CN-122001018-A - Virtual power plant resource aggregation scheduling method and system based on time scale cooperation

CN122001018ACN 122001018 ACN122001018 ACN 122001018ACN-122001018-A

Abstract

The invention discloses a virtual power plant resource aggregation scheduling method and a system based on time scale cooperation, which relate to the technical field of resource aggregation scheduling, predict electric power parameters in a monitoring period, construct a day front layer optimization model related to the total cost of the predicted monitoring period to obtain a day front plan, start a rolling optimization model according to preset frequency in the monitoring period, revise the day front plan according to the rolling optimization model, and construct autonomous response based on chained autonomous response and elastic time window protocol.

Inventors

PEI TAO
Song Daijie
PAN TAO
FENG NAIXI
Leng Jixiang
FEI HONGZHENG
XIA DEZHI
ZHU YUANBING
WANG KANG

Assignees

山东华信电气股份有限公司

Dates

Publication Date: 20260508
Application Date: 20260127

Claims (9)

1. The virtual power plant resource aggregation scheduling method based on time scale cooperation is characterized by comprising the following steps of: S1, predicting electric power parameters in a monitoring period, constructing a day-ahead layer optimization model about the total cost of the predicted monitoring period, and performing power balance constraint, resource operation constraint, energy storage SOC dynamic constraint and power grid interaction constraint on the day-ahead layer optimization model to further obtain a day-ahead plan; s2, starting a rolling optimization model according to a preset frequency in a monitoring period, performing rolling optimization model condition constraint on the rolling optimization model, and correcting a day-ahead plan according to the rolling optimization model; S3, constructing an autonomous response based on the chained autonomous response and the elastic time window protocol, establishing a cost minimization model, and performing power plant resource aggregation scheduling based on the cost minimization model.
2. The virtual power plant resource aggregation scheduling method based on time scale cooperation according to claim 1, wherein authorized prediction is used for monitoring periodic power parameters, a day-ahead layer optimization model is established, power acquisition total cost is solved through an interior point method, model constraint covers power balance, resource operation, energy storage dynamics and power grid interaction, then a rolling optimization model is established based on a current power use state, optimization coefficients are targeted, constraint comprises two types of boundary inheritance constraint, day-ahead layer decision variables and resource real-time dynamic potential intervals are associated, finally autonomous response is established based on chain autonomous response and an elastic time window protocol, the chain autonomous response is used for uniformly distributing power according to an energy storage state through a power distribution model, and the elastic time window protocol comprises energy total amount and cost minimization related constraint.
3. The method for aggregated scheduling of virtual power plant resources based on time scale coordination according to claim 2, wherein in step S1, after authorization, the power parameter in a monitoring period is predicted, the monitoring period is set to be T on , namely, the monitoring period comprises T on time periods, and a day-ahead layer optimization model is established: ; solving for alpha by the interior point method, where alpha represents the total cost of power harvesting in one monitoring cycle, Representing the total cost of power taken from the main grid during period t, The unit power cost of drawing power from the main grid for period t, For the amount of power taken from the main grid for period t, The total amount of benefit obtained from providing auxiliary services to the grid for period t, The benefit per unit of power obtained by providing auxiliary services to the grid for period t, The power obtained by the auxiliary service is provided to the grid for period t, Regularization term for period t, which is used to penalize the total output of the virtual power plant And dynamic baseline load Is used for the deviation of (a), Preset trade-off parameters for regularization term.
4. The method for virtual power plant resource aggregate scheduling based on time scale collaboration of claim 3, wherein the day-ahead optimization model is constrained by a day-ahead optimization model, and the day-ahead optimization model constraint comprises a power balance constraint, a resource operation constraint, an energy storage SOC dynamic constraint and a power grid interaction constraint.
5. The method for aggregated scheduling of virtual power plant resources based on time scale collaboration of claim 4, wherein in step S2, a rolling optimization model is built based on a current power utilization state, and the rolling optimization model and beta are built based on the state, 、 And In relation to the use of a liquid crystal display device, Where beta represents the optimization coefficient obtained based on the rolling optimization model, In order to optimize the first object of the present invention, As a current cost of electricity in the market, Power costs that are currently optimizable; Representing an optimization target II, wherein gamma is the weight of the optimization target II, Is a preset SOC reference value.
6. The method for aggregated scheduling of virtual power plant resources based on time scale collaboration of claim 5, wherein the rolling optimization model condition constraint comprises a first boundary inheritance constraint and a second boundary inheritance constraint, and the expression of the first boundary inheritance constraint is: ; Wherein the method comprises the steps of As a decision variable for the inner layer of the day, The lower limit of the decision variable given for the daily inner layer daily schedule, The upper limit of the decision variable is given for the daily schedule of the daily inner layer; The expression of the second boundary inheritance constraint is: ; And representing the latest dynamic potential interval of the real-time calculation of the resource edge layer.
7. The method for aggregated scheduling of virtual power plant resources based on time scale collaboration of claim 6, wherein in step S3, an autonomous response is constructed based on a chained autonomous response and an elastic time window protocol, wherein the chained autonomous response comprises: Performing power distribution based on SOC balance, and establishing a power distribution model: ; Wherein the method comprises the steps of Indicating the power to which the kth energy storage unit is allocated, The discharge is indicated to be such that, Indicating charging; representing the total power instruction issued by the upper layer scheduling system, Indicating that the clusters need to be totally discharged, Indicating the total charge required by the cluster; Representing the current charge state of the kth energy storage unit, and when the kth energy storage unit is full of electricity 1, When the k energy storage unit is empty Is 0; Is a preset equalization factor; as an index weight term, the weight of the index is calculated, The deviation distance between the current SOC and the ideal intermediate value is measured; is a global normalization factor.
8. The method for aggregating and scheduling virtual power plant resources based on time scale collaboration of claim 6, wherein the flexible time window protocol comprises: constructing an energy total constraint expression: ; Wherein the method comprises the steps of Representing the actual response power of resource j at time t, t a representing the lower limit of the total regulated energy that resource j needs to complete within time window [ a, b ], t b representing the upper limit of the total regulated energy that resource j needs to complete within time window [ a, b ], Indicating the total regulated energy that resource j needs to complete over time window [ a, b ]; construction cost minimization constraint expression: ; Wherein the method comprises the steps of Representing the time-varying response cost function of resource j, In inverse proportion to the power order density, Representing the user response power curve obtained after authorization.
9. The virtual power plant resource aggregation scheduling system based on time scale coordination is applied to the virtual power plant resource aggregation scheduling method based on time scale coordination as claimed in any one of claims 1 to 8, and is characterized by comprising a day-ahead plan construction module, a real-time model plan optimization module and an autonomous response derivation module; The day-ahead plan construction module is used for predicting power parameters in one monitoring period, constructing a day-ahead layer optimization model about the total cost of the predicted monitoring period, and carrying out power balance constraint, resource operation constraint, energy storage SOC dynamic constraint and power grid interaction constraint on the day-ahead layer optimization model to further obtain a day-ahead plan; The real-time model plan optimization module is used for starting a rolling optimization model according to a preset frequency in a monitoring period, carrying out rolling optimization model condition constraint on the rolling optimization model, and correcting a day-ahead plan according to the rolling optimization model; The autonomous response export module is used for constructing autonomous response based on chained autonomous response and elastic time window protocol, establishing a cost minimization model and carrying out power plant resource aggregation scheduling based on the cost minimization model.

Description

Virtual power plant resource aggregation scheduling method and system based on time scale cooperation Technical Field The invention relates to the technical field of resource aggregation scheduling, in particular to a virtual power plant resource aggregation scheduling method and system based on time scale cooperation. Background The virtual power plant is used as an important carrier for integrating distributed resources and participating in electric market trading and power grid dispatching, plays a key role in promoting energy transformation and improving power grid flexibility, and the scientificity of resource aggregation dispatching directly determines operation benefit and power grid safety level. The core technical problem faced in the current virtual power plant resource scheduling field is that the multi-time scale scheduling links are not enough in cooperation, and the organic connection of global planning and real-time response is difficult to realize, so that a series of operation defects are caused. Due to the defect of a multi-time scale scheduling coordination mechanism, the global planning of a day front layer and the real-time adjustment of a day inner layer in the existing scheduling scheme are mutually split in real time response of an autonomous layer, and effective linkage cannot be formed. The method is characterized in that a scheme is formulated on the basis of static data in a day-ahead planning link, dynamic changes of real-time running states of resources and power grid requirements are not fully considered, so that deviation exists between the plan and actual execution conditions, and the scheme is difficult to land due to the fact that accurate connection constraint with the day-ahead plan is lacking in a day-ahead adjusting link, and the decision is easy to exceed an actual running potential interval of the resources. Meanwhile, the insufficient cooperation makes the scheduling strategy difficult to consider the multi-dimensional requirement, or simply chases the economic benefit to neglect the impact of power fluctuation on the stability of the power grid, so that the project feasibility of the scheme is reduced, or the balanced distribution and elastic regulation mechanism is lacking in the resource response stage, so that the load distribution imbalance and the response cost of equipment such as energy storage are high. The problem of insufficient coordination of multi-time scale scheduling severely restricts the improvement of the resource aggregation efficiency of the virtual power plant, cannot fully exert the value of distributed resources, and is difficult to meet the comprehensive requirements of a power grid on scheduling safety, economy and flexibility. Disclosure of Invention The invention aims to provide a virtual power plant resource aggregation scheduling method and system based on time scale cooperation, so as to solve the problems in the background technology. In order to solve the technical problems, the invention provides a virtual power plant resource aggregation scheduling method based on time scale cooperation, which comprises the following steps: S1, predicting electric power parameters in a monitoring period, constructing a day-ahead layer optimization model about the total cost of the predicted monitoring period, and performing power balance constraint, resource operation constraint, energy storage SOC dynamic constraint and power grid interaction constraint on the day-ahead layer optimization model to further obtain a day-ahead plan; s2, starting a rolling optimization model according to a preset frequency in a monitoring period, performing rolling optimization model condition constraint on the rolling optimization model, and correcting a day-ahead plan according to the rolling optimization model; S3, constructing an autonomous response based on the chained autonomous response and the elastic time window protocol, establishing a cost minimization model, and performing power plant resource aggregation scheduling based on the cost minimization model. And (3) carrying out authorized prediction and monitoring of periodic power parameters, establishing a day-ahead layer optimization model, solving power by an interior point method to obtain total cost, wherein model constraint covers power balance, resource operation, energy storage dynamics and power grid interaction, then constructing a rolling optimization model based on a current power use state, taking an optimization coefficient as a target, wherein the constraint comprises two types of boundary inheritance constraint, associating a day-ahead layer decision variable with a resource real-time dynamic potential interval, and finally constructing an autonomous response based on a chained autonomous response and an elastic time window protocol, wherein the chained autonomous response is used for uniformly distributing power according to the energy storage state through a power distribution model, and the