An electric-car-hydrogen collaborative optimization scheduling method for carbon emission control and wind-solar power consumption

Abstract

The application discloses an electricity-carbon-hydrogen collaborative optimization scheduling method for carbon emission management and wind-solar power consumption, first, a dynamic model of "knowing carbon by electricity" is constructed by fusing multi-source data, and key carbon sources are identified; second, the uncertainty of wind-solar power output is described by using the Copula theory and clustering method; then, an electricity-carbon-hydrogen multi-energy flow system state model is established; then, a robust optimization model is constructed with the minimum carbon emission and the minimum interactive power as the target; finally, a multi-objective intelligent algorithm is used for solving and closed-loop verification. The application innovatively deeply fuses carbon emission online perception, carbon source tracking evaluation and wind-solar hydrogen system robust scheduling, and solves the problem of carbon-electricity collaborative optimization.

Description

Technical Field

[0001] This invention belongs to the field of power system optimization scheduling and carbon emission control technology, specifically involving a method and system for coordinated optimization scheduling of electricity, carbon and hydrogen for carbon emission control and wind and solar power integration. It is particularly applicable to the coordinated optimization problem of power system carbon emission reduction and fluctuating new energy integration under the background of high proportion of renewable energy access. Background Technology

[0002] Existing technologies suffer from the following shortcomings: 1) Disconnection between carbon emission control and power dispatch: Traditional dispatch models prioritize economic efficiency, lacking precise tracking and online constraints of carbon emission sources, leaving carbon reduction in a passive accounting state. 2) Prominent contradiction between renewable energy consumption and system stability: Wind and solar power output exhibits strong uncertainty and volatility. Traditional deterministic dispatch or single robust optimization methods struggle to mitigate power fluctuations while ensuring consumption, and are not linked to carbon reduction targets. 3) Lack of a collaborative optimization framework integrating multi-source data: The technology fails to deeply integrate the carbon emission perception capability of "knowing carbon through electricity," the distribution network carbon source tracking and assessment capability, and the flexible adjustment capability of the wind, solar, and hydrogen multi-energy flow system to form a closed-loop optimization system, resulting in limited dispatch effectiveness. Summary of the Invention

[0003] Based on the above, this invention provides a collaborative optimization scheduling method for electricity-carbon-hydrogen for carbon emission control and wind-solar integration. It innovatively integrates online carbon emission sensing, carbon source tracking and assessment, and robust scheduling of wind-solar-hydrogen systems, thus solving the problem of carbon-electricity collaborative optimization.

[0004] The technical solution of this invention is: a collaborative optimization scheduling method for electricity-carbon-hydrogen for carbon emission control and wind and solar energy consumption, comprising the following steps:

[0005] S1: Integrate multi-source data from the power production process, including energy type, power generation, load power, transmission and distribution losses, and equipment operating status data. Establish a carbon emission model based on dynamic carbon emission factors, estimate the carbon emissions of each link in the system in real time, and use cluster analysis and entropy weight method to identify key carbon source links in the distribution network and assess their carbon emission intensity.

[0006] S2: Input historical wind power output data and historical photovoltaic output data, use the Copula function to construct the joint probability distribution of wind power and photovoltaic to quantify the spatiotemporal correlation, generate an initial scene set through Monte Carlo sampling, and use an improved K-means clustering algorithm to reduce the scene to obtain a typical wind and solar power joint output scene set and its occurrence probability.

[0007] S3: Construct a multi-energy flow coupled system including wind power, photovoltaic, electrolyzer, hydrogen storage tank, fuel cell and local load, and define the system state space model, where the state variables include the state of charge of the hydrogen energy storage system and the power of the electrolyzer, and the output variables include the grid interaction power and the real-time carbon emission intensity of the system.

[0008] S4: With the dual objectives of minimizing the total carbon emissions of the system and minimizing the power interaction with the grid during the scheduling period, an optimization objective function is established. Considering the typical wind and solar combined power output scenario set, a min-max robust optimization framework is established to cope with uncertainties, and a system constraint set is constructed, including power balance constraints, upper and lower limit constraints of equipment operation, opportunity constraints of hydrogen energy storage system, and carbon emission intensity constraints.

[0009] S5: An improved multi-objective genetic algorithm based on the NSGA-II framework is used to solve the robust optimization model in a rolling manner, and the optimal scheduling instruction set is output, including the electrolyzer hydrogen production power, fuel cell power generation power and grid interaction power plan.

[0010] S6: Execute the scheduling instruction set, collect actual system operation data and carbon emission monitoring data, verify the actual data with the model prediction values, and dynamically update the parameters of the carbon emission model and the probability distribution of the wind-solar combined output scenario set.

[0011] Preferably, the dynamic carbon emission factor is dynamically adjusted based on the proportion of thermal power, hydropower, and new energy power generation in the real-time power generation structure.

[0012] Preferably, the Copula function is a Frank-Copula function.

[0013] Preferably, the improved K-means clustering algorithm uses the elbow method to determine the optimal number of clusters, and the number of clusters is set to 4.

[0014] Preferably, the confidence level of the opportunity constraint of the hydrogen energy storage system is set to 95%, and the carbon emission intensity constraint is a time-limited upper limit constraint.

[0015] Preferably, the population size of the improved multi-objective genetic algorithm is set to 200, the maximum number of iterations is set to 300, and an adaptive crossover and mutation operator is adopted.

[0016] Preferably, the dual objectives are transformed into a single objective and solved using a weighted summation method, with the weighting coefficients dynamically adjusted according to the emphasis on carbon emission reduction and power fluctuation during the scheduling phase.

[0017] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0018] The beneficial effects of this invention are as follows: First, this invention constructs a dynamic model of "knowing carbon through electricity" by integrating multi-source data and identifies key carbon sources; second, it uses Copula theory and clustering methods to characterize the uncertainty of wind and solar power output; then, it establishes a state model of the electricity-carbon-hydrogen multi-energy flow system; next, it constructs a robust optimization model with the objectives of minimizing carbon emissions and minimizing interactive power; finally, it uses a multi-objective intelligent algorithm to solve the problem and perform closed-loop verification. This invention innovatively integrates online carbon emission sensing, carbon source tracking and assessment, and robust scheduling of wind-solar-hydrogen systems, solving the problem of carbon-electricity synergistic optimization. Verification using measured data from Xinjiang shows that this method reduces system carbon emissions by 18.5% while decreasing wind and solar curtailment rates by up to 94.21% and grid-connected power fluctuations by 57.83%, achieving synergistic optimization of carbon emission control and efficient renewable energy consumption. Attached Figure Description

[0019] Figure 1 This is a flowchart of the implementation method of the present invention;

[0020] Figure 2 The results of generating typical scenarios of wind and solar power combined output according to the present invention;

[0021] Figure 3 This is a comparison chart of the results of the electro-carbon-hydrogen synergistic optimization scheduling of the present invention with those of traditional methods. Detailed Implementation

[0022] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0023] This invention provides a collaborative optimization scheduling method for electricity-carbon-hydrogen energy transfer aimed at carbon emission control and wind / solar energy integration. This method achieves dual-objective optimization by integrating multi-source data, modeling wind / solar uncertainties, constructing a multi-energy flow system, establishing a robust optimization model, employing intelligent algorithms for solution, and implementing closed-loop verification and updates. Reference Figure 1 The specific steps of the present invention will be described in detail below:

[0024] Step S1: Multi-source electricity-carbon data fusion and carbon source identification

[0025] This step integrates multi-source data from the power generation process to establish a dynamic carbon emission model and identify key carbon source processes.

[0026] S11: Input multi-source data

[0027] Multi-source data is collected from the power generation process, including energy type (such as thermal power, hydropower, wind power, and photovoltaic power), power generation, load power, transmission and distribution losses, and equipment operating status data. This data is acquired in real time through smart meters, SCADA systems, or IoT devices, with a sampling frequency ranging from 1 minute to 1 hour, depending on the dispatch cycle. The data is in structured format and stored in a database for subsequent processing.

[0028] S12: Establish a carbon emission model based on electricity consumption.

[0029] The system estimates carbon emissions at each stage based on a dynamic carbon emission factor. The dynamic carbon emission factor is dynamically adjusted according to the proportion of thermal power, hydropower, and renewable energy in the real-time power generation structure to reflect the carbon intensity of the current power generation mix.

[0030] Specifically, the dynamic carbon emission factor is calculated as follows:

[0031] Let the real-time total power generation be Thermal power generation capacity is Hydropower generation capacity is The power generation capacity of new energy sources is (Including wind power and solar power). Then the proportion of thermal power... Hydropower account for The proportion of new energy Dynamic carbon emission factor The calculation formula is:

[0032]

[0033] in, It is the benchmark carbon emission factor for thermal power. It is the baseline carbon emission factor for hydropower. These are the benchmark carbon emission factors for new energy sources, and these benchmark values ​​can be adjusted according to region or standard.

[0034] carbon emissions The estimate is:

[0035]

[0036] in, This refers to the time interval. In one example, the model updates in real time, adjusting factors every 5-15 minutes to improve estimation accuracy.

[0037] S13: Identifying Key Carbon Sources

[0038] Cluster analysis and entropy weight method were used to identify key carbon source links in the power distribution network, including transformer losses, line losses and power distribution equipment losses.

[0039] Cluster analysis: The K-means algorithm is used to cluster the loss data to identify high-loss nodes.

[0040] Entropy weight method: Calculates the entropy weight of each stage to assess carbon emission intensity. Carbon emission intensity Defined as:

[0041]

[0042] in It represents the carbon emissions corresponding to losses in each process. It refers to load power. The weight of each component is determined using the entropy weight method, and components with high weights are processed first.

[0043] Step S2: Uncertainty Modeling of Combined Wind and Solar Power Output

[0044] This step addresses the uncertainties in wind and solar power output and generates a set of typical scenarios.

[0045] S21: Construct the joint probability distribution

[0046] Input historical wind power output data and historical photovoltaic power output data (time series data, 15-minute resolution), and use the Copula function to construct the joint probability distribution of wind power and photovoltaic power to quantify the spatiotemporal correlation.

[0047] Copula function selection: Use the Frank-Copula function, whose expression is:

[0048]

[0049] in, and It is the marginal distribution function of wind power and photovoltaic power output (using empirical distribution or normal distribution). These are relevant parameters, obtained by fitting from historical data using the maximum likelihood estimation method. The Frank-Copula method can effectively capture nonlinear dependencies.

[0050] S22: Generate a set of typical scenes

[0051] An initial scene set, such as 1000 scenes, is generated by Monte Carlo sampling. Then, an improved K-means clustering algorithm is used to reduce the number of scenes to obtain a typical set of wind and solar power combined output scenes and their occurrence probabilities.

[0052] Improved K-means clustering: The elbow method is used to determine the optimal number of clusters. The elbow method calculates the sum of squares within each cluster under different numbers of clusters. ,choose Points with a slower rate of decline are selected as the optimal number of clusters. In this implementation, the number of clusters is set to 4 to balance computational efficiency and scene representativeness.

[0053] In one example, after scene reduction, four typical scenarios are obtained, each with a corresponding probability of occurrence, such as 25.4%, 42.0%, 1.8%, and 30.8%. Figure 2 As shown.

[0054] Step S3: State modeling of the electric-carbon-hydrogen co-system

[0055] This step involves constructing a multi-energy flow coupled system and defining a state-space model.

[0056] S31: Constructing a multi-energy flow coupled system

[0057] The system comprises wind power, solar power, an electrolyzer, a hydrogen storage tank, a fuel cell, and local loads. The components are coupled via electrical and hydrogen flow: wind and solar power directly supply the loads or the electrolyzer; the electrolyzer converts excess electrical energy into hydrogen, which is stored in the hydrogen storage tank; the fuel cell converts hydrogen into electrical energy when needed to supply the loads; and grid interconnection power is used to balance supply and demand.

[0058] S32: Define the state-space model

[0059] State variables include the state of charge of the hydrogen energy storage system. The output variables include the power of the electrolyzer and the power of the grid interaction and the real-time carbon emission intensity of the system.

[0060] Equations of state:

[0061]

[0062] in, It is the state of energy storage (SOC) of hydrogen at time t (range 0-1). It refers to the hydrogen production capacity of the electrolyzer. It refers to the power generation capacity of the fuel cell. and It is efficiency (typical value 0.6-0.7). This refers to the hydrogen storage tank capacity (kWh).

[0063] Output equation:

[0064] Grid interaction power:

[0065]

[0066] Real-time carbon emission intensity Model calculation based on S12.

[0067] Step S4: Construction of a Multi-Objective Robust Optimization Scheduling Model

[0068] This step establishes the objective function and constraint set for optimization to address uncertainties.

[0069] S41: Establish the optimization objective function

[0070] The dual objectives are to minimize the total carbon emissions of the system and the power grid interaction within the scheduling period T.

[0071]

[0072] in, It represents the carbon emissions at time t.

[0073] Objective transformation: The objective is transformed into a single objective for solution using a weighted summation method.

[0074]

[0075] Weighting coefficient and Dynamic adjustments based on scheduling phases: During key carbon emission reduction periods (such as peak load periods), increase... During periods sensitive to power fluctuations, increase The weight range is 0-1, and .

[0076] S42: Establish a robust min-max optimization framework

[0077] Considering the typical wind and solar power output scenarios generated by S2, a min-max robust optimization framework is established to handle the worst-case uncertainty. The optimization problem is formulated as follows:

[0078]

[0079] in, These are decision variables (such as electrolyzer power and fuel cell power). It is a scene of scenic contribution. It is a typical set of scenarios.

[0080] S43: The system constraint set includes the following constraints:

[0081] Power balance constraints: as described in S32.

[0082] Equipment operating limits: for example, electrolytic cell power. fuel cell power .

[0083] Opportunity constraints for hydrogen energy storage systems: The confidence level is set at 95%, i.e.:

[0084] in and This is the SOC safety range (e.g., 0.2-0.8).

[0085] Carbon emission intensity constraints: Time-specific upper limits, such as carbon emission intensity not exceeding a certain limit during peak hours. .

[0086] Step S5: Rolling optimization solution and instruction issuance

[0087] This step uses an improved multi-objective genetic algorithm to solve the optimization model and outputs scheduling instructions.

[0088] S51: Improved multi-objective genetic algorithm for solving

[0089] An improved multi-objective genetic algorithm based on the NSGA-II framework is used to solve the robust optimization model in a rolling manner.

[0090] The population size is set to 200, and the maximum number of iterations is set to 300. An adaptive crossover and mutation operator is used: crossover probability... Initially 0.9, dynamically adjusted based on population diversity; mutation probability. The initial value is 0.1, which decreases as the number of iterations increases. The solution is recalculated for each scheduling cycle (e.g., 1 hour), taking into account the latest data and scenario.

[0091] S52: Output scheduling instruction set

[0092] Each scheduling cycle outputs the optimal scheduling instruction set, including the electrolyzer hydrogen production power, fuel cell power generation power, and grid interaction power plan. The instructions are issued to the executing equipment in a time-series format.

[0093] Step S6: Closed-loop verification and model update

[0094] This step executes scheduling instructions, collects actual data, and dynamically updates the model.

[0095] S61: Execute instructions and collect data

[0096] The system executes scheduling instructions and collects actual system operation data (such as actual wind and solar power output and load power) and carbon emission monitoring data (from carbon monitoring equipment). The data sampling frequency is consistent with the scheduling cycle.

[0097] S62: Checksum Dynamic Update

[0098] The actual data is compared with the model predictions, and error metrics (such as root mean square error) are calculated. The parameters of the "Carbon Detection Based on Electricity" model are dynamically updated (e.g., adjusting the baseline carbon emission factor) and the probability distribution of wind and solar scenarios (e.g., refitting the Copula parameters). The update frequency is daily or weekly, depending on the stability of the data.

[0099] Application Case: An Explanation Using Measured Data from Xinjiang

[0100] Depend on Figure 3 It can be seen that the Copula-based method generates 1000 initial scenarios through Monte Carlo simulation, and then performs an inverse transformation to obtain the physical scenarios. The application of the enhanced k-means clustering algorithm produces four probability-weighted typical scenarios (25.4%, 42.0%, 1.8%, and 30.8%). Only the extreme scenario (1.8%) is selected for method comparison study. It is found that the traditional method results in concentrated power curtailment (maximum 2.1 MW) at 11:00 due to the sudden change in wind and solar power generation, forming a steep power peak. The method of this invention, through robust optimization and active adjustment of energy storage plan, distributes the power curtailment event as a smooth fluctuation between 09:00 and 15:00 (<0.8MW).

[0101]

[0102] As shown in Table 1, the method significantly reduced wind and solar curtailment during typical daily operation (the maximum curtailed power decreased from 9.85MW to 0.3MW), smoothed grid-connected power fluctuations (the standard deviation of fluctuation decreased by 52.0%), and achieved a reduction of approximately 18.5% in system carbon emission intensity. This verifies the synergistic effectiveness of the method in improving renewable energy consumption and strengthening carbon emission control.

[0103] Therefore, this invention adopts the above-mentioned electricity-carbon-hydrogen coordinated optimization scheduling method for carbon emission control and wind and solar power consumption. Through multi-source electricity-carbon data fusion and carbon source tracking, wind and solar uncertainty modeling, coordinated state description of the electricity-carbon-hydrogen system, multi-objective robust optimization and closed-loop verification, it successfully achieves the coordinated optimization goal of power system carbon emission reduction and efficient consumption of new energy under strong uncertainty.

[0104] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A collaborative optimization scheduling method for electricity-carbon-hydrogen energy for carbon emission control and wind and solar energy consumption, characterized in that, Includes the following steps: S1: Integrate multi-source data from the power production process, including energy type, power generation, load power, transmission and distribution losses, and equipment operating status data. Establish a carbon emission model based on dynamic carbon emission factors, estimate the carbon emissions of each link in the system in real time, and use cluster analysis and entropy weight method to identify key carbon source links in the distribution network and assess their carbon emission intensity. S2: Input historical wind power output data and historical photovoltaic output data, use the Copula function to construct the joint probability distribution of wind power and photovoltaic to quantify the spatiotemporal correlation, generate an initial scene set through Monte Carlo sampling, and use an improved K-means clustering algorithm to reduce the scene to obtain a typical wind and solar power joint output scene set and its occurrence probability. S3: Construct a multi-energy flow coupled system including wind power, photovoltaic, electrolyzer, hydrogen storage tank, fuel cell and local load, and define the system state space model, where the state variables include the state of charge of the hydrogen energy storage system and the power of the electrolyzer, and the output variables include the grid interaction power and the real-time carbon emission intensity of the system. S4: With the dual objectives of minimizing the total carbon emissions of the system and minimizing the power interaction with the grid during the scheduling period, an optimization objective function is established. Considering the typical wind and solar combined power output scenario set, a min-max robust optimization framework is established to cope with uncertainties, and a system constraint set is constructed, including power balance constraints, upper and lower limit constraints of equipment operation, opportunity constraints of hydrogen energy storage system, and carbon emission intensity constraints. S5: An improved multi-objective genetic algorithm based on the NSGA-II framework is used to solve the robust optimization model in a rolling manner, and the optimal scheduling instruction set is output, including the electrolyzer hydrogen production power, fuel cell power generation power and grid interaction power plan. S6: Execute the scheduling instruction set, collect actual system operation data and carbon emission monitoring data, verify the actual data with the model prediction values, and dynamically update the parameters of the carbon emission model and the probability distribution of the wind-solar combined output scenario set.

2. The method for coordinated optimization scheduling of electricity, carbon, and hydrogen for carbon emission control and wind and solar energy consumption as described in claim 1, characterized in that, The dynamic carbon emission factor is dynamically adjusted based on the proportion of thermal power, hydropower, and new energy power generation in the real-time power generation structure.

3. The method for coordinated optimization scheduling of electricity, carbon, and hydrogen for carbon emission control and wind and solar energy consumption as described in claim 1, characterized in that, The Copula function is the Frank-Copula function.

4. The method for coordinated optimization scheduling of electricity, carbon, and hydrogen for carbon emission control and wind and solar energy consumption as described in claim 1, characterized in that, The improved K-means clustering algorithm uses the elbow method to determine the optimal number of clusters, and the number of clusters is set to 4.

5. The method for coordinated optimization scheduling of electricity, carbon, and hydrogen for carbon emission control and wind and solar energy consumption as described in claim 1, characterized in that, The confidence level of the opportunity constraint for the hydrogen energy storage system is set at 95%, and the carbon emission intensity constraint is a time-limited upper limit constraint.

6. The method for coordinated optimization scheduling of electricity, carbon, and hydrogen for carbon emission control and wind and solar energy consumption as described in claim 1, characterized in that, The improved multi-objective genetic algorithm has a population size of 200, a maximum number of iterations of 300, and adopts adaptive crossover and mutation operators.

7. The method for coordinated optimization scheduling of electricity, carbon, and hydrogen for carbon emission control and wind and solar energy consumption as described in claim 1, characterized in that, The dual objectives are transformed into a single objective by a weighted summation method, and the weighting coefficients are dynamically adjusted according to the emphasis on carbon emission reduction and power fluctuation during the scheduling phase.

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

Images