Virtual power plant scheduling method and system based on distributed resource operation domain description

Abstract

The invention provides a virtual power plant scheduling method and system based on distributed resource operation domain description, and relates to the technical field of virtual power plant optimization scheduling, and the method comprises the steps: determining a virtual power plant distributed resource operation domain; considering the influence of demand response load active power and reactive power adjustment on the virtual power plant output model, and constructing a comprehensive load model; dividing a feasible region of a variable in a bilinear term in the comprehensive load model into several non-intersecting regions by adopting a segmented McCormick relaxation method, and converting the regions into a mixed integer linear programming form; the operation constraint condition of the virtual power plant is considered, a mixed integer linear programming form is combined, boundary points of the operation domain of the virtual power plant are solved, and the range of the distributed resource operation domain of the virtual power plant is depicted; and in the distributed resource operation domain range of the virtual power plant, constructing a virtual power plant optimization scheduling model by taking the sum of the total operation cost and the demand response compensation as a target, and applying the obtained optimal operation scheduling strategy to the virtual power plant for execution.

Description

Technical Field

[0001] The present disclosure relates to the technical field of virtual power plant optimization scheduling, and in particular to a virtual power plant scheduling method and system based on distributed resource operation domain characterization. Background Art

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art.

[0003] Although distributed resources such as micro gas turbines and energy storage devices can use their flexible regulation potential to provide regulation services for the power grid, their small size and scattered location make them unable to be directly dispatched by the power grid, so their regulation potential cannot be effectively utilized. At the same time, renewable energy such as distributed photovoltaics has the characteristics of intermittency and volatility, which may affect the safe and stable operation of the power grid, such as causing voltage over-limit and power flow distribution fluctuations. Virtual power plants can aggregate large-scale distributed resources and conduct centralized scheduling to achieve reasonable optimization and utilization of resources. Virtual power plant-distribution network collaborative optimization will give full play to the adjustable potential of flexible resources in virtual power plants and play an important role in building a more flexible and efficient new distribution system.

[0004] The research on the characterization and optimal scheduling of the flexible operating domain of virtual power plants is still in its initial stage. In the actual application process, it faces problems such as inaccurate characterization of the operating domain and scheduling schemes that do not meet safety requirements. Technically, the means of characterizing the operating domain of virtual power plants is reflected in the integration of the operating domains of single resources to form the overall operating domain of the resource cluster. The size of the operating domain of a virtual power plant reflects the flexibility and response speed of resources participating in regulation. In the existing scheme, Minkowski modeling is performed on the original operating domain of single distributed resources to obtain a non-deterministic polynomial problem, and an approximate operating domain model with certain rules is found to simplify the calculation process, and the operating domain boundary is solved through optimization algorithms such as genetic algorithm, improved particle swarm algorithm, and gray wolf algorithm.

[0005] However, existing solutions still have the following problems: 1) The external approximation method cannot accurately describe the flexible operation domain of the virtual power plant. At the same time, most of the research on the optimal scheduling of virtual power plants focuses on improving the efficiency of renewable energy and controlling the operation and maintenance costs of virtual power plants. It does not consider the voltage limit and reactive power flow imbalance existing in the actual topology of virtual power plants, resulting in certain safety hazards in the application of scheduling schemes.

[0006] 2) Existing studies on the characterization of the operating domain and optimal scheduling lack consideration of the load-side demand response, and the modeling of the adjustable capacity of flexible loads in virtual power plants needs further study; 3) The existing methods for characterizing the boundaries of the operating domain focus on the boundary characterization of voltage and current, and ignore key constraints such as the climbing capability of resource equipment in the characterization of the operating domain, and are not aggregated into an image form; in addition, the processing of the relaxation domain size and model processing is not accurate enough, and the calculation speed is relatively slow. Summary of the invention

[0007] In order to solve the above problems, the present disclosure proposes a virtual power plant scheduling method and system based on the characterization of the distributed resource operation domain, constructs an output model of flexible distributed resources in the virtual power plant, and characterizes the distributed resource operation domain. Secondly, considering the demand response behavior of the load, a comprehensive load model under the static characteristics of the voltage is constructed, and the feasible domain of a variable in the bilinear term is divided into several non-overlapping regions based on the piecewise McCormick relaxation method. The model is converted into a mixed integer linear programming form, and the virtual power plant-superior power grid coupling node connection power is used as the target observation variable. By adjusting the direction parameters, the boundary points of the PQ flexible operation domain of the virtual power plant are obtained to characterize the range of the flexible operation domain.

[0008] According to some embodiments, the present disclosure adopts the following technical solutions: The virtual power plant scheduling method based on the characterization of distributed resource operation domain includes: Based on the operating characteristics and mutual influence of the virtual power plant operating equipment, a virtual power plant output model of distributed resources in the virtual power plant is constructed to determine the operating domain of the distributed resources in the virtual power plant; Considering the impact of demand response load active power and reactive power adjustment on the virtual power plant output model, a comprehensive load model under voltage static characteristics is constructed; Based on the comprehensive load model, the piecewise McCormick relaxation method is used to divide the feasible domain of a variable in the bilinear term of the comprehensive load model into several non-overlapping regions, and the comprehensive load model is transformed into a mixed integer linear programming form. Considering the operation constraints of virtual power plants, combined with mixed integer linear programming, the power of the distribution network-virtual power plant coupling node is used as the target observation variable of the virtual power plant operation domain. By adjusting the direction parameter to minimize, the boundary points of the virtual power plant operation domain are obtained to characterize the scope of the virtual power plant distributed resource operation domain. Within the distributed resource operation domain of the virtual power plant, an optimization scheduling model of the virtual power plant is constructed with the sum of total operating cost and demand response compensation as the target, and the solver is used to solve the optimal operation scheduling strategy, which is applied to the virtual power plant execution.

[0009] According to some embodiments, the present disclosure adopts the following technical solutions: The virtual power plant dispatching system based on the characterization of the distributed resource operation domain includes: a virtual power plant model building module, which is used to build a virtual power plant output model of distributed resources in the virtual power plant based on the operation characteristics and mutual influence of the virtual power plant operation equipment, and determine the distributed resource operation domain of the virtual power plant; The relaxation linearization module is used to consider the impact of demand response load active power and reactive power adjustment on the virtual power plant output model and construct a comprehensive load model under voltage static characteristics. Based on the comprehensive load model, the piecewise McCormick relaxation method is used to divide the feasible domain of a variable in the bilinear term in the comprehensive load model into several non-overlapping regions, and the comprehensive load model is converted into a mixed integer linear programming form. The operating domain boundary characterization module is used to consider the operating constraints of the virtual power plant. In combination with the mixed integer linear programming form, the distribution network-virtual power plant coupling node connection power is used as the target observation variable of the virtual power plant operating domain. By adjusting the direction parameter to minimize, the boundary points of the virtual power plant operating domain are obtained to characterize the scope of the virtual power plant distributed resource operating domain. The scheduling solution module is used to build a virtual power plant optimization scheduling model within the distributed resource operation domain of the virtual power plant with the sum of total operating cost and demand response compensation as the target, and use the solver to solve and apply the obtained optimal operation scheduling strategy to the virtual power plant execution.

[0010] According to some embodiments, the present disclosure adopts the following technical solutions: A computer program product includes a computer program, and when the computer program is executed by a processor, the virtual power plant scheduling method based on distributed resource operation domain characterization is implemented.

[0011] According to some embodiments, the present disclosure adopts the following technical solutions: A non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the virtual power plant scheduling method based on distributed resource operation domain characterization is implemented.

[0012] According to some embodiments, the present disclosure adopts the following technical solutions: An electronic device comprises: a processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device executes the virtual power plant scheduling method based on the distributed resource operation domain characterization.

[0013] Compared with the prior art, the present invention has the following beneficial effects: The virtual power plant scheduling method based on the characterization of the distributed resource operation domain disclosed in the present invention constructs an output model of flexible distributed resources in the virtual power plant based on the operating characteristics and mutual influence of wind turbines, photovoltaic power generation units, micro gas turbine units, and energy storage units, characterizes the distributed resource operation domain, characterizes the PQ flexible operation domain of the virtual power plant containing flexible distributed resources and adjustable loads, reduces the operating cost of the virtual power plant, and realizes the unity of safety and economy of the operation of the virtual power plant.

[0014] The virtual power plant dispatching method disclosed in the present invention, based on the characterization of the distributed resource operation domain, considers the demand response behavior of the load, constructs a comprehensive load model under the voltage static characteristics, considers the existence of non-convex bilinear terms in the model, divides the feasible domain of a variable in the bilinear term into several non-overlapping regions based on the piecewise McCormick relaxation method, and converts the model into a mixed integer linear programming form. According to the branch flow model, the constraints of the active power, reactive power, voltage, and current of the virtual power plant are set, and the nonlinear terms are processed based on the second-order cone relaxation method, which reduces the complexity of the model solution while ensuring the solution accuracy.

[0015] The virtual power plant scheduling method based on the characterization of the distributed resource operation domain disclosed in the present invention takes into account the operation constraints of the above-mentioned virtual power plant, and adopts the communication power of the virtual power plant-upper-level power grid coupling node as the target observation variable. By adjusting the direction parameters, the boundary points of the virtual power plant PQ flexible operation domain are obtained, the range of the flexible operation domain is characterized, and an observable virtual power plant operation domain image is obtained, thereby avoiding the privacy leakage of the distributed resource equipment parameters inside the virtual power plant during the coordinated optimization scheduling with the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings constituting a part of the present disclosure are used to provide a further understanding of the present disclosure. The illustrative embodiments of the present disclosure and their descriptions are used to explain the present disclosure and do not constitute an improper limitation on the present disclosure.

[0017] Figure 1 Schematic diagram of segmented McCormick relaxation of an embodiment of the present disclosure; Figure 2 Describes the scope of the flexible operation domain of the virtual power plant PQ in the embodiment of the present disclosure; Figure 3 This is a flowchart of the application of the virtual power plant scheduling method based on distributed resource operation domain characterization according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0018] The present disclosure is further described below in conjunction with the accompanying drawings and embodiments.

[0019] It should be noted that the following detailed descriptions are all illustrative and are intended to provide further explanation of the present disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present disclosure belongs.

[0020] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present disclosure. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0021] Example 1 In one embodiment of the present disclosure, a virtual power plant scheduling method based on distributed resource operation domain characterization is provided, and the steps include: Step 1: Based on the operating characteristics and mutual influence of the virtual power plant operating equipment, a virtual power plant output model of distributed resources in the virtual power plant is constructed to determine the operating domain of the distributed resources of the virtual power plant; Step 2: Consider the impact of demand response load active power and reactive power adjustment on the virtual power plant output model, and construct a comprehensive load model under voltage static characteristics; Step 3: Based on the comprehensive load model, the piecewise McCormick relaxation method is used to divide the feasible domain of a variable in the bilinear term in the comprehensive load model into several non-overlapping regions, and the comprehensive load model is converted into a mixed integer linear programming form; Step 4: Considering the operating constraints of the virtual power plant, combined with the mixed integer linear programming form, the distribution network-virtual power plant coupling node connection power is used as the target observation variable of the virtual power plant operation domain. By adjusting the direction parameter to minimize, the boundary points of the virtual power plant operation domain are obtained to characterize the scope of the virtual power plant distributed resource operation domain; Step 5: Within the distributed resource operation domain of the virtual power plant, a virtual power plant optimization scheduling model is constructed with the sum of total operating cost and demand response compensation as the target, and the optimal operation scheduling strategy is applied to the virtual power plant using the solver.

[0022] As an embodiment, the virtual power plant scheduling method based on the characterization of the distributed resource operation domain disclosed in the present invention, based on the operating characteristics and mutual influence of wind turbines, photovoltaic power generation units, micro gas turbine units, and energy storage units, constructs an output model of flexible distributed resources in the virtual power plant to characterize the distributed resource operation domain. Secondly, considering the demand response behavior of the load, a comprehensive load model under the voltage static characteristics is constructed, which includes a constant power model, a constant current model, and a constant impedance model. Considering the existence of non-convex bilinear terms in the model, the feasible domain of a variable in the bilinear term is divided into several non-intersecting regions based on the piecewise McCormick relaxation method, and the model is converted into a mixed integer linear programming form. Thirdly, according to the branch flow model, the constraints of active power, reactive power, voltage, and current of the virtual power plant are set, and the nonlinear terms are processed based on the second-order cone relaxation method. Finally, considering the operating constraints of the above-mentioned virtual power plant, the virtual power plant-superior power grid coupling node connection power is used as the target observation variable, and the boundary points of the PQ flexible operation domain of the virtual power plant are obtained by adjusting the direction parameters to characterize the range of the flexible operation domain. In the flexible operation domain of the virtual power plant, the sum of the total operating cost and the demand response compensation is used as the objective function, and the mixed integer linear programming method is used to solve the virtual power plant optimization scheduling model considering demand response, and the obtained optimal operation scheduling strategy of the equipment is applied to the virtual power plant execution. The specific implementation process is as follows: Step 1: Based on the operating characteristics and mutual influence of the virtual power plant operating equipment, a virtual power plant output model of distributed resources in the virtual power plant is constructed to determine the operating domain of the distributed resources of the virtual power plant; Specifically, the virtual power plant operating equipment includes wind turbines, photovoltaic power generation units, micro gas turbines, energy storage units and demand response. The power flow balance constraints of the virtual power plant topology are analyzed based on the branch power flow model. At the same time, the node voltage and branch current safety constraints are considered to build a virtual power plant output model of distributed resources in the virtual power plant, and determine the distributed resource operation domain of the virtual power plant, including: Step 1.1 Operational and security constraints of flexible distributed resources (operating equipment) Step 1.1.1 Wind turbine The relationship between the power generation of a wind turbine and the wind speed and the operating constraints are as follows: (1) (2) (3) (4) in, , is the active and reactive power of the wind turbine generator set, , are the upper and lower limits of active power, , are the upper and lower limits of reactive power, is the power factor angle of the wind turbine generator set, is the rated power of the wind turbine, is the real-time wind speed, is the rated wind speed, and are the cut-in wind speed and the cut-off wind speed respectively.

[0023] Step 1.1.2 Photovoltaic power generation unit Ignoring the effect of temperature, the power generated by the photovoltaic power generation unit and the operating constraints can be expressed as: (5) (6) (7) in, , is the active and reactive power of the photovoltaic power generation unit, , are the upper and lower limits of active output of photovoltaic equipment, is the power factor angle of the wind turbine generator set, is the efficiency of the photovoltaic system, is the rated photovoltaic power, is the total irradiance of the sun actually incident on the photovoltaic array, It is the total radiation of the sun incident on the photovoltaic array under standard test conditions.

[0024] Step 1.1.3 Gas Turbine Unit Gas turbine units use natural gas as their main fuel, and the relationship between their actual active power and natural gas consumption is: (8) in, is the actual active power, The conversion efficiency of micro gas turbine power generation is is the calorific value of natural gas, is the volume of natural gas.

[0025] The micro gas turbine should also meet the following output constraints and climbing constraints: (9) (10) (11) (12) in, It is a 0-1 variable of the gas turbine power generation state, 1 means the gas turbine is in operation state, 0 means the gas turbine is in shutdown state; is the active power of the micro gas turbine, is the reactive power of the gas turbine, , are the upper and lower limits of active power, , are the upper and lower limits of reactive power, , The upper and lower limits of the climbing slope.

[0026] The operating cost of a gas turbine includes its fuel cost and start-up and shutdown cost, which can be expressed as follows: (13) (14) (15) in, For fuel costs, a , b , c is the cost coefficient, is the start-stop cost, Comprehensive cost of operating a gas turbine.

[0027] Step 1.1.4 Without considering the internal structure of the energy storage battery, the main constraints involved are its charging power, discharging power and state of charge, which can be expressed by the following formula.

[0028] During charging: (16) During discharge: (17) in, For the The remaining power of the time storage battery, For the The remaining power of the time storage battery, is the self-discharge rate of the energy storage battery, is the rated capacity of the energy storage battery, and are the charging power and discharging power of the energy storage battery respectively, and They are the charging efficiency and discharging efficiency of the energy storage battery respectively. The charging efficiency is usually 0.65~0.95, and the discharging efficiency is usually 1.

[0029] The maximum charge and discharge power allowed by the energy storage battery increases with The charge and discharge power at any time is not only related to It is also affected by the maximum continuous charging power of the energy storage battery. and maximum continuous discharge power The maximum continuous charging power and maximum continuous discharging power of the energy storage battery can be expressed as: (18) (19) in, The maximum allowable charging multiple is usually 0.5~1; The maximum allowable discharge multiple is usually 1~1.5; is the rated power of the battery.

[0030] The maximum charging power of the energy storage battery can be expressed by the following formula: (20) The maximum discharge power of the energy storage battery can be expressed by the following formula: (twenty one) in, The maximum remaining power is usually 0.8~0.9; It is the minimum remaining power, usually 0.1~0.2.

[0031] Step 2: Considering the impact of demand response load active power and reactive power adjustment on the virtual power plant output model, a comprehensive load model under voltage static characteristics is constructed; Step 2.1 Demand response behavior modeling Considering the impact of demand response load active and reactive power adjustment on the operation status of the virtual power plant, whether it is a response model based on electricity price or an incentive-based response model, the effect on the power grid can be characterized as the increment of node injection power: (twenty two) in, , , and are load demand and demand response, respectively; is the set of demand response nodes, t For demand response time, j It is the access node for demand response load.

[0032] The comprehensive load considered in the branch flow model is transformed accordingly. Obviously, in the load demand, the traditional idea of ​​only considering the constant power load is not accurate enough in the stage where technology is increasingly advanced and refined simulation is increasingly urgent. Therefore, a comprehensive load modeling under the static characteristics of voltage is constructed. The ZIP model is divided into a constant power model, a constant current model, and a constant impedance model. For the clarity of the expression, the time period mark is omitted .

[0033] (twenty three) (twenty four) (25) (26) in, , and is the ratio of constant impedance load, constant current load and constant power load; , is the power requirement at rated voltage, , are the upper and lower bounds of active power, , are the upper and lower bounds of reactive power, For Node j Voltage amplitude, , Node j The upper and lower bounds of the voltage amplitude are converted and rewritten according to the relaxation model to be related to the square of the voltage The relationship is as follows: (27) (28) For constant current load Since the voltage per unit value is near the rated voltage 1.0, the square term is also near the rated value 1.0. , , and the actual interval is smaller, not exceeding 0.1. Taylor series expansion is performed, and we get: (29) Ignoring the higher-order terms in the above formula, we get: (30) Will Substituting the variables in the above formula gives: (31) Step 2.2 Modeling of power flow balance constraints Factors such as the location layout and load distribution of flexible resources in a virtual power plant will directly affect the stability and reliability of power supply, so the topological structure of the virtual power plant is considered. The structure of the virtual power plant is basically the same as that of the distribution network, and the power flow balance constraints it meets are also similar to those of the distribution network, as shown in the following formula.

[0034] (32) (33) (34) (35) (36) in, i , j , k The node number in the virtual power plant, is the set of virtual power plant nodes, Node i is the set of branch end nodes of the head end node, Node i is the set of branch headend nodes of the terminal node, , The time periods t node i The net active power and net reactive power injected are , , The time periods t Flowing through the branch ij The active power, reactive power, current amplitude, , Branch ij The resistance and reactance of , , The time periods t Flowing through the branch ki The active power, reactive power, and current amplitude of , Branch ki The resistance and reactance of For the period t node i The voltage amplitude, , Node i Voltage upper and lower limits, For branch ij Current limit, , It is the node set and branch set of the virtual power plant.

[0035] Power after accounting for flexibility resources and load , It can be expressed as: (37) (38) in, , Access Node i The active and reactive power output of the photovoltaic power generation unit, , Access Node i The active and reactive power output of the wind power generation unit, , Access Node i The active (reactive) power output of the micro gas turbine, , are the charging and discharging active powers of the battery, , are the active and reactive power of the load after demand response, , They are the interactive active and reactive powers of the virtual power plant-distribution network coupling node, respectively.

[0036] Step 3: Based on the comprehensive load model, the piecewise McCormick relaxation method is used to divide the feasible domain of a variable in the bilinear term in the comprehensive load model into several non-overlapping regions, and the comprehensive load model is converted into a mixed integer linear programming form; Specifically, there are nonlinear constraints in both the demand response behavior model and the power flow balance constraint. According to the characteristics of different types of nonlinear terms, linearization methods based on piecewise McCormick relaxation and second-order cone relaxation are proposed respectively. They include: Step 3.1 Linearization of the demand response behavior model based on piecewise McCormick relaxation Due to the existence of bilinear terms and , the original demand response model is a non-convex model. The present invention first introduces auxiliary variables to establish a reconstructed model after reducing the bilinear terms, and then uses McCormick relaxation to convexify the bilinear terms and obtain the target lower bound of the reconstructed model. In order to improve the quality of McCormick relaxation, the piecewise McCormick method is used to derive tighter upper and lower bounds of the bilinear terms to reduce the volume of the feasible domain after McCormick relaxation. The piecewise McCormick method divides the domain of a variable in the bilinear term into several non-overlapping regions, and determines the optimal region, thereby tightening the bounds of the selected variable. In this way, the enhanced lower bound solution of the original problem is obtained. Specifically as follows: Step 3.1.1 Reconstruct the model By introducing auxiliary variables: (39) (40) in, , is an auxiliary variable.

[0037] Equation (31) can be transformed into: (41) In order to further transform the upper and lower bound constraints related to active and reactive power into constraints related to auxiliary variables, the following constraints are used: (42) (43) Therefore, the model of demand response behavior can be reformulated in the form of the following reconstructed model: (44) In the reconstructed model, non-convex quadratic constraints are transformed into linear constraints and independent bilinear constraints, and the reconstructed model is equivalent to the basic model.

[0038] Step 3.1.2 McCormick convex relaxation The McCormick relaxation method is used to replace the independent bilinear constraints in the reconstructed model, and the formula is as follows: (45) (46) Under McCormick relaxation, the reconstruction model becomes a convex problem, where the Karush-Kuhn-Tucker condition is necessary and sufficient (e.g., under Slater conditions). Therefore, a global minimum can be obtained in the relaxed McCormick model. This global minimum can be regarded as a lower bound of the reconstruction model. However, McCormick relaxation still produces relatively large errors in bilinear constraints.

[0039] Step 3.1.3 Segmented McCormick method In the piecewise McCormick method, the domain of a variable in the bilinear term is partitioned into a number of disjoint regions and the optimal region is determined, thereby tightening the bounds on the selected variable. The typical partitioning scheme is uniform partitioning, where the problem size increases linearly with the number of partitions. Other partitioning schemes with adaptive segment lengths or partition-dependent bounds can also be used for this problem to improve performance.

[0040] The variables that need to be partitioned will affect the quality of relaxation. In the virtual power plant demand response problem, the variables that can be selected are: node voltage, power demand, or a combination of the two. The present disclosure will choose to partition the power demand.

[0041] set up and Represents the partitions s Medium Variable The upper and lower bounds of the binary variable Assign to each partition s .if The value belongs to the partition s ,but ,otherwise, . Another variant of the bilinear term Decompose into , ,in, S yes s A collection of indexes.

[0042] Then the model after segmented McCormick processing can be expressed as: (47) (48) (49) (50) (51) (52) Similarly, setting and Represents the partitions s Medium Variable The upper and lower bounds of the binary variable Assign to each partition s .if The value belongs to the partition s ,but ,otherwise, . Another variant of the bilinear term Decompose into , ,in, S yes s A collection of indexes.

[0043] (53) (54) (55) (56) by Taking the piecewise transformation of as an example, in the above formula, if the binary variable, then the s All variables in the partitions, that is and Decide and In contrast, variables in all other partitions are forced to zero. In other words, if ,but s All constraints in the partition will be enforced, while constraints in all other partitions will be ignored. Increasing the number of binary variables will improve relaxation performance, but the resulting increase in mixed integer computations must also be addressed. Overall, the algorithm works well when the number of partitions is set to 3. Figure 1 shown.

[0044] Step 3.1.4 Boundary shrinkage algorithm The piecewise McCormick method tightens the upper and lower bounds of the active power, thereby enhancing the lower bound solution of the reconstructed model. In order to further reduce the violation error of the bilinear constraint, it is expected that a feasible solution can be found near the lower bound solution. Therefore, the present disclosure proposes a boundary contraction algorithm to iteratively strengthen the variable boundary and approach the optimal result with fewer violations of the bilinear constraints. The core idea of ​​the boundary contraction algorithm is: at the nth iteration, the upper and lower bounds of the decision variables are updated according to the n−1 iteration results and a series of hyperparameters ε (0<ε<1). The principle of setting the hyperparameter sequence is to gradually reduce the value of ε in order to tighten the boundary. In order to ensure the feasibility of the original problem, the updated boundary should be the intersection of the updated boundary and the original boundary. When the average relative error of the bilinear constraint reaches an acceptable level, the algorithm terminates. The purpose of the above method is to find a tighter lower bound of the objective function.

[0045] Step 3.2 Linearization of virtual power plant power flow model based on second-order cone relaxation Second-order cone programming is a special convex programming method with the characteristics of high solving efficiency and high accuracy.

[0046] (57) Among them, the variable ;coefficient , , , K is a second-order cone or rotated second-order cone constraint function, as shown below.

[0047] Second-order cone: (58) Rotate the second-order cone: (59) Cone Optimization Variables , , and then put it into the virtual power plant power flow model (32) (33) (34) to get: (60) (61) (62) Optimize the cone constraint relaxation: (63) Convert the above equation into a standard second-order cone form: (64) The optimal solution after the relaxation of the second-order cone programming must satisfy , otherwise the relaxation may be invalid at the optimal solution. If the voltage of any node at the optimal solution does not reach the set upper limit, and the reverse power flow of the radial network branch does not exist (satisfying the following formula), the second-order cone programming relaxation is accurate.

[0048] (65) Simplifying the above formula, we get: (66) because ,so ; Similarly, holds. So the second-order cone relaxation is exact.

[0049] Step 4: Considering the operating constraints of the virtual power plant, combined with the mixed integer linear programming form, the distribution network-virtual power plant coupling node connection power is used as the target observation variable of the virtual power plant operation domain. By adjusting the direction parameter to minimize, the boundary points of the virtual power plant operation domain are obtained to characterize the scope of the virtual power plant distributed resource operation domain; Specifically, based on the linearized model above, a method based on fitting to characterize the operation domain of the virtual power plant is proposed, and the adjustable capacity of the virtual power plant is described based on the piecewise linearization method, including: Step 4.1 Run domain characterization It can be seen from the above linearized model that the virtual power plant operation domain is based on satisfying the operation constraints and studies the feasibility space with the distribution network-virtual power plant transmission power as the key decision variable. In order to have a deeper understanding of the virtual power plant operation domain, its characteristics are analyzed as follows: 1) The virtual power plant operation domain uses the transmission power adjustment range at the virtual power plant grid connection point to represent the power adjustment range of internal resource equipment at each moment, under the premise of meeting the virtual power plant operation constraints, thus protecting the privacy of the internal data of the virtual power plant; 2) The virtual power plant operation point corresponds to a certain operating state of the virtual power plant. The virtual power plant operation domain is the set of operable states of the virtual power plant. The actual virtual power plant operation point is uniquely regulated by the distribution network. 3) The constructed virtual power plant operation domain model is universal and can adapt to virtual power plants of different types or configurations.

[0050] The key to solving the virtual power plant operation domain is: 1) The approximate accuracy of the flexible operation domain model needs to be guaranteed; 2) The computational efficiency of the flexible operation domain needs to be ensured.

[0051] Since the actual operating domain of the virtual power plant is unknown, in order to more accurately obtain the approximate data of its boundary sample points and ensure the balance between solution efficiency and accuracy, the present invention is based on the above-mentioned linearized model, and uses a radial iterative search algorithm to find the boundary points of the operating domain of the virtual power plant, and then uses the convex hull fitting method to obtain an observable virtual power plant operating domain image.

[0052] First, the operating constraints of the virtual power plant mentioned above are considered, and the power distribution network-virtual power plant coupling node connection power is adopted. , As the target observation variable of the virtual power plant operation domain, the direction parameter is minimized by adjusting , , to obtain the boundary points of the virtual power plant operation domain, the objective function As shown in the following formula, (67) in, , is the direction parameter, and its search direction and step size can be set according to the following formula: , They are the sets of equality and inequality constraints respectively: (68) in, is the initial search angle, is the step length, N is the total number of searches, and its value is related to the selection of the search step.

[0053] After obtaining a series of operating domain boundary sample points, the convex hull fitting method can be used to obtain the virtual power plant operating domain image, which is the minimum convex set containing all boundary sample points. Convex hull fitting is essentially a piecewise linearization of the nonlinear operating domain boundary, which can be implemented using MATLAB's ConvexHull function, which has a higher fitting accuracy.

[0054] Based on a series of boundary sample points, the convex hull piecewise linear fitting formula is used to describe the flexible adjustment capability of the virtual power plant, as shown in the following formula: (69) in, , They are the active power and reactive power obtained by the root node of the virtual power plant, is the set of convex hull vertices, For the z convex hull vertices; , Respectively z The active and reactive components of the convex hull vertices; For the corresponding z Segment coefficients; r is the number of convex hull vertices.

[0055] Step 5: Within the distributed resource operation domain of the virtual power plant, a virtual power plant optimization scheduling model is constructed with the sum of total operating cost and demand response compensation as the target, and the optimal operation scheduling strategy is applied to the virtual power plant execution using the solver.

[0056] Specifically, the virtual power plant optimization dispatch model is composed of an objective function of comprehensive cost and flexibility resource constraints, a demand response model after piecewise McCormick relaxation, a power flow balance constraint after second-order cone relaxation, and a safety constraint. The implementation process includes: Step 5.1: Construct the objective function In the construction and planning process of virtual power plants, economic efficiency is the most important consideration. Therefore, one of the main goals of optimizing scheduling is to minimize the comprehensive cost. The calculation formula of the comprehensive cost is as follows: (70) in, For the comprehensive cost, The equipment purchase cost is For operating costs, is the cost of purchasing electricity from the distribution grid, Reimbursement costs for demand response.

[0057] (1) Equipment purchase cost Acquisition cost Refers to the total initial investment cost of the main equipment of the virtual power plant manager. In order to comprehensively consider the impact of equipment life on the virtual power plant, the purchase cost is converted into an annualized equivalent investment amount. The formula is as follows: (71) (72) in, It is a collection of photovoltaic power generation units, wind turbines and other equipment. Yes Equipment The acquisition cost, Yes Equipment The number of purchases, is the capital recovery factor, is the base interest rate, is the lifespan of the virtual power plant.

[0058] (2) Operating costs Running costs It refers to the cost incurred by the power generation unit during the operation of the virtual power plant. The formula is as follows: (73) in, , , are the cost coefficients of photovoltaic power generation unit, wind turbine generator set and energy storage unit respectively. is the operating cost of the micro gas turbine, , Node i The output power of photovoltaic power generation units and wind turbines, , Node i The charging and discharging active power of the energy storage unit, A collection of virtual power plant nodes.

[0059] (3) Cost of purchasing electricity Cost of purchasing electricity It refers to the cost of the virtual power plant purchasing electricity from the distribution network, which is expressed by the active power cost at the virtual power plant-distribution network coupling node: (74) in, for t The electricity price of the distribution network during the period, , is the interactive active power at the virtual power plant-distribution network coupling node. Based on the convex hull piecewise linearization constraint (69), the calculation formula can be obtained based on a series of convex hull vertices as follows: (75) (4) Demand response compensation costs (76) in, is the unit power demand response compensation, is the original load demand, is the load after demand response.

[0060] The above objective function composed of comprehensive cost and flexible resource constraints, demand response model after piecewise McCormick relaxation, power flow balance constraints after second-order cone relaxation, and safety constraints together constitute the optimization scheduling model based on the virtual power plant operation domain. This model is a mixed integer linear programming model that can be directly solved using the CPLEX commercial solver.

[0061] As an example, Figure 1 As shown, the virtual power plant scheduling method based on distributed resource operation domain characterization disclosed in the present invention should include: Select a typical day to set the total scheduling cycle to 24 hours and the time interval to 1 hour. In the initialization stage, first enter the basic parameters of the supply-side flexibility resources, including the technical parameters of photovoltaic power generation units, wind turbines, micro gas turbines and energy storage units. Secondly, enter the photovoltaic power generation, wind power generation, load demand and electricity price forecast curves of typical days to provide data support for subsequent optimization scheduling. Finally, enter the topological structure parameters of the virtual power plant, including the upper and lower limit constraints of node voltage and line flow, to ensure the safe operation of the power system. On this basis, the segmented McComick relaxation method is used to linearize the demand response model to improve the solution efficiency of the model. At the same time, the power flow balance model is linearized by the second-order cone relaxation method to further simplify the complexity of the optimization problem. After completing the linearization of the model, the system will characterize the flexible operation domain of the virtual power plant and clarify the feasible operation range of each device in the scheduling process. Subsequently, the CPLEX solver is used to solve the linearized virtual power plant optimization scheduling model to obtain the optimal operation strategy of each device. These strategies will be sent to the corresponding equipment for execution, enabling the virtual power plant to operate economically, efficiently and safely on a typical day.

[0062] Example 2 In one embodiment of the present disclosure, a virtual power plant scheduling system based on distributed resource operation domain characterization is provided, including: A virtual power plant model building module is used to build a virtual power plant output model of distributed resources in the virtual power plant based on the operating characteristics and mutual influence of the virtual power plant operating equipment, and determine the operating domain of the distributed resources of the virtual power plant; The relaxation linearization module is used to consider the impact of demand response load active power and reactive power adjustment on the virtual power plant output model and construct a comprehensive load model under voltage static characteristics. Based on the comprehensive load model, the piecewise McCormick relaxation method is used to divide the feasible domain of a variable in the bilinear term in the comprehensive load model into several non-overlapping regions, and the comprehensive load model is converted into a mixed integer linear programming form. The operating domain boundary characterization module is used to consider the operating constraints of the virtual power plant. In combination with the mixed integer linear programming form, the distribution network-virtual power plant coupling node connection power is used as the target observation variable of the virtual power plant operating domain. By adjusting the direction parameter to minimize, the boundary points of the virtual power plant operating domain are obtained to characterize the scope of the virtual power plant distributed resource operating domain. The scheduling solution module is used to build a virtual power plant optimization scheduling model within the distributed resource operation domain of the virtual power plant with the sum of total operating cost and demand response compensation as the target, and use the solver to solve and apply the obtained optimal operation scheduling strategy to the virtual power plant execution.

[0063] Example 3 In one embodiment of the present disclosure, a computer program product is provided, including a computer program, which, when executed by a processor, implements the virtual power plant scheduling method based on distributed resource operation domain characterization.

[0064] Example 4 In one embodiment of the present disclosure, a non-transitory computer-readable storage medium is provided, wherein the non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the virtual power plant scheduling method based on distributed resource operation domain characterization is implemented.

[0065] Example 5 In one embodiment of the present disclosure, an electronic device is provided, including: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes the virtual power plant scheduling method based on the distributed resource operation domain characterization.

[0066] The present disclosure is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present disclosure. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0067] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0068] Although the above describes the specific implementation methods of the present disclosure in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present disclosure. Technical personnel in the relevant field should understand that on the basis of the technical solution of the present disclosure, various modifications or variations that can be made by those skilled in the art without creative work are still within the scope of protection of the present disclosure.

Claims

1. A virtual power plant scheduling method based on distributed resource operation domain characterization, characterized in that: include: Based on the operating characteristics and mutual influence of the virtual power plant operating equipment, a virtual power plant output model of distributed resources in the virtual power plant is constructed to determine the operating domain of the distributed resources in the virtual power plant; Considering the impact of demand response load active power and reactive power adjustment on the virtual power plant output model, a comprehensive load model under voltage static characteristics is constructed; Based on the comprehensive load model, the piecewise McCormick relaxation method is used to divide the feasible domain of a variable in the bilinear term of the comprehensive load model into several non-overlapping regions, and the comprehensive load model is transformed into a mixed integer linear programming form. Considering the operation constraints of virtual power plants, combined with mixed integer linear programming, the power of the distribution network-virtual power plant coupling node is used as the target observation variable of the virtual power plant operation domain. By adjusting the direction parameter to minimize, the boundary points of the virtual power plant operation domain are obtained to characterize the scope of the virtual power plant distributed resource operation domain. Within the distributed resource operation domain of the virtual power plant, an optimization scheduling model of the virtual power plant is constructed with the sum of total operating cost and demand response compensation as the target, and the solver is used to solve the optimal operation scheduling strategy, which is applied to the virtual power plant execution.

2. The virtual power plant scheduling method based on distributed resource operation domain characterization according to claim 1 is characterized in that: The virtual power plant operating equipment includes wind turbines, photovoltaic power generation units, micro gas turbine units, energy storage units and demand response. The flow balance constraints of the virtual power plant topology are analyzed based on the branch flow model. While considering the node voltage and branch current safety constraints, a virtual power plant output model of distributed resources in the virtual power plant is constructed to determine the distributed resource operation domain of the virtual power plant.

3. The virtual power plant scheduling method based on distributed resource operation domain characterization according to claim 1 is characterized in that: Considering the impact of active power and reactive power adjustment of demand response load on the output model of virtual power plant, a comprehensive load model under voltage static characteristics is constructed. There are nonlinear constraints in both demand response behavior and power flow balance constraints in the comprehensive load model. According to the characteristics of different types of nonlinear terms, the demand response behavior model is linearized based on the piecewise McCormick relaxation linearization method. Auxiliary variables are introduced to establish a reconstructed model after reducing bilinear terms. The piecewise McCormick relaxation is used to convexify the bilinear terms, and the target lower bound of the reconstructed model is obtained. Tighter upper and lower bounds of the bilinear terms are derived to reduce the volume of the feasible domain after McCormick relaxation. The domain of a variable in the bilinear term is divided into several non-overlapping regions, and the optimal region is determined, thereby tightening the boundaries of the selected variables. The power flow balance constraints and safety constraints are relaxed by the second-order cone relaxation linearization method and converted into a mixed integer linear programming form.

4. The virtual power plant scheduling method based on distributed resource operation domain characterization according to claim 1 is characterized in that: Based on the mixed integer linear programming form, the boundary points of the virtual power plant operation domain are found through the radial iterative search algorithm, and then the convex hull fitting method is used to obtain the observable virtual power plant operation domain image. The operation constraints of the virtual power plant are considered, and the distribution network-virtual power plant coupling node connection power is used as the target observation variable of the virtual power plant operation domain. The boundary points of the virtual power plant operation domain are obtained by minimizing the direction parameters. After the boundary sample points of the operation domain are obtained, the convex hull fitting method is used to obtain the virtual power plant operation domain image, which is the minimum convex set containing all boundary sample points.

5. The virtual power plant scheduling method based on distributed resource operation domain characterization according to claim 1 is characterized in that: Within the distributed resource operation domain of the virtual power plant, the virtual power plant optimization scheduling model is constructed with the sum of total operating cost and demand response compensation as the target. The objective function is as follows: in, For the comprehensive cost, The equipment purchase cost is For operating costs, is the cost of purchasing electricity from the distribution grid, Reimbursement costs for demand response.

6. The virtual power plant scheduling method based on distributed resource operation domain characterization according to claim 1 is characterized in that: The virtual power plant optimization scheduling model is composed of an objective function consisting of comprehensive cost and flexibility resource constraints, a demand response model after segmented McCormick relaxation, a power flow balance constraint after second-order cone relaxation, and a safety constraint.

7. A virtual power plant dispatching system based on distributed resource operation domain characterization is characterized by: include: A virtual power plant model building module is used to build a virtual power plant output model of distributed resources in the virtual power plant based on the operating characteristics and mutual influence of the virtual power plant operating equipment, and determine the operating domain of the distributed resources of the virtual power plant; The relaxation linearization module is used to consider the impact of demand response load active power and reactive power adjustment on the virtual power plant output model and construct a comprehensive load model under voltage static characteristics. Based on the comprehensive load model, the piecewise McCormick relaxation method is used to divide the feasible domain of a variable in the bilinear term in the comprehensive load model into several non-overlapping regions, and the comprehensive load model is converted into a mixed integer linear programming form. The operating domain boundary characterization module is used to consider the operating constraints of the virtual power plant. In combination with the mixed integer linear programming form, the distribution network-virtual power plant coupling node connection power is used as the target observation variable of the virtual power plant operating domain. By adjusting the direction parameter to minimize, the boundary points of the virtual power plant operating domain are obtained to characterize the scope of the virtual power plant distributed resource operating domain. The scheduling solution module is used to build a virtual power plant optimization scheduling model within the distributed resource operation domain of the virtual power plant with the sum of total operating cost and demand response compensation as the target, and use the solver to solve and apply the obtained optimal operation scheduling strategy to the virtual power plant execution.

8. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the virtual power plant scheduling method based on distributed resource operation domain characterization as described in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by the processor, the virtual power plant scheduling method based on distributed resource operation domain characterization as described in any one of claims 1-6 is implemented.

10. An electronic device, characterized in that: include: A processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device executes the virtual power plant scheduling method based on distributed resource operation domain characterization as described in any one of claims 1-6.

Images