Multi-energy coupling system optimization scheduling method and device, electronic equipment and storage medium

Abstract

The invention discloses a multi-energy coupling system optimization scheduling method and device, electronic equipment and a storage medium, which are used for solving the technical problem of how to realize more effective and more comprehensive multi-energy coupling system coordination scheduling solution under the condition of considering uncertainty. The method comprises the following steps: carrying out uncertainty modeling based on a day-ahead prediction scene on the multi-energy coupling system to obtain an uncertainty set and a confidence coefficient set; day-ahead scheduling and real-time scheduling are considered at the same time, and a system scheduling model of the multi-energy coupling system is constructed; converting the system scheduling model into a robust optimization scheduling model according to the uncertainty set and the confidence coefficient set; and solving the robust optimization scheduling model by adopting a three-level master-slave problem framework to obtain a real-time scheduling result of the multi-energy coupling system.

Description

Technical Field

[0001] The present invention relates to the technical field of power system dispatching, and in particular to a multi-energy coupling system optimization dispatching method, device, electronic equipment and storage medium. Background Art

[0002] As energy transition deepens, the proportion of renewable energy, represented by photovoltaic power generation, and electrified transportation, represented by electric vehicles, in the energy mix continues to increase. Multi-energy coupling systems, integrating a high proportion of new energy and other flexible resources, have become a key development direction for modern power systems. However, the intermittent nature of renewable energy and the fluctuating load of electric vehicles pose challenges to the stable operation of power systems.

[0003] Gas turbines, capable of rapid startup and shutdown, and flexible power regulation, can effectively mitigate power fluctuations in photovoltaic power generation and electric vehicle electricity consumption. Therefore, the coordinated scheduling of multi-energy coupled systems that incorporate renewable energy and uncertain electricity loads is becoming a focus of industry attention.

[0004] To address uncertainties in power systems, various optimization methods have been employed, such as scenario-based optimization and robust optimization. However, these two common optimization methods, on the one hand, fail to account for rare extreme events, and on the other hand, it is difficult to prove the validity of the applied probability distribution model. Consequently, the reliability of their decisions falls short of system operational requirements. On the other hand, robust optimization only considers the boundary information of uncertainty, resulting in overly conservative solutions. Summary of the Invention

[0005] The present invention provides a multi-energy coupling system optimization scheduling method, device, electronic device and storage medium, which are used to solve or partially solve the technical problem of how to achieve more effective and comprehensive multi-energy coupling system coordinated scheduling solution under consideration of uncertainty.

[0006] The present invention provides a multi-energy coupling system optimization scheduling method, the method comprising:

[0007] Uncertainty modeling of multi-energy coupling systems based on day-ahead prediction scenarios is performed to obtain uncertainty sets and confidence sets;

[0008] Considering both day-ahead scheduling and real-time scheduling, a system scheduling model of the multi-energy coupling system is constructed;

[0009] converting the system scheduling model into a robust optimization scheduling model according to the uncertainty set and the confidence set;

[0010] A three-level master-slave problem framework is adopted to solve the robust optimization scheduling model and obtain the real-time scheduling results of the multi-energy coupling system.

[0011] Optionally, the uncertainty set includes a power generation uncertainty set and a power consumption uncertainty set, and the confidence set includes a power generation confidence set and a power consumption confidence set; and performing uncertainty modeling on the multi-energy coupling system based on a day-ahead prediction scenario to obtain the uncertainty set and the confidence set includes:

[0012] Taking the sum of the predicted value and the error value as the output, the photovoltaic power generation output model and the tram power output model of the multi-energy coupling system are constructed respectively;

[0013] Using an improved minimum volume closed ellipsoid algorithm, uncertainty modeling is performed on the prediction errors of the photovoltaic power generation output model and the electric vehicle power output model based on the day-ahead prediction scenario, thereby obtaining corresponding power generation uncertainty sets and power consumption uncertainty sets.

[0014] Introducing 1-norm and infinity-norm constraints on probability distribution to construct a first initial confidence set for the power generation uncertainty set and a second initial confidence set for the power consumption uncertainty set;

[0015] By introducing binary variables and continuous variables, the first initial confidence set and the second initial confidence set are respectively subjected to linear constraint conversion processing to obtain corresponding revised power generation confidence set and power consumption confidence set.

[0016] Optionally, the improved minimum volume closed ellipsoid algorithm is used to perform uncertainty modeling based on the day-ahead prediction scenario on the prediction errors of the photovoltaic power generation output model and the electric vehicle power consumption output model, respectively, to obtain corresponding power generation uncertainty sets and power consumption uncertainty sets, including:

[0017] Obtain historical data on photovoltaic power generation and historical data on electric vehicle electricity consumption;

[0018] Performing different day-ahead scenario forecasts based on the power generation history data and the power consumption history data to obtain corresponding power generation forecast data and power consumption forecast data;

[0019] Substituting the power generation history data and the power generation forecast data into the photovoltaic power generation output model to obtain power generation forecast error data, and substituting the power consumption history data and the power consumption forecast data into the electric vehicle power output model to obtain power consumption forecast error data;

[0020] Establishing a high-dimensional ellipsoid set according to the power generation prediction error data and the power consumption prediction error data, respectively, to obtain a corresponding power generation ellipsoid set and a power consumption ellipsoid set;

[0021] According to the minimum volume closed ellipsoid algorithm, in combination with orthogonal transformation and inverse transformation, the first original convex polyhedron of the power generation ellipsoid set and the second original convex polyhedron of the power consumption ellipsoid set are inverted;

[0022] By introducing a scaling factor, the first original convex polyhedron and the second original convex polyhedron are respectively corrected to the ellipsoidal range, so as to obtain the corresponding corrected power generation uncertainty set and power consumption uncertainty set.

[0023] Optionally, the simultaneously considering day-ahead scheduling and real-time scheduling to construct a system scheduling model for the multi-energy coupling system includes:

[0024] Taking the minimization of the total operating cost in the day-ahead scheduling phase and the real-time scheduling phase as the optimization goal, the system objective function is constructed;

[0025] A system scheduling model of the multi-energy coupling system is constructed according to the system objective function and preset multiple constraints; the preset multiple constraints include basic constraints in the day-ahead scheduling stage and relative constraints in the real-time scheduling stage.

[0026] Optionally, the multi-energy coupling system includes an electric power network, a natural gas network, and a gas-fired generator set; the gas-fired generator set provides upper and lower reserve capacities for the multi-energy coupling system; and converting the system scheduling model into a robust optimization scheduling model based on the uncertainty set and the confidence set includes:

[0027] Using the gas generator set as an energy coupling element between the power network and the natural gas network;

[0028] converting the upper and lower reserve capacities into a gas consumption uncertainty set of the natural gas network;

[0029] Constructing a pipeline natural gas flow model of the natural gas network, and transforming the pipeline natural gas flow model by combining Wendroff difference and linearization constraints to obtain pressure constraints and flow constraints of gas nodes;

[0030] Combining the gas consumption uncertainty set, the pressure constraint, and the flow constraint to construct a natural gas network decision variable;

[0031] Based on the system objective function and in combination with the natural gas network decision variables, constructing day-ahead scheduling decision variables and real-time scheduling decision variables;

[0032] According to the power generation uncertainty set, the power consumption uncertainty set, the power generation confidence set, the power consumption confidence set, the day-ahead scheduling decision variables and the real-time scheduling decision variables, the system scheduling model is converted into a robust optimization scheduling model in a compact matrix form.

[0033] Optionally, the adopting a three-level master-slave problem framework to solve the robust optimization scheduling model to obtain a real-time scheduling result of the multi-energy coupling system includes:

[0034] The day-ahead scheduling solution is considered as the main problem, and the real-time power correction scheduling of all units that deal with the uncertainty of photovoltaic power generation and electric vehicle power consumption in the real-time scheduling stage is considered as the sub-problem.

[0035] Auxiliary binary variables and continuous variables are introduced to convert the nonlinear constraints and absolute value constraints of the multi-energy coupling system into linear constraints, so as to convert the robust optimization scheduling model into a mixed integer linear programming problem in the form of minimum-maximum-minimum;

[0036] The main problem and subproblems in the mixed integer linear programming problem are solved according to the C&CG algorithm, and the optimal solution value output by the subproblem is used as the real-time scheduling result of the multi-energy coupling system.

[0037] Optionally, all subproblems to be solved are composed of a series of parallel subproblems and one secondary subproblem; solving the main problem and subproblems in the mixed integer linear programming problem according to the C&CG algorithm, and using the optimal solution value output by the subproblem as the real-time scheduling result of the multi-energy coupling system, includes:

[0038] Step S1: setting the initial conditions required by the C&CG algorithm; the initial conditions include an initial upper limit value, an initial lower limit value, a maximum number of iterations, and a difference between upper and lower limits;

[0039] Step S2: Solve the main problem and update the lower limit value based on the solved optimal solution for day-ahead scheduling;

[0040] Step S3: Solve the series of parallel sub-problems in parallel, and solve the secondary sub-problems based on the solved real-time scheduling optimal solutions to obtain the optimal solution of the linear programming, and update the upper limit value according to the optimal solution of the linear programming;

[0041] Step S4: subtract the updated lower limit from the updated upper limit, and determine whether the resulting difference is less than the upper and lower limit difference;

[0042] Step S5: If not, and the current number of iterations has not reached the maximum number of iterations, then re-execute steps S2 to S4 after updating the main problem based on the newly added scenario; if, and / or, the current number of iterations reaches the maximum number of iterations, then extract the optimal solution value of the sub-problem from the upper limit value obtained from the last update as the real-time scheduling result of the multi-energy coupling system.

[0043] The present invention also provides a multi-energy coupling system optimization scheduling device, comprising:

[0044] Uncertainty modeling unit, used to perform uncertainty modeling of the multi-energy coupling system based on the day-ahead prediction scenario, and obtain the uncertainty set and confidence set;

[0045] a scheduling model building unit, configured to simultaneously consider day-ahead scheduling and real-time scheduling to build a system scheduling model for the multi-energy coupling system;

[0046] A model conversion unit, configured to convert the system scheduling model into a robust optimization scheduling model according to the uncertainty set and the confidence set;

[0047] The optimization scheduling unit is used to solve the robust optimization scheduling model using a three-level master-slave problem framework to obtain a real-time scheduling result of the multi-energy coupling system.

[0048] The present invention further provides an electronic device, comprising a processor and a memory:

[0049] The memory is used to store program code and transmit the program code to the processor;

[0050] The processor is configured to execute the multi-energy coupling system optimization scheduling method as described above according to the instructions in the program code.

[0051] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium is used to store program code, and the program code is used to execute the multi-energy coupling system optimization scheduling method as described in any one of the above items.

[0052] It can be seen from the above technical solutions that the present invention has the following advantages:

[0053] A method for optimizing the scheduling of a multi-energy coupling system is presented. First, uncertainty modeling is performed on the multi-energy coupling system based on a day-ahead forecast scenario to obtain an uncertainty set and confidence set. This precise modeling of uncertainty factors enables more accurate prediction of the fluctuations of various resources in the multi-energy coupling system, providing reliable data support for its optimal scheduling. A system scheduling model for the multi-energy coupling system is constructed by simultaneously considering both day-ahead and real-time scheduling. By considering coordinated scheduling at different time scales, the system can flexibly adjust energy allocation within different time periods, further improving its flexibility and responsiveness. Based on the uncertainty set and confidence set, the system scheduling model is transformed into a robust optimization scheduling model. This fully accounts for the uncertainty factors affecting the system during subsequent optimization, resulting in a more accurate and effective scheduling strategy. Finally, a three-level master-slave problem framework is employed to solve the robust optimization scheduling model and obtain real-time scheduling results for the multi-energy coupling system. This model solution, using the three-level master-slave problem framework, not only improves the computational efficiency of large-scale systems but also outputs a more optimal scheduling solution, demonstrating superior economic performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0055] Figure 1 A schematic diagram of the application scenario architecture of a multi-energy coupling system optimization scheduling method;

[0056] Figure 2 A flowchart of the steps of a multi-energy coupling system optimization scheduling method;

[0057] Figure 3 This is a schematic diagram of the overall process of an optimization scheduling method for a multi-energy coupling system;

[0058] Figure 4 This is a structural diagram of an optimization scheduling device for a multi-energy coupling system;

[0059] Figure 5 This is a structural diagram of a coordinated scheduling system for a multi-energy coupling system;

[0060] Figure 6 The figure is a structural diagram of an electronic device for realizing coordinated scheduling of multi-energy coupling systems. DETAILED DESCRIPTION

[0061] Embodiments of the present invention provide a multi-energy coupling system optimization scheduling method, device, electronic device and storage medium for solving or partially solving the technical problem of how to achieve more effective and comprehensive multi-energy coupling system coordinated scheduling solution under consideration of uncertainty.

[0062] In order to make the purpose, features, and advantages of the present invention more obvious and easy to understand, 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 embodiments described below are only part of the embodiments of the present invention, not all of 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.

[0063] As an example, in a multi-energy coupled system, gas turbines with the ability to quickly start and shut down and flexibly adjust power can effectively alleviate power fluctuations in photovoltaic power generation and electric vehicle electricity consumption. Therefore, how to coordinate and dispatch multi-energy coupled systems that incorporate renewable energy and uncertain electricity loads is becoming a focus of industry attention.

[0064] To address uncertainties in power systems, various optimization methods have been employed, such as scenario-based optimization and robust optimization. However, these two common optimization methods, on the one hand, fail to account for rare extreme events, and on the other hand, it is difficult to prove the validity of the applied probability distribution model. Consequently, the reliability of their decisions falls short of system operational requirements. On the other hand, robust optimization only considers the boundary information of uncertainty, resulting in overly conservative solutions.

[0065] Further research and analysis by the present inventors revealed that current optimization processes generally fail to consider the spatiotemporal correlations between outputs such as photovoltaic power generation and electric vehicle electricity consumption, another key factor leading to overly conservative strategies. Furthermore, the coordinated scheduling of power and natural gas networks at different operating timescales is also worth considering. However, the gas flow dynamics, represented by a set of partial differential equations, are difficult to integrate with the power system for joint modeling.

[0066] Therefore, for the optimal scheduling of multi-energy coupled systems, the probability distribution information of photovoltaic power generation and electric vehicle electricity consumption should be integrated into the initial uncertainty sets of their respective optimizations, and the temporal and spatial correlation uncertainties of each should be considered separately. Currently, comprehensive research on these two distribution characteristics is insufficient. Furthermore, further research is needed on the coordinated modeling of the system at different time scales to account for the operational coupling and differences of the two subsystems.

[0067] Therefore, one of the core inventive aspects of the present invention lies in simultaneously considering transient gas flow modeling, spatiotemporal correlations in photovoltaic power generation, and the uncertainty and distribution of electric vehicle electricity consumption, providing a data-driven coordinated optimization scheduling solution for a multi-energy coupling system that considers the spatiotemporal correlation and distributional uncertainty of renewable energy. First, the spatiotemporal correlations of photovoltaic power generation and electric vehicle electricity consumption are determined based on the minimum volume closed convex hull uncertainty set. Simultaneously, probabilistic distribution confidence sets for photovoltaic power generation and electric vehicle electricity consumption scenarios are constructed using 1-norm and infinity-norm constraints. Thus, by establishing a modified convex hull uncertainty set and confidence set for photovoltaic power generation and electric vehicle electricity consumption, the spatiotemporal correlations and distributional uncertainties of photovoltaic power generation output and electric vehicle electricity consumption are comprehensively considered. Next, a Wendrooff difference scheme and linearization techniques are used to transform a series of partial differential equations representing the gas flow dynamics of natural gas, constructing a data-driven distributed robust optimization scheduling model using a master-slave problem framework and a three-level non-dual decomposition method. This approach not only considers the correlations and distributional characteristics of the uncertainties of photovoltaic power generation and electric vehicle electricity consumption, but also reduces the conservativeness of decision-making. The modeling differences and coupling relationships between power networks and gas systems at different operating timescales are also considered, making the scheduling strategy more practical. By adopting the technical solution of this invention, not only can the computational efficiency of large-scale systems be improved, but the renewable energy absorption rate can also be significantly increased. While effectively addressing the impact of uncertainty, it enables multi-energy coupled systems to achieve more optimal scheduling solutions and demonstrate excellent economic performance.

[0068] The optimization scheduling method of multi-energy coupling system based on time-space correlation uncertainty provided by the embodiment of the present invention can be applied to Figure 1 In the application scenario shown.

[0069] Photovoltaic power generation and gas-fired generators are integrated into the traditional power grid to provide electricity. Gas wells transmit gas through natural gas pipelines to gas-consuming equipment and gas-fired generators, forming a natural gas network. These power and natural gas networks are integrated into a multi-energy coupling system that provides electricity and heat to the combined energy load, including electric vehicles. All three are connected to the same communication system and can receive command information for related scheduling.

[0070] Scheduling instructions and data for the multi-energy coupling system are transmitted to the server configured for this device via a communication system. The server then passes the instructions and information to the processor and memory. During actual processing, relevant data, such as historical data on photovoltaic power generation, electric vehicle electricity consumption, and natural gas pipeline flow, is preprocessed using linear transformations such as the Minimum Volume Enclosing Ellipsoid (MVEE) algorithm, orthogonal transformation, inverse transformation, and the Wendroff difference scheme. Day-ahead scheduling data is predicted using the C&CG algorithm (Column-and-Constraint Generation) and the non-dual decomposition method. Based on this data, power correction is used to address the uncertainty of photovoltaic power generation and electric vehicle electricity consumption during the real-time scheduling phase.

[0071] It should be noted that the data storage system can be integrated on this server, or it can be placed on a hard disk, cloud or other network server. In the embodiments of the present invention, the term "server" should be understood as a broad computing device that can be implemented in various forms. Specifically, the server can be a physical server with independent hardware resources (including but not limited to processors, memory, storage devices, etc.). It can also be a server cluster composed of multiple physical servers that work together to provide enhanced processing power and reliability. It can also be a component of a distributed system that is interconnected through a network and collaborates to complete complex computing tasks. Or it can be a cloud server located in a remote data center that allows users to dynamically allocate computing resources according to actual needs without considering the physical location and hardware configuration details.

[0072] Those skilled in the art should understand that any one of the above-mentioned storage methods and server implementation forms or their combinations can be used to execute the method and system described in the present invention, and the selection between different implementation forms will not have a substantial impact on the technical solution of the present invention.

[0073] Reference Figure 2 , shows a flowchart of the steps of a multi-energy coupling system optimization scheduling method provided by an embodiment of the present invention. Figure 1 The electronic device is used as an example for illustration. The method may specifically include the following steps:

[0074] Step 201: uncertainty modeling is performed on the multi-energy coupling system based on the day-ahead prediction scenario to obtain an uncertainty set and a confidence set;

[0075] In a specific implementation, the uncertainty set may further include a power generation uncertainty set and a power consumption uncertainty set, and the confidence set may include a power generation confidence set and a power consumption confidence set. The process of performing uncertainty modeling on the multi-energy coupling system based on the day-ahead forecast scenario and obtaining the uncertainty set and confidence set may include the following sub-steps S2011 to S2014:

[0076] Step S2011: Using the sum of the predicted value and the error value as output, respectively construct a photovoltaic power generation output model and a tram power output model of the multi-energy coupling system;

[0077] Step S2012: Using an improved minimum volume closed ellipsoid algorithm, uncertainty modeling is performed on the prediction errors of the photovoltaic power generation output model and the electric vehicle power output model based on the day-ahead prediction scenario to obtain corresponding power generation uncertainty sets and power consumption uncertainty sets;

[0078] Furthermore, step S2012 can be implemented by executing the following sub-steps S2012-1 to S2012-6:

[0079] Step S2012-1: Obtaining historical data on photovoltaic power generation and historical data on electric vehicle electricity consumption;

[0080] Step S2012-2: performing day-ahead forecasts for different scenarios based on the historical power generation data and the historical power consumption data, respectively, to obtain corresponding power generation forecast data and power consumption forecast data;

[0081] Step S2012-3: Substitute the historical power generation data and the power generation forecast data into the photovoltaic power generation output model to obtain power generation forecast error data, and substitute the historical power consumption data and the power consumption forecast data into the electric vehicle power output model to obtain power consumption forecast error data;

[0082] Step S2012-4: establishing a high-dimensional ellipsoid set based on the power generation prediction error data and the power consumption prediction error data, respectively, to obtain a corresponding power generation ellipsoid set and a power consumption ellipsoid set;

[0083] Step S2012-5: Inverting the first original convex polyhedron of the power generation ellipsoid set and the second original convex polyhedron of the power consumption ellipsoid set using the minimum volume closed ellipsoid algorithm, combined with orthogonal transformation and inverse transformation;

[0084] Step S2012-6: By introducing a scaling factor, the first original convex polyhedron and the second original convex polyhedron are respectively corrected in terms of ellipsoidal range to obtain corresponding corrected power generation uncertainty sets and power consumption uncertainty sets.

[0085] Step S2013: introducing the 1-norm and infinity-norm constraints on the probability distribution, and constructing a first initial confidence set for the power generation uncertainty set and a second initial confidence set for the power consumption uncertainty set;

[0086] Step S2014: by introducing binary variables and continuous variables, linear constraint conversion processing is performed on the first initial confidence set and the second initial confidence set respectively, so as to obtain corresponding revised power generation confidence set and power consumption confidence set.

[0087] Through modeling, the spatiotemporal correlation and uncertainty of photovoltaic power generation and tram electricity consumption are converted into constraints in the subsequent data-driven distributed robust optimization scheduling model.

[0088] Based on the aforementioned steps, in order to enable those skilled in the art to better understand the technical solution of the present invention, the process of constructing the time-space-related uncertainty constraints of new energy power generation will be described in detail using photovoltaic power generation as an example.

[0089] Firstly, considering the spatiotemporal correlation and distribution uncertainty of photovoltaic power generation output, a modified photovoltaic power generation convex hull uncertainty set and a confidence set are established.

[0090] According to the traditional uncertainty method, light energy is defined as the sum of the predicted value and the error value. Then, regarding the uncertainty modeling of photovoltaic power generation, the photovoltaic power generation output can be expressed as the sum of the predicted value and the error value:

[0091]

[0092] Where, Output actual value for photovoltaic power generation; Output predicted values for photovoltaic power generation; is the prediction error value; is the number of planning cycles; is the total number of different error scenarios considered.

[0093] Then, based on the correlation between the prediction errors of adjacent photovoltaic power plants at the same time point and the prediction errors of the same photovoltaic power plant at adjacent time points, and the uncertainty between random variables at different spatiotemporal scales, a minimum volume closed convex hull uncertainty set can be established. Specifically, a high-dimensional ellipsoid set can be first established based on the historical data of photovoltaic prediction errors. This ellipsoid set contains all historical scenarios of photovoltaic prediction errors. Then, the ellipsoid center point and the deviation direction of the ellipsoid symmetry axis of the ellipsoid set are determined using the minimum volume closed ellipsoid algorithm MVEE. The original convex polyhedron is obtained by inversion based on orthogonal transformation and inverse transformation. At this time, the obtained model cannot completely cover all historical scenarios belonging to the ellipsoid. Therefore, the original convex polyhedron can be further modified according to the scale factor to achieve the purpose of scaling the original convex polyhedron to expand the ellipsoid range through the scale factor, and finally obtain the modified convex hull uncertainty set.

[0094] Among them, the prediction error value The uncertainty set can be expressed as the following three sets describe:

[0095]

[0096] Where, and are the lower and upper bounds of the error, respectively.

[0097] Taking into account the temporal and spatial correlation, a high-dimensional ellipsoid set of photovoltaic prediction error historical data (i.e., power generation prediction error data) can be established as follows:

[0098]

[0099] Where, is the center point of the ellipsoid; positive definite matrix is the deviation direction of the ellipsoid's symmetry axis.

[0100] Since the above ellipsoid set contains all historical scenarios of photovoltaic power generation prediction errors, the The uncertainty vector of a historical scenario can be expressed as:

[0101]

[0102] The minimum volume closed ellipsoid algorithm is used to determine the center point of the ellipsoid and the deviation direction of the ellipsoid's symmetry axis by solving the following optimization problem:

[0103]

[0104] Where, is the volume of the N-dimensional unit sphere; is the number of historical scenarios.

[0105] The original convex polyhedron can be constructed by the following orthogonal transformation and inverse transformation:

[0106]

[0107]

[0108] Where, express the corresponding scene in the axial ellipsoid; Represented by orthogonal decomposition The resulting transformation matrix; is a diagonal matrix; Indicates the number of vertices; The axial ellipsoid Vertex coordinate values; Indicates that the original convex polyhedron is obtained by inverse transformation. Vertex coordinate values. Therefore, the original convex polyhedron can be expressed as:

[0109]

[0110] The polyhedron The scale factor is used to scale and correct all historical scenes that belong to the ellipsoid. Therefore, a scale factor is introduced Extensions Region, can be a polyhedron The vertex coordinates are:

[0111]

[0112] Where, It can be obtained by solving the following minimum optimization problem:

[0113]

[0114] By solving the above minimum optimization problem, the proportional factor used for correction can be solved For each historical scene point, a minimum scale factor is determined so that the point can be scaled to the polyhedron. After solving the minimum scale factor value corresponding to all historical scene points, the maximum value of these scale factors is selected as the overall scale factor , thus ensuring that all historical scene points are covered. Through rigorous mathematical optimization, the minimum scale factor that can cover all historical scenes is determined at once, ensuring the completeness of coverage while avoiding overly conservative scaling.

[0115] Then the modified convex polyhedron It can be expressed as:

[0116]

[0117] According to convex set theory, convex polyhedron Any possible photovoltaic power generation error scenario can be expressed as an extreme scenario In the embodiment of the present invention, a day-ahead prediction of the photovoltaic power generation scenario is performed, and the corresponding error extreme scenario real-time description model can be obtained as follows:

[0118]

[0119]

[0120] In the formula, the first Historical error scenario The probability is set to ; The probability of occurrence is the sum of the probability values of the recent historical error scenarios, which are also considered as scenarios The initial probability distribution value (ie, the true probability distribution value) is recorded as .

[0121] According to probability theory, the 1-norm and infinity-norm constraints on probability distribution are introduced to construct a 95% confidence set interval to integrate the distribution information of photovoltaic power generation uncertainty. Specifically, the following confidence set can be constructed:

[0122]

[0123]

[0124] Where, represents the scenario probability estimate used; and The right side of the above inequality can be regarded as the confidence level, that is, the parameter predetermined by the probability theory formula. In the embodiment of the present invention, the confidence level is set to 95%, respectively. and The corresponding tolerance value can be obtained by the following method:

[0125]

[0126] There are nonlinear constraints in the above confidence set constraints. Therefore, in the embodiment of the present invention, by introducing 、 、 、 These binary variables, 、 Continuous variables and formula conversion methods are used to linearize them:

[0127]

[0128] Since the modeling principles and process for the uncertainty constraints of electric vehicle electricity consumption are the same as those for photovoltaic power generation, they will not be elaborated here. After performing a day-ahead forecast for the electric vehicle electricity consumption scenario, the corresponding real-time description model for the extreme error scenario can be obtained as shown below:

[0129]

[0130] Similarly, the first Historical error scenario The probability is set to ;at this time The probability of occurrence is the sum of the probability values of the recent historical error scenarios, which are also considered as scenarios The initial probability distribution value of .

[0131] Similarly, for the electric vehicle electricity usage scenario, the 1-norm and infinity-norm constraints on the probability distribution can be introduced according to the method given in the above embodiment, a 95% confidence set can be constructed, and the nonlinear constraints can be linearized.

[0132] Step 202 , considering both day-ahead scheduling and real-time scheduling, constructing a system scheduling model for the multi-energy coupling system;

[0133] The present invention also aims at the coordination degree of the multi-energy coupling system and establishes a data-driven distributed robust optimization scheduling model to realize the coupling relationship between the power system and the natural gas system under different time scales.

[0134] In its implementation, it is necessary to consider both day-ahead and real-time scheduling simultaneously, building a system scheduling model for a multi-energy coupling system. Specifically, the multi-energy coupling system model, considering different time scales, primarily involves constrained modeling of the power grid, natural gas grid, and gas turbines. The system objective function is constructed with the optimization objective of minimizing total operating costs during both the day-ahead and real-time scheduling phases. Based on the system objective function and pre-set multiple constraints, a system scheduling model for the multi-energy coupling system is constructed. These pre-set multiple constraints primarily include basic constraints during the day-ahead scheduling phase and relative constraints during the real-time scheduling phase.

[0135] The dispatching costs during the day-ahead dispatching phase include the generation costs, start / stop costs, and up / download reserve costs based on the PV generation forecast. The operating costs during the real-time dispatching phase are the worst-case unit regulation costs, downtime costs, and load shedding costs for the PV generation distribution and the tram power consumption distribution. The system objective function is specifically expressed as:

[0136]

[0137] Where, is the unit fuel price; For equipment exist Output power at the moment; and is the unit startup and shutdown cost; and They are respectively the start and stop status indication signs; and are the unit upper and unit lower reserve capacities, respectively; and For equipment exist Up / down reserve capacity at all times; and Adjust prices for upward and downward reserves of units separately; and Equipment exist Moment Scene Adjust the power up and down under; and They are the PV restriction penalty price and the load shedding penalty price respectively; and Respectively expressed in Extreme environment at all times Photovoltaic power generation under photovoltaic reduction and bus exist The amount of load reduction at the .

[0138] In an embodiment of the present invention, the power network imposes constraints on the minimum up / down operating time, startup and shutdown status, minimum / maximum output and upper and lower reserve capacity of the unit, unit ramp rate, line transmission capacity, power balance, up / down correction power limit, and real-time extinction load reduction of the unit according to the day-ahead / real-time scheduling situation.

[0139] The basic constraints of the power system during the day-ahead dispatch phase (i.e., under forecasting conditions) are as follows:

[0140]

[0141] Where, and Respectively represent devices exist Start / shutdown before time; and Representation device Minimum startup / shutdown time; Representation device exist The on / off status at the moment; and Equipment Minimum and maximum output power; For equipment exist Output power at the moment; and Equipment Descent and ascent power limits; and They are the power system's upward / downward reserve requirements; Bus transmission capacity; Bus exist Tidal flow distribution coefficient at ; for Photovoltaic prediction of photovoltaic power plants at all times; for time The power load at the location; for Time-based tram electricity consumption forecast.

[0142] The relative constraints in the real-time scheduling phase are:

[0143]

[0144] Where, 、 Indicates the scene Down, Time unit Up / down adjustment of power; Indicates that in the scene Down time load reduction on roads; Indicates that in the scene Photovoltaic Photovoltaic power generation at the moment; Indicates that in the scene Photovoltaic The amount of photovoltaic loss at the time; Indicates the scene Get off the tram at When the power consumption.

[0145] Step 203: transforming the system scheduling model into a robust optimization scheduling model according to the uncertainty set and the confidence set;

[0146] Combining the above, we can see that a multi-energy coupling system can include a power grid, a natural gas grid, and gas-fired generators. Gas-fired generators consume natural gas to produce electricity and provide up and down reserve capacity for the power system's day-ahead scheduling to account for the uncertainties of photovoltaic power generation and electric vehicle electricity consumption. In other words, gas-fired generators provide up and down reserve capacity for the multi-energy coupling system.

[0147] In some embodiments, the system scheduling model is converted into a robust optimization scheduling model based on the uncertainty set and the confidence set, which can be specifically:

[0148] Based on the characteristics of gas-fired units consuming natural gas to produce electricity, they are used as energy coupling elements between the power network and the natural gas network;

[0149] Based on the characteristic that gas-fired units provide upper and lower reserve capacity for power system day-ahead scheduling to cope with the uncertainty of photovoltaic power generation, the upper and lower reserve capacity of gas-fired units is converted into the uncertainty set of gas consumption in the natural gas network to achieve system-level coupling between the power system and the natural gas system.

[0150] Because natural gas flow in pipelines takes much longer than tidal currents in power systems, a hydrodynamic model expressed by partial differential equations is used to describe the transient characteristics of gas. A pipeline natural gas flow model for the natural gas network is constructed. The pipeline natural gas flow model is linearized using Wendroff differencing and linearization constraints to obtain pressure and flow constraints at gas nodes.

[0151] Combine the uncertainty set of gas consumption, pressure constraints and flow constraints to construct the decision variables of the natural gas network;

[0152] Based on the system objective function and combined with the natural gas network decision variables, the day-ahead scheduling decision variables and real-time scheduling decision variables are constructed;

[0153] According to the generation uncertainty set, consumption uncertainty set, generation confidence set, consumption confidence set, day-ahead scheduling decision variables and real-time scheduling decision variables, the system scheduling model is transformed into a robust optimization scheduling model in a compact matrix form.

[0154] The above embodiments focus on the constraint modeling of the power network. The following describes in detail the key steps of building a robust optimization scheduling model in combination with the constraint content of the natural gas network and gas turbine units.

[0155] The natural gas network part constructs the natural gas flow in the pipeline, the natural gas production limit of the gas well, the pressure limit of the gas node and the flow balance constraint of the gas node.

[0156] The fluid dynamics model of natural gas flow in the pipeline (i.e., pipeline natural gas flow model) can be expressed by the following partial differential equation:

[0157]

[0158] Where, Indicates the natural gas pressure in the pipeline; is the velocity of natural gas in the pipe; is the spatial coordinate variable; is the natural gas flow rate in the pipeline; is the cross-sectional area of the pipe; is the pipe diameter.

[0159] Considering that the airflow direction in each pipeline remains unchanged during the entire scheduling period, the Wendroff difference calculation equation is used to calculate the partial differential terms, and the following equations can be obtained:

[0160]

[0161] Where, is the time variation; Here, represents the speed of sound; is the pipe length; is the pipe friction coefficient.

[0162] The natural gas production limit constraint of the gas well is:

[0163]

[0164] Where, and Respectively represent the minimum and maximum production of the gas well; express Gas well production at a given moment.

[0165] The pressure limit constraint of the gas node is:

[0166]

[0167] Where, and represent the minimum and maximum pressures of the gas nodes, respectively; express Gas node pressure at the moment.

[0168] The gas node flow balance equation for the node is as follows:

[0169]

[0170] Where, Representation device exist Gas load at any moment; Representation device exist Gas consumption at a given moment.

[0171] As can be seen from the previous content, as an energy coupling element between the power network and the natural gas network, the gas generator set must participate in the constraints of both the electrical network and the natural gas network.

[0172] Gas turbines provide up / down reserve capacity in the power grid to handle the uncertainty of photovoltaic power generation. Natural gas consumption can be considered as an uncertainty set, expressed as:

[0173]

[0174] Where HHV is the higher heating value of natural gas.

[0175] To ensure the availability of gas unit reserve configuration, When the upper / lower bounds are reached, all relevant constraints of the gas network should be satisfied. The decision variables in the gas network are expressed as:

[0176]

[0177] Where, express Gas node at time The square of the pressure; Indicates Always connected to the node Gas flow in the pipe is square.

[0178] This embodiment of the present invention constructs a master-slave architecture framework and solves it using the C&CG algorithm. For subproblems involving maximum-minimum values, a non-dual decomposition method is used to convert them into linear programming problems, thereby forming a three-layer hierarchical algorithm to solve the scheduling problem.

[0179] Among them, according to the day-ahead dispatch constraints of the power system, natural gas constraints, and the upper / lower reserve capacity of the gas-fired units as the decision variables for day-ahead dispatch, the day-ahead dispatch decision variables can be expressed as:

[0180]

[0181] Establish photovoltaic scenarios based on real-time dispatch constraints of the power system Real-time scheduling under . The real-time scheduling decision variables can be summarized as:

[0182]

[0183] The data-driven robust optimization scheduling model can be expressed in the form of a compact matrix:

[0184]

[0185] Where, 、 、 、 is a constant coefficient vector; 、 、 、 、 is a constant coefficient matrix; day-ahead scheduling decision variables is the main decision variable of the model; real-time scheduling decision variable is the decision variable for the sub-problem.

[0186] The main problem of decomposition of the above model and subproblems They are:

[0187]

[0188] Where, yes Auxiliary variables introduced in ; yes The probability value of the worst-case distribution of real-time PV power at the first iteration; is a constant coefficient matrix; is a constant coefficient vector; Indicates the optimal cut or feasible cut The new decision vector.

[0189] Step 204 : Using a three-level master-slave problem framework to solve the robust optimization scheduling model, and obtain a real-time scheduling result of the multi-energy coupling system.

[0190] The embodiment of the present invention also proposes a three-level master-slave problem framework based on the combination of the non-dual decomposition method and the C&CG algorithm to solve the proposed robust optimization scheduling model.

[0191] In some embodiments, a three-level master-slave problem framework is used to solve the robust optimization scheduling model to obtain the real-time scheduling results of the multi-energy coupling system, which can be specifically:

[0192] The day-ahead scheduling solution is considered as the main problem, and the real-time power correction scheduling of all units that deal with the uncertainty of photovoltaic power generation and electric vehicle power consumption in the real-time scheduling stage is considered as the sub-problem.

[0193] Auxiliary binary variables and continuous variables are introduced to transform the nonlinear constraints and absolute value constraints of the multi-energy coupled system into linear constraints, so as to transform the robust optimization scheduling model into a mixed integer linear programming problem with the minimum-maximum-minimum form.

[0194] The main problem and subproblems in the mixed integer linear programming problem are solved according to the C&CG algorithm, and the optimal solution value output by the subproblem is used as the real-time scheduling result of the multi-energy coupling system.

[0195] It should be pointed out that in the robust optimization scheduling model, the decomposition of sub-problems is not based on each power generation unit, but on different photovoltaic output scenarios. Each sub-problem This corresponds to the system power correction optimization problem under a specific photovoltaic output scenario. This scenario-based decomposition allows the system to process multiple possible photovoltaic output conditions in parallel, thereby improving the algorithm's computational efficiency.

[0196] Taking photovoltaic power generation as an example, power correction to deal with the uncertainty of photovoltaic power generation means that in the real-time scheduling stage, when there is a deviation between the actual photovoltaic output and the predicted value, the system needs to adjust the output plan of each power generation unit (including conventional units, natural gas equipment, etc.) to maintain the supply and demand balance and safe and stable operation of the system, and make necessary power adjustments to each power generation unit relative to the day-ahead plan in each scenario.

[0197] Specifically, in the day-ahead scheduling phase, the system formulates a preliminary scheduling plan based on the PV output forecast. In the real-time scheduling phase, facing the actual deviation of PV output, the system needs to decide how to adjust the output of each power generation resource. The processing is in a specific photovoltaic output scenario Under the condition of minimizing the objective (i.e. minimizing the cost), the output of each dispatchable unit is calculated.

[0198] In the implementation of the present invention, the main problem and sub-problems are solved according to the C&CG algorithm. The maximum-minimum problem in the sub-problem is decomposed using the non-dual method, so a three-level master-slave problem decomposition method integrated with C&CG can be obtained. According to the non-dual decomposition method, the above sub-problems It can be viewed as a series of parallel subproblems and a linear program :

[0199]

[0200]

[0201] Where, To solve The resulting decision vector Optimal value. That is, all subproblems to be solved A series of parallel subproblems and a secondary sub-problem Based on the large number of uncertain scenarios in the model, the C&CG algorithm is used to iteratively generate constraints and decision variables (i.e., "columns") for key scenarios to approximate the optimal solution to the original problem, effectively reducing computational complexity while ensuring solution quality.

[0202] Furthermore, the execution process of solving the main problem and subproblems in the mixed integer linear programming problem according to the C&CG algorithm and using the optimal solution value output by the subproblem as the real-time scheduling result of the multi-energy coupling system may include the following sub-steps S1 to S5:

[0203] Step S1: Setting the initial conditions required by the C&CG algorithm;

[0204] Initialize the relevant parameters of the C&CG algorithm according to the model scale. Including: initial lower limit value (lower limit ), initial upper limit value (upper limit ), (iteration counter ), upper and lower limit difference (convergence region ), maximum number of iterations, etc.

[0205] Step S2: Solve the main problem and update the lower limit value based on the solved optimal solution for day-ahead scheduling;

[0206] Solve the main problem based on the corresponding data information and constraints And get the optimal solution Update the lower limit accordingly. .

[0207] Step S3: Solve a series of parallel sub-problems in parallel, and solve the secondary sub-problems based on the solved real-time scheduling optimal solutions to obtain the optimal solution of the linear programming, and update the upper limit value according to the optimal solution of the linear programming;

[0208] Solve sub-problems in parallel based on corresponding data information and constraints , to obtain each sub-problem The optimal solution for real-time scheduling is obtained, and the optimal solution with the smallest error is saved as .

[0209] according to Optimal solution Solution ,get Optimal solution . The objective function is recorded as (i.e. the optimal solution of linear programming). Update upper limit value .

[0210] Step S4: subtract the updated lower limit from the updated upper limit, and determine whether the resulting difference is less than the difference between the upper and lower limits;

[0211] Determine whether the difference between the upper limit and the lower limit is less than the given convergence region , to determine whether the scheduling operation is completed.

[0212] Step S5: If not, and the current number of iterations has not reached the maximum number of iterations, then re-execute steps S2 to S4 after updating the main problem based on the newly added scenario; if, and / or, the current number of iterations reaches the maximum number of iterations, then extract the optimal solution value of the sub-problem from the upper limit value obtained from the last update as the real-time scheduling result of the multi-energy coupling system.

[0213] When the difference result is less than the specified convergence region , and / or, the current number of iterations reaches the maximum number of iterations, then the calculation is completed and returned As the optimal solution, the program ends and returns to the initial preparation state of the program. Otherwise, auxiliary variables are generated , and generate corresponding decision variables and constraints for the newly added scenario (to ensure that the newly added scenario can also be covered by the solution proposed in the embodiment of the present invention), add it to the main problem model, and then update the number of iterations. And return to step S2 to continue solving. Its constraints are:

[0214]

[0215] In order to make those skilled in the art understand the technical solution of the present invention more intuitively, Figure 3 The overall flow chart of the multi-energy coupling system optimization scheduling method is shown.

[0216] In an embodiment of the present invention, a data-driven method for optimizing the scheduling of a multi-energy coupling system that considers the spatiotemporal correlation and distribution uncertainty of renewable energy is proposed. When modeling uncertainty, not only the uncertainty of renewable resources such as photovoltaic power generation under spatiotemporal correlation is considered, but also the uncertainty of flexible loads such as trams under spatiotemporal correlation is considered. Furthermore, in the case of multi-energy coupling, the coordinated scheduling at different time scales is considered. On the one hand, by accurately modeling the uncertainty factors, the fluctuation of renewable energy, trams and other flexible loads can be more accurately predicted, thereby providing reliable data support for the optimized scheduling of the multi-energy coupling system. On the other hand, by considering the coordinated scheduling at different time scales, the system can flexibly adjust energy distribution within different time periods, further improving the flexibility and response speed of the system. By adopting the technical solution of the present invention, when optimizing the scheduling of the multi-energy coupling system, multiple energy sources such as solar energy and natural gas are effectively integrated, thereby improving energy utilization efficiency and the stability of system operation.

[0217] Reference Figure 4 , shows a structural block diagram of a multi-energy coupling system optimization scheduling device provided by an embodiment of the present invention, which may specifically include:

[0218] The uncertainty modeling unit 401 is used to perform uncertainty modeling on the multi-energy coupling system based on the day-ahead prediction scenario to obtain an uncertainty set and a confidence set;

[0219] A scheduling model building unit 402 is configured to build a system scheduling model for the multi-energy coupling system by considering both day-ahead scheduling and real-time scheduling;

[0220] A model conversion unit 403 is configured to convert the system scheduling model into a robust optimization scheduling model according to the uncertainty set and the confidence set;

[0221] The optimization scheduling unit 404 is used to solve the robust optimization scheduling model using a three-level master-slave problem framework to obtain a real-time scheduling result of the multi-energy coupling system.

[0222] In an optional embodiment, the uncertainty set includes a power generation uncertainty set and a power consumption uncertainty set, and the confidence set includes a power generation confidence set and a power consumption confidence set; the uncertainty modeling unit 401 includes:

[0223] An output model building unit is used to build a photovoltaic power generation output model and a tram power output model of the multi-energy coupling system respectively using the sum of the predicted value and the error value as output;

[0224] an uncertainty set construction subunit, configured to perform uncertainty modeling on the prediction errors of the photovoltaic power generation output model and the electric vehicle power consumption output model based on a day-ahead prediction scenario using an improved minimum volume closed ellipsoid algorithm, thereby obtaining corresponding power generation uncertainty sets and power consumption uncertainty sets;

[0225] an initial confidence set construction subunit, configured to introduce 1-norm and infinity-norm constraints on probability distribution, and respectively construct a first initial confidence set for the power generation uncertainty set and a second initial confidence set for the power consumption uncertainty set;

[0226] The confidence set conversion unit is used to perform linear constraint conversion processing on the first initial confidence set and the second initial confidence set respectively by introducing binary variables and continuous variables to obtain corresponding revised power generation confidence sets and power consumption confidence sets.

[0227] In an optional embodiment, the uncertainty set construction subunit includes:

[0228] A historical data acquisition unit, used to acquire historical data on photovoltaic power generation and historical data on electric vehicle electricity consumption;

[0229] a day-ahead scenario prediction unit, configured to perform day-ahead scenario predictions based on the power generation history data and the power consumption history data, respectively, to obtain corresponding power generation prediction data and power consumption prediction data;

[0230] a prediction error data calculation unit, configured to substitute the power generation history data and the power generation prediction data into the photovoltaic power generation output model to obtain power generation prediction error data, and substitute the power consumption history data and the power consumption prediction data into the electric vehicle power consumption output model to obtain power consumption prediction error data;

[0231] an ellipsoid set construction unit, configured to respectively establish a high-dimensional ellipsoid set according to the power generation prediction error data and the power consumption prediction error data, and obtain a corresponding power generation ellipsoid set and a power consumption ellipsoid set;

[0232] The original convex polyhedron inversion unit is used to invert the first original convex polyhedron of the power generation ellipsoid set and the second original convex polyhedron of the power consumption ellipsoid set according to the minimum volume closed ellipsoid algorithm, combined with orthogonal transformation and inverse transformation;

[0233] The ellipsoid range correction unit is used to perform ellipsoid range correction on the first original convex polyhedron and the second original convex polyhedron respectively by introducing a proportional factor to obtain corresponding corrected power generation uncertainty set and power consumption uncertainty set.

[0234] In an optional embodiment, the scheduling model building unit 402 includes:

[0235] A system objective function construction unit is used to construct a system objective function with the minimization of the total operating cost in the day-ahead scheduling phase and the real-time scheduling phase as the optimization goal;

[0236] The system scheduling model construction subunit is used to construct the system scheduling model of the multi-energy coupling system according to the system objective function and preset multiple constraints; the preset multiple constraints include the basic constraints of the day-ahead scheduling stage and the relative constraints of the real-time scheduling stage.

[0237] In an optional embodiment, the multi-energy coupling system includes an electric power network, a natural gas network, and a gas generator set; the gas generator set provides upper and lower reserve capacities for the multi-energy coupling system; the model conversion unit 403 includes:

[0238] an energy coupling element determination unit, configured to use the gas generator set as an energy coupling element between the power network and the natural gas network;

[0239] a gas consumption uncertainty set conversion unit, configured to convert the upper and lower reserve capacities into a gas consumption uncertainty set of the natural gas network;

[0240] a pipeline natural gas flow model processing unit, configured to construct a pipeline natural gas flow model of the natural gas network, and transform the pipeline natural gas flow model by combining Wendroff difference and linearization constraints to obtain pressure constraints and flow constraints of gas nodes;

[0241] A natural gas network decision variable construction unit, configured to construct a natural gas network decision variable by combining the gas consumption uncertainty set, the pressure constraint, and the flow constraint;

[0242] A scheduling decision variable construction unit, configured to construct a day-ahead scheduling decision variable and a real-time scheduling decision variable based on the system objective function and in combination with the natural gas network decision variable;

[0243] A robust optimization scheduling model conversion unit is used to convert the system scheduling model into a robust optimization scheduling model in a compact matrix form based on the power generation uncertainty set, the power consumption uncertainty set, the power generation confidence set, the power consumption confidence set, the day-ahead scheduling decision variables and the real-time scheduling decision variables.

[0244] In an optional embodiment, the optimization scheduling unit 404 includes:

[0245] The problem division unit is used to solve the day-ahead scheduling problem as the main problem, and the real-time power correction scheduling of all units that handle the uncertainty of photovoltaic power generation and tram power consumption in the real-time scheduling stage as the sub-problem;

[0246] a linear constraint conversion unit, configured to introduce auxiliary binary variables and continuous variables, and convert the nonlinear constraints and absolute value constraints of the multi-energy coupling system into linear constraints, so as to convert the robust optimization scheduling model into a mixed integer linear programming problem in a minimum-maximum-minimum form;

[0247] A problem-solving unit is used to solve the main problem and sub-problems in the mixed integer linear programming problem according to the C&CG algorithm, and use the optimal solution value output by the sub-problem as the real-time scheduling result of the multi-energy coupling system.

[0248] In an optional embodiment, all subproblems to be solved are composed of a series of parallel subproblems and a secondary subproblem; the problem-solving unit includes:

[0249] An initial condition setting unit is used to execute step S1: setting the initial conditions required by the C&CG algorithm; the initial conditions include an initial upper limit value, an initial lower limit value, a maximum number of iterations, and a difference between upper and lower limits;

[0250] The main problem solving subunit is used to execute step S2: solve the main problem and update the lower limit value based on the solved optimal solution for day-ahead scheduling;

[0251] The subproblem solving subunit is configured to execute step S3: solving the series of parallel subproblems in parallel, solving the secondary subproblems based on the solved real-time scheduling optimal solutions, obtaining a linear programming optimal solution, and updating the upper limit value according to the linear programming optimal solution;

[0252] The upper and lower limit difference judgment unit is used to execute step S4: subtract the updated lower limit from the updated upper limit, and judge whether the obtained difference is less than the upper and lower limit difference;

[0253] The result output unit is used to execute step S5: if not, and the current number of iterations has not reached the maximum number of iterations, then re-execute steps S2 to S4 after updating the main problem based on the newly added scenario; if, and / or, the current number of iterations reaches the maximum number of iterations, then extract the optimal solution value of the sub-problem from the upper limit value obtained from the last update as the real-time scheduling result of the multi-energy coupling system.

[0254] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the aforementioned method embodiment.

[0255] It should be noted that in order to enable those skilled in the art to better distinguish data of the same type but with different actual meanings, some technical features are distinguished by the first and second in the embodiments of the present invention. The first and second are only used for data distinction and have no other special meanings. It can be understood that the present invention does not impose any restrictions on this.

[0256] like Figure 5 As shown, an embodiment of the present invention further provides a multi-energy coupling system coordinated scheduling system, including:

[0257] The data receiving module 100 is used to receive real-time operating data from the power grid, natural gas grid, integrated loads including electric vehicles, and gas turbine coupling components. This real-time operating data includes historical output data from photovoltaic power plants and electric vehicles, flow data from the natural gas pipeline network, and operating parameters of gas turbines, providing a data foundation for subsequent uncertainty modeling and optimized scheduling.

[0258] The data preprocessing module 200 cleans, filters, and standardizes the raw data transmitted by the data receiving module 100, identifies and removes abnormal data, calculates the spatiotemporal correlation of the forecast errors for photovoltaic power generation and electric vehicle power consumption, and constructs a set of uncertainty scenarios for each. These include a modified convex hull uncertainty set and a confidence set, and converts the dynamic characteristics of the natural gas system into a mathematical expression suitable for the optimization model.

[0259] Model operation module 300 is used to construct and solve a robust optimization scheduling model for a multi-energy coupling system based on a data-driven distributed robust optimization method. This module considers the spatiotemporal correlation of the uncertainties of photovoltaic power generation and electric vehicle electricity consumption, as well as the operating characteristics of the power and natural gas systems at different time scales. It implements a three-level master-slave problem framework based on a combination of a non-dual decomposition method and the C&CG algorithm to address optimization problems in both the day-ahead and real-time scheduling phases.

[0260] Data storage module 400 is used to store intermediate results of model calculations and final dispatch plans. This data includes day-ahead dispatch plans, real-time dispatch corrections, photovoltaic power generation and electric vehicle power consumption forecast error distribution parameters for each time period, and natural gas network flow status. Data storage module 400 can also be used to establish a historical database to provide reference for future dispatch optimization.

[0261] The information sending module 500 is used to send the final dispatching instructions to the various control units of the power system and the natural gas system to guide the gas units, photovoltaic power plants and other equipment to operate according to the optimization results, and at the same time feedback the system operation status information to provide a basis for the dynamic adjustment of the dispatching strategy.

[0262] In some embodiments, the data receiving module 100 may also be used to:

[0263] Receive output power data of multiple photovoltaic power plants at different time points to establish a spatiotemporal correlation dataset of photovoltaic power generation prediction errors; receive electric power consumption data of trams in multiple regions at different time points to establish a spatiotemporal correlation dataset of tram power consumption prediction errors; receive parameters such as pressure, flow, and temperature at each node in the natural gas network; receive load demands and grid constraints at each node in the power system; and monitor the operating status and parameter changes of gas turbines in real time.

[0264] In some embodiments, the data storage module 400 may also be used to:

[0265] Establish a historical database of photovoltaic power generation prediction errors to support spatiotemporal correlation analysis; store the operating parameters and status information of each component of the multi-energy coupling system; record iterative data and intermediate results in the scheduling optimization process; implement version management of scheduling plans, support query and comparison of historical plans; establish a scheduling effect evaluation indicator system to provide a basis for system performance optimization.

[0266] In some embodiments, the information sending module 500 may also be used to:

[0267] The format of dispatch instructions is converted according to the communication protocol requirements of different control objects; dispatch instructions are sent in a hierarchical manner to ensure that key equipment receives instructions first; an instruction execution confirmation mechanism is established to monitor the implementation of dispatch plans; in the event of system abnormalities, emergency adjustment instructions are sent to ensure safe and stable operation of the system; and operational feedback data from each system is collected to provide real-time information for dynamic adjustment of dispatch strategies.

[0268] In some embodiments, the coordination and scheduling system further includes a scheduling evaluation module, specifically configured to:

[0269] The effectiveness of the dispatching plan after execution is evaluated, and indicators such as energy utilization efficiency, system operating costs, renewable energy absorption rate, tram charging congestion, and user satisfaction are calculated to provide improvement directions for subsequent dispatching optimization.

[0270] Therefore, through the collaborative work of the above modules, this system can effectively handle the spatiotemporal correlation and distribution uncertainty of distributed photovoltaic and large-scale tram access, realize the coordinated optimization scheduling of power systems and natural gas systems at different time scales, improve energy utilization efficiency, reduce system operating costs, enhance the system's ability to cope with uncertainty, and provide reliable technical support for the efficient operation of multi-energy coupling systems.

[0271] An embodiment of the present invention further provides an electronic device, the device including a processor and a memory:

[0272] The memory is used to store program codes and transmit the program codes to the processor;

[0273] The processor is configured to execute the multi-energy coupling system optimization scheduling method according to any embodiment of the present invention according to the instructions in the program code.

[0274] More specifically, if Figure 6 As shown, an embodiment of the present invention provides an electronic device for implementing coordinated scheduling of a multi-energy coupling system. The electronic device 10 includes a communication device 20, a memory 40, and a processor 30. The memory 40 stores a computer program that, when executed by the processor 30, causes the processor 30 to perform the steps of the multi-energy coupling system optimization scheduling method described in the above embodiment.

[0275] Communication device 20 is used to receive relevant information transmitted from the power grid, natural gas grid, integrated energy system including electric vehicles, and gas turbine coupling elements. This information includes historical and forecasted data on photovoltaic power plants and electric vehicle electricity consumption, flow data from the natural gas pipeline network, and operating parameters of gas turbines. Furthermore, after the optimization calculations are completed, communication device 20 is responsible for sending dispatch instructions to the various control units in the power and natural gas systems, directing the systems to operate according to the optimization results.

[0276] Memory 40 contains embedded computer programs for uncertainty modeling, multi-energy coupling system optimization model construction, and a three-level master-slave problem-solving framework. It also stores historical data on forecast errors, natural gas network parameters, and power system operational constraints. The database in memory 40 supports analysis of the spatiotemporal correlations of the forecast errors, providing the data foundation for constructing the revised convex hull uncertainty and confidence sets for photovoltaic power generation and electric vehicle electricity consumption.

[0277] Processor 30 is responsible for executing the computer program in memory 40, implementing the core functions of the coordinated scheduling method for multi-energy coupled systems. These functions include: constructing a set of high-dimensional ellipsoids based on historical data; determining the deviation direction of the ellipsoid center and axis of symmetry using the minimum volume closed ellipsoid algorithm; modifying the original convex polyhedron using a scaling factor; constructing a 95% confidence interval; establishing a coupling relationship model that considers the different time scales of the power and natural gas systems; solving the distributed robust optimization scheduling model using a three-level master-slave problem framework combining a non-dual decomposition method and the C&CG algorithm; and outputting the final day-ahead scheduling plan and real-time scheduling corrections. Processor 30 controls the sequence and speed of the entire calculation process to ensure that the optimization solution is completed within the specified time.

[0278] An embodiment of the present invention further provides a computer-readable storage medium, which is used to store program code, and the program code is used to execute the multi-energy coupling system optimization scheduling method of any embodiment of the present invention.

[0279] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0280] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0281] The units described as separate components 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 these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0282] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0283] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0284] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-energy coupling system optimization scheduling method, characterized in that: include: Uncertainty modeling of multi-energy coupling systems based on day-ahead prediction scenarios is performed to obtain uncertainty sets and confidence sets; Considering both day-ahead scheduling and real-time scheduling, a system scheduling model of the multi-energy coupling system is constructed; converting the system scheduling model into a robust optimization scheduling model according to the uncertainty set and the confidence set; A three-level master-slave problem framework is adopted to solve the robust optimization scheduling model and obtain the real-time scheduling results of the multi-energy coupling system.

2. The multi-energy coupling system optimization scheduling method according to claim 1, characterized in that: The uncertainty set includes a power generation uncertainty set and a power consumption uncertainty set, and the confidence set includes a power generation confidence set and a power consumption confidence set; The uncertainty modeling of the multi-energy coupling system based on the day-ahead prediction scenario to obtain the uncertainty set and the confidence set includes: Taking the sum of the predicted value and the error value as the output, the photovoltaic power generation output model and the tram power output model of the multi-energy coupling system are constructed respectively; Using an improved minimum volume closed ellipsoid algorithm, uncertainty modeling is performed on the prediction errors of the photovoltaic power generation output model and the electric vehicle power output model based on the day-ahead prediction scenario, thereby obtaining corresponding power generation uncertainty sets and power consumption uncertainty sets. Introducing 1-norm and infinity-norm constraints on probability distribution to construct a first initial confidence set for the power generation uncertainty set and a second initial confidence set for the power consumption uncertainty set; By introducing binary variables and continuous variables, the first initial confidence set and the second initial confidence set are respectively subjected to linear constraint conversion processing to obtain corresponding revised power generation confidence set and power consumption confidence set.

3. The multi-energy coupling system optimization scheduling method according to claim 2, characterized in that: The improved minimum volume closed ellipsoid algorithm is used to perform uncertainty modeling on the prediction errors of the photovoltaic power generation output model and the electric vehicle power output model based on the day-ahead prediction scenario, respectively, to obtain corresponding power generation uncertainty sets and power consumption uncertainty sets, including: Obtain historical data on photovoltaic power generation and historical data on electric vehicle electricity consumption; Performing different day-ahead scenario forecasts based on the power generation history data and the power consumption history data to obtain corresponding power generation forecast data and power consumption forecast data; Substituting the power generation history data and the power generation forecast data into the photovoltaic power generation output model to obtain power generation forecast error data, and substituting the power consumption history data and the power consumption forecast data into the electric vehicle power output model to obtain power consumption forecast error data; Establishing a high-dimensional ellipsoid set according to the power generation prediction error data and the power consumption prediction error data, respectively, to obtain a corresponding power generation ellipsoid set and a power consumption ellipsoid set; According to the minimum volume closed ellipsoid algorithm, in combination with orthogonal transformation and inverse transformation, the first original convex polyhedron of the power generation ellipsoid set and the second original convex polyhedron of the power consumption ellipsoid set are inverted; By introducing a scaling factor, the first original convex polyhedron and the second original convex polyhedron are respectively corrected to the ellipsoidal range, so as to obtain the corresponding corrected power generation uncertainty set and power consumption uncertainty set.

4. The multi-energy coupling system optimization scheduling method according to claim 2, characterized in that: The method of simultaneously considering day-ahead scheduling and real-time scheduling to construct a system scheduling model for the multi-energy coupling system includes: Taking the minimization of the total operating cost in the day-ahead scheduling phase and the real-time scheduling phase as the optimization goal, the system objective function is constructed; A system scheduling model of the multi-energy coupling system is constructed according to the system objective function and preset multiple constraints; the preset multiple constraints include basic constraints in the day-ahead scheduling stage and relative constraints in the real-time scheduling stage.

5. The multi-energy coupling system optimization scheduling method according to claim 4, characterized in that: The multi-energy coupling system includes an electric power network, a natural gas network and a gas-fired generator set; the gas-fired generator set provides upper and lower reserve capacity for the multi-energy coupling system; The converting the system scheduling model into a robust optimization scheduling model according to the uncertainty set and the confidence set includes: Using the gas generator set as an energy coupling element between the power network and the natural gas network; converting the upper and lower reserve capacities into a gas consumption uncertainty set of the natural gas network; Constructing a pipeline natural gas flow model of the natural gas network, and transforming the pipeline natural gas flow model by combining Wendroff difference and linearization constraints to obtain pressure constraints and flow constraints of gas nodes; Combining the gas consumption uncertainty set, the pressure constraint, and the flow constraint to construct a natural gas network decision variable; Based on the system objective function and in combination with the natural gas network decision variables, constructing day-ahead scheduling decision variables and real-time scheduling decision variables; According to the power generation uncertainty set, the power consumption uncertainty set, the power generation confidence set, the power consumption confidence set, the day-ahead scheduling decision variables and the real-time scheduling decision variables, the system scheduling model is converted into a robust optimization scheduling model in a compact matrix form.

6. The multi-energy coupling system optimization scheduling method according to claim 5, characterized in that: The three-level master-slave problem framework is used to solve the robust optimization scheduling model to obtain the real-time scheduling result of the multi-energy coupling system, including: The day-ahead scheduling solution is considered as the main problem, and the real-time power correction scheduling of all units that deal with the uncertainty of photovoltaic power generation and electric vehicle power consumption in the real-time scheduling stage is considered as the sub-problem. Auxiliary binary variables and continuous variables are introduced to convert the nonlinear constraints and absolute value constraints of the multi-energy coupling system into linear constraints, so as to convert the robust optimization scheduling model into a mixed integer linear programming problem in the form of minimum-maximum-minimum; The main problem and subproblems in the mixed integer linear programming problem are solved according to the C&CG algorithm, and the optimal solution value output by the subproblem is used as the real-time scheduling result of the multi-energy coupling system.

7. The multi-energy coupling system optimization scheduling method according to claim 6, characterized in that: All subproblems to be solved are composed of a series of parallel subproblems and a secondary subproblem; solving the main problem and subproblems in the mixed integer linear programming problem according to the C&CG algorithm, and using the optimal solution value output by the subproblem as the real-time scheduling result of the multi-energy coupling system, includes: Step S1: setting the initial conditions required by the C&CG algorithm; the initial conditions include an initial upper limit value, an initial lower limit value, a maximum number of iterations, and a difference between upper and lower limits; Step S2: Solve the main problem and update the lower limit value based on the solved optimal solution for day-ahead scheduling; Step S3: Solve the series of parallel sub-problems in parallel, and solve the secondary sub-problems based on the solved real-time scheduling optimal solutions to obtain the optimal solution of the linear programming, and update the upper limit value according to the optimal solution of the linear programming; Step S4: subtract the updated lower limit from the updated upper limit, and determine whether the resulting difference is less than the upper and lower limit difference; Step S5: If not, and the current number of iterations has not reached the maximum number of iterations, then re-execute steps S2 to S4 after updating the main problem based on the newly added scenario; if, and / or, the current number of iterations reaches the maximum number of iterations, then extract the optimal solution value of the sub-problem from the upper limit value obtained from the last update as the real-time scheduling result of the multi-energy coupling system.

8. A multi-energy coupling system optimization scheduling device, characterized in that: include: Uncertainty modeling unit, used to perform uncertainty modeling of the multi-energy coupling system based on the day-ahead prediction scenario, and obtain the uncertainty set and confidence set; a scheduling model building unit, configured to simultaneously consider day-ahead scheduling and real-time scheduling to build a system scheduling model for the multi-energy coupling system; A model conversion unit, configured to convert the system scheduling model into a robust optimization scheduling model according to the uncertainty set and the confidence set; The optimization scheduling unit is used to solve the robust optimization scheduling model using a three-level master-slave problem framework to obtain a real-time scheduling result of the multi-energy coupling system.

9. An electronic device, characterized in that: The device includes a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the multi-energy coupling system optimization scheduling method according to any one of claims 1 to 7 according to the instructions in the program code.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium is used to store program code, and the program code is used to execute the multi-energy coupling system optimization scheduling method according to any one of claims 1 to 7.

Images