels and load rates, leading to low accuracy in clearing calculation results. Summary of the Invention

[0004] Therefore, it is necessary to provide a joint clearing calculation method, apparatus, device, and storage medium to address the aforementioned technical problems.

[0005] In a first aspect, this application provides a joint clearing calculation method, including:

[0006] Obtain day-ahead unit parameters;

[0007] The day-ahead unit parameters are input into the pre-established day-ahead unit combination model, and optimized by combining the first dynamic carbon emission factor sequence calculated by the factor calculation model to obtain the day-ahead joint clearing result.

[0008] Using the unit start-up and shutdown status in the day-ahead joint clearing results as constraints, and combining the second dynamic carbon emission factor sequence calculated by the factor calculation model, the obtained intraday unit parameters are substituted into the pre-established intraday constraint scheduling model for calculation to obtain the intraday joint clearing results.

[0009] In one embodiment, the step of calculating the first dynamic carbon emission factor sequence includes:

[0010] Acquire the static and dynamic characteristics of the generator set; the static characteristics include the generator set type parameters, air-cooling type parameters, and unit capacity parameters, and the dynamic characteristics include the relative load data sequence corresponding to the historical power generation duration of the generator set; the relative load data is composed of the day-ahead power generation and installed capacity of the unit to be optimized.

[0011] Feature embedding calculations are performed on static and dynamic features along the time dimension to obtain working condition features;

[0012] The operating condition characteristics are input into the factor calculation model, and the operating condition characteristics are jointly learned and mapped through the factor calculation model to obtain the first dynamic carbon emission factor sequence.

[0013] In one embodiment, operating condition characteristics are input into a factor calculation model, and the operating condition characteristics are jointly learned and mapped through the factor calculation model to obtain a first dynamic carbon emission factor sequence, including:

[0014] The working condition features are input into the multi-head sub-attention layer in the factor calculation model to obtain the query vector, key vector and value vector output by the multi-head sub-attention layer;

[0015] Using the scaled dot product attention algorithm, combined with the query vector, key vector and value vector corresponding to each attention head, the correlation weight between each time step is calculated to obtain the attention weight matrix. The attention weight matrix is ​​used to characterize the degree of influence of each historical moment on the current moment.

[0016] The attention weight matrices are concatenated, and the dynamic carbon emission factor is obtained based on the concatenated attention weight matrices.

[0017] In one embodiment, day-ahead unit parameters are input into a pre-established day-ahead unit combination model, and optimized using a first dynamic carbon emission factor sequence calculated through a factor calculation model to obtain day-ahead joint clearing results, including:

[0018] Obtain the constraints of the day-ahead unit combination model;

[0019] Substituting the day-ahead unit parameters and the first dynamic carbon emission factor sequence into the day-ahead unit combination model, and combining the constraints of the day-ahead unit combination model, the day-ahead joint clearing result is obtained with the objective of minimizing the total day-ahead operating cost. The day-ahead joint clearing result includes the start-up and shutdown status of each generator unit and the operating power of the generator unit.

[0020] In one embodiment, the solution yields the day-ahead joint clearing results, including:

[0021] Obtain the day-ahead total operating cost calculation function; the day-ahead total operating cost calculation function includes the electricity market cost item, carbon emission cost item, curtailment loss cost item, power shortage loss cost item, and unit start-up and shutdown cost item;

[0022] Substitute the day-ahead unit parameters and the first dynamic carbon emission factor sequence into the total cost calculation function, combine the constraints of the day-ahead unit combination model, and call the linear programming solver to calculate the day-ahead joint clearing result.

[0023] In one embodiment, the unit start-up and shutdown status in the day-ahead joint clearing results is used as a constraint condition. Combined with the second dynamic carbon emission factor sequence calculated through a factor calculation model, the intraday unit parameters are substituted into a pre-established intraday constraint scheduling model for calculation, resulting in the intraday joint clearing results, including:

[0024] Obtain the constraints of the intraday constraint scheduling model;

[0025] Substituting the intraday unit parameters and the second dynamic carbon emission factor sequence into the intraday constrained scheduling model, and combining the constraints of the intraday constrained scheduling model, with the goal of minimizing the total intraday operating cost, the intraday joint clearing result is obtained. The intraday joint clearing result includes the corrected operating power and frequency regulation call amount of each generator unit.

[0026] In one embodiment, the intraday joint clearing result is obtained by solving for:

[0027] Obtain the function for calculating total intraday operating costs. The function includes intraday cost calculation items, frequency regulation revenue items, and frequency regulation cost items; the frequency regulation revenue items include spot market revenue and frequency regulation revenue.

[0028] Substitute the initial parameters and dynamic carbon emission factor into the daily total operating cost calculation function, and substitute the unit start-up and shutdown status and generator operating power from the day-ahead joint clearing results into the constraints of the daily constrained scheduling model. Call the linear programming solver to calculate the daily joint clearing results.

[0029] Secondly, this application also provides a joint clearing calculation device, comprising:

[0030] The acquisition module is used to acquire day-ahead unit parameters;

[0031] The day-ahead calculation module is used to input the day-ahead unit parameters into the pre-established day-ahead unit combination model, and perform optimization calculations by combining the first dynamic carbon emission factor sequence calculated through the factor calculation model to obtain the day-ahead joint clearing results.

[0032] The intraday calculation module uses the unit start-up and shutdown status in the day-ahead joint clearing results as constraints, and combines it with the second dynamic carbon emission factor sequence calculated by the factor calculation model. The obtained intraday unit parameters are then substituted into the pre-established intraday constraint scheduling model for calculation to obtain the intraday joint clearing results.

[0033] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0034] Obtain day-ahead unit parameters;

[0035] The day-ahead unit parameters are input into the pre-established day-ahead unit combination model, and optimized by combining the first dynamic carbon emission factor sequence calculated by the factor calculation model to obtain the day-ahead joint clearing result.

[0036] Using the unit start-up and shutdown status in the day-ahead joint clearing results as constraints, and combining the second dynamic carbon emission factor sequence calculated by the factor calculation model, the obtained intraday unit parameters are substituted into the pre-established intraday constraint scheduling model for calculation to obtain the intraday joint clearing results.

[0037] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0038] Obtain day-ahead unit parameters;

[0039] The day-ahead unit parameters are input into the pre-established day-ahead unit combination model, and optimized by combining the first dynamic carbon emission factor sequence calculated by the factor calculation model to obtain the day-ahead joint clearing result.

[0040] Using the unit start-up and shutdown status in the day-ahead joint clearing results as constraints, and combining the second dynamic carbon emission factor sequence calculated by the factor calculation model, the obtained intraday unit parameters are substituted into the pre-established intraday constraint scheduling model for calculation to obtain the intraday joint clearing results.

[0041] The aforementioned joint clearing calculation method, apparatus, equipment, and storage medium acquire day-ahead unit parameters; input these parameters into a pre-established day-ahead unit combination model; and perform optimization calculations using a first dynamic carbon emission factor sequence calculated through a factor calculation model to obtain the day-ahead joint clearing result. Using the unit start-up and shutdown status in the day-ahead joint clearing result as constraints, and combining it with a second dynamic carbon emission factor sequence calculated through the factor calculation model, the acquired intraday unit parameters are substituted into a pre-established intraday constrained scheduling model for calculation to obtain the intraday joint clearing result. This application dynamically outputs the carbon emission factor of each unit at each moment through a factor calculation model, truly reflecting the physical characteristics of thermal / coal-fired power units, and further improving the accuracy of subsequent clearing calculations. Furthermore, this application embeds the dynamic carbon emission factor sequence into the day-ahead unit combination model and the intraday constrained scheduling model, enabling the joint optimization of carbon emission costs and power generation costs in the same optimization model, thereby achieving a joint optimal balance between economic efficiency and low carbon emissions. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a diagram illustrating the application environment of the joint clearing calculation method in one embodiment;

[0044] Figure 2 This is a flowchart illustrating the joint clearing calculation method in one embodiment;

[0045] Figure 3 This is a flowchart illustrating the calculation of the first dynamic carbon emission factor sequence in one embodiment;

[0046] Figure 4 This is a flowchart illustrating a method for calculating the first dynamic carbon emission factor sequence in another embodiment;

[0047] Figure 5 This is a flowchart illustrating the calculation of the day-ahead joint clearing results in one embodiment;

[0048] Figure 6 This is a flowchart illustrating the calculation of intraday joint clearing results in one embodiment;

[0049] Figure 7 This is a structural block diagram of a joint clearing calculation device in one embodiment. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0051] The technical background of this application will be explained below.

[0052] With the continuous advancement of the construction of new power systems, the power system operation mode is rapidly transforming from traditional planned dispatch to market-based dispatch. Against this backdrop, the spot market for electricity and ancillary service markets such as frequency regulation and peak shaving jointly participate in unit output decisions, resulting in stronger dynamism and multi-market coupling in unit operating conditions and marginal emission characteristics. However, existing methods for calculating the carbon factor are mostly based on static average emission levels or solely on the results of electricity market clearing, failing to effectively reflect the impact of ancillary services such as frequency regulation on actual unit output allocation and carbon emission characteristics, leading to insufficient accuracy and timeliness in carbon emission calculation results. Therefore, there is an urgent need to construct a carbon factor calculation method that comprehensively considers the electricity-carbon market synergy mechanism and can accurately characterize the relationship between actual unit output and emissions under multi-market coupling conditions, aiming to provide solid data support for low-carbon dispatch and optimized operation of the carbon market in the power system.

[0053] In related technologies, a method for calculating the carbon emission factor of electricity has been proposed. First, the total power generation of various types of generating units within a specific time scale is statistically analyzed. Then, based on the unit power generation carbon emission factor corresponding to different types of units, a weighted average method is used to obtain the overall carbon emission factor of the system. The core idea is to start from the power generation side, treating each unit as an independent emission unit, and using the proportion of power generation as the weight to characterize the system's carbon emission level. The advantage of this method is its simplicity and low data acquisition cost. However, it fails to consider the impact of changes in unit operating status and multi-market, multi-dimensional scheduling factors on actual emissions, thus having significant limitations in accurately characterizing the dynamic carbon emission characteristics of the system.

[0054] In summary, the above description highlights the main problems with the existing technology: the accuracy of the electrical energy clearing calculation results is low, which is caused by insufficient accuracy and timeliness in the calculation of the electrical carbon factor.

[0055] Based on this, this application proposes a joint clearing calculation method, including: inputting the acquired operating condition characteristics of the generator sets into a preset factor calculation model; performing joint learning mapping on the operating condition characteristics through the factor calculation model to obtain the dynamic carbon emission factor of the generator sets; inputting the dynamic carbon emission factor into a pre-established day-ahead unit combination model for optimization calculation to obtain the day-ahead joint clearing result; and using the unit start-up / shutdown status and generator set operating power in the day-ahead joint clearing result as constraints, inputting them into a pre-established intraday constraint scheduling model for rolling calculation to obtain the intraday joint clearing result. This application can improve the accuracy of joint clearing calculations, as detailed below.

[0056] The joint clearing calculation method provided in this application embodiment can be applied to, for example, Figure 1In the application environment shown, the data acquisition module 102 communicates with the server 104 via a network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104 or placed on the cloud or other network servers. The acquired generator set operating characteristics are input into a preset factor calculation model. The operating characteristics are jointly learned and mapped through the factor calculation model to obtain the generator set's dynamic carbon emission factor. The dynamic carbon emission factor is then input into a pre-established day-ahead unit combination model for optimization calculation to obtain the day-ahead joint clearing result. The unit start-stop status and generator set operating power in the day-ahead joint clearing result are used as constraints and input into a pre-established intraday constraint scheduling model for rolling calculation to obtain the intraday joint clearing result. The data acquisition module 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart vehicle devices, projection devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Headset devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. Server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0057] In one exemplary embodiment, such as Figure 2 As shown, a joint clearing calculation method is provided, which is applied to... Figure 1 Taking the server in the example of this, the explanation includes:

[0058] S210, obtain day-ahead unit parameters.

[0059] Among them, day-ahead unit parameters refer to all the input data required to calculate the next day's power generation plan before the operating day, including but not limited to: physical characteristic parameters of thermal / coal-fired power units, forecast and boundary parameters, market economic price parameters, etc. These parameters are known and are used to describe the external environment and physical characteristics of the thermal / coal-fired power units themselves.

[0060] In this embodiment of the application, the aforementioned day-ahead unit parameters can be obtained through a data acquisition device and a monitoring system.

[0061] S220 inputs the day-ahead unit parameters into the pre-established day-ahead unit combination model, and performs optimization calculations by combining the first dynamic carbon emission factor sequence calculated through the factor calculation model to obtain the day-ahead joint clearing result.

[0062] The aforementioned day-ahead unit combination model is a commonly used mathematical optimization model in this field. This model aims to minimize the total operating cost of all units and calculates the start-up and shutdown status and planned output of each unit in each time period. This model generally aims to minimize the total system operating cost.

[0063] The aforementioned factor calculation model is a pre-trained deep learning model, which can be implemented using a Transformer neural network structure in practical applications. This model establishes a mapping relationship between the operating parameters of thermal power / coal-fired power units and the dynamic carbon emission factor sequence at each time point. The first dynamic carbon emission factor sequence is a set of carbon emission intensities corresponding to unit power generation at multiple consecutive time points for the same thermal power / coal-fired power unit. Each factor value in this sequence is not a pre-calculated constant, but rather a decision variable to be optimized that dynamically changes with the load rate and actual output level of the thermal power / coal-fired power unit at different time points. Unlike the carbon emission factors using fixed regional averages in related technologies, the dynamic carbon emission factor sequence in this application can reflect a more realistic carbon emission level of thermal power / coal-fired power units under actual operating conditions.

[0064] In this embodiment of the application, a first dynamic carbon emission factor sequence is first generated through a factor calculation model. The operating characteristics of each unit can be input into the trained factor calculation model. By performing vector embedding and multi-head self-attention mechanism processing on these characteristics, the corresponding dynamic carbon emission factor sequence for each time period is output. The factor value at each moment in the sequence reflects the unit carbon emission intensity at this specific output level of the unit at that moment.

[0065] After obtaining the first dynamic carbon emission factor sequence, the day-ahead unit parameters and the first dynamic carbon emission factor sequence are input into the day-ahead unit combination model for optical optimization calculation. During the solution process, the solver repeatedly tries different unit start-up and shutdown combinations and output allocation schemes. Each attempt calls the factor calculation model again to obtain the dynamic carbon emission factor corresponding to the current output scheme of each unit, thereby obtaining new carbon emission cost feedback. After iterative convergence, the day-ahead unit combination model outputs the joint clearing results for the day-ahead stage. The most important part of the day-ahead joint clearing results is the start-up and shutdown status of each unit in each time period. This start-up and shutdown status will serve as a constraint condition for the subsequent intraday constrained scheduling model. The aforementioned day-ahead joint clearing results include at least the start-up and shutdown status of each unit in each time period, the output plan of each unit, and the frequency regulation capacity reservation value.

[0066] S230 uses the unit start-up and shutdown status in the day-ahead joint clearing results as a constraint condition, and combines it with the second dynamic carbon emission factor sequence calculated by the factor calculation model. The obtained intraday unit parameters are then substituted into the pre-established intraday constraint scheduling model for calculation to obtain the intraday joint clearing results.

[0067] Among them, the unit start-up and shutdown conditions refer to the status indicators of whether each thermal power / coal-fired power unit is in operation or shut down at any given time period.

[0068] Similarly, intraday unit parameters are the unit operation information obtained over time on the operating day, including but not limited to boundary parameters calculated based on the day-ahead unit model, measured parameters obtained based on the operation of thermal / coal-fired power units, market economic price parameters, etc. Intraday parameters have higher accuracy and real-time performance than day-ahead parameters, reflecting the real-time status of the power system operation.

[0069] The intraday constraint scheduling model refers to a calculation model that re-optimizes the power output of operating units when the start-up and shutdown status of the units is already determined. This model is a common mathematical optimization model in this field. With the goal of minimizing power generation cost, it re-optimizes the output value of each operating unit under the premise of satisfying constraints, and finally issues power generation instructions to each unit.

[0070] In this embodiment of the application, after entering the operating day, as the acquired unit operating information is updated, the factor calculation model is called again. Through the unit operating condition characteristics updated within the day, the factor calculation model calculates the second dynamic carbon emission factor sequence for the day stage through forward inference. This sequence reflects the carbon emission intensity estimate at each moment after being corrected according to the latest dynamic operating condition characteristics.

[0071] Then, the unit start-up and shutdown status from the previous day's joint clearing results are used as constraints. Combined with the second dynamic carbon emission factor sequence and intraday unit parameters, these are input into the intraday constrained scheduling model for calculation. Under fixed unit start-up and shutdown conditions, this model aims to minimize the sum of power generation costs and carbon emission costs. While satisfying all constraints, it re-optimizes the output values ​​of units in each operating state. The intraday constrained scheduling model also dynamically calls the factor calculation model during the iterative solution process to obtain carbon emission factor feedback corresponding to the current processing scheme until convergence. Finally, the intraday joint clearing results are output, which at least include the corrected real-time output of each unit. These results serve as the basis for final implementation by each power plant.

[0072] In related technologies, regional average carbon emission factors are often used, but this method cannot distinguish the differences in carbon emissions between different units operating at different load rates within the same region. This application uses a factor calculation model to dynamically output the carbon emission factor of each unit at each time point, truly reflecting the physical characteristics of thermal / coal-fired power units, and further improving the accuracy of subsequent clearing calculations. Furthermore, this application embeds the dynamic carbon emission factor sequence into the day-ahead unit combination model and the intraday constraint scheduling model, enabling the optimization of carbon emission costs and power generation costs in the same optimization model, thereby achieving a joint optimal balance between economic efficiency and low carbon emissions.

[0073] In one exemplary embodiment, such as Figure 3 As shown, the steps for calculating the first dynamic carbon emission factor sequence include:

[0074] S310, acquire the static and dynamic characteristics of the generator set; the static characteristics include the generator set type parameters, air-cooling type parameters, and unit capacity parameters, and the dynamic characteristics include the relative load data sequence corresponding to the historical power generation duration recorded by the generator set; the relative load data is composed of the day-ahead power generation and installed capacity of the unit to be optimized.

[0075] The operating condition characteristics are mainly composed of a combination of static and dynamic characteristics of the unit. The static characteristics include unit type parameters. Air-cooled form parameters Unit capacity parameters In this embodiment, the coupling relationship between static feature factors can be mined and the physical boundary can be fitted through neural networks. Since the temporal changes in load directly affect fuel efficiency, start-stop frequency, and load rate fluctuations, and thus determine the carbon emission level per unit of electricity and its dynamic pattern, this embodiment also selects the specific load data corresponding to the historical power generation duration. As a dynamic feature input, the load ratio data is usually expressed as a percentage or decimal, which is the ratio of the actual power generation of a thermal / coal-fired power unit to its rated installed capacity at a certain moment. The value of this ratio directly determines the carbon emission intensity of the unit.

[0076] It should be noted that in the day-ahead stage calculation of this embodiment, the specific load data is not the final known parameter, but a decision variable to be optimized. That is, the output value that the day-ahead unit combination model continuously adjusts during the solution process. In other words, the specific load data is in the form of a variable before optimization convergence, and the final numerical solution is only determined after the day-ahead unit combination model calculation is completed.

[0077] S320 performs feature embedding calculations on static and dynamic features in the time dimension to obtain operating condition features.

[0078] In this embodiment, the obtained static and dynamic features of the generator set are spliced ​​together to obtain:

[0079]

[0080] Then, using vector embedding, the original data, composed of static features and dynamic features at each time step, is combined in the time dimension. Perform feature mapping and embedding into tensor sequences. , where each e tThe vector representation of the operating condition characteristics at time t in the embedding space is used to obtain the aforementioned operating condition characteristics E, which serve as the input to the factor calculation model.

[0081] S330: Input the operating condition characteristics into the factor calculation model, and perform joint learning mapping on the operating condition characteristics through the factor calculation model to obtain the first dynamic carbon emission factor sequence.

[0082] In this embodiment, by introducing the day-ahead power generation to be optimized as the numerator of the load ratio data into the input features, the carbon emission factor is no longer a fixed constant, but a dynamic variable that is updated synchronously with the adjustment of the output scheme during the optimization iteration of the day-ahead unit combination model. This ensures that in the subsequent optimization process of the day-ahead unit combination model, any change in the output scheme will receive corresponding carbon emission factor feedback, and the calculation of carbon emission costs truly reflects the actual level of the units under the current exploratory operating conditions. Thus, it achieves the following: the adjustment of the output scheme drives the update of the carbon emission factor, and the change in carbon emission cost after the update of the carbon emission factor, in turn, affects the adjustment of the next output scheme, thereby forming a complete low-carbon optimization closed loop.

[0083] In one exemplary embodiment, such as Figure 4 As shown, the operating condition features are input into the factor calculation model, and the operating condition features are jointly learned and mapped through the factor calculation model to obtain the first dynamic carbon emission factor sequence, including:

[0084] S410, input the working condition features into the multi-head sub-attention layer in the factor calculation model to obtain the query vector, key vector and value vector output by the multi-head sub-attention layer.

[0085] In this embodiment, the input operating condition features are deeply modeled through a multi-head self-attention layer in the factor calculation model to capture the long-range dependence and local changes in the dynamic evolution of carbon emission factors, thereby obtaining the query vector, key vector, and value vector corresponding to each sub-attention head:

[0086]

[0087] in, , , These are the learnable parameter matrices for each attention head, Q. hA K hA and V hA These are the query vector, key vector, and value vector, respectively.

[0088] S420 uses the scaled dot product attention algorithm to combine the query vector, key vector and value vector corresponding to each attention head to calculate the correlation weight between each time step, and obtains the attention weight matrix. The attention weight matrix is ​​used to characterize the degree of influence of each historical moment on the current moment.

[0089] By using Scaled Dot-Product Attention, the correlation between time steps is calculated, specifically:

[0090]

[0091] Where, d K =d / h A , representing the dimension of the key vector in each attention head. Divide the calculation result by Its main function is to mitigate and stabilize the gradient. The above R... hA This is the attention weight matrix calculated by the h-th attention head. In summary, the attention weight matrix can be obtained.

[0092] S430, the attention weight matrix is ​​concatenated, and the dynamic carbon emission factor is obtained based on the concatenated attention weight matrix.

[0093] In this embodiment, the attention weight matrices of all attention heads are concatenated, then linearly mapped, and after softmax activation, the aforementioned dynamic carbon emission factor E can be obtained. f :

[0094]

[0095] in, This module provides a learnable parameter matrix. It captures the interaction between multi-scale nonlinear load evolution and static parameters in different subspaces, enhancing the model's ability to model the time-series response of load-driven carbon emission factors. This is particularly suitable for units with complex start-up and shutdown processes, high-frequency fluctuations, or unsteady-state operation. Multi-head attention enables hierarchical modeling of dynamic carbon emission drivers, providing a deep, structured time-series feature representation for carbon emission factor construction.

[0096] Understandably, the above example of calculating the first dynamic carbon emission factor sequence illustrates a specific method for calculating dynamic carbon emission factors based on a factor calculation model. Similarly, when calculating the second dynamic carbon emission factor sequence, the same factor calculation model and the same input data structure are used, i.e., the static characteristics of the generator units and the specific load data sequences for each moment of the day are obtained. However, the difference lies in that the numerator in the specific load data sequence of the day is replaced with the unit output value tentatively allocated by the daytime safety-constrained economic dispatch model in the current iteration round. By inputting the operating condition characteristic sequence of the daytime stage into the same factor calculation model for joint learning mapping, the second dynamic carbon emission factor sequence of the daytime stage can be output accordingly.

[0097] In an exemplary embodiment, day-ahead unit parameters are input into a pre-established day-ahead unit combination model, and optimized using a first dynamic carbon emission factor sequence calculated through a factor calculation model to obtain day-ahead joint clearing results, including:

[0098] Obtain the constraints of the day-ahead unit combination model;

[0099] Substituting the day-ahead unit parameters and the first dynamic carbon emission factor sequence into the day-ahead unit combination model, and combining the constraints of the day-ahead unit combination model, the day-ahead joint clearing result is obtained with the objective of minimizing the total day-ahead operating cost. The day-ahead joint clearing result includes the start-up and shutdown status of each generator unit and the operating power of the generator unit.

[0100] The constraints of the day-ahead unit combination model include power balance constraints, DC power flow constraints, unit output upper and lower limit constraints, unit ramping constraints, unit start-up and shutdown constraints, energy storage charging and discharging constraints, and energy storage state of charge constraints. Specifically:

[0101] Power balance constraints include:

[0102]

[0103] in, This represents the power flow from node j to node i at time t. This represents a collection of conventional generating units. It represents a collection of new energy generating units. This represents the real-time output power of the new energy unit g at node i at time t. This represents the incoming power to node i at time t. Let N represent the total load of node i at time t, and N be the total number of nodes. This represents the actual active power output of conventional generator set g connected to node i at time t. This represents the active power reserved or actually used at node i at time t to meet the needs of auxiliary services such as system frequency regulation and peak shaving. This represents the active power of the energy storage device (such as electrochemical energy storage) connected to node i at time t.

[0104] DC power flow constraints include:

[0105]

[0106] in, This represents the injected power flow at node i at time t. , Let i and j represent the phase angles of nodes i and j, respectively. This represents the branch reactance between nodes.

[0107] The upper and lower limits of the unit output are constrained as follows:

[0108]

[0109] Among them, u t,i,g This indicates the operating status of unit g at node i at time t; , These represent the minimum and maximum power output of unit g at node i, respectively.

[0110] The unit's ramp-up constraints are:

[0111]

[0112] in, Indicates coal-fired power units. and These represent the uphill and downhill limits for coal-fired power units, respectively, in MW / h.

[0113] The unit start-up and shutdown constraints are as follows:

[0114]

[0115]

[0116] in, and Let represent the minimum shutdown and startup times of coal-fired power unit g at node i, respectively. Indicates the current calculation time. Indicates at time A binary variable representing the start / stop status of coal-fired power unit g connected to node i, where a value of 1 indicates that the unit is in a state of start / stop. The unit is always powered on; a value of 0 indicates that the unit is in a state of constant operation. It is always in a shutdown state, similarly. For the same unit at the previous moment The start and stop states are binary variables.

[0117] Energy storage charging and discharging constraints include:

[0118]

[0119]

[0120]

[0121] in, , These represent the charging or discharging state of energy storage g at node i at time t, respectively. The charging / discharging state is 1, and the discharging state is 0 otherwise. , Let g represent the charging and discharging power of energy storage g at node i at time t, respectively. Indicates the maximum charging power of the energy storage. Indicates the maximum discharge power of the energy storage. This indicates the remaining energy storage capacity.

[0122] Energy storage state of charge constraints include:

[0123]

[0124]

[0125]

[0126] in, This represents the amount of energy stored in energy storage g at node i at time t. , Let represent the charging and discharging efficiencies of energy storage g at node i, respectively. , Let g represent the energy stored at node i at time 0 and time T, respectively. Indicates the initial energy storage capacity. This indicates the rated capacity of the energy storage.

[0127] In this embodiment, the day-ahead unit parameters and the first dynamic carbon emission factor sequence are input into the day-ahead unit combination model, and combined with the constraints given above, and with the goal of minimizing the total day-ahead operating cost, that is, minimizing the operating cost of all units in the system, the day-ahead joint clearing result is obtained.

[0128] In one exemplary embodiment, such as Figure 5 As shown, the solution yields the day-ahead joint clearing results, including:

[0129] S510, obtain the day-ahead total operating cost calculation function; the day-ahead total operating cost calculation function includes the electricity market cost item, carbon emission cost item, curtailment loss cost item, power shortage loss cost item, and unit start-up and shutdown cost item.

[0130] Among them, the functions for calculating total operating cost are:

[0131]

[0132] Furthermore, there are:

[0133]

[0134]

[0135]

[0136]

[0137]

[0138] in, , , , , Let represent the electricity market cost, carbon emission cost, curtailment loss cost, power shortage loss cost, and unit start-up and shutdown cost at time t, respectively. This represents the set of generator sets, where N represents the system node number. This represents the electricity price declared by unit g at node i at time t. This represents the declared output of unit g at node i at time t. This indicates the trading price corresponding to a unit of carbon emissions. For example, the closing price of the national carbon market on June 30, 2025, was 75.02 yuan / ton. This represents the dynamic carbon emission factor of unit g, which is the substitution term in the first dynamic carbon emission factor sequence mentioned above. This represents the cost per unit of abandoned electricity, typically determined using the feed-in tariff for renewable energy. This represents the amount of electricity discarded at node i at time t. The cost per unit of power shortage is typically related to the GDP generated per unit of load in a region. This represents the power shortage at time t at node i. This indicates the start / stop status of unit g at node i at time t. This indicates that the generator unit has started up. This indicates that the unit is shut down. , These represent the start-up cost and shutdown cost per unit capacity of a conventional generating unit (g), respectively. It is understandable that the total day-ahead operating cost can be calculated using the obtained day-ahead unit parameters.

[0139] S520 substitutes the day-ahead unit parameters and the first dynamic carbon emission factor sequence into the total cost calculation function, combines the constraints of the day-ahead unit combination model, calls the linear programming solver, and calculates the day-ahead joint clearing result.

[0140] In this embodiment, the total cost calculation function with day-ahead unit parameters substituted into it, along with all the constraints corresponding to the day-ahead unit combination model, are input into the linear programming solver.

[0141] After startup, the solver first constructs the feasible region space of the decision variables based on the constraints, and then searches within the feasible region for the combination of decision variables that minimizes the total cost function. In each iteration of the solver's trial and error, the solver generates a complete set of trial values ​​for the decision variables (including the start-up and shutdown status of each unit at each time period, output allocation, frequency regulation capacity reservation, and the power curtailment and power shortage at each node), and recalculates the specific value of the total cost function based on the current trial values. If the current trial solution has a lower cost than the previously recorded optimal solution and satisfies all constraints, it is replaced with the new optimal record. Subsequently, the solver automatically adjusts the assignment direction of the decision variables according to the optimization algorithm and generates the next set of trial values. The above iterative process is repeated until the solver's convergence criterion is met (such as the change in the objective function between two adjacent iterations being less than a preset threshold, or all feasible branches have been traversed).

[0142] After the solver converges, the optimal combination of decision variables is extracted and parsed from the solver's result cache, and the day-ahead joint clearing result is output according to the scheduling service requirement format. This result includes at least: the start-up and shutdown status variable values ​​of each unit in each time period (i.e., the final determined start-up and shutdown schedule) and the planned output values ​​of each unit in each time period (i.e., the day-ahead output allocation scheme).

[0143] In this embodiment, the total cost calculation function includes both the electricity market cost and the carbon emission cost, rather than calculating them separately or in stages. This means that when deciding on the start-up, shutdown, and output of the generator unit, the optimization solver must simultaneously balance the two objectives of minimizing the electricity purchase cost and minimizing the carbon emission allowance cost, and the final optimal solution is the optimal solution that combines both objectives.

[0144] In an exemplary embodiment, the unit start-up and shutdown status in the day-ahead joint clearing results is used as a constraint condition. Combined with the second dynamic carbon emission factor sequence calculated through a factor calculation model, the intraday unit parameters are substituted into a pre-established intraday constraint scheduling model for calculation, resulting in the intraday joint clearing results, including:

[0145] Obtain the constraints of the intraday constraint scheduling model;

[0146] Substituting the intraday initial parameters and the second dynamic carbon emission factor sequence into the intraday constrained scheduling model, and combining the constraints of the intraday constrained scheduling model, with the goal of minimizing the total intraday operating cost, the intraday joint clearing result is obtained. The intraday joint clearing result includes the corrected operating power and frequency regulation call amount of each generator unit.

[0147] The constraints of the intraday constrained scheduling model include the power balance constraints, DC power flow constraints, unit output upper and lower limit constraints, unit ramping constraints, energy storage charging and discharging constraints, energy storage state of charge constraints, as well as unit frequency regulation constraints and frequency regulation ancillary service upper limit constraints. Except for the unit frequency regulation constraints and frequency regulation ancillary service upper limit constraints, the remaining constraints have been specifically explained above. The unit frequency regulation constraints and frequency regulation ancillary service upper limit constraints are as follows:

[0148] The unit frequency regulation constraints are:

[0149]

[0150] in, , These represent the frequency regulation capacity and frequency regulation upper limit of unit g at node i at time t, respectively.

[0151] The upper limit constraints for FM ancillary services are:

[0152]

[0153] in, , Let g represent the frequency regulation capacity of energy storage g and generator g at node i at time t. , These represent the system's frequency modulation capacity requirement and frequency modulation mileage requirement at time t, respectively. , These represent the frequency regulation mileage multipliers for energy storage power station g and generator unit g, respectively. Frequency regulation mileage for energy storage power stations , Frequency regulation mileage of the generator unit .

[0154] In this embodiment, the intraday unit parameters and the second dynamic carbon emission factor sequence are input into the intraday constrained scheduling model. Combined with the constraints given above, the intraday constrained scheduling model uses the generator start-up and shutdown status determined by the day-ahead reliability unit combination model as the boundary input. With the goal of minimizing grid operating costs and maximizing market participants' profits, the model calculates the operating status of each unit, the nodal marginal price, and the frequency regulation marginal price, and solves to obtain the day-ahead joint clearing result.

[0155] In one exemplary embodiment, such as Figure 6 As shown, the solution yields the intraday joint clearing results, including:

[0156] S610, obtain the intraday total operating cost calculation function, which includes intraday cost calculation item, frequency regulation revenue item, and frequency regulation cost item; the frequency regulation revenue item includes spot market revenue and frequency regulation revenue.

[0157] The function for calculating the total daily operating cost is as follows:

[0158]

[0159] Among them are:

[0160]

[0161]

[0162] And specifically,

[0163]

[0164]

[0165]

[0166]

[0167] in, The day-ahead total operating cost can be calculated using the day-ahead unit combination model given above. , This represents the frequency regulation cost of the generating unit and energy storage at time t. , , , These represent the revenue from energy storage frequency regulation, the revenue from thermal power frequency regulation, the revenue from the energy storage spot market, and the revenue from the thermal power spot market, respectively. This represents the amount of energy discharged during frequency regulation. This indicates the amount of energy stored during frequency regulation. This indicates the spot electricity price for the corresponding settlement period; This indicates the amount of electricity discharged from energy storage in the spot market. This indicates the amount of electricity charged for energy storage in the spot market; This indicates the deviation or adjustment of electrical power generated by frequency regulation in thermal power units. This indicates the spot electricity price for the corresponding settlement period. This indicates the additional fuel costs incurred by thermal power units due to frequency regulation operation; This indicates the amount of electricity generated by thermal power units and fed into the grid. This indicates the settlement price in the spot market. This indicates the fuel cost for thermal power generation; Represents a set of generator sets. , These represent the price of the unit's frequency regulation capacity and the price of its frequency regulation mileage, respectively. This represents the frequency regulation mileage of unit g at node i at time t. Represents an energy storage collection. , These represent the price of frequency regulation capacity and the price of frequency regulation mileage for energy storage, respectively. The frequency regulation mileage of the energy storage g at node t is represented.

[0168] S620 substitutes the initial parameters and dynamic carbon emission factor into the daily total operating cost calculation function, substitutes the unit start-up and shutdown status and generator operating power in the day-ahead joint clearing results into the constraints of the daily constrained scheduling model, calls the linear programming solver, and calculates the daily joint clearing results.

[0169] In the intraday rolling optimization calculation, the trial values ​​of unit output at each moment of the day are input into the factor calculation model. After forward inference, the second dynamic carbon emission factor sequence is output, which, along with the intraday updated intraday unit parameters, is substituted into the intraday total operating cost calculation function. Simultaneously, the start-up and shutdown states of each unit output from the day-ahead unit combination model are used as fixed boundaries, and the day-ahead planned output is used as the initial value. These, along with constraints such as power balance and network security, constitute the constraint set of the intraday optimization model. With the objective function and constraints already assigned values, a linear programming solver is invoked for iterative solution. Each time a trial output scheme is proposed, the factor calculation model is re-invoked to update the carbon emission factor until the objective function converges, and the intraday joint clearing result is output.

[0170] The following is a preferred embodiment of the joint clearing calculation method.

[0171] 1. A factor calculation model is established. An initial factor calculation model is trained using historical data. After training, the model's internal parameters are fixed. By inputting any set of operating condition features, the carbon emission factor sequence for multiple corresponding time points can be forward-inferred. Furthermore, the factor calculation model is trained using supervised learning. The training set consists of a large number of historical samples. Each sample includes the static characteristic parameters of the thermal power unit within a continuous time period, the specific load data sequence for each time point within that period as input features, and the actual carbon emission factor directly measured by the monitoring system at the corresponding time point as the true label value. The model iteratively updates the network weights by minimizing the error between the predicted carbon emission factor and the measured label until convergence.

[0172] 2. On the day before the operating day, obtain the day-ahead unit parameters. For each unit, input its static characteristics and dynamic characteristics (composed of specific load data) into the factor calculation model. The factor calculation model infers and outputs the first dynamic carbon emission factor sequence. Substitute this first dynamic carbon emission factor sequence into the day-ahead total operating cost calculation function, and combine it with the constraints of the day-ahead unit combination model, call the mixed integer linear programming solver for iterative optimization. Finally, output the day-ahead joint clearing results, including the start-up and shutdown status of each unit at each time period, the unit output plan, and the frequency regulation capacity reservation results.

[0173] 3. On the operating day, the unit start-up and shutdown status output by the day-ahead model is used as a fixed boundary and remains unchanged. Static features and dynamic features composed of load ratio data are input into the factor calculation model. The load ratio data is not equal to the load ratio data in step two; the load ratio data in this step is the unit output trial value at each moment of the day. The above static and dynamic features are input into the same factor calculation model to obtain the second dynamic carbon emission factor sequence. This carbon emission factor sequence is substituted into the intraday total cost calculation function. Under the premise of locked start-up and shutdown status, and combined with the constraints of the intraday model, a linear programming solver is invoked for iterative optimization. The final intraday joint clearing result is output, including the corrected unit output commands, frequency regulation call amounts, and market marginal prices, serving as the final execution basis for distribution to each power plant.

[0174] In summary, during the day-ahead phase, the start-up and shutdown status, day-ahead power generation plan, and frequency regulation capacity reservation results for each type of unit are determined based on the day-ahead unit combination model. During the intraday phase, based on the intraday constrained scheduling model, and combined with updated load forecasts, renewable energy output forecasts, grid operating status, and frequency regulation demand, rolling adjustments are made to unit output, energy storage charging and discharging, renewable energy consumption, and frequency regulation service calls. During the aforementioned market clearing process, a dynamic carbon emission calculation method for coal-fired power units is simultaneously introduced. Based on changes in operating conditions such as coal-fired power unit output level, load factor, ramp-up status, start-up and shutdown status, and frequency regulation call status, the unit carbon emission intensity of coal-fired power units is dynamically adjusted. This adjustment is then combined with the unit output results obtained from market clearing to calculate carbon emissions and electrical carbon factors under different time periods and operating scenarios.

[0175] Through the above technical solution, this application can simultaneously reflect the impact of electricity market clearing, frequency regulation auxiliary service call and coal-fired power unit operating condition changes on carbon emission levels during the calculation of the electricity carbon factor, realize the synchronous characterization of market operation behavior and carbon emission response, and make the calculation results of the electricity carbon factor closer to the actual operating state of the power system.

[0176] Compared with traditional methods that rely on static average carbon emission factors or only on power generation for carbon emission accounting, this application can effectively avoid the shortcomings of traditional methods in reflecting market dispatch behavior, frequency regulation service call process and dynamic operating characteristics of coal-fired power units. It improves the dynamism, timeliness, refinement and accuracy of power carbon factor calculation, and provides more reliable technical support for power system low-carbon dispatch, carbon emission accounting, carbon market compliance and power carbon collaborative optimization decision-making.

[0177] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0178] Based on the same inventive concept, this application also provides a joint clearing calculation apparatus for implementing the joint clearing calculation method described above. The solution provided by this apparatus is similar to the implementation scheme described in the above method; therefore, the specific limitations in one or more joint clearing calculation apparatus embodiments provided below can be found in the limitations of the joint clearing calculation method described above, and will not be repeated here.

[0179] In one exemplary embodiment, such as Figure 7 As shown, a joint clearing calculation device is provided, comprising:

[0180] Module 71 is used to acquire day-ahead unit parameters;

[0181] The day-ahead calculation module 72 is used to input the day-ahead unit parameters into the pre-established day-ahead unit combination model, and perform optimization calculations by combining the first dynamic carbon emission factor sequence calculated by the factor calculation model to obtain the day-ahead joint clearing result.

[0182] The intraday calculation module 73 is used to take the unit start-up and shutdown status in the day-ahead joint clearing results as constraints, combine it with the second dynamic carbon emission factor sequence calculated by the factor calculation model, and substitute the obtained intraday unit parameters into the pre-established intraday constraint scheduling model for calculation to obtain the intraday joint clearing results.

[0183] Each module in the aforementioned joint clearing computing device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0184] In one exemplary embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement any of the joint clearing calculation methods described above.

[0185] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements any of the joint clearing calculation methods described above.

[0186] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements any of the joint clearing calculation methods described above.

[0187] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0188] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0189] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A joint clearing calculation method, characterized in that, The method includes: Obtain day-ahead unit parameters; The day-ahead unit parameters are input into a pre-established day-ahead unit combination model, and optimized by combining the first dynamic carbon emission factor sequence calculated by the factor calculation model to obtain the day-ahead joint clearing result. Using the unit start-up and shutdown status in the day-ahead joint clearing results as constraints, and combining the second dynamic carbon emission factor sequence calculated by the factor calculation model, the obtained intraday unit parameters are substituted into the pre-established intraday constraint scheduling model for calculation to obtain the intraday joint clearing results.

2. The method according to claim 1, characterized in that, The steps for calculating the first dynamic carbon emission factor sequence include: The static and dynamic characteristics of the generator set are obtained. The static characteristics include the generator set type parameters, air-cooling type parameters, and generator set capacity parameters. The dynamic characteristics include the specific load data sequence corresponding to the historical power generation duration of the generator set. The specific load data is constructed based on the day-ahead power generation and installed capacity of the generator set to be optimized. The static and dynamic features are subjected to feature embedding calculations in the time dimension to obtain the working condition features; The operating condition features are input into the factor calculation model, and the operating condition features are jointly learned and mapped through the factor calculation model to obtain the first dynamic carbon emission factor sequence.

3. The method according to claim 2, characterized in that, The step of inputting the operating condition features into the factor calculation model, and performing joint learning mapping on the operating condition features through the factor calculation model to obtain the first dynamic carbon emission factor sequence includes: The working condition features are input into the multi-head sub-attention layer in the factor calculation model to obtain the query vector, key vector and value vector output by the multi-head sub-attention layer; Using the scaled dot product attention algorithm, combined with the query vector, key vector and value vector corresponding to each attention head, the correlation weight between each time step is calculated to obtain the attention weight matrix. The attention weight matrix is ​​used to characterize the degree of influence of each historical moment on the current moment. The attention weight matrix is ​​concatenated, and the dynamic carbon emission factor is obtained based on the concatenated attention weight matrix.

4. The method according to any one of claims 1 to 3, characterized in that, The step of inputting the day-ahead unit parameters into a pre-established day-ahead unit combination model, and performing optimization calculations based on the first dynamic carbon emission factor sequence calculated through a factor calculation model to obtain the day-ahead joint clearing result includes: Obtain the constraints of the day-ahead unit combination model; The day-ahead unit parameters and the first dynamic carbon emission factor sequence are substituted into the day-ahead unit combination model. Combined with the constraints of the day-ahead unit combination model, the day-ahead joint clearing result is obtained with the goal of minimizing the total day-ahead operating cost. The day-ahead joint clearing result includes the start-up and shutdown status of each generator unit and the operating power of the generator units.

5. The method according to claim 4, characterized in that, The solution yields the day-ahead joint clearing result, including: Obtain the day-ahead total operating cost calculation function; the day-ahead total operating cost calculation function includes the electricity market cost item, carbon emission cost item, curtailment loss cost item, power shortage loss cost item, and unit start-up and shutdown cost item; The day-ahead unit parameters and the first dynamic carbon emission factor sequence are substituted into the day-ahead total operating cost calculation function. Combined with the constraints of the day-ahead unit combination model, the linear programming solver is called to calculate the day-ahead joint clearing result.

6. The method according to any one of claims 1 to 3, characterized in that, The step involves using the unit start-up and shutdown status in the day-ahead joint clearing results as constraints, combining this with the second dynamic carbon emission factor sequence calculated through the factor calculation model, and substituting the intraday unit parameters into a pre-established intraday constraint scheduling model for calculation to obtain the intraday joint clearing results, including: Obtain the constraints of the intraday constraint scheduling model; The intraday initial parameters and the second dynamic carbon emission factor sequence are substituted into the intraday constrained scheduling model. Combined with the constraints of the intraday constrained scheduling model, the intraday joint clearing result is obtained with the goal of minimizing the total intraday operating cost. The intraday joint clearing result includes the corrected operating power and frequency regulation call amount of each generator set.

7. The method according to claim 6, characterized in that, The solution yields the intraday joint clearing result, including: Obtain the intraday total operating cost calculation function, which includes an intraday cost calculation item, a frequency regulation revenue item, and a frequency regulation cost item; the frequency regulation revenue item includes spot market revenue and frequency regulation revenue. The intraday initial parameters and the dynamic carbon emission factor are substituted into the intraday total operating cost calculation function. The unit start-up and shutdown status and generator operating power in the day-ahead joint clearing results are substituted into the constraints of the intraday constrained scheduling model. The linear programming solver is called to calculate the intraday joint clearing results.

8. A joint clearing calculation device, characterized in that, The device includes: The acquisition module is used to acquire day-ahead unit parameters; The day-ahead calculation module is used to input the day-ahead unit parameters into a pre-established day-ahead unit combination model, and perform optimization calculations in combination with the first dynamic carbon emission factor sequence calculated by the factor calculation model to obtain the day-ahead joint clearing result. The intraday calculation module is used to take the unit start-up and shutdown status in the previous day joint clearing result as a constraint condition, and combine it with the second dynamic carbon emission factor sequence calculated by the factor calculation model. The obtained intraday unit parameters are substituted into the pre-established intraday constraint scheduling model for calculation to obtain the intraday joint clearing result.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images