Battery charging network deployment method using renewable energy sources

Abstract

The invention discloses a battery charging network deployment method using renewable energy. The method comprises the following steps: S1, collecting power grid state data and renewable energy supply data in real time through a sensor network; s2, predicting the power grid load and the renewable energy source supply based on the power grid state data and the renewable energy source supply data by using a long short-term memory network; s3, adopting regression analysis to determine an optimal scheduling plan of the charging demand of the electric vehicle based on a long short-term memory network prediction result; and S4, according to the real-time change of the power grid and the renewable energy source supply, a charging scheduling plan is dynamically updated, and the charging scheduling plan comprises adjustment of charging power and time. According to the invention, through intelligent and automatic charging network management, power grid changes are effectively adapted, energy utilization is optimized, environmental influences are reduced, and a solid technical support is provided for sustainable and intelligent power grid management.

Description

Technical Field

[0001] The present invention relates to the field of battery technology, and in particular to a battery charging network deployment method utilizing renewable energy. Background Art

[0002] Currently, with the popularity of electric vehicles and the increase in renewable energy, grid management faces new challenges and opportunities. In particular, in terms of electric vehicle charging, it is necessary to effectively manage the grid load while optimizing charging costs and efficiency. Existing technologies mainly rely on traditional grid management methods, which have limitations in dealing with electric vehicle charging needs and renewable energy supply. For example, traditional charging networks are usually unable to effectively allocate resources during peak grid load periods, resulting in the risk of grid overload and inefficiency.

[0003] Existing technologies usually rely on static charging plans and limited data analysis capabilities to deal with the risk of grid overload during peak hours and underutilization of resources during off-peak hours. This approach is inefficient in predicting grid load and renewable energy supply, and cannot dynamically adjust charging strategies to adapt to real-time changes in the grid. In addition, existing charging methods often do not fully utilize renewable energy, resulting in energy waste and over-reliance on traditional power resources.

[0004] Therefore, how to provide a method that can effectively manage grid load, optimize electric vehicle charging strategies, and make full use of renewable energy has become an urgent problem that technicians in this field need to solve. In particular, a method is needed that can monitor grid status and electricity price changes in real time and automatically adjust charging strategies to reduce grid pressure during peak hours, reduce charging costs, and improve energy efficiency and reduce carbon emissions. In addition, the method must also be able to optimize the use of renewable energy, improve the overall operating efficiency of the grid, and provide electric vehicle users with more economical, efficient, and environmentally friendly charging solutions. Summary of the invention

[0005] One purpose of the present invention is to propose a battery charging network deployment method using renewable energy. The present invention effectively adapts to grid changes, optimizes energy utilization, and reduces environmental impact through intelligent and automated charging network management, thereby providing solid technical support for achieving sustainable and intelligent grid management.

[0006] A method for deploying a battery charging network using renewable energy according to an embodiment of the present invention comprises the following steps:

[0007] S1, collect grid status data and renewable energy supply data in real time through sensor networks;

[0008] S2. Use long short-term memory networks to predict grid load and renewable energy supply based on grid status data and renewable energy supply data;

[0009] S3, using regression analysis to determine the optimal scheduling plan for electric vehicle charging demand based on the prediction results of the long short-term memory network;

[0010] S4. Dynamically update the charging scheduling plan according to the real-time changes in the power grid and renewable energy supply, wherein the charging scheduling plan includes adjusting the charging power and time.

[0011] Optionally, the S1 specifically includes:

[0012] S11. Using a sensor network to collect grid status data and renewable energy supply data in real time, where the grid status data includes grid load, grid frequency and grid voltage, and the renewable energy supply data includes wind and solar power generation;

[0013] S12. The sensor network transmits the collected data to a data processing center through an effective data transmission mechanism.

[0014] Optionally, the S2 specifically includes:

[0015] S21. Constructing long short-term memory network:

[0016] i t =σ(W i *[h t-1 ,x t ]+b i )

[0017] f t =σ(W f *[h t-1 ,x t ]+b f )

[0018] o t =σ(W o *[h t-1 ,x t ]+b o )

[0019] c t =f t *c t-1 +i t *tanh(W c *[h t-1 ,x t ]+b c )

[0020] Among them, the input gate i tRepresents the grid load data, and the forget gate f t : Indicates that outdated grid load patterns are forgotten, and the output gate is o t Predict the upcoming grid load, unit status c t Represents key information about grid load and renewable energy supply history;

[0021] S22, the long short-term memory network model integrates the grid state and renewable energy supply data and the interface of the key components of the grid management method, wherein the interface of the key components of the grid management method also includes an electricity price module and a demand response controller;

[0022] S23, the long short-term memory network model uses adaptive algorithm optimization to adjust the learning rate and memory unit configuration to adapt to the dynamic changes of the power grid;

[0023] S24. Adjust the charging station scheduling, including charging time and power, based on the output results of the long short-term memory network model.

[0024] Optionally, the S22 specifically includes:

[0025] S221, the long short-term memory network model is configured to receive the power grid state data N s and renewable energy supply data R s As input;

[0026] S222, the long short-term memory network model integrates the electricity price data P and the demand response signal D from the power grid management method;

[0027] S223. The integrated data fusion process can be expressed as:

[0028] X t =f(N s ,R s ,P,D)

[0029] Among them, X t represents the comprehensive input vector of the long short-term memory network model at time t, f represents the data fusion function, N s Indicates the power grid status data, R s represents renewable energy supply data, P represents electricity price data, and D represents demand response signal;

[0030] S224, long short-term memory network model applies the comprehensive input vector X t Forecast grid load and renewable energy supply.

[0031] Optionally, the S23 specifically includes:

[0032] S231. In a long short-term memory network model, a customized adaptive learning rate algorithm for grid load and renewable energy supply forecasting is used;

[0033] S232, the adaptive learning rate algorithm realizes dynamic adjustment of the learning rate through the following formula:

[0034] η t+1 =η t *g(ΔE t ,ΔE t-1 )

[0035] Among them, η t+1 and η t are the learning rates in the consecutive training stages t+1 and t, ΔE t Indicates the current forecast stage and actual grid load, ΔE t-1 is the error in the previous prediction stage, and g is an adjustment function based on the change in error;

[0036] S233. Through the learning rate adjustment mechanism, the model responds promptly to rapid changes in the grid status and renewable energy supply.

[0037] Optionally, the S3 specifically includes:

[0038] S31. Determine the optimal scheduling plan for electric vehicle charging demand using a regression analysis method based on the prediction results of grid load and renewable energy supply provided by a long short-term memory network;

[0039] S32, the regression analysis includes evaluating the relationship between grid load, electricity price, renewable energy supply and electric vehicle charging demand, and calculating the optimal charging time and power;

[0040] S33. The customized regression model uses the following formula to consider the relationship between electric vehicle charging and grid load:

[0041] y=β 0 +β 1 XL+β 2 f(X E )+β 3 X P +γZ+ε

[0042] Where y represents the charging power or charging time of the charging station; X L Indicates the grid load; X E represents the renewable energy supply, f(X E ) is a nonlinear transformation of renewable energy supply, indicating its impact on charging demand; X Prepresents the factors that affect the charging strategy; γZ represents the specific features of the LSTM prediction results or other relevant grid information; β 0 ,β 1 ,β 2 ,β 3 and γ represent the influence of each variable on the charging strategy; ε represents the difference between the model prediction and the actual value.

[0043] Optionally, the S32 specifically includes:

[0044] S321, using a multivariate linear regression model to calculate the optimal charging time and power of the electric vehicle charging station, the model taking into account the grid load, electricity price, renewable energy supply and the impact on future grid load forecast;

[0045] S322, the form of the multivariate linear regression model is:

[0046] P opt =α 0 +α 1 L+α 2 E+α 3 R+α 4 F(L,R)+ε

[0047] T opt =β 0 +β 1 L+β 2 E+β 3 R+β 4 F(L,R)+ε

[0048] Among them, P opt represents the optimal charging power, T opt represents the optimal charging time, L is the grid load, E is the electricity price, R represents the renewable energy supply, and F(L,R) represents the function of grid load and renewable energy supply;

[0049] S323, by optimizing the model parameters α and β, the model predicts the optimal charging power P under specific grid conditions and resource availability. opt and charging time T opt .

[0050] Optionally, the S321 specifically includes:

[0051] S3211. Analyze the data of power grid load L, electricity price E, and renewable energy supply R using a multivariate linear regression model to obtain data on optimal charging time and power of electric vehicle charging stations;

[0052] S3212, the model analyzes the relationship between charging demand and grid capacity according to the grid load L, and optimizes the operation of the charging station during the high-load period of the grid;

[0053] S3213. Considering the electricity price E, the model adjusts the charging strategy to utilize the low electricity price period to reduce the charging cost;

[0054] S3214: Input the renewable energy supply R into the model, and adjust the charging plan to maximize the use of renewable energy.

[0055] Optionally, the S4 specifically includes:

[0056] S41, optimal charging time T calculated according to S3 opt and power P opt Develop charging scheduling plans for electric vehicle charging stations;

[0057] S42. Implement an automatic adjustment mechanism that dynamically adjusts the charging time and power of the charging station according to real-time grid load and renewable energy supply data;

[0058] S43, adjusting includes evaluating the deviation between the current grid state and the predicted state, and optimizing the charging power and time accordingly;

[0059] S44, the adjustment mechanism uses the following formula to make real-time adjustments:

[0060] P adj =P opt *(1+δ P *ΔL)

[0061] T adj =T opt *(1+δ T *ΔE)

[0062] Among them, P adj Indicates the adjusted charging power and T adj represents time, δ P and δ T is the adjustment coefficient, ΔL is the real-time change of grid load, and ΔE is the real-time change of electricity price.

[0063] The beneficial effects of the present invention are:

[0064] The present invention provides a battery charging network deployment method that combines advanced data analysis, prediction algorithms and smart grid technology. It uses long short-term memory networks (LSTM) and regression analysis to accurately predict grid load and renewable energy supply, thereby achieving efficient grid load management and intelligent charging methods. The method can dynamically reduce non-emergency power consumption during peak grid demand and reduce grid pressure, while optimizing energy utilization during low-load periods, especially giving priority to the use of renewable energy for charging. This not only improves energy efficiency and reduces dependence on traditional power resources, but also reduces carbon emissions. In addition, the method provides electric vehicle users with a more economical, efficient and environmentally friendly charging solution. Overall, the present invention effectively adapts to grid changes, optimizes energy utilization, and reduces environmental impact through intelligent and automated charging network management, providing solid technical support for achieving sustainable and intelligent grid management. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0066] Figure 1 A schematic diagram of the overall architecture of a battery charging network deployment method using renewable energy proposed by the present invention;

[0067] Figure 2 The configuration and functional flow of the long short-term memory network (LSTM) model in this method are demonstrated, including the data input and prediction process of grid load and renewable energy supply. DETAILED DESCRIPTION

[0068] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0069] refer to Figure 1 , a hydraulic support pin hole alignment guidance method based on deep vision algorithm, comprising the following steps:

[0070] S1, collect grid status data and renewable energy supply data in real time through sensor networks;

[0071] In this implementation, S1 specifically includes:

[0072] S11. Using a sensor network to collect grid status data and renewable energy supply data in real time, where the grid status data includes grid load, grid frequency and grid voltage, and the renewable energy supply data includes wind and solar power generation;

[0073] S12. The sensor network transmits the collected data to a data processing center through an effective data transmission mechanism.

[0074] S2. Use long short-term memory networks to predict grid load and renewable energy supply based on grid status data and renewable energy supply data;

[0075] In this implementation, S2 specifically includes:

[0076] S21. Constructing long short-term memory network:

[0077] i t =σ(W i *[h t-1 ,x t ]+b i )

[0078] f t =σ(W f *[h t-1 ,x t ]+b f )

[0079] o t =σ(W o *[h t-1 ,x t ]+b o )

[0080] c t =f c *c t-1 +i t *tanh(W c *[h t-1 ,x t ]+b c )

[0081] Among them, the input gate i t Represents the grid load data, and the forget gate f t : Indicates that outdated grid load patterns are forgotten, and the output gate is o t Predict the upcoming grid load, unit status c t Represents key information about grid load and renewable energy supply history;

[0082] S22, the long short-term memory network model integrates the grid state and renewable energy supply data and the interface of the key components of the grid management method, wherein the interface of the key components of the grid management method also includes an electricity price module and a demand response controller;

[0083] S23, the long short-term memory network model uses adaptive algorithm optimization to adjust the learning rate and memory unit configuration to adapt to the dynamic changes of the power grid;

[0084] S24. Adjust the charging station scheduling, including charging time and power, based on the output results of the long short-term memory network model.

[0085] In this implementation manner, the S22 specifically includes:

[0086] S221, the long short-term memory network model is configured to receive the power grid state data N s and renewable energy supply data R s As input;

[0087] S222, the long short-term memory network model integrates the electricity price data p and the demand response signal D from the power grid management method;

[0088] S223. The integrated data fusion process can be expressed as:

[0089] X t =f(N s ,R s ,P,D)

[0090] Among them, X t represents the comprehensive input vector of the long short-term memory network model at time t, f represents the data fusion function, N s Indicates the power grid status data, R s represents renewable energy supply data, P represents electricity price data, and D represents demand response signal;

[0091] S224, long short-term memory network model applies the comprehensive input vector X t Forecast grid load and renewable energy supply.

[0092] In this implementation manner, the S23 specifically includes:

[0093] S231. In a long short-term memory network model, a customized adaptive learning rate algorithm for grid load and renewable energy supply forecasting is used;

[0094] S232, the adaptive learning rate algorithm realizes dynamic adjustment of the learning rate through the following formula:

[0095] η t+1 =η t *g(ΔE t ,ΔE t-1 )

[0096] Among them, η t+1 and η t are the learning rates in the consecutive training stages t+1 and t, ΔE t Indicates the current forecast stage and the actual grid load, ΔE t-1is the error in the previous prediction stage, and g is an adjustment function based on the change in error;

[0097] S233. Through the learning rate adjustment mechanism, the model responds promptly to rapid changes in the grid status and renewable energy supply.

[0098] S3, using regression analysis to determine the optimal scheduling plan for electric vehicle charging demand based on the prediction results of the long short-term memory network;

[0099] In this implementation, S3 specifically includes:

[0100] S31. Determine the optimal scheduling plan for electric vehicle charging demand using a regression analysis method based on the prediction results of grid load and renewable energy supply provided by a long short-term memory network;

[0101] S32, the regression analysis includes evaluating the relationship between grid load, electricity price, renewable energy supply and electric vehicle charging demand, and calculating the optimal charging time and power;

[0102] S33. The customized regression model uses the following formula to consider the relationship between electric vehicle charging and grid load:

[0103] y=β 0 +β 1 X L +β 2 f(X E )+β 3 X P +γZ+ε

[0104] Where y represents the charging power or charging time of the charging station; X L Indicates the grid load; X E represents the renewable energy supply, f(X E ) is a nonlinear transformation of renewable energy supply, indicating its impact on charging demand; X P represents the factors that affect the charging strategy; γZ represents the specific features of the LSTM prediction results or other relevant grid information; β 0 ,β 1 ,β 2 ,β 3 and γ represent the influence of each variable on the charging strategy; ε represents the difference between the model prediction and the actual value.

[0105] S32, the regression analysis includes evaluating the relationship between grid load, electricity price, renewable energy supply and electric vehicle charging demand, and calculating the optimal charging time and power;

[0106] In this implementation manner, the S32 specifically includes:

[0107] S321, using a multivariate linear regression model to calculate the optimal charging time and power of the electric vehicle charging station, the model taking into account the grid load, electricity price, renewable energy supply and the impact on future grid load forecast;

[0108] S322, the form of the multivariate linear regression model is:

[0109] P opt =α 0 +α 1 L+α 2 E+α 3 R+α 4 F(L,R)+ε

[0110] T opt =β 0 +β 1 L+β 2 E+β 3 R+β 4 F(L,R)+ε

[0111] Among them, P opt represents the optimal charging power, T opt represents the optimal charging time, L is the grid load, E is the electricity price, R represents the renewable energy supply, and F(L,R) represents the function of grid load and renewable energy supply;

[0112] S323, by optimizing the model parameters α and β, the model predicts the optimal charging power P under specific grid conditions and resource availability. opt and charging time T opt .

[0113] S321, using a multivariate linear regression model to calculate the optimal charging time and power of the electric vehicle charging station, the model taking into account the grid load, electricity price, renewable energy supply and the impact on future grid load forecast;

[0114] In this implementation manner, the S321 specifically includes:

[0115] S3211. Analyze the data of power grid load L, electricity price E, and renewable energy supply R using a multivariate linear regression model to obtain data on optimal charging time and power of electric vehicle charging stations;

[0116] S3212, the model analyzes the relationship between charging demand and grid capacity according to the grid load L, and optimizes the operation of the charging station during the high-load period of the grid;

[0117] S3213. Considering the electricity price E, the model adjusts the charging strategy to utilize the low electricity price period to reduce the charging cost;

[0118] S3214: Input the renewable energy supply R into the model, and adjust the charging plan to maximize the use of renewable energy.

[0119] S4. Dynamically update the charging scheduling plan according to the real-time changes in the power grid and renewable energy supply, wherein the charging scheduling plan includes adjusting the charging power and time.

[0120] In this implementation, S4 specifically includes:

[0121] S41, optimal charging time T calculated according to S3 opt and power P opt Develop charging scheduling plans for electric vehicle charging stations;

[0122] S42. Implement an automatic adjustment mechanism that dynamically adjusts the charging time and power of the charging station according to real-time grid load and renewable energy supply data;

[0123] S43, adjusting includes evaluating the deviation between the current grid state and the predicted state, and optimizing the charging power and time accordingly;

[0124] S44, the adjustment mechanism uses the following formula to make real-time adjustments:

[0125] P adj =P opt *(1+δ P *ΔL)

[0126] T adj =T opt *(1+δ T *ΔE)

[0127] Among them, P adj Indicates the adjusted charging power and T adj represents time, δ P and δ T is the adjustment coefficient, ΔL is the real-time change of grid load, and ΔE is the real-time change of electricity price.

[0128] Embodiment 1:

[0129] As a densely populated metropolis with a large number of vehicles, Beijing faces increasingly severe challenges in grid management and energy utilization. The grid load reaches its peak during the peak hours of 7:00 to 9:00 in the morning and 17:00 to 20:00 in the evening on weekdays, while the grid load drops significantly during the low hours of 22:00 to 5:00 the next day. At the same time, Beijing has a large number of solar and wind power generation facilities, and solar power generation can reach a fairly high level, especially during sunny days.

[0130] The method was tested in July, a month in the summer of 2023. Through long short-term memory networks (LSTM) and regression analysis, the method collected and analyzed grid status data and renewable energy supply data in real time. For example, the data collected from grid sensors showed that the grid load during the morning peak on weekdays in July averaged 1,200 MW, while it reached 1,300 MW during the evening peak. In terms of renewable energy, the average daytime solar power generation on clear days can reach 700 MW.

[0131] This method uses regression analysis to calculate the optimal charging time and power for electric vehicles. During the evening peak period, the method recommends reducing the charging activity of the charging station, while during the daytime peak solar power generation period, the method increases the charging activity of the charging station.

[0132] Table 1 Grid load and charging cost efficiency improvement data

[0133]

[0134] Referring to Table 1 above, after one month of operation, the method successfully reduced the grid load during the peak hours of weekdays by an average of 15%, that is, the grid load during the evening peak hours dropped from 1,300 MW to 1,105 MW. At the same time, due to the optimization of charging time, the average charging cost of electric vehicle users dropped from 1.5 yuan per kilowatt-hour to 1.2 yuan, a reduction of about 20%. In addition, the utilization rate of solar power generation increased by 10%, and about 70% of the charging demand on sunny days was met by solar energy, and carbon emissions were reduced by 30% compared with traditional charging methods.

[0135] Through this intelligent and automated charging network deployment, this method not only solves the overload problem of Beijing's power grid during peak hours, but also improves the overall efficiency of power grid operation and the utilization rate of renewable energy. It provides electric vehicle users with a more economical, efficient and environmentally friendly charging solution, while providing power grid operators with a more flexible and sustainable power grid management method, demonstrating the significant effects and advantages of this method in practical applications.

[0136] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for deploying a battery charging network using renewable energy, characterized in that: The steps include: S1, collect grid status data and renewable energy supply data in real time through sensor networks; S2. Use long short-term memory networks to predict grid load and renewable energy supply based on grid status data and renewable energy supply data; S3, using regression analysis to determine the optimal scheduling plan for electric vehicle charging demand based on the prediction results of the long short-term memory network; S4. Dynamically update the charging scheduling plan according to the real-time changes in the power grid and renewable energy supply, wherein the charging scheduling plan includes adjusting the charging power and time.

2. A method for deploying a battery charging network using renewable energy according to claim 1, characterized in that: The S1 specifically includes: S11. Using a sensor network to collect grid status data and renewable energy supply data in real time, where the grid status data includes grid load, grid frequency and grid voltage, and the renewable energy supply data includes wind and solar power generation; S12. The sensor network transmits the collected data to a data processing center through an effective data transmission mechanism.

3. A method for deploying a battery charging network using renewable energy according to claim 1, characterized in that: The S2 specifically includes: S21. Constructing long short-term memory network: i t =σ(W i *[h t-1 ,x t ]+b i ) f t =σ(W f *[h t-1 ,x t ]+b f ) the t =σ(W o *[h t-1 ,x t ]+b o ) c t =f t *c t-1 +i t *tanh(W c *[h t-1 ,x t ]+b c ) Among them, the input gate i t Represents the grid load data, the forget gate f t : Indicates that outdated grid load patterns are forgotten, and the output gate is o t Predict the upcoming grid load, unit status c t Represents key information about grid load and renewable energy supply history; S22, the long short-term memory network model integrates the grid state and renewable energy supply data and the interface of the key components of the grid management method, wherein the interface of the key components of the grid management method also includes an electricity price module and a demand response controller; S23, the long short-term memory network model uses adaptive algorithm optimization to adjust the learning rate and memory unit configuration to adapt to the dynamic changes of the power grid; S24. Adjust the charging station scheduling, including charging time and power, based on the output results of the long short-term memory network model.

4. A method for deploying a battery charging network using renewable energy according to claim 1, characterized in that: The S22 specifically includes: S221, the long short-term memory network model is configured to receive the power grid state data N s and renewable energy supply data R s As input; S222, the long short-term memory network model integrates the electricity price data P and the demand response signal D from the power grid management method; S223. The integrated data fusion process can be expressed as: X t =f(N s ,R s ,P,D) Among them, X t represents the comprehensive input vector of the long short-term memory network model at time t, f represents the data fusion function, N s Indicates the power grid status data, R s represents renewable energy supply data, P represents electricity price data, and D represents demand response signal; S224, long short-term memory network model applies the comprehensive input vector X t Forecast grid load and renewable energy supply.

5. A method for deploying a battery charging network using renewable energy according to claim 1, characterized in that: The S23 specifically includes: S231. In a long short-term memory network model, a customized adaptive learning rate algorithm for grid load and renewable energy supply forecasting is used; S232, the adaptive learning rate algorithm realizes dynamic adjustment of the learning rate through the following formula: or t+1 =the t *g(ΔE t ,NO t-1 ) Among them, η t+1 and η t are the learning rates in the consecutive training stages t+1 and t, ΔE t Indicates the current forecast stage and the actual grid load, ΔE t-1 is the error in the previous prediction stage, and g is an adjustment function based on the change in error; S233. Through the learning rate adjustment mechanism, the model responds promptly to rapid changes in the grid status and renewable energy supply.

6. A method for deploying a battery charging network using renewable energy according to claim 1, characterized in that: The S3 specifically includes: S31. Determine the optimal scheduling plan for electric vehicle charging demand using a regression analysis method based on the prediction results of grid load and renewable energy supply provided by a long short-term memory network; S32, the regression analysis includes evaluating the relationship between grid load, electricity price, renewable energy supply and electric vehicle charging demand, and calculating the optimal charging time and power; S33. The customized regression model uses the following formula to consider the relationship between electric vehicle charging and grid load: y=β0+β1X L +β2f(X E )+β3X P +γZ+v Where y represents the charging power or charging time of the charging station; X L Indicates the grid load; X E represents the renewable energy supply, f(X E ) is a nonlinear transformation of renewable energy supply, indicating its impact on charging demand; X P represents the factors affecting the charging strategy; γZ represents the specific characteristics of the LSTM prediction results or other relevant power grid information; β0, β1, β2, β3 and γ represent the influence of each variable on the charging strategy; ε represents the difference between the model prediction and the actual value.

7. A method for deploying a battery charging network using renewable energy according to claim 1, characterized in that: The S32 specifically includes: S321, using a multivariate linear regression model to calculate the optimal charging time and power of the electric vehicle charging station, the model taking into account the grid load, electricity price, renewable energy supply and the impact on future grid load forecast; S322, the form of the multivariate linear regression model is: P opt =α0+α1L+α2E+α3R+α4F(L,R)+ε T opt =β0+β1L+β2E+β3R+β4F(L,R)+ε Among them, P opt represents the optimal charging power, T opt represents the optimal charging time, L is the grid load, E is the electricity price, R represents the renewable energy supply, and F(L, R) represents the function of grid load and renewable energy supply; S323, by optimizing the model parameters α and β, the model predicts the optimal charging power P under specific grid conditions and resource availability. opt and charging time T opt .

8. A method for deploying a battery charging network using renewable energy according to claim 1, characterized in that: The S321 specifically includes: S3211. Analyze the data of power grid load L, electricity price E, and renewable energy supply R using a multivariate linear regression model to obtain data on optimal charging time and power of electric vehicle charging stations; S3212, the model analyzes the relationship between charging demand and grid capacity according to the grid load L, and optimizes the operation of the charging station during the high-load period of the grid; S3213. Considering the electricity price E, the model adjusts the charging strategy to utilize the low electricity price period to reduce the charging cost; S3214: Input the renewable energy supply R into the model, and adjust the charging plan to maximize the use of renewable energy.

9. A method for deploying a battery charging network using renewable energy according to claim 1, characterized in that: The S4 specifically includes: S41, optimal charging time T calculated according to S3 opt and power P opt Develop charging scheduling plans for electric vehicle charging stations; S42. Implement an automatic adjustment mechanism that dynamically adjusts the charging time and power of the charging station according to real-time grid load and renewable energy supply data; S43, adjusting includes evaluating the deviation between the current grid state and the predicted state, and optimizing the charging power and time accordingly; S44, the adjustment mechanism uses the following formula to make real-time adjustments: P adj =P opt *(1+δ P *ΔL) T adj =T opt *(1+δ T *ΔE) Among them, P adj Indicates the adjusted charging power and T adj represents time, δ P and δ T is the adjustment coefficient, ΔL is the real-time change of grid load, and ΔE is the real-time change of electricity price.

Images