Wind and light renewable energy project space-time analysis method based on remote sensing data and GIS technology

Abstract

The invention discloses a wind and light renewable energy project space-time analysis method based on remote sensing data and a deep learning algorithm, and belongs to the field of renewable energy. The method comprises the following steps: 1) collecting and processing data; 2) land suitability evaluation: based on the collected data, analyzing by using a GIS tool, and determining an area suitable for developing a wind and light renewable energy source project; 3) meteorological prediction model construction: training and predicting future annual data by using an LSTM model in combination with historical meteorological data; 4) constructing a wind and light renewable energy project power generation model; and 5) space-time analysis method application: according to the determined suitable area, combining meteorological data, and utilizing the constructed power generation model to carry out space-time analysis on the renewable energy project. According to the method, remote sensing data, GIS analysis and a deep learning algorithm are applied to the field of wind and light renewable energy sources, a space-time analysis method of renewable energy source projects is formed, multi-source data are fused, and rapid screening of the renewable energy source projects can be achieved.

Description

Technical Field

[0001] The present invention belongs to the field of renewable energy, and more specifically, relates to a spatiotemporal analysis method for wind and solar renewable energy projects based on remote sensing data and GIS technology. Background Art

[0002] The installed capacity of renewable energy is increasing rapidly, and the newly added installed capacity in the power industry is mainly composed of renewable energy installed capacity. As important components of renewable energy projects, wind power and photovoltaic power are increasingly being planned and constructed. In this context, higher requirements are put forward for the feasibility analysis of wind and solar renewable energy projects, which not only involves energy production, management and optimization, but also has a close relationship with construction planning, weather forecasting and other aspects.

[0003] At present, the spatiotemporal analysis of wind and solar renewable energy projects mostly relies on the solar or wind resource conditions of the region throughout the year, which are usually the integration of data for the whole year. When faced with detailed spatiotemporal analysis of wind and solar renewable energy projects, such results expose several limitations and constraints. First, the analysis results are too generalized in terms of time, and detailed analysis results for every hour of the whole year cannot be obtained; second, the traditional method has bottlenecks in spatial analysis, which limits users' ability to conduct instant spatial analysis and real-time decision-making for more detailed wind and solar renewable energy projects; finally, the method fails to predict future weather and climate changes, especially in scenarios where extreme weather occurs frequently. Summary of the invention

[0004] In view of the defects of the prior art, the purpose of the present invention is to provide a spatiotemporal analysis method for wind and solar renewable energy projects based on remote sensing data and GIS technology.

[0005] To achieve the above purpose, the present invention provides a spatiotemporal analysis method for wind and solar renewable energy projects based on remote sensing data and GIS technology, comprising the following steps:

[0006] Step 1: Collect land use type raster data, environmental protection area type raster data, topography raster data, and meteorological data. Based on the regional GIS model, collect land use type data (land cover type), environmental protection area type data, topography data (slope data), and meteorological data (solar radiation and wind speed, etc.), and process and convert the data into a specific format.

[0007] Step 2: Conduct land suitability assessment for wind and solar renewable energy projects. Based on the land use type raster data, environmental protection area type raster data, and topography raster data collected in step 1, define the constraints of wind and solar renewable energy projects, use GIS tools to conduct land suitability assessment, and screen out areas suitable for the development of wind and solar renewable energy projects.

[0008] Step 3: Build a weather forecast model. Use historical weather data to train the LSTM model, input the weather data collected in step 1 into the trained LSTM model to predict the weather data for the next year; assign the predicted weather data for the next year to the area in step 2.

[0009] The hyperparameters used during model training are shown in the following table.

[0010] Hyperparameters value LSTM Cell 32–32,64–64,128–128,256–256,512–512 Training rounds 100 Optimizer AdamW Learning Rate Adaptation during training Loss Function Mean squared error Activation function of LSTM layer tanh Batch size 32

[0011] Step 4: Construct a power generation model for wind and solar renewable energy projects. Simulate the installation of standard capacity photovoltaic panels and wind turbines, combine the power generation mechanism and characteristics of photovoltaic panels and wind turbines to establish a power generation model for renewable energy projects, and calculate the hourly power generation of wind and solar renewable energy projects.

[0012] Furthermore, step 2 specifically includes:

[0013] Step 2-1: Using the constraints of the environmental protection zone type, the “raster mask” function in the GIS tool is used to exclude the environmental protection zone raster data that is not suitable for the development of wind and solar renewable energy projects;

[0014] Step 2-2: Using the constraints of topography and geomorphology, the “reclassification” and “raster masking” functions in GIS tools are used to exclude topographic raster data that are not suitable for the development of wind and solar renewable energy projects;

[0015] Step 2-3: Using the constraints of land use types, use the "raster overlay analysis" function in the GIS tool to exclude the land use type raster data that is not suitable for the development of wind and solar renewable energy projects;

[0016] Step 2-4: Use GIS tools to integrate the screened data to obtain areas suitable for the development of wind and solar renewable energy projects.

[0017] The land use type constraints suitable for the development of wind and solar renewable energy projects in step 2 are as follows:

[0018] Land use type Photoelectric suitability Wind power suitability farmland Unsuitable suitable forest Unsuitable Unsuitable shrub suitable suitable grassland suitable suitable Water Unsuitable Unsuitable Snow suitable Unsuitable Bare Land suitable suitable Impervious surface Unsuitable Unsuitable Wetlands Unsuitable Unsuitable

[0019] The constraints of the environmental protection zone suitable for the development of wind and solar renewable energy projects in step 2 are as follows:

[0020]

[0021] In addition, the land suitability for photovoltaic renewable energy projects is limited to a slope of less than 5, and the land suitability for wind power renewable energy projects is limited to a slope of less than 20.

[0022] In step 3, the long short-term memory network (LSTM) is used to predict the required parameters of the mathematical model of the hourly power generation of wind and solar renewable energy projects. The long short-term memory network forgets the gate f through the interacting gate units. t , input gate i t , output gate O t And the unit state plays a role. The specific calculation steps and formulas are as follows:

[0023] Step 3-1, calculate the forget gate, which determines which memories from the previous time step should be discarded (i.e., the cell state C of the previous time step) t-1 ). It uses the Sigmoid activation function to process the current input X t and the hidden state h at the previous moment t-1 , output a vector f with a value range between 0 and 1 t , indicating the degree of forgetting. t Contains weather data for the current time step.

[0024] f t =δ(W t ×[h t-1 ,X t ]+b i )

[0025] W t is the weight matrix of the forget gate, b i is the bias term of the forget gate, and δ is the Sigmoid activation function.

[0026] Step 3-2, calculate the input gate i t and candidate cell state C t ′ The input gate uses the Sigmoid function to determine the current input X t and the hidden state h at the previous time step t-1 How much information should be stored in the candidate cell state? The candidate cell state is generated using the tanh function. t ′ , which represents the new information to be added to the cell state. t Contains weather data for the current time step.

[0027] i t =δ(W t ×[h t-1 ,X t ]+b i )

[0028] C t ′ =tanh(W c ×[ht-1 ,X t ]+b c )

[0029] W c is the weight matrix of the input gate, b c is the bias term of the forget gate.

[0030] Step 3-3, Update cell status C t The new cell state C t is obtained by combining the cell state C of the previous time step t-1 and the new candidate state C t ′ To update. t Determines how much of the previous state to keep, input gate i t Determines how many new candidate states to add. The updated cell state is as follows:

[0031] C t =f t ×C t-1 +i t ×C t ′

[0032] Step 3-4, calculate the output gate O t and the hidden state h t The output gate controls the current cell state C t How much information should be output as the current hidden state h t The output gate is calculated by the Sigmoid activation function to process the current input X t and the hidden state h at the previous time step t-1 , and get a value. Then, this value is combined with the updated cell state C t Multiply the tanh result of to generate the final hidden state.

[0033] O t =δ(W α ×[h t-1 ,X t ]+b α )

[0034] W α is the weight matrix of the output gate, b α is the bias term of the output gate. The hidden state h t The calculation formula is:

[0035] h t =O t ×tanh(C t )

[0037] In the above steps, the meteorological data X t The meteorological data is introduced as input into each gated unit of the LSTM model. t The input is sent to the forget gate, input gate, and output gate, affecting the calculation results of each gate. By using past meteorological data, the LSTM model can learn the temporal relationship between meteorological parameters and predict future meteorological data. The final hidden state h t Represents predicted future weather data. These prediction results will be used in subsequent power generation prediction models to ensure that wind and solar renewable energy projects can be optimized based on the latest weather forecasts.

[0038] The specific mathematical model calculation process of hourly power generation of wind power renewable energy projects is as follows:

[0039] Hourly electricity generation E of onshore wind farms wind The calculation formula is:

[0040] E wind =P wind ×CF wind ×η conv ×η array

[0041] P wind =A wind ×ρ wind

[0042]

[0043] Among them, P wind is the wind power capacity potential, ρ wind Indicates wind power density, CF wind is the wind turbine's power generation capacity factor, η conv and η array They are energy conversion efficiency (0.95) and array efficiency coefficient (0.9), A wind is the pixel area where wind turbines are installed, D wind is the rotor diameter of the wind turbine (103m), MP wind is the maximum power of the wind turbine (2.5MW). wind Reference is made to General Electric's power curve.

[0044] The specific calculation process of the mathematical model for the hourly power generation of photovoltaic renewable energy projects is as follows:

[0045] Assuming the standard power capacity of the installation is 161.9W / m 2 The photovoltaic panels. Define the hourly power generation E of the photovoltaic power plant solar The calculation formula is:

[0046]

[0047] Among them, P solar Represents the photovoltaic capacity potential, I panel represents the solar radiation captured by the photovoltaic panel, μ shade Represents the shadow coefficient, μ temp Represents the temperature coefficient, μ loss Represents the power loss from power generation to grid connection (19.44%).

[0048] The photovoltaic capacity potential is calculated as follows:

[0049] P solar =A solar ×ψ×SP

[0050] Among them, A solar is the pixel area where the photovoltaic panel is installed, ψ is the ratio of the effective photovoltaic panel to the pixel area where the photovoltaic panel is installed, and SP is the standard power capacity (161.9W / m 2 ).

[0051] The effective area ratio is calculated as follows:

[0052]

[0053] Among them, L p represents the length of the photovoltaic panel, ω is the tilt angle of the photovoltaic panel, β is the solar altitude angle, and φ is the solar azimuth angle.

[0054]

[0055] The solar radiation captured by the photovoltaic panel is panel By direct radiation I direct , diffuse radiation I diff , reflected radiation I ref It consists of three parts. direct represents the direct solar radiation per hour, S diff represents the hourly diffuse solar radiation, S total is the total solar radiation, ρ G represents the surface albedo (0.2).

[0056] The calculation formula of the photovoltaic panel tilt angle is as follows:

[0057]

[0058] Where θ represents the latitude where the solar panel is installed, a0(1.3793), a1(1.2011), a2(-0.0144), a3(1.3793)

[0059] The formula for calculating the solar altitude angle is as follows:

[0060]

[0061] The formula for calculating the solar azimuth is as follows:

[0062]

[0063] Solar declination

[0064] δ=b0+b1cosd A +c1sind A +b2cos2d A +c2sin2d A +b3cos3d A +c3sin3d A

[0066]

[0067] Among them, b0(0.00692), b1(-0.400), b2(-0.006768), b3(-0.00270), c1(0.0703), c2(0.000907), c3(0.00148). A Indicates the angle of the day, used to indicate the change in the position of the sun in a year, indicating the angle of the number of days in a year in the solar year cycle. d Represents (1-365 days of the year).

[0068] The formula for calculating the solar hour angle is as follows:

[0069]

[0070] Among them, b0(0.00692), b1(-0.400), b2(-0.006768), b3(-0.00270), c1(0.0703), c2(0.000907), c3(0.00148). A Indicates the angle of the day, used to indicate the change in the position of the sun in a year, indicating the angle of the number of days in a year in the solar year cycle. d represents (1-365 days in a year), t hour Indicates (1-24 hours in a day), t min Indicates (1-60 minutes in an hour).

[0071] The shading coefficient calculation formula is as follows:

[0072]

[0073] The temperature coefficient calculation formula is as follows:

[0074]

[0075] Among them, T atm It represents the hourly atmospheric surface temperature at 2 meters above the ground, T cell is the normal operating temperature of the battery (44°C), η T is the temperature coefficient (-0.41% per °C).

[0076] Through the above technical scheme conceived by the present invention, compared with the prior art, the spatiotemporal analysis method of wind and solar renewable energy projects based on remote sensing data and GIS technology provided by the present invention collects various data such as land use type, environmental protection area, topography, meteorological environment, etc., and uses GIS tools to evaluate land suitability and determine areas suitable for the development of wind and solar renewable energy projects; combined with the LSTM model, through the training of historical meteorological data, the meteorological data of future years is predicted; the power generation model of wind and solar renewable energy projects is constructed to simulate the power generation of photovoltaic panels and wind turbines of standard capacity; finally, combined with suitable areas and predicted meteorological data, the power generation model is applied to perform spatiotemporal analysis, thereby realizing the rapid screening and optimization layout of wind and solar renewable energy projects. The present invention combines remote sensing data, GIS analysis and deep learning algorithms to provide an efficient and intelligent solution for the spatiotemporal analysis of wind and solar renewable energy projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 This is a flow chart of the spatiotemporal analysis method for wind and solar renewable energy projects based on remote sensing data and GIS technology. DETAILED DESCRIPTION

[0078] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0079] like Figure 1 As shown, the present invention is a spatiotemporal analysis method for wind and solar renewable energy projects based on remote sensing data and GIS technology, comprising the following steps:

[0080] Step 1: Collect and process land use type raster data, environmental protection area type raster data, topography raster data, and meteorological data. Based on the regional GIS model, collect land use type data (land cover type), environmental protection area type data, topography data (slope data), and meteorological data (solar radiation and wind speed, etc.), and convert the data into a specific format.

[0081] Step 2: Conduct land suitability assessment for wind and solar renewable energy projects. Based on the land use type raster data, environmental protection area type raster data, and topography raster data collected in step 1, define the constraints of wind and solar renewable energy projects, use GIS tools to conduct land suitability assessment, and screen out areas suitable for the development of wind and solar renewable energy projects.

[0082] Step 3: Use historical meteorological data to train the LSTM model, input the meteorological data collected in step 1 into the trained LSTM model to predict the meteorological data of the future year; assign the predicted meteorological data of the future year to the area of ​​step 2.

[0083] Step 4: Construct a power generation model for wind and solar renewable energy projects. Simulate the installation of standard capacity photovoltaic panels and wind turbines, combine the power generation mechanism and characteristics of photovoltaic panels and wind turbines to establish a power generation model for renewable energy projects, and calculate the hourly power generation of wind and solar renewable energy projects.

[0084] The following are specific embodiments of the present invention.

[0085] First, collect 30m×30m land cover data of our country;

[0086] Secondly, based on the land suitability assessment results and restricted areas, areas that are not suitable for the development of wind and solar renewable energy projects are excluded;

[0087] Then use the LSTM model to combine historical weather and meteorological data to train the model and predict future weather and meteorological data;

[0088] Finally, based on the mathematical model for calculating the hourly power generation of wind and solar renewable energy projects and the geometric weather data, the renewable energy power generation of each suitable area is calculated, and the final map of photovoltaic and wind power renewable energy projects suitable for development in China is obtained.

[0089] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A spatiotemporal analysis method for wind and solar renewable energy projects based on remote sensing data and GIS technology, characterized in that: The following steps are involved: Step 1: Collect land use type raster data, environmental protection area type raster data, topography raster data, and meteorological data; Step 2: Based on the land use type raster data, environmental protection area type raster data, and topographic raster data collected in step 1, define the constraints of wind and solar renewable energy projects, use GIS tools to conduct land suitability assessment, and screen out areas suitable for the development of wind and solar renewable energy projects; Step 3: Use historical meteorological data to train the LSTM model, input the meteorological data collected in step 1 into the trained LSTM model to predict the meteorological data of the future year; assign the predicted meteorological data of the future year to the area of ​​step 2; Step 4: Simulate the installation of photovoltaic panels and wind turbines of standard capacity to build a renewable energy project power generation model; Step 5: Based on the area determined in step 3, the power generation model constructed in step 4 is used to conduct spatiotemporal analysis of renewable energy projects.

2. The method according to claim 1, characterized in that: Step 2 specifically includes: Step 2-1: Using the constraints of the environmental protection zone type, the "raster mask" function in the GIS tool is used to exclude the environmental protection zone raster data that is not suitable for the development of wind and solar renewable energy projects; Step 2-2: Using the constraints of topography and geomorphology, the "reclassification" and "raster masking" functions in the GIS tool are used to exclude the topography and geomorphology raster data that are not suitable for the development of wind and solar renewable energy projects; Step 2-3: Using the constraints of land use types, use the "raster overlay analysis" function in the GIS tool to exclude the land use type raster data that is not suitable for the development of wind and solar renewable energy projects; Step 2-4: Use GIS tools to integrate the screened data to obtain areas suitable for the development of wind and solar renewable energy projects.

3. The method according to claim 1, characterized in that Current weather data X in step 3 t It is introduced as input into the forget gate, input gate and output gate of the LSTM model, and the final hidden state h t Represents the predicted future weather data, as follows: Step 3-1: Use the Sigmoid activation function to calculate the forget gate, which processes the current input X t and the hidden state h at the previous moment t-1 , output a vector f with a value range between 0 and 1 t , indicating the degree of forgetfulness; f t =δ(W t ×[h t-1 ,X t ]+b i ) Among them, W t is the weight matrix of the forget gate, b i is the bias term of the forget gate, and δ is the Sigmoid activation function; Step 3-2: Use the Sigmoid activation function to calculate the input gate i t , use the tanh function to generate candidate cell states C t ′ ; The input gate determines the current input X t and the hidden state h at the previous time step t-1 How much information should be stored in the candidate cell state C t ′ Represents new information to be added to the cell state; i t =δ(W t ×[h t-1 ,X t ]+b i ) C t ′ =tanh(W c ×[h t-1 ,X t ]+b c ) Among them, W c is the weight matrix of the input gate, b c is the bias term of the forget gate; Step 3-3, combined with the cell state C of the previous time step t-1 and the new candidate state C t ′ Update cell status C t ; Forget gate f t Determines how much of the previous state to keep, input gate i t Determines how many new candidate states to add; the updated cell states are as follows: C t =f t ×C t-1 +i t ×C t ′ Step 3-4: Calculate the output gate O through the Sigmoid activation function t and the hidden state h t , the output gate O t Control the current cell state C t How much information should be output as the current hidden state h t , the O t For: t =δ(W α ×[h t-1 ,X t ]+b α ), where W α is the weight matrix of the output gate, b α Is the bias term of the output gate; processes the current input X t and the hidden state h at the previous time step t-1 , get a value, and compare it with the updated cell state C t Multiply the tanh result of the hidden state h to generate the final hidden state h t ; The final hidden state h t For: h t =O t ×tanh(C t ).

4. The method according to claim 3, characterized in that Hourly electricity generation of wind power renewable energy projects E wind The calculation formula is: AND wind =P wind ×CF wind ×η conv ×η array P wind =A wind ×ρ wind Among them, P wind is the wind power capacity potential, ρ wind Indicates wind power density, CF wind is the wind turbine's power generation capacity factor, η conv and η array are energy conversion efficiency and array efficiency coefficient, A wind is the pixel area where wind turbines are installed, D wind is the rotor diameter of the wind turbine, MP wind is the maximum power of the wind turbine.

5. The method according to claim 1, characterized in that Hourly electricity generation of photovoltaic renewable energy projects E solar The calculation formula is: Among them, P solar Represents the photovoltaic capacity potential, I panel represents the solar radiation captured by the photovoltaic panel, μ shade Represents the shadow coefficient, μ temp Represents the temperature coefficient, μ loss Represents the power loss from power generation to grid connection.

6. An electronic device, characterized in that: include: A computer readable storage medium and a processor; The computer-readable storage medium is used to store executable instructions; The processor is used to read the executable instructions stored in the computer-readable storage medium and execute the method according to any one of claims 1 to 5.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to execute the method according to any one of claims 1 to 5.

8. A computer program product comprising a computer program or instructions, characterized in that When the computer program or instruction is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images