New energy consumption multi-target power system source load interaction optimization scheduling method

Abstract

The invention relates to the technical field of power systems. The multi-target power system source-load interaction optimization scheduling method for new energy consumption is provided, and the method comprises the steps: carrying out the analysis processing of inter-region new energy output space-time complementation characteristics in a multi-space-time scale input data set through a preset standardized coordination interface unit, and generating a consumption complementation factor; performing dynamic correction processing on the local scheduling instruction based on the consumption complementary factor and the source-load interaction constraint condition, and generating a cross-regional power mutual aid plan; a preset improved multi-objective evolutionary algorithm is adopted to solve the four-dimensional optimization target model, and a Pareto frontier solution set is generated; and based on the Pareto frontier solution set and the cross-regional power mutual aid plan, processing is carried out through a preset multi-target balance decision-making unit, and an execution scheme is generated, so that the multi-target collaborative optimization capability is improved, the flexible load regulation potential is released, and the robustness of the cross-regional mutual aid plan is enhanced.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a multi-objective power system source-load interaction optimization scheduling method for renewable energy consumption. Background Technology

[0002] With the ongoing transformation of the energy structure, the integration of a high proportion of renewable energy into the power system has become a core path to achieving the "dual carbon" goals. Against this backdrop, multi-objective collaborative optimization scheduling technology has become a key means to enhance the capacity for renewable energy absorption and ensure the safe and economical operation of the power grid.

[0003] Traditional renewable energy dispatching technologies use static weighting or priority ranking to handle multi-objective optimization problems, but they cannot effectively characterize the dynamic conflict relationships between targets such as absorption rate, cost, emissions, and stability, leading to the failure of target coordination. The modeling of source-load interaction constraints ignores the capacity-delay correlation characteristics and time-period elasticity of flexible loads, resulting in the load adjustment potential not being fully explored. Cross-regional dispatching coordination relies on fixed transmission quota strategies, and the lack of quantification of the spatiotemporal complementarity characteristics of renewable energy and the probability of absorption gap transmission leads to insufficient robustness of mutual assistance plans. Summary of the Invention

[0004] Based on this, it is necessary to provide a multi-objective power system source-load interaction optimization scheduling method to address the above-mentioned technical problems, so as to improve the multi-objective collaborative optimization capability, release the potential of flexible load regulation, and enhance the robustness of inter-regional mutual assistance plans.

[0005] Firstly, this application provides a multi-objective power system source-load interaction optimization scheduling method for renewable energy consumption, which includes:

[0006] The system synchronously integrates and processes the new energy output forecast data collected from the preset new energy power stations, the load demand data collected from the load monitoring system, the power grid topology data stored in the power grid dispatch center, the unit operation parameters collected from the unit monitoring system, and the cross-regional power transmission channel boundary data provided by the cross-regional coordination platform to generate a multi-temporal and spatial scale input dataset.

[0007] Based on a multi-temporal and spatial scale input dataset, joint modeling is performed on the targets of new energy absorption rate, system operating cost, pollution emission, and operational stability to generate a four-dimensional optimized target model.

[0008] Based on the four-dimensional optimization objective model, collaborative constraint modeling is performed on the source-side conventional unit group operation parameters, new energy cluster output characteristics and load-side flexible load resource regulation characteristics in the multi-temporal and spatial scale input dataset to generate source-load interaction constraint conditions.

[0009] The spatiotemporal complementarity characteristics of interregional renewable energy output in the multi-temporal scale input dataset are analyzed and processed by the preset standardized coordination interface unit to generate the absorption complementarity factor; based on the absorption complementarity factor and source-load interaction constraints, the local scheduling instructions are dynamically modified to generate the cross-regional power mutual assistance plan.

[0010] A pre-defined improved multi-objective evolutionary algorithm is used to solve the four-dimensional optimization objective model to generate a Pareto front solution set. Based on the Pareto front solution set and the cross-regional power mutual assistance plan, an execution scheme is generated through a pre-defined multi-objective equilibrium decision unit.

[0011] In one embodiment, based on the Pareto front solution set and the cross-regional power mutual assistance plan, an execution plan is generated through a preset multi-objective equilibrium decision unit, including:

[0012] Robust gap analysis is performed on the Pareto front solution set to generate a perturbation-resistant solution set subspace;

[0013] By embedding channel capacity constraints into the solution space of the cross-regional power mutual assistance plan to counteract disturbances, a feasible scheduling solution set is generated.

[0014] The following formula is used to perform a three-axis projection process on the feasible scheduling solution set, considering economic efficiency, environmental friendliness, and stability, to generate a multi-dimensional decision feature vector:

[0015]

[0016] in, Represents a multidimensional decision feature vector. Indicates the index of the feasible scheduling solution. Indicates the first A feasible scheduling solution This represents the value of the economic objective function. This represents the mean of the economic objective function. The standard deviation represents the economic objective function. This represents the value of the environmental protection objective function. This represents the mean of the environmental protection objective function. The standard deviation of the environmental protection objective function is represented by... This represents the value of the stability objective function. The mean of the stability objective function is represented by . The standard deviation represents the stability objective function;

[0017] Based on preset dynamic preference weights, the multidimensional decision feature vectors are subjected to non-dominated sorting to generate a balanced optimization sequence.

[0018] The balanced optimization sequence is processed by instruction reorganization through a preset spatiotemporal consistency fusion unit to generate an execution plan.

[0019] In one embodiment, a feasible scheduling solution set is generated by embedding channel capacity constraints into the cross-regional power mutual assistance plan's anti-disturbance solution set subspace, including:

[0020] Multidimensional feature analysis of transmission channels is performed on the inter-regional power mutual assistance plan to generate a spatiotemporal capacity constraint tensor.

[0021] The spatiotemporal capacity constraint tensor is mapped to the perturbation-resistant solution set subspace for dimension alignment, generating an extended constraint solution set.

[0022] By using a preset margin adaptive scaling unit, the extended constraint solution set is elastically adjusted to ensure safety boundaries, thereby generating a reinforced boundary solution set.

[0023] Based on the probability distribution of new energy fluctuations, the enhanced boundary solution set is subjected to limit-crossing risk filtering to generate a risk-immune solution set.

[0024] Constraint satisfaction verification is performed on the risk-immune solution set to generate a feasible scheduling solution set.

[0025] In one embodiment, the spatiotemporal complementarity characteristics of inter-regional new energy output in a multi-temporal-scale input dataset are analyzed and processed through a preset standardized coordination interface unit to generate a complementary absorption factor, including:

[0026] Time offset correlation analysis was performed on the spatiotemporal complementarity characteristics of new energy output between regions to generate a cross-regional time complementarity coefficient matrix.

[0027] Spatial fluctuation covariance calculation is performed on the spatiotemporal complementarity characteristics of new energy output between regions to generate a cross-regional spatial complementarity covariance tensor.

[0028] Based on the cross-regional time complementary coefficient matrix, the probability mapping of the absorption gap is processed to generate a time-dimensional absorption potential feature vector.

[0029] Power fluctuation smoothness assessment is performed based on cross-regional spatial complementary covariance tensor to generate spatial dimension absorption potential feature vector;

[0030] Using the following formula, the time-dimensional absorption potential feature vector and the spatial-dimensional absorption potential feature vector are input into a preset dynamic coupling unit for feature cross-entropy fusion processing to generate an absorption complementarity factor:

[0031]

[0032] in, Indicates the absorption of complementary factors. This represents a feature vector representing the time-dimensional absorption potential. The feature vector representing the spatial dimension absorption potential. The information entropy function is represented by... express Divergence.

[0033] In one embodiment, a time-dimensional absorption potential feature vector is generated by performing probability mapping processing on the cross-regional time complementarity coefficient matrix to absorb the absorption gap, including:

[0034] The cross-regional time complementarity coefficient matrix is ​​processed by inter-regional absorption gap correlation analysis to generate a gap transmission topology network.

[0035] Based on the probability of grid congestion, the path reliability weighting processing of the gap propagation topology network is performed to generate a probability propagation channel set.

[0036] By performing gap propagation simulation processing on the probability propagation channel set through preset time series Monte Carlo units, a spatiotemporal propagation probability distribution is generated.

[0037] The spatiotemporal transmission probability distribution is enhanced with fluctuation sensitivity processing to generate a robust transmission feature layer.

[0038] The robust transfer feature layer is input into a preset feature compression unit for dimensional reduction and regularization to generate a time-dimensional absorption potential feature vector.

[0039] In one embodiment, based on a four-dimensional optimization objective model, collaborative constraint modeling is performed on the source-side conventional unit group operating parameters, new energy cluster output characteristics, and load-side flexible load resource adjustment characteristics in the multi-temporal and spatiotemporal scale input dataset to generate source-load interaction constraint conditions, including:

[0040] The start-stop combination and ramp rate are coupled to process the operating parameters of the conventional unit group on the source side to generate a set of unit operating constraints.

[0041] Based on the output characteristics of new energy clusters, a set of constraints for new energy consumption is generated by mapping the curtailment penalty coefficient.

[0042] The interruptible load component in the load-side flexible load resource regulation characteristics is subjected to capacity-delay correlation processing to generate a load reduction constraint set.

[0043] The shiftable load component in the load-side flexible load resource adjustment characteristics is subjected to time-period elastic closure processing to generate a load transfer constraint set.

[0044] The power balance and regulation rate matching processes are applied to the set of constraints for unit operation, renewable energy consumption, load reduction, and load transfer to generate source-load interaction constraints.

[0045] In one embodiment, based on a multi-temporal-scale input dataset, the targets for renewable energy absorption rate, system operating cost, pollution emission, and operational stability are jointly modeled to generate a four-dimensional optimization target model, including:

[0046] The output deviation is integrally processed to generate the new energy consumption rate target to generate the consumption rate optimization function.

[0047] The system operating cost target is coupled with fuel cost and start-stop loss to generate an operating cost control function;

[0048] The pollution emission target is converted to carbon and sulfur equivalent to generate a pollution emission suppression function.

[0049] The operational stability target is constrained by voltage fluctuation rate and line load rate to generate a stability enhancement function.

[0050] The coupling strength of the absorption rate optimization function, operating cost control function, pollution emission suppression function and stability enhancement function is evaluated by time-varying correlation, and a target correlation matrix is ​​generated.

[0051] Based on the target correlation matrix, adaptive weight allocation is performed to generate a dynamic set of weight coefficients.

[0052] The following formula is used to nonlinearly superimpose the absorption rate optimization function, operating cost control function, pollution emission suppression function, stability enhancement function, and dynamic weight coefficient set to generate a four-dimensional optimization target model:

[0053]

[0054] in, This represents a four-dimensional optimization objective model. Indicates the first The dynamic weighting coefficients of each objective function. Indicates the first The value of the objective function, Represents the nonlinear coupling strength coefficient. This represents the hyperbolic tangent activation function. This represents the scaling factor of the hyperbolic tangent function.

[0055] Secondly, this application also provides a multi-objective power system source-load interaction optimization dispatching system for renewable energy consumption, which includes:

[0056] The multi-source data integration module is used to synchronously integrate and process the new energy output forecast data collected by the preset new energy power plants, the load demand data collected by the load monitoring system, the power grid topology data stored in the power grid dispatch center, the unit operation parameters collected by the unit monitoring system, and the cross-regional power transmission channel boundary data provided by the cross-regional coordination platform to generate a multi-temporal and spatial scale input dataset.

[0057] The four-dimensional target modeling module is used to jointly model and process the new energy consumption rate target, system operating cost target, pollution emission target and operation stability target based on multi-temporal and spatiotemporal scale input datasets, and generate a four-dimensional optimized target model.

[0058] The source-load constraint modeling module is used to perform collaborative constraint modeling on the source-side conventional unit group operation parameters, new energy cluster output characteristics and load-side flexible load resource adjustment characteristics in multi-temporal and spatial scale input datasets based on a four-dimensional optimization objective model, and generate source-load interaction constraint conditions.

[0059] The cross-regional coordination and optimization module is used to analyze the spatiotemporal complementarity characteristics of inter-regional new energy output in the multi-temporal and spatiotemporal scale input dataset through a preset standardized coordination interface unit, and generate a complementary absorption factor; based on the complementary absorption factor and source-load interaction constraints, the module dynamically corrects the local scheduling instructions and generates a cross-regional power mutual assistance plan.

[0060] The multi-objective decision module is used to solve the four-dimensional optimization objective model using a preset improved multi-objective evolutionary algorithm to generate a Pareto front solution set; based on the Pareto front solution set and the cross-regional power mutual assistance plan, it is processed by a preset multi-objective equilibrium decision unit to generate an execution plan.

[0061] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of any of the methods in the first aspect of this application.

[0062] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any of the methods in the first aspect of this application.

[0063] The multi-objective power system source-load interaction optimization scheduling method for renewable energy consumption provided in this application synchronously integrates renewable energy output forecasting, load demand, grid topology, unit operating parameters, and cross-regional transmission channel boundary data to generate a multi-temporal and spatial scale input dataset. Based on this dataset, a four-dimensional optimization objective model is generated by jointly modeling four objectives: renewable energy consumption rate, system operating cost, pollution emissions, and operational stability. This breaks through the limitations of traditional static weighting or priority ranking, which cannot characterize the dynamic conflicts between objectives, and effectively improves the multi-objective collaborative optimization capability.

[0064] Based on a four-dimensional optimization objective model, collaborative constraint modeling is performed on the operating parameters of conventional generating units on the source side, the output characteristics of renewable energy clusters, and the adjustment characteristics of flexible load resources on the load side to generate source-load interaction constraints. This captures the adjustment characteristics of flexible loads, fully explores their adjustment space, and releases their adjustment potential. A standardized coordination interface unit is used to analyze the spatiotemporal complementarity characteristics of renewable energy output between regions to generate absorption complementarity factors. Combined with source-load interaction constraints, local dispatch instructions are dynamically modified to generate a cross-regional power mutual aid plan. An improved multi-objective evolutionary algorithm is then used to solve the four-dimensional optimization objective model to obtain the Pareto front solution set. An execution scheme is generated by combining the cross-regional power mutual aid plan with a multi-objective equilibrium decision unit, allowing the cross-regional mutual aid plan to fully adapt to the spatiotemporal complementarity characteristics of renewable energy and the transmission law of absorption gaps, significantly enhancing the robustness of the cross-regional mutual aid plan. Attached Figure Description

[0065] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0066] Figure 1 This is a flowchart of a multi-objective power system source-load interaction optimization scheduling method for renewable energy consumption in one embodiment of the present invention;

[0067] Figure 2 This is a flowchart of a method for generating a feasible scheduling solution set by embedding channel capacity constraints into the solution set subspace of the cross-regional power mutual assistance plan to combat disturbances in one embodiment of the present invention.

[0068] Figure 3 This is a structural diagram of a multi-objective power system source-load interaction optimization scheduling system for renewable energy consumption, as described in one embodiment of the present invention. Detailed Implementation

[0069] To make the above-mentioned objects, features, and advantages of this application more apparent and understandable, the specific embodiments of this application 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 this application. However, this application can be implemented 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 application. Therefore, this application is not limited to the specific embodiments disclosed below.

[0070] First, the application scenarios of the embodiments of this application are described. In the embodiments of this application, a multi-objective power system source-load interaction optimization scheduling method for renewable energy consumption is provided, applicable to scenarios such as daily dispatching of regional power grids with high proportions of renewable energy access, cross-regional renewable energy consumption coordination, and power system operation optimization with flexible loads.

[0071] The illustrative example provided in this application is that the multi-objective power system source-load interaction optimization scheduling method for renewable energy consumption can also be applied to other related application scenarios such as the safe and economical operation and control of the power grid in renewable energy-rich areas and the mutual assistance and allocation of power resources in multiple regions. This is only an example and does not limit the specific application scenarios.

[0072] like Figure 1 As shown, this application provides a multi-objective power system source-load interaction optimization scheduling method for renewable energy consumption, which includes:

[0073] S101: Synchronously integrate and process the new energy output forecast data collected from the preset new energy power plants, the load demand data collected from the load monitoring system, the power grid topology data stored in the power grid dispatch center, the unit operation parameters collected from the unit monitoring system, and the cross-regional power transmission channel boundary data provided by the cross-regional coordination platform to generate a multi-temporal and spatial scale input dataset.

[0074] For example, the timeliness and coverage of the new energy output forecast data collected by the new energy power plant are verified; the monitoring accuracy and data integrity of the load demand data collected by the load monitoring system are confirmed; the storage integrity and update frequency of the grid topology data stored in the power grid dispatch center are checked; the parameter validity and collection frequency of the unit operation parameters collected by the unit monitoring system are verified; and the reliability of the data source and the integrity of the attributes of the cross-regional transmission channel boundary data provided by the cross-regional coordination platform are verified.

[0075] Following unified power system data standards, the grid topology data stored in the grid dispatch center undergoes structural standardization and transformation. Unit operating parameters collected by the unit monitoring system are formatted and encoded in a unified way. The attributes of cross-regional transmission channel boundary data provided by the cross-regional coordination platform are standardized. Simultaneously, the data representation formats of renewable energy output forecast data collected by renewable energy power plants and load demand data collected by the load monitoring system are standardized. Based on preset time granularity division rules and spatial partitioning standards, the above five types of data undergo time-dimensional synchronization calibration and spatial-dimensional correlation matching to ensure that all types of data form corresponding connections in terms of time nodes and spatial scope.

[0076] Verify the logical relationships and numerical rationality among various data types, correct any biased or conflicting data items, and ensure data consistency and compatibility. Collaborately integrate verified, standardized, spatiotemporally aligned, and consistency-corrected renewable energy power plant output forecast data, load demand data collected by the load monitoring system, grid topology data stored in the power grid dispatch center, unit operating parameters collected by the unit monitoring system, and cross-regional transmission channel boundary data provided by the cross-regional coordination platform to generate a multi-temporal-scale input dataset.

[0077] S102: Based on a multi-temporal and spatial scale input dataset, the new energy absorption rate target, system operation cost target, pollution emission target and operation stability target are jointly modeled and processed to generate a four-dimensional optimized target model.

[0078] For example, based on a multi-temporal and spatiotemporal scale input dataset, the core influencing factors of the renewable energy absorption rate target are analyzed, and renewable energy output forecast data is correlated with absorption-related constraints to construct a basic representation function for the renewable energy absorption rate target. Key correlation data for the system operating cost target are extracted, and unit operating parameters and resource consumption information are linked to establish a quantitative calculation function for the system operating cost target. The core driving factors of the pollution emission target are analyzed, and a constraint control function for the pollution emission target is generated by combining unit operating parameters and environmental protection standards.

[0079] This paper identifies key influencing parameters of operational stability objectives and constructs a security assurance function for these objectives based on power grid topology data and operational status information. It analyzes the inherent correlations among four types of objectives, clarifies the collaborative logic and conflict mechanisms, and designs coupling rules adapted to the characteristics of multiple objectives. Combining the spatiotemporal dimensional features of multi-scale input datasets, a dynamic response mechanism is configured for each objective function to adapt to different spatiotemporal scenarios. Finally, the four types of objective functions, coupling rules, and dynamic response mechanisms are organically integrated, and after logical consistency verification and objective adaptability verification, a four-dimensional optimized objective model is generated.

[0080] S103: Based on the four-dimensional optimization objective model, collaborative constraint modeling is performed on the source-side conventional unit group operation parameters, new energy cluster output characteristics and load-side flexible load resource adjustment characteristics in the multi-temporal and spatial scale input dataset to generate source-load interaction constraint conditions.

[0081] For example, based on a four-dimensional optimization objective model, operating parameters of the source-side conventional generating units are extracted from the multi-temporal-scale input dataset. Combined with operating attributes such as unit start-up / shutdown status and regulation rate, basic operating constraints of the source-side conventional generating units are constructed. The output characteristics of the renewable energy cluster in the multi-temporal-scale input dataset are analyzed, and the renewable energy absorption rate target is correlated with output fluctuation characteristics to establish boundary constraints for renewable energy cluster absorption.

[0082] This study explores the regulation characteristics of flexible load resources on the load side within multi-temporal and spatiotemporal scale input datasets, distinguishes between interruptible and shiftable load types, clarifies the load regulation capacity and time period adaptation requirements, and generates regulation behavior constraints for flexible loads on the load side. Guided by a four-dimensional optimization objective model, it coordinates three types of constraints and verifies the power balance relationship and regulation rate adaptability. It integrates various constraints that have undergone collaborative verification, and generates source-load interaction constraints through logical consistency analysis and constraint strength adaptation adjustments.

[0083] S104: The spatiotemporal complementarity characteristics of interregional renewable energy output in the multi-temporal scale input dataset are analyzed and processed through the preset standardized coordination interface unit to generate the absorption complementarity factor; based on the absorption complementarity factor and source-load interaction constraints, the local scheduling instructions are dynamically modified to generate the cross-regional power mutual assistance plan.

[0084] For example, by using a pre-defined standardized coordination interface unit, the spatiotemporal dimension features of the spatiotemporal complementarity of new energy output between regions in the multi-spatial-temporal scale input data are extracted, the temporal correlation pattern and spatial coordination law of output fluctuation are analyzed, the analysis results are integrated to form an index for quantifying the absorption and complementarity capability, and an absorption and complementarity factor is generated.

[0085] Based on the absorption complementarity factor, the cross-regional adaptability of local scheduling instructions is evaluated. Combined with the source-load interaction constraints, the cross-regional power transmission parameters that need to be adjusted are identified. The parameters are dynamically calibrated according to the absorption complementarity factor and the constraint compliance is verified. The calibrated and verified parameters are integrated to correct the local scheduling instructions and generate a cross-regional power mutual assistance plan.

[0086] S105: The four-dimensional optimization objective model is solved using a pre-set improved multi-objective evolutionary algorithm to generate a Pareto front solution set; based on the Pareto front solution set and the cross-regional power mutual assistance plan, the execution scheme is generated through a pre-set multi-objective equilibrium decision unit.

[0087] For example, a population adapted to the four-dimensional optimization objective model is initialized using a pre-defined improved multi-objective evolutionary algorithm. The fitness of individual population members is evaluated according to the objective function and constraint boundary of the four-dimensional optimization objective model. The population characteristics are optimized through iterative evolution, and a set of multi-objective non-dominated optimal solutions is selected to generate a Pareto front solution set. Based on the Pareto front solution set, robustness analysis is performed to select disturbance-resistant solutions. Combined with the inter-regional power mutual assistance plan and the embedded transmission channel capacity constraints, the economic, environmental, and stability characteristics of each solution are quantified by a pre-defined multi-objective equilibrium decision unit. The solutions are then non-dominated according to dynamic preference weights, and the compatibility of the ranking results with the inter-regional power mutual assistance plan is verified. The results are integrated and the scheduling instructions are reorganized to generate an execution plan.

[0088] An embodiment of this application provides a multi-objective power system source-load interaction optimization scheduling method for renewable energy consumption. This method synchronously integrates renewable energy output forecasts, load demand, grid topology, unit operating parameters, and cross-regional transmission channel boundary data to generate a multi-temporal and spatial scale input dataset. Based on this dataset, it jointly models four major objectives—renewable energy consumption rate, system operating cost, pollution emissions, and operational stability—to generate a four-dimensional optimization objective model. This method breaks through the limitations of traditional static weighting or priority ranking, which cannot characterize the dynamic conflicts between objectives, and effectively improves the multi-objective collaborative optimization capability.

[0089] Based on a four-dimensional optimization objective model, collaborative constraint modeling is performed on the operating parameters of conventional generating units on the source side, the output characteristics of renewable energy clusters, and the adjustment characteristics of flexible load resources on the load side to generate source-load interaction constraints. This captures the adjustment characteristics of flexible loads, fully explores their adjustment space, and releases their adjustment potential. A standardized coordination interface unit is used to analyze the spatiotemporal complementarity characteristics of renewable energy output between regions to generate absorption complementarity factors. Combined with source-load interaction constraints, local dispatch instructions are dynamically modified to generate a cross-regional power mutual aid plan. An improved multi-objective evolutionary algorithm is then used to solve the four-dimensional optimization objective model to obtain the Pareto front solution set. An execution scheme is generated by combining the cross-regional power mutual aid plan with a multi-objective equilibrium decision unit, allowing the cross-regional mutual aid plan to fully adapt to the spatiotemporal complementarity characteristics of renewable energy and the transmission law of absorption gaps, significantly enhancing the robustness of the cross-regional mutual aid plan.

[0090] In one embodiment, based on the Pareto front solution set and the cross-regional power mutual assistance plan, an execution plan is generated through a preset multi-objective equilibrium decision unit, including:

[0091] (1) Robust gap analysis is performed on the Pareto front solution set to generate an anti-disturbance solution set subspace.

[0092] For example, typical disturbance scenarios affecting power system operation are identified, including the fluctuation range of renewable energy output and the magnitude of load demand changes. For each solution in the Pareto front solution set, the change of its objective function value under each disturbance scenario is simulated. The difference in target performance before and after the disturbance for each solution is calculated, i.e., the robustness gap. Solutions with robustness gaps within a preset threshold are selected, and these solutions are integrated to generate a disturbance-resistant solution set subspace.

[0093] Robust gap analysis refers to the analytical process of simulating disturbance scenarios, quantifying the performance changes of the solution before and after the disturbance, and evaluating the solution's anti-interference capability.

[0094] (2) Through the cross-regional power mutual assistance plan, channel capacity constraint embedding is performed on the solution set subspace of the anti-disturbance solution set to generate a feasible scheduling solution set.

[0095] For example, constraint parameters such as the maximum transmission capacity, minimum transmission capacity, and transmission loss characteristics of each transmission channel are extracted from the inter-regional power sharing plan. A channel capacity constraint model is constructed based on these parameters. Each solution in the disturbance-resistant solution set subspace is substituted into the constraint model for verification, identifying solutions that do not meet the channel capacity constraints. These non-constrained solutions are adjusted, including modifying the inter-regional power transmission amount and optimizing transmission time allocation. The adjusted solutions are then verified again to ensure compliance with the channel capacity constraints, and all compliant solutions are integrated to generate a feasible scheduling solution set.

[0096] Among them, channel capacity constraint embedding refers to the process of converting the operating limitation parameters of the transmission channel into constraint conditions and applying them to the process of solution screening and adjustment to ensure that the solution meets the physical limitations of cross-regional power transmission.

[0097] (3) Using the following formula, the feasible scheduling solution set is subjected to economic-environmental-stability three-axis projection processing to generate a multi-dimensional decision feature vector:

[0098]

[0099] in, Represents a multidimensional decision feature vector. Indicates the index of the feasible scheduling solution. Indicates the first A feasible scheduling solution This represents the value of the economic objective function. This represents the mean of the economic objective function. The standard deviation represents the economic objective function. This represents the value of the environmental protection objective function. This represents the mean of the environmental protection objective function. The standard deviation of the environmental protection objective function is represented by... This represents the value of the stability objective function. The mean of the stability objective function is represented by . The standard deviation represents the stability objective function.

[0100] For example, for each feasible scheduling solution in the feasible scheduling solution set, its corresponding economic objective function value, environmental objective function value, and stability objective function value are calculated. The mean and standard deviation of the economic objective function values, environmental objective function values, and stability objective function values ​​of all solutions in the feasible scheduling solution set are statistically analyzed. The mean of the economic objective function value is subtracted from the economic objective function value of each feasible scheduling solution, and then divided by the standard deviation of the economic objective function value to obtain the standardized economic characteristic value. The standardized environmental characteristic value and standardized stability characteristic value are calculated using the same method. The standardized economic characteristic value, standardized environmental characteristic value, and standardized stability characteristic value of each feasible scheduling solution are combined sequentially to generate a multi-dimensional decision feature vector.

[0101] Among them, the three-axis projection processing of economy, environmental protection and stability refers to the process of extracting features from the solution from three dimensions of economy, environmental protection and stability, eliminating the influence of dimensions through standardization, and mapping the solution to a three-dimensional feature space.

[0102] (4) Based on the preset dynamic preference weights, the multidimensional decision feature vectors are subjected to non-dominated sorting to generate a balanced optimization sequence.

[0103] For example, a preset dynamic preference weight is obtained, which reflects the relative importance of economy, environmental protection, and stability in the current scheduling scenario. Based on the preset dynamic preference weight, a comprehensive evaluation index for multi-dimensional decision feature vectors is constructed. The multi-dimensional decision feature vectors corresponding to solutions in the feasible scheduling solution set are compared pairwise according to the comprehensive evaluation index to determine the dominance relationship between solutions; that is, if one solution has better feature values ​​in all dimensions than another solution, then the former dominates the latter. Non-dominated solutions that are not dominated by other solutions are selected as first-level solutions. The above comparison process is repeated among the remaining solutions to determine second-level, third-level, and so on solutions. The solutions are sorted in descending order of level, combined with the comprehensive evaluation index of solutions at the same level, to generate a balanced optimization sequence.

[0104] Non-dominated ranking refers to a ranking method in multi-objective optimization that classifies solutions into different levels of superiority by comparing the dominance relationships between solutions. Dynamic preference weights are weight coefficients that are dynamically adjusted according to the scheduling scenario and are used to measure the relative importance of economy, environmental protection, and stability.

[0105] (5) The balanced optimization sequence is processed by instruction reorganization through a preset spatiotemporal consistency fusion unit to generate an execution plan.

[0106] For example, the scheduling instructions corresponding to each solution in the equilibrium optimization sequence are analyzed, including unit output plans, load adjustment schemes, and inter-regional power transmission arrangements for each time period. A pre-set spatiotemporal consistency fusion unit verifies the continuity of the scheduling instructions in the time dimension, ensuring that changes in instructions between adjacent time periods comply with time constraints such as unit ramp-up rates and load adjustment rates. The spatial correlation of the scheduling instructions is also verified, ensuring that power transmission and load allocation in different regions are adapted to spatial constraints such as grid topology and transmission channel capacity. Instructions that do not conform to spatiotemporal consistency are corrected, including adjusting output values ​​at time period transition points and optimizing inter-regional power allocation ratios. The corrected scheduling instructions are then integrated to generate a structured execution plan, including specific operational requirements for each time period and region.

[0107] Among them, the preset spatiotemporal consistency fusion unit refers to the processing unit used to verify and correct the adaptability of scheduling instructions in terms of temporal continuity and spatial correlation, so as to ensure that the instructions conform to the spatiotemporal rules of power grid operation.

[0108] like Figure 2 As shown, a feasible scheduling solution set is generated by embedding channel capacity constraints into the subspace of the anti-disturbance solution set through a cross-regional power mutual assistance plan, including:

[0109] S201: Perform multi-dimensional feature analysis on the inter-regional power mutual assistance plan to generate a spatiotemporal capacity constraint tensor.

[0110] For example, the components of the inter-regional power mutual assistance plan are decomposed, and capacity characteristics such as the maximum transmission capacity, minimum transmission capacity, and transmission loss characteristics of each transmission channel are extracted. The capacity change patterns of each channel during different scheduling periods are analyzed to obtain temporal characteristics, and the origin and destination regions of each channel and its correlation with the grid topology are clarified to extract spatial characteristics. The above capacity characteristics, temporal characteristics, and spatial characteristics are mapped to the capacity dimension, time dimension, and spatial dimension of a tensor, respectively, to construct a multi-dimensional tensor structure containing dynamic constraint information of channel capacity in the time and spatial dimensions, i.e., generating a spatiotemporal capacity constraint tensor.

[0111] Among them, multidimensional feature analysis of transmission channels refers to the process of extracting detailed attributes of transmission channels in terms of capacity, time, and space from inter-regional power exchange plans and performing structured analysis. Spatiotemporal capacity constraint tensor refers to a multidimensional data structure that integrates the capacity constraint information of transmission channels in time and space dimensions in tensor form.

[0112] S202: Map the spatiotemporal capacity constraint tensor to the perturbation-resistant solution set subspace for dimension alignment to generate an extended constraint solution set.

[0113] For example, the dimensional composition of the disturbance rejection solution set subspace is analyzed, clarifying the scheduling time period division in the time dimension, the regional power grid distribution in the spatial dimension, and the numerical representation of transmission and output in the power dimension. Then, the time dimension of the spatiotemporal capacity constraint tensor is adjusted to ensure its time period division is completely consistent with the disturbance rejection solution set subspace; the spatial dimension is adjusted to match its region definition; and the power dimension is adjusted to unify its constraint units with the disturbance rejection solution set subspace. The dimensionally adjusted spatiotemporal capacity constraint tensor is then used as a new constraint condition and matched with the scheduling solutions in the disturbance rejection solution set subspace to generate an extended constraint solution set.

[0114] Dimensional alignment refers to adjusting the dimensional structure of the spatiotemporal capacity constraint tensor and the disturbance-resistant solution set subspace to ensure complete matching in terms of time, space, power, and other dimensions. Extended constraint solution set refers to a new solution set formed by embedding the transmission channel constraints represented by the spatiotemporal capacity constraint tensor into the disturbance-resistant solution set subspace.

[0115] S203: The extended constraint solution set is elastically adjusted by a preset margin adaptive scaling unit to generate a reinforced boundary solution set.

[0116] For example, a preset margin adaptive scaling unit collects current power system operating indicators such as the fluctuation range of renewable energy output, the rate of change of load demand, and the operating status of grid equipment. Based on these operating indicators, a safety margin adjustment coefficient is calculated. If the system operating risk is high, the safety margin is increased; if the system operating risk is low, the safety margin is decreased. The constraint boundaries of each scheduling solution in the extended constraint solution set are dynamically scaled according to the adjustment coefficient, and an enhanced boundary solution set is generated after adjustment.

[0117] Among them, the preset margin adaptive scaling unit refers to a processing unit capable of collecting system operating indicators, calculating adjustment coefficients, and dynamically adjusting constraint safety margins. Flexible adjustment of safety boundaries refers to the process of dynamically adjusting the constraint boundary safety margins according to the real-time operating status of the system to adapt to actual safety requirements. Enhanced boundary solution set refers to the solution set of constraint boundaries that is more flexible and adaptable after dynamic adjustment of safety margins.

[0118] S204: Based on the probability distribution of new energy fluctuations, the enhanced boundary solution set is subjected to over-limit risk filtering to generate a risk-immune solution set.

[0119] For example, a probability distribution of renewable energy output fluctuations is constructed using historical data statistics or predictive models. This distribution includes the probability of different fluctuation amplitudes. For each scheduling solution in the reinforced boundary solution set, the transmission channel capacity occupancy is simulated under various renewable energy fluctuation scenarios, and the probability of channel capacity exceeding limits under each scenario is calculated. Scheduling solutions in which the probability of transmission channel capacity exceeding limits is 0 under all renewable energy fluctuation scenarios are selected, and these scheduling solutions constitute a risk-immune solution set.

[0120] Among them, the probability distribution of renewable energy fluctuations refers to the statistical and modeling results of the fluctuation amplitude and probability of renewable energy output. Over-limit risk filtering refers to the process of selecting scheduling solutions that will not cause transmission channel capacity over-limit under all fluctuation scenarios, based on the renewable energy fluctuation probability distribution. The risk-immune solution set refers to the scheduling solution set that can guarantee that the transmission channel capacity will not exceed the limit under various renewable energy fluctuation scenarios.

[0121] S205: Perform constraint satisfaction verification on the risk-immune solution set to generate a feasible scheduling solution set.

[0122] For example, according to the categories of power system operation constraints, each scheduling solution in the risk-immune solution set is verified one by one: The start-up and shutdown status of the conventional generating units on the source side is checked to see if they meet the unit start-up and shutdown constraints, and whether the ramp rate meets the unit ramp rate constraints. The absorption rate of the renewable energy cluster is confirmed to meet the renewable energy absorption constraint requirements. The interruptible capacity and shiftable time periods of the flexible load on the load side are verified to meet the load regulation constraints. Power transmission under the grid topology is verified to meet grid constraints such as line load rate and voltage fluctuation. All scheduling solutions that pass the above-mentioned constraint verifications are integrated to generate a feasible scheduling solution set.

[0123] Among them, constraint satisfaction verification refers to the process of checking the scheduling solution one by one for various power system operation constraints, such as source-side units, renewable energy consumption, load-side load, and grid topology. The feasible scheduling solution set refers to the scheduling solution set that has passed the verification of all power system operation constraints.

[0124] In one embodiment, the spatiotemporal complementarity characteristics of inter-regional new energy output in a multi-temporal-scale input dataset are analyzed and processed through a preset standardized coordination interface unit to generate a complementary absorption factor, including:

[0125] (1) Perform time offset correlation analysis on the spatiotemporal complementarity characteristics of new energy output between regions to generate a cross-regional time complementarity coefficient matrix.

[0126] For example, the temporal granularity of the analysis is determined, and the time series of new energy power output from different regions are extracted from the multi-temporal-scale input dataset. A time offset range is set, and the time series of new energy power output from a certain region is aligned with the time series of new energy power output from other regions according to each offset. The power output correlation coefficient for each pair of regions at different time offsets is calculated, and the correlation coefficients are filled into a matrix using the region as the row index and the time offset as the column index to generate a cross-regional time complementarity coefficient matrix.

[0127] Among them, time offset correlation analysis is the process of analyzing the correlation of new energy output time series in different regions after offset alignment by setting a time offset. The cross-regional time complementarity coefficient matrix is ​​a structured data in matrix form that records the correlation coefficients of new energy output in different regions under different time offsets.

[0128] (2) Perform spatial fluctuation covariance calculation on the spatiotemporal complementary characteristics of new energy output between regions to generate cross-regional spatial complementary covariance tensors.

[0129] For example, the spatial scope of the region is divided, and the fluctuation data of new energy power output of each region within the same time period is extracted from the multi-temporal scale input dataset. The covariance of the new energy power output fluctuation between any two regions is calculated to reflect the degree of coordination between the fluctuations between regions. Using the region pair as the first dimension, time segmentation as the second dimension, and fluctuation frequency band as the third dimension, the calculated covariance value is filled into a three-dimensional tensor structure to generate a cross-regional spatial complementary covariance tensor.

[0130] Among them, spatial fluctuation covariance calculation is the process of calculating the covariance of the fluctuation of new energy power output between different regions to quantify the spatial fluctuation synergy. The cross-regional spatial complementarity covariance tensor is a multi-dimensional data structure that integrates the fluctuation covariance of new energy power output under different regional pairs, different time segments, and different fluctuation frequency bands in the form of a three-dimensional tensor.

[0131] (3) Based on the cross-regional time complementary coefficient matrix, the probability mapping of the absorption gap is processed to generate the time dimension absorption potential feature vector.

[0132] For example, the correlation coefficients of each region pair in the cross-regional time complementarity coefficient matrix and the transmission of absorption gaps are analyzed. A mapping rule is established from the correlation coefficients to the transmission probability of absorption gaps, converting each coefficient in the matrix into a corresponding transmission probability. The transmission probability features of each region are extracted according to the time offset sequence, and the probability values ​​of each region within the entire time offset range are combined into a vector to generate a time-dimensional absorption potential feature vector.

[0133] The probability mapping of absorption gap transmission is a process of converting the correlation coefficients in the cross-regional time complementarity coefficient matrix into the probability of absorption gap transmission between regions. The time-dimensional absorption potential feature vector is a vector that characterizes the potential of each region to receive or transmit absorption gaps from the perspective of time offset, and the vector elements correspond to the transmission probability characteristics under different time offsets.

[0134] (4) Power fluctuation smoothness assessment is performed based on cross-regional spatial complementary covariance tensor to generate spatial dimension absorption potential feature vector.

[0135] For example, the covariance values ​​of each region pair across the entire time segment and frequency band are extracted from the cross-regional spatial complementary covariance tensor, and the power fluctuation smoothness index of the region combination is calculated. The average smoothness of each region combined with all other regions is evaluated to quantify the region's ability to mitigate power fluctuations through spatial cooperation. The average smoothness indices of each region are arranged into a vector according to the region order to generate a feature vector of spatial dimensional absorption potential.

[0136] Among them, the power fluctuation smoothness assessment is based on the cross-regional spatial complementarity covariance tensor, which evaluates the ability of different regional combinations to smooth out power fluctuations in new energy sources. The spatial dimension absorption potential feature vector is a vector representing the potential of each region to absorb new energy sources through complementarity with other regions from the perspective of spatial synergy. The vector elements correspond to the average smoothness index of each region.

[0137] (5) Using the following formula, input the time dimension absorption potential feature vector and the spatial dimension absorption potential feature vector into the preset dynamic coupling unit, perform feature cross-entropy fusion processing, and generate absorption complementarity factor:

[0138]

[0139] in, Indicates the absorption of complementary factors. This represents a feature vector representing the time-dimensional absorption potential. The feature vector representing the spatial dimension absorption potential. The information entropy function is represented by... express Divergence.

[0140] For example, a pre-defined dynamic coupling unit receives a temporal dimension absorption potential feature vector and a spatial dimension absorption potential feature vector, calculates their tensor product, and then processes this tensor product using an information entropy function to quantify the uncertainty of feature crossover. Simultaneously, the KL divergence between the temporal and spatial dimension absorption potential feature vectors is calculated to measure their distributional differences. The information entropy processing result is added to the KL divergence to obtain a fusion value, which is the absorption complementarity factor.

[0141] The preset dynamic coupling unit is a fusion processing unit capable of receiving feature vectors in both time and space dimensions, performing tensor product calculations, information entropy processing, and KL divergence calculations. Feature cross-entropy fusion is a process of fusing the temporal and spatial absorption potential features by combining the information entropy of the tensor product with the KL divergence between feature vectors. The absorption complementarity factor is obtained through the above fusion process and is a quantitative factor that comprehensively characterizes the complementary absorption capacity of new energy sources between regions in both time and space dimensions.

[0142] In one embodiment, a time-dimensional absorption potential feature vector is generated by performing probability mapping processing on the cross-regional time complementarity coefficient matrix to absorb the absorption gap, including:

[0143] (1) Perform inter-regional gap reduction correlation analysis on the cross-regional time complementarity coefficient matrix to generate a gap transmission topology network.

[0144] For example, the correlation coefficients of all region pairs in the cross-regional time complementarity coefficient matrix are extracted, and the coefficient values ​​are converted into the correlation strength of the inter-regional absorption gap according to a preset correlation strength conversion rule. Each region is treated as a network node, and the correlation strength of the region pair is used as the edge weight. When the correlation strength exceeds a preset threshold, a weighted edge is established between the corresponding region nodes. By integrating all nodes and weighted edges, a network structure representing the inter-regional absorption gap transmission relationship is constructed, generating a gap transmission topology network.

[0145] Among them, the inter-regional absorption gap correlation analysis is an analytical process that quantifies the degree of mutual influence between absorption gaps in different regions based on the cross-regional time complementarity coefficient matrix. The gap transmission topology network presents the inter-regional absorption gap transmission relationship in a network form, including nodes (regions) and weighted edges (correlation strength).

[0146] (2) Based on the grid congestion probability, the path reliability weighting processing of the gap transmission topology network is performed to generate a probability transmission channel set.

[0147] For example, the grid congestion probability of each transmission path is collected (which can be obtained from historical operating data or grid congestion prediction models). The reliability of each path is calculated. The calculated reliability is used as a weight and assigned to the corresponding transmission path in the gap transmission topology network. Transmission paths with reliability weights higher than a preset threshold are selected, and the identifiers, starting regions, ending regions, and corresponding reliability weights of the above paths are compiled to form a set containing multiple valid transmission paths and their reliability information, generating a probability transmission channel set.

[0148] The path reliability weighting process involves calculating path reliability based on grid congestion probability and using this reliability as a weight to adjust the importance of paths in the gap propagation topology. The probability propagation channel set is a collection of transmission paths that meet reliability requirements and their reliability weights, used to characterize effective channels that can propagate gaps.

[0149] (3) The probability transmission channel set is processed by the pre-set time series Monte Carlo unit to absorb the gap propagation simulation and generate the spatiotemporal transmission probability distribution.

[0150] For example, a pre-defined Monte Carlo time series unit generates multiple sets of initial time series values ​​for the absorption gap, each set including different gap sizes and occurrence periods. For each initial time series, based on the paths and their reliability weights in the probability propagation channel set, the propagation process of the absorption gap between regions is simulated. At each propagation step, a propagation path is randomly selected according to the path reliability weight, and the probability of each region receiving the absorption gap in each time period is recorded. The results of all simulation sets are integrated to form a three-dimensional distribution structure with region as the spatial dimension, time as the temporal dimension, and probability values ​​as elements, generating a spatiotemporal propagation probability distribution.

[0151] The preset time-series Monte Carlo unit is a processing unit capable of generating time-series simulation scenarios and performing absorption gap propagation simulations. The absorption gap propagation simulation uses random sampling to simulate the time-varying process of absorption gaps between regions. The spatiotemporal propagation probability distribution integrates the simulation results, characterizing the distribution structure of the probability of different regions receiving absorption gaps at different times.

[0152] (4) Perform fluctuation sensitivity enhancement processing on the spatiotemporal transmission probability distribution to generate a robust transmission feature layer.

[0153] For example, the fluctuation amplitude of each probability value in the spatiotemporal transmission probability distribution is calculated in both the time and regional dimensions. Key spatiotemporal points (i.e., points sensitive to fluctuations) whose fluctuation amplitudes exceed a preset range are identified. The probability values ​​of these key spatiotemporal points are enhanced using a preset enhancement strategy, while retaining the basic features of non-sensitive points whose fluctuation amplitudes are within the preset range. The enhanced key features and basic features are then hierarchically integrated to form a multi-level feature structure, generating a robust transmission feature layer.

[0154] Among them, the fluctuation sensitivity enhancement processing is the process of identifying key features in the spatiotemporal transmission probability distribution that are sensitive to fluctuations and strengthening their representation. The robust transmission feature layer is a multi-level feature structure that can stably reflect the transmission law of the absorption gap after the fluctuation sensitivity enhancement processing.

[0155] (5) Input the robust transfer feature layer into the preset feature compression unit for dimensional reduction and regularization to generate a time dimension absorption potential feature vector.

[0156] For example, the preset feature compression unit uses a feature selection algorithm to reduce the dimensionality of the robust transfer feature layer, removing redundant and irrelevant features and reducing the dimensionality of the feature space. The reduced features are then regularized to transform the feature values ​​to a uniform numerical range. Regularized feature values ​​for each region are extracted in chronological order over the entire time range, and the feature values ​​for each region are combined into a vector according to the time series to generate a time-dimensional absorption potential feature vector.

[0157] The preset feature compression unit is a unit capable of performing dimensionality reduction and regularization on features. Dimensionality reduction and regularization are processes that reduce feature dimensions and unify the range of feature values. The time-dimensional absorption potential feature vector is a vector that characterizes the transmission potential of absorption gaps in each region from a time perspective, and its elements correspond to feature values ​​at different times.

[0158] In one embodiment, based on a four-dimensional optimization objective model, collaborative constraint modeling is performed on the source-side conventional unit group operating parameters, new energy cluster output characteristics, and load-side flexible load resource adjustment characteristics in the multi-temporal and spatiotemporal scale input dataset to generate source-load interaction constraint conditions, including:

[0159] (1) The start-stop combination and ramp rate coupling processing of the operating parameters of the conventional unit group on the source side are performed to generate the unit operation constraint set.

[0160] For example, the start-up and shutdown status, minimum continuous operating time, minimum continuous downtime, maximum output limit, minimum output limit, upper limit of ramp rate, and lower limit of ramp rate are extracted from the source-side conventional unit group operating parameters in the multi-temporal and spatiotemporal scale input dataset. The ramp rate variation patterns of the units from shutdown to startup, startup to stable operation, and stable operation to shutdown are analyzed under different start-up and shutdown combination modes, such as independent start-up and shutdown of a single unit and joint start-up and shutdown of multiple units, to establish dynamic correlation rules between start-up and shutdown status and ramp rate. Based on the correlation rules and the extracted parameters, a set containing unit start-up and shutdown status transition constraints, output range constraints, and ramp rate constraints (such as matching corresponding rate limits for different start-up and shutdown stages) is constructed to generate a unit operation constraint set.

[0161] The coupled processing of start-stop combinations and ramp rates involves analyzing the impact of different start-stop combination modes on ramp rates at various operating stages and establishing dynamic correlation rules. The unit operation constraint set is a set of constraints that include the operational limitations of the source-side conventional unit group in terms of start-stop, output, and ramp rates.

[0162] (2) Based on the output characteristics of the new energy cluster, the curtailment penalty coefficient is mapped to generate a new energy consumption constraint set.

[0163] For example, the predicted output sequence, actual output fluctuation range, historical maximum absorption, and grid capacity ceiling are extracted from the output characteristics of new energy clusters in the multi-temporal and spatiotemporal scale input dataset. A mapping rule for the curtailment penalty coefficient is established: the larger the curtailment (the difference between predicted output and actual absorption), the higher the corresponding penalty coefficient, and the penalty coefficient increases stepwise as the curtailment exceeds a preset reasonable range. Combining the predicted output sequence and the grid capacity ceiling, a lower limit constraint on new energy absorption, a correlation constraint between curtailment and the penalty coefficient are established, and these constraints are integrated to generate a new energy absorption constraint set.

[0164] The curtailment penalty coefficient mapping process involves associating the curtailed electricity volume of a renewable energy cluster with a penalty coefficient according to preset rules. The renewable energy consumption constraint set is a set of constraints that includes limitations such as the lower limit of renewable energy consumption and the association between curtailment and penalty coefficients.

[0165] (3) Perform capacity-delay correlation processing on the interruptible load component in the load-side flexible load resource adjustment characteristics to generate a load reduction constraint set.

[0166] For example, the maximum allowable reduction capacity, minimum necessary reduction capacity, maximum response delay of interruption command, and longest duration of a single interruption are extracted from the load-side flexible load resource regulation characteristics of the interruptible load component. The relationship between the reduction capacity of the interruptible load and the interruption response delay is analyzed; that is, a larger reduction capacity requires a longer response delay to ensure user experience, while a smaller reduction capacity corresponds to a shorter response delay. Capacity and delay adaptation rules are established. Based on the adaptation rules and the extracted parameters, a set including reduction capacity range constraints, response delay constraints, and continuous interruption time constraints is constructed to generate a load reduction constraint set.

[0167] Among them, capacity-delay correlation processing is the process of analyzing the adaptation relationship between interruptible load reduction capacity and interruption response delay and establishing correlation rules. The load reduction constraint set is a set of constraints that include limitations on interruptible load in terms of reduction capacity, response delay, duration, etc.

[0168] (4) Perform time-period elastic closure processing on the shiftable load component in the load-side flexible load resource adjustment characteristics to generate a load transfer constraint set.

[0169] For example, the total load, allowed transfer time interval, maximum load transferable amount in each time period, maximum load transferable amount out of each time period, and load transfer flexibility coefficient of the transferable load component are extracted from the load-side flexible load resource adjustment characteristics. Time-period flexibility closure processing is then performed to ensure that the total load of the transferable load remains unchanged before and after the transfer, while limiting the load transferable amount in each time period to not exceed the corresponding maximum limit. The flexibility of the allowed transfer time interval is adjusted according to the flexibility coefficient. Based on the above constraints, a set containing total load conservation constraints, transfer amount constraints for each time period, and transfer time interval constraints is constructed to generate a load transfer constraint set.

[0170] The time-period flexible closure process, while maintaining the total load of the transferable load constant, combines an elastic coefficient to limit the transfer time period and the transfer amount in each time period. The load transfer constraint set is a set of constraints that include limitations on the transferable load in terms of total load conservation, transfer amount, and transfer time period.

[0171] (5) The power balance and regulation rate matching processing of the unit operation constraint set, new energy consumption constraint set, load reduction constraint set and load transfer constraint set are performed to generate source-load interaction constraint conditions.

[0172] For example, the total power output of the conventional generating units on the source side and the actual consumption of renewable energy are calculated. The total load demand on the load side after reduction and transfer is calculated. The power balance relationship between the total power on the source side and the total demand on the load side is verified. If there is an imbalance, the output or load parameter ranges in each constraint set are adjusted. The ramp rate in the unit operation constraint set, the output fluctuation rate in the renewable energy consumption constraint set, and the load adjustment rate in the load reduction and transfer constraint set are analyzed. The matching between each rate is verified. If there is a mismatch, the limits of the rate-related constraints are corrected. All constraints that have undergone power balance verification and adjustment rate matching processing are integrated to form a unified constraint system including source-side and load-side interaction constraints, generating source-load interaction constraint conditions.

[0173] The power balance and regulation rate matching process involves verifying the balance between total power on the source side and total demand on the load side, the matching of regulation rates at each stage, and adjusting parameters. The source-load interaction constraint condition is a unified constraint system that integrates constraints related to generating units, renewable energy, and load, reflecting the interaction limitations on both sides. The output of conventional generating units on the source side is based on the set of unit operation constraints. The actual consumption of renewable energy is based on the set of renewable energy consumption constraints. The load reduction on the load side is based on the set of load reduction constraints.

[0174] In one embodiment, based on a multi-temporal-scale input dataset, the targets for renewable energy absorption rate, system operating cost, pollution emission, and operational stability are jointly modeled to generate a four-dimensional optimization target model, including:

[0175] (1) Perform output deviation integral processing on the new energy consumption rate target to generate the consumption rate optimization function.

[0176] For example, the predicted output and actual absorption output data of the new energy cluster in the multi-temporal and spatial scale input dataset are extracted, the output deviation of each time period is calculated, the above output deviation is integrated within the scheduling period, the correlation between the integral result and the absorption rate is represented by a function, a function with the goal of minimizing the integral of the output deviation is constructed, and the absorption rate optimization function is generated.

[0177] The output deviation integral processing involves integrating the deviation between the predicted output of new energy and the actual output absorbed over time. The absorption rate optimization function is a function that characterizes the optimization objective of the new energy absorption rate, with minimizing the output deviation integral as its core objective.

[0178] (2) The fuel cost and start-stop loss are coupled to generate the operating cost control function.

[0179] For example, the fuel consumption characteristics and start-up loss parameters of conventional units on the source side are extracted from the multi-temporal and spatiotemporal scale input dataset. The coupling relationship between fuel cost and start-up loss of different unit combinations is analyzed. The fuel cost and start-up loss are integrated into the total operating cost. A function with the goal of minimizing the total operating cost is constructed to generate the operating cost control function.

[0180] The coupling treatment of fuel cost and start-up losses involves analyzing the relationship between unit fuel cost and start-up losses and integrating them into the total operating cost. The operating cost control function is a function that characterizes the system's operating cost control objective, with minimizing the total operating cost as its core objective.

[0181] (3) The pollution emission target is converted into carbon and sulfur equivalents to generate a pollution emission suppression function.

[0182] For example, the pollutant emission characteristics of conventional units on the source side in the multi-temporal and spatiotemporal scale input dataset are extracted, different types of pollutants are converted into a unified pollution equivalent according to a preset equivalent conversion rule, the total pollution equivalent is calculated, a function is constructed with the goal of minimizing the total pollution equivalent, and a pollution emission suppression function is generated.

[0183] Among them, carbon and sulfur oxide equivalent conversion treatment is the process of converting different types of pollutants into a uniform pollution equivalent. The pollution emission suppression function is a function that characterizes the target of pollution emission suppression, with minimizing the total pollution equivalent as its core.

[0184] (4) The voltage fluctuation rate and line load rate are combined to constrain the operation stability target, and a stability enhancement function is generated.

[0185] For example, voltage fluctuation data and line load rate data of the power grid are extracted from the multi-temporal and spatial scale input dataset. The joint distribution law of voltage fluctuation rate and line load rate is analyzed. The constraints of the two are represented by functions, and a function is constructed to satisfy the voltage and line constraints, thereby generating a stability enhancement function.

[0186] The joint constraint treatment of voltage fluctuation rate and line load rate involves analyzing the joint law of voltage fluctuation rate and line load rate and functionalizing its constraint conditions. The stability enhancement function is a function that characterizes the objective of enhancing operational stability, with the satisfaction of voltage and line constraints as its core.

[0187] (5) The coupling strength of the absorption rate optimization function, the operating cost control function, the pollution emission suppression function and the stability enhancement function is evaluated by time-varying correlation to generate the target correlation matrix.

[0188] For example, the time-varying correlation of four objective functions in different scheduling periods is analyzed, the correlation coefficient between each pair of objective functions is calculated, and the correlation coefficients are filled into a matrix with the objective functions as rows and columns to generate an objective correlation matrix.

[0189] The time-varying correlation coupling strength assessment process involves analyzing the correlation between the four objective functions at different time periods and calculating the correlation coefficients. The objective correlation matrix is ​​a matrix that records the pairwise correlation coefficients between the four objective functions.

[0190] (6) Perform adaptive weight allocation based on the target correlation matrix to generate a dynamic set of weight coefficients.

[0191] For example, based on the correlation coefficients between the objective functions in the objective correlation matrix, an adaptive weight allocation rule is set, the dynamic weight coefficients of each objective function under different scheduling scenarios are calculated, and these coefficients are organized into a set to generate a dynamic weight coefficient set.

[0192] The adaptive weight allocation process involves calculating the dynamic weights of each objective function according to preset rules based on the correlation coefficients of the objective correlation matrix. The dynamic weight coefficient set is a collection containing the dynamic weight coefficients of the four objective functions under different scenarios.

[0193] (7) Using the following formula, the absorption rate optimization function, operating cost control function, pollution emission suppression function, stability enhancement function, and dynamic weight coefficient set are nonlinearly superimposed to generate a four-dimensional optimization target model:

[0194]

[0195] in, This represents a four-dimensional optimization objective model. Indicates the first The dynamic weighting coefficients of each objective function. Indicates the first The value of the objective function, Represents the nonlinear coupling strength coefficient. This represents the hyperbolic tangent activation function. This represents the scaling factor of the hyperbolic tangent function.

[0196] For example, the four objective functions are multiplied by the corresponding dynamic weight coefficients in the dynamic weight coefficient set, and then the four objective functions are nonlinearly superimposed through a nonlinear coupling term (combining the hyperbolic tangent function and the coupling strength coefficient). All terms are integrated to obtain the comprehensive optimization objective function, generating a four-dimensional optimization objective model.

[0197] The nonlinear superposition process combines dynamic weights and nonlinear coupling terms to integrate the four objective functions into a comprehensive optimization objective. The four-dimensional optimization objective model is a comprehensive optimization model that integrates four objectives: new energy consumption, operating costs, pollution emissions, and operational stability.

[0198] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0199] In one embodiment, such as Figure 3 As shown, this application also provides a multi-objective power system source-load interaction optimization dispatching system 300 for renewable energy consumption, the system 300 including:

[0200] The multi-source data integration module 301 is used to synchronously integrate the new energy output forecast data collected by the preset new energy power plants, the load demand data collected by the load monitoring system, the power grid topology data stored in the power grid dispatch center, the unit operation parameters collected by the unit monitoring system, and the cross-regional power transmission channel boundary data provided by the cross-regional coordination platform to generate a multi-temporal and spatial scale input dataset.

[0201] The four-dimensional target modeling module 302 is used to jointly model the new energy consumption rate target, system operating cost target, pollution emission target and operation stability target based on multi-temporal and spatiotemporal scale input dataset, and generate a four-dimensional optimized target model.

[0202] The source-load constraint modeling module 303 is used to perform collaborative constraint modeling on the source-side conventional unit group operation parameters, new energy cluster output characteristics and load-side flexible load resource adjustment characteristics in the multi-temporal and spatial scale input dataset based on the four-dimensional optimization objective model, and generate source-load interaction constraint conditions.

[0203] The cross-regional coordination and optimization module 304 is used to analyze and process the spatiotemporal complementary characteristics of inter-regional new energy output in the multi-temporal and spatiotemporal scale input dataset through a preset standardized coordination interface unit, and generate a complementary absorption factor; based on the complementary absorption factor and source-load interaction constraints, it dynamically corrects the local scheduling instructions and generates a cross-regional power mutual assistance plan.

[0204] The multi-objective decision module 305 is used to solve the four-dimensional optimization objective model using a preset improved multi-objective evolutionary algorithm to generate a Pareto front solution set; based on the Pareto front solution set and the cross-regional power mutual assistance plan, it is processed by a preset multi-objective equilibrium decision unit to generate an execution plan.

[0205] Specifically, the multi-source data integration module 301 collects new energy output forecast data from new energy power plants, load demand data from load monitoring systems, grid topology data from power grid dispatch centers, unit operating parameters from unit monitoring systems, and cross-regional transmission channel boundary data from cross-regional coordination platforms. It performs time synchronization, spatial matching, and format unification processing on the above data to generate multi-temporal and spatial scale input datasets.

[0206] The four-dimensional target modeling module 302 extracts relevant data from the multi-temporal and spatiotemporal scale input dataset, constructs four target functions: absorption rate optimization, operating cost control, pollution emission suppression, and stability enhancement, analyzes the time-varying correlation of the functions to generate a target correlation matrix, calculates dynamic weight coefficients based on the matrix to generate a dynamic weight coefficient set, and integrates the four functions with dynamic weights and nonlinear coupling terms to generate a four-dimensional optimization target model.

[0207] The source-load constraint modeling module 303 extracts relevant parameters and characteristics from the multi-temporal and spatial scale input dataset, and constructs unit operation constraint sets, new energy consumption constraint sets, load reduction constraint sets and load transfer constraint sets respectively. It verifies the matching of source-load power balance and regulation rate and adjusts parameters, and integrates the four types of constraint sets to generate source-load interaction constraint conditions.

[0208] The cross-regional coordination and optimization module 304 extracts relevant data from the multi-temporal and spatial scale input dataset, generates a cross-regional temporal complementarity coefficient matrix and a cross-regional spatial complementarity covariance tensor, further generates temporal and spatial dimension absorption potential feature vectors and fuses them to obtain absorption complementarity factors, calibrates and corrects local scheduling instructions based on absorption complementarity factors and source-load interaction constraints, and generates a cross-regional power mutual assistance plan.

[0209] The multi-objective decision module 305 uses a preset improved multi-objective evolutionary algorithm to solve the four-dimensional optimization objective model, generate a Pareto front solution set, perform robustness analysis and channel constraint verification on the solution set to generate a feasible scheduling solution set, and generate an execution plan after feature vector extraction, non-dominated sorting and instruction reorganization verification.

[0210] In one embodiment, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0211] In one embodiment, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method embodiments.

[0212] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0213] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.

Claims

1. A multi-objective power system source-load interaction optimization scheduling method for renewable energy consumption, characterized in that, The method includes: The system synchronously integrates and processes the new energy output forecast data collected from the preset new energy power stations, the load demand data collected from the load monitoring system, the power grid topology data stored in the power grid dispatch center, the unit operation parameters collected from the unit monitoring system, and the cross-regional power transmission channel boundary data provided by the cross-regional coordination platform to generate a multi-temporal and spatial scale input dataset. Based on the multi-temporal scale input dataset, the new energy absorption rate target, system operating cost target, pollution emission target and operation stability target are jointly modeled and processed to generate a four-dimensional optimized target model. Based on the four-dimensional optimization objective model, the source-side conventional unit group operation parameters, new energy cluster output characteristics and load-side flexible load resource adjustment characteristics in the multi-temporal and spatial scale input dataset are subjected to collaborative constraint modeling to generate source-load interaction constraint conditions. The spatiotemporal complementarity characteristics of interregional renewable energy output in the multi-temporal scale input dataset are analyzed and processed by a pre-set standardized coordination interface unit to generate a complementary absorption factor; based on the complementary absorption factor and the source-load interaction constraints, the local scheduling instructions are dynamically modified to generate a cross-regional power mutual assistance plan. The four-dimensional optimization objective model is solved using a pre-defined improved multi-objective evolutionary algorithm to generate a Pareto front solution set. Based on the Pareto front solution set and the cross-regional power mutual assistance plan, an execution scheme is generated through a pre-defined multi-objective equilibrium decision unit.

2. The multi-objective power system source-load interaction optimization scheduling method for new energy consumption according to claim 1, characterized in that, The process, based on the Pareto front solution set and the cross-regional power mutual assistance plan, is handled by a pre-set multi-objective equilibrium decision unit to generate an execution plan, including: Robust gap analysis is performed on the Pareto front solution set to generate a perturbation-resistant solution set subspace; The channel capacity constraint embedding process is performed on the anti-disturbance solution set subspace through the cross-regional power mutual assistance plan to generate a feasible scheduling solution set. The feasible scheduling solution set is subjected to a three-axis projection process of economy, environmental protection, and stability using the following formula to generate a multi-dimensional decision feature vector: in, Represents a multidimensional decision feature vector. Indicates the index of the feasible scheduling solution. Indicates the first A feasible scheduling solution This represents the value of the economic objective function. This represents the mean of the economic objective function. The standard deviation represents the economic objective function. This represents the value of the environmental protection objective function. This represents the mean of the environmental protection objective function. The standard deviation of the environmental protection objective function is represented by... This represents the value of the stability objective function. The mean of the stability objective function is represented by . The standard deviation represents the stability objective function; The multidimensional decision feature vector is subjected to non-dominated sorting based on preset dynamic preference weights to generate a balanced optimization sequence. The balanced optimization sequence is processed by instruction reorganization through a preset spatiotemporal consistency fusion unit to generate the execution scheme.

3. The multi-objective power system source-load interaction optimization scheduling method for new energy consumption according to claim 2, characterized in that, The step of performing channel capacity constraint embedding processing on the anti-disturbance solution set subspace through the cross-regional power mutual assistance plan to generate a feasible scheduling solution set includes: The cross-regional power mutual assistance plan is subjected to multi-dimensional feature analysis of the transmission channels to generate a spatiotemporal capacity constraint tensor; The spatiotemporal capacity constraint tensor is mapped to the perturbation-resistant solution set subspace for dimension alignment, thereby generating an extended constraint solution set. The extended constraint solution set is elastically adjusted for safety boundaries by a preset margin adaptive scaling unit to generate a reinforced boundary solution set. Based on the probability distribution of new energy fluctuations, the enhanced boundary solution set is subjected to limit-crossing risk filtering to generate a risk-immune solution set. The risk-immune solution set is subjected to constraint satisfaction verification to generate the feasible scheduling solution set.

4. The multi-objective power system source-load interaction optimization scheduling method for new energy consumption according to claim 1, characterized in that, The step involves analyzing the spatiotemporal complementarity characteristics of inter-regional new energy output in the multi-temporal-scale input dataset through a pre-defined standardized coordination interface unit to generate a complementary absorption factor, including: Time offset correlation analysis was performed on the spatiotemporal complementarity characteristics of new energy output between the regions to generate a cross-regional time complementarity coefficient matrix. Spatial fluctuation covariance calculation is performed on the spatiotemporal complementarity characteristics of new energy output between the regions to generate a cross-regional spatial complementarity covariance tensor. Based on the cross-regional time complementary coefficient matrix, a probability mapping process for the absorption gap transmission is performed to generate a time-dimensional absorption potential feature vector. Based on the cross-regional spatial complementary covariance tensor, power fluctuation smoothness assessment is performed to generate a spatial dimension absorption potential feature vector. Using the following formula, the time-dimensional absorption potential feature vector and the spatial-dimensional absorption potential feature vector are input into a preset dynamic coupling unit for feature cross-entropy fusion processing to generate the absorption complementarity factor: in, Indicates the absorption of complementary factors. This represents a feature vector representing the time-dimensional absorption potential. The feature vector representing the spatial dimension absorption potential. The information entropy function is represented by... express Divergence.

5. The multi-objective power system source-load interaction optimization scheduling method for renewable energy consumption according to claim 4, characterized in that, The step of performing probability mapping processing on the absorption gap propagation based on the cross-regional time complementarity coefficient matrix to generate a time-dimensional absorption potential feature vector includes: The cross-regional time complementarity coefficient matrix is ​​subjected to inter-regional absorption gap correlation analysis to generate a gap transmission topology network; Based on the probability of grid congestion, the gap transmission topology network is subjected to path reliability weighting to generate a probability transmission channel set. The probability propagation channel set is subjected to gap propagation simulation processing by a preset time series Monte Carlo unit to generate a spatiotemporal propagation probability distribution. The spatiotemporal transmission probability distribution is subjected to fluctuation sensitivity enhancement processing to generate a robust transmission feature layer; The robust transfer feature layer is input into a preset feature compression unit for dimensional reduction and regularization to generate the time dimension absorption potential feature vector.

6. The multi-objective power system source-load interaction optimization scheduling method for renewable energy consumption according to claim 1, characterized in that, Based on the four-dimensional optimization objective model, the source-side conventional unit group operating parameters, new energy cluster output characteristics, and load-side flexible load resource adjustment characteristics in the multi-temporal and spatiotemporal scale input dataset are subjected to collaborative constraint modeling to generate source-load interaction constraints, including: The operating parameters of the conventional unit group on the source side are coupled with start-stop combination and ramp rate to generate a set of unit operation constraints; Based on the output characteristics of the new energy cluster, a set of constraints for new energy consumption is generated by mapping the curtailment penalty coefficient. The interruptible load component in the load-side flexible load resource adjustment characteristics is subjected to capacity-delay correlation processing to generate a load reduction constraint set; The shiftable load component in the load-side flexible load resource adjustment characteristics is subjected to time-period elastic closure processing to generate a load transfer constraint set; The power balance and regulation rate matching processes are performed on the set of unit operation constraints, the set of new energy consumption constraints, the set of load reduction constraints, and the set of load transfer constraints to generate the source-load interaction constraints.

7. The multi-objective power system source-load interaction optimization scheduling method for renewable energy consumption according to claim 1, characterized in that, Based on the multi-temporal and spatiotemporal scale input dataset, the new energy absorption rate target, system operating cost target, pollution emission target, and operational stability target are jointly modeled to generate a four-dimensional optimized target model, including: The output deviation is integrally processed to generate the new energy absorption rate target to generate an absorption rate optimization function; The system operating cost target is coupled with fuel cost and start-stop loss to generate an operating cost control function; The pollution emission target is converted to carbon and sulfur equivalent to generate a pollution emission suppression function; The operational stability target is subjected to joint constraint processing of voltage fluctuation rate and line load rate to generate a stability enhancement function; The coupling strength of the absorption rate optimization function, the operating cost control function, the pollution emission suppression function, and the stability enhancement function is evaluated by time-varying correlation to generate a target correlation matrix. Based on the target correlation matrix, adaptive weight allocation is performed to generate a dynamic set of weight coefficients. The following formula is used to nonlinearly superimpose the absorption rate optimization function, the operating cost control function, the pollution emission suppression function, the stability enhancement function, and the dynamic weight coefficient set to generate the four-dimensional optimization target model: in, This represents a four-dimensional optimization objective model. Indicates the first The dynamic weighting coefficients of each objective function. Indicates the first The value of the objective function, Represents the nonlinear coupling strength coefficient. This represents the hyperbolic tangent activation function. This represents the scaling factor of the hyperbolic tangent function.

8. A multi-objective power system source-load interaction optimization dispatching system for renewable energy consumption, characterized in that, The system includes: The multi-source data integration module is used to synchronously integrate and process the new energy output forecast data collected by the preset new energy power plants, the load demand data collected by the load monitoring system, the power grid topology data stored in the power grid dispatch center, the unit operation parameters collected by the unit monitoring system, and the cross-regional power transmission channel boundary data provided by the cross-regional coordination platform to generate a multi-temporal and spatial scale input dataset. The four-dimensional target modeling module is used to jointly model the new energy absorption rate target, system operating cost target, pollution emission target and operation stability target based on the multi-temporal scale input dataset, and generate a four-dimensional optimized target model. The source-load constraint modeling module is used to perform collaborative constraint modeling on the source-side conventional unit group operating parameters, new energy cluster output characteristics and load-side flexible load resource adjustment characteristics in the multi-temporal-scale input dataset based on the four-dimensional optimization target model, and generate source-load interaction constraint conditions. The cross-regional coordination and optimization module is used to analyze the spatiotemporal complementarity characteristics of inter-regional new energy output in the multi-temporal and spatiotemporal scale input dataset through a preset standardized coordination interface unit, and generate a complementary absorption factor; based on the complementary absorption factor and the source-load interaction constraints, the module dynamically corrects the local scheduling instructions and generates a cross-regional power mutual assistance plan. The multi-objective decision module is used to solve the four-dimensional optimization objective model using a preset improved multi-objective evolutionary algorithm to generate a Pareto front solution set; based on the Pareto front solution set and the cross-regional power mutual assistance plan, it is processed by a preset multi-objective equilibrium decision unit to generate an execution plan.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the multi-objective power system source-load interaction optimization scheduling method for new energy consumption as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the multi-objective power system source-load interaction optimization scheduling method for new energy consumption as described in any one of claims 1 to 7.

Images