A new energy installed capacity scale and space-time layout determination method based on contribution degree attribution

Abstract

The present application relates to a kind of new energy installed capacity scale and space-time layout determination method based on contribution degree attribution, belong to new energy installed capacity planning and power system dispatching technical field, including: the construction grid system total energy consumption and carbon emission space-time quantification model, including with system energy consumption and carbon emission total minimum target function and grid constraint condition;Based on the given grid system parameter value and the current total scale of new energy installed capacity utilizes space-time quantification model to determine the first optimal expansion scale of new energy installed capacity;Determine multiple key parameters in grid system parameter based on the model;Contribution degree attribution analysis is carried out to multiple key parameters, and the attribution index of each key parameter is determined;When grid condition, i.e. the change value of key parameter value occurs, the second optimal expansion scale of new energy installed capacity is determined based on the change value of key parameter and attribution index.This method realizes the continuity of new energy installed capacity scale by contribution degree attribution analysis to grid parameter, and realizes the continuity of new energy installed capacity scale.

Description

Technical Field

[0001] This invention relates to the fields of new energy installed capacity planning and power system dispatching technology, and in particular to a method for determining the scale and spatiotemporal layout of new energy installed capacity based on contribution attribution. Background Technology

[0002] The proportion of new energy sources such as wind power and photovoltaics in the installed capacity of the power system continues to increase, and new energy has become an important component in promoting energy structure transformation and achieving low-carbon development. As the penetration rate of new energy continues to rise, the scale of new energy installed capacity and its rational allocation in different regions and time periods have a significant impact on the safe and stable operation of the power system and the effectiveness of low-carbon transformation. Therefore, how to scientifically determine the regional new energy installed capacity scale and formulate feasible planning schemes while ensuring the safe and stable operation of the power system has become a key issue in new energy planning and power system operation optimization.

[0003] Existing methods for new energy planning and investment decision-making are mostly based on installed capacity ratio constraints, empirical planning indicators, or static economic calculations, and generally suffer from the following shortcomings: First, most methods fail to conduct refined quantitative analysis of changes in new energy installed capacity from the perspective of power system operation, making it difficult to characterize the impact of changes in new energy installed capacity on the system's total energy consumption and carbon emission levels, and thus difficult to accurately determine the reasonable range of new energy installed capacity. Second, existing methods often fail to adequately consider the constraints of coordinated operation of multiple types of power sources and flexible resources, such as coal power, gas power, hydropower, nuclear power, and energy storage, leading to discrepancies between planning results and actual system operation, affecting the feasibility of new energy installed capacity planning schemes. Third, although some methods introduce sensitivity analysis or influencing factor analysis, they mostly remain at the level of ex-post analysis or qualitative interpretation, lacking a systematic characterization of the formation mechanism of new energy installed capacity, making it difficult to effectively transform the analysis results into a basis for subsequent planning decisions. When grid operating conditions change, complex scale calculations often need to be repeated, restricting the application efficiency and continuous guidance capability of planning methods in engineering practice. Summary of the Invention

[0004] Based on the above analysis, this invention aims to disclose a method for determining the scale and spatiotemporal layout of new energy installed capacity based on contribution attribution. By constructing a spatiotemporal quantitative model of the total energy consumption and carbon emissions of the power grid system, the marginal impact of the scale of new energy installed capacity on the energy consumption and carbon emission levels of the power grid system is quantified. Furthermore, the contribution attribution index of key parameters of the power grid system to the power grid is obtained through analysis. Thus, the expansion scale of new energy installed capacity can be continuously guided and planned based on the changes in key parameter values ​​and the attribution index.

[0005] This invention provides a method for determining the installed capacity and spatiotemporal layout of new energy sources based on contribution attribution, which specifically includes the following steps:

[0006] S1. Construct a spatiotemporal quantitative model of total energy consumption and carbon emissions of the power grid system. The model includes an objective function that aims to minimize the total energy consumption and carbon emissions of the system and set power grid constraints. S2. Based on the given values ​​of power grid system parameters and the current total scale of new energy installed capacity, the first optimal expansion scale of new energy installed capacity is determined using the spatiotemporal quantification model. S3. Determine several key parameters in the power grid system parameters based on the spatiotemporal quantification model of total energy consumption and carbon emissions; S4. Perform contribution attribution analysis on multiple key parameters to determine the attribution index of each key parameter; S5. When the grid conditions, i.e. the key parameter values, change, the second optimal expansion scale of new energy installations is determined based on the change values ​​of the key parameters, the attribution index, and the first optimal expansion scale.

[0007] Furthermore, the determination of the first optimal expansion scale of new energy installations using the spatiotemporal quantification model based on the given grid system parameter values ​​and the current total scale of new energy installations includes: Based on the given values ​​of the power grid system parameters, the spatiotemporal quantization model is solved based on the current total scale of new energy installed capacity and multiple pre-expansion scales, respectively, to obtain the sum of system energy consumption and carbon emissions corresponding to the current total scale and the total scale of new energy installed capacity after each pre-expansion. The marginal benefit of new energy installations corresponding to each pre-expansion scale is calculated based on the sum of the energy consumption and carbon emissions of each system. The marginal benefit of new energy installations refers to the reduction in the total energy consumption and carbon emissions of the system brought about by a unit expansion scale. The pre-expansion scale corresponding to the maximum marginal benefit of new energy installations is taken as the first optimal expansion scale.

[0008] Furthermore, the determination of multiple key system parameters in the power grid system parameters based on the spatiotemporal quantification model of total energy consumption and carbon emissions includes: Determine the value range of each parameter of the power grid system; Multiple sets of parameter value combinations for the power grid system are randomly generated based on the value range of each parameter. Based on each of the aforementioned value combinations and the current total scale of new energy installations, the optimal total scale of new energy installations after pre-expansion is obtained using the aforementioned spatiotemporal quantization model for each of the aforementioned value combinations. Based on the combinations of values ​​and the corresponding optimal total scale after the pre-expansion of new energy installed capacity, the global sensitivity analysis method is used to obtain the influence effect of each system parameter on the scale of new energy installed capacity expansion. Multiple key system parameters were determined according to the described impact effects, from largest to smallest.

[0009] Furthermore, the contribution attribution analysis of the multiple key parameters to determine the attribution index of each key parameter includes: Construct a proxy model to predict the optimal total scale of new energy installed capacity after pre-expansion; The proxy model is trained based on each of the aforementioned value combinations and the corresponding optimal total scale after the pre-expansion of new energy installed capacity. The attribution indices of each key parameter are calculated using the Sobol global variance decomposition method based on the trained surrogate model.

[0010] Furthermore, the proxy model is trained by minimizing the following loss function: ; in, For loss function, This represents the optimal total scale of new energy installed capacity after pre-expansion for the j-th combination of values. The model predicts the total installed capacity for the j-th combination of values; N is the number of samples.

[0011] Furthermore, determining the second optimal expansion scale for new energy installations based on the changes in key system parameters, the attribution index, and the first optimal expansion scale includes: Calculate the normalized change magnitude of each key parameter. If it exceeds the first threshold, calculate the second optimal expansion scale using the spatiotemporal quantification model based on the change magnitude of each key parameter, the current total scale of new energy installations, and the first optimal expansion scale. Otherwise, proceed with the following steps: Step a: Based on the change magnitude and the attribution index, determine the impact score and total impact score of each system key parameter. When the impact score of any system key parameter is greater than the second threshold or the total impact score is greater than the third threshold, use the surrogate model to predict the optimal expansion scale based on the changed key parameter values; otherwise, terminate the operation. Step b: If the percentage increase of the predicted optimal expansion scale relative to the sum of the first optimal expansion scale and the current total scale exceeds the fourth threshold, then the second optimal expansion scale is calculated using the spatiotemporal quantification model based on the changes in each key parameter, the current total scale of new energy installed capacity, and the first optimal expansion scale; otherwise, the optimal expansion scale is the second optimal expansion scale.

[0012] Furthermore, the determination of the impact score and total impact score of each system key parameter based on the magnitude of change and the attribution index includes: ; ; in, Indicates key system parameters Impact score; The magnitude of the change; for Attribution index; The overall impact score; This represents the number of key parameters in the system.

[0013] Furthermore, the step of using the surrogate model to predict and determine the optimal expansion scale based on the changed key parameter values ​​includes: Based on the changed values ​​of each key parameter, the predicted value of the total scale after the pre-expansion is obtained by using the surrogate model. If the difference between the predicted total scale after pre-expansion and the first optimal expansion scale exceeds the fifth threshold, the predicted total scale after pre-expansion will be used as the candidate optimal expansion scale; otherwise, the first optimal expansion scale will be determined as the candidate optimal expansion scale.

[0014] Furthermore, the marginal benefit of the new energy installation capacity is calculated using the following formula: ; in, This represents the total system energy consumption and carbon emissions corresponding to the current total scale of new energy installations. This represents the total system energy consumption and carbon emissions corresponding to the total scale after expansion. To expand the scale, that is, to add new energy installed capacity.

[0015] Furthermore, the objective function is expressed as: ; in, This represents the total number of time periods in the scheduling cycle. Indicates a time period; , , , They represent Energy consumption of coal-fired power units, energy consumption of gas-fired power units, carbon emissions of coal-fired power units, and carbon emissions of gas-fired power units during different time periods; , , , These represent the corresponding weighting coefficients.

[0016] The present invention can achieve at least one of the following beneficial effects: By constructing a spatiotemporal quantitative model of new energy installed capacity, the variation law of total system energy consumption and carbon emissions under different pre-expansion scales of new energy installed capacity is quantified, and the optimal expansion scale of new energy installed capacity in the planning area is determined. Furthermore, through global sensitivity analysis and attribution analysis, key parameters in the power grid system parameters are identified, and the contribution attribution index of each key parameter is analyzed and determined. On this basis, the scale of new energy installed capacity and its configuration in different regions and time periods are guided and adjusted, thus providing a technical method with both interpretability and engineering feasibility for regional new energy planning and construction decisions, avoiding the uncertainty of determining the installed capacity scale by relying on experience or a single indicator in existing technologies.

[0017] By calculating the total system energy consumption and carbon emissions corresponding to the total installed capacity of new energy after each pre-expansion, the reduction in the total system energy consumption and carbon emissions corresponding to the total installed capacity of new energy after each pre-expansion is obtained, which is the marginal benefit of new energy installation, providing a clear quantitative basis for new energy planning.

[0018] By employing a global sensitivity analysis method, we can initially screen the system parameters that affect the optimal scale of new energy expansion, identify key parameters that have a significant impact on the optimal scale of new energy expansion, thereby reducing the parameter dimensionality of subsequent analysis and laying the foundation for in-depth correlation analysis.

[0019] By employing a refined attribution analysis method, the variance sources of the optimal new energy expansion scale are decomposed, and the contribution of each key parameter and its interaction to the determination of the optimal new energy investment scale is quantitatively characterized. This enables a causal explanation of the relationship between parameter changes and changes in the optimal new energy expansion scale, significantly improving the interpretability and scientific nature of new energy planning decisions.

[0020] By characterizing the response of new energy installed capacity to changes in key system parameters based on attribution analysis results, when grid operating conditions or planning conditions change, a proxy model can be used to quickly judge, adjust or correct the new energy installed capacity based on the magnitude of changes in key parameters and their contribution, reducing the need for frequent repeated scale traversal calculations; and triggering scale re-solution when the parameter changes are large or the planning accuracy requirements are increased, thereby enabling continuous guidance for new energy installed capacity planning under changing grid conditions.

[0021] The method of this invention can be flexibly configured with parameters according to the resource endowment, load characteristics and network structure of different regional power systems. It is applicable to regional power system planning scenarios with different new energy penetration rates and different planning scales, and has good engineering applicability and scalability.

[0022] In this invention, the above-described technical solutions can be combined with each other to achieve more preferred combinations. Other features and advantages of this invention will be set forth in the following description, and some advantages may become apparent from the description or be learned by practicing the invention. The objects and other advantages of this invention can be realized and obtained from what is particularly pointed out in the description and drawings. Attached Figure Description

[0023] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts.

[0024] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0025] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

[0026] One embodiment of the present invention discloses a method for determining the scale and spatiotemporal layout of new energy installed capacity based on contribution attribution, used to determine the expansion scale and spatiotemporal layout of new energy (including wind power and photovoltaic) installed capacity in the power grid system of a region to be planned. Specifically, it includes steps S1-S5.

[0027] S1. Construct a spatiotemporal quantitative model of total energy consumption and carbon emissions of the power grid system. The model includes an objective function that aims to minimize the total energy consumption and carbon emissions of the system and set power grid constraints.

[0028] Specifically, the power grid system in the planned area includes coal-fired power units, gas-fired power units, hydroelectric power units, nuclear power units, new energy units, and energy storage devices. Among them, new energy units include wind turbines and photovoltaic units, and energy storage devices include battery energy storage devices and pumped storage devices.

[0029] The system energy consumption described in this invention mainly refers to the energy consumption of units using non-renewable energy sources, specifically including the energy consumption of coal-fired power units and gas-fired power units. The carbon emissions are mainly generated by the power generation of coal-fired power units and gas-fired power units.

[0030] Specifically, the objective function that aims to minimize the total system energy consumption and carbon emissions is expressed as: ; in, This represents the total number of time periods in the scheduling cycle, such as 8760 time periods throughout the year. Indicates a time period; , , , They represent Energy consumption of coal-fired power units, energy consumption of gas-fired power units, carbon emissions of coal-fired power units, and carbon emissions of gas-fired power units during different time periods; , , , These represent the corresponding weight coefficients, which are known quantities.

[0031] Furthermore, the energy consumption of coal-fired power units and gas-fired power units adopts a piecewise linear model, and the consumption parameters are set differently according to the technical characteristics of the units. The expression is as follows: ; in, express and ; These correspond to coal-fired power and gas-fired power, respectively; They are respectively a collection of coal-fired power units and a collection of gas-fired power units; For coal-fired or gas-fired power units The unit output operating consumption coefficient is a known quantity; For the unit exist The output during a given time period is a variable; For the unit The fixed operating consumption is a known quantity.

[0032] Furthermore, the carbon emissions of the power grid system are mainly generated by coal-fired power units and gas-fired power units. The expressions for the carbon emissions of coal-fired power units or gas-fired power units are as follows: ; in, express and ; These correspond to coal-fired power and gas-fired power, respectively; They are respectively a collection of coal-fired power units and a collection of gas-fired power units; For coal-fired or gas-fired power units The carbon emission coefficient per unit output is a known quantity.

[0033] Specifically, the set grid constraints include node power constraints, wind and solar power output constraints, coal and gas power unit output constraints, hydropower unit output constraints, nuclear power unit output constraints, and energy storage device power constraints.

[0034] The node power constraint is expressed as: ; in, These correspond to coal-fired power and gas-fired power, respectively; Indicates an energy storage device. These correspond to pumped hydro storage and battery energy storage, respectively. These correspond to wind power, photovoltaic power, hydropower, and nuclear power, respectively. Let t represent the output of wind power / solar power / hydropower / nuclear power. The upper and lower limits of the output of wind power, solar power, and hydropower are known quantities, while the output of wind power, solar power, and hydropower are variables, and the output of nuclear power is a known quantity. Let be the discharge power of the energy storage device during time period t, and be a variable; Let be the charging power of the energy storage device during time period t, and be a variable. The line power flowing into the node is a known quantity, obtained based on historical data; The line power flowing out of the node is a known quantity, obtained based on historical data; The node load power is a known quantity, obtained based on historical data.

[0035] The output constraints for wind power and solar power are expressed as follows: ; in, , representing wind power and photovoltaic power respectively; The minimum output coefficient for wind / solar power is a known quantity. The existing wind power / solar power installed capacity is a known quantity; The wind power / solar power output coefficient for time period t (such as the output ratio corresponding to wind speed and solar intensity) is a known quantity determined based on historical data. The predicted output of wind / solar power during period t.

[0036] The output constraints of coal-fired and gas-fired power units are expressed as follows: ; ; ; in, This is a flag indicating the start / stop status of the generator unit; it is a variable. 0 means stop, 1 means start; , For the unit The minimum and maximum output forces are known quantities; , For the unit The downward and upward climbing rates are known quantities; For the unit The rated installed capacity is a known quantity; Let be the output of unit i during time period t, and be a variable.

[0037] The output constraint of the hydropower unit is expressed as: ; ; in, The minimum output of the hydropower unit is a known quantity. Let be the output of the hydropower unit during time period t, and be a variable. The maximum output of the hydropower unit during time period t is a known quantity determined based on historical data. The existing installed hydropower capacity is a known quantity. Let be the hydropower energy consumption coefficient for time period t, which is a known quantity.

[0038] The power output constraint of a nuclear power unit is expressed as: ; in, The energy consumption coefficient of the nuclear power unit is a known quantity (usually determined based on the actual operation of the power grid). The existing installed nuclear power capacity is a known quantity. Let be the output of the nuclear power unit during time period t, and be a known quantity.

[0039] The power constraint of energy storage devices is expressed as: ; ; ; ; ; in, , Let be the discharge and charging power energy consumption coefficients of the energy storage device during time period t, respectively, which are known quantities; The existing installed capacity of the energy storage device is a known quantity. , Let be the discharge and charging efficiencies of the energy storage device, respectively, which are known quantities; Let MWh be the state of charge of the energy storage device during time period t, which is the total amount of electrical energy currently stored in the energy storage device (MWh), and it is a variable. This represents the maximum discharge duration of the energy storage device.

[0040] This embodiment constructs a spatiotemporal quantitative model of new energy installed capacity, which is used to quantify the variation law of the total system energy consumption and carbon emissions under different new energy installed capacity pre-expansion scales, laying a theoretical foundation for determining the optimal expansion scale of new energy installed capacity in the planning area.

[0041] S2. Based on the given power grid system parameter values ​​and the current total scale of new energy installed capacity, the first optimal expansion scale of new energy installed capacity is determined using the spatiotemporal quantification model.

[0042] The power grid system parameters refer to the parameters defined as known quantities and variables in step S1. The given values ​​of the power grid parameters refer to the specific values ​​of the known quantities determined based on the current construction and operation of the power grid system. Among them, parameters such as node load, wind power / solar power output coefficient, and upper and lower limits of wind power and solar power output are in time series form.

[0043] Step S2 specifically includes S21-S23.

[0044] S21. Based on the given values ​​of the power grid system parameters, solve the spatiotemporal quantization model based on the current total scale of new energy installed capacity and multiple pre-expansion scales respectively, to obtain the sum of system energy consumption and carbon emissions corresponding to the current total scale and the total scale of new energy installed capacity after each pre-expansion.

[0045] S22. Calculate the marginal benefit of new energy installations corresponding to each pre-expansion scale based on the sum of the energy consumption and carbon emissions of each system; the marginal benefit of new energy installations refers to the reduction in the total energy consumption and carbon emissions of the system brought about by a unit expansion scale. S23. The pre-expansion scale corresponding to the maximum marginal benefit of new energy installed capacity is taken as the first optimal expansion scale. The first optimal expansion scale is the scale of expansion of the power grid system based on the current total scale.

[0046] In step S21, the spatiotemporal quantization model is solved based on the current total scale of renewable energy installed capacity in the power grid system and the given values ​​of power grid system parameters. This yields the optimal operating results of the power system corresponding to the current total scale throughout the entire time period, including the status and output of thermal power units, gas power units, energy storage devices, hydropower units, wind power, and photovoltaic units in each time period. The total system energy consumption and carbon emissions are also calculated. It should be noted that during the solution process, the upper and lower limits of wind power and photovoltaic unit output are known quantities determined through prediction, limiting the range of wind power and photovoltaic unit output. The solution yields the actual grid-connected output of wind power and photovoltaic units in each time period.

[0047] Furthermore, in step S21, multiple pre-expansion scales are determined by a segmented incremental approach. Based on the total installed capacity of new energy after the pre-expansion, and given the current grid system parameter values, the optimal operating results of the grid system corresponding to each total installed capacity of new energy after the pre-expansion, as well as the total system energy consumption and carbon emissions, are obtained.

[0048] It should be noted that, since the spatial distribution of wind power and photovoltaic units corresponding to the various pre-expansion scales in the planned area is different, the optimal operating results of the corresponding power system obtained from the solution will also be different throughout the time period due to the different spatial distribution. The status and output of each unit, including wind power and photovoltaic units, are different at different times, that is, they correspond to different spatiotemporal layouts.

[0049] In step S22, the marginal benefit of the new energy installation is calculated using the following formula: ; in, This represents the total system energy consumption and carbon emissions corresponding to the current total scale of new energy installations. This represents the total system energy consumption and carbon emissions corresponding to the total scale after expansion. To expand the scale, that is, to add new energy installed capacity.

[0050] In step S23, the pre-expansion scale corresponding to the maximum marginal benefit of new energy installations is taken as the first optimal expansion scale, denoted as... .

[0051] In this embodiment, by calculating the total system energy consumption and carbon emissions corresponding to the total scale of new energy installed capacity after each pre-expansion, the reduction in the total system energy consumption and carbon emissions corresponding to the total scale of new energy installed capacity after each pre-expansion is obtained, which is the marginal benefit of new energy installed capacity, providing a clear quantitative basis for new energy planning.

[0052] S3. Determine several key parameters in the power grid system parameters based on the spatiotemporal quantification model of total energy consumption and carbon emissions. Specifically, this includes S31-S35.

[0053] S31. Determine the range of values ​​for each parameter of the power grid system.

[0054] Specifically, reasonable ranges of values ​​for power grid system parameters are determined based on historical statistical data or engineering experience.

[0055] S32. Based on the value range of each parameter, randomly generate multiple sets of parameter value combinations for the power grid system.

[0056] Specifically, multiple sets of parameter value combinations are generated using a random sampling method. Each parameter changes independently during the sampling process, without pre-setting any specific correlation between the parameters.

[0057] S33. Based on each of the aforementioned value combinations and the current total scale of new energy installed capacity, the optimal total scale of new energy installed capacity after pre-expansion is obtained using the spatiotemporal quantization model. See step S2 for details.

[0058] S34. Based on the combinations of values ​​and the corresponding optimal total scale of new energy installed capacity after pre-expansion, the global sensitivity analysis method is used to obtain the influence effect of each system parameter on the optimal total scale of new energy installed capacity after pre-expansion.

[0059] Specifically, based on each of the aforementioned value combinations and the corresponding optimal expansion scale of new energy installed capacity, a mapping relationship is constructed between the system parameter vector and the optimal total scale after the pre-expansion of new energy installed capacity, expressed as: ; in, This represents the optimal total scale of new energy installed capacity after pre-expansion corresponding to a set of the aforementioned value combinations. This represents a vector of system parameters that influence the optimal expansion scale of new energy installations, determined by the selected system parameters. ( The selected system parameters include at least the wind / solar power output factor, load level parameters (obtained based on the load curve, which is the change range between the currently selected load curve and the specified load curve), the installed capacity and related parameters of coal-fired and gas-fired power units of different technical levels, and the relevant parameters of energy storage devices.

[0060] Furthermore, regarding the parameter vector The parameters were perturbed, and the Morris global sensitivity analysis method was used to analyze the parameters. The basic effects are calculated using the following method: ; in, For parameters The basic effect; For the first One system parameter; The parameter perturbation step size.

[0061] Furthermore, parameters are calculated for multiple random sampling trajectories based on the aforementioned basic effect. Average absolute basic effect: ; in, Indicates parameters The average absolute fundamental effect, i.e., the parameter The impact on the optimal expansion scale of new energy installed capacity.

[0062] S35. Determine multiple key system parameters according to the impact effect from largest to smallest.

[0063] Specifically, the number of key system parameters is determined based on the planning requirements of the power grid system or empirical values.

[0064] In this embodiment, a global sensitivity analysis method is used to initially screen the system parameters that affect the optimal scale of new energy expansion, identify the key parameters that have a significant impact on the optimal scale of new energy expansion, thereby reducing the parameter dimensionality of subsequent analysis and laying the foundation for in-depth correlation analysis.

[0065] S4. Perform contribution attribution analysis on multiple key parameters to determine the attribution index for each key parameter. Specifically, this includes: S41. Construct a proxy model to predict the optimal total scale after the pre-expansion of new energy installed capacity, expressed as: ;in, A set of key parameters. The total installed capacity of new energy sources predicted by the surrogate model is expressed in MW. For example, the surrogate model can be a Gaussian regression model or a DNN neural network, etc.

[0066] S42. Based on each of the aforementioned value combinations and the corresponding optimal total scale after the pre-expansion of new energy installed capacity, the proxy model is trained to obtain the trained proxy model.

[0067] Specifically, the proxy model is trained by minimizing the following loss function: ; in, For loss function, For the first The optimal total scale of new energy installed capacity after pre-expansion corresponds to the aforementioned combination of values. For the first The model predicts the total installed capacity corresponding to each of the given value combinations; N is the number of samples.

[0068] S43. The attribution index of each key system parameter is calculated using the Sobol global variance decomposition method based on the trained agent model.

[0069] Specifically, the population variance of the Sobol global variance decomposition method is expressed as: ; in, This represents the overall variance of the proxy model; Indicates key parameters The independent contribution variance; Indicates key parameters and The contribution variance of pairwise parameter interactions; express The contribution variance of the interaction of key parameters.

[0070] Furthermore, key parameters Attribution index Defined as: ;in, Indicates non .

[0071] In this embodiment, by employing a refined attribution analysis method, the variance sources of the optimal new energy expansion scale are decomposed, and the contribution of each key parameter and its interaction to the determination of the optimal new energy investment scale is quantitatively characterized. This enables the explanation of the causal relationship between parameter changes and changes in the optimal new energy expansion scale, significantly improving the interpretability and scientific nature of new energy planning decisions.

[0072] S5. When the grid conditions, i.e. the key parameter values, change, the second optimal expansion scale of new energy installations is determined based on the change values ​​of the key parameters, the attribution index, and the first optimal expansion scale.

[0073] In practical applications, the scale of the power grid is continuously developing, and correspondingly, the construction of new energy sources will also be dynamic and continuous. When expanding the installed capacity of new energy sources for the first time, the first optimal expansion scale is obtained through steps S1 and S2. After the actual expansion of new energy installed capacity is carried out according to the first optimal expansion scale, when the grid conditions (values ​​of various key parameters) change, it is necessary to determine the scale of further expansion. At this time, the total scale of new energy sources in the power grid system is the sum of the current total scale and the first optimal expansion scale.

[0074] Step S5 specifically includes: Calculate the normalized change magnitude of each key parameter. If it exceeds the first threshold, calculate the second optimal expansion scale using the spatiotemporal quantification model based on the change magnitude of each key parameter, the current total scale of new energy installations, and the first optimal expansion scale. Otherwise, proceed with the following steps: Step a: Based on the change magnitude and the attribution index, determine the impact score and total impact score of each system key parameter. When the impact score of a system key parameter is greater than the second threshold or the total impact score is greater than the third threshold, use the surrogate model to predict the optimal expansion scale based on the changed key parameter values; otherwise, terminate the operation. Step b: If the percentage increase of the predicted optimal expansion scale relative to the sum of the first optimal expansion scale and the current total scale exceeds the fourth threshold, then the second optimal expansion scale is calculated using the spatiotemporal quantification model based on the changes in each key parameter, the current total scale of new energy installed capacity, and the first optimal expansion scale; otherwise, the optimal expansion scale is the second optimal expansion scale.

[0075] Furthermore, the key parameters are calculated using the following formula. Normalized variation of change values : ; in, and Key parameters The maximum and minimum values ​​of the range are determined based on historical statistical data or engineering experience; and They are respectively The values ​​before and after the change. Preferably, the range of the first threshold corresponding to the normalized change magnitude is: .

[0076] When the normalization change exceeds the first threshold, it indicates that the power grid conditions have changed significantly. Therefore, it is necessary to reuse the spatiotemporal quantization model, taking the changed values ​​of each key parameter and the current total installed capacity of new energy (the sum of the current total capacity and the first optimal expansion capacity) as input, and use the spatiotemporal quantization model to solve for the second optimal expansion capacity and the corresponding optimal operating results of the power system throughout the entire time period.

[0077] Furthermore, in step a, determining the impact score and total impact score of each system key parameter based on the magnitude of change and the attribution index includes: ; ; in, Indicates key system parameters Impact score; The magnitude of the change; for Attribution index; The overall impact score; This represents the number of key parameters in the system.

[0078] Preferably, the range of values ​​for the second threshold corresponding to the impact of a single system key parameter on the score is as follows: The range of values ​​for the third threshold corresponding to the total impact score is: .

[0079] When the impact score of any key system parameter exceeds the second threshold or the total impact score exceeds the third threshold, it indicates a significant change in grid conditions. To conserve computing power and improve efficiency, the surrogate model can be used to predict the optimal expansion scale based on the changed key parameter values. Otherwise, the change in grid conditions is minor and can be ignored; no expansion planning is needed, and the operation should be terminated. It should be noted that in practical applications, for cases where grid conditions change only slightly, the scale of new energy installed capacity can be fine-tuned, i.e., a very small expansion can be carried out based on engineering experience.

[0080] Furthermore, determining the optimal expansion scale using the proxy model includes: Based on the changes in key system parameters, the surrogate model is used to predict the total scale, which is expressed as follows: Specifically, the changed values ​​of the key system parameters are calculated based on the changed values ​​and the original values ​​of the key system parameters. The changed values ​​of the key system parameters are then used as given inputs to predict the total scale using the surrogate model. If the difference between the predicted total capacity and the current total installed capacity of new energy in the power grid system (the sum of the current total capacity and the first optimal expansion capacity) exceeds the fifth threshold (preferably within 1% of the current total installed capacity of new energy in the power grid system), then... If the expansion rate is 5%, then the difference is taken as the candidate optimal expansion scale; otherwise, the candidate optimal expansion scale is zero. The candidate optimal expansion scale is expressed as... .

[0081] Furthermore, in step b, the percentage increase of the predicted candidate optimal expansion size relative to the sum of the first optimal expansion size and the current total size is calculated. If this increase exceeds a fourth threshold (preferably 5%), then... If the change value of each key parameter is 15%, then the value of each key parameter is determined based on the change value of each key parameter, and used as input based on the sum of the first optimal expansion scale of new energy installed capacity and the current total scale, and the second optimal expansion scale is calculated using the spatiotemporal quantification model.

[0082] In step S5, while solving for the second optimal expansion scale, the optimal operating result of the power grid system corresponding to the second solution scale is obtained over the entire time period. Based on the second optimal expansion scale and the optimal operating result of the power grid system over the entire time period, the final new energy installed capacity scale and spatiotemporal layout are determined.

[0083] Furthermore, in subsequent planning after implementing the second optimal expansion scale, the approach in step S5 can still be referenced to provide continuous dynamic guidance for the expansion of the power grid system, which will not be elaborated further here.

[0084] In this embodiment, by characterizing the response pattern of new energy installed capacity to changes in key system parameters based on attribution analysis results, when the grid operating conditions or planning conditions change, the new energy installed capacity can be judged, adjusted or corrected according to the magnitude of change of key parameters and their contribution. This reduces the need for frequent and repeated scale traversal calculations, reduces the amount of computation, and improves the planning efficiency for the expansion of new energy installed capacity. Furthermore, it can trigger a re-solution of the scale when the parameter change is large or the planning accuracy requirements are increased, thereby enabling continuous guidance for the planning of new energy installed capacity under changing grid conditions.

[0085] The method in this embodiment can be flexibly configured with parameters according to the resource endowment, load characteristics and network structure of different regional power systems. It is applicable to regional power system planning scenarios with different new energy penetration rates and different planning scales, and has good engineering applicability and scalability.

[0086] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for determining the scale and spatiotemporal layout of new energy installed capacity based on contribution attribution, characterized in that, Includes the following steps: S1. Construct a spatiotemporal quantitative model of total energy consumption and carbon emissions of the power grid system. The model includes an objective function that aims to minimize the total energy consumption and carbon emissions of the system and set power grid constraints. S2. Based on the given values ​​of power grid system parameters and the current total scale of new energy installed capacity, the first optimal expansion scale of new energy installed capacity is determined using the spatiotemporal quantification model. S3. Determine several key parameters in the power grid system parameters based on the spatiotemporal quantification model of total energy consumption and carbon emissions; S4. Perform contribution attribution analysis on multiple key parameters to determine the attribution index of each key parameter; S5. When the grid conditions, i.e. the key parameter values, change, the second optimal expansion scale of new energy installations is determined based on the change values ​​of the key parameters, the attribution index, and the first optimal expansion scale.

2. The method for determining the scale and spatiotemporal layout of new energy installed capacity according to claim 1, characterized in that, The determination of the first optimal expansion scale of new energy installations using the spatiotemporal quantification model based on given power grid system parameter values ​​and the current total scale of new energy installations includes: Based on the given values ​​of the power grid system parameters, the spatiotemporal quantization model is solved based on the current total scale of new energy installed capacity and multiple pre-expansion scales, respectively, to obtain the sum of system energy consumption and carbon emissions corresponding to the current total scale and the total scale of new energy installed capacity after each pre-expansion. The marginal benefit of new energy installations corresponding to each pre-expansion scale is calculated based on the sum of the energy consumption and carbon emissions of each system. The marginal benefit of new energy installations refers to the reduction in the total energy consumption and carbon emissions of the system brought about by a unit expansion scale. The pre-expansion scale corresponding to the maximum marginal benefit of new energy installations is taken as the first optimal expansion scale.

3. The method for determining the scale and spatiotemporal layout of new energy installed capacity according to claim 2, characterized in that, The key system parameters determined based on the spatiotemporal quantification model of total energy consumption and carbon emissions include: Determine the value range of each parameter of the power grid system; Multiple sets of parameter value combinations for the power grid system are randomly generated based on the value range of each parameter. Based on each of the aforementioned value combinations and the current total scale of new energy installations, the optimal total scale of new energy installations after pre-expansion is obtained using the aforementioned spatiotemporal quantization model for each of the aforementioned value combinations. Based on the combinations of values ​​and the corresponding optimal total scale after the pre-expansion of new energy installed capacity, the global sensitivity analysis method is used to obtain the influence effect of each system parameter on the scale of new energy installed capacity expansion. Multiple key system parameters were determined according to the described impact effects, from largest to smallest.

4. The method for determining the scale and spatiotemporal layout of new energy installed capacity according to claim 3, characterized in that, The contribution attribution analysis of the multiple key parameters, and the determination of the attribution index for each key parameter, includes: Construct a proxy model to predict the optimal total scale of new energy installed capacity after pre-expansion; The proxy model is trained based on each of the aforementioned value combinations and the corresponding optimal total scale after the pre-expansion of new energy installed capacity. The attribution indices of each key parameter are calculated using the Sobol global variance decomposition method based on the trained surrogate model.

5. The method for determining the scale and spatiotemporal layout of new energy installed capacity according to claim 4, characterized in that, The proxy model is trained by minimizing the following loss function: ； in, For loss function, This represents the optimal total scale of new energy installed capacity after pre-expansion for the j-th combination of values. The model predicts the total installed capacity for the j-th combination of values; N is the number of samples.

6. The method for determining the scale and spatiotemporal layout of new energy installed capacity according to claim 3, characterized in that, The determination of the second optimal expansion scale for new energy installations based on the changes in key system parameters, the attribution index, and the first optimal expansion scale includes: Calculate the normalized change magnitude of each key parameter. If it exceeds the first threshold, calculate the second optimal expansion scale using the spatiotemporal quantification model based on the change magnitude of each key parameter, the current total scale of new energy installations, and the first optimal expansion scale. Otherwise, proceed with the following steps: Step a: Based on the change magnitude and the attribution index, determine the impact score and total impact score of each system key parameter. When the impact score of any system key parameter is greater than the second threshold or the total impact score is greater than the third threshold, use the surrogate model to predict the optimal expansion scale based on the changed key parameter values; otherwise, terminate the operation. Step b: If the percentage increase of the predicted optimal expansion scale relative to the sum of the first optimal expansion scale and the current total scale exceeds the fourth threshold, then the second optimal expansion scale is calculated using the spatiotemporal quantification model based on the changes in each key parameter, the current total scale of new energy installed capacity, and the first optimal expansion scale; otherwise, the optimal expansion scale is the second optimal expansion scale.

7. The method for determining the scale and spatiotemporal layout of new energy installed capacity according to claim 6, characterized in that, The determination of the impact scores and total impact scores for each system's key parameters based on the magnitude of change and the attribution index includes: ； ； in, Indicates key system parameters Impact score; The magnitude of the change; for Attribution index; The overall impact score; This represents the number of key parameters in the system.

8. The method for determining the scale and spatiotemporal layout of new energy installed capacity according to claim 7, characterized in that, The process of using the proxy model to predict and determine the optimal expansion scale based on the changed key parameter values ​​includes: Based on the changed values ​​of each key parameter, the predicted value of the total scale after the pre-expansion is obtained by using the surrogate model. If the difference between the predicted total scale after pre-expansion and the first optimal expansion scale exceeds the fifth threshold, the predicted total scale after pre-expansion will be used as the candidate optimal expansion scale; otherwise, the first optimal expansion scale will be determined as the candidate optimal expansion scale.

9. The method for determining the scale and spatiotemporal layout of new energy installed capacity according to claim 2, characterized in that, The marginal benefit of the new energy installation capacity is calculated using the following formula: ； in, This represents the total system energy consumption and carbon emissions corresponding to the current total scale of new energy installations. This represents the total system energy consumption and carbon emissions corresponding to the total scale after expansion. To expand the scale, that is, to add new energy installed capacity.

10. The method for determining the installed capacity and spatiotemporal layout of new energy sources according to any one of claims 1-9, characterized in that, The objective function is expressed as: ； in, This represents the total number of time periods in the scheduling cycle. Indicates a time period; , , , They represent Energy consumption of coal-fired power units, energy consumption of gas-fired power units, carbon emissions of coal-fired power units, and carbon emissions of gas-fired power units during different time periods; , , , These represent the corresponding weighting coefficients.

Images