CN-121984045-A - Energy storage multi-time scale optimization regulation and control system and method for receiving-end urban power grid

CN121984045ACN 121984045 ACN121984045 ACN 121984045ACN-121984045-A

Abstract

The invention relates to a multi-time scale optimization regulation system and method for energy storage of a receiving-end urban power grid, which adopts a layered architecture and comprises a daytime coordination layer, a real-time execution layer and a multi-main-body coordination layer, wherein the daytime coordination layer is used for determining an energy storage daytime output target, the real-time execution layer is used for optimally distributing real-time available capacity of energy storage, and the multi-main-body coordination layer adopts a multi-agent deep reinforcement learning mode to control a plurality of energy storage devices to execute corresponding actions. Compared with the prior art, the invention can realize the energy storage optimization regulation and control of the receiving-end urban power grid oriented to safe-economic-low-carbon multi-target drive, and effectively optimize the income and reduce the emission in the carbon-electricity-green evidence coupling market.

Inventors

ZHANG HUI
FEI DANXIONG
CAI JIANFENG
HE XINQIN
CHEN ZEYUAN
ZHANG XINRAN
FAN WENWEN
ZHANG YANSHI
JIANG HAOMIN

Assignees

国网上海市电力公司

Dates

Publication Date: 20260505
Application Date: 20251208

Claims (10)

1. The energy storage multi-time scale optimization regulation and control system of the receiving-end urban power grid is characterized by adopting a layered architecture, comprising a daytime coordination layer, a real-time execution layer and a multi-main-body coordination layer, wherein the daytime coordination layer is used for determining an energy storage daytime output target; the real-time execution layer is used for optimizing and distributing the real-time available capacity of the energy storage; the multi-main-body cooperative layer adopts a multi-agent deep reinforcement learning mode to control the plurality of energy storage devices to execute corresponding actions.
2. The energy storage multi-time scale optimization regulation system of the receiving-end urban power grid according to claim 1, wherein the daytime coordination layer is provided with a multi-objective comprehensive optimization model, and the multi-objective optimization model comprises a comprehensive objective function for simultaneously minimizing the system operation cost, the operation risk and the carbon emission, and a multi-objective constraint condition, wherein the comprehensive objective function comprises an economical sub-objective, a safe sub-objective and a low-carbon sub-objective.
3. The energy storage multi-time scale optimization and regulation system of the receiving-end urban power grid according to claim 2, wherein the comprehensive objective function is specifically: Wherein, the In order to achieve the goal of the integration, Respectively an economy sub-target, a safety sub-target and a low-carbon sub-target, Weight coefficients corresponding to economic, safety and low-carbon sub-targets respectively and satisfy ; The multi-target constraint condition specifically comprises: Wherein, the In order to store the state of charge of the energy, The charging power and the discharging power are respectively provided, For the output of the renewable energy source, Is a load demand.
4. A receiver-side urban power grid energy storage multi-time scale optimization regulation system according to claim 3, wherein the economical sub-objective is specifically: Wherein T is the number of time steps in the optimization period, For the electricity price at the time t, For the grid purchase power at time t, For the price of the carbon, the carbon is available, Carbon emissions generated for time t; the security sub-object is specifically: Wherein, the For the real-time voltage and frequency of the system, For the rated voltage and frequency of the power supply, Is a weight factor, and is used for reflecting the security sensitivity; the low-carbon sub-target is specifically: Wherein, the For grid carbon emissions when the stored energy is not in use, And the energy storage is the corresponding carbon emission after participating in peak shaving.
5. The energy storage multi-time scale optimization and regulation system of the receiving-end urban power grid according to claim 1, wherein the real-time execution layer is provided with a real-time constraint optimization function and a real-time safety constraint condition, and adopts a QP quadratic programming projection mode to carry out actions Projecting to the closest action meeting real-time security constraints 。
6. The energy storage multi-time scale optimization and regulation system of the receiving-end urban power grid according to claim 5, wherein the real-time constraint optimization function is specifically: Wherein, the For the number of nodes of the urban power network at the receiving end, Is the measured value of the voltage and frequency of the node i, For the node reference voltage and the system nominal frequency, For adjusting the coefficients, for controlling the relative importance of the constraints in the optimization.
7. The energy storage multi-time scale optimization and regulation system of the receiving-end urban power grid according to claim 6, wherein the real-time safety constraint condition is specifically as follows: Wherein, the The upper and lower voltage limits are allowed for the nodes, For the range of the allowable deviation of the frequency, And the upper limit of the change rate of the energy storage charging and discharging power is the power climbing constraint.
8. The energy storage multi-time scale optimization and regulation system of the receiving-end urban power grid according to claim 1, wherein the multi-main-body cooperative layer specifically realizes distributed optimization of energy storage groups by coupling local rewards and global performance and taking global performance maximization as a target.
9. The energy storage multi-time scale optimization and regulation system of the receiving-end urban power grid according to claim 8, wherein the global performance is specifically: Wherein, the For a local reward for the energy storage unit i, Is SOC distribution variance, is used for measuring the utilization balance of energy storage resources, For balancing factors, for penalizing energy maldistribution, As charge/discharge energy of the energy storage unit i, As a term of the voltage deviation of the node, In order to achieve the current carbon number, Is a rewarding factor for balancing economic benefits, safety constraints and carbon costs.
10. The energy storage multi-time scale optimization regulation and control method for the receiving-end urban power grid is characterized by comprising the following steps of: S1, building a layered architecture comprising a daytime coordination layer, a real-time execution layer and a multi-main-body coordination layer in a simulation environment; s2, solving a sum and an optimization model by a daytime coordination layer through multi-objective planning or a Pareto optimization-based multi-objective evolutionary algorithm, and outputting power distribution and SOC reference tracks of each energy storage unit in a time period T; s3, taking the output of the daytime coordination layer as a scheduling reference, adopting a linear prediction control (MPC) or a fast constraint optimization algorithm by a real-time execution layer, and carrying out second-level correction on the stored energy output power by utilizing safety constraint; s4, global information is acquired, and a cooperation strategy among the plurality of energy storage devices is output through multi-agent reinforcement learning and used for controlling the working state of each energy storage device.

Description

Energy storage multi-time scale optimization regulation and control system and method for receiving-end urban power grid Technical Field The invention relates to the technical field of energy storage regulation and control, in particular to a multi-time scale optimization regulation and control system and method for energy storage of a receiving-end urban power grid. Background In the low-carbon transformation process of a novel power system, the receiving-end urban power grid faces the problems of high ratio of new energy to external electricity, fast load increase, large fluctuation, small adjustable load resources and the like, the extra-high voltage direct current blocking risk and the new energy consumption demand are superposed, the system operation scene is complex, the overvoltage and the oscillation risk of the receiving-end urban power grid are outstanding, the moment of inertia gradually becomes a scarce resource, the frequency stability margin is continuously reduced, and the voltage-frequency safety problem caused by the problem becomes a new challenge facing the low-carbon transformation of the receiving-end urban power grid. In recent years, new energy with strong randomness, volatility and intermittence is rapidly developed, the output of a traditional unit represented by coal electricity is limited, energy storage resources with low carbon attribute, large adjustment range, high adjustment speed and long duration are adopted to participate in the balance of an electric power system, and a great deal of research on the optimal configuration and optimal regulation of the energy storage resources is carried out at present. SCHICK C and other scholars utilize a market clearing model to study the reduction effect of energy storage on peak load and price. LARSEN M and other scholars obtain the optimal capacity and the optimal operation strategy of the energy storage investment of an independent system operator (INDEPENDENT SYSTEM operator, ISO) by solving an investment-operation double-layer optimization model. KAZEMI M and other scholars consider uncertainty of market price and market energy deployment and propose a combined bidding strategy of independent energy storage in energy and standby markets. GUSTAVO D V and other scholars introduce the charge state constraint of energy storage in the market clearing model, and research the influence of the bidding structure on the service provided by the energy storage system. Scholars such as the zodiac Yun Peng consider the coupling constraint when the energy storage provides energy and auxiliary service and put forward independent energy storage sequential and joint clear model considering opportunity cost. Li Yaowang and other scholars propose a low-carbon optimization method and a demand response mechanism of the power distribution network combined with a carbon emission flow theory, and fully mine emission reduction potential of the electricity utilization side. Hu Jingzhe reasonably divides the carbon emission responsibility between the generator set and the load node, guides the user to correct the self electricity consumption behavior according to the node carbon emission factor, and mobilizes the enthusiasm of the user for energy conservation and emission reduction. However, the interaction rule among the markets of the carbon-electricity-evidence is complex and changeable, in the real-time engineering, the daily/real-time market signals, the carbon price and the safety constraint of the power system are obviously different on the time scale, namely, the daily planning needs to consider economical efficiency and long-term SOC (State of Charge), and the frequency and voltage safety needs to be ensured in real time. Traditional MPC (Model Predictive Control ) or rule-based methods are not robust enough with high uncertainty, while pure RL (Reinforcement Learning ) lacks long-term constraint guarantees and interpretability. The energy storage of the receiving-end urban power grid cannot be comprehensively and accurately regulated and controlled, and the energy storage safety is difficult to maintain under real-time extreme disturbance, and meanwhile, the economic benefit and the carbon emission reduction are maximized. Disclosure of Invention The invention aims to overcome the defects of the prior art and provide a multi-time scale optimization regulation system and method for energy storage of a receiving-end urban power grid, which can realize the optimization regulation and control of the energy storage of the receiving-end urban power grid oriented to safe-economic-low-carbon multi-target driving, and effectively optimize benefits and reduce emission in a carbon-electricity-green evidence coupling market. The energy storage multi-time scale optimization regulation system of the receiving-end urban power grid adopts a layered architecture, and comprises a daytime coordination layer, a real-time execution layer and a multi-main-body coordinati