A method for constructing a climate-energy market supply and demand model based on artificial intelligence

Abstract

The present invention proposes a method for constructing a climate-energy market supply and demand model based on artificial intelligence. The method obtains climate data with high correlation with power generation in the area to be analyzed, inputs the trained climate-energy supply model, and obtains a supply power generation forecast value. The method also obtains climate data with high correlation with residential power consumption in the area to be analyzed, inputs the trained climate-energy demand model, and obtains a demand power consumption forecast value. Optimization scheduling is performed based on the supply power consumption forecast value and the demand power consumption forecast value, thereby effectively reducing the impact of the instability of new energy power generation on the power grid and improving the comfort of residential power consumption and the stability of industrial power consumption.

Description

Technical Field

[0001] The present invention relates to a method for constructing a climate-energy market supply and demand model, and in particular to a method for constructing a climate-energy market supply and demand model based on artificial intelligence. Background Art

[0002] Since the Industrial Revolution, the global climate has been warming at an unprecedented rate, which has had a significant impact on the global ecosystem and human society. As the impact gradually increases, the study of global climate change has received more and more attention from research institutions and governments.

[0003] In recent years, my country has vigorously developed new energy, but the output of new energy is greatly affected by the climate, especially in summer and winter. In summer, it is mostly controlled by warm high pressure and prevailing downward airflow, which is very conducive to ground warming and air drying. Therefore, high temperature, drought, little rain, and light wind are prone to occur. High temperature and drought are not conducive to hydropower generation, and light wind is not conducive to wind power generation. The output of photovoltaic power generation is limited by excessively high daytime temperatures, and the output is almost zero at night. The hot summer weather also drives the power load of refrigeration. As the power supply end of new energy, wind power, photovoltaic power generation, and hydropower have outputs far lower than the power load and cannot meet the demand of the power load. In winter, affected by cold air, cold waves, low temperatures and freezing weather are prone to occur, which can cause wind turbines to freeze or shut down for protection, limiting wind power output. Winter is also the dry season, so hydropower output is also limited. Snowfall affects photovoltaic power output, and there is almost no output at night. At night, due to the low temperature, the heating load is driven, which is the electricity load period. As new energy power supply ends, wind power, photovoltaic power generation, and hydropower have an output far lower than the electricity load and are far from meeting electricity demand.

[0004] With the intensification of global climate change and the continuous growth of energy demand, traditional energy market supply and demand models have been unable to accurately predict and respond to complex and changing market environments. Existing technologies lack effective means to combine climate factors with energy market dynamics, resulting in irrational allocation of energy resources. Therefore, forecasting the power generation of different types of new energy sources and regional electricity loads, and studying the impact of climate factors on the supply and demand of multi-regional and multi-energy markets, is conducive to the refined management and control of risks in the energy market, and can effectively reduce the impact of the instability of new energy power generation on the power grid, and improve the comfort of residential electricity use and the stability of industrial electricity use. Summary of the invention

[0005] The purpose of the present invention is to solve the problem of unreasonable energy resource allocation in the above-mentioned prior art, and to provide a method for constructing a climate-energy market supply and demand model based on artificial intelligence, so as to achieve the effect of improving the comfort of residential electricity use and the stability of industrial electricity use.

[0006] The present invention is achieved through the following technical solutions:

[0007] The present invention provides a method for constructing a climate-energy market supply and demand model based on artificial intelligence:

[0008] Step S1: Collect historical climate data and actual power generation data and actual residential power consumption data for the same period as a training set;

[0009] Step S2: Based on the Spearman rank correlation coefficient, select climate data with high correlation with power generation as input of the climate-energy supply model, and train the climate-energy supply model;

[0010] Step S3: Based on the Spearman rank correlation coefficient, select climate data with high correlation with residential electricity consumption as input of the climate-energy demand model, and train the climate-energy demand model;

[0011] Step S4: Acquire climate data with high correlation with power generation in the area to be analyzed, input the trained climate-energy supply model, obtain the predicted value of power generation supplied in the area, acquire climate data with high correlation with residential power consumption in the area to be analyzed, input the trained climate-energy demand model, and obtain the predicted value of power consumption demand in the area;

[0012] Step S5: Optimize scheduling based on the supply power generation forecast value and the demand power consumption forecast value.

[0013] Furthermore, the climate-energy supply model includes: a CNN spatiotemporal feature extraction module and a fully connected neural network module, wherein the climate data highly correlated with power generation is input into the CNN spatiotemporal feature extraction module for feature extraction, the features are input into the fully connected neural network module for prediction, and a supply power generation prediction value is output;

[0014] The CNN spatiotemporal feature extraction module includes a sequentially connected input layer, a convolutional layer, a temporal feature extraction module, and a spatial feature extraction module. The output of the convolutional layer is denoted as Fc, the temporal feature extraction module is denoted as Ft, and the output of the spatial feature extraction module is denoted as Fs.

[0015] Among them, Ft is obtained by the following method:

[0016]

[0017] TAB(Fc)=Fc⊙Upsample(MLP(GAP(Fc))),

[0018] Among them, MLP() represents multi-layer perceptron, GAP() represents global average pooling, and Upsample() represents upsampling;

[0019] Fs is obtained as follows:

[0020]

[0021]

[0022] Among them, conv() represents the convolution layer, dwconv1() represents the 3X3 depth separation convolution layer, dwconv2() represents the 5X5 depth separation convolution layer, and pw() represents the point-by-point convolution layer.

[0023] Furthermore, the climate-energy demand model is a CNN-LightGBM model.

[0024] Furthermore, climate data includes climate element data and climate risk index data.

[0025] Furthermore, climate element data include: temperature (2-meter air temperature, dew point temperature), precipitation, humidity, wind speed (10 meters and 100 meters), wind direction, sea level pressure, radiation intensity, runoff, evaporation, snowfall, sunshine duration, etc.

[0026] Furthermore, the climate risk index data includes: flooding, drought, typhoon, high temperature, low temperature and freezing, etc.

[0027] The present invention proposes a method for constructing a climate-energy market supply and demand model based on artificial intelligence, which collects historical climate data and actual power generation data and actual electricity consumption data of residents in the same period as training sets; based on the Spearman rank correlation coefficient, selects climate data with high correlation with power generation as input of a climate-energy supply model, and trains the climate-energy supply model; based on the Spearman rank correlation coefficient, selects climate data with high correlation with residents' electricity consumption as input of a climate-energy demand model, and trains the climate-energy demand model; obtains climate data with high correlation with power generation in a to-be-analyzed area, inputs the trained climate-energy supply model, obtains a predicted value of supply power generation in the area, obtains climate data with high correlation with residents' electricity consumption in the to-be-analyzed area, inputs the trained climate-energy demand model, and obtains a predicted value of demand power consumption in the area; optimizes scheduling based on the predicted value of supply power generation and the predicted value of demand power consumption, thereby effectively reducing the impact of the instability of new energy power generation on the power grid and improving the comfort of residents' electricity consumption and the stability of industrial electricity consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the specific implementation of the present invention or the technical solution in the prior art, the drawings required for describing the specific implementation or the prior art will be briefly introduced below. Obviously, the drawings described below are only one implementation of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0029] Figure 1 A flowchart of a method for building an AI-based climate-energy market supply and demand model;

[0030] Figure 2 This is a schematic diagram of the structure of the climate-energy supply model;

[0031] Figure 3 It is a structural diagram of the time feature extraction module;

[0032] Figure 4 It is a structural diagram of the spatial feature extraction module;

[0033] Figure 5 It is a structural diagram of the TAB module;

[0034] Figure 6 Figure 2 is a schematic diagram of the structure of the SAB module. DETAILED DESCRIPTION

[0035] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the technical scheme in the specific implementation mode of the present invention is clearly and completely described below to further illustrate the present invention. Obviously, the specific implementation mode described is only a part of the implementation mode of the present invention, rather than all styles.

[0036] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0037] As attached Figure 1 As shown, the present invention provides a method for constructing a climate-energy market supply and demand model based on artificial intelligence:

[0038] Step S1: Collect historical climate data and actual power generation data and actual electricity consumption data of residents in the same period as training sets;

[0039] Step S2: Based on the Spearman rank correlation coefficient, select climate data with high correlation with power generation as input of the climate-energy supply model, and train the climate-energy supply model;

[0040] Step S3: Based on the Spearman rank correlation coefficient, select climate data with high correlation with residential electricity consumption as input of the climate-energy demand model, and train the climate-energy demand model;

[0041] Step S4: Acquire climate data with high correlation with power generation in the area to be analyzed, input the trained climate-energy supply model, obtain the supply power generation forecast value, acquire climate data with high correlation with residential power consumption in the area to be analyzed, input the trained climate-energy demand model, and obtain the demand power consumption forecast value;

[0042] Step S5: Optimize scheduling based on the supply power generation forecast value and the demand power consumption forecast value.

[0043] As attached Figure 2 As shown, the climate-energy supply model includes: a CNN spatiotemporal feature extraction module and a fully connected neural network module, the climate data highly correlated with power generation is input into the CNN spatiotemporal feature extraction module for feature extraction, the feature is input into the fully connected neural network module for prediction, and the supply power generation prediction value is output;

[0044] The CNN spatiotemporal feature extraction module includes an input layer, a convolutional layer, a temporal feature extraction module, and a spatial feature extraction module connected in sequence. The output of the convolutional layer is denoted as Fc, the output of the temporal feature extraction module is denoted as Ft, and the output of the spatial feature extraction module is denoted as Fs.

[0045] Among them, as attached Figure 3 As shown in FIG. 1 , the time feature extraction module is obtained by:

[0046]

[0047] As attached Figure 5 As shown, TAB(Fc)=Fc⊙Upsample(MLP(GAP(Fc))),

[0048] Among them, MLP() represents multi-layer perceptron, GAP() represents global average pooling, and Upsample() represents upsampling;

[0049] As attached Figure 4 As shown, the spatial feature extraction module Fs is obtained by:

[0050]

[0051] As attached Figure 6 As shown,

[0052]

[0053] Among them, conv() represents the convolution layer, dwconv1() represents the 3X3 depth separation convolution layer, dwconv2() represents the 5X5 depth separation convolution layer, and pw() represents the point-by-point convolution layer.

[0054] Furthermore, the climate-energy demand model is a CNN-LightGBM model.

[0055] Furthermore, climate data includes climate element data and climate risk index data.

[0056] Furthermore, climate element data include: temperature (2-meter air temperature, dew point temperature), precipitation, humidity, wind speed (10 meters and 100 meters), wind direction, sea level pressure, radiation intensity, runoff, evaporation, snowfall, sunshine duration, etc.

[0057] Furthermore, the climate risk index data includes: flooding, drought, typhoon, high temperature, and low temperature freezing.

[0058] The present invention proposes a method for constructing a climate-energy market supply and demand model based on artificial intelligence, which collects historical climate data and actual power generation data and actual electricity consumption data of residents in the same period as training sets; based on the Spearman rank correlation coefficient, selects climate data with high correlation with power generation as input of a climate-energy supply model, and trains the climate-energy supply model; based on the Spearman rank correlation coefficient, selects climate data with high correlation with residents' electricity consumption as input of a climate-energy demand model, and trains the climate-energy demand model; obtains climate data with high correlation with power generation in a to-be-analyzed area, inputs the trained climate-energy supply model, obtains a predicted value of supply power generation in the area, obtains climate data with high correlation with residents' electricity consumption in the to-be-analyzed area, inputs the trained climate-energy demand model, and obtains a predicted value of demand power consumption in the area; optimizes scheduling based on the predicted value of supply power generation and the predicted value of demand power consumption, thereby effectively reducing the impact of the instability of new energy power generation on the power grid and improving the comfort of residents' electricity consumption and the stability of industrial electricity consumption.

[0059] The above describes the main technical features and basic principles of the present invention and the related advantages. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the concept or basic features of the present invention. Therefore, no matter from which point of view, the above specific embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the attached claims rather than the above description, and it is intended that all changes that fall within the meaning and scope of the equivalent elements of the claims are included in the present invention.

[0060] In addition, it should be understood that although the present specification is described according to various implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each implementation method may also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

Claims

1. A method for constructing a climate-energy market supply and demand model based on artificial intelligence, characterized in that: The method comprises the following steps: Step S1: Collect historical climate data and actual power generation data and actual residential power consumption data for the same period as a training set; Step S2: Based on the Spearman rank correlation coefficient, select climate data with high correlation with power generation as input of the climate-energy supply model, and train the climate-energy supply model; Step S3: Based on the Spearman rank correlation coefficient, select climate data with high correlation with residential electricity consumption as input of the climate-energy demand model, and train the climate-energy demand model; Step S4: Acquire climate data with high correlation with power generation in the area to be analyzed, input the trained climate-energy supply model, obtain the supply power generation forecast value, acquire climate data with high correlation with residential power consumption in the area to be analyzed, input the trained climate-energy demand model, and obtain the demand power consumption forecast value; Step S5: Optimizing scheduling based on the supply power generation forecast value and the demand power consumption forecast value; The climate-energy supply model includes: a CNN spatiotemporal feature extraction module and a fully connected neural network module, wherein the climate data highly correlated with power generation is input into the CNN spatiotemporal feature extraction module for feature extraction, the features are input into the fully connected neural network module for prediction, and the predicted value of power generation is output; The CNN spatiotemporal feature extraction module includes a sequentially connected input layer, a convolutional layer, a temporal feature extraction module, and a spatial feature extraction module. The output of the convolutional layer is denoted as Fc, the temporal feature extraction module is denoted as Ft, and the output of the spatial feature extraction module is denoted as Fs. Among them, Ft is obtained by the following method: TAB(Fc)=Fc⊙Upsample(MLP(GAP(Fc))), Among them, MLP() represents multi-layer perceptron, GAP() represents global average pooling, and Upsample() represents upsampling; Fs is obtained as follows: Among them, conv() represents the convolution layer, dwconv1() represents the 3X3 depth separation convolution layer, dwconv2() represents the 5X5 depth separation convolution layer, and pw() represents the point-by-point convolution layer.

2. The method for constructing a climate-energy market supply and demand model based on artificial intelligence according to claim 1, characterized in that: The climate-energy demand model is the CNN-LightGBM model.

3. The method for constructing a climate-energy market supply and demand model based on artificial intelligence according to claim 1, characterized in that: Climate data includes climate element data and climate risk index data.

4. The method for constructing a climate-energy market supply and demand model based on artificial intelligence according to claim 3, characterized in that: Climate element data include: 2-meter air temperature, dew point temperature, precipitation, humidity, 10-meter wind speed, 100-meter wind speed, wind direction, sea level pressure, radiation intensity, runoff, evaporation, snowfall, and sunshine duration.

5. The method for constructing a climate-energy market supply and demand model based on artificial intelligence according to claim 3, characterized in that: Climate risk index data include: flooding, drought, typhoon, high temperature, and low temperature freezing.

Images