Distributed optimal scheduling method and system for AC / DC power distribution network

Abstract

The invention provides a distributed optimal scheduling method and system for an AC / DC power distribution network. The method comprises the following steps: acquiring load data, distributed power supply output data and energy storage state data of a local node based on a first local intelligent controller; performing a local optimization decision based on the first local intelligent controller according to the local optimization objective function in combination with the load data, the distributed power supply output data and the energy storage state data to obtain a preliminary local scheduling decision; based on each first local intelligent controller, performing consistency collaborative optimization according to the first interaction information and second interaction information of the second local intelligent controller, and determining a target local scheduling decision; and performing distributed collaborative optimization according to the target local scheduling decision of the target local intelligent controller based on each regional coordinator, determining a node scheduling decision of a node corresponding to the target local intelligent controller, and realizing peak-valley difference optimal scheduling of the AC / DC power distribution network. The overall operation efficiency and stability of the power grid are improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and particularly to a distributed optimal scheduling method and system for an AC-DC distribution network. Background Art

[0002] With the continuous development of the power system and the wide access of various distributed power sources, energy storage devices, DC loads, etc. in the distribution network, the operation of the AC-DC distribution network has become increasingly complex. During the operation of the power system, the problem of peak-valley difference of the load brings many challenges to the safe, stable and economic operation of the power grid.

[0003] At present, the main distributed peak-valley difference methods for AC-DC distribution networks are centralized control-based scheduling methods, that is, relying on the control center of the power grid to collect various information within the entire distribution network area, including load data of each node, power generation power of distributed power sources, and the state of energy storage devices, etc., and then formulating a unified scheduling plan through complex optimization algorithms to regulate the output of distributed power sources, the charging and discharging behavior of energy storage, etc., in an attempt to flatten the peak-valley difference. On the one hand, it highly depends on the transmission of a large amount of accurate real-time information, and has extremely high requirements for the bandwidth and stability of the communication network. Once communication failures, data packet loss or delays occur, the information obtained by the control center will be distorted, resulting in the failure of the formulated scheduling strategy and being unable to effectively respond to the peak-valley difference changes. On the other hand, the centralized control mode lacks detailed consideration of the characteristics of each local area and different nodes, is prone to the situation of "one-size-fits-all", and is difficult to give full play to the advantages of each distributed power source and energy storage in different scenarios, and cannot achieve refined peak-valley difference optimization. Summary of the Invention

[0004] To solve the technical problems of the existing methods, the present invention provides a distributed optimal scheduling method and system for an AC-DC distribution network, so as to accurately and finely optimize and schedule the peak-valley difference of the AC-DC distribution network, better respond to load changes, realize the reasonable allocation and utilization of resources, and improve the overall operation efficiency and stability of the AC-DC distribution network.

[0005] In the first aspect, the present invention provides a distributed optimal scheduling method for an AC-DC distribution network, which is implemented based on a distributed optimal scheduling system for an AC-DC distribution network. The distributed optimal scheduling system for an AC-DC distribution network at least includes local intelligent controllers and area coordinators, and each area coordinator is connected to the local intelligent controllers within a preset range through a communication link; the local intelligent controllers are deployed at each node of the AC-DC distribution network, and the nodes include distributed power source nodes, energy storage device nodes and load nodes; the method includes: Collecting load data, distributed power source output data and energy storage state data of the local node where each first local intelligent controller is located; Based on each first local intelligent controller making local optimization decisions according to the local optimization objective function in combination with load data, distributed power generation output data, and energy storage state data, to obtain the preliminary local scheduling decisions of each first local intelligent controller; Based on each first local intelligent controller performing consistency collaborative optimization according to the first interaction information and the second interaction information of the second local intelligent controller, to determine the target local scheduling decisions of each first local intelligent controller; Based on each regional coordinator performing distributed collaborative optimization according to the target local scheduling decisions of the target local intelligent controller, to determine the node scheduling decisions of the nodes corresponding to the target local intelligent controller, and to achieve peak-valley difference optimized scheduling of the AC-DC distribution network; The local optimization objective function is a function constructed with the objectives of minimizing the peak-valley difference, maximizing the local energy self-sufficiency rate, and minimizing the energy storage life loss, in combination with the load characteristics of the local load nodes, the power regulation characteristics of the distributed power generation nodes, and the charge-discharge characteristics of the energy storage device nodes; the second local intelligent controller is the controller corresponding to the node adjacent to the local node; the interaction information includes the preliminary local scheduling decisions and the local resource margin information; the target local intelligent controller is the first local intelligent controller connected to each regional coordinator.

[0006] In a second aspect, the present invention also provides an AC-DC distribution network distributed optimization scheduling system, which at least includes local intelligent controllers and regional coordinators, and each regional coordinator is connected to the local intelligent controllers within a preset range through a communication link; the local intelligent controllers are deployed at each node of the AC-DC distribution network, and the nodes include distributed power generation nodes, energy storage device nodes, and load nodes, and are used to implement the AC-DC distribution network distributed optimization scheduling method as described in the first aspect; The local intelligent controller is used for: Collecting the load data, distributed power generation output data, and energy storage state data of the local node where it is located; Making local optimization decisions according to the local optimization objective function in combination with the load data, distributed power generation output data, and energy storage state data, to obtain the preliminary local scheduling decisions of each first local intelligent controller Performing consistency collaborative optimization according to the first interaction information and the second interaction information of the second local intelligent controller, to determine the target local scheduling decisions of each first local intelligent controller; The regional coordinator is used for: Performing distributed collaborative optimization according to the target local scheduling decisions of the target local intelligent controller, to determine the node scheduling decisions of the nodes corresponding to the target local intelligent controller, and to achieve peak-valley difference optimized scheduling of the AC-DC distribution network; The local optimization objective function aims to minimize the peak-valley difference, maximize the local energy self-sufficiency rate, and minimize the energy storage life loss. It is a function constructed by combining the load characteristics of local load nodes, the power regulation characteristics of distributed power generation nodes, and the charge and discharge characteristics of energy storage device nodes. The second local intelligent controller is the controller corresponding to the node adjacent to the local node. The interaction information includes the preliminary local scheduling decision and the local resource margin information. The target local intelligent controller is the first local intelligent controller connected to each area coordinator.

[0007] In a third aspect, the present invention also provides an electronic device, including: a memory for storing a computer software program; a processor for reading and executing the computer software program to implement the AC-DC distribution network distributed optimization scheduling method as described in any one of the above.

[0008] In a fourth aspect, the present invention also provides a non-transitory computer-readable storage medium, in which a computer software program is stored. When the computer software program is executed by a processor, it implements the AC-DC distribution network distributed optimization scheduling method as described in any one of the above.

[0009] In a fifth aspect, the present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the AC-DC distribution network distributed optimization scheduling method as described in any one of the above.

[0010] For the AC-DC distribution network distributed optimization scheduling method provided by the embodiments of the present invention, the first local intelligent controller can comprehensively consider various characteristics of local nodes, and make optimization decisions based on the constructed local optimization objective function and the load data, distributed power generation output data, and energy storage state data of local nodes. Therefore, it can provide reasonable local scheduling decisions for the efficient and stable operation of the AC-DC distribution network according to its own actual situation. Further, the area coordinator performs distributed collaborative optimization on the local scheduling decisions of the local intelligent controllers to determine the optimal scheduling decisions for each local intelligent controller, achieving accurate and fine optimization scheduling of the peak-valley difference of the AC-DC distribution network, better coping with load changes, realizing reasonable allocation and utilization of resources, and improving the overall operation efficiency and stability of the AC-DC distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 is a flowchart of the AC-DC distribution network distributed optimization scheduling method provided by the embodiments of the present invention; Figure 2 is a structural diagram of the AC-DC distribution network distributed optimization scheduling system provided by the embodiments of the present invention; Figure 3 is an embodiment diagram of the electronic device provided by the embodiments of the present invention; Figure 4Embodiment diagram of the computer-readable storage medium provided by the embodiments of the present invention. Detailed implementation manners

[0012] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present invention.

[0013] In the description of the present invention, the terms "first" and "second" are only used for descriptive purposes, and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of the described features. In the description of the present invention, "a plurality of" means two or more, unless otherwise specifically defined.

[0014] In the description of the present invention, the term "for example" is used to mean "serving as an example, illustration, or explanation". Any embodiment described as "for example" in the present invention is not necessarily construed as being more preferred or more advantageous than other embodiments. In order for any person skilled in the art to implement and use the present invention, the following description is given. In the following description, details are set forth for the purpose of explanation. It should be understood that those skilled in the art can recognize that the present invention can be implemented without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid unnecessary details from obscuring the description of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope that conforms to the principles and features disclosed in the present invention.

[0015] Refer to Figure 1 , Figure 1 is a flowchart of the distributed optimal scheduling method for an AC / DC distribution network provided by the present invention. In the embodiments of the present invention, the execution subject of the distributed optimal scheduling method for the AC / DC distribution network is an optimal scheduling system. The optimal scheduling system includes at least a local intelligent controller and a regional coordinator. Each regional coordinator is connected to the local intelligent controllers within a preset range through a communication link. The local intelligent controllers are deployed at each node of the AC / DC distribution network. The nodes include distributed power nodes, energy storage device nodes, and load nodes. Therefore, the distributed optimal scheduling method for the AC / DC distribution network includes: Step 10, based on each first local intelligent controller, collect the load data, distributed power output data, and energy storage state data of the local node where it is located.

[0016] Optionally, for each first local intelligent controller, the first local intelligent controller collects key data on its local node, and the key data specifically includes load data , representing the load conditions at different times, in kilowatts; distributed power generation output data , representing the power generation power of distributed power sources at different times, in kilowatts; energy storage state data , representing the power storage of the energy storage, in kilowatt-hours.

[0017] Furthermore, in order to ensure the stability of the AC / DC distribution network, it is necessary to monitor the stability of each node at different times. Therefore, the first local intelligent controller performs node energy balance analysis based on the collected data to obtain the node energy balance coefficient, that is, to analyze the dynamic energy balance situation of the local node at different times. Among them, the node energy balance coefficient The specific analysis formula is as follows: .

[0018] Among them, represents the rate of change of the energy storage power with time.

[0019] Furthermore, based on the energy storage state data, a comprehensive evaluation and analysis of the energy storage life and efficiency is performed to obtain the comprehensive energy storage impact factor. The specific analysis formula of the comprehensive energy storage impact factor is as follows: .

[0020] Among them, represents the charging efficiency of the energy storage device (the value range is between 0 and 1), represents the discharging efficiency of the energy storage device (the value range is between 0 and 1).

[0021] Furthermore, based on the distributed power generation output data, a stability and load matching analysis of the distributed power generation output is performed to obtain the distributed power generation output stability coefficient , and the specific analysis formula is as follows: .

[0022] Among them, represents the length of the time window of a preset length, such as 1 hour; represents the average output of the distributed power source within this time window.

[0023] Furthermore, the first local intelligent controller calculates the comprehensive stability coefficient according to the point energy balance coefficient , the comprehensive energy storage impact factor and the distributed power generation output stability coefficient , where the comprehensive stability coefficient has the following specific formula: .

[0024] If the comprehensive stability coefficient , it is determined that the current node is in a stable state. If the comprehensive stability coefficient , it is determined that the current node is in an unstable state and needs further maintenance.

[0025] Step 20: Based on each first local intelligent controller, local optimization decisions are made according to the local optimization objective function in combination with load data, distributed power generation output data, and energy storage state data, and preliminary local scheduling decisions for each first local intelligent controller are obtained.

[0026] Optionally, a pre-constructed local optimization objective function is stored in the first local intelligent controller of the embodiment of the present invention. Among them, the local optimization objective function is a function constructed with the objectives of minimizing the peak-valley difference, maximizing the local energy self-sufficiency rate, and minimizing the energy storage life loss, in combination with the load characteristics of local load nodes, the power regulation characteristics of distributed power generation nodes, and the charge and discharge characteristics of energy storage device nodes. The specific process of constructing the local optimization objective function is as described in Steps 201 to 204.

[0027] Among them, the load characteristics of load nodes include the load characteristics of industrial user concentration areas and the load characteristics of residential user concentration areas. The power regulation characteristics of distributed power generation nodes include the output of photovoltaic power generation and the output of wind power generation. The charge and discharge characteristics of energy storage device nodes include the charge and discharge efficiency of lithium batteries and the charge and discharge efficiency of flow batteries.

[0028] Furthermore, the first local intelligent controller uses a MINLP (Mixed Integer Nonlinear Programming) solver to perform local optimization decisions according to the local optimization objective function in combination with the collected load data, distributed power generation output data, and energy storage state data. The MINLP solver will search for a set of decision variable values that make the local optimization objective function reach the optimal value within the feasible solution space that satisfies all constraint conditions, and obtain the preliminary local scheduling decision of the first local intelligent controller.

[0029] Step 30: Based on each first local intelligent controller, consistency collaborative optimization is performed according to the first interaction information and the second interaction information of the second local intelligent controller, and the target local scheduling decision of each first local intelligent controller is determined.

[0030] Optionally, the second local intelligent controller in the embodiments of the present invention is a controller corresponding to a node adjacent to the local node. The interaction information includes a preliminary local scheduling decision and local resource margin information, such as the remaining adjustable power margin of distributed power sources and the remaining charge-discharge capacity margin of energy storage.

[0031] Further, the first local intelligent controller performs consistency collaborative optimization based on its own first preliminary local scheduling decision and first local resource margin information, as well as the second preliminary local scheduling decision and second local resource margin information of the second local intelligent controller, to determine the target local scheduling decision of the first local intelligent controller, as specifically described in steps 301 to 303.

[0032] Step 40: Based on each regional coordinator performing distributed collaborative optimization according to the target local scheduling decision of the target local intelligent controller, determine the node scheduling decision of the node corresponding to the target local intelligent controller, so as to achieve peak-valley difference optimal scheduling of the AC-DC distribution network.

[0033] Further, for each regional coordinator, the regional coordinator takes the first local intelligent controller connected to it as the target local intelligent controller, performs distributed collaborative optimization according to the target local scheduling decision of the target local intelligent controller, and determines the node scheduling decision of the node corresponding to the target local intelligent controller. Among them, the node scheduling decision includes the output behavior of distributed power sources and the charge-discharge behavior of energy storage. Therefore, the regional coordinator controls the target local intelligent controller to execute the output behavior of distributed power sources and the charge-discharge behavior of energy storage in the node scheduling decision, so as to achieve peak-valley difference optimal scheduling of the AC-DC distribution network, as specifically described in steps 401 to 405.

[0034] The first local intelligent controller in the embodiments of the present invention can comprehensively consider various characteristics of local nodes, make optimization decisions based on the constructed local optimization objective function and the load data, distributed power output data, and energy storage state data of local nodes, and can provide reasonable local scheduling decisions for the efficient and stable operation of the AC-DC distribution network according to its own situation. Further, the regional coordinator performs distributed collaborative optimization on the local scheduling decisions of local intelligent controllers to determine the optimal scheduling decision for each local intelligent controller, achieving accurate and refined optimal scheduling of the peak-valley difference of the AC-DC distribution network, better coping with load changes, realizing reasonable allocation and utilization of resources, and improving the overall operation efficiency and stability of the AC-DC distribution network.

[0035] In one embodiment, the descriptions of steps 201 to 204 are as follows: Step 201: Determine the total load characteristic based on the first load curve corresponding to the load characteristic of the industrial user concentrated area and the second load curve corresponding to the load characteristic of the residential user concentrated area.

[0036] Optionally, for the load characteristics in the industrial user concentration area, with a 24-hour cycle, the time variable , in hours, and its first load curve can be expressed as: .

[0037] Among them, represents the reference capacity of the industrial load, characterizing the approximate scale of the industrial load in this area; represents the parameter of the industrial load's decline rate at night, characterizing the characteristic that the load rapidly decreases due to the shutdown of industrial users at night.

[0038] Optionally, for the load characteristics in the residential user concentration area, its second load curve can be expressed as: .

[0039] Among them, represents the reference capacity of the residential load, represents the influence parameters of the residential load during the noon and night low load periods.

[0040] Therefore, the total load characteristic in the comprehensive area can be expressed as: .

[0041] Among them, and are the weight coefficients of the industrial load and the residential load in the total load respectively.

[0042] Step 202: Determine the distributed power output based on the first power regulation characteristic corresponding to the output of the photovoltaic power source and the second power regulation characteristic corresponding to the output of the wind power source.

[0043] Optionally, the distributed power source node characteristics in the embodiments of the present invention include the output of the photovoltaic power source and the output of the wind power source. Among them, the output of the photovoltaic power source is related to the light intensity (unit is ), the ambient temperature (unit is ), etc. Its first power regulation characteristic can be expressed as: .

[0044] Among them, represents the rated power of the photovoltaic module under standard test conditions (STC), and represent the light intensity and temperature under standard test conditions respectively, represents the light intensity threshold.

[0045] Optional, wind power output Depends on wind speed (Unit is ) and the fan cut-in speed , rated wind speed , Cut-out wind speed Therefore, the second power regulation characteristic of the wind power output can be expressed as: .

[0046] Therefore, the comprehensive distributed power output It can be expressed as: .

[0047] in, and They are the weight coefficients of photovoltaic power and wind power in the distributed power supply combination, satisfying .

[0048] Step 203 : determining the energy capacity change rate based on the lithium battery capacity change corresponding to the charge and discharge efficiency of the lithium battery and the energy storage capacity change rate corresponding to the charge and discharge efficiency of the flow battery.

[0049] Optionally, the node characteristics of the energy storage device in the embodiment of the present invention include the charge and discharge efficiency of the lithium battery and the charge and discharge efficiency of the flow battery, wherein the charge and discharge efficiency of the lithium battery varies under different charge and discharge powers and states of charge (SOC). It can be expressed as: .

[0050] in, It indicates the benchmark charging efficiency of lithium batteries when they are charged at an ideal low power and in the best state of charge. Indicates the charging power of the lithium battery. Indicates the maximum charging power of the lithium battery. Indicates the real-time charge status of the lithium battery. Indicates the optimal state of charge of the lithium battery. and The maximum and minimum state of charge allowed for lithium batteries are shown respectively.

[0051] Similarly, the discharge efficiency can be obtained .

[0052] Therefore, the charge of the lithium battery changes It can be expressed as: .

[0053] Among them, represents the discharge power.

[0054] Optionally, for the charge-discharge efficiency of the flow battery, the charge efficiency of the flow battery is related to the charge electrolyte flow rate (in the unit of ) and the charge current density (in the unit of ), and the charge efficiency can be expressed as: .

[0055] Among them, represents the reference charge efficiency of the flow battery, and respectively represent the optimal electrolyte flow rate and the optimal current density of the flow battery, , , and are the corresponding limit parameters.

[0056] Similarly, the discharge efficiency can be obtained.

[0057] Therefore, the change rate of the stored energy can be expressed as: .

[0058] Among them, represents the charge power of the flow battery, represents the discharge power of the flow battery, represents the discharge efficiency of the flow battery.

[0059] Therefore, the comprehensive charge-discharge characteristics, that is, the change rate of the stored energy can be expressed as: .

[0060] Among them, and respectively represent the weight coefficients of the lithium battery energy storage and the flow battery energy storage in the total energy storage, satisfying .

[0061] Step 204, construct a local optimization objective function based on the total load characteristics, distributed power generation output, and the change rate of the stored energy.

[0062] Further, a peak-valley difference minimization objective function is constructed according to the total load characteristics, an energy self-sufficiency rate maximization objective function is constructed according to the distributed power generation output and the total load characteristics, and a storage life loss minimization objective function is constructed according to the rate of change of energy quantity. Further, according to the peak-valley difference minimization objective function, the energy self-sufficiency rate maximization objective function, and the storage life loss minimization objective function, a local optimization objective function is constructed, as specifically described in Steps 2041 to 2044.

[0063] The first local intelligent controller in the embodiment of the present invention can comprehensively consider various characteristics of the local node, and make an optimization decision based on the constructed local optimization objective function, as well as the load data, distributed power generation output data, and energy storage state data of the local node. Therefore, it can provide a reasonable local scheduling decision for the efficient and stable operation of the AC / DC distribution network according to its actual situation.

[0064] In one embodiment, the descriptions of Steps 2041 to 2044 are as follows: Step 2041: Determine the peak-valley difference based on the total load characteristics, and construct a peak-valley difference minimization objective function based on the peak-valley difference.

[0065] Optionally, for the peak-valley difference minimization objective function, first calculate the peak-valley difference of the load curve within a certain time period (such as one day), and the specific formula is as follows: .

[0066] Further, in order to better incorporate into the optimization objective function, normalize the peak-valley difference and construct a peak-valley difference minimization objective function. The specific formula of the peak-valley difference minimization objective function is as follows: .

[0067] Wherein, represents the average load within the time period .

[0068] Step 2042: Based on the total load characteristics and the distributed power generation output, respectively determine the total energy generated by the local distributed power generation and the energy consumed by the local total load, and construct a local energy self-sufficiency rate maximization objective function based on the total energy generated by the local distributed power generation and the energy consumed by the local total load.

[0069] Optionally, for the local energy self-sufficiency rate maximization objective function, the local energy self-sufficiency rate is the ratio of the total energy generated by the local distributed power generation to the energy consumed by the local total load. Within the time period , the calculation formula of the total energy generated by the local distributed power generation is as follows: .

[0070] Local total load energy consumption The calculation formula is as follows: .

[0071] Therefore, the objective function for maximizing local energy self-sufficiency is constructed as follows: .

[0072] in, Represents the adjustment coefficient.

[0073] Step 2043: Determine a first life loss index of lithium battery energy storage and a second life loss index of flow battery energy storage based on the energy and charge change rate, and construct an energy storage life loss minimization objective function based on the first life loss index and the second life loss index.

[0074] Optionally, for the objective function of minimizing the energy storage life loss, it is considered that the energy storage life loss is related to the number of charge and discharge times, charge and discharge depth, and characteristics of different energy storage technologies.

[0075] For lithium battery energy storage, the first life loss indicator of lithium battery energy storage is It can be expressed as: .

[0076] in, Indicates that lithium batteries are The change in charge and discharge power at each moment, Indicates the rated capacity of the lithium battery. Indicates the coefficient that affects the relationship between the life loss and state of charge of lithium batteries. represents the average state of charge, Indicates that lithium batteries are The cumulative value of charge and discharge times at the moment.

[0077] For liquid flow battery energy storage, the second life loss index of liquid flow battery energy storage It can be expressed as: .

[0078] in, Indicates that flow batteries The change in charge and discharge power at each moment, Indicates the rated capacity of the flow battery, The coefficient that affects the relationship between the life loss and state of charge of the flow battery, represents the average state of charge, Indicates that flow batteries The cumulative value of charge and discharge cycles at a certain moment.

[0079] Therefore, the comprehensive energy storage life loss index can be expressed as: .

[0080] Among them, and are the weight coefficients of lithium battery energy storage and flow battery energy storage considering life loss respectively, satisfying .

[0081] Therefore, the objective function for minimizing energy storage life loss is: .

[0082] Step 2044, based on the objective function of minimizing the peak-valley difference, the objective function of maximizing the local energy self-sufficiency rate, and the objective function of minimizing the energy storage life loss, construct the local optimization objective function.

[0083] Combining the above three objective functions, the constructed local optimization objective function is: .

[0084] Optionally, in the local optimization decision-making process of the embodiment of the present invention, constraint conditions are set. Among them, for the power balance constraint: at any moment , it is satisfied that the total output of the distributed power source and the sum of the charge and discharge power changes of the energy storage device are equal to the local load demand, that is: .

[0085] The distributed power source power regulation range constraint includes photovoltaic power source and wind power source constraints. Among them, for the photovoltaic power source constraint: . For the wind power source: . Among them, and are the maximum allowable power outputs of the photovoltaic power source and the wind power source respectively.

[0086] The energy storage device charge and discharge power and capacity constraints include lithium battery energy storage constraints and flow battery energy storage constraints. Among them, for the lithium battery energy storage constraint: , , . For the flow battery energy storage constraint: , , . Among them, , , , , , , and represent the corresponding maximum and minimum charge / discharge power and capacity limits respectively.

[0087] Furthermore, the embodiments of the present invention adopt the Mixed-Integer Nonlinear Programming (MINLP) method to solve the above-mentioned constrained optimization problem. The power output values of distributed power sources (such as of photovoltaic power sources, of wind power sources, etc.), the charge / discharge power values of energy storage devices (such as , of lithium batteries, , of flow batteries, etc.) and related discrete control variables (such as the start / stop status of distributed power sources, etc.) are used as decision variables.

[0088] By constructing the local optimization objective function and the inequality and equality constraints corresponding to the above-mentioned constraint conditions, a MINLP solver (such as Bonmin, Couenne, etc.) is used for solving. The solver will search for a set of decision variable values that make the local optimization objective function reach the optimal value within the feasible solution space that satisfies all constraint conditions. The initial power output value of the distributed power source and the initial charge / discharge power of the energy storage device corresponding to this set of values are the preliminary local scheduling decisions of each first local intelligent controller .

[0089] The first local intelligent controller in the embodiments of the present invention can comprehensively consider various characteristics of the local node, and make an optimization decision based on the constructed local optimization objective function and the load data, distributed power source output data, and energy storage state data of the local node. Therefore, it can provide reasonable local scheduling decisions for the efficient and stable operation of the AC / DC distribution network according to its actual situation.

[0090] In one embodiment, the descriptions of steps 301 to 303 are as follows: Step 301, based on the first decision variable vector and the first preliminary local scheduling decision vector of the first local intelligent controller, and the second decision variable vector and the second preliminary local scheduling decision vector of the second local intelligent controller, determine the consistency degree of the decision variables between the first local intelligent controller and the second local intelligent controller.

[0091] Optionally, there are nodes in the AC / DC distribution network area of the embodiments of the present invention, and each node corresponds to a local intelligent controller. Therefore, the set of local nodes corresponding to the first local intelligent controller can be expressed as , the set of second local intelligent controllers corresponding to the nodes adjacent to the local node can be expressed as , the set of local nodes and the set of nodes are associated, and the nodes corresponding to the elements in the set of nodes are adjacent to the nodes in the set of local nodes . For each first local intelligent controller , the corresponding first decision variable vector can be expressed as , where represents the number of decision variables. For example, can be the set value of the distributed power output, can be the set value of the charge and discharge power of the energy storage system.

[0092] Optionally, the first preliminary local scheduling decision vector of the first local intelligent controller obtained based on the local optimization objective function in the embodiments of the present invention can be expressed as , and the first local resource margin information vector can be expressed as , represents the number of dimensions included in the resource margin information, such as the remaining adjustable power margin of the distributed power source, the remaining charge and discharge capacity margin of the energy storage system, etc. Correspondingly, for the second local intelligent controller , the corresponding second decision variable vector can be expressed as , the second preliminary local scheduling decision vector can be expressed as , and the second local resource margin information vector can be expressed as .

[0093] Further, according to the first decision variable vector and the first preliminary local scheduling decision vector of the first local intelligent controller, and the second decision variable vector and the second preliminary local scheduling decision vector of the second local intelligent controller, the consistency degree of the decision variables between the first local intelligent controller and the second local intelligent controller is determined. The specific formula for the consistency degree of the decision variables is as follows: .

[0094] Where represents the consistency degree of the decision variables between the first local intelligent controller and the second local intelligent controller , represents the -th element in the first decision variable vector, represents the -th element in the second decision variable vector, represents the The th element, representing the

[0095] th element in the second preliminary local scheduling decision vector.

[0096] Furthermore, based on the first decision variable vector and the first local resource margin information of the first local intelligent controller, determine the degree of association between the decision variables and the local resource margin information in the first local intelligent controller. The specific formula for the degree of association is as follows: .

[0097] Where, represents the degree of association between the local resource margin information and the decision variables in the first local intelligent controller , represents the th element in the first decision variable vector, represents the upper limit value of the th element in the first decision variable vector.

[0098] Step 303: Construct a target decision function based on the degree of consistency and the degree of association, and perform consistency collaborative optimization on the target decision function based on the second local resource margin information of the second local intelligent controller to obtain the target local scheduling decision of each first local intelligent controller.

[0099] Furthermore, in order to measure whether the entire regional power grid reaches an overall optimal consistency state, an embodiment of the present invention constructs a target decision function , where represents the overall vector of decision variables covering all node intelligent controllers, and the specific formula of the target decision function is as follows: .

[0100] Where, represents the preset weight coefficient between the first local intelligent controller and the second local intelligent controller .

[0101] Furthermore, an embodiment of the present invention optimizes the target decision function through a distributed gradient descent algorithm. Therefore, calculate the target decision function with respect to the first local intelligent controller gradients of each decision variable, and for the first local intelligent controller iterate on the gradients of each decision variable to obtain an optimal iteration result, and obtain the target local scheduling decision of the first local intelligent controller, as specifically described in steps 3031 to 3033.

[0102] In the embodiment of the present invention, consistency collaborative optimization is performed according to the first interaction information of the first local intelligent controller and the second interaction information of the second local intelligent controller. Therefore, reasonable local scheduling decisions can be provided for the efficient and stable operation of the AC-DC distribution network according to its own actual situation and the actual situation of adjacent nodes, thereby improving the overall operation efficiency and stability of the AC-DC distribution network.

[0103] In one embodiment, the descriptions of steps 3031 to 3033 are as follows: Step 3031: Calculate the gradient of the objective decision function with respect to each decision variable in the first decision variable vector of the first local intelligent controller according to the second local resource margin information.

[0104] Optionally, according to the second local resource margin information of the second local intelligent controller calculate the objective decision function with respect to the first local intelligent controller for the first decision variable vector in each decision variable. Therefore, for the first local intelligent controller , its first decision variable vector in the th element the gradient calculation formula is specifically as follows: .

[0105] Where represents a preset weight coefficient.

[0106] Step 3032: Iteratively optimize the gradient of each decision variable and obtain the convergence index in each iterative optimization process. The convergence index characterizes whether each decision variable is gradually tending towards the overall optimal consistency state of the AC-DC distribution network.

[0107] Furthermore, based on the above gradient calculation, at each iteration , the first local intelligent controller for the decision variable vector the update formula is specifically as follows: .

[0108] Where represents the learning rate, represents at the The decision variable vector at the -th iteration, which represents that at the -th iteration, the gradient vector of the objective decision function with respect to the decision variable vector of the -th node is calculated; represents the constraint vector, which is used to ensure that the updated decision variables satisfy the local resource margin and other actual constraint conditions, and its elements can be expressed as: .

[0109] In one embodiment, if the updated distributed power output setting value exceeds the adjustable power range allowed by the local resource margin or exceeds its own maximum power limit and other constraint conditions, then the corresponding constraint vector element is 0, preventing this update and ensuring that the update of the decision variables is always carried out within a reasonable and feasible range.

[0110] Optionally, at the initial stage , each first local intelligent controller starts iteration with its preliminary local scheduling decision vector as the initial value. During each iteration, each first local intelligent controller first exchanges interaction information with adjacent second local intelligent controllers , that is, their respective preliminary local scheduling decision vectors and local resource margin information vectors, and then calculates the weight coefficients with each adjacent node under the current iteration according to the received information. Among them, the calculation of the weight coefficient can comprehensively consider multiple factors, such as the electrical connection strength between nodes (measured by relevant electrical parameters such as conductance), the communication link quality (comprehensively measured by indicators such as signal strength and packet loss rate), and the correlation of historical interaction information. Therefore, the calculation formula of the weight coefficient is as follows: .

[0111] Among them, represents the electrical connection strength related parameter between node and node , represents the communication link quality related parameter between node and node , represents the correlation index calculated based on historical interaction information, , and represent the corresponding weight coefficients, .

[0112] Further, calculate and update its own decision variable vector according to the above update formula of distributed gradient descent , and at the same time, perform constraint checking and adjustment according to the constraint vector to ensure that the updated decision variables meet the local resource margin and other actual constraint conditions. In one embodiment, if the remaining adjustable power margin of the distributed power source is insufficient, then when updating the output setting value of the distributed power source, it cannot exceed this margin range. This can be achieved by adding a margin constraint condition to the update formula. For example, when updating the output setting value of the distributed power source, it is necessary to satisfy , where

[0113] is the margin of the corresponding distributed power source in the margin information .

[0114] Among them, represents the convergence index

[0115] Step 3033, if the convergence index is less than the preset convergence threshold, then determine the decision variable vector obtained after the current iteration as the target local scheduling decision of the first local intelligent controller

[0116] Further, determine whether the convergence index is less than the preset convergence threshold. Among them, the preset convergence threshold in the embodiment of the present invention . If it is determined that the convergence index is less than the preset convergence threshold , it is determined that the iteration has converged. At this time, the decision variable vector of each first local intelligent controller is the finally determined target local scheduling decision

[0117] In the embodiment of the present invention, the target decision function is optimized in a consistent manner according to the second local resource margin information of the second local intelligent controller. Therefore, it is possible to provide a reasonable local scheduling decision for the efficient and stable operation of the AC-DC distribution network according to the actual situation of adjacent nodes, thereby improving the overall operation efficiency and stability of the AC-DC distribution network

[0118] In one embodiment, the descriptions of steps 401 to 405 are as follows Step 401, obtain the total duration of the entire scheduling process and divide the total duration into multiple time segments

[0119] In the embodiment of the present invention, the AC-DC distribution network is divided into a region, and each region is managed by a region coordinator. For the th region, the first set of local intelligent controllers connected to its region coordinator is , that is, the th region contains target local intelligent controllers, that is, it can be understood that the th region is connected to target local intelligent controllers .

[0120] Optionally, in the embodiment of the present invention, the total duration considered in the entire scheduling process is set to , and the total duration is divided into equally spaced time segments, and the duration of each time segment is , that is . For each time segment , the corresponding moment is .

[0121] Step 402: Based on the time segment and the load data of the local nodes of the target local intelligent controllers at the current moment, determine the total peak-valley difference within the region to be minimized.

[0122] Optionally, each target local intelligent controller corresponds to a local node, and the information of this local node includes: the load data of the local node at the current moment , the output data of the distributed power source at the current moment , and the energy storage state data of the energy storage device at the current moment . .

[0123] Among them, the load data of the local node at the current moment is composed of the superposition of multiple different types of load components. For example, the load curve includes the base load and the peak load. Therefore, the load data can be expressed as , where , represents the base load, represents the peak load, and the peak load is related to different electricity consumption characteristic periods such as residents and industries.

[0124] Furthermore, based on the time segment and the load data of the local nodes of the target local intelligent controllers at the current moment, determine the total peak-valley difference within the region to be minimized. The specific formula is as follows: .

[0125] in, Represents the total peak-to-valley difference within the minimized region.

[0126] Step 403 : Based on the time segment and the output data of the distributed power source of the target local intelligent controller at the current moment, determine the deviation between the actual output of the distributed power source and the output set value in the target local scheduling decision.

[0127] Optionally, distributed power generation at the current moment Output data Including distributed power generation at the current moment Output setting value (from target local dispatch decision) and actual output The output characteristics are affected by many factors, such as external environment such as light intensity and wind speed, as well as the limitations of its own adjustment ability.

[0128] Furthermore, based on the time segment and the output data of the distributed power source of the target local intelligent controller at the current moment, the deviation between the actual output of the distributed power source and the output set value in the target local scheduling decision is determined. The specific formula is as follows: .

[0129] It represents the deviation between the actual output of the distributed generation and the output set value in the target local dispatch decision.

[0130] Step 404 : Based on the time segment and the energy storage status data of the energy storage device of the target local intelligent controller at the current moment, determine the degree of deviation between the actual power of the energy storage device and the power set value in the target local scheduling decision.

[0131] Optionally, the energy storage device at the current moment Energy storage status data Including the energy storage device at the current moment Charge and discharge power setting value (from target local dispatch decision) and actual charge and discharge power , actual charge and discharge power The power state is , and has minimum power and maximum power limit.

[0132] Furthermore, based on the time segment and the energy storage status data of the target local intelligent controller's energy storage device at the current moment, the degree of deviation between the actual power of the energy storage device and the power set value in the target local scheduling decision is determined. The specific formula is as follows: .

[0133] Among them, It represents the deviation degree of the actual power of the energy storage device from the power setting value in the target local scheduling decision.

[0134] Step 405: Construct a comprehensive objective function based on the total peak-valley difference, deviation, and deviation degree, and update the target local scheduling decision of the target local intelligent controller based on the distributed cooperative optimization algorithm to minimize the comprehensive objective function, so as to obtain the node scheduling decision of the node corresponding to the target local intelligent controller.

[0135] Furthermore, in order to achieve the peak-valley difference optimal scheduling of the AC / DC distribution network, the embodiment of the present invention constructs a comprehensive objective function according to the total peak-valley difference, deviation, and deviation degree Therefore, the comprehensive objective function Comprehensively considers various factors such as peak-valley difference reduction, reasonable resource utilization, and scheduling stability. The formula of the comprehensive objective function is specifically as follows: .

[0136] Among them, represents the minimum total peak-valley difference within the region, represents the deviation between the actual output of the distributed power source and the output setting value in the target local scheduling decision, represents the deviation degree of the actual power of the energy storage device from the power setting value in the target local scheduling decision, represents the coordination of scheduling behaviors between different target local intelligent controllers, represents a preset weight coefficient.

[0137] In the embodiment of the present invention, the coordination of scheduling behaviors between different target local intelligent controllers is considered , and the specific formula is as follows: .

[0138] Among them, represents the th target local intelligent controller and the th target local intelligent controller, and the association weight between them can be determined according to factors such as the electrical distance and resource complementarity between the th target local intelligent controller and the th target local intelligent controller. <>

[0139] Optionally, the area coordinator in the embodiment of the present invention adopts a distributed cooperative optimization algorithm to update the target local scheduling decision of each target local intelligent controller (node) to minimize the comprehensive objective function , obtain the node scheduling decision for the node corresponding to the target local intelligent controller, specifically as described in steps 4051 to 4053.

[0140] In the embodiment of the present invention, the area coordinator performs distributed collaborative optimization on the local scheduling decisions of local intelligent controllers, determines the optimal scheduling decision for each local intelligent controller, realizes accurate and refined optimal scheduling of the peak-valley difference of the AC / DC distribution network, better responds to load changes, realizes reasonable allocation and utilization of resources, and improves the overall operation efficiency and stability of the AC / DC distribution network.

[0141] In one embodiment, the descriptions of steps 4051 to 4053 are as follows: Step 4051, update each scheduling decision variable in the target local scheduling decision of the target local intelligent controller, and obtain the first convergence index in each update process.

[0142] Optionally, in each iteration of the embodiment of the present invention when, for the th target local intelligent controller (node) in the area, the update rule for the scheduling decision variables (such as the actual output of distributed power sources, the actual charge-discharge power of energy storage, etc.) in the target local scheduling decision of the th target local intelligent controller is as follows: For the update rule of distributed power source output: For the distributed power source output update rule: .

[0143] Among them, represents the actual output of the distributed power source of the th target local intelligent controller at the rd iteration and time segment , represents the learning rate, represents the comprehensive objective function with respect to the partial derivative of the output of the distributed power source of the th target local intelligent controller at the th iteration and time segment ; represents the collaborative weight of the distributed power source output between the th target local intelligent controller and the th target local intelligent controller in the area, represents the set of nodes that have a specific association (such as electrical adjacency, resource complementary demand, etc.) with the th target local intelligent controller, represents the association weight of the distributed power source output.

[0144] For the energy storage charge and discharge power update rule: 。

[0145] Among them, represents the actual charge and discharge power of the distributed power source of the th target local intelligent controller at the th iteration and time segment . represents the learning rate, represents the comprehensive objective function with respect to the charge and discharge power of the distributed power source of the th target local intelligent controller at the th iteration and time segment . represents the collaborative weight of the energy storage charge and discharge power between the th target local intelligent controller and the th target local intelligent controller in the region. represents the node set with a specific association (such as complementary energy storage resources, load characteristic association, etc.) with the th target local intelligent controller. represents the association weight of the energy storage charge and discharge power.

[0146] Furthermore, obtain the first convergence index in each update process. Among them, the calculation formula of the first convergence index is as follows: .

[0147] Step 4052, obtain the constraint conditions in the distributed collaborative optimization process.

[0148] Optionally, in the distributed collaborative optimization process of the embodiments of the present invention, the following constraint conditions need to be satisfied to ensure the feasibility and rationality of the scheduling decision: 1. Power balance constraint For each target local intelligent controller (node ) at any time segment , there is: .

[0149] Among them, represents the power of the target local intelligent controller interacting with the outside (such as adjacent nodes, superior power grid, etc.) at the time segment . The inflow is positive and the outflow is negative.

[0150] 2. Distributed power source output range constraint The actual output of the distributed power source needs to be within its allowable minimum and maximum output ranges, i.e.: .

[0151] Among them, and are the minimum and maximum output limits of the distributed power source respectively.

[0152] 3. Energy storage charge and discharge power and capacity constraints The charge and discharge power of the energy storage device should meet its rated charge and discharge power limits. When charging: .

[0153] When discharging: .

[0154] At the same time, the power of the energy storage should always be between its minimum and maximum capacities, i.e.: .

[0155] Among them, and are the rated charging power and rated discharging power of the energy storage device respectively, and are its minimum and maximum power capacities respectively.

[0156] Step 4053, if under the condition of meeting the constraint conditions and the first convergence index is less than the first preset convergence threshold, then based on the actual output of the distributed power source and the actual charge and discharge power of the energy storage when the distributed collaborative optimization algorithm converges, determine the node scheduling decision of the node corresponding to the target local intelligent controller.

[0157] Optionally, the convergence judgment mechanism of the embodiment of the present invention is: judge whether the convergence index of the distributed collaborative optimization algorithm converges, that is, whether the convergence index is less than the preset convergence threshold .

[0158] When is less than the preset convergence threshold , it is determined that the distributed collaborative optimization algorithm converges.

[0159] Furthermore, after determining that the distributed collaborative optimization algorithm converges, at this time, the actual output of the distributed power source and the actual charge and discharge power of the energy storage of each target local intelligent controller are the final node scheduling decisions of the target local intelligent controller. Therefore, the node scheduling decision includes the output behavior of the distributed power source and the charge and discharge behavior of the energy storage.

[0160] Further, the area coordinator controls each target local intelligent controller to perform corresponding scheduling actions according to the output behavior of the corresponding distributed power source and the charge and discharge behavior of the energy storage, that is, the distributed power source generates electricity according to the determined output behavior, and the energy storage performs charge and discharge operations according to the determined charge and discharge power, so as to realize the optimal scheduling of the peak-valley difference of the AC / DC distribution network.

[0161] In the embodiment of the present invention, the area coordinator performs distributed collaborative optimization on the local scheduling decisions of the local intelligent controllers, determines the optimal scheduling decision for each local intelligent controller, realizes accurate and fine optimization scheduling of the peak-valley difference of the AC / DC distribution network, better responds to load changes, realizes reasonable allocation and utilization of resources, and improves the overall operation efficiency and stability of the AC / DC distribution network.

[0162] Further, the AC / DC distribution network distributed optimization scheduling system provided by the present invention is described below. The AC / DC distribution network distributed optimization scheduling system described below can be mutually corresponding and referred to the AC / DC distribution network distributed optimization scheduling method described above.

[0163] Optionally, referring to Figure 2 , Figure 2 is the structural diagram of the AC / DC distribution network distributed optimization scheduling system provided by the present invention. The AC / DC distribution network distributed optimization scheduling system at least includes an optimization scheduling management middleware, local intelligent controllers and area coordinators. Each area coordinator is connected to the local intelligent controllers within a preset range through a communication link. The optimization scheduling management middleware is connected to all area coordinators to uniformly manage all area coordinators. The local intelligent controllers are deployed at each node of the AC / DC distribution network. The nodes include distributed power source nodes, energy storage device nodes and load nodes.

[0164] The local intelligent controller is used for: Collect the load data, distributed power source output data and energy storage state data of the local node where it is located; Perform local optimization decisions according to the local optimization objective function in combination with the load data, distributed power source output data and energy storage state data to obtain the preliminary local scheduling decision of each first local intelligent controller Perform consistency collaborative optimization according to the first interaction information and the second interaction information of the second local intelligent controller to determine the target local scheduling decision of each first local intelligent controller; The area coordinator is used for: Perform distributed collaborative optimization according to the target local scheduling decision of the target local intelligent controller to determine the node scheduling decision of the node corresponding to the target local intelligent controller, so as to realize the optimal scheduling of the peak-valley difference of the AC / DC distribution network.

[0165] In the embodiment of the present invention, the first local intelligent controller can comprehensively consider various characteristics of local nodes, make an optimization decision based on the constructed local optimization objective function and the load data, distributed power output data, and energy storage state data of local nodes, and can provide reasonable local scheduling decisions for the efficient and stable operation of the AC-DC distribution network according to its own situation. Further, the regional coordinator performs distributed collaborative optimization on the local scheduling decisions of the local intelligent controller to determine the optimal scheduling decision for each local intelligent controller, achieving accurate and fine optimization scheduling of the peak-valley difference of the AC-DC distribution network, better coping with load changes, realizing the reasonable allocation and utilization of resources, and improving the overall operation efficiency and stability of the AC-DC distribution network.

[0166] Please refer to Figure 3 , Figure 3 which is the embodiment diagram of the electronic device provided by the embodiment of the present invention. As Figure 3 shown, the embodiment of the present invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored on the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, the following steps are implemented: Collecting the load data, distributed power output data, and energy storage state data of the local nodes where each first local intelligent controller is located; Based on each first local intelligent controller making a local optimization decision according to the local optimization objective function in combination with the load data, distributed power output data, and energy storage state data, obtaining the preliminary local scheduling decision of each first local intelligent controller; Based on each first local intelligent controller performing consistency collaborative optimization according to the first interaction information and the second interaction information of the second local intelligent controller to determine the target local scheduling decision of each first local intelligent controller; Based on each regional coordinator performing distributed collaborative optimization according to the target local scheduling decision of the target local intelligent controller to determine the node scheduling decision of the node corresponding to the target local intelligent controller, realizing the peak-valley difference optimization scheduling of the AC-DC distribution network.

[0167] Please refer to Figure 4 , Figure 4 which is the embodiment diagram of the computer-readable storage medium provided by the embodiment of the present invention. As Figure 4 shown, this embodiment provides a computer-readable storage medium 400, on which a computer program 311 is stored. When the computer program 311 is executed by a processor, the following steps are implemented: Collecting the load data, distributed power output data, and energy storage state data of the local nodes where each first local intelligent controller is located; Based on each first local intelligent controller making local optimization decisions according to the local optimization objective function in combination with load data, distributed power generation output data, and energy storage state data, to obtain the preliminary local scheduling decisions of each first local intelligent controller; Based on each first local intelligent controller performing consistency collaborative optimization according to the first interaction information and the second interaction information of the second local intelligent controller, to determine the target local scheduling decisions of each first local intelligent controller; Based on each area coordinator performing distributed collaborative optimization according to the target local scheduling decisions of the target local intelligent controller, to determine the node scheduling decisions of the nodes corresponding to the target local intelligent controller, so as to achieve the peak-valley difference optimal scheduling of the AC-DC distribution network.

[0168] On the other hand, the present invention also provides a computer program product, the computer program product includes a computer program, the computer program can be stored on a non-transitory computer-readable storage medium, when the computer program is executed by a processor, the computer can execute the AC-DC distribution network distributed optimal scheduling method provided by each of the above methods, and this method includes: Based on each first local intelligent controller collecting the load data, distributed power generation output data, and energy storage state data of the local node where it is located; Based on each first local intelligent controller making local optimization decisions according to the local optimization objective function in combination with load data, distributed power generation output data, and energy storage state data, to obtain the preliminary local scheduling decisions of each first local intelligent controller; Based on each first local intelligent controller performing consistency collaborative optimization according to the first interaction information and the second interaction information of the second local intelligent controller, to determine the target local scheduling decisions of each first local intelligent controller; Based on each area coordinator performing distributed collaborative optimization according to the target local scheduling decisions of the target local intelligent controller, to determine the node scheduling decisions of the nodes corresponding to the target local intelligent controller, so as to achieve the peak-valley difference optimal scheduling of the AC-DC distribution network.

[0169] The system embodiments described above are merely illustrative, where the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0170] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0171] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. And these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A distributed optimal scheduling method for AC / DC distribution networks, characterized in that, Based on the implementation of the AC / DC distribution network distributed optimal scheduling system, the AC / DC distribution network distributed optimal scheduling system at least includes local intelligent controllers and area coordinators, and each area coordinator is connected to the local intelligent controllers within a preset range through a communication link; The local intelligent controllers are deployed at each node of the AC / DC distribution network, and the nodes include distributed power source nodes, energy storage device nodes, and load nodes; the method includes: Based on each first local intelligent controller, collecting the load data, distributed power output data, and energy storage state data of the local node where it is located; Based on each first local intelligent controller, making a local optimal decision according to the local optimal objective function in combination with the load data, distributed power output data, and energy storage state data, and obtaining the preliminary local scheduling decision of each first local intelligent controller; Based on each first local intelligent controller, performing consistency collaborative optimization according to the first interaction information and the second interaction information of the second local intelligent controller, and determining the target local scheduling decision of each first local intelligent controller; Based on each area coordinator, performing distributed collaborative optimization according to the target local scheduling decision of the target local intelligent controller, and determining the node scheduling decision of the node corresponding to the target local intelligent controller, so as to achieve the peak-valley difference optimal scheduling of the AC / DC distribution network; The local optimal objective function is a function constructed with the goals of minimizing the peak-valley difference, maximizing the local energy self-sufficiency rate, and minimizing the energy storage life loss, in combination with the load characteristics of the local load nodes, the power regulation characteristics of the distributed power source nodes, and the charge and discharge characteristics of the energy storage device nodes; the second local intelligent controller is the controller corresponding to the node adjacent to the local node; the interaction information includes the preliminary local scheduling decision and the local resource surplus information; the target local intelligent controller is the first local intelligent controller connected to each area coordinator.

2. The distributed optimal scheduling method for AC / DC distribution network according to claim 1, wherein The performing distributed collaborative optimization according to the target local scheduling decision of the target local intelligent controller and determining the node scheduling decision of the node corresponding to the target local intelligent controller includes: Obtaining the total duration of the entire scheduling process and dividing the total duration into multiple time segments; Based on the time segment and the load data of the local node of the target local intelligent controller at the current moment, determining the total peak-valley difference within the minimized area; Based on the time segment and the output data of the distributed power source of the target local intelligent controller at the current moment, determining the deviation between the actual output of the distributed power source and the output set value in the target local scheduling decision; Based on the time segment and the energy storage state data of the energy storage device of the target local intelligent controller at the current moment, determining the deviation degree between the actual power of the energy storage device and the power set value in the target local scheduling decision; Constructing a comprehensive objective function based on the total peak-valley difference, the deviation, and the deviation degree, and updating the target local scheduling decision of the target local intelligent controller based on the distributed collaborative optimization algorithm to minimize the comprehensive objective function, so as to obtain the node scheduling decision of the node corresponding to the target local intelligent controller.

3. The distributed optimal scheduling method for AC / DC distribution network according to claim 2, wherein, Updating the target local scheduling decision of the target local intelligent controller based on the distributed cooperative optimization algorithm to minimize the comprehensive objective function, and obtaining the node scheduling decision of the node corresponding to the target local intelligent controller, including: Updating each scheduling decision variable in the target local scheduling decision of the target local intelligent controller, and obtaining the first convergence index in each update process; Obtaining the constraint conditions in the distributed cooperative optimization process; If the first convergence index is less than the first preset convergence threshold under the condition of satisfying the constraint conditions, determining the node scheduling decision of the node corresponding to the target local intelligent controller based on the actual output of the distributed power source and the actual charge and discharge power of the energy storage of the target local intelligent controller when the distributed cooperative optimization algorithm converges; the node scheduling decision includes the output behavior of the distributed power source and the charge and discharge behavior of the energy storage.

4. The distributed optimal scheduling method for AC / DC distribution network according to claim 1, characterized in that Performing consistency cooperative optimization according to the first interaction information and the second interaction information of the second local intelligent controller to determine the target local scheduling decision of each first local intelligent controller, including: Determining the degree of consistency of decision variables between the first local intelligent controller and the second local intelligent controller based on the first decision variable vector and the first preliminary local scheduling decision vector of the first local intelligent controller, and the second decision variable vector and the second preliminary local scheduling decision vector of the second local intelligent controller; Determining the degree of association between the decision variables and the local resource margin information in the first local intelligent controller based on the first decision variable vector and the first local resource margin information of the first local intelligent controller; Constructing a target decision function based on the degree of consistency and the degree of association, and performing consistency cooperative optimization on the target decision function based on the second local resource margin information of the second local intelligent controller to obtain the target local scheduling decision of each first local intelligent controller.

5. The distributed optimal scheduling method for AC-DC distribution network according to claim 4, wherein Performing consistency cooperative optimization on the target decision function based on the second local resource margin information of the second local intelligent controller to obtain the target local scheduling decision of each first local intelligent controller, including: Calculating the gradient of the target decision function with respect to each decision variable in the first decision variable vector of the first local intelligent controller according to the second local resource margin information; Performing iterative optimization on the gradient of each decision variable, and obtaining the second convergence index in each iterative optimization process; the second convergence index characterizes whether each decision variable gradually tends to the consistent state of the overall optimal of the AC / DC distribution network; If the second convergence index is less than the second preset convergence threshold, determining the decision variable vector obtained after the current iteration as the target local scheduling decision of the first local intelligent controller.

6. The distributed optimal scheduling method for AC / DC distribution network according to claim 1, characterized in that The specific process of constructing the local optimization objective function includes: Determining the total load characteristic based on the first load curve corresponding to the load characteristic of the industrial user concentration area and the second load curve corresponding to the load characteristic of the residential user concentration area; Determining the distributed power source output based on the first power regulation characteristic corresponding to the output of the photovoltaic power source and the second power regulation characteristic corresponding to the output of the wind power source; Determine the energy change rate based on the battery charge and discharge efficiency of the lithium battery corresponding to the battery power change and the charge and discharge efficiency of the flow battery corresponding to the energy storage power change rate. Construct the local optimization objective function based on the total load characteristics, the distributed power generation output, and the energy change rate.

7. The distributed optimal scheduling method for AC / DC distribution network according to claim 6, wherein The constructing the local optimization objective function based on the total load characteristics, the distributed power generation output, and the energy change rate includes: Determine the peak-valley difference based on the total load characteristics, and construct a peak-valley difference minimization objective function based on the peak-valley difference. Based on the total load characteristics and the distributed power generation output, respectively determine the total energy generated by the local distributed power generation and the energy consumed by the local total load, and construct a local energy self-sufficiency rate maximization objective function based on the total energy generated by the local distributed power generation and the energy consumed by the local total load. Determine the first life loss index of the lithium battery energy storage and the second life loss index of the flow battery energy storage based on the energy change rate, and construct a storage life loss minimization objective function based on the first life loss index and the second life loss index. Construct the local optimization objective function based on the peak-valley difference minimization objective function, the local energy self-sufficiency rate maximization objective function, and the storage life loss minimization objective function.

8. A distributed optimal scheduling system for AC-DC distribution network, characterized in that, The system includes at least a local intelligent controller and a regional coordinator, and each regional coordinator is connected to the local intelligent controllers within a preset range through a communication link; the local intelligent controllers are deployed at each node of the AC / DC distribution network, and the nodes include distributed power generation nodes, energy storage device nodes, and load nodes, and are used to implement the AC / DC distribution network distributed optimization scheduling method according to any one of claims 1 to 7. The local intelligent controller is used for: Collect the load data, distributed power generation output data, and energy storage state data of the local node where it is located. Make a local optimization decision according to the local optimization objective function in combination with the load data, distributed power generation output data, and energy storage state data, and obtain the preliminary local scheduling decision of each first local intelligent controller. Perform consistency collaborative optimization according to the first interaction information and the second interaction information of the second local intelligent controller to determine the target local scheduling decision of each first local intelligent controller. The regional coordinator is used for: Perform distributed collaborative optimization according to the target local scheduling decision of the target local intelligent controller to determine the node scheduling decision of the node corresponding to the target local intelligent controller, and achieve peak-valley difference optimization scheduling of the AC / DC distribution network. The local optimization objective function is a function constructed with the goals of minimizing the peak-valley difference, maximizing the local energy self-sufficiency rate, and minimizing the storage life loss, in combination with the load characteristics of the local load nodes, the power regulation characteristics of the distributed power generation nodes, and the charge and discharge characteristics of the energy storage device nodes. The second local intelligent controller is the controller corresponding to the node adjacent to the local node; the interaction information includes the preliminary local scheduling decision and the local resource margin information; the target local intelligent controller is the first local intelligent controller connected to each regional coordinator.

9. An electronic device, comprising: The memory is used to store computer software programs. A processor for reading and executing the computer software program, characterized in that when the processor executes the computer software program, it implements the AC-DC distribution network distributed optimal scheduling method according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing a computer software program, characterized in that, When the computer software program is executed by the processor, it implements the AC-DC distribution network distributed optimal scheduling method according to any one of claims 1 to 7.

Images