A low-carbon optimal dispatching method, system, device and storage medium for a power system

Abstract

This invention relates to the field of power system dispatching technology, and provides a method, system, equipment, and storage medium for low-carbon optimized dispatching of power systems. The method includes acquiring day-ahead system forecast data, deep peak-shaving interval parameters, and basic system operating parameters for the target dispatching period. Based on the day-ahead system forecast data, deep peak-shaving interval parameters, and basic system operating parameters, a multi-objective low-carbon optimized dispatching model is solved, with the optimization objectives of minimizing source-side low-carbon dispatching costs, minimizing load-side low-carbon dispatching costs, minimizing total system carbon emissions, and minimizing renewable energy curtailment rate. This model is based on a two-way carbon responsibility sharing mechanism that considers load-side following characteristics and the contribution of source-side peak-shaving depth to thermal power units and load nodes to share carbon emission responsibilities. The resulting solution yields the target dispatching strategy. This invention can accurately assess load-side emission reduction contributions, eliminate the carbon penalty effect of deep unit dispatching, improve renewable energy absorption capacity, and reduce overall system carbon emissions and operating costs.

Description

Technical Field

[0001] This invention relates to the field of power system dispatching technology, and in particular to a low-carbon optimized dispatching method, system, equipment and storage medium for power systems. Background Technology

[0002] The power industry has a large carbon emission volume and is a key area for deep decarbonization. In addition, the high proportion of renewable energy integration makes the net load of the power system exhibit a typical "duck curve" characteristic, which poses challenges to peak-shaving capacity and operational flexibility.

[0003] While existing low-carbon optimization dispatching technologies for power systems can reduce carbon emissions and improve peak-shaving capacity to some extent, they still have several limitations: 1) The carbon responsibility sharing mechanism, which uses cosine similarity to measure the degree of load following renewable energy output, cannot capture the load time shift characteristics caused by demand response. This leads to a serious underestimation of the load-side emission reduction contribution of users who actively migrate their electricity consumption periods to areas with high renewable energy generation, weakening the incentive effect for users to participate in demand response and hindering the full realization of the carbon reduction potential of load-side response; 2) Directly using the equivalent carbon emission intensity of units that do not differentiate between the special characteristics of deep peak-shaving conditions as the basis for carbon emission rating results in the underestimation of the emission reduction contribution of users who participate in deep peak-shaving. Units with increased carbon emission intensity face higher carbon emission penalties, which discourages them from actively reducing output and making room for renewable energy consumption. This contradicts the objective law of achieving overall system emission reduction through deep peak shaving and restricts the full realization of the potential of deep peak shaving. 3) At the source-load coordinated control level, source-side peak shaving and load-side demand response are often treated separately. In extreme load or high-proportion renewable energy access scenarios, even if the source side enters the most difficult operating conditions of deep peak shaving or even oil injection peak shaving, the load side cannot perceive and respond in real time. As a result, the consumption space freed up by deep peak shaving is not fully utilized by the load side, and the problem of renewable energy curtailment is difficult to solve fundamentally, thus restricting the full realization of the potential of source-load coordinated emission reduction. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a low-carbon optimized scheduling method, system, equipment, and storage medium for power systems. By employing a source-load two-way carbon responsibility sharing mechanism that considers load-side following characteristics and source-side peak-shaving depth contributions, it achieves deep source-load coordinated control, accurately assesses the load-side time-shifted emission reduction contribution, eliminates source-side deep-shaving carbon penalties, and realizes accurate and fair sharing of carbon responsibility between thermal power units and load nodes, thereby improving the source-load coordinated emission reduction effect and the capacity for renewable energy absorption.

[0005] In a first aspect, embodiments of the present invention provide a low-carbon optimized dispatching method for a power system, the method comprising: The system acquires day-ahead system forecast data, deep peak-shaving interval parameters, and basic system operating parameters for the target scheduling cycle. The day-ahead system forecast data includes new energy node output forecast data, thermal power unit node output forecast data, load node demand forecast data, and load node transferable load forecast data. The deep peak-shaving interval parameters include peak-shaving intervals without oil injection and peak-shaving intervals with oil injection. Based on the day-ahead system forecast data, the deep peak-shaving interval parameters, and the system's basic operating parameters, a pre-constructed multi-objective low-carbon optimization scheduling model is solved to obtain the target scheduling strategy. The target low-carbon optimization scheduling model takes minimizing the source-side low-carbon scheduling cost, the load-side low-carbon scheduling cost, the total system carbon emissions, and the renewable energy curtailment rate as its optimization objectives, and is constructed based on a source-load two-way carbon responsibility sharing mechanism. The source-load two-way carbon responsibility sharing mechanism is to allocate carbon emission responsibilities to thermal power units and load nodes by considering the load-side following characteristics and the source-side peak-shaving depth contribution.

[0006] Furthermore, the construction steps of the multi-objective low-carbon optimization scheduling model include: Based on the real-time output data of thermal power units and the parameters of the deep peak shaving interval, carbon emission penalty exemption analysis is performed based on the contribution of source-side peak shaving depth to obtain the set of real-time carbon emission correction factors for the units. Based on the historical load demand data of the load nodes at the scheduling time and the corresponding historical total output data of new energy sources, load following characteristics analysis is performed to obtain a set of real-time load following characteristics factors; Based on the load real-time following characteristic factor set and the unit real-time carbon emission correction factor set, a scheduling cycle carbon responsibility cost analysis is performed based on the tiered carbon trading mechanism to obtain the source-side carbon responsibility allocation cost and the load-side carbon responsibility allocation cost. The source-side low-carbon dispatch cost is obtained based on the source-side carbon responsibility allocation cost, the net cost of deep peak shaving of thermal power units, and the operating cost of thermal power units. The load-side low-carbon dispatch cost is obtained based on the load-side carbon responsibility sharing cost, the load-side electricity purchase cost, and the load-side demand response adjustment cost. The multi-objective low-carbon optimization scheduling model is constructed based on a multi-objective optimization function and preset constraints, which are derived from the source-side low-carbon scheduling cost, the load-side low-carbon scheduling cost, the total carbon emissions of the system, and the renewable energy curtailment rate. The preset constraints include power balance constraints, unit operation constraints, energy storage state of charge constraints, load demand response constraints, and renewable energy output constraints.

[0007] Furthermore, the load following characteristic factor set includes the real-time following characteristic factors of each load node; The step of performing load following characteristic analysis based on the historical load demand data of the load nodes at the scheduling time and the corresponding historical total output data of new energy sources to obtain the load real-time following characteristic factor set includes: The historical load demand data of each load node in the historical load demand data of the load node is dynamically time-warped and analyzed with the historical total output data of the new energy to obtain the following pattern distance of the corresponding load node. The historical load demand data of each load node in the historical load demand data of the load node is averaged with the historical total output data of the new energy source over the entire time period to obtain the comprehensive average value of the corresponding load node. The normalized morphological distance of the corresponding load node is obtained by multiplying the comprehensive average value of each load node with the total number of scheduling period periods. The load following characteristic factor set is calculated based on the normalized morphological distance of all load nodes and the preset attenuation rate adjustment coefficient, using a preset exponential attenuation function.

[0008] Further, the step of performing scheduling cycle carbon responsibility cost analysis based on the real-time load following characteristic factor set and the real-time unit carbon emission correction factor set, and obtaining the source-side carbon responsibility allocation cost and the load-side carbon responsibility allocation cost based on the tiered carbon trading mechanism, includes: Based on the complex power tracking theory, real-time carbon emission component data of the unit's load-affecting nodes and real-time carbon emission of the total network loss of the system are obtained. Carbon emission rating analysis is performed based on the real-time carbon emission correction factor set of the unit and the real-time output data of the thermal power unit to obtain the carbon emission rating coefficient set of the unit. By fusing and analyzing the set of carbon emission rating coefficients for the generating units and the set of load following characteristic factors, a set of bidirectional source-load allocation coefficients is obtained. Carbon emission responsibility allocation is calculated based on the source-load bidirectional allocation coefficient set, the real-time carbon emission component data of the unit's load-affected nodes, and the real-time carbon emission of the total network loss of the system, to obtain the unit carbon emission responsibility allocation data and the load node carbon emission responsibility allocation data. Based on the carbon emission responsibility allocation data of the generating units and the carbon emission responsibility allocation data of the load nodes, a carbon responsibility cost analysis is performed based on the tiered carbon trading mechanism to obtain the source-side carbon responsibility allocation cost and the load-side carbon responsibility allocation cost.

[0009] Further, the step of performing carbon emission rating analysis based on the real-time carbon emission correction factor set of the unit and the real-time output data of the thermal power unit to obtain the carbon emission rating coefficient set of the unit includes: Based on the real-time output data of the thermal power unit, the carbon content of coal, the carbon oxidation rate and the carbon capture rate, the carbon emission intensity data of the thermal power unit is calculated based on the secondary coal consumption curve. Based on the real-time carbon emission correction factor set of the unit, the carbon emission intensity of the thermal power unit is corrected to obtain the corrected carbon emission intensity data of the thermal power unit. Based on the corrected carbon emission intensity data of the thermal power unit and the real-time output data of the thermal power unit, the equivalent carbon emission intensity of the unit's cycle is analyzed to obtain the equivalent carbon emission intensity data of the thermal power unit. The carbon emission intensity of the thermal power unit is assessed based on the equivalent carbon emission intensity data and the preset graded carbon emission intensity limit parameters to obtain the set of carbon emission rating coefficients for the unit.

[0010] Furthermore, the step of obtaining the load-side demand response adjustment cost includes: Based on the complex power tracking theory, the total carbon emissions data of the load nodes are obtained, and the real-time carbon potential data of the load nodes is obtained based on the total carbon emissions data of the load nodes and the real-time load data of the load nodes. Based on the capacity-weighted average of the real-time carbon emission correction factors of each unit in the real-time carbon emission correction factor set, the deep adjustment collaborative incentive factor is obtained, and based on the deep adjustment collaborative incentive factor, the time-of-use electricity price and the real-time carbon potential data of the load node, the real-time incentive data of the load node is obtained. Based on the deep-tuning collaborative excitation factor and the load following characteristic factor set, elastic demand response analysis is performed to obtain the real-time load elasticity coefficient set; Based on the real-time load elasticity coefficient set and the real-time load node excitation data, demand response adjustment analysis is performed to obtain load demand response adjustment data. The load-side demand response adjustment cost is obtained based on the load demand response adjustment data.

[0011] Furthermore, the step of solving the pre-constructed multi-objective low-carbon optimization scheduling model to obtain the target scheduling strategy includes: An initial population is generated based on tent mapping, and the fitness of the initial population is evaluated according to the objective optimization function in the multi-objective low-carbon optimization scheduling model to obtain an initial fitness set. Based on the initial fitness set, non-dominated solutions are extracted from the initial population to generate an initial external archive set; The initial population is updated according to the update mechanism of discoverers, followers, and vigilants to obtain the current iterative population; in the update mechanism, the position of the discoverer is updated using a dynamic adaptive inertia weight that decreases linearly based on the number of iterations. Add Cauchy mutation perturbation to the continuously stagnant individuals in the current iterative population, perform non-dominated sorting on the union of the resulting new iterative population and the initial external archive set, and update and maintain the resulting current external archive set based on the hypercube grid method. Determine whether the current iteration count has reached the maximum iteration count. If so, obtain the target scheduling strategy based on the current external file set. Otherwise, continue iterative search based on the current external file set until the maximum iteration count is reached, and obtain the target scheduling strategy.

[0012] Secondly, embodiments of the present invention provide a low-carbon optimized dispatching system for a power system, the system comprising: The data acquisition module is used to acquire day-ahead system forecast data, deep peak shaving interval parameters, and basic system operating parameters for the target scheduling cycle. The day-ahead system forecast data includes new energy node output forecast data, thermal power unit node output forecast data, load node demand forecast data, and load node transferable load forecast data. The deep peak shaving interval parameters include non-oil peak shaving intervals and oil peak shaving intervals. The low-carbon optimization scheduling module is used to solve a pre-constructed multi-objective low-carbon optimization scheduling model based on the day-ahead system forecast data, the deep peak-shaving interval parameters, and the system's basic operating parameters to obtain the target scheduling strategy. The target low-carbon optimization scheduling model aims to minimize the source-side low-carbon scheduling cost, the load-side low-carbon scheduling cost, the total system carbon emissions, and the renewable energy curtailment rate. It is constructed based on a source-load two-way carbon responsibility sharing mechanism. The source-load two-way carbon responsibility sharing mechanism allocates carbon emission responsibilities to thermal power units and load nodes by considering the load-side following characteristics and the source-side peak-shaving depth contribution.

[0013] Thirdly, embodiments of the present invention also provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method.

[0014] Fourthly, embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described method.

[0015] This invention provides a method, system, equipment, and storage medium for low-carbon optimized scheduling of power systems. The method acquires day-ahead system forecast data, including output forecast data of new energy nodes, output forecast data of thermal power unit nodes, demand forecast data of load nodes, and load transferable load forecast data of load nodes for the target scheduling period; parameters of deep peak shaving intervals, including those without oil-fired peak shaving intervals and those with oil-fired peak shaving intervals; and basic system operating parameters. Based on the day-ahead system forecast data, deep peak shaving interval parameters, and basic system operating parameters, the method solves a multi-objective low-carbon optimized scheduling model, which is pre-constructed with the optimization objectives of minimizing source-side low-carbon scheduling costs, minimizing load-side low-carbon scheduling costs, minimizing total system carbon emissions, and minimizing the new energy curtailment rate. This model is based on a two-way carbon responsibility sharing mechanism that considers load-side following characteristics and the contribution of source-side peak shaving depth to allocate carbon emission responsibilities between thermal power units and load nodes, to obtain the technical solution for the target scheduling strategy. Compared with existing technologies, this low-carbon optimized dispatching method for power systems can share carbon emission responsibilities between thermal power units and load nodes based on a two-way carbon responsibility sharing mechanism that considers load-side following characteristics and source-side peak-shaving depth contributions. This not only accurately identifies load time shifts caused by demand response and accurately assesses load-side emission reduction contributions, but also eliminates the penalty effect of deep unit dispatch, incentivizes units to actively conduct deep dispatch, improves the renewable energy absorption rate, and reduces the overall carbon emissions and operating costs of the system. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the low-carbon optimized scheduling method for power systems in an embodiment of the present invention. Figure 2 This is a schematic diagram comparing the equivalent carbon emission intensity of thermal power units and the carbon emission responsibility allocation value per unit of power generation obtained by the method of the present invention and the traditional method under different output conditions in the embodiments of the present invention. Figure 3 This is a schematic diagram comparing the net load curves of the system before and after using the source-load spatiotemporal coordinated control method provided by the present invention in an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of the low-carbon optimized dispatching system for the power system in an embodiment of the present invention; Figure 5 This is an internal structural diagram of the computer device in an embodiment of the present invention; The attached figures are labeled as follows: 1. Data acquisition module; 2. Low-carbon optimization scheduling module. Detailed Implementation

[0017] To make the objectives, technical solutions, and beneficial effects of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. Obviously, the embodiments described below are only part of the embodiments of this invention and are used to illustrate the invention, but are not intended to limit the scope of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0018] In one embodiment, such as Figure 1 As shown, a low-carbon optimized dispatching method for power systems is provided, including: S11. Obtain the day-ahead system forecast data, deep peak-shaving interval parameters, and basic system operating parameters for the target scheduling period. The target scheduling period can be understood as the time range within which the current power system needs to perform scheduling optimization, and may include multiple scheduling periods, such as one day as a scheduling cycle, with one scheduling cycle comprising 24 scheduling periods. The day-ahead system forecast data can be understood as the system operating status data for the target scheduling period, obtained through existing power system day-ahead scheduling optimization technology, based on the day-ahead system operating status data. This may include new energy node output forecast data, thermal power unit node output forecast data, and load node demand forecast data. The deep peak-shaving interval parameters include the non-oil-fueled peak-shaving interval and the oil-fueled peak-shaving interval. Furthermore, the non-oil-fuel peak shaving range can be understood as the real-time output range that allows thermal power units to participate in non-oil-fuel peak shaving, while the oil-fuel peak shaving range can be understood as the real-time output range that allows thermal power units to participate in oil-fuel peak shaving. The corresponding basic system operating parameters can be understood as the basic operating information required for the power system to perform dispatch optimization, which may include thermal power unit rated capacity data, energy storage system rated capacity, thermal power unit output limit data, thermal power unit ramp power limit data, energy storage state of charge limit data, demand response adjustment limit data (including the adjustable range of demand response for each load node), response satisfaction threshold, carbon emission quota data (including carbon emission quotas for each thermal power unit and each load node), and new energy output limits, etc.

[0019] S12. Based on the day-ahead system forecast data, the deep peak-shaving interval parameters, and the system's basic operating parameters, the pre-constructed multi-objective low-carbon optimization scheduling model is solved to obtain the target scheduling strategy. The target low-carbon optimization scheduling model takes the minimum source-side low-carbon scheduling cost, the minimum load-side low-carbon scheduling cost, the minimum total system carbon emissions, and the minimum renewable energy curtailment rate as optimization objectives, and is constructed based on a source-load two-way carbon responsibility sharing mechanism. The source-load two-way carbon responsibility sharing mechanism is to share carbon emission responsibilities between thermal power units and load nodes, taking into account the load-side following characteristics and the source-side peak-shaving depth contribution.

[0020] The multi-objective low-carbon optimization scheduling model in this embodiment uses the active power output sequence of all units (including thermal power units and new energy units), the start-up state sequence of all units, the charging and discharging sequence of the energy storage system, and the load demand response adjustment sequence of all load nodes within the target scheduling period as decision variables. It is constructed based on four optimization objectives: minimizing source-side low-carbon scheduling costs, minimizing load-side low-carbon scheduling costs, minimizing total system carbon emissions, and minimizing the new energy curtailment rate. Specifically, the construction steps of the multi-objective low-carbon optimization scheduling model include: Based on the real-time output data of thermal power units and the parameters of the deep peak shaving interval, a carbon emission penalty exemption analysis is performed based on the contribution of source-side peak shaving depth to obtain a set of real-time carbon emission correction factors for the units. Among them, the real-time output data of thermal power units includes the real-time active power output of all thermal power units at each scheduling moment within the target scheduling cycle. The corresponding carbon emission penalty exemption analysis can be understood as considering that the traditional unit carbon emission rating mechanism (the higher the carbon emission intensity, the heavier the carbon responsibility is allocated, so that the units that actively carry out deep peak shaving will actually suffer higher carbon emission penalties) will produce deep peak shaving penalty problems under deep peak shaving conditions (the output of thermal power units is greatly reduced, the total coal consumption decreases, but the coal consumption rate increases sharply with the decrease in output, resulting in a significant increase in carbon emission intensity). This contradicts the objective law that deep peak shaving is conducive to freeing up space for the consumption of new energy sources, thereby promoting the overall emission reduction of the system. Therefore, a carbon emission penalty analysis process is proposed to give corresponding exemption discounts to the corresponding carbon emission intensity based on the peak shaving stage (including the conventional non-oil peak shaving stage and the oil peak shaving stage) of the current output of thermal power units.

[0021] To ensure consistency in subsequent terminology, the critical output of a thermal power unit transitioning from its conventional operating range to the non-oil-injection peak-shaving range is defined as the lower limit of conventional peak-shaving; the critical output of a unit transitioning from the non-oil-injection peak-shaving range to the oil-injection peak-shaving range is defined as the lower limit of non-oil-injection deep peak-shaving; and the unit's minimum technical output is defined as the lower limit of oil-injection deep peak-shaving. The area above the lower limit of conventional peak-shaving is the conventional operating range, and the two critical points correspond to the non-oil-injection peak-shaving range and the oil-injection peak-shaving range, respectively.

[0022] Specifically, the real-time carbon emission correction factor for each thermal power unit in the real-time carbon emission correction factor set can be expressed as: In the formula, The maximum exemption coefficient for the peak-shaving period without oil injection is the upper limit, and its value range is: The preferred value is 0.30; This represents the upper limit of the maximum exemption coefficient for the oil injection peak-shaving zone. The range of values ​​is The preferred value is 0.55; for thermal power units Real-time active power output; and thermal power units The lower limit of peak-shaving output without oil injection and the lower limit of peak-shaving output with oil injection; for Real-time carbon emission correction factor for centralized thermal power units Real-time carbon emission correction factor: When the real-time active power output of a thermal power unit is greater than the lower limit of peak-shaving output without oil injection (when the unit is in the normal range and does not generate additional peak-shaving contribution), no exemption is granted; when the real-time active power output of a thermal power unit is greater than or equal to the lower limit of peak-shaving output with oil injection and less than or equal to the lower limit of peak-shaving output without oil injection (entering the peak-shaving range without oil injection), the lower the real-time active power output and the greater the space given up for new energy, the exemption coefficient increases linearly to [a certain value]. When the real-time active power output of a thermal power unit is less than the lower limit of its oil injection peak-shaving output (entering the oil injection peak-shaving range), it will incur additional oil injection costs and higher coal consumption rate losses. The exemption coefficient is [not specified in the original text]. Further linearly increase to This demonstrates full recognition of the contributions of thermal power units under extreme operating conditions.

[0023] The real-time carbon emission correction factor for thermal power units provided in this embodiment is strongly coupled with the boundary parameters of the non-oil-injection peak shaving interval and the oil-injection peak shaving interval in the deep peak shaving interval parameters. The piecewise linear exemption coefficient in the corresponding piecewise linear calculation design is positively correlated with the peak shaving depth of the unit, which can correct the fixed energy consumption item in the coal consumption rate (see the relevant description in the calculation process of carbon emission intensity data of thermal power units below). ( A linear approximate compensation is made for the carbon emission penalty caused by the fixed energy consumption coefficient of thermal power units, thereby encouraging thermal power units to actively adjust their emissions.

[0024] Based on the historical load demand data of load nodes at the scheduling time and the corresponding historical total output data of renewable energy, load following characteristic analysis is performed to obtain a set of real-time load following characteristic factors. Among them, the historical load demand data of load nodes at the scheduling time includes the historical load demand curves (historical load demand data) of all load nodes within a certain time range before the scheduling time (the same as the total number of time periods within the target scheduling cycle), and the historical total output data of renewable energy can be understood as the historical total output curve of renewable energy within a certain time range before the scheduling time (renewable energy historical total output data). The corresponding set of real-time load following characteristic factors includes the real-time following characteristic factors of each load node at the corresponding scheduling time.

[0025] To accurately identify the load time shift contribution caused by the demand response of each load node, enhance the incentive effect of user participation in demand response, and fully leverage the carbon reduction potential of load-side response, this embodiment preferably employs the Dynamic Time Warping (DTW) algorithm to calculate the following characteristics between the load demand and renewable energy output of each load node. Specifically, the step of analyzing the load following characteristics based on the historical load demand data of the load nodes at the scheduling time and the corresponding historical total renewable energy output data to obtain the load real-time following characteristic factor set includes: The historical load demand data of each load node in the historical load demand data of the load node is dynamically time-warped and analyzed with the historical total output data of the new energy to obtain the following pattern distance of the corresponding load node; wherein, the following pattern distance can be understood as the pattern distance between the historical load demand curve of each load node and the historical total output curve of the new energy obtained by optimally aligning the two through the warping path.

[0026] In practical applications, assume the current scheduling period corresponding load nodes Historical load demand data is , and They are the load nodes. exist Time and The load demand at any time, The total number of time periods in the target scheduling cycle; the historical total output data of new energy sources is... , and They are the load nodes. exist Time and The total output of new energy sources at any given time can be used to obtain the cumulative distance matrix for each load node based on the DTW algorithm. The corresponding recurrence relation is: In the formula, Cumulative distance matrix Medium load node The first in the historical load demand data The first time period in the historical total output data of new energy The cumulative bending distance between time periods and The range of values ​​is Boundary conditions are ; Optimal bending path from arrive The path with minimum cumulative cost must satisfy boundary, continuity, and monotonicity constraints; the corresponding load nodes... The following pattern distance can be expressed as: In the formula, For load nodes The following pattern distance can identify load nodes through time axis curvature. The historical load curve still matches the historical total output curve of new energy sources after a time shift. This can effectively solve the problem that the existing cosine similarity can only measure the consistency of the direction of the vector inner product and is completely insensitive to the time axis misalignment. As a result, the emission reduction contribution of users who actively migrate their electricity consumption period to the period of high new energy generation is seriously underestimated. This can effectively improve the incentive effect of users participating in demand response and make it easier to give full play to the carbon reduction potential of load response when source and load are coordinated for low carbon.

[0027] The historical load demand data of each load node in the historical load demand data of the aforementioned load nodes are averaged over all time periods with the historical total output data of the new energy sources to obtain the comprehensive average value of the corresponding load node; that is, the load node The overall mean can be expressed as: In the formula, For load nodes The overall mean can be used to eliminate the influence of power magnitude during subsequent normalization.

[0028] The following pattern distance of the corresponding load node is normalized by multiplying the comprehensive average value of each load node by the total number of scheduling period segments, resulting in the normalized pattern distance. The normalized pattern distance can be expressed as: In the formula, For load nodes Normalized morphological distance; This represents the total number of time slots in the scheduling cycle (the total number of scheduling time slots in the target scheduling cycle).

[0029] The load following characteristic factor set is calculated based on the normalized morphological distance of all load nodes and the preset attenuation rate adjustment coefficient, using a preset exponential attenuation function; wherein, the load following characteristic factor of each load node in the load following characteristic factor set can be expressed as: In the formula, The preset attenuation rate adjustment coefficient has a value range of [value range missing]. The preferred value is 5; For load nodes The load following characteristic factor has a value range of 100%. The smaller the normalized shape distance, the larger the corresponding load following characteristic factor, indicating that the shape of the historical load curve is more consistent with the historical total output curve of new energy and the degree of following is higher, and the load node needs to bear less carbon responsibility; conversely, the larger the normalized shape distance, the closer the corresponding load following characteristic factor is to 0, and the load node needs to bear more carbon responsibility.

[0030] The load real-time following characteristic factor construction mechanism provided in this embodiment can accurately capture the time shift characteristics caused by the load demand response of each load node, effectively correct the underestimated emission reduction contribution of user behavior, and thus enhance the incentive effect of user participation in demand response.

[0031] Based on the load real-time following characteristic factor set and the unit real-time carbon emission correction factor set, a scheduling cycle carbon responsibility cost analysis is performed based on a tiered carbon trading mechanism to obtain the source-side carbon responsibility allocation cost and the load-side carbon responsibility allocation cost. The scheduling cycle carbon responsibility cost analysis can be understood as follows: within the entire target scheduling cycle, according to the load real-time following characteristic factor of each load node and the real-time carbon emission correction factor of each thermal power unit, the real-time carbon emissions of each load node and the total real-time carbon emissions of the system's network loss are allocated to each thermal power unit and each load node. Then, according to the tiered carbon trading mechanism, the carbon responsibility allocation costs of each thermal power unit and each load node are statistically summarized in segments to obtain the corresponding source-side carbon responsibility allocation cost and load-side carbon responsibility allocation cost.

[0032] Specifically, the step of performing a scheduling cycle carbon responsibility cost analysis based on the real-time load following characteristic factor set and the real-time unit carbon emission correction factor set, and obtaining the source-side carbon responsibility allocation cost and the load-side carbon responsibility allocation cost, includes: Based on complex power point tracking (CPLT) theory, real-time carbon emission component data of load nodes affected by generating units and real-time carbon emissions of total system network losses are obtained. CPLT theory can be understood as an extension of traditional power flow tracking, which only focuses on active power, to CPLT encompassing both active and reactive power, thus enabling more accurate identification of carbon emission responsibility through a collaborative analysis of electricity and carbon emissions. This embodiment introduces the existing complex proportional sharing principle for CPLT to account for the impact of reactive power on network losses, constructing basic carbon emission accounting models for load nodes and total system network losses. These models serve as the accounting basis for the basic carbon emissions corresponding to each load node and the total system network loss, and as the analytical basis for subsequent carbon emission responsibility allocation between thermal power units and load nodes using a source-load two-way carbon responsibility sharing mechanism.

[0033] In practical applications, it is assumed that the scheduling time in the power system... node Injection Node The complex power is In the middle, flow to the node The components are: In the formula, For scheduling time node via node Injected complex power; For scheduling time node Total injected complex power; For scheduling time node Flow to Node The complex power. Define the tracking matrix. Its elements are: In the formula, To the node The set of nodes for input power; For scheduling time node Total injected complex power; For scheduling time node Flow to Node Complex power; tracking coefficient matrix Its elements For scheduling time node For nodes The proportion of power contribution.

[0034] In this embodiment, the real-time carbon emission component data of the load nodes affected by the generating units includes the real-time basic carbon emission component of each thermal power unit for each load node (in actual applications, the real-time carbon emission intensity of new energy generating units is zero and is not considered); the real-time carbon emission of the total system network loss can be understood as the real-time carbon emission corresponding to the total network loss of the power system obtained by processing the lossy network as an equivalent lossless network. In practical applications, both the real-time carbon emission component data of the load nodes affected by the generating units and the real-time carbon emission of the total system network loss can be calculated based on the tracking coefficient matrix obtained above and the real-time carbon emission intensity of each thermal power unit. The specific calculation process is as follows: In the formula, To take the real part of the complex number; For scheduling time Load Node The load power; For scheduling time Load Node Total injected complex power; The combined power output of thermal power units; For scheduling time The real-time carbon emission intensity of thermal power units can be analyzed and calculated based on the secondary coal consumption curve. The specific calculation process can be described in the following section on the acquisition process of carbon emission intensity data of thermal power units in the unit carbon emission rating coefficient collection, which will not be detailed here. For scheduling time thermal power units For each load node Real-time baseline carbon emissions.

[0035] To accurately allocate carbon emissions from network losses, a virtual load node with a load value equal to the line loss is introduced into each branch, treating the lossy network as an equivalent lossless network. Assuming the scheduling time... branch road Network loss recovery power From thermal power units The components are: In the formula, For scheduling time branch road Inject complex power at the head end; branch road First node number; For tracking coefficient matrix Middle First Node Corresponding thermal power units Element; For scheduling time branch road From thermal power units The network loss complex power component; correspondingly, the real-time carbon emissions of the total network loss of the system can be expressed as: In the formula, The total number of branch roads; This represents the total number of thermal power units. For scheduling time The system's total network loss and real-time carbon emissions.

[0036] Carbon emission rating analysis is performed based on the real-time carbon emission correction factor set and the real-time output data of the thermal power units to obtain a set of carbon emission rating coefficients for the units; wherein, the set of carbon emission rating coefficients for the units includes the carbon emission rating coefficients of each thermal power unit at the current dispatch time; specifically, the step of performing carbon emission rating analysis based on the real-time carbon emission correction factor set and the real-time output data of the thermal power units to obtain a set of carbon emission rating coefficients for the units includes: Based on the real-time output data, coal carbon content, carbon oxidation rate, and carbon capture rate of the thermal power units, carbon emission intensity data of the thermal power units is calculated based on the secondary coal consumption curve; wherein, the carbon emission intensity data of the thermal power units includes the basic carbon emission intensity of each thermal power unit, which can be expressed as: In the formula, , and thermal power units The carbon content, carbon oxidation rate, and carbon capture rate of the coal, with the carbon capture rate of ordinary thermal power units taken as 0; and They are respectively and C The molar masses were taken as 44 g / mol and 12 g / mol, respectively; , and For thermal power units Coal consumption characteristic parameters; For scheduling time thermal power units Real-time active power output; For scheduling time thermal power units Real-time carbon emission intensity.

[0037] In the above formula for calculating the real-time carbon emission intensity of thermal power units, the coal consumption function corresponding to the secondary coal consumption curve... Divided by real-time active power This is the coal consumption rate, which, when expanded, gives the coal consumption rate as _____. Among them, fixed energy consumption items With output The decreasing and monotonically increasing trend reflects the physical fact that thermal power units, regardless of their active power output, must maintain fixed energy consumption such as boiler ignition and auxiliary equipment operation. Therefore, the lower the real-time active power output of a thermal power unit, the larger the proportion of fixed energy consumption in total power generation, the higher the coal consumption rate, and the corresponding real-time carbon emission intensity. And it rose accordingly.

[0038] Based on the real-time carbon emission correction factor set of the unit, the carbon emission intensity of the thermal power unit is corrected to obtain the corrected carbon emission intensity data of the thermal power unit; wherein, the corrected carbon emission intensity data of the thermal power unit includes the corrected carbon emission intensity of each thermal power unit, which can be expressed as: Based on the correction factor, the unit exist The corrected carbon intensity at any given time is: In the formula, For scheduling time thermal power units The corrected carbon emission intensity is the effective carbon emission intensity after deducting the deep adjustment contribution exemption based on the corresponding real-time carbon emission correction factor, and is used for subsequent allocation calculations. In practical applications, when thermal power units are in deep adjustment mode, And because Strongly coupled with the boundary parameters of the two-stage peak-shaving interval and linearly amplified with the peak-shaving depth, the corrected effective carbon emission intensity meets the requirements. ,Right now Effective suppression within the deep peak-shaving range can fundamentally eliminate the risk that, under the traditional static carbon accounting paradigm, thermal power units, after entering the deep peak-shaving range, will fall into a deep peak-shaving penalty trap where the deeper the peak-shaving, the heavier the carbon penalty, and be assigned higher carbon responsibilities under the tiered carbon trading mechanism. This would severely suppress the willingness of thermal power units to engage in deep peak-shaving.

[0039] Based on the corrected carbon emission intensity data and the real-time output data of the thermal power units, an equivalent carbon emission intensity analysis of the unit cycle is performed to obtain the equivalent carbon emission intensity data of the thermal power units; wherein, the equivalent carbon emission intensity data of the thermal power units includes the equivalent carbon emission intensity of each thermal power unit within the target scheduling cycle, which can be expressed as: In the formula, For thermal power units The equivalent carbon emission intensity.

[0040] Based on the equivalent carbon emission intensity data of the thermal power units and the preset graded carbon emission intensity limit parameters, a graded assessment of the unit carbon emissions is conducted to obtain the set of unit carbon emission rating coefficients. The preset graded carbon emission intensity limit parameters can be set into three levels—low-carbon, medium-carbon, and high-carbon—based on industry-wide thermal power unit carbon emission rating principles. This ensures that each level corresponds to a different rating coefficient mapping relationship, allowing low-carbon units to obtain lower rating coefficients (bearing less carbon responsibility) and high-carbon units to obtain higher rating coefficients (bearing more carbon responsibility), forming a differentiated incentive structure for carbon responsibility allocation. The carbon emission rating coefficients of each thermal power unit in the corresponding set of unit carbon emission rating coefficients can be obtained based on the following piecewise function mapping: In the formula, , and These are the carbon emission intensity limits corresponding to the low-carbon, medium-carbon, and high-carbon levels in the preset graded carbon emission intensity limit parameters, respectively. ; As the industry benchmark carbon emission intensity; For thermal power units The carbon emission rating coefficient of actively adjusted thermal power units is lower due to the reduction in equivalent carbon emission intensity. The mechanism provided in this embodiment for correcting the equivalent carbon emission intensity of the unit based on the peak shaving depth can effectively suppress the corrected effective carbon emission intensity within the deep shaving range, resulting in less carbon responsibility and a significant reduction in the carbon trading penalty cost calculated subsequently. This breaks the inverse bond between deep shaving of the unit and high carbon fines from the carbon accounting level, and achieves a positive incentive for deep shaving behavior of the unit.

[0041] The set of unit carbon emission rating coefficients and the set of load following characteristic factors are fused and analyzed to obtain the set of source-load bidirectional allocation coefficients. This set includes the source-load bidirectional allocation coefficients between each thermal power unit and each load node. Each source-load bidirectional allocation coefficient can be understood as a comprehensive allocation coefficient constructed by introducing a weighted correction method based on physical tracking, considering that the tracking coefficient matrix obtained based on complex power tracking theory already gives the physical power contribution ratio of each thermal power unit to each load node. However, this physical ratio does not differentiate between source-side active emission reduction behavior (deep peak shaving) and load-side active emission reduction behavior (demand response time shift), and lacks incentive attributes. This ensures a fair allocation effect where the greater the contribution of source-load allocation, the less carbon responsibility is borne. The source-load bidirectional allocation coefficient between each thermal power unit and each load node can be expressed as follows: In the formula, , These are weighting coefficients, formulated based on the key targets of emission reduction responsibilities. The range of values ​​is , The range of values ​​is Preferred selection , It remains unchanged during the scheduling period; For load nodes The corresponding load following characteristic factor; For thermal power units With load nodes The dynamic adjustment capability of the source-load bidirectional allocation coefficient is borne by the carbon emission rating coefficient and the load following characteristic factor as the scheduling status changes. This makes the source-load bidirectional allocation coefficient have a positive incentive attribute for active emission reduction behavior on both the source and load sides, so as to achieve fair allocation: the larger the source-load bidirectional allocation coefficient, the greater the contribution of thermal power units to the carbon responsibility of the load node; conversely, the load node bears more carbon responsibility.

[0042] Carbon emission responsibility allocation is calculated based on the source-load bidirectional allocation coefficient set, the real-time carbon emission component data of the unit's impact on load nodes, and the real-time carbon emission of the total system network loss. This yields unit carbon emission responsibility allocation data and load node carbon emission responsibility allocation data. The unit carbon emission responsibility allocation data includes the carbon emission responsibility allocation value for each thermal power unit, and the load node carbon emission responsibility allocation data includes the carbon emission responsibility allocation value for each load node. In practical applications, the specific calculation formula for carbon emission responsibility allocation is as follows: In the formula, For thermal power units The allocation factor for the real-time carbon emissions of the total network loss of the system is equal to that of the thermal power unit. The proportion of active power in the total power generation of the system; For load nodes The allocation factor for the real-time carbon emissions of the total network loss of the system is the value of the load node. The proportion of real-time active power load demand in the total system load demand; This represents the number of load nodes. This represents the number of thermal power units. and Within the target scheduling period thermal power units and load nodes The carbon emission responsibility allocation value.

[0043] Figure 2 This diagram illustrates a comparison between the equivalent carbon emission intensity and carbon emission responsibility allocation per unit of power generation obtained by the method of this invention and the conventional method under different power output conditions. Figure 2 As shown, with the reduction in unit output, the equivalent carbon emission intensity of thermal power units calculated by both the method in this embodiment and the traditional method continues to increase, and increases significantly in the lowest output range, indicating that the objective law of increased coal consumption rate under deep peak shaving remains unchanged. However, in terms of carbon emission responsibility allocation, the carbon responsibility of thermal power units continues to increase after entering the deep peak shaving range under the traditional method, while the carbon responsibility of thermal power units begins to decrease after entering the deep peak shaving range without oil injection, and further decreases after entering the oil injection peak shaving range, and is significantly lower than that of the traditional method in the deepest peak shaving range. This result shows that the carbon emission responsibility allocation mechanism provided in this embodiment can effectively transmit the benefits of deep peak shaving exemption to the carbon responsibility allocation link while maintaining the objectivity of physical carbon emission intensity, thereby eliminating the "deep peak shaving penalty" effect and enhancing the initiative of units to actively shave peaks.

[0044] Based on the carbon emission responsibility allocation data of the generating units and the carbon emission responsibility allocation data of the load nodes, a carbon responsibility cost analysis is performed based on the tiered carbon trading mechanism to obtain the source-side carbon responsibility allocation cost and the load-side carbon responsibility allocation cost. The tiered carbon trading mechanism can be understood as dividing the difference between the carbon emission responsibility of the thermal power units or load nodes and their corresponding carbon emission quotas into several tiered intervals, with each tier corresponding to an increasing carbon price. This non-linear penalty structure drives the thermal power units or load nodes to pay higher marginal carbon costs when they exceed their emission quotas, incentivizing them to take priority emission reduction actions to reduce carbon emissions to within the corresponding carbon emission quotas, and imposing stronger economic constraints on thermal power units or load nodes that exceed their quotas.

[0045] In this embodiment, the source-side carbon responsibility allocation cost can be understood as the total carbon responsibility allocation cost of all thermal power units in the target scheduling period, and the load-side carbon responsibility allocation cost is the total carbon responsibility allocation cost of all load nodes in the target scheduling period. Since both the source-side and load-side carbon responsibility allocation costs are statistically calculated based on the tiered carbon trading mechanism and have the same calculation principle, the following explanation will only use the calculation formula of the source-side carbon responsibility allocation cost as an example: In the formula, In the formula: The initial price for carbon trading; The interval length; The percentage increase in prices; For thermal power units The carbon emission quota is obtained through the carbon emission quota data in the system's basic operating parameters; For thermal power units The cost of sharing carbon responsibility; Costs of carbon responsibility are shared at the source.

[0046] In practical applications, simply replacing the carbon emission responsibility allocation value of the thermal power unit in the above calculation formula with that of the load node will yield the calculation formula for the required load-side carbon responsibility allocation cost, which will not be elaborated here.

[0047] The carbon responsibility cost analysis method provided in this embodiment, which combines a two-way carbon responsibility sharing mechanism based on load-side following characteristics and source-side peak-shaving depth contribution to allocate carbon emission responsibilities between thermal power units and load nodes with a tiered carbon trading mechanism, can effectively improve the fairness and accuracy of source-side and load-side carbon responsibility sharing cost analysis, and provide reliable technical support for improving the source-load coordinated emission reduction effect.

[0048] The source-side low-carbon dispatch cost is obtained based on the source-side carbon responsibility allocation cost, the net cost of deep peak shaving for thermal power units, and the operating cost of thermal power units; wherein, the net cost of deep peak shaving for thermal power units is calculated based on the total cost of peak shaving without oil injection, the total cost of peak shaving with oil injection, and the total peak shaving compensation revenue, and can be expressed as: , , In the formula, and These are the unit cost coefficients for no oil injection and peak shaving with oil injection, respectively. ; and These are the total costs for peak shaving without oil injection and the total costs for peak shaving with oil injection, respectively. , and The first The output value, duration, and compensation coefficient of the peak shaving depth are all considered; the greater the peak shaving depth, the higher the compensation coefficient. The number of peak-shaving segments within the target scheduling cycle; For total peak-shaving compensation revenue; This represents the net cost of deep peak shaving for thermal power units.

[0049] In this embodiment, the operating cost of thermal power units can be understood as the sum of the coal consumption cost and start-up and shutdown cost of all thermal power units within the target scheduling period, and can be expressed as: In the formula, For scheduling time thermal power units The start / stop status is indicated by 1 for running and 0 for stopping. For thermal power units The cost of a single start-up is divided into hot start costs based on the duration of downtime. and cold start costs ; and These are the coal consumption cost of thermal power units and the start-up and shutdown cost of thermal power units, respectively. This refers to the operating cost of thermal power units.

[0050] By summing the source-side carbon responsibility allocation cost, the net cost of deep peak shaving for thermal power units, and the operating cost of thermal power units obtained above, the source-side low-carbon dispatch cost can be obtained, which can be expressed as: In the formula, To reduce the cost of low-carbon dispatching at the source.

[0051] Based on the load-side carbon responsibility allocation cost, load-side electricity purchase cost, and load-side demand response adjustment cost, the load-side low-carbon dispatch cost is obtained; wherein, the load-side demand response adjustment cost can be understood as being obtained through dynamic elastic demand analysis of electricity-carbon coupling, considering the real-time transmission of deep peak-shaving conditions of source-side thermal power to load-side demand response and the joint adjustment of load following characteristic factors; specifically, the steps for obtaining the load-side demand response adjustment cost include: Based on the complex power tracking theory, the total carbon emissions data of the load nodes are obtained, and the real-time carbon potential data of the load nodes is obtained based on the total carbon emissions data of the load nodes and the real-time load data of the load nodes. The total carbon emissions data of the load nodes may include the total carbon emissions of each load node, and the total carbon emissions of each load node is the cumulative value of the real-time baseline carbon emission components of the thermal power unit for each load node obtained by the aforementioned complex power tracking theory, which can be expressed as: In the formula, For load nodes Total carbon emissions.

[0052] The real-time load data for each load node includes the real-time load demand of each load node, and the corresponding real-time carbon potential data for each load node includes the real-time carbon potential of each load node. The real-time carbon potential of each load node can be expressed as: In the formula, and Scheduling time Load Node Real-time carbon potential and real-time load demand.

[0053] Based on the capacity-weighted average of the real-time carbon emission correction factors of each unit in the set of real-time carbon emission correction factors, a deep-shifting coordinated incentive factor is obtained. Furthermore, based on the deep-shifting coordinated incentive factor, time-of-use electricity prices, and the real-time carbon potential data of the load nodes, real-time incentive data for the load nodes is obtained. The deep-shifting coordinated incentive factor represents the overall deep peak-shaving intensity of the quantified scheduling system, and amplifies the load-side incentive signal under extreme deep-shifting conditions through nonlinear mapping. It can be expressed as: In the formula, This is the amplification factor for the synergistic excitation, and its value range is... The preferred value is 8; For thermal power units The rated capacity is used for capacity weighting to reflect the dominant effect of deep regulation contribution from large-capacity units; For scheduling time The deep-tuned synergistic incentive factor, the square term makes The response to the degree of deep adjustment is nonlinear: when all thermal power units in the system are in the normal operating range... , Synergistic incentives have no additional effects; when a large number of thermal power units enter the peak-shaving range without oil injection or even with deep oil injection, The weighted mean increases. Due to the drastic amplification of the squared terms (up to a maximum of... The signal broadcasts a strong coordinated regulation signal to the load side, which incentivizes the load to significantly increase the transfer amount during the most difficult period of deep regulation. The deep regulation coordinated excitation factor constitutes a mathematical bridge for the transmission of the deep regulation state from the source side to the load side. Together with the reverse channel of the load following characteristic factor obtained earlier, which feeds back from the load side to the carbon allocation of the source side, it forms a closed-loop system of spatiotemporal coordinated regulation between the source and load.

[0054] In this embodiment, the real-time excitation data for load nodes includes the real-time excitation signals of each load node. These real-time excitation signals are comprehensive excitation guidance signals obtained by fusing the real-time carbon potential, time-of-use electricity price, and deep-adjustment synergistic excitation factors of the load nodes, and can be expressed as: In the formula, , These are the electricity price guidance weights and carbon potential guidance weights, respectively, satisfying... , The range of values ​​is , The range of values ​​is This is to reflect the dominant role of carbon potential guidance in low-carbon scheduling; For scheduling time Time-of-use electricity pricing; The average electricity price for the target scheduling period; For scheduling time The system average carbon potential; For scheduling time Load Node The real-time excitation signal is used to characterize the strength of two types of guiding components in a deviation-normalized form: when the real-time carbon potential of the load node is higher than the system average carbon potential ( When the carbon potential deviation term is positive, it drives the load of that load node to switch out of the high-carbon period; when the real-time carbon potential of the load node is lower than the system average carbon potential, the carbon potential deviation term is negative, driving the load of the load node to switch into the low-carbon (usually the period of high renewable energy generation) period; the electricity price deviation term has the same logic, driving the load to switch out during the high-price period and driving the load to switch in during the low-price period.

[0055] In the above formula for calculating the real-time excitation signal of the load node, the amplification and adjustment of the carbon potential deviation term by the deep-tuning collaborative excitation factor enables the source-side deep-tuning information to be embedded into the load-side guidance signal in real time, thereby realizing the spatiotemporal guidance of source-load collaboration.

[0056] Based on the deep-tuning collaborative excitation factor and the load following characteristic factor set, elastic demand response analysis is performed to obtain a real-time load elasticity coefficient set. This set includes the real-time elasticity coefficients of each load node, and these coefficients are jointly adjusted by the deep-tuning collaborative excitation factor and the load following characteristic factors of each load node, and can be expressed as follows: in, The reference elastic coefficient has a range of values ​​of [value range missing]. The preferred value is ; To adjust the collaborative elasticity amplification factor, the value range is: ; The value range is the elastic amplification factor for the following characteristics. ; For scheduling time Load Node The real-time elasticity coefficient consists of two parts: the deep-adjustment collaborative incentive factor corresponds to external environmental incentives; the more severe the overall deep adjustment of the system, the more sensitive the load-side response should be, and the larger the absolute value of the elasticity should be; the load following characteristic factor corresponds to the subject endowment reward; the better the following characteristics of the load node, the more positive its historical behavior, and the more elastic reward should be given to strengthen positive incentives; the multiplication of the two parts constitutes a joint adjustment structure, so that the elasticity coefficient has differentiated response capabilities at both the system level affected by the deep-adjustment collaborative incentive factor and the node level affected by the load following characteristic factor.

[0057] Based on the real-time load resilience coefficient set and the real-time excitation data of the load nodes, demand response adjustment analysis is performed to obtain load demand response adjustment data; wherein, the load demand response adjustment data includes the load demand response adjustment amount of each load node, which can be expressed as: In the formula, For scheduling time Load Node The baseline load forecast is derived from the load node demand forecast data in the day-ahead system forecast data. For scheduling time Load Node The load demand response adjustment amount.

[0058] In this embodiment, the large-area deep adjustment on the source side in the spatiotemporal coordinated regulation of source and load increases the capacity-weighted average of the real-time carbon emission correction factor. After nonlinear mapping, the deep adjustment coordinated excitation factor is amplified sharply. The source-side peak-shaving dilemma is transmitted to the load side in real time through the real-time elastic coefficient, driving the transferable load to gather on a large scale during the midday low-carbon and low-price period. This effectively avoids the risk of the source-side deep adjustment decision and the load-side demand response being disconnected under the existing conventional scheduling paradigm, resulting in a sharp drop in net load during the midday peak renewable energy generation period, causing a severe trough, triggering large-scale curtailment of solar power, and forcing the units to be in extreme oil injection peak-shaving conditions for a long time.

[0059] Figure 3 This diagram illustrates a comparison of the system net load curves before and after implementing the source-load spatiotemporal coordinated control method provided by this invention. Net load is defined as the remaining load after subtracting the output of new energy sources from the total system load. The original net load curve is the predicted net load curve before implementing demand response and source-load coordinated control. The net load curve of this invention is the actual operating net load curve after implementing the source-load spatiotemporal coordinated control method. Figure 3 As shown, the original net load curve exhibited a significant trough during periods of high renewable energy generation, with the trough bottom penetrating the lower limit of deep fuel injection regulation. The system required units to enter deep fuel injection regulation mode to maintain power balance, and the load was high during the evening peak period, resulting in a significant peak-to-trough difference. After adopting the method of this invention, the deep fuel injection regulation status on the source side is transmitted to the load side in real time through a collaborative incentive mechanism. Dynamic demand response guides some load to shift from the evening peak to periods of high renewable energy generation, causing the net load trough bottom to rise back to the non-deep fuel injection regulation range and avoiding triggering deep fuel injection regulation. Simultaneously, the evening peak load is reduced, the system peak-to-trough difference is significantly narrowed, and the operating curve becomes smoother. This result verifies that the system's deep fuel injection pressure can be reduced and the renewable energy absorption capacity improved through bidirectional collaborative regulation of source and load.

[0060] Based on the load demand response adjustment data, the load-side demand response adjustment cost is obtained; wherein, the load-side demand response adjustment cost can be expressed as: In the formula, For load nodes Adjustment cost per unit of demand response; The duration of the scheduling period; Adjust costs to meet load-side demand.

[0061] The load-side electricity purchase cost in this embodiment can be expressed as: In the formula, For scheduling time Time-of-use electricity pricing; The cost of purchasing electricity for the load side.

[0062] By summing up the load-side carbon responsibility allocation cost, load-side electricity purchase cost, and load-side demand response adjustment cost obtained above, the load-side low-carbon dispatch cost can be obtained, which can be expressed as: In the formula, Cost sharing for carbon responsibility on the Dutch side; To reduce the cost of low-carbon dispatching on the load side.

[0063] The total carbon emissions of the system in this embodiment can be calculated based on the carbon emission intensity of the thermal power unit and the real-time output data of the thermal power unit, and can be expressed as: In the formula, This represents the total carbon emissions of the system.

[0064] The renewable energy curtailment rate in this embodiment can be calculated based on the renewable energy output and the maximum available renewable energy output, and can be expressed as: In the formula, For scheduling time The maximum available power output of new energy sources; For scheduling time The new energy source is generating power in real time.

[0065] The multi-objective low-carbon optimization scheduling model is constructed based on a multi-objective optimization function and preset constraints, which are derived from the source-side low-carbon scheduling cost, the load-side low-carbon scheduling cost, the total carbon emissions of the system, and the renewable energy curtailment rate. The preset constraints include: 1) Power balance constraints: In the formula, For scheduling time The system's externally purchased power; For scheduling time The system network power loss.

[0066] 2) Unit operating constraints: In the formula, For scheduling time thermal power units The start / stop status; and thermal power units The minimum and maximum technical output are obtained from the thermal power unit output limit data in the system's basic operating parameters. and thermal power units The downward ramp rate limit and the upward ramp rate limit are obtained from the ramp power limit data of thermal power units in the system's basic operating parameters.

[0067] 3) Energy storage state of charge constraints: In the formula, For scheduling time The state of charge of an energy storage system; and These are the lower and upper limits of the allowable state of charge, respectively, obtained from the energy storage state of charge limit data in the system's basic operating parameters; and Scheduling time The energy storage charging power and energy storage discharging power; , and These are the charging efficiency, discharging efficiency, and rated energy capacity of the energy storage system, respectively, and can also be obtained from the energy storage state of charge limit data in the system's basic operating parameters.

[0068] 4) Load demand response constraints: , In the formula, , They are the load nodes. The lower and upper limits of the single-period demand response load adjustment amount can be obtained from the demand response adjustment amount limit data in the system's basic operating parameters. In response to the satisfaction index; The minimum satisfaction threshold is set by the system operator based on user participation intentions and obtained through the system's basic operating parameters.

[0069] 5) Constraints on new energy output: In the formula, For scheduling time The maximum available power output of new energy sources can be determined through the system's basic operating parameters.

[0070] The above methods and steps yield a multi-objective optimization model that takes the entire power dispatching system as the unified research object, with the optimization objectives of minimizing low-carbon dispatching costs on the source side, low-carbon dispatching costs on the load side, total carbon emissions of the system, and curtailment rate of new energy sources. The decision variables are the active power output sequence of thermal power units, the start-up state sequence of thermal power units, the charging and discharging sequence of energy storage systems, and the corresponding adjustment sequence of load demand at load nodes. In principle, the multi-objective low-carbon optimization dispatching model constructed in this embodiment can be obtained using any existing multi-objective optimization algorithm. However, to avoid premature convergence, uneven solution set distribution, and low solution efficiency in existing multi-objective optimization algorithms when dealing with high-dimensional nonlinear constraints in power systems, this embodiment preferably introduces an improved multi-objective sparrow search algorithm based on the standard sparrow search algorithm. This improved algorithm includes using Tent chaotic mapping to replace random initialization to generate the initial population, introducing a dynamically adaptive inertial weight that decreases linearly with the number of iterations in the discoverer position update formula, performing Cauchy mutation perturbation on continuously stagnant individuals, and employing an improved strategy of maintaining the Pareto non-dominated solution set using an external archive set based on the hypercube grid method. This improved algorithm is then used to solve the above multi-objective low-carbon optimization dispatching model.

[0071] Specifically, the step of solving the pre-constructed multi-objective low-carbon optimization scheduling model to obtain the target scheduling strategy includes: An initial population is generated based on tent mapping, and its fitness is evaluated according to the objective optimization function in the multi-objective low-carbon optimization scheduling model to obtain an initial fitness set. In practical applications, after initializing parameters such as population size, maximum number of iterations, decision variable dimensions, discoverer ratio, vigilant ratio, follower ratio, safety threshold, number of grids, Cauchy mutation trigger stagnation generation threshold, comprehensive fitness improvement threshold, and dynamic weight upper and lower limits, the initial population is generated using Tent chaotic mapping according to the following formula to improve the ergodicity and diversity of the population and overcome the problems of poor initial solution quality and uneven distribution caused by random initialization. In the formula, For the bifurcation parameter, take ; For the first The chaotic sequence values ​​of the next iteration; initial value exist Random selection within, and To ensure the chaotic properties of the sequence; map the Tent chaotic sequence to the first... The feasible region of each decision variable The formula is: In the formula, For the first The individual The corresponding Tent chaotic sequence value; , The first Lower and upper bounds of each decision variable; This represents the initial position of the individual.

[0072] Each individual in the initial population is substituted into one of the four objective functions in the multi-objective low-carbon optimization scheduling model for fitness evaluation. The obtained four objective function values The corresponding fitness values ​​are then aggregated to obtain the required initial fitness set.

[0073] Based on the initial fitness set, non-dominated solutions are extracted from the initial population to generate an initial external archive set; wherein the method for generating the initial external archive set can be implemented with reference to relevant existing technologies, and will not be described in detail here.

[0074] The initial population is updated according to the update mechanism of discoverers, followers, and watchdogs to obtain the current iteration population; in the update mechanism, the position of the discoverer is updated using a dynamic adaptive inertia weight that decreases linearly based on the number of iterations; in practical applications, the dynamic adaptive inertia weight can be expressed as: In the formula, This represents the current iteration number; This represents the maximum number of iterations. , These are the upper and lower limits of the inertia weight, respectively; For the first Dynamic adaptive inertia weights for each iteration.

[0075] The discoverer position update formula corresponding to the dynamic adaptive inertia weight is: In the formula, and The first The second iteration and the first In the nth iteration Among the individuals, the first The position of each decision variable; It is a random number; This is a warning value; This is a safety threshold; These are random numbers that follow a standard normal distribution. It is a row vector consisting entirely of 1s.

[0076] Subsequently, the follower positions are updated according to the following rules: In the formula, For the first The current global worst position in the next iteration is selected from the initial external archive set; The location of the optimal discoverer; Population size; Randomly select elements or The row vector, .

[0077] When a vigilant person senses danger, they perform a jump escape maneuver: In the formula, For the first The current global optimal position in the next iteration is randomly selected from the initial external archive set based on a grid density roulette strategy. It follows a normal distribution with a mean of 0 and a variance of 1; It is a random number; To prevent extremely small positive numbers with a denominator of zero; For individuals The overall fitness value is obtained by weighted summation of the four objective function values ​​corresponding to the individual after each individual's extreme value normalization (subtracting the minimum objective function value corresponding to all individuals of the objective from the individual's objective function value, and then dividing by the difference between the maximum and minimum objective function values ​​among all individuals of the objective). and These represent the maximum and minimum overall fitness values ​​of the current population, respectively.

[0078] Add Cauchy mutation perturbation to continuously stagnant individuals in the current iterative population, perform non-dominated sorting on the union of the resulting new iterative population and the initial external archive set, and update and maintain the resulting current external archive set based on the hypercube grid method; wherein, continuously stagnant individuals can be understood as individuals whose overall fitness improvement is below a threshold, and Cauchy mutation perturbation is added to them in the following manner: In the formula: Individual before mutation The position vector; For a standard Cauchy distribution random number, its probability density function is: ; This represents element-wise multiplication; For the mutated individual The position vector, by utilizing the thick-tailed characteristic of the Cauchy distribution, can generate a large perturbation with a high probability, enabling the algorithm to escape local optima traps with a larger step size.

[0079] After obtaining an iterative new population by adding Cauchy mutation perturbation to continuously stagnant individuals and an initial external archive set, and performing non-dominated sorting to obtain the external archive set, the external archive set is updated and maintained based on the hypercube grid method. That is, after each iteration, the dominance relationship between the non-dominated solution candidates in the current population and the existing members in the external archive set is compared. New non-dominated solutions are added to the external archive set, while old members dominated by new solutions are removed. When the number of solutions in the external archive set exceeds the preset capacity, the individual with the highest grid density is deleted first to ensure the uniformity of the solution set distribution, thus obtaining the required current external archive set.

[0080] The algorithm determines whether the current iteration count has reached the maximum iteration count. If so, it obtains the target scheduling strategy based on the current external archive set. Otherwise, it continues iteratively searching based on the current external archive set until the maximum iteration count is reached, thus obtaining the target scheduling strategy. In practical applications, if the termination condition is met, all non-dominated solutions in the current external archive set are output as the Pareto optimal solution set. Otherwise, it returns to the discoverer update step and continues iterating until the termination condition is met, obtaining the Pareto optimal solution set. Finally, based on actual scheduling requirements, the non-dominated solutions in the Pareto optimal solution set are comprehensively evaluated and ranked using the entropy-weighted TOPSIS method. The optimal compromise solution that simultaneously considers the four optimization objectives is selected as the target scheduling strategy.

[0081] Furthermore, to verify the effectiveness of the improved multi-objective sparrow search algorithm proposed in this embodiment, a simulation comparison was performed on a standard IEEE 30-node system with the standard multi-objective sparrow search algorithm. The scheduling cycle was 24 hours, and the time resolution was 1 hour. This embodiment optimizes the solution after initializing the parameters of the improved multi-objective sparrow search algorithm according to the following principles: the population size is set to a range of values. The preferred value is 40; the maximum number of iterations can be set within a certain range. The preferred value is 300; the proportion of discoverers (20%), the proportion of vigilants (10%), the proportion of followers (70%), and the safety threshold range. The range of values ​​for the external archive set capacity is: The preferred value is 100; the range of the number of grid cells is... The preferred value is 10; the range of the Cauchy mutation triggering stagnation algebra threshold is... The preferred value is 8; the range of the overall fitness improvement threshold is... The preferred value is The upper and lower limits of dynamic weights are: , After selecting the target scheduling strategy from the Pareto optimal solution set using the entropy-weighted TOPSIS method, the values ​​of each objective function are compared as follows: In the solution results of the improved multi-objective sparrow search algorithm proposed in this embodiment, the source-side low-carbon scheduling cost, load-side low-carbon scheduling cost, total system carbon emissions, and renewable energy curtailment rate are 653,000 yuan, 1,444,000 yuan, 4,152.11 tons, and 18.458%, respectively; compared with the solution results of the standard multi-objective sparrow search algorithm, the source-side low-carbon scheduling cost, load-side low-carbon scheduling cost, total system carbon emissions, and renewable energy curtailment rate are 656,000 yuan, 1,442,000 yuan, 4,160.20 tons, and 20.342%, respectively. The results show that, at the objective scheduling strategy level, the improved multi-objective sparrow search algorithm reduces the total carbon emissions of the system by approximately 8.09 tons and the renewable energy curtailment rate by approximately 1.884 percentage points. The source-side low-carbon scheduling cost is slightly reduced, while the load-side low-carbon scheduling cost is slightly increased, but the increase is small, demonstrating a better multi-objective trade-off capability. Overall, this implementation effectively improves the algorithm's global search capability and ability to escape local optima by introducing chaotic initialization, dynamic weighting mechanism, and Cauchy mutation strategy, thereby obtaining a higher quality and more uniformly distributed Pareto optimal solution set in the multi-objective optimization process.

[0082] This invention provides day-ahead system forecast data for acquiring new energy node output forecast data, thermal power unit node output forecast data, load node demand forecast data, and load node transferable load forecast data for the target scheduling period; deep peak-shaving interval parameters, including non-oil-operated peak-shaving intervals and oil-operated peak-shaving intervals; and basic system operating parameters. Based on the day-ahead system forecast data, deep peak-shaving interval parameters, and basic system operating parameters, it pre-optimizes the system with the goals of minimizing source-side low-carbon scheduling costs, minimizing load-side low-carbon scheduling costs, minimizing total system carbon emissions, and minimizing new energy curtailment rates. This is based on consideration of load-side following characteristics and source-side... The multi-objective low-carbon optimization scheduling model, which is based on the source-load two-way carbon responsibility sharing mechanism that allocates carbon emission responsibility between thermal power units and load nodes according to the contribution of peak shaving depth on the load side, is used to solve the technical solution of the target scheduling strategy. This two-way carbon responsibility sharing mechanism, which considers the load-side following characteristics and the contribution of peak shaving depth on the source side, can not only accurately identify the load time shift caused by demand response and accurately assess the emission reduction contribution of the load side, but also eliminate the penalty effect of deep scheduling by the unit, incentivize the unit to actively perform deep scheduling, improve the renewable energy consumption rate, and reduce the overall carbon emissions and operating costs of the system.

[0083] It should be noted that although the steps in the flowchart above are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order requirement for the execution of these steps, and they can be executed in other orders.

[0084] In one embodiment, such as Figure 4 As shown, a low-carbon optimized dispatching system for a power system is provided, the system comprising: Data acquisition module 1 is used to acquire day-ahead system forecast data, deep peak shaving interval parameters, and basic system operating parameters for the target scheduling cycle; the day-ahead system forecast data includes new energy node output forecast data, thermal power unit node output forecast data, load node demand forecast data, and load node transferable load forecast data; the deep peak shaving interval parameters include non-oil peak shaving intervals and oil peak shaving intervals. The low-carbon optimization scheduling module 2 is used to solve the pre-constructed multi-objective low-carbon optimization scheduling model based on the day-ahead system forecast data, the deep peak-shaving interval parameters, and the system's basic operating parameters to obtain the target scheduling strategy. The target low-carbon optimization scheduling model takes the minimum source-side low-carbon scheduling cost, the minimum load-side low-carbon scheduling cost, the minimum total system carbon emissions, and the minimum renewable energy curtailment rate as optimization objectives, and is constructed based on a source-load two-way carbon responsibility sharing mechanism. The source-load two-way carbon responsibility sharing mechanism is to share carbon emission responsibilities between thermal power units and load nodes, taking into account the load-side following characteristics and the source-side peak-shaving depth contribution.

[0085] Specific limitations regarding the low-carbon optimized dispatch system for power systems can be found in the limitations of the low-carbon optimized dispatch method for power systems described above; the corresponding technical effects are equivalent and will not be repeated here. Each module in the aforementioned low-carbon optimized dispatch system for power systems 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 corresponding operations of each module.

[0086] Figure 5 An internal structural diagram of a computer device is shown in one embodiment. This computer device may specifically be a terminal or a server. Figure 5As shown, the computer device includes a processor, memory, network interface, display, camera, and input device connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it can implement a low-carbon optimized scheduling method for the power system. The display screen can be an LCD screen or an e-ink screen. The input device can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0087] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the computer device to which the present invention is applied. Specific computing devices may include more or fewer components than those shown in the figure, or combine certain components, or have the same component arrangement.

[0088] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method described above.

[0089] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0090] In summary, the low-carbon optimized scheduling method, system, computer equipment, and storage medium provided by the embodiments of the present invention can achieve a two-way carbon responsibility sharing mechanism for thermal power units and load nodes based on considering load-side following characteristics and source-side peak-shaving depth contributions. This mechanism can not only accurately identify load time shifts caused by demand response and accurately assess load-side emission reduction contributions, but also eliminate the penalty effect of deep scheduling by units, incentivize units to actively perform deep scheduling, improve the renewable energy absorption rate, and reduce the overall carbon emissions and operating costs of the system.

[0091] The various embodiments in this specification are described in a progressive manner. For directly identical or similar parts of the embodiments, refer to each other. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. It should be noted that the technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0092] The embodiments described above are merely preferred embodiments of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various improvements and substitutions without departing from the principles of the present invention, and these improvements and substitutions should also be considered within the scope of protection of the present invention. Therefore, the scope of protection of this invention should be determined by the scope of the claims.

Claims

1. A method for low-carbon optimal dispatching of a power system, characterized in that, The method includes: The system acquires day-ahead system forecast data, deep peak-shaving interval parameters, and basic system operating parameters for the target scheduling cycle. The day-ahead system forecast data includes new energy node output forecast data, thermal power unit node output forecast data, load node demand forecast data, and load node transferable load forecast data. The deep peak-shaving interval parameters include peak-shaving intervals without oil injection and peak-shaving intervals with oil injection. Based on the day-ahead system forecast data, the deep peak-shaving interval parameters, and the system's basic operating parameters, a pre-constructed multi-objective low-carbon optimization scheduling model is solved to obtain the target scheduling strategy. The target low-carbon optimization scheduling model takes minimizing the source-side low-carbon scheduling cost, the load-side low-carbon scheduling cost, the total system carbon emissions, and the renewable energy curtailment rate as its optimization objectives, and is constructed based on a source-load two-way carbon responsibility sharing mechanism. The source-load two-way carbon responsibility sharing mechanism considers the load-side following characteristics and the contribution of the source-side peak-shaving depth to allocate carbon emission responsibilities to thermal power units and load nodes. The construction steps of the multi-objective low-carbon optimization scheduling model include: Based on the real-time output data of thermal power units and the parameters of the deep peak shaving interval, carbon emission penalty exemption analysis is performed based on the contribution of source-side peak shaving depth to obtain the set of real-time carbon emission correction factors for the units. Based on the historical load demand data of the load nodes at the scheduling time and the corresponding historical total output data of new energy sources, load following characteristics analysis is performed to obtain a set of real-time load following characteristics factors; Based on the load real-time following characteristic factor set and the unit real-time carbon emission correction factor set, a scheduling cycle carbon responsibility cost analysis is performed based on the tiered carbon trading mechanism to obtain the source-side carbon responsibility allocation cost and the load-side carbon responsibility allocation cost. The source-side low-carbon dispatch cost is obtained based on the source-side carbon responsibility allocation cost, the net cost of deep peak shaving of thermal power units, and the operating cost of thermal power units. The load-side low-carbon dispatch cost is obtained based on the load-side carbon responsibility sharing cost, the load-side electricity purchase cost, and the load-side demand response adjustment cost. The multi-objective low-carbon optimization scheduling model is constructed based on a multi-objective optimization function and preset constraints, which are derived from the source-side low-carbon scheduling cost, the load-side low-carbon scheduling cost, the total carbon emissions of the system, and the renewable energy curtailment rate. The preset constraints include power balance constraints, unit operation constraints, energy storage state of charge constraints, load demand response constraints, and renewable energy output constraints.

2. The low-carbon optimized dispatching method for power systems as described in claim 1, characterized in that, The load following characteristic factor set includes the real-time following characteristic factors of each load node. The step of performing load following characteristic analysis based on the historical load demand data of the load nodes at the scheduling time and the corresponding historical total output data of new energy sources to obtain the load real-time following characteristic factor set includes: The historical load demand data of each load node in the historical load demand data of the load node is dynamically time-warped and analyzed with the historical total output data of the new energy to obtain the following pattern distance of the corresponding load node. The historical load demand data of each load node in the historical load demand data of the load node is averaged with the historical total output data of the new energy source over the entire time period to obtain the comprehensive average value of the corresponding load node. The normalized morphological distance of the corresponding load node is obtained by multiplying the comprehensive average value of each load node with the total number of scheduling period periods. The load following characteristic factor set is calculated based on the normalized morphological distance of all load nodes and the preset attenuation rate adjustment coefficient, using a preset exponential attenuation function.

3. The low-carbon optimized dispatching method for power systems as described in claim 1, characterized in that, The step of performing carbon liability cost analysis for the scheduling cycle based on the real-time load following characteristic factor set and the real-time unit carbon emission correction factor set, and obtaining the source-side carbon liability allocation cost and the load-side carbon liability allocation cost, includes: Based on the complex power tracking theory, real-time carbon emission component data of the unit's load-affecting nodes and real-time carbon emission of the total network loss of the system are obtained. Carbon emission rating analysis is performed based on the real-time carbon emission correction factor set of the unit and the real-time output data of the thermal power unit to obtain the carbon emission rating coefficient set of the unit. By fusing and analyzing the set of carbon emission rating coefficients for the generating units and the set of load following characteristic factors, a set of bidirectional source-load allocation coefficients is obtained. Carbon emission responsibility allocation is calculated based on the source-load bidirectional allocation coefficient set, the real-time carbon emission component data of the unit's load-affected nodes, and the real-time carbon emission of the total network loss of the system, to obtain the unit carbon emission responsibility allocation data and the load node carbon emission responsibility allocation data. Based on the carbon emission responsibility allocation data of the generating units and the carbon emission responsibility allocation data of the load nodes, a carbon responsibility cost analysis is performed based on the tiered carbon trading mechanism to obtain the source-side carbon responsibility allocation cost and the load-side carbon responsibility allocation cost.

4. The low-carbon optimized dispatching method for power systems as described in claim 3, characterized in that, The step of performing carbon emission rating analysis based on the real-time carbon emission correction factor set of the unit and the real-time output data of the thermal power unit to obtain the carbon emission rating coefficient set of the unit includes: Based on the real-time output data of the thermal power unit, the carbon content of coal, the carbon oxidation rate and the carbon capture rate, the carbon emission intensity data of the thermal power unit is calculated based on the secondary coal consumption curve. Based on the real-time carbon emission correction factor set of the unit, the carbon emission intensity of the thermal power unit is corrected to obtain the corrected carbon emission intensity data of the thermal power unit. Based on the corrected carbon emission intensity data of the thermal power unit and the real-time output data of the thermal power unit, the equivalent carbon emission intensity of the unit's cycle is analyzed to obtain the equivalent carbon emission intensity data of the thermal power unit. The carbon emission intensity of the thermal power unit is assessed based on the equivalent carbon emission intensity data and the preset graded carbon emission intensity limit parameters to obtain the set of carbon emission rating coefficients for the unit.

5. The low-carbon optimized dispatching method for power systems as described in claim 2, characterized in that, The steps for obtaining the load-side demand response adjustment cost include: Based on the complex power tracking theory, the total carbon emissions data of the load nodes are obtained, and the real-time carbon potential data of the load nodes is obtained based on the total carbon emissions data of the load nodes and the real-time load data of the load nodes. Based on the capacity-weighted average of the real-time carbon emission correction factors of each unit in the real-time carbon emission correction factor set, the deep adjustment collaborative incentive factor is obtained, and based on the deep adjustment collaborative incentive factor, the time-of-use electricity price and the real-time carbon potential data of the load node, the real-time incentive data of the load node is obtained. Based on the deep-tuning collaborative excitation factor and the load following characteristic factor set, elastic demand response analysis is performed to obtain the real-time load elasticity coefficient set; Based on the real-time load elasticity coefficient set and the real-time load node excitation data, demand response adjustment analysis is performed to obtain load demand response adjustment data. The load-side demand response adjustment cost is obtained based on the load demand response adjustment data.

6. The low-carbon optimized dispatching method for power systems as described in claim 1, characterized in that, The step of solving the pre-constructed multi-objective low-carbon optimization scheduling model to obtain the target scheduling strategy includes: An initial population is generated based on tent mapping, and the fitness of the initial population is evaluated according to the objective optimization function in the multi-objective low-carbon optimization scheduling model to obtain an initial fitness set. Based on the initial fitness set, non-dominated solutions are extracted from the initial population to generate an initial external archive set; The initial population is updated according to the update mechanism of discoverers, followers, and vigilants to obtain the current iterative population; in the update mechanism, the position of the discoverer is updated using a dynamic adaptive inertia weight that decreases linearly based on the number of iterations. Add Cauchy mutation perturbation to the continuously stagnant individuals in the current iterative population, perform non-dominated sorting on the union of the resulting new iterative population and the initial external archive set, and update and maintain the resulting current external archive set based on the hypercube grid method. Determine whether the current iteration count has reached the maximum iteration count. If so, obtain the target scheduling strategy based on the current external archive set. Otherwise, continue iterative search based on the current external archive set until the maximum iteration count is reached, and obtain the target scheduling strategy.

7. A low-carbon optimized dispatching system for power systems, characterized in that, The power system low-carbon optimization dispatching method as described in claim 1, wherein the system comprises: The data acquisition module is used to acquire day-ahead system forecast data, deep peak shaving interval parameters, and basic system operating parameters for the target scheduling cycle. The day-ahead system forecast data includes new energy node output forecast data, thermal power unit node output forecast data, load node demand forecast data, and load node transferable load forecast data. The deep peak shaving interval parameters include non-oil peak shaving intervals and oil peak shaving intervals. The low-carbon optimization scheduling module is used to solve a pre-constructed multi-objective low-carbon optimization scheduling model based on the day-ahead system forecast data, the deep peak-shaving interval parameters, and the system's basic operating parameters to obtain the target scheduling strategy. The target low-carbon optimization scheduling model aims to minimize the source-side low-carbon scheduling cost, the load-side low-carbon scheduling cost, the total system carbon emissions, and the renewable energy curtailment rate. It is constructed based on a source-load two-way carbon responsibility sharing mechanism. The source-load two-way carbon responsibility sharing mechanism allocates carbon emission responsibilities to thermal power units and load nodes by considering the load-side following characteristics and the source-side peak-shaving depth contribution.

8. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, 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 6.

9. 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 6.

Images