Single user short-term load prediction method under novel power system background

Abstract

The invention relates to a short-term load prediction method for a single user under the background of a novel power system, and the method comprises the steps: reading user load data in a power utilization information collection system, and decomposing the user load data into a conventional load and a distributed power supply load; meteorological data are collected, photovoltaic power generation is predicted based on the meteorological data, and distributed power supply load prediction is obtained; carrying out traditional load prediction by considering demand side response; and adding the distributed power supply load prediction and the traditional load prediction to obtain single user short-term load prediction. In order to realize accurate prediction of the total load of a single user, the total load is decomposed into a conventional load and a distributed power supply which operates in a'self-generation and self-use and residual internet surfing 'mode; for traditional conventional load prediction, influence of market electricity price and policies needs to be considered on the basis of influence factors such as electricity consumption behaviors and meteorological conditions, and strong weather association attributes are focused on distributed power supply prediction.

Description

Technical Field

[0001] The invention belongs to the field of distribution network load forecasting, and in particular to a single user short-term load forecasting method under the background of a new power system. Background Art

[0002] For the single user load forecasting scenario, the user's main electricity consumption habits are highly personalized and differentiated, and the user's load pattern shows strong variability and uncertainty, making forecasting modeling more difficult. In the context of the new power system, on the one hand, the access of flexible and controllable resources on the demand side such as distributed power sources and energy storage will gradually change the user load from rigid to flexible. On the other hand, different users respond to demand in different ways under the influence of various incentive policies in the power market, which will greatly change their original electricity consumption patterns and load characteristics. Therefore, load forecasting for single users in the context of the new power system has become more difficult. Summary of the invention

[0003] The purpose of the present invention is to overcome the shortcomings of the prior art and propose a single-user short-term load forecasting method under the background of a new power system, by decomposing the total load into conventional load, self-generation and self-use, and distributed power sources operating in surplus grid-connected mode; for the prediction of traditional conventional load, it is necessary to focus on the impact of market electricity prices and policies on the basis of influencing factors such as electricity consumption behavior and meteorological conditions, and the prediction of distributed power sources focuses on strong weather correlation attributes.

[0004] The present invention solves the technical problem by adopting the following technical solutions:

[0005] A method for short-term load forecasting of a single user under the background of a new power system includes the following steps:

[0006] Step 1: Read user load data in the power consumption information collection system, and decompose the user load data into traditional load and distributed power load;

[0007] Step 2: Collect meteorological data, predict photovoltaic power generation based on the meteorological data, and obtain distributed power load prediction;

[0008] Step 3: Consider the demand-side response to perform traditional load forecasting;

[0009] Step 4: Add the distributed power load forecast in step 2 and the traditional load forecast in step 3 to obtain the short-term load forecast for a single user.

[0010] Moreover, the specific implementation method of step 1 is: reading the user load data of the gateway total meter and the photovoltaic power generation meter in the electricity consumption collection information system, and decomposing the total load of a single household into traditional load and distributed power supply load.

[0011] Furthermore, the step 2 comprises the following steps:

[0012] Step 2.1, using the historical numerical weather forecast NWP as the meteorological numerical forecast source and the public meteorological service forecast as the weather state forecast source, normalize the historical photovoltaic power generation curve, and use the LSTM network prediction model to establish the mapping relationship between meteorological factors and photovoltaic power;

[0013] Step 2.2: Obtain future short-term meteorological information as model input to obtain distributed generation load forecast.

[0014] Furthermore, the step 3 comprises the following steps:

[0015] Step 3.1, using CEEMD algorithm to decompose the traditional load into basic load and demand response load;

[0016] Step 3.2: Use the time series model to predict base load;

[0017] Step 3.3, use the maximum information coefficient method to analyze the correlation between electricity price and demand response load, and use the LSTM network to predict the demand response load;

[0018] Step 3.4: Add the base load obtained in step 3.2 and the demand response load in step 3.3 to obtain a traditional load forecast.

[0019] The advantages and positive effects of the present invention are:

[0020] The present invention reads user load data in the power consumption information collection system, and decomposes the user load data into traditional load and distributed power load; collects meteorological data, predicts photovoltaic power generation based on meteorological data, and obtains distributed power load prediction; considers demand-side response to perform traditional load prediction; adds distributed power load prediction and traditional load prediction to obtain a single user short-term load prediction. In order to achieve accurate prediction of the total load of a single user, the present invention decomposes the total load into conventional load and distributed power sources operating in the "self-generation and self-use, surplus power access to the Internet" mode; for the prediction of traditional conventional load, it is necessary to focus on the impact of market electricity prices and policies on the basis of influencing factors such as electricity consumption behavior and meteorological conditions, and the prediction of distributed power sources focuses on strong weather-related attributes. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a schematic diagram of the electricity consumption information collection system of the present invention. DETAILED DESCRIPTION

[0022] The present invention is further described in detail below with reference to the accompanying drawings.

[0023] A method for short-term load forecasting of a single user under the background of a new power system includes the following steps:

[0024] Step 1: Read user load data in the power consumption information collection system, and decompose the user load data into traditional load and distributed power load.

[0025] like Figure 1 As shown, Table 1 in the figure is the gateway total table, and Table 2 in the figure is the gateway total table. The gateway total table and the photovoltaic power generation meter user load data are read in the electricity consumption collection information system, and the total load of a single household is decomposed into traditional load and distributed power supply load.

[0026] Step 2: Collect meteorological data, predict photovoltaic power generation based on the meteorological data, and obtain a distributed power load forecast.

[0027] Step 2.1, using the historical numerical weather forecast NWP as the meteorological numerical forecast source and the public meteorological service forecast as the weather state forecast source, normalize the historical photovoltaic power generation curve, and use the LSTM network prediction model to establish the mapping relationship between meteorological factors and photovoltaic power;

[0028] Step 2.2: Obtain future short-term meteorological information as model input to obtain distributed generation load forecast.

[0029] Step 3: Perform traditional load forecasting considering demand-side response.

[0030] Step 3.1, using CEEMD algorithm to decompose the traditional load into basic load and demand response load;

[0031] Step 3.2: Use the time series model to predict base load;

[0032] Step 3.3, use the maximum information coefficient method to analyze the correlation between electricity price and demand response load, and use the LSTM network to predict the demand response load;

[0033] Step 3.4: Add the base load obtained in step 3.2 and the demand response load in step 3.3 to obtain a traditional load forecast.

[0034] Step 4: Add the distributed power load forecast in step 2 and the traditional load forecast in step 3 to obtain the short-term load forecast for a single user.

[0035] It should be emphasized that the embodiments described in the present invention are illustrative rather than restrictive. Therefore, the present invention includes but is not limited to the embodiments described in the specific implementation manner. Any other implementation manners derived by those skilled in the art based on the technical solution of the present invention also fall within the scope of protection of the present invention.

Claims

1. A method for short-term load forecasting of a single user under the background of a new power system, characterized by: The following steps are involved: Step 1: Read user load data in the power consumption information collection system, and decompose the user load data into traditional load and distributed power load; Step 2: Collect meteorological data, predict photovoltaic power generation based on the meteorological data, and obtain distributed power load prediction; Step 3: Consider the demand-side response to perform traditional load forecasting; Step 4: Add the distributed power load forecast in step 2 and the traditional load forecast in step 3 to obtain the short-term load forecast for a single user.

2. A method for short-term load forecasting of a single user under the background of a new power system according to claim 1, characterized in that: The specific implementation method of step 1 is: reading the user load data of the gateway total meter and the photovoltaic power generation meter in the electricity consumption collection information system, and decomposing the total load of a single household into traditional load and distributed power supply load.

3. A method for short-term load forecasting of a single user under the background of a new power system according to claim 1, characterized in that: The step 2 comprises the following steps: Step 2.1, using the historical numerical weather forecast NWP as the meteorological numerical forecast source and the public meteorological service forecast as the weather state forecast source, normalize the historical photovoltaic power generation curve, and use the LSTM network prediction model to establish the mapping relationship between meteorological factors and photovoltaic power; Step 2.2: Obtain future short-term meteorological information as model input to obtain distributed generation load forecast.

4. A method for short-term load forecasting of a single user under the background of a new power system according to claim 1, characterized in that: The step 3 comprises the following steps: Step 3.1, using CEEMD algorithm to decompose the traditional load into basic load and demand response load; Step 3.2: Use the time series model to predict base load; Step 3.3, use the maximum information coefficient method to analyze the correlation between electricity price and demand response load, and use the LSTM network to predict the demand response load; Step 3.4: Add the base load obtained in step 3.2 and the demand response load in step 3.3 to obtain a traditional load forecast.

Images