CN-121984234-A - New energy system, new energy system energy scheduling method and local controller

CN121984234ACN 121984234 ACN121984234 ACN 121984234ACN-121984234-A

Abstract

The application discloses a new energy system, a new energy system energy scheduling method and a local controller, which relate to the technical field of new energy, wherein the new energy system accurately collects electricity utilization data of electric equipment by utilizing an intelligent socket and is used as a unified load sensing inlet, the local controller flexibly selects a target node from the local controller, the available edge gateway and network side equipment according to whether the new energy system has an available edge gateway and the communication state between the local and network side equipment, so that total electricity utilization load prediction data is obtained from the target node, a scheduling instruction is generated by combining the running data of a direct current source, and flexible switching among three prediction modes of the network side, the edge and the local according to the current communication state and resource conditions is realized, and model prediction precision and system robustness are considered.

Inventors

YANG FAN
FAN RENKAI
HE PENG

Assignees

南京光献科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260123

Claims (12)

1. The new energy system is characterized by comprising a local controller, an inverter, a direct current source and at least one intelligent socket; The direct current side of the inverter is connected with the direct current source, and the direct current source comprises at least one of a photovoltaic panel and an energy storage battery; The alternating-current side grid-connected port of the inverter is connected with a power grid; the alternating current side load port of the inverter is connected with at least one electric device; the intelligent socket collects electricity utilization data of the electric equipment electrically connected with the intelligent socket; the local controller obtains the running data of the direct current source, obtains total power load prediction data from a target node according to whether an available edge gateway exists in the new energy system and the communication state between the local and network side equipment, and generates a scheduling instruction according to the total power load prediction data and the running data of the direct current source; the target node is the local controller, the available edge gateway or the network side equipment; The total power consumption load prediction data are obtained by summarizing the load prediction data of each electric equipment, the load prediction data of the electric equipment are obtained by predicting the sub-term power time sequence of the electric equipment by a load prediction model corresponding to the electric equipment, and the sub-term power time sequence of each electric equipment is obtained by decomposing the total power time sequence corresponding to the power consumption data acquired by all intelligent sockets by a load decomposition model.
2. The new energy system of claim 1, wherein the local controller obtains the total power usage load prediction data from the available edge gateway if the available edge gateway is present; the available edge gateway is in communication connection with all the intelligent sockets to acquire electricity consumption data acquired by all the intelligent sockets, operates a light load decomposition model and light load prediction models corresponding to the electric equipment to acquire total electricity consumption load prediction data, and sends the total electricity consumption load prediction data to the local controller.
3. The new energy system of claim 2, wherein the available edge gateway is integrated in the inverter.
4. The new energy system according to claim 1, wherein the local controller obtains electricity consumption data collected by all the smart sockets when the available edge gateway is not present and the communication state between the local and the network side devices does not meet a preset condition, and operates a lightweight load decomposition model and a lightweight load prediction model corresponding to each electric device to obtain the total electricity consumption load prediction data.
5. The new energy system according to any one of claims 2 to 4, wherein the target node sends historical sample data to the network side device when a communication state between the target node and the network side device meets a preset condition, the historical sample data including the total power consumption load prediction data, power consumption load prediction data corresponding to each electric consumer, and actual total power consumption load data corresponding to the total power consumption load prediction data; the network side equipment periodically utilizes the historical sample data to optimize a network side load decomposition model and network side load prediction models corresponding to the electric equipment, performs light weight processing on the optimized network side load decomposition model and the network side load prediction model to obtain an optimized light weight load decomposition model and a light weight load prediction model, and sends the optimized light weight load decomposition model and the light weight load prediction model to the target node.
6. The new energy system according to claim 1, wherein the local controller obtains all the electricity consumption data collected by the smart sockets and sends the total power time sequence corresponding to all the electricity consumption data collected by the smart sockets to the network side device when the available edge gateway is not available and the communication state between the local and the network side device meets a preset condition; And the network side equipment operates a network side load decomposition model and a network side load decomposition model corresponding to each electric equipment to obtain the total power consumption load prediction data corresponding to the total power time sequence, and sends the total power consumption load prediction data to the local controller.
7. The new energy system according to claim 1, wherein the local controller generates a scheduling instruction according to the total power time sequence and the operation data of the direct current source in a case that no edge gateway is available and a communication state between a local and the network side device does not satisfy a preset condition.
8. The utility model provides a new energy system energy scheduling method which is characterized in that is applied to the local controller in the new energy system, new energy system still includes dc-to-ac converter, direct current source and at least one smart jack, the direct current side of dc-to-ac converter is connected the direct current source, the direct current source includes at least one of photovoltaic board and energy storage battery, the electric wire netting is connected to the ac-to-ac side grid-connected mouth of dc-to-ac converter, at least one consumer is connected to the ac-to-ac side load mouth of dc-to-ac converter, the smart jack gathers with the power consumption data of the consumer of smart jack electricity connection, new energy system energy scheduling method includes: Acquiring operation data of the direct current source; Acquiring total power consumption load prediction data from a target node according to whether an available edge gateway exists in the new energy system and the communication state between local and network side equipment, wherein the target node is the local controller, the available edge gateway or the network side equipment, the total power consumption load prediction data are summarized by load prediction data of each electric equipment, the load prediction data of the electric equipment are obtained by predicting a sub-term power time sequence of the electric equipment by a load prediction model corresponding to the electric equipment, and the sub-term power time sequence of each electric equipment is obtained by decomposing a total power time sequence corresponding to power consumption data acquired by all intelligent sockets by a load decomposition model; and generating a scheduling instruction according to the total power load prediction data and the running data of the direct current source.
9. The method for energy scheduling of a new energy system according to claim 8, wherein the obtaining the total power load prediction data from the target node according to whether there is an edge gateway available in the new energy system and a communication state between the local and network side devices includes: And under the condition that the available edge gateway exists, receiving the total power consumption load prediction data sent by the available edge gateway, and after the total power consumption load prediction data are obtained by the available edge gateway, operating a light-weight load decomposition model and light-weight load prediction models corresponding to the electric equipment.
10. The method for energy scheduling of a new energy system according to claim 8, wherein the obtaining the total power load prediction data from the target node according to whether there is an edge gateway available in the new energy system and a communication state between the local and network side devices includes: and under the condition that the communication states between the available edge gateway and the local network side equipment do not meet the preset conditions, acquiring power consumption data acquired by all the intelligent sockets, and operating a light load decomposition model and a light load prediction model corresponding to each electric equipment to obtain the total power consumption load prediction data.
11. The method for energy scheduling of a new energy system according to claim 8, wherein the obtaining the total power load prediction data from the target node according to whether there is an edge gateway available in the new energy system and a communication state between the local and network side devices includes: Under the condition that the available edge gateway is not available and the communication state between the local and the network side equipment meets the preset condition, acquiring all the power consumption data acquired by the intelligent sockets, and transmitting the total power time sequence corresponding to all the power consumption data acquired by the intelligent sockets to the network side equipment; and receiving the total power consumption load prediction data sent by the network side equipment, wherein the total power consumption load prediction data is obtained by the network side equipment operation network side load decomposition model and the network side load decomposition model corresponding to each electric equipment.
12. A local controller configured to perform the new energy system energy scheduling method of any one of claims 8 to 11.

Description

New energy system, new energy system energy scheduling method and local controller Technical Field The application relates to the technical field of new energy, in particular to a new energy system, an energy scheduling method of the new energy system and a local controller. Background With the development of photovoltaic technology, more and more families install new energy systems comprising photovoltaic panels and energy storage batteries, and normal operation of the new energy systems is guaranteed by coordinating power among photovoltaic power generation, charging and discharging of the energy storage batteries and household loads. With the development of artificial intelligence technology, the utilization of an AI model to predict the power load of a new energy system so as to improve the energy scheduling accuracy of the new energy system gradually becomes a research hotspot in the field. Because the cloud has enough computing resources, the AI model is generally deployed in the cloud, but is limited by factors such as network bandwidth, communication delay and the like, and the prediction result of the cloud cannot always meet the local real-time scheduling requirement. Disclosure of Invention In view of the above problems, the present application provides a new energy system, a new energy system energy scheduling method and a local controller, which flexibly switch among three prediction modes of a network side, an edge and a local according to a current communication state and a resource condition, and both model prediction accuracy and system robustness are considered. The specific scheme is as follows: the first aspect of the application provides a new energy system, which comprises a local controller, an inverter, a direct current source and at least one intelligent socket; The direct current side of the inverter is connected with the direct current source, and the direct current source comprises at least one of a photovoltaic panel and an energy storage battery; The alternating-current side grid-connected port of the inverter is connected with a power grid; the alternating current side load port of the inverter is connected with at least one electric device; the intelligent socket collects electricity utilization data of the electric equipment electrically connected with the intelligent socket; the local controller obtains the running data of the direct current source, obtains total power load prediction data from a target node according to whether an available edge gateway exists in the new energy system and the communication state between the local and network side equipment, and generates a scheduling instruction according to the total power load prediction data and the running data of the direct current source; the target node is the local controller, the available edge gateway or the network side equipment; The total power consumption load prediction data are obtained by summarizing the load prediction data of each electric equipment, the load prediction data of the electric equipment are obtained by predicting the sub-term power time sequence of the electric equipment by a load prediction model corresponding to the electric equipment, and the sub-term power time sequence of each electric equipment is obtained by decomposing the total power time sequence corresponding to the power consumption data acquired by all intelligent sockets by a load decomposition model. In one possible implementation, the local controller obtains the total power load prediction data from the available edge gateway if the available edge gateway is present; the available edge gateway is in communication connection with all the intelligent sockets to acquire electricity consumption data acquired by all the intelligent sockets, operates a light load decomposition model and light load prediction models corresponding to the electric equipment to acquire total electricity consumption load prediction data, and sends the total electricity consumption load prediction data to the local controller. In one possible implementation, the available edge gateway is integrated in the inverter. In one possible implementation, the local controller obtains power consumption data collected by all the intelligent sockets under the condition that the available edge gateway is not available and the communication state between the local controller and the network side equipment does not meet the preset condition, and operates a light load decomposition model and a light load prediction model corresponding to each electric equipment to obtain the total power consumption load prediction data. In one possible implementation, the target node sends historical sample data to the network side device when the communication state between the target node and the network side device meets a preset condition, wherein the historical sample data comprises the total power consumption load prediction data, the power consumption load prediction data corre