A virtual power plant-based user-side energy storage coordination optimization method

Abstract

The application relates to the technical field of energy storage regulation, in particular to a user-side energy storage collaborative optimization method based on a virtual power plant. The method comprises the following steps: aggregating all user-side energy storage resources based on a virtual power plant platform, setting multiple energy storage nodes according to an aggregation result; establishing multiple scheduling periods, generating expected adjustable resource parameters of each energy storage node in a current scheduling period according to a preset response prediction model; obtaining expected energy storage demand in the current scheduling period, generating a primary control strategy according to all expected adjustable resource parameters and the expected energy storage demand; obtaining a feedback data packet according to a preset feedback time node, judging whether to generate a primary correction instruction according to the feedback data packet; establishing multiple node categories according to different user-side energy storage equipment types, constructing corresponding response sub-models according to characteristic parameters of the node categories, realizing accurate prediction of response capabilities of all user-side energy storage nodes, and improving collaborative energy storage efficiency and collaborative control of the user side.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of energy storage regulation, in particular to a user-side energy storage collaborative optimization method based on a virtual power plant. BACKGROUND

[0002] With the acceleration of the transformation of global energy structure to clean and low-carbon, the penetration rate of distributed renewable energy (DER) represented by photovoltaic and wind power continues to increase at the user side (such as industrial parks, commercial buildings, and residential communities). At the same time, in order to improve the energy self-sufficiency rate, reduce the electricity cost, and participate in demand response, the installation scale of user-side electrochemical energy storage systems (ESS) also presents an explosive growth. These distributed resources have the characteristics of location dispersion, relatively small capacity, and different operating characteristics.

[0003] At present, most of the user-side energy storage systems mainly serve local demand, such as realizing peak-valley price difference arbitrage, improving photovoltaic self-generation and self-use rate, and guaranteeing key load power supply. This single independent operation mode is difficult to fully exert the system-level value of energy storage, and the overall economic and social benefits are not optimal. The capacity of a single user energy storage is limited, and the response ability to the grid or power market signal is weak, which is difficult to meet the requirements of large-scale and high-reliability aggregated response. SUMMARY

[0004] The purpose of the present application is to solve the above technical problems, and the present application provides a user-side energy storage collaborative optimization method based on a virtual power plant, aiming to improve the collaborative energy storage efficiency and collaborative control of the user side.

[0005] In some embodiments of the present application, all user-side energy storage resources are aggregated based on the virtual power plant technology, and multiple node categories are established according to different user-side energy storage device types. Corresponding response sub-models are constructed according to the characteristic parameters of each node category, the response ability of all user-side energy storage nodes is accurately predicted, and the collaborative energy storage efficiency and collaborative control of the user side are improved.

[0006] In some embodiments of the present application, by establishing multi-level regulation, the expected energy storage demand is initially allocated according to the response characteristics of different node categories, and then the operating state of the energy storage nodes in each node category is allocated again, the control efficiency of the user-side energy storage nodes is improved, and the overall operating income is improved.

[0007] In some embodiments of the present application, a user-side energy storage collaborative optimization method based on a virtual power plant is provided, which comprises:

[0008] All user-side energy storage resources are aggregated based on a virtual power plant platform, and multiple energy storage nodes are set according to the aggregation result;

[0009] Multiple scheduling cycles are established, and the expected adjustable resource parameters of each energy storage node in the current scheduling cycle are generated based on the preset response prediction model.

[0010] Obtain the expected energy storage demand for the current scheduling period, and generate a primary control strategy based on all expected adjustable resource parameters and expected energy storage demand;

[0011] Obtain feedback data packets according to preset feedback time nodes, and determine whether to generate a first-level correction instruction based on the feedback data packets;

[0012] When setting up multiple energy storage nodes, the following are included:

[0013] Establish a sequence of energy storage nodes A, A=(a1,a2…a ... i …a n ), where a i Let be the i-th energy storage node; n is the number of energy storage nodes.

[0014] In some embodiments of this application, the preset response prediction model includes:

[0015] Multiple equipment characteristic indicators are generated based on all energy storage node parameters;

[0016] Multiple node categories are generated based on all device characteristic indicators;

[0017] Establish a node category sequence B, B=(b1, b2, ..., bb2) i …b m ), where b i Let m be the node category of the i-th node; m is the number of node categories.

[0018] Construct response sub-models for each node category in sequence;

[0019] Establish a response sub-model sequence P, P = (p1, p2, ..., p...) i …p m ), where p i For the response sub-model of the i-th node category;

[0020] Construct a response prediction model based on the response sub-model sequence P.

[0021] In some embodiments of this application, the step of sequentially constructing the response sub-model for each node category includes:

[0022] Based on the node category sequence B, b is set sequentially. i For the target node category;

[0023] Generate multiple perturbation feature indices for the target node category;

[0024] Construct an initial evaluation model for the target node category;

[0025] obtain all associated energy storage nodes of the target node category;

[0026] establish a sequence A1 of associated energy storage nodes of the target node category;

[0027] A1=(a 11 ,a 12 …a 1i …a 1n1 ), wherein a 1i is the i-th associated energy storage node; and n1 is the number of associated energy storage nodes of the target node category;

[0028] according to the sequence A1 of associated energy storage nodes, sequentially set a 1i as the target associated energy storage node;

[0029] obtain a historical operation package of the target associated energy storage node, and generate a disturbance deviation value c of the target associated energy storage node according to the historical operation package;

[0030] preset a disturbance deviation threshold C1;

[0031] if c>C1, generate a compensation sub-strategy of the target associated energy storage node;

[0032] sequentially determine whether each associated energy storage node generates a compensation sub-strategy;

[0033] generate a response sub-model of the target program category according to all compensation sub-strategies and an initial evaluation model.

[0034] In some embodiments of the present application, when the disturbance deviation value c of the target associated energy storage node is generated, it includes:

[0035] c=[ η i *(v i -v' i )];

[0036] wherein θ1 is the number of disturbance characteristic indexes of the target node category; η i is the influence factor of the i-th disturbance characteristic index in the target node category; v i is the reference value of the i-th disturbance characteristic index generated based on the historical operation package; and v' i is the standard reference value of the i-th disturbance characteristic index in the target node category.

[0037] In some embodiments of the present application, when the expected adjustable resource parameter of each energy storage node in the current scheduling period is generated, it includes:

[0038] set a first prediction time node of the current scheduling period;

[0039] The response prediction model generates the response capability parameters of each energy storage node at a first prediction time node;

[0040] The response capability sub-tables of each node category are generated according to all the response capability parameters;

[0041] The expected adjustable resource table of the current scheduling period is generated according to all the response capability sub-tables.

[0042] In some embodiments of the present application, when generating the primary control strategy, the following steps are included:

[0043] The expected energy storage demand is generated based on the expected fluctuation parameters of the power grid;

[0044] The primary allocation strategy is generated according to the expected energy storage demand and the response capability sub-tables of each node category;

[0045] The energy storage sub-demand of each node category is generated according to the primary allocation strategy;

[0046] The secondary allocation strategies of each node category are generated in sequence;

[0047] The primary control strategy is generated according to all the secondary allocation strategies.

[0048] In some embodiments of the present application, when generating the secondary allocation strategy of each node category, the following steps are included:

[0049] The node category number sequence B is set in sequence according to the node category number sequence B; i to be allocated;

[0050] The associated energy storage node sequence A2 of the node category to be allocated is established;

[0051] A2=(a 21 ,a 22 …a 2i …a 2n2 ), wherein a 2i is the i th associated energy storage node of the node category to be allocated; n2 is the number of associated energy storage nodes of the node category to be allocated;

[0052] The response evaluation values of each associated energy storage node in the associated energy storage node sequence A2 are generated;

[0053] The response scheduling order of the node category to be allocated is generated according to all the response evaluation values;

[0054] The energy storage sub-demand of the node category to be allocated is obtained;

[0055] The secondary allocation strategy of the node category to be allocated is generated according to the response scheduling order and the energy storage sub-demand.

[0056] In some embodiments of the present application, when generating the response evaluation value of each associated energy storage node in the associated energy storage node sequence A2, the following steps are included:

[0057] According to the associated energy storage node sequence A2, a 2i is set in turn

[0058] The response evaluation value f of the associated energy storage node to be evaluated is generated.

[0059] f = e1*Q1*c' + e2*Q2*[ β i *s i ];

[0060] Wherein e1 is a preset first fixed coefficient; e2 is a preset second fixed coefficient; Q1 is a preset first weight coefficient; Q2 is a preset second fixed coefficient; c' is the perturbation deviation value of the associated energy storage node to be evaluated; θ2 is the number of response resource indicators; β i is the influence factor of the i-th response resource indicator; s i is the reference value of the i-th response resource indicator of the associated energy storage node to be evaluated.

[0061] In some embodiments of the present application, when determining whether to generate a first correction instruction according to the feedback data packet, the following steps are included:

[0062] A plurality of feedback time nodes are preset.

[0063] The feedback data packet of the current feedback time node is obtained.

[0064] A correction evaluation value d of the current feedback time node is generated according to the feedback data packet.

[0065] d = e3*Q3*[ λ i *k i ]+e4*Q4*[ r i *j i ];

[0066] Wherein e3 is a preset third weight coefficient; e4 is a preset fourth weight coefficient; Q3 is a preset third fixed coefficient; Q4 is a preset fourth fixed coefficient; is the number of energy storage demand indicators; λ i is the influence factor of the i-th energy storage demand indicator; k i is the fluctuation value of the i-th energy storage demand; r i is the influence factor of the i-th energy storage node in the energy storage node sequence A; j i is the response capacity change value of the i-th energy storage node.

[0067] A preset correction evaluation value threshold D1 is set.

[0068] If d>D1, a first correction instruction is generated at the current feedback time node.

[0069] In some embodiments of the application, the first correction instruction comprises:

[0070] Real-time state parameters of each energy storage node are obtained.

[0071] An expected adjustable resource table is updated according to all real-time state parameters in response to the prediction model.

[0072] An actual energy storage demand of the current feedback time node is obtained.

[0073] A second control strategy is generated according to the updated expected adjustable resource table and the actual energy storage demand.

[0074] Compared with the prior art, the user-side energy storage collaborative optimization method based on a virtual power plant according to an embodiment of the application has the following beneficial effects:

[0075] All user-side energy storage resources are aggregated based on the virtual power plant technology, and multiple node categories are established according to different user-side energy storage device types. Corresponding response sub-models are constructed according to characteristic parameters of each node category, precise prediction of response capabilities of all energy storage nodes on the user side is realized, and user-side collaborative energy storage efficiency and collaborative control are improved.

[0076] Through multi-level regulation, the expected energy storage demand is initially allocated according to the response characteristics of different node categories, and then secondarily allocated according to the operating states of energy storage nodes in each node category, the control efficiency of user-side energy storage nodes is improved, and the overall operating income is improved. BRIEF DESCRIPTION OF DRAWINGS

[0077] Figure 1 is a flowchart of a user-side energy storage collaborative optimization method based on a virtual power plant according to a preferred embodiment of the application. DETAILED DESCRIPTION

[0078] The specific embodiments of the application will be further described in detail below in combination with the drawings and embodiments. The following embodiments are used to illustrate the application, but are not used to limit the scope of the application.

[0079] In the description of the present application, it needs to be understood that the terms "center", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application.

[0080] The terms "first", "second" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise stated, the meaning of "a plurality of" is two or more.

[0081] In the description of the present application, it needs to be explained that, unless otherwise explicitly specified and limited, the terms "mounting", "connection", "connection" should be understood broadly, for example, it can be fixed connection, or detachable connection, or integral connection; it can be mechanical connection, or electrical connection; it can be directly connected, or indirectly connected through intermediate medium, or the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0082] As Figure 1 shown, a user-side energy storage collaborative optimization method based on virtual power plant of the preferred embodiment of the present application comprises:

[0083] S101: aggregating all user-side energy storage resources based on the virtual power plant platform, and setting a plurality of energy storage nodes according to the aggregation result;

[0084] S102: establishing a plurality of scheduling periods, and generating expected adjustable resource parameters of each energy storage node in the current scheduling period according to a preset response prediction model;

[0085] S103: obtaining expected energy storage demand of the current scheduling period, and generating a primary control strategy according to all expected adjustable resource parameters and expected energy storage demand;

[0086] S104: obtaining feedback data packets according to preset feedback time nodes, and determining whether to generate a primary correction instruction according to the feedback data packets;

[0087] Among them, when setting a plurality of energy storage nodes, it comprises:

[0088] establishing an energy storage node sequence A, A=(a1, a2…a i …a n ), wherein ai Let be the i-th energy storage node; n is the number of energy storage nodes.

[0089] Specifically, based on virtual power plant technology, all user-side energy storage resources are aggregated to establish multiple energy storage nodes, where each energy storage node represents a user-side energy storage device.

[0090] Specifically, the preferred duration of the scheduling cycle is 1 day.

[0091] Specifically, when setting up a response prediction model, the following are included:

[0092] Multiple equipment characteristic indicators are generated based on all energy storage node parameters;

[0093] Multiple node categories are generated based on all device characteristic indicators;

[0094] Establish a node category sequence B, B=(b1, b2, ..., bb2) i …b m ), where b i Let m be the node category of the i-th node; m is the number of node categories.

[0095] Construct response sub-models for each node category in sequence;

[0096] Establish a response sub-model sequence P, P = (p1, p2, ..., p...) i …p m ), where p i For the response sub-model of the i-th node category;

[0097] Construct a response prediction model based on the response sub-model sequence P.

[0098] Specifically, multiple device characteristic indicators are established based on different user-side energy storage device parameters, thereby constructing multiple node categories. These device characteristic indicators include, but are not limited to, parameters such as device category, device response time, and device operation randomness. Each node category represents the same type of energy storage device, such as electric vehicle charging pile devices, distributed energy storage devices, central air conditioning devices, and industrial and commercial load control devices.

[0099] Specifically, response sub-models are established based on the characteristics of different node categories. For example, a single-unit model of a distributed energy storage device is built based on distributed energy storage parameters, PCS parameters, and other data, thereby generating a corresponding response sub-model. Similarly, a single-unit model is built based on the equipment parameters of a charging pile, thereby generating a corresponding response sub-model for the charging pile. These response sub-models can perform day-ahead and real-time assessments of the response capabilities of each node category based on its characteristics.

[0100] In a preferred embodiment of this application, the construction of response sub-models for each node category in sequence includes:

[0101] Based on the node category sequence B, b is set sequentially. i For the target node category;

[0102] Generate multiple perturbation feature indices for the target node category;

[0103] Construct an initial evaluation model for the target node category;

[0104] Obtain all associated energy storage nodes of the target node category;

[0105] Establish the associated energy storage node sequence A1 for the target node category;

[0106] A1=(a 11 ,a 12 …a 1i …a 1n1 ), where a 1i Let be the i-th associated energy storage node; n1 is the number of associated energy storage nodes of the target node category;

[0107] Based on the associated energy storage node sequence A1, a is set sequentially. 1i Associate energy storage nodes with the target;

[0108] Obtain the historical operation packets of the target associated energy storage node, and generate the disturbance deviation value c of the target associated energy storage node based on the historical operation packets;

[0109] Preset disturbance deviation threshold C1;

[0110] If c>C1, generate a compensation sub-strategy for the target associated energy storage node;

[0111] Sequentially determine whether each associated energy storage node generates a compensation sub-strategy;

[0112] Generate a response sub-model for the target program category based on all compensation sub-strategies and the initial evaluation model.

[0113] Specifically, the LightGBM ensemble algorithm is used to construct initial evaluation models for each node category.

[0114] Specifically, the initial evaluation model for the target node category is to obtain the corresponding historical similar day's operating curve based on the operating characteristics of the target node category, construct a prediction baseline, and calculate the expected operating status at each moment in the current cycle through a prediction algorithm.

[0115] Specifically, the prediction baseline in the initial evaluation model is calculated as follows:

[0116]

[0117] in, is the baseline load value of the k period of the jth day; is the load value of the k period of the dth day before the jth day; N is the total number of typical days. The number of typical days needs to be set according to the characteristics of the target node category. The typical days need to be selected from the consecutive operation days before the current period, and need to be of the same type (workday or public holiday) as the current period. If a date on which energy storage optimization scheduling has been performed is encountered, the previous 1 day of the same type is taken as the typical day.

[0118] Specifically, the initial evaluation model of the target node category can also generate the response capability of each time point in the remaining time of the current period according to the real-time operation state and the predicted baseline of the energy storage node.

[0119] Specifically, when generating the disturbance deviation value c of the target associated energy storage node, the following steps are included:

[0120] c=[ η i *(v i -v' i )];

[0121] Where θ1 is the number of disturbance characteristic indexes of the target node category; η i is the influence factor of the i th disturbance characteristic index in the target node category; v i is the reference value of the i th disturbance characteristic index based on the historical operation package; v' i is the standard reference value of the i th disturbance characteristic index in the target node category.

[0122] Specifically, the disturbance characteristic indexes include but are not limited to device cumulative operation time, device failure rate, device remaining life, etc. The standard reference value of each disturbance characteristic value is set according to the corresponding device parameters (optimal operating state) when the single model of the target node category is established. The greater the disturbance deviation value, the greater the probability of deviation in the initial evaluation model of the target node category when predicting the response capability of the target associated energy storage node.

[0123] Specifically, the greater the reference value of each disturbance characteristic index, the worse the operating state of the energy storage device corresponding to the current target associated energy storage node. The worse the ability to coordinate energy storage response.

[0124] Specifically, the influence factor of each disturbance deviation index can be set according to the interference degree of the response capability prediction. The greater the interference degree, the greater the corresponding influence factor.

[0125] Specifically, the compensation sub-strategy refers to generating a corresponding compensation value by analyzing the evaluation difference of the initial evaluation model for the target associated energy storage node, and modifying the original prediction result by the compensation value.

[0126] It can be understood that in the above embodiments, all user-side energy storage resources are aggregated based on virtual power plant technology, and multiple node categories are established according to different user-side energy storage device types, corresponding response sub-models are constructed according to the characteristic parameters of each node category, the response capability of all user-side energy storage nodes is accurately predicted, and the collaborative energy storage efficiency and collaborative control of the user side are improved.

[0127] In the preferred embodiments of the present application, when generating the expected adjustable resource parameters of each energy storage node in the current scheduling period, the following steps are included:

[0128] Set a first prediction time node of the current scheduling period;

[0129] Generate the response capability parameters of each energy storage node at the first prediction time node according to the response prediction model;

[0130] Generate a response capability sub-table of each node category according to all response capability parameters;

[0131] Generate an expected adjustable resource table of the current scheduling period according to all response capability sub-tables.

[0132] Specifically, the first prediction time node is set based on the day-ahead prediction principle, generally within 24-72 hours before the start of the current adjustment period. The state of each energy storage node in the current adjustment period is predicted in advance.

[0133] Specifically, the expected running state of a single energy storage node at each time in the current scheduling period is generated according to the response prediction model, thereby generating the response capability parameters of the energy storage node, which include energy storage day-ahead peak shaving response capability, energy storage day-ahead valley filling response capability, energy storage real-time peak shaving capability, and real-time valley filling capability. Each capability includes adjustable duration and scheduling power.

[0134] Specifically, when generating the primary control strategy, the following steps are included:

[0135] Generate the expected energy storage demand based on the expected fluctuation parameters of the power grid;

[0136] Generate a primary allocation strategy according to the expected energy storage demand and the response capability sub-table of each node category;

[0137] Generate an energy storage sub-demand of each node category according to the primary allocation strategy;

[0138] Generate a secondary allocation strategy of each node category in turn;

[0139] Generate the primary control strategy according to all secondary allocation strategies.

[0140] Specifically, by fusion processing all resources within a single node category, a corresponding response capability sub-table is generated, and based on the optimization principle and the actual response characteristics of each node category, the expected energy storage demand is initially allocated to generate a corresponding first allocation strategy. Then, according to the operating state of the energy storage nodes in each node category, secondary allocation is performed to improve the control efficiency of the user-side energy storage nodes and improve the overall operating income.

[0141] Specifically, according to the allocation result, the working parameters of each energy storage node are set, and a first control strategy is generated according to all working parameters.

[0142] In the preferred embodiment of the present application, when generating the secondary allocation strategy of each node category, the following steps are included:

[0143] According to the node category sequence B, b i is set in sequence.

[0144] A sequence of associated energy storage nodes A2 of the to-be-allocated node category is established;

[0145] A2=(a 21 ,a 22 …a 2i …a 2n2 ), wherein a 2i is the i-th associated energy storage node of the to-be-allocated node category; and n2 is the number of associated energy storage nodes of the to-be-allocated node category.

[0146] The response evaluation value of each associated energy storage node in the sequence of associated energy storage nodes A2 is generated;

[0147] The response scheduling order of the to-be-allocated node category is generated according to all response evaluation values;

[0148] The energy storage sub-demand of the to-be-allocated node category is obtained;

[0149] The secondary allocation strategy of the to-be-allocated node category is generated according to the response scheduling order and the energy storage sub-demand.

[0150] Specifically, the larger the response evaluation value, the higher the collaborative energy storage efficiency of the corresponding associated energy storage node, and the to-be-allocated node category needs to be preferentially selected to share the energy storage sub-demand required to be borne. Thus, the collaborative energy storage scheduling efficiency within a single node category is improved.

[0151] Specifically, when generating the response evaluation value of each associated energy storage node in the sequence of associated energy storage nodes A2, the following steps are included:

[0152] According to the sequence of associated energy storage nodes A2, a 2i is set in sequence.

[0153] generate a response evaluation value f of the to-be-evaluated associated energy storage node;

[0154] f = e1*Q1*c' + e2*Q2*[ β i *s i ];

[0155] Wherein, e1 is a preset first fixed coefficient; e2 is a preset second fixed coefficient; Q1 is a preset first weight coefficient; Q2 is a preset second fixed coefficient; c' is a disturbance deviation value of the to-be-evaluated associated energy storage node; θ2 is a response resource index number; β i is an influence factor of the i th response resource index; s i is a reference value of the i th response resource index of the to-be-evaluated associated energy storage node.

[0156] Specifically, the response resource index includes but is not limited to callable duration, schedulable power and other parameters. The greater the reference value of each response resource index, the stronger the response ability of the current to-be-evaluated associated energy storage node, and the higher the operation efficiency of the device.

[0157] Specifically, by presetting the first fixed coefficient, the disturbance deviation value c' of the to-be-evaluated associated energy storage node is processed, so that the greater the value of c', the greater the corresponding response evaluation value f.

[0158] Specifically, by presetting the first fixed coefficient and the second fixed coefficient, all parameters in the model are processed, so that each parameter in the model is in the same value range.

[0159] In the preferred embodiment of the present application, when determining whether to generate a first correction instruction according to the feedback data packet, the following steps are included:

[0160] Preset a plurality of feedback time nodes;

[0161] Obtain the feedback data packet of the current feedback time node;

[0162] Generate a correction evaluation value d of the current feedback time node according to the feedback data packet;

[0163] d = e3*Q3*[ λ i *k i ]+e4*Q4*[ r i *j i ];

[0164] Wherein, e3 is a preset third weight coefficient; e4 is a preset fourth weight coefficient; Q3 is a preset third fixed coefficient; Q4 is a preset fourth fixed coefficient; is an energy storage demand index number; λ iis the influence factor of the i th energy storage demand index; k i is the fluctuation value of the i th energy storage demand; r i is the influence factor of the i th energy storage node in the energy storage node sequence A; j i is the response capability variation value of the i th energy storage node;

[0165] A preset correction evaluation value threshold D1 is set.

[0166] If d > D1, a first-level correction instruction is generated at the current feedback time node.

[0167] Specifically, the energy storage demand index includes but is not limited to various parameters affecting power generation efficiency, such as light intensity variation affecting photovoltaic power generation, temperature, wind speed variation affecting wind power, and other parameters. The variation value of each energy storage demand index is set according to the difference between the reference value when generating the expected energy storage demand and the real-time reference value of the current feedback time node. The greater the difference, the greater the fluctuation value.

[0168] Specifically, the variation value of the response capability of each energy storage node is set according to the difference between the real-time response capability and the parameters in the expected adjustable resource table. The greater the difference, the greater the variation value.

[0169] Specifically, the correction evaluation value threshold can be set according to historical parameters. When the real-time correction evaluation value is greater than the preset correction evaluation value threshold, it means that the current first-level control strategy cannot complete the current energy storage coordination and scheduling task, and needs to be corrected in time to ensure the collaborative energy storage efficiency and collaborative control efficiency of the user side.

[0170] Specifically, all parameters in the model are normalized by presetting third and fourth fixed coefficients, so that each parameter in the model is in the same value range.

[0171] Specifically, the first-level correction instruction includes:

[0172] Obtain the real-time state parameters of each energy storage node;

[0173] The response prediction model updates the expected adjustable resource table according to all real-time state parameters;

[0174] Obtain the actual energy storage demand of the current feedback time node;

[0175] Generate a second-level control strategy according to the updated expected adjustable resource table and the actual energy storage demand.

[0176] Specifically, the response prediction model generates the response capability of each time point in the remaining time of the current period according to the real-time operating state and the prediction baseline of the energy storage node, and the updated expected adjustable resource table includes the response capability parameters of each energy storage node in the remaining time of the current scheduling period.

[0177] Specifically, the generation mode of the secondary control strategy is the same as that of the primary control strategy, and both are secondary allocation modes.

[0178] It can be understood that in the above embodiments, by establishing a plurality of feedback time nodes, the energy storage fluctuations in the current scheduling period are timely warned, and the corresponding control strategy is adjusted to ensure the collaborative energy storage efficiency and collaborative control efficiency of the user side.

[0179] According to the first concept of the present application, all user-side energy storage resources are aggregated based on virtual power plant technology, and a plurality of node categories are established according to different user-side energy storage device types. Corresponding response sub-models are constructed according to the characteristic parameters of each node category, the response capability of all user-side energy storage nodes is accurately predicted, and the collaborative energy storage efficiency and collaborative control of the user side are improved.

[0180] According to the second concept of the present application, by establishing multi-level regulation, the expected energy storage demand is initially allocated according to the response characteristics of different node categories, and then secondarily allocated according to the operating state of the energy storage nodes in each node category, the control efficiency of the user-side energy storage nodes is improved, and the overall operating income is improved.

[0181] The above is only the preferred embodiment of the present application, and it should be pointed out that for ordinary skilled persons in the technical field, several improvements and replacements can be made without departing from the technical principles of the present application, and these improvements and replacements should also be considered as the protection scope of the present application.

Claims

1. A virtual power plant based user side energy storage collaborative optimization method, characterized in that, The application comprises the following steps: Aggregating all user-side energy storage resources based on a virtual power plant platform, and setting multiple energy storage nodes according to the aggregation result; Establishing multiple scheduling periods, and generating expected adjustable resource parameters of each energy storage node in the current scheduling period according to a preset response prediction model; Obtaining expected energy storage demand in the current scheduling period, and generating a primary control strategy according to all expected adjustable resource parameters and the expected energy storage demand; Obtaining a feedback data packet according to a preset feedback time node, and judging whether to generate a primary correction instruction according to the feedback data packet; When the multiple energy storage nodes are set, the following steps are included: Establish a sequence of energy storage nodes A, A=(a1,a2…a ... i …a n ), where a i Let i be the i-th energy storage node; n is the number of energy storage nodes; When the preset response prediction model is set, the following steps are included: Generating multiple device characteristic indexes according to all energy storage node parameters; Generating multiple node categories according to all device characteristic indexes; Establish a node category sequence B, B=(b1, b2, ..., bb2) i …b m ), where b i Let m be the node category of the i-th node; m is the number of node categories. Sequentially constructing response sub-models of each node category; A response sub-model sequence P is established, P=(p1, p2…p i …p m ), wherein p i is a response sub-model of the i-th node category; Constructing the response prediction model according to the response sub-model sequence P; When the response sub-models of each node category are sequentially constructed, the following steps are included: Based on the number of node categories, column B is set in turn b i Target node category; Generating multiple disturbance characteristic indexes of the target node category; Constructing an initial evaluation model of the target node category; Obtaining all associated energy storage nodes of the target node category; Establishing an associated energy storage node sequence A1 of the target node category; A1=(a 11 ,a 12 …a 1i …a 1n1 ), wherein a 1i is the i-th associated energy storage node; n1 is the number of associated energy storage nodes of the target node class; According to the associated energy storage node sequence A1, a is set in turn 1i Target associated energy storage node; Obtaining a historical operation packet of the target associated energy storage node, and generating a disturbance deviation value c of the target associated energy storage node according to the historical operation packet; Setting a disturbance deviation value threshold C1; If c>C1, generating a compensation sub-strategy of the target associated energy storage node; Sequentially judging whether each associated energy storage node generates a compensation sub-strategy; Generating a response sub-model of the target node category according to all compensation sub-strategies and the initial evaluation model; When the expected adjustable resource parameters of each energy storage node in the current scheduling period are generated, the following steps are included: Setting a first prediction time node of the current scheduling period; The response prediction model generates response capability parameters of each energy storage node at the first prediction time node; Generating response capability sub-tables of each node category according to all response capability parameters; Generating an expected adjustable resource table of the current scheduling period according to all response capability sub-tables, When the primary control strategy is generated, the following steps are included: Generating expected energy storage demand based on grid expected fluctuation parameters; Generating a primary allocation strategy according to the expected energy storage demand and the response capability sub-tables of each node category; Generating energy storage sub-demand of each node category according to the primary allocation strategy; Sequentially generating secondary allocation strategies of each node category; Generating the primary control strategy according to all secondary allocation strategies; When whether to generate a primary correction instruction is judged according to the feedback data packet, the following steps are included: Setting multiple feedback time nodes; Obtaining a feedback data packet of the current feedback time node; Generating a correction evaluation value d of the current feedback time node according to the feedback data packet; d = e3*Q3*[ λ i *k i ]+ e4*Q4*[ r i *j i ]; Wherein, e3 is a preset third weight coefficient; e4 is a preset fourth weight coefficient; Q3 is a preset third fixed coefficient; Q4 is a preset fourth fixed coefficient; is the number of energy storage demand indicators; λ i is the influence factor of the i-th energy storage demand indicator; k i is the fluctuation value of the i-th energy storage demand; r i is the influence factor of the i-th energy storage node in the energy storage node sequence A; j i is the response capability variation value of the i-th energy storage node; Setting a correction evaluation value threshold D1; If d>D1, the current feedback time node generates a primary correction instruction.

2. The virtual power plant based user side energy storage collaborative optimization method according to claim 1, wherein, When the disturbance deviation value c of the target associated energy storage node is generated, the following steps are included: c=[ η i *(v i -v' i )]; wherein θ1 is the number of perturbation characteristic indexes of the target node category; η i is the influence factor of the i-th perturbation characteristic index in the target node category; v i is the reference value of the i-th perturbation characteristic index generated based on the historical operation package; v' i is the standard reference value of the i-th perturbation characteristic index in the target node category.

3. The virtual power plant based user side energy storage collaborative optimization method according to claim 2, wherein, When the secondary allocation strategies of each node category are generated, the following steps are included: According to the node category number sequence B, b is set in turn i is the node category to be allocated; Establishing an associated energy storage node sequence A2 of the node category to be allocated; A2= (a 21 ,a 22 …a 2i …a 2n2 ), wherein a 2i is the i-th associated energy storage node of the node category to be allocated; n2 is the number of associated energy storage nodes of the node category to be allocated; Generating response evaluation values of each associated energy storage node in the associated energy storage node sequence A2; Generating a response scheduling order of the node category to be allocated according to all response evaluation values; Obtaining energy storage sub-demand of the node category to be allocated; According to the response scheduling sequence and the energy storage sub-demand, a two-level allocation strategy of the to-be-allocated node category is generated.

4. The virtual power plant based user side energy storage collaborative optimization method of claim 3, wherein, When generating the response evaluation value of each associated energy storage node in the associated energy storage node sequence A2, the following are included: According to the associated energy storage node sequence A2, a is set in turn 2i is the associated energy storage node to be evaluated; Generating the response evaluation value f of the to-be-evaluated associated energy storage node; f = e1*Q1*c' + e2*Q2*[ β i *s i ] Wherein, e1 is a preset first fixed coefficient; e2 is a preset second fixed coefficient; Q1 is a preset first weight coefficient; Q2 is a preset second fixed coefficient; c' is a disturbance deviation value of the associated energy storage node to be evaluated; θ2 is the number of response resource indicators; β i is an influence factor of the i th response resource indicator; s i is a reference value of the i th response resource indicator of the associated energy storage node to be evaluated.

5. The virtual power plant based user side energy storage co-optimization method of claim 4, wherein, The first correction instruction includes: Obtaining real-time state parameters of each energy storage node; The response prediction model updates the expected adjustable resource table according to all real-time state parameters; Obtaining the actual energy storage demand of the current feedback time node; According to the updated expected adjustable resource table and the actual energy storage demand, a two-level control strategy is generated.

Images