New energy and electric vehicle charging pile inertia aggregation method and system and computer program product

Abstract

The invention discloses a new energy and electric vehicle charging pile inertia aggregation method and system and a computer program product, and belongs to the technical field of smart power grids, and the method comprises the steps: obtaining an inertia aggregation upper limit of an inertia aggregator at each node; taking the minimum inertia aggregation cost of the inertia aggregator as a target, constructing an optimal aggregation model of the inertia aggregator, and obtaining the optimal inertia aggregation cost of the inertia aggregator under the known middle standard amount; and according to the inertia aggregator optimal aggregation model, obtaining a quotation sample data set, establishing a curve fitting model based on the deep residual network in combination with an adaptive activation function, and inputting the quotation sample data set into the curve fitting model to calculate a quotation curve of the inertia aggregator under the unknown bid winning amount. According to the method, the calculation amount is reduced while the quotation curve of the inertia aggregator is predicted more accurately under the consideration of the influence of the distributed electric vehicle, the reliability and efficiency of inertia service are practically improved, and the method has remarkable practical value.

Description

Technical Field

[0001] The present application relates to the field of smart grids, and particularly to a method, system, and computer program product for aggregating the inertia of new energy and electric vehicle charging piles. Background Art

[0002] With the grid connection of a high proportion of renewable energy, the synchronous rotational inertia of the power system continues to decrease, and the frequency stability problem becomes increasingly prominent. Traditional inertia support relies on synchronous generator sets, but new energy units and distributed resources (such as electric vehicles) cannot directly provide physical inertia and need to achieve equivalent inertia response through virtual inertia control technology. During an inertia service event, the power grid often signs contracts with multiple inertia aggregators to indirectly supply inertia services to underlying electric vehicles and distributed photovoltaics. The underlying electric vehicles and distributed photovoltaics report their inertia supply unit prices to the inertia aggregators according to demand, and the inertia aggregators will calculate the quoted cost according to an optimization algorithm to participate in inertia market transactions. In recent years, as a large-scale distributed energy storage resource, electric vehicles can participate in system inertia support through vehicle-to-grid (V2G), but their dispersion, dynamics, and multi-level power grid structure characteristics pose challenges to the efficient aggregation of inertia resources.

[0003] Currently, there has been in-depth research on energy storage in control methods such as improving power grid frequency stability, suppressing the frequency change rate of frequency disturbances, and reducing steady-state frequency deviation, which provides a theoretical reference for electric vehicles and their supporting photovoltaics to participate in power grid inertia services. When an electric vehicle cluster participates in power grid inertia services, it must meet the charging needs of users. Different from the state of charge (SOC) control target of energy storage batteries, the virtual inertia operation control of an electric vehicle cluster essentially involves the aggregated regulation of the charging and discharging powers of a large number of individual electric vehicles. Since the charging and discharging requirements of each individual electric vehicle are different, a suitable control strategy for participating in inertia services should be formulated within the adjustable range of the charging and discharging of each electric vehicle. However, due to the large number of nodes in the multi-level power grid and the connection of a large number of electric vehicle charging piles and distributed photovoltaics at the underlying voltage level, traditional inertia supply does not consider the problem of distributed electric vehicle inertia resources; if an optimization algorithm for quoted cost is performed on each node, the workload is large and problems such as data errors are likely to occur. Summary of the Invention

[0004] The present application aims to provide a method, system, and computer program product for aggregating the inertia of new energy and electric vehicle charging piles to solve the problem that distributed electric vehicle inertia resources are not considered in traditional inertia supply.

[0005] To achieve the above object, the technical solution of the present application is: A method for aggregating the inertia of new energy and electric vehicle charging piles, including, obtaining the inertia aggregation upper limit of the inertia aggregator at each node; Taking the minimum inertia aggregation cost of the inertia aggregator as the goal, an optimal aggregation model of the inertia aggregator is constructed to obtain the optimal inertia aggregation cost of the inertia aggregator under the known winning bid quantity; According to the optimal aggregation model of the inertia aggregator, a quotation sample data set is obtained, a curve fitting model based on a deep residual network combined with an adaptive activation function is established, and the quotation sample data set is input into the curve fitting model to calculate the quotation curve of the inertia aggregator under the unknown winning bid quantity.

[0006] Optionally, obtaining the inertia aggregation upper limit of the inertia aggregator at each node includes: Calculating the upper limit of inertia supply facing power deficit at each node; Calculating the upper limit of inertia supply facing power redundancy at each node; Taking the minimum value between the aggregable inertia upper limit facing power deficit and the aggregable inertia upper limit facing power redundancy at each node as the inertia aggregation upper limit of the inertia aggregator at each node.

[0007] Optionally, the objective function for calculating the upper limit of inertia supply facing power deficit at each node is expressed as follows: Among them, represents the voltage level, The lower the value, the lower the voltage level; b represents the node number connecting the upper and lower power grids; t represents the time sequence number; H aggEV.up,i-1,b,t is at time the inertia aggregated from the lower voltage level connected to the node at voltage level upward to voltage level converted into the aggregable inertia upper limit facing power deficit is the injection power of node at voltage level at time is the maximum frequency change rate constraint of node b; is at time the non-adjustable load in the voltage level connected to the node, which is a fixed value; is at time the equivalent load aggregated from the lower voltage level connected to the node at voltage level upward to voltage level is at time the equivalent power generation of the lower voltage level connected to the node at voltage level upward to voltage level When At time an adjustable load at the lowest voltage level is connected to the node, at time a collective equivalent load on the lowest voltage level is connected to the node, at time a collective equivalent power generation on the lowest voltage level is connected to the node; The objective function for calculating the upper limit of inertia supply for power redundancy at each node is expressed as follows: where H aggEV.down,i-1,b,t is time the inertia of the lower voltage level connected to the node aggregated upward to voltage level converted into the upper limit of aggregable inertia for power redundancy ; is the injection power of node at voltage level at time ; is the non-adjustable load in voltage level connected to the node at time a fixed value; is when is time the non-adjustable load at the lowest voltage level connected to the node, is time the equivalent load of electric vehicles on the lowest voltage level connected to the node; The constraints for calculating the upper limit of inertia supply for power deficit at each node and calculating the upper limit of inertia supply for power redundancy at each node include: the active power balance constraint of the power grid, the reactive power constraint of the power grid, the current relationship constraint between grid nodes 、 the voltage relationship constraint between grid nodes 、 the charge and discharge power constraint of distributed electric vehicles participating in inertia control 、 the power generation constraint of distributed photovoltaics participating in inertia control 、 the active power output constraint of generators at each level、 Reactive power output constraints of generators at each level 、 Active power constraints of branches between nodes at each level 、 Reactive power constraints of branches between nodes at each level 、 Voltage constraints of each node 、 Branch current constraints between nodes at each level.

[0008] Optionally, the objective function of the optimal aggregation model of the inertia aggregator is expressed as follows: Among them, is the bid price of the k-th distributed electric vehicle, represents the adjustable power generation or power consumption supplied by the k-th distributed electric vehicle; is the bid price of the k-th supporting distributed photovoltaic, represents the adjustable power generation or power consumption supplied by the k-th supporting distributed photovoltaic; represents the maximum allowable rate of change of frequency required at node b; represents the supply inertia of the k-th distributed electric vehicle at time t, represents the supply inertia of the k-th supporting distributed photovoltaic at time t; The constraint conditions of the optimal aggregation model of the inertia aggregator include: upper limit constraints on the adjustable inertia of each distributed electric vehicle and each supporting distributed photovoltaic, inertia aggregation constraints balanced with the winning bid quantity, lower-level power grid power flow constraints, and maximum rate of change of frequency constraints of the new power system.

[0009] Optionally, modify the adopted deep residual network, including: Use the self-gating activation function to replace the rectified linear unit function; Add a batch normalization layer after residual addition; Use a fully connected layer to replace the convolution operation; Autonomously optimize the curve fitting accuracy by iteratively updating hyperparameters through the genetic algorithm.

[0010] Optionally, obtain a bid sample data set according to the optimal aggregation model of the inertia aggregator, establish a curve fitting model based on the deep residual network combined with the adaptive activation function, and input the bid sample data set into the curve fitting model to calculate the bid curve of the inertia aggregator under unknown winning bid quantities, including: Step S31: Obtain a bid sample data set according to the optimal aggregation model of the inertia aggregator and preprocess the bid sample data set; Step S32: Initialize the model structure and parameters of the deep residual network; Step S33: Input the bid sample data set into the curve fitting model for bid curve fitting; Step S34: Determine whether the quotation curve is completely fitted. If so, proceed to step S36; if not, proceed to step S35; Step S35: Use the genetic algorithm to change the number of residual blocks N and the hidden layer dimension d for iterative optimization, and return to step S32; Step S36: Output the final quotation curve.

[0011] Optionally, input the quotation sample data set into the curve fitting model for quotation curve fitting, including: Processed through the first fully connected layer, batch normalization, and self-gating activation function to obtain the intermediate feature state inside the residual block, expressed as follows: Where, W t,1 Represents the residual block t The weight matrix of the first fully connected layer inside; b t,1 Represents the residual block t The bias vector of the first fully connected layer inside; h t-1 Represents the (t - 1)-th residual block; BN represents the batch normalization operation; h mid Represents the residual block t The intermediate feature state inside, which is the bridge connecting two layers; Processed through the second fully connected layer to obtain the output features of the second fully connected layer, expressed as follows: Where, h res Represents the output features of the second fully connected layer inside the residual block t ; W t,2 Represents the residual block t The weight matrix of the second fully connected layer inside; b t,2 Represents the residual block t The bias vector of the second fully connected layer inside; Connect the residual blocks to obtain the quotation curve, expressed as follows: Where, y pred Represents the quotation curve processed by the deep residual network once; h N Represents the last residual block; W out Represents the weight matrix of the output layer; b out Represents the bias vector of the output layer.

[0012] Optionally, the conditions for the quotation curve to complete fitting include: the loss on the validation set is less than or equal to a set threshold , which is expressed as follows: Among them, represents the loss on the validation set, represents the set threshold.

[0013] A new energy and electric vehicle charging pile inertia aggregation system for performing the new energy and electric vehicle charging pile inertia aggregation method described in any one of the above, includes: an inertia aggregation upper limit calculation module of an inertia aggregator, an optimal aggregation cost module of an inertia aggregator, and an inertia aggregator quotation curve calculation module; The inertia aggregation upper limit calculation module of the inertia aggregator, the optimal aggregation cost module of the inertia aggregator, and the inertia aggregator quotation curve calculation module are connected in sequence.

[0014] A computer program product includes a computer program, and when the computer program is executed by a processor, it implements the steps of the new energy and electric vehicle charging pile inertia aggregation method described in any one of the above.

[0015] The present application provides a new energy and electric vehicle charging pile inertia aggregation method, system and computer program product, which solves the problem that distributed electric vehicle inertia resources are not considered in traditional inertia supply. By using an optimization algorithm considering the influence of distributed electric vehicles, the maximum inertia supply upper limit of each node is calculated, and then optimal inertia aggregation is performed from the perspective of an inertia aggregator under the known winning bid quantity. Finally, an artificial intelligence method is used to fit the quotation curve of the inertia aggregator participating in the inertia market. In the problem of maximizing the interests of the inertia aggregator participating in inertia trading, an optimal aggregation model of the inertia aggregator is constructed to calculate the winning bid quantity for the inertia aggregator to achieve inertia trading at the lowest cost, and then the quotation of the inertia aggregator is predicted more accurately. In the problem of calculating the quotation of the inertia aggregator participating in inertia trading, several key nodes are first selected for traditional quotation calculation, and then the improved deep residual network of the present application is used to perform curve fitting with the key nodes to obtain the quotation curve, which solves the problem of huge calculation amount caused by the need to optimize and calculate each node in the traditional method, significantly reduces the calculation amount, and improves the quotation accuracy. The present application more accurately predicts the quotation curve of the inertia aggregator considering the influence of distributed electric vehicles while reducing the calculation amount, effectively improves the reliability and efficiency of inertia services, and has significant practical value.

[0016] To make the above features and advantages of the application more obvious and understandable, specific embodiments are hereinafter given and described in detail in conjunction with the accompanying drawings as follows. Description of the Drawings

[0017] Figure 1 Flow chart of the inertia aggregation method for new energy and electric vehicle charging piles provided by this application.

[0018] Figure 2 Flow chart for obtaining the upper limit of inertia aggregation of inertia aggregators at each node.

[0019] Figure 3 Flow chart of step S3.

[0020] Figure 4 Module diagram of the inertia aggregation system for new energy and electric vehicle charging piles provided by this application. Specific implementation mode

[0021] To make the objectives and technical solutions of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of this application. Obviously, the described embodiments are some, but not all, of the embodiments of this application. All other embodiments obtained by those of ordinary skill in the art based on the described embodiments of this application without creative efforts fall within the scope of protection of this application.

[0022] Due to the dispersed characteristics of electric vehicles, it is necessary to aggregate all resources participating in inertia services using charging piles or parking areas (such as parking lots, etc.) and supporting distributed photovoltaics at the lowest voltage level to form an electric vehicle cluster and then connect it to the power grid. The electric vehicle cluster includes distributed electric vehicles and supporting distributed photovoltaics.

[0023] In a specific embodiment of this application, please refer to Figure 1 , Figure 1 Flow chart of the inertia aggregation method for new energy and electric vehicle charging piles provided by this application. The inertia aggregation method for new energy and electric vehicle charging piles provided by this application includes: step S1 to step S3.

[0024] Step S1: Obtain the upper limit of inertia aggregation of inertia aggregators at each node; Step S2: With the goal of minimizing the inertia aggregation cost of the inertia aggregator, construct an optimal aggregation model for the inertia aggregator, and obtain the optimal inertia aggregation cost of the inertia aggregator under the known winning bid quantity; Step S3: According to the optimal aggregation model of the inertia aggregator, obtain a quotation sample data set, establish a curve fitting model based on the deep residual network (ResNet) combined with an adaptive activation function, and input the quotation sample data set into the curve fitting model to calculate the quotation curve of the inertia aggregator under the unknown winning bid quantity.

[0025] Among them, new energy includes photovoltaic energy.

[0026] In step S1, please refer toFigure 1 In step S1, obtain the upper limit of inertia aggregation of the inertia aggregators at each node.

[0027] To achieve the inertia incentive of the power grid for load-side aggregations such as distributed electric vehicle chargers and supporting distributed photovoltaics, and at the same time facilitate the participation of load-side inertia aggregators in the inertia service volume market, it is necessary to first obtain the maximum supportable amount of inertia supply from all aggregations.

[0028] When a disturbance occurs, the traditional power system realizes the conservation of mechanical-electromagnetic power in the power system before and after the disturbance through the response of the rotor mechanical motion of the synchronous generator, that is, fills the disturbance power through the response of the rotor mechanical motion. This process is called inertia response. In the new power system, distributed power sources can achieve equivalent inertia response through virtual inertia control technology, that is, fill the disturbance power through the response of the power generation or consumption power. Therefore, the virtual inertia of the new power system resources is related to the occupied power generation or consumption, and is specifically expressed as follows: (1) Where P c,n represents the power generation or consumption power occupied when the nth disturbance occurs, H agg represents inertia, P c,max represents the supplied inertia H agg the maximum power generation or consumption power occupied, represents the angular frequency change rate under the historical maximum disturbance, and represents the frequency change rate under the historical maximum disturbance.

[0029] It can be seen from formula (1) that to require the upper limit of inertia aggregation of the inertia aggregators at each node, that is, to find the upper limit of the maximum power generation / consumption power that can be occupied at each node, that is, to find the maximum adjustable load of the equivalent electric vehicle aggregation at each node. The upper limit of inertia aggregation is obtained according to the output power that needs to be controlled by the virtual inertia provided by the aggregations connected to each node, that is, by changing the power consumption load of the electric vehicle and the photovoltaic output power, reducing the power consumption power by controlling the distributed electric vehicle, increasing the photovoltaic output power to achieve equivalent power generation power, increasing the power consumption power by controlling the distributed electric vehicle, and reducing the photovoltaic output power to achieve power consumption power, so as to achieve the effect of virtual inertia.

[0030] Since the power grid flow will also be different with the change of the number of electric vehicle aggregations connected to the power grid or the declared power generation of users, it is not that the more the access quantity, the better. The optimized value of the upper limit of inertia aggregation of the inertia aggregators at each node is obtained through the objective function of solving the optimization problem.

[0031] Specifically, please refer toFigure 2 , Figure 2 It is a flowchart for obtaining the upper limit of inertia aggregation of inertia aggregators at each node. Obtaining the upper limit of inertia aggregation of inertia aggregators at each node includes: step S11 to step S13.

[0032] In step S11, calculate the upper limit of inertia supply facing power deficit at each node.

[0033] As an example, to find the upper limit of inertia supply facing power deficit at each node, that is, to find the maximum reducible power consumption of the electric vehicle population at each node, that is, to find the minimum value of the active power injected into this node by the upper voltage level. The objective function is expressed as follows: (2) Wherein, represents the voltage level, The lower the value, the lower the voltage level; b represents the node number connecting the upper and lower power grids; t represents the time sequence number; H aggEV.up,i-1,b,t is at time the inertia aggregated upward from the lower voltage level connected to the node to the voltage level converted into the upper limit of aggregable inertia facing power deficit by formula (1) ; is the injection power of node at voltage level at time is the maximum frequency change rate constraint of node b; is at time the non-adjustable load in the voltage level connected to the node, which is a fixed value; is at time the equivalent load of the lower voltage level connected to the node aggregated upward to the voltage level which is an adjustable load is at time the equivalent power generation of the lower voltage level connected to the node aggregated upward to the voltage level . When , is at time the non-adjustable load of the lowest voltage level connected to the node, is at time the equivalent load of the population on the lowest voltage level connected to the node, is at time The equivalent power generation of the aggregation group at the lowest voltage level is accessed at the node.

[0034] Furthermore, establish the constraint conditions for the inertia supply facing the power deficit at each node con1.1 ~Constraint conditions con1.13 .

[0035] As an example, the active power balance constraint of the power grid con1.1 is expressed as follows: (3) Wherein, and represent the node number of the network level; is the time the active power injection of the node, is the time the non-adjustable active load of the node itself, is the time the active power output of the generator on the node, is the time the equivalent aggregated load on the node, is the time the equivalent aggregated power generation of the node.

[0036] The reactive power constraint of the power grid con1.2 is expressed as follows: (4) Wherein, represents the node number of the network level; is the time the reactive power injection of the node, is the time the non-adjustable reactive load of the node itself, is the time the reactive power output of the generator on the node, is the time the equivalent aggregated reactive load on the node, is the time the equivalent aggregated reactive power of the node.

[0037] The current relationship constraint between power grid nodes con1.3It is expressed as follows: (5) Wherein, is the moment 、 the current value between nodes, 、 is the moment 、 the active power and reactive power flowing between nodes.

[0038] The voltage relationship constraint between power grid nodes con1.4 is expressed as follows: (6) Wherein, is the moment the voltage value of the node, is the moment the voltage value of the node, 、 is 、 the resistance and reactance values between nodes. The charge and discharge power constraint for the electric vehicle group to participate in inertia control con1.5 is expressed as follows: (7) Wherein, is the charge and discharge power of the electric vehicle group, are the minimum and maximum values of the charge and discharge power of the electric vehicle group.

[0039] The power generation power constraint for the inertia control participated by the distributed photovoltaic power generation con1.6 is expressed as follows: (8) Wherein, is the distributed photovoltaic power generation power, are the minimum and maximum values of the distributed photovoltaic power generation power.

[0040] The active power output constraint of generators at each level con1.7 is expressed as follows: (9) Wherein, is i the active power output of the generator at the is i the lower limit of the active power output of the generator at the isi The upper limit of the active power output of the generator at each level.

[0041] The reactive power output constraints of the generators at each level con1.8 are expressed as follows: (10) where is i the reactive power output of the generator at level is i the lower limit of the reactive power output of the generator at level is i the upper limit of the reactive power output of the generator at level.

[0042] The active power constraints of the branches between the nodes at each level con1.9 are expressed as follows: (11) where is the time i the node at level , node the active power of the branch between is i the lower limit of the active power of the branch between the node at level , node the upper limit of the active power of the branch between is i the upper limit of the active power of the branch between the node at level , node .

[0043] The reactive power constraints of the branches between the nodes at each level con1.10 are expressed as follows: (12) where is the time i the node at level , node the reactive power of the branch between is i the lower limit of the reactive power of the branch between the node at level , node the upper limit of the reactive power of the branch between is i the upper limit of the reactive power of the branch between the node at level , node .

[0044] The voltage constraint of node k is expressed as follows: con1.11 ​ (13) in, for Time Node The voltage, For Node The voltage lower limit, For Node The upper voltage limit.

[0045] node Voltage Constraint con1.12 It is expressed as follows: (14) in, for Time Node The voltage, For Node The voltage lower limit, For Node The upper voltage limit.

[0046] Branch current constraints between nodes at each level con1.13 It is expressed as follows: (16) in, for Time Node ,node The branch current between For Node ,node The lower limit of branch current between For Node ,node The upper limit of branch current between .

[0047] In step S12, the upper limit of inertia supply for power redundancy at each node is calculated.

[0048] As an example, the upper limit of the inertia supply for power redundancy under each node is calculated, that is, the equivalent maximum power load of the electric vehicle cluster at each node is calculated, that is, the maximum active power injected into each node by the upper voltage level is calculated. The objective function is expressed as follows: (17) in, H aggEV.down,i-1,b,t for time The lower voltage level connected at the node is aggregated upward to the voltage level The inertia of is converted into the upper limit of the aggregated inertia for power redundancy by formula (1): ; is the injection power of the node at voltage level at time of ; is time the voltage level connected at node The non-adjustable load in is a fixed value; is time The equivalent load of the lower voltage level connected at node aggregated upward to voltage level is an adjustable load. When is is time the non-adjustable load of the lowest voltage level connected at node is time the equivalent load of the electric vehicle group on the lowest voltage level connected at node

[0049] Furthermore, the constraint conditions for inertia supply facing power redundancy at each node are established con2.1 ~Constraint conditions con2.13 . It should be noted that in addition to the objective function, the upper limit of inertia supply facing power redundancy at each node is different from that of inertia supply facing power deficit at each node. In the constraint conditions, since the upper limit of inertia supply facing power redundancy at each node is to find the maximum power consumption load of the equivalent electric vehicle group at each node, the increase in load may lead to an increase in the power flow burden of the lower-level power grid. Therefore, not only the power flow constraint of the upper-level power grid but also the power flow constraint of the lower-level power grid need to be considered.

[0050] As an example, the active power balance constraint of the power grid con2.1 is expressed as follows: (18) where and represent the node numbers of the i th level of the power grid; is time the injected active power of node is time the non-adjustable active load of node itself is time the active power output of the generator on node is time the equivalent aggregated load on node is Moment Equivalent aggregated power generation at the node

[0051] Reactive power constraint of the power grid con2.2 Is expressed as follows (19) Wherein Represents the node number of the i Level of the power grid Is Moment Reactive power injected at the node Is Moment Unadjustable reactive load of the node itself Is Moment Reactive power output of the generator at the node Is Moment Equivalent aggregated reactive load at the node Is Moment Equivalent aggregated reactive power at the node

[0052] Current relationship constraint between power grid nodes con2.3 Is expressed as follows (20) Wherein Is Moment And Current value between nodes And Is Moment And Active power and reactive power flowing between nodes

[0053] Voltage relationship constraint between power grid nodes con2.4 Is expressed as follows (21) Wherein Is Moment Voltage value of the node Is Moment Voltage value of the node And Is And Resistance and reactance values between nodes

[0054] Charging and Discharging Power Constraints for the Inertia Control Participated by Electric Vehicle Aggregates con2.5 It is expressed as follows: (22) Wherein, is the charging and discharging power of the electric vehicle aggregates, are the minimum and maximum values of the charging and discharging power of the electric vehicle aggregates.

[0055] Power Generation Constraints for the Inertia Control Participated by Distributed Photovoltaic Systems con2.6 It is expressed as follows: (23) Wherein, is the power generation of the distributed photovoltaic system, are the minimum and maximum values of the power generation of the distributed photovoltaic system.

[0056] Active Power Output Constraints of Generators at Each Level con2.7 It is expressed as follows: (24) Wherein, is i the active power output of the generator at level is i the lower limit of the active power output of the generator at level is i the upper limit of the active power output of the generator at level

[0057] Reactive Power Output Constraints of Generators at Each Level con2.8 It is expressed as follows: (25) Wherein, is i the reactive power output of the generator at level is i the lower limit of the reactive power output of the generator at level is i the upper limit of the reactive power output of the generator at level

[0058] Active Power Constraints of Branches between Nodes at Each Level con2.9 It is expressed as follows: (26) Wherein, is the active power of the branch between node i at level and node at time is i the node at level , the active power lower limit of the branch between nodes is the active power upper limit of the branch between node i at layer and node . Further, the power flow constraint of the lower-level power grid is calculated using formula (26).

[0059] The reactive power constraint of the branch between nodes at each layer con2.10 is expressed as follows: (27) where is the reactive power of the branch between node i at layer and node , is i the lower limit of the reactive power of the branch between node at layer and node , i is the upper limit of the reactive power of the branch between node at layer

[0060] and node k . Further, the power flow constraint of the lower-level power grid is calculated using formula (27). con2.11 The voltage constraint of node where is the voltage of node at time , is the lower voltage limit of node ,

[0061] and con2.12 is the upper voltage limit of node (29) where is the voltage of node at time , is the lower voltage limit of node ,

[0062] con2.13 The current constraint of the branch between nodes at each layer con2.13It is expressed as follows: (30) Wherein, is the time node , the branch current between nodes ; is , the lower limit of the branch current between nodes ; is , the upper limit of the branch current between nodes .

[0063] In step S13, take the minimum value between the upper limit of the aggregable inertia facing power deficit and the upper limit of the aggregable inertia facing power redundancy of each node as the upper limit of the inertia aggregation of the inertia aggregator under each node.

[0064] As an example, the maximum reducible power consumption and the maximum increasable power consumption of an electric vehicle group are usually inconsistent. When calculating the upper limit of the aggregable inertia, find out the maximum reducible power consumption and the maximum increasable power consumption of the electric vehicle group respectively. Take the minimum value between the maximum reducible power consumption of each node and the maximum increasable power consumption of each node, that is, the minimum value between the upper limit of the aggregable inertia facing power deficit and the upper limit of the aggregable inertia facing power redundancy of each node, as the upper limit of the maximum available power generation or consumption, that is, the upper limit of the inertia aggregation of the inertia aggregator under each node, which is specifically expressed as follows: (31) Wherein, represents the grid voltage level, the lower the value, the lower the voltage level; b represents the node number; t represents the time sequence number; and respectively represent the upper limit of the aggregable inertia facing power deficit and the upper limit of the aggregable inertia facing power redundancy.

[0065] In step S2, please refer to Figure 1 step S2 therein, and construct an optimal aggregation model of the inertia aggregator with the goal of minimizing the inertia aggregation cost of the inertia aggregator, and obtain the optimal inertia aggregation cost of the inertia aggregator under the known winning bid amount.

[0066] As an example, different electric vehicle charging piles and supporting photovoltaics will have different virtual inertias, and the quotes provided will also be different. Take the winning bid amount of inertia trading with the lowest cost to ensure that the inertia aggregator can efficiently obtain the benefits of inertia trading.

[0067] Taking the minimum inertia aggregation cost of the inertia aggregator as the objective function, an optimal aggregation model for the inertia aggregator is constructed, and the objective function is expressed as follows: (31) where, is the quotation of the k-th distributed electric vehicle, represents the adjustable power generation or power consumption supplied by the k-th distributed electric vehicle; is the quotation of the k-th supporting distributed photovoltaic, represents the adjustable power generation or power consumption supplied by the k-th supporting distributed photovoltaic; represents the maximum allowable frequency change rate required by node b; represents the supply inertia of the k-th distributed electric vehicle at time t, represents the supply inertia of the k-th supporting distributed photovoltaic at time t.

[0068] Furthermore, the constraint conditions for constructing the optimal aggregation model of the inertia aggregator include: the upper limit constraint of the adjustable inertia of each distributed electric vehicle and each supporting distributed photovoltaic, the inertia aggregation constraint balanced with the winning bid quantity, the lower-level power grid power flow constraint, and the maximum frequency change rate constraint of the new power system.

[0069] Taking the inertia aggregator as an example, when aggregating each distributed electric vehicle group, it is necessary to consider the upper limit of its adjustable inertia, and the adjustment amount of each distributed electric vehicle should be within this upper limit to ensure the normal operation of the power source. The upper limit constraint condition of the adjustable inertia of each distributed electric vehicle con3.1 is expressed as follows: (32) where, is the upper limit of the adjustable capacity of the k-th distributed electric vehicle at time t, which is determined by the inertia aggregation upper limit of the inertia aggregator under each node obtained in step S1; is the lower limit of the adjustable capacity of the k-th distributed electric vehicle group at time t.

[0070] Taking the inertia aggregator as an example, when aggregating each supporting distributed photovoltaic, it is necessary to consider the upper limit of its adjustable inertia, and the adjustment amount of each supporting distributed photovoltaic should be within this upper limit to ensure the normal operation of the power source. The upper limit constraint condition of the adjustable inertia of each supporting distributed photovoltaic con3.2 is expressed as follows: (33) where, is the upper limit of the adjustable capacity of the k-th supporting distributed photovoltaic at time t, which is determined by the inertia aggregation upper limit of the inertia aggregator under each node obtained in step S1; is the lower limit of the adjustable capacity of the k-th supporting distributed photovoltaic at time t.

[0071] As an example, when the inertia aggregator participates in the spot market bidding, it is required that the inertia aggregator provides an inertia supply balanced with its winning bid through the integration and coordination of multiple distributed electric vehicle clusters, and the inertia aggregation constraint conditions balanced with the winning bid con3.3 are expressed as follows: (34) wherein, is the adjustable power generation or power consumption supplied by the kth electric vehicle cluster, including: the adjustable power generation or power consumption supplied by the kth distributed electric vehicle and the adjustable power generation or power consumption supplied by the kth supporting distributed photovoltaic ; represents i the winning bid volume of the inertia aggregator at the -1 level participating in the inertia market transaction.

[0072] As an example, when the inertia aggregator aggregates inertia in the lower - layer power grid, it constraints the active power and reactive power of the branch between node k and node , and the lower - layer power grid power flow constraint conditions con3.4 ~constraint conditions con3.5 are expressed as follows: (35) (36) wherein, represents t the active power of the branch between node k and node at time represents the lower limit of the active power of the branch between node k and node , represents the upper limit of the active power of the branch between node k and node , represents t the reactive power of the branch between node k and node at time represents the lower limit of the reactive power of the branch between node k and node , represents the upper limit of the reactive power of the branch between node k and node .

[0073] For example, when the inertia provided by the inertia supplier acts together with the inertia of the new power system, it should be able to meet the frequency stability requirements of the new power system and the constraint condition of the maximum frequency change rate of the new power system con3.6 It is expressed as follows: (37) Among them, represents the system disturbance, represents the system rated frequency, represents the total active power output of the system, represents the original inertia of the system, represents the set maximum frequency change rate of the system.

[0074] For example, after constructing the optimal aggregation model of the inertia aggregator, the genetic algorithm is used to solve the optimal aggregation model of the inertia aggregator to obtain the optimal aggregation cost of the inertia aggregator under the known winning bid quantity.

[0075] In step S3, please refer to Figure 1 step S3 in it. According to the optimal aggregation model of the inertia aggregator, obtain the quotation sample data set, establish a curve fitting model based on the deep residual network combined with the adaptive activation function, and input the quotation sample data set into the curve fitting model to calculate the quotation curve of the inertia aggregator under the unknown winning bid quantity.

[0076] Before conducting inertia trading, the inertia aggregator needs to make an inertia quotation. However, the cost calculation of the optimal aggregation model of the inertia aggregator requires knowing the actual winning bid quantity of the inertia trading. Therefore, the inertia aggregator needs to obtain the cost-inertia supply curve to support the quotation of the inertia trading. The cost-inertia supply curve is optimized from each optimal aggregation cost. In theory, the curve has countless points, and it is impossible to optimize the inertia aggregation cost each time. Therefore, in this application, an improved deep residual network is adopted, and some optimized values of the inertia aggregation cost are selected as key points to form a quotation sample data set as training data, and a curve fitting model is established through the improved deep residual network to obtain the quotation curve of the cost-inertia supply.

[0077] For example, in the curve fitting scenario, the following improvements are made to the standard deep residual network in this application: Activation function replacement: The self-gating activation function (Swish function) is used to replace the rectified linear unit function (ReLU function); Add a batch normalization (BN) layer after the residual addition to avoid training instability caused by the dimensional difference between the input and the residual; Remove the convolution operation and replace it with a fully connected layer to adapt to the one-dimensional input-output structure; Iteratively update the hyperparameters through the genetic algorithm to autonomously optimize the curve fitting accuracy.

[0078] Specifically, please refer to Figure 3 , Figure 3 which is the flowchart of step S3. Based on an improved deep residual network combined with a curve fitting model of an adaptive activation function, step S3 includes: steps S31 to S36.

[0079] Step S31: Obtain a quotation sample data set according to the optimal aggregation model of inertia aggregators, and preprocess the quotation sample data set.

[0080] As an example, the quotation sample data set is represented as follows: (38) Wherein, represents the q th group of inertia winning scalars, represents the quotation of the inertia aggregator under the q th group of inertia winning scalars and Q represents the set of all data points.

[0081] Furthermore, preprocessing the quotation sample data set includes: normalization processing; the normalization processing is represented as follows: (39) Wherein, represents the normalized inertia winning scalar, represents the normalized quotation, μ x represents the mean value of inertia winning scalars, μ y represents the mean value of quotations; δ x represents the maximum value of inertia winning scalars; μ y represents the maximum value of quotations.

[0082] Furthermore, divide the preprocessed quotation sample data set into a training set, a validation set, and a test set according to a ratio.

[0083] In an embodiment of the present application, the preprocessed quotation sample data set is divided into a training set, a validation set, and a test set according to a ratio of 8:1:1.

[0084] Step S32: Initialize the model structure and parameters of the deep residual network.

[0085] As an example, input network hyperparameters into the deep residual network, including: the number of input residual blocks N, the input hidden layer dimension d, and the input adaptive activation function .

[0086] As an example, the genetic algorithm is used to optimize the network hyperparameters, and the number of residual blocks N and the hidden layer dimension d are changed.

[0087] As an example, the parameter initialization includes: initializing the parameters of the deep residual network using HE initialization, which is expressed as follows: (40) Among them, W represents the weight parameter of the deep residual network; N represents the normal distribution; 0 represents the mean of the normal distribution, that is, the mean of the initialized weight parameter is 0; represents the standard deviation of the normal distribution, d in represents the input hidden layer dimension.

[0088] Step S33: Input the offer sample data set into the curve fitting model for offer curve fitting.

[0089] As an example, the input layer inputs the preprocessed inertia bid scalar and maps it to a high-dimensional space, which is expressed as follows: (41) Among them, W 0 represents the weight matrix, and maps the preprocessed inertia bid scalar to the d-dimensional feature space; b 0 represents the bias vector.

[0090] Furthermore, perform residual block calculation. Taking the t-th residual block as an example (t = 1, 2,..., N), the internal calculation process of the t-th residual block includes: Step S331 to Step S333: Step S331: Process through the first fully connected layer, batch normalization, and self-gating activation function to obtain the intermediate feature state inside the residual block, which is expressed as follows: (42) Among them, W t,1 represents the residual block t the weight matrix of the first fully connected layer inside; b t,1 represents the residual block t the bias vector of the first fully connected layer inside; h t-1 represents the (t - 1)-th residual block; BN represents the batch normalization operation; h mid represents the intermediate feature state inside the residual block t and is the bridge connecting two layers.

[0091] Step S332: After passing through the second fully connected layer, obtain the output features of the second fully connected layer, which are expressed as follows: (43) where, h res represents the output features of the second fully connected layer within the residual block t ; W t,2 represents the weight matrix of the second fully connected layer within the residual block t ; b t,2 represents the bias vector of the second fully connected layer within the residual block t .

[0092] Step S333: Connect the residual blocks to obtain the quote curve.

[0093] As an example, the t-th residual block h t is expressed as follows: (44) Furthermore, the output layer outputs the quote curve, which is expressed as follows: (45) where, y pred represents the quote curve processed by the deep residual network once; h N represents the last residual block; W out represents the weight matrix of the output layer; b out represents the bias vector of the output layer.

[0094] Step S34: Determine whether the quote curve has completed fitting. If so, enter Step S36; if not, enter Step S35.

[0095] As an example, the mean squared error is used for loss calculation, which is expressed as follows: (46) where, y pred,i represents the predicted value of the i-th quote curve output by the deep residual network; y norm,i represents the standard value of the i-th quote curve output by the deep residual network in the validation set; M represents the number of samples; represents the loss on the validation set.

[0096] Furthermore, the conditions for the quote curve to complete fitting include: on the validation set The loss is less than or equal to the set threshold value , which is expressed as follows: (46) Step S35: Use the genetic algorithm to change the number N of residual blocks and the dimension d of the hidden layer for iterative optimization, and return to step S32.

[0097] Step S36: Output the final bid curve.

[0098] This application also provides a new energy and electric vehicle charging pile inertia aggregation system for executing the above-mentioned new energy and electric vehicle charging pile inertia aggregation method. Please refer to Figure 4 , Figure 4 is the module diagram of the new energy and electric vehicle charging pile inertia aggregation system provided by this application. The new energy and electric vehicle charging pile inertia aggregation system provided by this application includes: an inertia aggregation upper limit calculation module 41 of the inertia aggregator, an optimal aggregation cost module 42 of the inertia aggregator, and a bid curve calculation module 43 of the inertia aggregator; the inertia aggregation upper limit calculation module 41 of the inertia aggregator, the optimal aggregation cost module 42 of the inertia aggregator, and the bid curve calculation module 43 of the inertia aggregator are connected in sequence.

[0099] As an example, the inertia aggregation upper limit calculation module 41 of the inertia aggregator is used to obtain the inertia aggregation upper limit of the inertia aggregator at each node; The optimal aggregation cost module 42 of the inertia aggregator is used to construct an optimal aggregation model of the inertia aggregator with the goal of minimizing the inertia aggregation cost of the inertia aggregator, and obtain the optimal inertia aggregation cost of the inertia aggregator under the known winning bid quantity; The bid curve calculation module 43 of the inertia aggregator is used to obtain a bid sample data set according to the optimal aggregation model of the inertia aggregator, establish a curve fitting model based on the deep residual network combined with the adaptive activation function, and input the bid sample data set into the curve fitting model to calculate the bid curve of the inertia aggregator under the unknown winning bid quantity.

[0100] This application also provides a computer program product, including a computer program, which when executed by a processor implements the steps of the above-mentioned new energy and electric vehicle charging pile inertia aggregation method.

[0101] The present application provides a method, system and computer program product for aggregating the inertia of new energy and electric vehicle charging piles, which solves the problem that distributed electric vehicle inertia resources are not considered in traditional inertia supply. By using an optimization algorithm considering the influence of distributed electric vehicles, the upper limit of the maximum inertia supply at each node is calculated. Subsequently, from the perspective of the inertia aggregator, the optimal inertia aggregation is carried out under the known winning bid quantity. Finally, based on the artificial intelligence method, the bid curve of the inertia aggregator participating in the inertia market is fitted. In the problem of maximizing the interests of the inertia aggregator participating in the inertia transaction, an optimal aggregation model of the inertia aggregator is constructed to calculate the winning bid quantity for the inertia aggregator to achieve inertia transactions at the lowest cost, thereby more accurately predicting the bid of the inertia aggregator. In the problem of calculating the bid of the inertia aggregator participating in the inertia transaction, several key nodes are first selected for traditional bid calculation, and then the improved deep residual network of the present application is used to fit the curve with the key nodes to obtain the bid curve, which solves the problem of huge computational complexity caused by the need to optimize and calculate each node in the traditional method, significantly reduces the computational complexity, and improves the bid accuracy. The present application more accurately predicts the bid curve of the inertia aggregator while considering the influence of distributed electric vehicles, reduces the computational complexity, effectively improves the reliability and efficiency of the inertia service, and has significant practical value.

[0102] Although the present application has been disclosed above by way of examples, it is not intended to limit the present application. Any person with ordinary knowledge in the technical field to which the present application pertains may make some modifications and refinements without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application shall be subject to that defined by the appended patent application scope.

Claims

1. A method for aggregating the inertia of a new energy and electric vehicle charging pile, characterized in that, including Obtain the inertia aggregation upper limit of the inertia aggregator at each node; Taking the minimum inertia aggregation cost of the inertia aggregator as the objective, construct an optimal aggregation model for the inertia aggregator, and obtain the optimal inertia aggregation cost of the inertia aggregator under the known winning bid quantity; According to the optimal aggregation model of the inertia aggregator, obtain the quotation sample data set, establish a curve fitting model based on the deep residual network combined with the adaptive activation function, and input the quotation sample data set into the curve fitting model to calculate the quotation curve of the inertia aggregator under the unknown winning bid quantity.

2. The inertia aggregation method for new energy and electric vehicle charging piles according to claim 1, wherein Obtain the inertia aggregation upper limit of the inertia aggregator at each node, including: Calculate the inertia supply upper limit facing power deficit at each node; Calculate the inertia supply upper limit facing power redundancy at each node; Take the minimum value between the aggregable inertia upper limit facing power deficit and the aggregable inertia upper limit facing power redundancy at each node as the inertia aggregation upper limit of the inertia aggregator at each node.

3. The inertia aggregation method for new energy and electric vehicle charging piles according to claim 2, wherein, The objective function for calculating the inertia supply upper limit facing power deficit at each node is expressed as follows: in, Indicates the voltage level, The lower the value, the lower the voltage level; b represents the node number connecting the upper and lower power grids; t represents the time sequence number; H aggEV.up,i-1,b,t for time The lower voltage level connected at the node is aggregated upward to the voltage level The inertia is converted into the upper limit of the aggregated inertia for power shortage ; for Voltage level at all times of The injected power of the node; is the maximum frequency change rate constraint of node b; for time Voltage level connected to the node The non-adjustable load in is a fixed value; for time The lower voltage level connected at the node is aggregated upward to the voltage level The equivalent load is an adjustable load. for time The lower voltage level connected at the node is aggregated upward to the voltage level The equivalent power generation power of hour, for time The node is connected to a non-adjustable load with the lowest voltage level. for time The equivalent load of the cluster connected to the lowest voltage level at the node, for time The equivalent power generation of the cluster connected to the lowest voltage level at the node; The objective function for calculating the inertia supply upper limit facing power redundancy at each node is expressed as follows: Among them, H aggEV.down,i-1,b,t is the moment when the lower voltage level connected at the node aggregates upward to the voltage level inertia, which is converted into the upper limit of aggregable inertia facing power redundancy ; is the injection power of the node at the voltage level at the moment is the moment when the voltage level connected at the node is the non-adjustable load, which is a fixed value; is the moment when the lower voltage level connected at the node aggregates upward to the voltage level equivalent load, which is an adjustable load; when it is is the moment when the non-adjustable load of the lowest voltage level connected at the node is the moment when the equivalent load of electric vehicles on the lowest voltage level connected at the node The constraint conditions for calculating the inertia supply upper limit facing power deficit at each node and calculating the inertia supply upper limit facing power redundancy at each node include: Active power balance constraint of the power grid, reactive power constraint of the power grid, current relationship constraint between grid nodes 、 Voltage relationship constraint between grid nodes 、 Charge and discharge power constraint for distributed electric vehicles participating in inertia control 、 Power generation power constraint for distributed photovoltaic participating in inertia control 、 Active power output constraint of generators at all levels 、 Reactive power output constraint of generators at all levels 、 Branch active power constraint between nodes at all levels 、 Branch reactive power constraint between nodes at all levels 、 Voltage constraint of each node 、 Branch current constraint between nodes at all levels.

4. The inertia aggregation method for new energy and electric vehicle charging piles according to claim 1, wherein The objective function of the optimal aggregation model of the inertia aggregator is expressed as follows: Among them, is the quotation of the k-th distributed electric vehicle, indicating the adjustable power generation or power consumption of the k-th distributed electric vehicle supply; is the quotation of the k-th supporting distributed photovoltaic, indicating the adjustable power generation or power consumption of the k-th supporting distributed photovoltaic supply; indicates the maximum frequency change rate required by node b; indicates the supply inertia of the k-th distributed electric vehicle at time t, indicating the supply inertia of the k-th supporting distributed photovoltaic at time t; The constraint conditions of the optimal aggregation model of the inertia aggregator include: the adjustable inertia upper limit constraints of each distributed electric vehicle and each supporting distributed photovoltaic, the inertia aggregation constraint balanced with the winning bid quantity, the lower-layer power grid power flow constraint, and the maximum frequency change rate constraint of the new power system.

5. The inertia aggregation method for new energy and electric vehicle charging piles according to claim 1, wherein Modify the adopted deep residual network, including: Use the self-gating activation function to replace the rectified linear unit; Add a batch normalization layer after the residual addition; Use a fully connected layer to replace the convolution operation; Iteratively update the hyperparameters through the genetic algorithm to autonomously optimize the curve fitting accuracy.

6. The inertia aggregation method for new energy and electric vehicle charging piles according to claim 5, wherein According to the optimal aggregation model of the inertia aggregator, obtain the quotation sample data set, establish a curve fitting model based on the deep residual network combined with the adaptive activation function, and input the quotation sample data set into the curve fitting model to calculate the quotation curve of the inertia aggregator under the unknown winning bid quantity, including: Step S31: Obtain the quotation sample data set according to the optimal aggregation model of the inertia aggregator, and preprocess the quotation sample data set; Step S32: Initialize the model structure and parameters of the deep residual network; Step S33: Input the quotation sample data set into the curve fitting model for quotation curve fitting; Step S34: Determine whether the quotation curve is completed for fitting. If so, enter step S36; if not, enter step S35; Step S35: Use the genetic algorithm to change the number of residual blocks N and the hidden layer dimension d for iterative optimization, and return to step S32; Step S36: Output the final quotation curve.

7. The method for aggregating the inertia of new energy and electric vehicle charging piles according to claim 6, wherein, Input the quotation sample data set into the curve fitting model for quotation curve fitting, including: After being processed by the first fully connected layer, batch normalization, and the self-gating activation function, obtain the intermediate feature state inside the residual block, which is expressed as follows: Among them, W t,1 represents the weight matrix of the first fully-connected layer in the residual block; t in the residual block; b t,1 represents the bias vector of the first fully-connected layer in the residual block; t in the residual block; h t-1 represents the (t - 1)-th residual block; BN represents the batch normalization operation; h mid represents the intermediate feature state in the residual block t which is a bridge connecting two layers; After passing through the second fully connected layer, obtain the output features of the second fully connected layer, which is expressed as follows: Among them, h res represents the output features of the second fully connected layer within the residual block; t represents the weight matrix of the second fully connected layer within the residual block; W t,2 represents the residual block t represents the bias vector of the second fully connected layer within the residual block; b t,2 represents the residual block t represents the bias vector of the second fully connected layer within the residual block; Connect the residual blocks to obtain the quotation curve, which is expressed as follows: Among them, y pred represents the quoted price curve processed by the deep residual network once; h N represents the final residual block; W out represents the weight matrix of the output layer; b out represents the bias vector of the output layer.

8. The inertia aggregation method for new energy and electric vehicle charging piles according to claim 6, wherein The conditions for the quotation curve to complete fitting include: the loss on the validation set is less than or equal to the set threshold , which is expressed as follows: Among them, represents the loss on the validation set, represents the set threshold.

9. A new energy and electric vehicle charging pile inertia aggregation system for performing the new energy and electric vehicle charging pile inertia aggregation method according to any one of claims 1 to 8, characterized in that, including: The inertia aggregation upper limit calculation module, the optimal aggregation cost module of the inertia aggregator, and the bid curve calculation module of the inertia aggregator of the inertia aggregator; The inertia aggregation upper limit calculation module, the optimal aggregation cost module of the inertia aggregator, and the bid curve calculation module of the inertia aggregator of the inertia aggregator are connected in sequence.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, the steps of the new energy and electric vehicle charging pile inertia aggregation method described in any one of claims 1 to 8 are implemented.

Images