New energy short-term intelligent optimization scheduling method under high-proportion new energy grid connection

Abstract

The invention relates to the technical field of new energy, and particularly discloses a new energy short-term intelligent optimization scheduling method under high-proportion new energy grid connection, and the method comprises the steps: data collection and preprocessing, multi-source data fusion prediction, hierarchical supply and demand balance, distributed collaborative optimization, and real-time regulation and feedback. According to the scheme, a multi-time-scale hierarchical supply and demand balance mechanism is adopted, fine management of different scheduling targets is achieved, the supply and demand relation is coordinated in combination with a dynamic weight adjustment mechanism, key load requirements are met preferentially, economy and stability are considered, and meanwhile the real-time performance and adaptability of scheduling decisions are improved; a distributed collaborative optimization structure is designed, the adaptability and robustness of a scheduling system to operation environment changes are effectively enhanced, a global scheduling problem is decomposed into a plurality of regional sub-problems, parallel solution and asynchronous update are achieved, the scheduling efficiency and the computing resource utilization rate are improved, and the requirements for real-time performance and high concurrency of large-scale new energy grid-connected scheduling are met.

Description

Technical Field

[0001] The present invention relates to the field of new energy technology, and specifically to a short-term intelligent optimization scheduling method for new energy with a high proportion of new energy connected to the grid. Background Art

[0002] The rapid development of renewable energy generation and the integration of a high proportion of renewable energy into the grid place higher demands on grid dispatching. Renewable energy output fluctuates significantly and is highly uncontrollable. Existing technologies primarily focus on analyzing and mitigating fluctuations in the overall grid system, but their short-term dispatch optimization capabilities in scenarios with a high proportion of renewable energy integration are limited. Furthermore, existing renewable energy dispatching solutions often employ centralized optimization methods, which are computationally complex and involve high communication overhead, making them difficult to adapt to the real-time requirements of large-scale renewable energy integration. Summary of the Invention

[0003] In response to the above situation, and to overcome the shortcomings of the existing technology, the present invention provides a short-term intelligent optimization scheduling method for new energy under the high proportion of new energy grid connection. In view of the large fluctuations and strong uncontrollability of new energy output, the existing technology mainly focuses on the volatility analysis and risk response of the overall power grid system, and has limited short-term scheduling optimization capabilities in the scenario of high proportion of new energy grid connection. This solution adopts a hierarchical supply and demand balance mechanism with multiple time scales to achieve refined management of different scheduling objectives, and combines a dynamic weight adjustment mechanism to coordinate the supply and demand relationship, giving priority to meeting the critical load demand, while taking into account economy and stability, and improving the real-time and adaptability of scheduling decisions. In view of the problem that existing new energy scheduling schemes mostly use centralized optimization methods with high computational complexity and large communication overhead, which are difficult to adapt to the real-time requirements of large-scale new energy access scenarios, this solution designs a distributed collaborative optimization structure to effectively enhance the adaptability and robustness of the scheduling system to changes in the operating environment. It decomposes the global scheduling problem into multiple regional sub-problems, realizes parallel solution and asynchronous update, improves scheduling efficiency and computing resource utilization, and adapts to the real-time and high concurrency requirements of large-scale new energy grid connection scheduling.

[0004] The technical solution adopted by the present invention is as follows: The present invention provides a method for short-term intelligent optimization scheduling of new energy under high-proportion new energy grid connection, the method comprising the following steps:

[0005] Step S1: Data collection and preprocessing, obtaining real-time operation data and meteorological data of new energy grid connection, and preprocessing the collected data;

[0006] Step S2: Multi-source data fusion prediction: Based on the time series decomposition algorithm, the real-time operating data of renewable energy power generation output is modally decomposed, and the renewable energy output prediction model is constructed and trained in combination with meteorological data to generate an ultra-short-term forecast value of renewable energy output;

[0007] Step S3: Hierarchical supply and demand balance: Based on ultra-short-term forecast values ​​and real-time operation data, the new energy grid connection scheduling task is divided into time scales. At each time scale, the supply and demand relationship is coordinated through a dynamic weight allocation strategy to generate a resource scheduling strategy;

[0008] Step S4: Distributed collaborative optimization, mapping the coordination strategy to the local area of ​​renewable energy grid connection, and aggregating the sub-problems of the local area into the global optimization problem of renewable energy grid connection. Consistency constraints are introduced to ensure that the solutions of each sub-problem are consistent in the global scope, and a global coordination strategy is generated.

[0009] Step S5: Real-time control and feedback: Based on the global coordination strategy, the equipment related to new energy power generation is controlled in real time. The deviation between the real-time operation data and the ultra-short-term forecast value is calculated and corrected through the closed-loop feedback mechanism.

[0010] Furthermore, in step S2, the multi-source data fusion prediction includes the following steps:

[0011] Step S21: Data decomposition, using a variational mode decomposition algorithm to decompose the trend component, periodic component, and residual component of renewable energy power generation output from real-time operation data;

[0012] Step S22: Feature extraction: extract frequency domain and time domain features from each decomposed modal signal, and construct a fusion feature vector based on meteorological data;

[0013] Step S23: Model training: Use a GRU (Gated Recurrent Unit) neural network to perform time series modeling on the fused feature vectors, build a new energy output prediction model, and iterate through multiple time steps to output the new energy output prediction value for the future time period;

[0014] Step S24: Forecast correction, comparing the predicted value output by the new energy output prediction model with the real-time operation data, calculating the error and correcting the predicted value to generate an ultra-short-term predicted value.

[0015] Furthermore, in step S3, the hierarchical supply and demand balance specifically includes the following steps:

[0016] Step S31: Time scale division, dividing the new energy grid connection dispatching task into three time scales: minute-level, hour-level, and day-level, corresponding to emergency dispatching tasks, mid-term balancing tasks, and long-term planning tasks, respectively;

[0017] Step S32: Dynamic weight allocation, setting weight coefficients according to the fluctuation characteristics of new energy at different time scales, and dynamically adjusting the weight coefficients through heuristic rules;

[0018] Step S33: Supply and demand matching. Based on the ultra-short-term forecast value and real-time operation data, combined with the dynamically adjusted weight coefficient, the supply and demand deviation at each time scale is calculated, the resource response is adjusted, and a preliminary supply and demand balance plan is generated. The formula used is as follows: ;

[0019] Where, represents the supply and demand deviation at each time scale, represents the time scale, represents the weight coefficient, represents the weighted supply and demand deviation;

[0020] Step S34: Optimize the solution. Set the optimization target and iteratively optimize the preliminary supply and demand balance solution. Introduce a penalty function to constrain the part that deviates from the optimization target value to obtain the optimized supply and demand balance solution. The formula used is as follows: ;

[0021] Where, represents the objective function, represents the minimization function, represents the scheduling period, represents the scheduling round, represents the system operating cost, and represents the trade-off parameter between the objective function and the penalty function, represents the penalty function;

[0022] Step S35: Strategy generation: generating a resource scheduling strategy based on the optimized supply and demand balance solution.

[0023] Furthermore, in step S4, the distributed collaborative optimization specifically includes the following steps:

[0024] Step S41: Regional mapping, decomposing the global resource scheduling problem of renewable energy grid connection into local sub-problems. Each sub-problem corresponds to a local area. The objective function and constraints of the sub-problem are obtained by parameter mapping of the coordination strategy.

[0025] Step S42: solving subproblems, using a distributed optimization algorithm to solve each subproblem and obtain a local optimization result;

[0026] Step S43: Consistency constraint modeling, introducing the Lagrange multiplier method to establish consistency constraints on shared variables between regions, ensuring that the solutions to each sub-problem are consistent globally, while reducing the deviation between local optimization and global optimization;

[0027] Step S44: Global optimization, summarizing and uniformly evaluating the local optimization results, and outputting a global coordination strategy.

[0028] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0029] (1) In view of the large fluctuations and strong uncontrollability of renewable energy output, existing technologies mainly focus on the volatility analysis and risk response of the overall power grid system. In the scenario of high proportion of renewable energy grid connection, the short-term dispatch optimization capability is limited. This solution adopts a multi-time scale hierarchical supply and demand balance mechanism to achieve refined management of different dispatch targets, and combines the dynamic weight adjustment mechanism to coordinate the supply and demand relationship, giving priority to meeting the critical load demand, taking into account both economy and stability, while improving the real-time and adaptability of dispatch decisions.

[0030] (2) In view of the problem that the existing new energy scheduling schemes mostly adopt centralized optimization methods, which have high computational complexity and large communication overhead and are difficult to adapt to the real-time requirements in large-scale new energy access scenarios, this scheme designs a distributed collaborative optimization structure to effectively enhance the adaptability and robustness of the scheduling system to changes in the operating environment, decompose the global scheduling problem into multiple regional sub-problems, realize parallel solution and asynchronous update, improve scheduling efficiency and computing resource utilization, and adapt to the real-time and high concurrency requirements of large-scale new energy grid-connected scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a flow chart of a method for short-term intelligent optimization scheduling of new energy with a high proportion of new energy grid connection proposed by the present invention;

[0032] Figure 2 Schematic diagram of the process of step S2;

[0033] Figure 3 Schematic diagram of the process of step S3;

[0034] Figure 4 Schematic diagram of the process of step S4.

[0035] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0036] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0037] Example 1, see Figure 1The present invention provides a method for short-term intelligent optimization scheduling of new energy with a high proportion of new energy grid connection, the method comprising the following steps:

[0038] Step S1: Data collection and preprocessing: acquiring real-time operating data and meteorological data from renewable energy power stations, load centers, and grid dispatch centers. Real-time operating data includes renewable energy output, load demand, and energy storage status information; meteorological data includes temperature, humidity, and wind speed variables. Preprocessing of the collected data is performed.

[0039] Step S2: Multi-source data fusion prediction: Based on the time series decomposition algorithm, the real-time operating data of renewable energy power generation output is modally decomposed, and the renewable energy output prediction model is constructed and trained in combination with meteorological data to generate an ultra-short-term forecast value of renewable energy output;

[0040] Step S3: Hierarchical supply and demand balance: Based on ultra-short-term forecast values ​​and real-time operation data, the new energy grid connection scheduling task is divided into time scales. At each time scale, the supply and demand relationship is coordinated through a dynamic weight allocation strategy to generate a resource scheduling strategy;

[0041] Step S4: Distributed collaborative optimization, mapping the coordination strategy to the local area of ​​renewable energy grid connection, and aggregating the sub-problems of the local area into the global optimization problem of renewable energy grid connection. Consistency constraints are introduced to ensure that the solutions of each sub-problem are consistent in the global scope, and a global coordination strategy is generated;

[0042] Step S5: Real-time control and feedback: Based on the global coordination strategy, the equipment related to new energy power generation is controlled in real time. The deviation between the real-time operation data and the ultra-short-term forecast value is calculated and corrected through the closed-loop feedback mechanism.

[0043] Example 2, see Figure 1 and Figure 2 This embodiment is based on the above embodiment. In step S2, multi-source data fusion prediction specifically includes the following steps:

[0044] Step S21: Data decomposition, using a variational mode decomposition algorithm to decompose the trend component, periodic component, and residual component of the renewable energy power generation output from the real-time operation data, wherein the trend component represents the long-term variation pattern, the periodic component represents the periodic fluctuation pattern, and the residual component represents random noise;

[0045] Step S22: Feature extraction: extract frequency domain and time domain features from each decomposed modal signal: extract amplitude features from trend components, extract frequency and phase features from periodic components, and construct a fusion feature vector by combining the temperature, humidity, and wind speed variables at the corresponding time in the meteorological data. The formula used is as follows: ;

[0046] Where, represents the fused feature vector, Indicates the moment, Indicates modality, represents the amplitude, Indicates frequency, Indicates the phase, 、 、 Respectively represent the temperature, humidity and wind speed in meteorological data;

[0047] Step S23: Model training, using a GRU neural network to perform time series modeling on the fused feature vector, building a new energy output prediction model, and outputting the new energy output prediction value for the future time period through iterative training at multiple time steps;

[0048] Step S24: Forecast correction: compare the predicted value output by the new energy output prediction model with the real-time operation data, calculate the error, and use the sliding weighted average method to correct the predicted value to generate an ultra-short-term prediction value. The formula used is as follows: ;

[0049] Where, express Real-time operation data at all times, express The predicted value at time, After correction The predicted value at time, represents the smoothing factor.

[0050] Example 3, see Figure 1 and Figure 3 This embodiment is based on the above embodiment. In step S3, hierarchical supply and demand balance is performed, specifically including the following steps:

[0051] Step S31: Time scale division, dividing the new energy grid connection dispatching task into three time scales: minute-level, hour-level, and day-level, corresponding to emergency dispatching tasks, mid-term balancing tasks, and long-term planning tasks, respectively;

[0052] Step S32: Dynamic weight allocation. Weight coefficients are set according to the fluctuation characteristics of renewable energy at different time scales. The weight coefficients are dynamically adjusted through heuristic rules: minute-level weights increase as the intensity of renewable energy fluctuations increases, hour-level weights balance the responsiveness and economy of renewable energy output, and daily weights are used for long-term power optimization. The formula used is as follows: ;

[0053] Where, represents the minute-level weight coefficient, represents the standard deviation of new energy output, represents the load mean, represents the electricity price elasticity coefficient;

[0054] Step S33: Supply and demand matching. Based on ultra-short-term forecast values ​​and real-time operating data, combined with dynamically adjusted weight coefficients, the supply and demand deviations at each time scale are calculated. Resource responses are adjusted, including frequency regulation, peak regulation, energy storage, and interruptible loads, to generate a preliminary supply and demand balance plan. The formula used is as follows: ;

[0055] Where, represents the supply and demand deviation at each time scale, represents the time scale, represents the weight coefficient, represents the weighted supply and demand deviation;

[0056] Step S34: Scheme optimization. Set the optimization goal to minimize the system operating cost, iteratively optimize the preliminary supply and demand balance scheme, introduce a penalty function to constrain the part that deviates from the optimization target value, and obtain the optimized supply and demand balance scheme. The formula used is as follows: ;

[0057] Where, represents the objective function, represents the minimization function, represents the scheduling period, represents the scheduling round, represents the system operating cost, and represents the trade-off parameter between the objective function and the penalty function, 、 、 represents the importance factor of the penalty term, 、 、 They represent the energy storage power limit, state of charge constraint and load failure penalty respectively;

[0058] Step S35: Strategy generation: generating a resource scheduling strategy based on the optimized supply and demand balance solution.

[0059] By executing the above operations, in view of the large fluctuations and strong uncontrollability of renewable energy output, the existing technology mainly focuses on the volatility analysis and risk response of the overall power grid system, and the limited short-term scheduling optimization capabilities in the scenario of high proportion of renewable energy grid connection. This solution adopts a hierarchical supply and demand balance mechanism with multiple time scales to achieve refined management of different scheduling targets, and combines the dynamic weight adjustment mechanism to coordinate the supply and demand relationship, give priority to meeting the key load needs, take into account both economy and stability, and improve the real-time and adaptability of scheduling decisions.

[0060] Example 4, see Figure 1 and Figure 4 This embodiment is based on the above embodiment. In step S4, distributed collaborative optimization specifically includes the following steps:

[0061] Step S41: Regional mapping, decomposing the global resource scheduling problem of renewable energy grid connection into local sub-problems, each sub-problem corresponding to a local area, and the objective function and constraints of the sub-problem are obtained by parameter mapping of the resource scheduling strategy;

[0062] Step S42: solving subproblems, using the alternating direction multiplier method to solve each subproblem, by alternating between updating the original variable, updating the global variable, and updating the dual variable to obtain a local optimization result;

[0063] Step S43: Consistency constraint modeling. The Lagrange multiplier method is introduced to establish consistency constraints on shared variables between regions. Shared variables include power, frequency, and voltage at the intersection of local regions. This ensures that the solutions to each sub-problem are consistent globally and reduces the deviation between local optimization and global optimization. The formula used is as follows: ;

[0064] Where, represents the optimization function, represents the subproblem, represents the optimization variable of the subproblem, represents the original objective function of the subproblem, represents the penalty factor, Represents a global variable, Indicates deviation feedback, Represents the square value of the L2 norm;

[0065] Step S44: Global optimization, summarizing and uniformly evaluating the local optimization results, and outputting a global coordination strategy.

[0066] By executing the above operations, this solution designs a distributed collaborative optimization structure to address the problem that existing new energy scheduling schemes mostly adopt centralized optimization methods, which have high computational complexity and large communication overhead, and are difficult to adapt to the real-time requirements in large-scale new energy access scenarios. It effectively enhances the adaptability and robustness of the scheduling system to changes in the operating environment, decomposes the global scheduling problem into multiple regional sub-problems, realizes parallel solution and asynchronous update, improves scheduling efficiency and computing resource utilization, and adapts to the real-time and high concurrency requirements of large-scale new energy grid-connected scheduling.

[0067] Example 5, see Figure 1 This embodiment is based on the above embodiment. In step S5, real-time control and feedback are specifically as follows: based on the global coordination strategy, specific control instructions are generated, including start-stop instructions for new energy power generation equipment, charge and discharge instructions for energy storage devices, and adjustment instructions for load-side resources, and the control instructions are sent to relevant equipment for execution; the deviation value between the real-time operating data after control and the ultra-short-term prediction value is calculated, and new control instructions are generated through a closed-loop feedback mechanism to further correct the global coordination strategy.

[0068] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0069] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

[0070] The present invention and its embodiments are described above. This description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if a person skilled in the art is inspired by this and, without departing from the purpose of the present invention, designs structures and embodiments similar to this technical solution without inventiveness, they shall fall within the scope of protection of the present invention.

Claims

1. A method for short-term intelligent optimization scheduling of new energy with a high proportion of new energy grid connection, characterized by: The method comprises the following steps: Step S1: Data collection and preprocessing, obtaining real-time operation data and meteorological data of new energy grid connection, and preprocessing the collected data; Step S2: Multi-source data fusion prediction: Based on the time series decomposition algorithm, the real-time operating data of renewable energy power generation output is modally decomposed, and the renewable energy output prediction model is constructed and trained in combination with meteorological data to generate an ultra-short-term forecast value of renewable energy output; Step S3: Hierarchical supply and demand balance: Based on ultra-short-term forecast values ​​and real-time operation data, the new energy grid connection scheduling task is divided into time scales. At each time scale, the supply and demand relationship is coordinated through a dynamic weight allocation strategy to generate a resource scheduling strategy; Step S4: Distributed collaborative optimization, mapping the coordination strategy to the local area of ​​renewable energy grid connection, and aggregating the sub-problems of the local area into the global optimization problem of renewable energy grid connection. Consistency constraints are introduced to ensure that the solutions of each sub-problem are consistent in the global scope, and a global coordination strategy is generated. Step S5: Real-time control and feedback: Based on the global coordination strategy, the equipment related to new energy power generation is controlled in real time. The deviation between the real-time operation data and the ultra-short-term forecast value is calculated and corrected through the closed-loop feedback mechanism.

2. The method for short-term intelligent optimization scheduling of new energy with a high proportion of new energy grid connection according to claim 1 is characterized by: In step S2, the multi-source data fusion prediction includes the following steps: Step S21: Data decomposition, using a variational mode decomposition algorithm to decompose the trend component, periodic component, and residual component of renewable energy power generation output from real-time operation data; Step S22: Feature extraction: extract frequency domain and time domain features from each decomposed modal signal, and construct a fusion feature vector based on meteorological data; Step S23: Model training, using a GRU neural network to perform time series modeling on the fused feature vector, building a new energy output prediction model, and outputting the new energy output prediction value for the future time period through iterative training at multiple time steps; Step S24: Forecast correction, comparing the predicted value output by the new energy output prediction model with the real-time operation data, calculating the error and correcting the predicted value to generate an ultra-short-term predicted value.

3. The method for short-term intelligent optimization scheduling of new energy with a high proportion of new energy grid connection according to claim 1 is characterized by: In step S3, the hierarchical supply and demand balance includes the following steps: Step S31: Time scale division, dividing the new energy grid-connected scheduling task into three time scales: minute level, hour level and day level; Step S32: Dynamic weight allocation, setting weight coefficients according to the fluctuation characteristics of new energy at different time scales, and dynamically adjusting the weight coefficients through heuristic rules; Step S33: Supply and demand matching: Based on ultra-short-term forecast values ​​and real-time operation data, combined with dynamically adjusted weight coefficients, calculate the supply and demand deviations at each time scale, adjust resource responses, and generate a preliminary supply and demand balance plan; Step S34: Optimize the solution, set the optimization target, iteratively optimize the preliminary supply and demand balance solution, introduce a penalty function to constrain the part that deviates from the optimization target value, and obtain the optimized supply and demand balance solution; Step S35: Strategy generation: generating a resource scheduling strategy based on the optimized supply and demand balance solution.

4. The method for short-term intelligent optimization scheduling of new energy with a high proportion of new energy grid connection according to claim 1 is characterized by: In step S4, the distributed collaborative optimization includes the following steps: Step S41: Regional mapping, decomposing the global resource scheduling problem of renewable energy grid connection into local sub-problems. Each sub-problem corresponds to a local area. The objective function and constraints of the sub-problem are obtained by parameter mapping of the coordination strategy. Step S42: solving subproblems, using a distributed optimization algorithm to solve each subproblem and obtain a local optimization result; Step S43: Consistency constraint modeling, introducing the Lagrange multiplier method to establish consistency constraints on shared variables between regions, ensuring that the solutions to each sub-problem are consistent globally, while reducing the deviation between local optimization and global optimization; Step S44: Global optimization, summarizing and uniformly evaluating the local optimization results, and outputting a global coordination strategy.

Images