Low-carbon economic dispatching method and system fusing ladder carbon trading based on carbon-energy-cost coupling coefficient matrix
By constructing a carbon-energy-cost coupling coefficient matrix and a time-sharing dynamic carbon emission factor, and combining flexible demand response and a two-stage robust optimization model, the problems of low carbon emission accounting accuracy and wind and solar power output fluctuations in the energy system of industrial parks were solved, thus achieving optimization and robustness of low-carbon economic operation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HUBEI FUBIAN SPACETIME ENERGY TECH CO LTD
- Filing Date
- 2026-04-28
- Publication Date
- 2026-06-12
AI Technical Summary
Traditional industrial park energy systems suffer from low accuracy in carbon emission accounting, a disconnect between carbon trading mechanisms and dispatch, and weak ability to cope with uncertainties, resulting in high electricity purchase costs or difficulty in meeting carbon emission standards. They also lack the ability to optimize multiple objectives and cope with fluctuations in wind and solar power output.
A low-carbon economic dispatch method that integrates carbon energy cost coupling coefficient matrix with tiered carbon trading is proposed. By constructing a three-dimensional coupling coefficient matrix and time-sharing dynamic carbon emission factors, combined with flexible demand response and a two-stage robust optimization model, the method achieves coordinated optimization of carbon emissions, energy utilization and economic costs.
It significantly reduced electricity purchase costs, improved carbon emission compliance rates, enhanced the system's robustness to uncertainties in wind and solar power output, and achieved the optimal balance for low-carbon economic operation.
Smart Images
Figure CN122198543A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of integrated energy system operation optimization technology, and in particular to a low-carbon economic dispatch method and system for cascade carbon trading based on the fusion of carbon energy cost coupling coefficient matrix. Background Technology
[0002] Driven by the "dual carbon" goals, industrial parks, as carriers of highly concentrated carbon emissions, face multiple challenges in the low-carbon economic operation of their integrated energy systems. Traditional industrial park energy system planning and scheduling methods have the following significant shortcomings: (1) Insufficient multi-objective optimization capability: Single-objective optimization or simple linear weighting is often used, which cannot explicitly depict the deep coupling and constraint relationship between "carbon energy cost". This can easily lead to high electricity purchase cost or difficulty in meeting carbon emission standards in actual operation, and cannot achieve joint optimization of carbon quota, tiered carbon trading cost and equipment capacity.
[0003] (2) Low accuracy of carbon emission accounting: Static carbon emission factors are generally used, ignoring the time-varying characteristics of the carbon emission intensity of power grid supply as wind and solar power output and load change in real time, resulting in a disconnect between carbon emission accounting and real-time scheduling.
[0004] (3) The carbon trading mechanism is disconnected from the scheduling: the dynamic allocation of carbon quotas and the tiered carbon trading costs are often used as ex-post evaluation indicators, failing to be effectively embedded in the scheduling objective function in the form of endogenous penalty terms, and thus failing to incentivize the system to actively reduce carbon emissions.
[0005] (4) Weak ability to cope with uncertainty: Traditional deterministic scheduling algorithms are difficult to handle the strong randomness of wind and solar power output and load fluctuations. They lack a refined and robust scheduling model specifically for pure wind, solar and energy storage clean energy scenarios, which can easily lead to real-time power mismatch risk.
[0006] Existing research has not yet combined an explicit carbon energy cost coupling architecture with a tiered carbon trading internalization mechanism to achieve the dual objectives of optimal comprehensive evaluation at the planning level and low-carbon robust response at the operation level for industrial parks. It has also not fully considered the dynamic linkage between dynamic carbon factors, time-of-use electricity prices, and energy storage charging and discharging behavior. Summary of the Invention
[0007] The purpose of this invention is to propose a low-carbon economic scheduling method and system for tiered carbon trading based on the fusion of carbon energy cost coupling coefficient matrix. Through a two-layer collaborative architecture of capacity configuration with explicit coupling of carbon energy cost and internalization of carbon costs at the operation layer, the optimal balance between low-carbon and economic operation of industrial parks can be achieved, while improving the system's robustness in the face of uncertainties in wind and solar power output.
[0008] To achieve the above objectives, this invention provides a low-carbon economic dispatch method for tiered carbon trading based on the fusion of carbon energy cost coupling coefficient matrices, comprising the following steps: Therefore, the low-carbon economic dispatch method and system of the carbon energy cost coupling coefficient matrix fusion tiered carbon trading described above have the following advantages: (1) The first three-dimensional carbon-energy-cost coupling evaluation mechanism is introduced: a three-dimensional coupling coefficient matrix based on Pearson correlation coefficient is introduced to conduct a quadratic scalar comprehensive evaluation of multiple objectives, which explicitly describes the synergistic and restrictive relationship between the three objectives of carbon emissions, energy utilization and economic cost, overcomes the limitations of the traditional linear weighting method, and significantly increases the proportion of renewable energy. (2) Achieve endogenous and dynamic carbon cost accounting: Use time-sharing dynamic carbon emission factors to replace static factors, and embed the tiered carbon trading cost as an endogenous progressive penalty term directly into the scheduling objective function. Combined with flexible demand response, the system can spontaneously reduce the amount of electricity purchased during high-carbon periods while pursuing economic optimization. (3) Construct a robust two-stage scheduling framework: In response to the random fluctuations in wind and solar power output, a two-stage robust optimization model based on budget uncertainty set is constructed. The model is solved efficiently by the C&CG algorithm, which effectively solves the risk of real-time power mismatch and achieves a high degree of unity between economy, low carbon emissions and operational robustness.
[0009] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description
[0010] Figure 1 This is a schematic diagram of the overall process of the low-carbon economic dispatch method of carbon energy cost coupling coefficient matrix fusion tiered carbon trading in this embodiment of the invention. Figure 2 This is a structural diagram of the integrated energy system of the industrial park in an embodiment of the present invention; Figure 3 This is a flowchart illustrating the reduction process for the K-means clustering scenario in an embodiment of the present invention. Figure 4 This is a framework diagram of a carbon energy cost coupled multi-objective capacity configuration model in an embodiment of the present invention; Figure 5 This is a flowchart illustrating the solution process of the Improved Multi-Objective Particle Swarm Optimization (IMOPSO) algorithm in this embodiment of the invention. Figure 6 This is a time-sharing dynamic carbon emission factor curve diagram in an embodiment of the present invention; Figure 7 This is a framework diagram of the low-carbon economic optimization scheduling model in an embodiment of the present invention; Figure 8 This is a flowchart illustrating the alternating solution of the principal and subproblems in the column constraint generation algorithm (C&CG) in this embodiment of the invention. Figure 9 This is a sensitivity analysis curve of photovoltaic uncertainty budget parameters in an embodiment of the present invention; Figure 10This is a sensitivity analysis curve of the uncertainty budget parameters of the wind turbine in an embodiment of the present invention; Figure 11 This is a graph showing the electrical load curves before and after considering demand response on a typical day in each of the four seasons in this embodiment of the invention. Figure 12 This is a typical daily power balance diagram for the four seasons before and after demand response in an embodiment of the present invention. Detailed Implementation
[0011] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.
[0012] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.
[0013] Example like Figure 1-12 As shown in the figure, this embodiment proposes a low-carbon economic dispatch method for tiered carbon trading based on the fusion of carbon energy cost coupling coefficient matrix. The complete process is divided into two levels: capacity optimization configuration at the planning level and day-ahead scheduling optimization at the operation level. The specific implementation steps are as follows: Step S1: Construct a set of typical daytime scenes for each of the four seasons: S11. Perform hourly probability density fitting on historical wind speed data using a Weibull distribution to obtain shape parameters for each time period. and scale parameters The wind speed simulation sequence was generated using Monte Carlo random sampling, and the Weibull distribution probability density function is: ; In the formula, This is the real-time wind speed.
[0014] S12. The wind power output simulation scenario is obtained by mapping the piecewise linear power curve of the wind turbine. The theoretical output power model of the wind turbine is as follows: ; In the formula, This refers to the rated power of the fan. To cut into wind speed, This is the rated wind speed.
[0015] S13. The median absolute deviation (MAD) method is used to detect and correct anomalies in historical photovoltaic power output data. The photovoltaic theoretical output power model with extended temperature correction is as follows: ; In the formula, This refers to the rated installed capacity of photovoltaic power. For real-time irradiance, Irradiance under standard test conditions. The overall efficiency coefficient of photovoltaic power. The photovoltaic temperature coefficient, Real-time temperature of photovoltaic modules. This is the standard test temperature.
[0016] S14. Slice the cleaned wind power, photovoltaic and load data by day, perform K-means clustering in spring, summer, autumn and winter respectively, and reduce them by probability weighting to single representative typical daily curves of each season to form a typical daily scene set of the four seasons.
[0017] S2, Carbon-energy-cost coupled multi-objective capacity optimization configuration: S21. Construct a multidimensional optimization objective function: (1) Minimize overall operating costs: ; In the formula, This represents the annualized cost of the total equipment investment. Annual fixed operation and maintenance costs, This represents the weighted average operating cost per typical day.
[0018] (2) Minimize carbon emission costs: ; ; In the formula, This represents the park's total annual carbon emissions. The annual average carbon emission factor of the power grid. For the annual electricity purchased from the grid in the park, The carbon emission factor for diesel power generation. This represents the annual diesel power generation capacity of the industrial park. This is the carbon emission cost price coefficient.
[0019] (3) To achieve the highest energy utilization rate, reverse normalization is used to transform the search for the maximum value into the search for the minimum value: ; In the formula, , They are respectively Actual electricity consumption of photovoltaic and wind power during the period , They are respectively Available power generation from photovoltaic and wind power during the specified time period.
[0020] S22. Construct a three-dimensional carbon energy-feed coupling coefficient matrix. With quadratic comprehensive evaluation scalar : matrix diagonal elements The self-weights of each objective, and the off-diagonal elements. For the goal With the goal The coupling coefficient between them is constructed using the following steps: The first step is to randomly sample the decision variables to generate simulation results and calculate the Pearson correlation coefficients between each pair of the three normalized objective values. ; The second step is to combine the subjective preference weight vector. Calculate diagonal elements Off-diagonal elements: ; The third step is to verify and correct the symmetry and positive definiteness of the obtained matrix to ensure the comprehensive evaluation scalar. It is a convex function over the entire domain.
[0021] Based on the coupling coefficient matrix Normalize the column vector of the three objective ranges. Mapped to a quadratic comprehensive evaluation scalar ,in , , These are the normalized values of comprehensive operating costs, carbon emission costs, and energy utilization efficiency, respectively. The expression is: ; In the formula, the diagonal terms Reflecting the contribution of a single objective to the overall evaluation, cross-items. Explicitly characterizing the coupling effect between two targets: when In collaborative relationships, the solution that achieves better results for both objectives receives an additional reward. The value decreases; when When there is a constraint relationship, a solution with both objectives being worse than the other is subject to double penalties. The value increased; The smaller the value, the better the overall performance of the solution.
[0022] S23. The improved multi-objective particle swarm optimization (IMOPSO) algorithm is used to solve for the Pareto optimal solution set. The algorithm improvement mechanism includes: using an inertial weight that adaptively and linearly decreases with the number of iterations. and collaboratively dynamically adjusted individual learning factors With social learning factors The expression is: ; ; ; In the formula, This represents the current iteration number. The maximum number of iterations, , These represent the maximum and minimum values of the inertia weight, respectively. , These represent the maximum and minimum values of the individual learning factor, respectively. , These represent the maximum and minimum values of the social learning factor, respectively.
[0023] S24. Introduce a Pareto non-dominated solution archive maintenance mechanism based on congestion distance weighting to remove solutions with excessive congestion and maintain the diversity of the solution set; S25. In the global guiding particle selection stage, a weighted roulette wheel betting strategy based on crowding distance is adopted to balance the global search capability and local convergence speed of the algorithm. S26. Combining the TOPSIS method and the weighted rank ratio method, the globally optimal device capacity configuration is selected from the Pareto optimal solution set.
[0024] S3. Constructing carbon trading and demand response models: S31. Constructing a carbon trading model for industrial parks: Determining initial carbon emission quotas for parks based on the baseline method. The actual carbon emissions are calculated using time-sharing dynamic carbon emission factors. The expression for the time-of-use dynamic carbon emission factor is: ; In the formula, for Carbon emission factors of power grid during a given period for Carbon emission correction factor for different time periods, during peak grid load periods (8:00-22:00). Low-load off-peak hours (22:00-8:00 the next day) or periods of high new energy output Normally .
[0025] The formula for calculating actual carbon emissions is: ; In the formula, for Power purchased by the park during the specified time period This is the scheduling time step.
[0026] when At that time, a progressive penalty settlement will be implemented according to the preset three-tier carbon price system, with the carbon price increasing by 50% for each higher tier; when At that time, the park sells its remaining carbon allowances to generate carbon emission reduction revenue.
[0027] S32. Construct a flexible user energy demand response model: (1) Transferable load model: The total energy of transferable load is conserved within a specified time window, and the compensation revenue during its scheduling cycle is... for: ; In the formula, The unit price for compensation of transferable load. for The transfer power of the load that can be transferred during a certain period.
[0028] (2) Load reduction model: its compensation revenue during the scheduling period for: ; In the formula, To reduce the unit compensation price per load, for User baseline load power during the time period for The percentage of load reduction during a given time period.
[0029] Total Revenue from Demand Response .
[0030] S4. Construct a two-stage robust optimization scheduling model: S41. Embedding the tiered carbon trading cost into the system's day-ahead scheduling objective function as an endogenously coupled progressive economic penalty term, the objective function for minimizing the system's overall daily operating cost is: ; In the formula, The cost of purchasing electricity under time-of-use pricing. Revenue from selling surplus electricity to the grid. The net cost of tiered carbon trading. For equipment operation and maintenance costs, To incur the cost of abandoning energy, This represents the total revenue from demand response.
[0031] S42. Establish core constraints: (1) Power balance constraint: ; In the formula, , They are respectively Power purchase and sale by the power grid during specific time periods , They are respectively Time-of-use energy storage charging and discharging power, for Total load after demand response period for Power curtailed during a given period.
[0032] (2) Dynamic constraints on the state of charge (SOC) of energy storage: ; In the formula, , These refer to energy storage charging and discharging efficiency, respectively. To achieve the rated capacity of energy storage, the State of Charge (SOC) must meet the following requirements. .
[0033] S43. To address the random fluctuations in wind and solar power output, construct a system based on budget uncertainty sets. Two-stage robust optimization scheduling model: ; In the formula, the first stage formulates a baseline scheduling scheme based on the predicted mean. The second phase focuses on actual efforts in the field of scenery. Once implemented, the energy storage charging and discharging, along with the output of the fast-response unit, will be adjusted. Interval absorption is implemented to eliminate prediction deviations and meet real-time power balance under the worst-case scenario.
[0034] S5 and C&CG algorithm solutions and scheduling strategy output: S51. Decompose the above three-level min-max-min problem into a main problem (MP) and subproblems (SP), where As the baseline scheduling variable for the first phase, The amount of energy contributed to wind and solar power is uncertain. For a budget uncertainty set, For the second stage of real-time adjustment variables; S52. Solve the main problem MP. Under the known worst-case scenario, obtain the optimal baseline solution and lower bound for the first stage. The expression for the main problem is: ; S53, Fix the optimal solution of the first stage Solve the subproblem SP in the uncertain set. The search above finds the most unfavorable deviation scenario and upper bound that maximizes the adjustment cost in the second stage. The subproblem expression is: ; In the formula, , , , This is the constraint coefficient matrix; S54. Determine whether the difference between the upper and lower bounds meets the preset convergence accuracy. If the conditions are met, the final low-carbon economic operation scheduling strategy is output; if not, the constraints corresponding to the newly identified worst-case scenario are added to the main problem, and the process returns to step S52 to continue iterating.
[0035] A low-carbon economic dispatch system based on the fusion of carbon energy cost coupling coefficient matrix and tiered carbon trading, used to implement a low-carbon economic dispatch method for tiered carbon trading based on the fusion of carbon energy cost coupling coefficient matrix, including: The data acquisition and scenario construction module is used to acquire historical data on industrial park load, wind power and photovoltaic output, complete data preprocessing and anomaly correction, and construct typical daily scenario sets for the four seasons. The carbon-energy-cost coupled capacity optimization module is used to construct a multi-objective optimization model of comprehensive operating cost, carbon emission cost and energy utilization rate, generate a three-dimensional carbon-energy-cost coupling coefficient matrix and map it into a comprehensive evaluation scalar, and use an improved multi-objective particle swarm algorithm to solve and select the globally optimal equipment capacity configuration. The carbon trading and demand response modeling module is used to determine the initial carbon emission quota of the park, generate time-sharing dynamic carbon emission factors, establish a tiered carbon trading cost model, and construct a flexible demand response model for transferable and reduceable loads. The robust optimization scheduling module is used to embed the tiered carbon trading cost as an endogenous penalty term into the day-ahead scheduling objective function, establish a unified daily comprehensive operating cost objective function and core constraints, and construct a two-stage robust optimization scheduling model based on the budget uncertainty set. The C&CG decomposition and solution module is used to decompose the two-stage robust optimization model into a main problem and sub-problems. Through alternating iterative solutions, it outputs a low-carbon economic operation scheduling strategy for industrial parks that meets the power balance requirements of the worst-case scenario.
[0036] One specific implementation method is as follows: To verify the optimization method combining carbon-energy-cost coupling configuration and two-stage robust scheduling proposed in Example 1, in which the planning layer is used to coordinate the multi-objective optimization configuration of carbon emissions, economy and energy efficiency, and the scheduling layer uses the C&CG algorithm to alternately solve the two-stage robust model to ensure the system's risk resistance capability, a simulation verification and strategy analysis of the system are carried out using an electrical manufacturing industrial park in Jiangsu as an example.
[0037] The park covers an area of 182 mu (approximately 12.8 hectares), is equipped with two 4000kVA transformers connected to the power grid, and has a maximum load of 1650kW. It is capable of configuring distributed photovoltaic, decentralized wind power, and industrial and commercial energy storage. Based on the complete method of this invention, the following verification scenarios and operating states were established and simulated for comparative analysis: First, analyze the capacity configuration results, such as Figure 4 and Figure 5 As shown. At this stage, under the system's baseline configuration (800kW photovoltaic, no wind turbines, no energy storage), renewable energy accounts for only 8.89%, with annual carbon emissions reaching 4.0193 million kWh. After introducing a three-dimensional coupling coefficient matrix and solving using IMOPSO, the globally optimal configuration was obtained: 3738kW photovoltaic power, 500kW wind turbines, and 4200kWh energy storage. At this stage, the largest coupling term demonstrated a strong synergy between economic cost and carbon emissions. The optimized configuration resulted in an average annual cost reduction of 29.7% and a 40.3% reduction in carbon emissions over the entire lifecycle. This demonstrates that fully considering the carbon-energy-cost coupling relationship during the planning stage can fundamentally change the energy structure of the industrial park.
[0038] Next, the robustness parameter selection for the scheduling layer is analyzed, and a two-step method is used to determine the robustness parameters. Under the engineering benchmark of a joint scheduling feasibility probability of no less than 90%, the budget uncertainty parameters for photovoltaic and wind turbines are adjusted (…). and Its sensitivity analysis is as follows: Figure 9 and Figure 10 As shown in the figure. Simulation results show that the daily comprehensive operating cost of the system increases monotonically with the increase of the robust parameters. When selecting the recommended parameter combination, a feasibility probability guarantee of 90.9% was achieved with a cost increase of only 7.1%, and the marginal cost is extremely low, which fully demonstrates the unique advantages of the two-stage framework in terms of both robustness and economy.
[0039] To further analyze the effectiveness of the scheduling mechanism, four progressive scenarios were designed and their operating costs and carbon emission performance were compared: Scenario 1: Deterministic economic dispatch baseline; Scenario 2: Add time-based dynamic carbon emission factors and tiered carbon trading costs to Scenario 1; Scenario 3: Add a flexible demand response mechanism based on Scenario 2; Scenario 4: Taking into account carbon trading and demand response, and introducing a two-stage robust optimization final strategy (corresponding to the complete method in Implementation Example 1).
[0040] Simulations were performed on these four optimized scheduling scenarios, and the specific scenario strategy analysis is as follows: Comparing Scenario 1 and Scenario 2, the system's daily operating cost decreased by 3.9%, and carbon emissions decreased by 7.0%, indicating that the internalization of tiered carbon trading effectively incentivized energy storage to charge and discharge during low-carbon periods. After introducing Scenario 3, demand response smoothed the load curve, significantly reducing the peak-to-valley difference (e.g., Figure 11 As shown), costs plummeted by 28.2%, at which point the load and photovoltaic output periods highly overlapped, promoting the consumption of new energy; Scenario 4, based on Scenario 3, introduced robust constraints to cope with the worst fluctuations in wind and solar power output, and reconstructed the power balance strategy (such as...). Figure 12 As shown in the figure, although this leads to a slight increase in operating costs of 7.0% and an increase in carbon emissions of 4.5%, it is exchanged for a power balance feasibility probability of over 90%. During special periods when electricity prices are high or default penalties are low, VPPs can choose to proactively bear a small amount of wind and solar curtailment or grid-purchased electricity costs. When resource regulation costs are lower than penalty costs, each energy storage and generator unit responds quickly to make up for the deviation.
[0041] Based on the column constraint generation algorithm (C&CG), the main-subproblem alternating solution process (e.g.) Figure 8 As shown in the figure, after several iterations of alternating solutions, the upper and lower bounds quickly close, and the output of each unit and the energy storage cost reach a constant value, which satisfies the convergence result of the two-stage robust optimization algorithm, proving the high solution efficiency of the low-carbon economy optimization scheduling algorithm under the worst scenario.
[0042] In summary, to address the risks of power mismatch caused by inaccurate wind and solar power output forecasting and load fluctuations during the operation of industrial parks, a low-carbon and economical operation strategy for integrated energy systems is constructed using a carbon-energy-cost coupled capacity configuration and a two-stage robust scheduling framework. The park operation platform is responsible for implementing the globally cost-optimal configuration and scheduling plan, and the IMOPSO and C&CG algorithms are used to solve the complex model. This method not only achieves an effective trade-off between the three objectives of carbon emissions, energy utilization, and economic costs at the planning level, but also internalizes carbon costs at the operational level, avoiding the error lag of static carbon factors. Simulation experiments in multiple scenarios also verify the high convergence efficiency of the C&CG algorithm and the reliability of the two-stage robust model in dealing with random fluctuations.
[0043] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.
Claims
1. A low-carbon economic dispatch method based on the fusion of carbon energy cost coupling coefficient matrix and tiered carbon trading, characterized in that, It includes two collaborative execution phases: capacity optimization configuration at the planning layer and day-ahead scheduling optimization at the operational layer. The specific steps are as follows: Step S1: Obtain historical data on industrial park load, wind power output, and photovoltaic power output, and construct a typical daily scenario set for the four seasons that takes into account the random fluctuation characteristics of source and load; Step S2: Taking the minimum overall operating cost, minimum carbon emission cost, and maximum energy utilization rate of the system throughout its entire life cycle as the multi-dimensional optimization objectives, a three-dimensional carbon-energy-cost coupling coefficient matrix is introduced, constructed based on the fusion of Pearson correlation coefficient and subjective preference weights. The normalized three objective values are mapped to a quadratic comprehensive evaluation scalar. A carbon-energy-cost coupled multi-objective capacity optimization configuration model is constructed, and the globally optimal equipment capacity configuration is obtained by solving the model. Step S3: Based on the globally optimal equipment capacity configuration, construct an industrial park carbon trading model that includes time-of-use dynamic carbon emission factors and tiered carbon trading costs, and at the same time establish a flexible user energy demand response model that takes into account transferable loads and loads that can be reduced. Step S4: Embed the tiered carbon trading cost into the system's day-ahead scheduling objective function in the form of an endogenously coupled progressive economic penalty term, with the goal of minimizing the system's overall daily operating cost. In addition, for the random fluctuation characteristics of wind and solar power output, construct a two-stage robust optimization scheduling model based on the budget uncertainty set. Step S5: The column constraint generation algorithm C&CG is used to decompose the two-stage robust optimization scheduling model into a main problem and sub-problems for alternating iterative solution. Under the worst-case power balance constraint, the low-carbon economic operation scheduling strategy of the industrial park is generated and output.
2. The low-carbon economic dispatch method for carbon energy cost coupling coefficient matrix fusion tiered carbon trading according to claim 1, characterized in that: In step S1, the specific method for constructing the typical daytime scene set for the four seasons is as follows: Step S11: Fit the historical wind speed data to the probability density function using a Weibull distribution for each time period to obtain the shape parameters. and scale parameters The wind speed simulation sequence was generated using Monte Carlo random sampling, and the Weibull distribution probability density function is: ； In the formula, Real-time wind speed; Step S12: Obtain the wind power output simulation scenario by mapping the piecewise linear power curve of the wind turbine. The theoretical output power model of the wind turbine is as follows: ； In the formula, This refers to the rated power of the fan. To cut into wind speed, Rated wind speed; Step S13: Anomaly detection and correction are performed on the historical photovoltaic power output data using the Median Absolute Deviation (MAD) method. The photovoltaic theoretical output power model using extended temperature correction is as follows: ； In the formula, This refers to the rated installed capacity of photovoltaic power. For real-time irradiance, Irradiance under standard test conditions. The overall efficiency coefficient of photovoltaic power. The photovoltaic temperature coefficient, Real-time temperature of photovoltaic modules. Standard test temperature; Step S14: Slice the cleaned wind power, photovoltaic and load data by day, perform K-means clustering in each of the four seasons, and reduce them by probability weighting to a single representative typical daily curve for each season, forming a typical daily scene set for each season.
3. The low-carbon economic dispatch method for carbon energy cost coupling coefficient matrix fusion tiered carbon trading according to claim 1, characterized in that: In step S2, the functional expression of the multidimensional optimization objective is: Minimize overall operating costs: ； In the formula, This represents the annualized cost of the total equipment investment. Annual fixed operation and maintenance costs, Weighted average operating cost per typical day; Minimize carbon emission costs: ； ； In the formula, This represents the park's total annual carbon emissions. The annual average carbon emission factor of the power grid. For the annual electricity purchased from the grid in the park, The carbon emission factor for diesel power generation. This represents the annual diesel power generation capacity of the industrial park. This is the carbon emission cost price coefficient; To achieve the goal of maximizing energy efficiency, inverse normalization is used to transform the search for maximum value into the search for minimum value: ； In the formula, , They are respectively Actual electricity consumption of photovoltaic and wind power during the period , They are respectively Available power generation from photovoltaic and wind power during the specified time period.
4. The low-carbon economic dispatch method for carbon energy cost coupling coefficient matrix fusion tiered carbon trading according to claim 1, characterized in that: In step S2, the three-dimensional carbon energy-feed coupling coefficient matrix With quadratic comprehensive evaluation scalar The construction method is as follows: matrix diagonal elements The self-weights of each objective, and the off-diagonal elements. For the goal With the goal The coupling coefficient between them is constructed using the following steps: The first step is to randomly sample the decision variables to generate simulation results and calculate the Pearson correlation coefficients between each pair of the three normalized objective values. ; The second step is to combine the subjective preference weight vector. Calculate diagonal elements Off-diagonal elements: ; The third step is to verify and correct the symmetry and positive definiteness of the obtained matrix to ensure the comprehensive evaluation scalar. It is a convex function over the entire domain; Based on the coupling coefficient matrix Normalize the column vector of the three objective ranges. Mapped to a quadratic comprehensive evaluation scalar ,in , , These are the normalized values of comprehensive operating costs, carbon emission costs, and energy utilization efficiency, respectively. The expression is: ； In the formula, the diagonal terms Reflecting the contribution of a single objective to the overall evaluation, cross-items. Explicitly characterize the coupling effect between two targets. The smaller the value, the better the overall performance of the solution.
5. The low-carbon economic dispatch method for carbon energy cost coupling coefficient matrix fusion tiered carbon trading according to claim 1, characterized in that: In step S2, the solution method for the multi-objective capacity optimization configuration model includes the following steps: Step S21: Solve the Pareto optimal solution set using the improved multi-objective particle swarm optimization algorithm (IMOPSO). The algorithm improvement mechanism includes: An inertia weight that decreases linearly with the number of iterations is adopted. and collaboratively dynamically adjusted individual learning factors With social learning factors The expression is: ； ； ； In the formula, This represents the current iteration number. The maximum number of iterations, , These are the maximum and minimum values of the inertia weight, respectively. , These represent the maximum and minimum values of the individual learning factor, respectively. , These represent the maximum and minimum values of the social learning factor, respectively. Step S22: Introduce a Pareto non-dominated solution archive maintenance mechanism based on congestion distance weighting to remove solutions with excessive congestion and maintain the diversity of the solution set; Step S23: In the global guided particle selection stage, a weighted roulette wheel betting strategy based on crowding distance is adopted to balance the global search capability and local convergence speed of the algorithm. Step S24: Combining the TOPSIS method and the weighted rank ratio method, select the globally optimal device capacity configuration from the Pareto optimal solution set.
6. The low-carbon economic dispatch method for carbon energy cost coupling coefficient matrix fusion tiered carbon trading according to claim 1, characterized in that: In step S3, the method for constructing the time-sharing dynamic carbon emission factor is as follows: Annual average carbon emission factor of the power grid Using the baseline value, a time period correction factor is introduced. The expression is: ； In the formula, for Carbon emission factors of power grid during a given period for Carbon emission correction factor for different time periods, peak load periods of the power grid Low-load off-peak periods or periods of high new energy output Normally ; The method for constructing a carbon trading model for industrial parks is as follows: Determining the initial carbon emission quota for the park based on the baseline method. The actual carbon emissions are calculated using time-sharing dynamic carbon emission factors. The actual carbon emissions calculation formula is: ； In the formula, for Power purchased by the park during different time periods The scheduling time step; when At that time, progressive penalty settlement will be carried out according to the preset tiered carbon price, with the carbon price increasing proportionally for each higher tier. when At that time, the park sells its remaining carbon allowances to generate carbon emission reduction revenue.
7. The low-carbon economic dispatch method for carbon energy cost coupling coefficient matrix fusion tiered carbon trading according to claim 1, characterized in that: In step S3, the method for constructing the flexible user energy demand response model is as follows: Transferable load model: The total energy of transferable loads is conserved within a specified time window, and the compensation revenue during its scheduling cycle is... for: ； In the formula, The unit price for compensation of transferable load. for The transfer power of the load that can be transferred during a given period; Load reduction model: its compensation revenue during the scheduling cycle for: ； In the formula, To reduce the unit compensation price per load unit, for User baseline load power for the time period for Periodic load reduction ratio; Total Revenue from Demand Response .
8. The low-carbon economic dispatch method for carbon energy cost coupling coefficient matrix fusion tiered carbon trading according to claim 1, characterized in that: In step S4, the day-ahead scheduling objective function and core constraints are as follows: The objective function for minimizing the system's daily overall operating cost is: ； In the formula, The cost of purchasing electricity under time-of-use pricing. Revenue from selling surplus electricity to the grid. The net cost of tiered carbon trading. For equipment operation and maintenance costs, To incur the cost of energy abandonment, Total revenue from demand response; The core constraints include: Power balance constraints: ； In the formula, , They are respectively Power purchase and sale by the power grid during specific time periods , They are respectively Time-of-use energy storage charging and discharging power, for Total load after demand response period for Power curtailed during a given period; Dynamic constraints on the state of charge (SOC) of energy storage: ； In the formula, , These refer to energy storage charging and discharging efficiency, respectively. This refers to the rated capacity of the energy storage. Based on budget uncertainty set Two-stage robust optimization scheduling model: ； In the formula, As the baseline scheduling variable for the first phase, The amount of energy contributed to wind and solar power is uncertain. For a budget uncertainty set, This is a variable for real-time adjustment in the second stage.
9. The low-carbon economic dispatch method for carbon energy cost coupling coefficient matrix fusion tiered carbon trading according to claim 1, characterized in that: In step S5, the solution process for the column constraint generation algorithm is as follows: Step S51: Decompose the three-level min-max-min problem of two-stage robust optimization scheduling into a main problem MP and a sub-problem SP; Step S52: Solve the main problem MP. Under the known worst-case scenario, obtain the optimal baseline solution and lower bound for the first stage. The expression for the main problem is: ； Step S53: Fix the optimal solution for the first stage Solve the subproblem SP in the uncertain set. The search above finds the most unfavorable deviation scenario and upper bound that maximizes the adjustment cost in the second stage. The subproblem expression is: ； In the formula, , , , This is the constraint coefficient matrix; Step S54: Determine whether the difference between the upper and lower bounds meets the preset convergence accuracy. If it does, output the final low-carbon economic operation scheduling strategy. If it does not, add the constraint corresponding to the newly identified worst scenario to the main problem and return to step S52 to continue iterating.
10. A low-carbon economic dispatch system for tiered carbon trading based on the fusion of carbon energy cost coupling coefficient matrix, characterized in that: The low-carbon economic dispatch method for implementing the carbon energy cost coupling coefficient matrix fusion tiered carbon trading as described in any one of claims 1-9 includes: The data acquisition and scenario construction module is used to acquire historical data on industrial park load, wind power and photovoltaic output, complete data preprocessing and anomaly correction, and construct typical daily scenario sets for the four seasons. The carbon-energy-cost coupled capacity optimization module is used to construct a multi-objective optimization model of comprehensive operating cost, carbon emission cost and energy utilization rate, generate a three-dimensional carbon-energy-cost coupling coefficient matrix and map it into a comprehensive evaluation scalar, and use an improved multi-objective particle swarm algorithm to solve and select the globally optimal equipment capacity configuration. The carbon trading and demand response modeling module is used to determine the initial carbon emission quota of the park, generate time-sharing dynamic carbon emission factors, establish a tiered carbon trading cost model, and construct a flexible demand response model for transferable and reduceable loads. The robust optimization scheduling module is used to embed the tiered carbon trading cost as an endogenous penalty term into the day-ahead scheduling objective function, establish a unified daily comprehensive operating cost objective function and core constraints, and construct a two-stage robust optimization scheduling model based on the budget uncertainty set. The C&CG decomposition and solution module is used to decompose the two-stage robust optimization model into a main problem and sub-problems. Through alternating iterative solutions, it outputs a low-carbon economic operation scheduling strategy for industrial parks that meets the power balance requirements of the worst-case scenario.