New energy hybrid power supply method and system based on energy unloading and storage medium

Abstract

The application discloses a new energy hybrid power supply method and system based on energy unloading, and a storage medium, and belongs to the field of new energy power supply planning. The method comprises the following steps: constructing a new energy hybrid power supply system model, including a plurality of cellular cells, each cellular cell comprising a base station, a plurality of photovoltaic components, and energy storage components connected with the photovoltaic components; constructing constraint conditions of the new energy hybrid power supply system model; constructing an objective function of the new energy hybrid power supply system model, wherein the objective function is to maximize the investment return rate and carbon efficiency; solving the new energy hybrid power supply system model to obtain the unloading path of the energy of the new energy hybrid power supply system at the current time, the deployment quantity of the base station, and the deployment parameters of the photovoltaic components. The deployment parameters and the energy unloading path of the new energy hybrid power supply system are dynamically adjusted, and the maximization of the system performance is realized.

Description

Technical Field

[0001] This invention belongs to the field of new energy power supply planning, and more specifically, relates to a new energy hybrid power supply method, system and storage medium based on energy offloading. Background Technology

[0002] In today's society, wireless cellular network architecture based on new energy sources has attracted much attention. The organic combination of new energy sources and wireless cellular networks constitutes a new energy hybrid power supply system. However, how to improve the performance of new energy hybrid power supply systems remains a challenge to be solved, including but not limited to improving system latency, energy efficiency, frequency efficiency, or total capacity.

[0003] Most existing technologies achieve a balance between renewable energy and wireless cellular networks by optimizing system configuration to maximize wireless network coverage and minimize energy consumption. For example, this involves adjusting the deployment location and number of wireless base stations, combined with the characteristics of the renewable energy system, to achieve optimal performance. Alternatively, resource scheduling within the wireless network system can improve the performance of the renewable energy hybrid power supply system. For instance, energy management models based on base station power consumption control model user services according to queuing theory and adjust base station transmission power in real time based on service traffic to improve system energy efficiency. Another approach is access methods based on received signal strength, where user terminals collect signal strength data from various base stations and select the appropriate base station to access by comparing the link quality of each wireless access network, thereby improving system energy efficiency.

[0004] These methods can all improve the performance of new energy hybrid power supply systems to some extent. However, these methods improve system performance by adjusting specific indicators and cannot reflect the impact of the time characteristics of new energy hybrid power supply systems on system performance, thus failing to maximize system performance. Summary of the Invention

[0005] To address the shortcomings and improvement needs of existing technologies, this invention provides a new energy hybrid power supply method, system, and storage medium based on energy offloading. Its purpose is to maximize system performance by dynamically adjusting the deployment parameters and energy offloading path of the new energy hybrid power supply system.

[0006] To achieve the above objectives, according to one aspect of the present invention, a new energy hybrid power supply method based on energy offloading is provided, comprising: A new energy hybrid power supply system model is constructed, including multiple cellular cells. Each cellular cell includes a base station, multiple photovoltaic modules, and an energy storage module connected to the photovoltaic modules. The energy storage module is used to offload excess energy to base stations in other cellular cells when the power generation of the photovoltaic modules exceeds the demand of the base station in the current cellular cell. The constraints for constructing the new energy hybrid power supply system model are used to constrain the capacity, charging and discharging state and charging and discharging power of the energy storage components, as well as the number of users served by each base station and the number of photovoltaic modules deployed at each base station. The objective function for constructing a new energy hybrid power supply system model is to maximize the return on investment and carbon efficiency. Solve the model of the new energy hybrid power supply system to obtain the offloading path of excess energy in the new energy hybrid power supply system at the current moment, the number of base stations deployed, and the deployment parameters of photovoltaic modules; Solving the model of a hybrid power supply system for new energy sources includes: S1. Constructing the state space of a new energy hybrid power supply system ,include: Return on investment at any time , Carbon efficiency at any time , The remaining capacity of the energy storage components at any time and Coverage rate of base stations at any given time; S2. Setting the operational space for the new energy hybrid power supply system, including: The unloading path at any given time, the number of base stations deployed, and the deployment parameters of the photovoltaic modules; S3. Based on the state corresponding to each action in the action space, construct the TD3 deep reinforcement learning network. Reward function at time step :

[0007]

[0008] in, , They represent Carbon efficiency at any time and Carbon efficiency at any given time. This indicates the set threshold. This represents the average rate of return on investment over the previous t time points; S4. Optimize the reward function through reverse iteration to maximize the reward function.

[0009] Furthermore, the energy storage component in Remaining capacity at time step for:

[0010] in, This indicates the self-discharge rate of the energy storage component when it is inactive. and These represent the energy exchange efficiency of the energy storage module during charging and discharging, respectively. This indicates the discharge power of the energy storage component; Indicates the first residential area The charging power of the energy storage component at any time is determined by the first The power generated by the photovoltaic modules in the community and the discharge power transferred from the energy storage modules in other communities; Indicates a unit time interval.

[0011] Furthermore, the first residential area Charging power of the instantaneous energy storage component for:

[0012] in, This indicates the total number of base stations deployed in the new energy hybrid power supply system. Indicates the first Community and the Distance between the residential areas The path loss factor represents energy transfer. The state function represents two states, 0 and 1. When, it indicates the first The community does not provide assistance to the first Energy transfer in the community, when When, it indicates the first The community to the first Energy transferred from the community; Indicates the first Base stations within the community Output power at any moment Indicates the first The power generation capacity of the photovoltaic modules in the community.

[0013] Furthermore, the objective function of the new energy hybrid power supply system model is:

[0014] ROI represents the rate of return on investment. For carbon efficiency, the expressions are as follows:

[0015]

[0016] In the formula, This indicates the number of days the new energy hybrid power supply system operates each year. Indicates the number of hours the device runs per day. This indicates the system's service life in years; Power supplied to the municipal power grid; This represents the carbon efficiency factor of renewable energy power supply in the i-th community. This represents the carbon efficiency factor of the mains power supply in the i-th cell; The unit price of mains electricity; The unit price of the carbon tax; This indicates the number of photovoltaic modules equipped in each base station. , and These represent the net present value costs of the base station, photovoltaic modules, and energy storage modules within each cell, respectively. This indicates the system's communication capacity within one day.

[0017] Furthermore, S4 includes: S41. The reward function is trained using the temporal difference algorithm. In each iteration, the two time steps of the TD3 deep reinforcement learning network at the current time step are calculated. Values ​​to predict the two values ​​at the next moment. Value, where each The value includes the current state and the corresponding uninstallation path. S42, predict the two next moments. Substitute the value into the performance evaluation function In the process, the corresponding evaluation value is calculated, and the parameters of the reward function are updated along the unloading path with the larger evaluation value; the evaluation function for:

[0018] in, As a discount factor, It is a constant. This represents the set of actions. , These represent the two predicted next time steps. value; S43. When the set number of iterations is reached or the gradient of the reward function reaches the set threshold, training is stopped, and the unloading path, the number of base stations deployed, and the deployment parameters of photovoltaic modules that maximize both return on investment and carbon efficiency at the current moment are obtained.

[0019] Furthermore, the capacity constraint of the energy storage component is:

[0020] in, Indicates that the energy storage components are in Remaining capacity at any given time Indicates the rated capacity of the energy storage component; The charging and discharging state constraints are as follows:

[0021]

[0022] in, This indicates the current charge / discharge state of the energy storage components. This indicates the discharge power of the energy storage component. , These represent the minimum and maximum discharge states of the energy storage component, respectively. The charging and discharging power constraint is:

[0023] in, Indicates the first residential area The charging power of the energy storage components at all times. This indicates the maximum rated power for charging the energy storage component.

[0024] Furthermore, the constraint on the number of users served by each base station is as follows:

[0025] in, This represents the number of users in the i-th cell. This indicates the maximum number of users that each cell base station can serve. This indicates the total number of users in a new energy hybrid power supply system. This represents the total number of base stations deployed in the system; The constraint on the number of photovoltaic modules deployed in each base station is as follows:

[0026] in, This indicates the number of photovoltaic modules deployed at each base station. This indicates the maximum number of photovoltaic modules deployed at each base station.

[0027] According to another aspect of the present invention, a new energy hybrid power supply system based on energy offloading is provided for performing the new energy hybrid power supply method based on energy offloading as described in any of the first aspects, wherein the new energy hybrid power supply system based on energy offloading includes: The model building module is used to build a new energy hybrid power supply system model, including multiple cells. Each cell includes a base station, multiple photovoltaic modules, and an energy storage module connected to the photovoltaic modules. The energy storage module is used to offload excess energy to base stations in other cells when the power generation of the photovoltaic modules exceeds the demand of the base station in the current cell. The constraint construction module is used to construct the constraint conditions of the new energy hybrid power supply system model, and to constrain the capacity, charging and discharging state and charging and discharging power of the energy storage components, as well as the number of users served by each base station and the number of photovoltaic modules deployed at each base station. The objective function construction module is used to construct the objective function of the new energy hybrid power supply system model, wherein the objective function is to maximize the return on investment and carbon efficiency. The model solving module is used to solve the model of the new energy hybrid power supply system, and obtain the unloading path of excess energy in the new energy hybrid power supply system at the current moment, the number of base stations deployed, and the deployment parameters of photovoltaic modules.

[0028] According to another aspect of the present invention, a computer-readable storage medium is provided, including a stored computer program; when the computer program is executed by a processor, it controls the device where the computer-readable storage medium is located to perform the new energy hybrid power supply method based on energy offloading as described in any one aspect.

[0029] In summary, the above-described technical solutions conceived in this invention can achieve the following beneficial effects: (1) The new energy hybrid power supply system model constructed in this invention, when the power generation of the photovoltaic modules in the current cell exceeds the demand of the base station in the current cell, unloads the excess energy to the base station of other cells through the energy storage modules connected to the photovoltaic modules. The model also constructs the capacity constraints, charge and discharge state constraints, charge and discharge power constraints, the number of users served by the base station, and the number of photovoltaic modules deployed in each base station, so that the energy storage modules maintain charge and discharge balance during the energy unloading process. With the goal of maximizing the return on investment and carbon efficiency, the model is solved to obtain the optimal unloading path for the excess energy of the photovoltaic modules in the current cell to be unloaded to the base station of other cells, as well as the number of base stations and the deployment parameters of the photovoltaic modules. The method of this invention, based on the optimal number of base stations and the deployment parameters of the photovoltaic modules, dynamically adjusts the optimal unloading path for the excess energy at the current moment, avoids energy waste, reduces the power supply from the grid to the base station, and maximizes the performance of the system.

[0030] (2) Furthermore, this invention constructs an energy storage module based on the characteristics of the energy storage module, its charging and discharging characteristics, and the output characteristics and trends of the photovoltaic module. The remaining capacity model at time step, and the... residential area The charging power of the energy storage components at all times fully reflects the coupling relationship between various physical quantities during the energy unloading process, making the constructed new energy hybrid power supply system model accurate and facilitating the maximization of system performance.

[0031] (3) Furthermore, the objective function constructed in this invention establishes the coupling relationship between parameters of energy storage components, base stations and photovoltaic components, and fully considers the comprehensive impact of multiple factors such as the cost of excess energy generated by photovoltaic components during the energy unloading process, carbon tax, and depreciation rates of base stations, energy storage components and photovoltaic components on the performance of the new energy hybrid power supply system; and, with return on investment and carbon efficiency as the main influencing factors of the objective function, the system's environmental impact and the degree of return on funds are also considered. The organic combination of these two indicators can better help decision-makers plan and make decisions based on the sustainability and economic benefits of the system.

[0032] (4) Furthermore, in solving the new energy hybrid power supply system model, the constructed reward function takes the objective function as a reference, comprehensively considering carbon efficiency and return on investment, through... Time and The difference in carbon efficiency at each time step is used to construct a reward for carbon efficiency. The reward for return on investment is designed based on deployment parameters. By averaging, the fluctuation of the reward function is reduced, avoiding the problem of local convergence of the optimal solution in TD3 deep reinforcement learning.

[0033] (5) Furthermore, when optimizing and updating the reward function, compared with the existing TD3 algorithm which directly uses the smaller of the two Q values ​​as the evaluation value, the performance evaluation function designed in this invention... It takes into account the impact of two Q values ​​on the reward function, avoiding underestimation bias when the two Q values ​​differ significantly, resulting in high exploration efficiency and reducing the number of model iterations.

[0034] (6) Furthermore, the constraints of the new energy hybrid power supply system model constructed by the present invention can ensure the charging and discharging balance of the energy storage components during the energy unloading process, prevent the voltage of the energy storage components from being too high or too low, improve the charging efficiency, ensure the rationality of energy transfer between the power generation of the photovoltaic components in the current community and the power transferred from other communities during the energy unloading process of the energy storage components; and ensure that each user can obtain sufficient bandwidth while also enabling all users in the system to receive services. Attached Figure Description

[0035] Figure 1 A schematic diagram of the new energy hybrid power supply method based on energy offloading provided by the present invention.

[0036] Figure 2 The graph shows the reward variation curves for system optimization between the ATD3 method provided in this embodiment of the invention and existing methods.

[0037] Figure 3 This is a comparison chart showing the return on investment of the ATD3 method provided in this embodiment of the invention and existing methods for optimizing the system under different base station coverage probabilities.

[0038] Figure 4 This is a comparison chart showing the carbon efficiency of the ATD3 method provided in this embodiment of the invention and existing methods optimized for different base station coverage probabilities in the system. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0040] like Figure 1 As shown, this invention provides a new energy hybrid power supply method based on energy offloading, which mainly includes the following steps: A new energy hybrid power supply system model is constructed. The model includes multiple cells, each cell including a base station, multiple photovoltaic modules, and energy storage modules connected to the photovoltaic modules. The energy storage modules are used to offload excess energy to base stations in other cells when the power generation of the photovoltaic modules in the current cell exceeds the demand of the base station in the current cell. The constraints for constructing the new energy hybrid power supply system model are used to constrain the capacity, charging and discharging state, and charging and discharging power of the energy storage components, the number of users served by the base station, and the number of photovoltaic components deployed at each base station, so as to achieve a balance in the charging and discharging of the energy storage components. The objective function for constructing a new energy hybrid power supply system model is to maximize the return on investment and carbon efficiency. Solve the model of the new energy hybrid power supply system to obtain the offloading path of excess energy, the number of base stations deployed, and the deployment parameters of photovoltaic modules at the current moment. The offloading path is used to indicate which cell's base station the excess energy is offloaded to. The deployment parameters of photovoltaic modules include the orientation angle, tilt angle, and number of photovoltaic modules.

[0041] Specifically, in the constructed new energy hybrid power supply system model, the power consumption transfer state of the new energy hybrid power supply system changes when the energy storage component is connected to the photovoltaic component. In this invention, based on the characteristics and charging / discharging patterns of the energy storage component, and the output characteristics and trends of the photovoltaic component, a model is constructed... Remaining capacity of the energy storage component Model:

[0042] In the formula, This represents a unit time interval in a discrete time series. This indicates the self-discharge rate of the energy storage component when it is inactive. and These represent the energy exchange efficiency of the energy storage component during charging and discharging, respectively. This indicates the discharge power of the energy storage component, which is a fixed value set in watts per hour. Indicates the first residential area The charging power of the energy storage component at any given time, among which, From the current community (the first) Power generation of photovoltaic modules in the community The discharge power transferred from energy storage components in other communities composition.

[0043] Specifically, the constructed for:

[0044] in, Indicates the first Community and the Distance between the residential areas The path loss factor represents energy transfer. The state function represents two states, 0 and 1. When, it indicates the first The community does not provide assistance to the first Energy transfer in the community, when When, it means the first The community to the first Energy transferred from the community; Indicates the first One base station Output power at any moment This indicates the power generation capacity of the photovoltaic modules in the current community.

[0045] This represents the total number of base stations deployed in a new energy hybrid power supply system. In this system, base stations and users are distributed in a two-dimensional space. Each follows an independent Poisson point process. and The deployment densities are respectively and The expression is as follows:

[0046]

[0047] in, This indicates the total number of base stations deployed in the new energy hybrid power supply system, which varies depending on the deployment situation. This represents the total number of users in the system, which is a fixed value. This represents the Poisson distribution of the i-th base station. Let represent the Poisson distribution of the i-th user.

[0048] Specifically, the constraints of the constructed new energy hybrid power supply system model include: capacity constraints of energy storage components, charging and discharging state constraints of energy storage components, charging and discharging power constraints of energy storage components, number of users served by base stations, and number of photovoltaic modules deployed at each base station.

[0049] The capacity constraints of the energy storage components are:

[0050] in, express The remaining capacity of the energy storage components at all times. This indicates the rated capacity of the energy storage component.

[0051] The charge / discharge state constraints of energy storage modules can prevent the voltage of the energy storage modules from being too high or too low, thereby improving charging efficiency. The charge / discharge state constraints are as follows:

[0052]

[0053] in, The current state of charge / discharge of an energy storage module refers to the ratio between the currently available electrical energy and the total electrical energy. Remaining capacity of the energy storage component With the rated capacity of energy storage components The ratio reflects the current energy storage status of the energy storage component. , These represent the minimum and maximum discharge states of the energy storage component, respectively.

[0054] To ensure the rationality of energy transfer between the photovoltaic power generation of the current community and the power transferred from other communities during the energy unloading process of the energy storage components, the charging and discharging power constraints of the energy storage components are as follows:

[0055] in, This indicates the maximum rated power for charging the energy storage component.

[0056] To ensure the average data reception rate for each user and prevent users from clustering under the same base station, a constraint is placed on the number of users served by each base station. This constraint ensures that each user receives sufficient bandwidth and that all users in the system can receive service. The constraint on the number of users served by each base station is as follows:

[0057] in, This represents the number of users in the i-th cell. This indicates the maximum number of users that each cell base station can serve. This represents the total number of users in the system. This indicates the total number of base stations deployed in the new energy hybrid power supply system.

[0058] The number of photovoltaic modules deployed at each base station is constrained to avoid excessive deployment that could lead to unstable power supply or exceed the base station's carrying capacity. The constraints are as follows:

[0059] in, This indicates the number of photovoltaic modules deployed at each base station. This indicates the maximum number of photovoltaic modules deployed at each base station.

[0060] Specifically, the objective function for constructing the new energy hybrid power supply system model is to maximize the return on investment and carbon efficiency, including: Using average annual profit as the return on investment, and amortizing the investment cost across... In the middle of the year, the annual return on investment (ROI) is calculated. In this invention, based on the constructed energy storage components, the cost of the portion of photovoltaic module power generation exceeding its own demand within a set time period is calculated, along with the corresponding carbon tax; in this embodiment of the invention, the set time period is one year. Based on the depreciation rates of the base station, energy storage components, and photovoltaic modules, the cost of deploying a new energy hybrid power supply system is calculated, thereby obtaining the current ROI.

[0061] In the formula, This indicates the number of days the new energy hybrid power supply system operates each year. Indicates the number of hours the machine runs per day. Indicates the system's service life in years; Indicates the first One base station Output power at any moment Power supplied to the municipal power grid; This represents the carbon efficiency factor of renewable energy power supply in the i-th community. This represents the carbon efficiency factor of the mains power supply in the i-th cell; The unit price of mains electricity; This is the unit price of carbon tax, used to measure the carbon tax payable per mg of carbon dioxide emissions, expressed in ¥ / mg. This indicates the number of photovoltaic modules equipped in each base station. , and These represent the net present value costs of the base station, photovoltaic modules, and energy storage modules within each cell, respectively.

[0062] in, , and They are respectively: , Pick , , ; In the formula, For purchase cost, The depreciation factor is expressed as follows:

[0063] in, This is the depreciation factor.

[0064] Municipal power supply for:

[0065] In the formula, the power generation of the photovoltaic modules in the current community is... for:

[0066] in, , , , These represent the amplitude, mean, standard deviation, and skewness parameters of a skewed normal distribution, respectively. The density function representing the standard normal distribution. It is the distribution function of the standard normal distribution; This represents the actual area of ​​the photovoltaic module that receives sunlight. The conversion efficiency of photovoltaic modules for power generation. The comprehensive coefficient is a correction coefficient value derived from the characteristics of photovoltaic modules, including: the light utilization rate of photovoltaic modules, surface contamination rate, integrated circuit loss rate, etc.

[0067] In this invention, the carbon efficiency of a wireless cellular network system based on hybrid power supply from new energy sources is discussed. for:

[0068] In the formula, This indicates the communication capacity of the new energy hybrid power supply system within one day. This represents the carbon efficiency factor of renewable energy power supply in the i-th community. This represents the carbon efficiency factor of the mains power supply in the i-th cell; Indicates the first One base station Output power at any moment Power supplied to the municipal power grid; It represents the unit time interval in a discrete time series.

[0069] Among them, in the new energy hybrid power supply system, for The capacity of a single base station at any given time is:

[0070] In the formula, This represents the total bandwidth of a single base station. Indicates the first The number of users connected to each base station Indicates the first The first base station The signal-to-interference-plus-noise ratio (SIR) of a user at time t, and the user access logic are as follows: Users select the optimal base station to access the wireless cellular network system. That is, users in the wireless cellular network system select the optimal base station for access based on the principle of maximum signal-to-interference-plus-noise ratio (SINR) to obtain the best signal quality, thereby maximizing the system's capacity. Among these, the first... The first base station The signal-to-interference-plus-noise ratio (SIR) of a user at time t for:

[0071] in, This represents the channel power gain. This is the path loss index. The variance is expressed as additive white Gaussian noise. This indicates the channel fading of the j-th user. Indicates the first The base station and the first Distance between users This represents the Poisson point process.

[0072] Therefore, the daily communication capacity of the new energy hybrid power supply system for:

[0073] Therefore, the objective function of the new energy hybrid power supply system model constructed in this invention is:

[0074] Specifically, in this embodiment of the invention, a hybrid energy supply system model is solved using a deep reinforcement learning method. Specifically, an improved TD3 deep reinforcement learning network (denoted as ATD3) is used for the solution, including: S1. Construct the state space of the new energy hybrid power supply system at the current moment. ,include: Return on investment at any time , Carbon efficiency at any time , Remaining capacity of the energy storage component and Coverage of time base stations In reinforcement learning, a state is the result of an agent's interaction with its environment. The state space is a set of various states, representing changes caused by different environments. The state space constructed in this invention not only considers the system's economic benefits and environmental protection level, but also fully takes into account factors such as the power capacity of energy storage components and the service quality of base stations, thus providing a more comprehensive description of the operation of the new energy hybrid power supply system. In the formula, , express Real-time energy storage components The remaining battery power, Take 1- ; , express Time users The coverage probability, Take 1- .

[0075] S2. Set the action space of the new energy hybrid power supply system at the current moment, including: the path at time t to offload excess energy from the photovoltaic modules in the current cell to the base stations of other cells. (i.e., unloading path), number of base stations deployed And the deployment parameters of photovoltaic modules; the action space is used to reflect the set of all possible actions that an agent can take when interacting with the environment; for the multi-objective optimization problem of this invention, the selected action space parameters can have a significant impact on both return on investment and carbon efficiency; in the embodiment of this invention, the action space at time t Represented as: ,in, , , Indicates energy storage components To energy storage components The transferred energy.

[0076] S3. Based on the state corresponding to each action in the action space, construct the reward function for the TD3 deep reinforcement learning network. ;in, , , They represent Carbon efficiency at any time and Carbon efficiency at any given time; The expression is:

[0077] In the formula, This indicates the set threshold, which is selected based on experience; Let $t$ be the average rate of return on investment over the first $t$ time points, expressed as:

[0078] in, This represents the rate of return on investment at time t.

[0079] The reward function designed in this invention comprehensively considers carbon efficiency and return on investment, taking into account that decisions at each time step have a greater impact on carbon efficiency and that the return on investment is related to the system's deployment parameters. Time and The difference in carbon efficiency at each moment is used to construct a reward for carbon efficiency. The reward for return on investment is designed based on deployment parameters. By averaging, the fluctuation of the reward function is reduced, and the local convergence of the optimal solution obtained by TD3 deep reinforcement learning is avoided.

[0080] S4. Optimize the reward function through reverse iteration to maximize the reward function, including: S41. The temporal difference algorithm is used to train the reward function. In each iteration, the two time steps of the TD3 deep reinforcement learning network at the current time step are calculated. value: , And predict the two next moments. value: , ; among them, each The value includes the current state and the corresponding uninstallation path. S42, predict the two next moments. value: , Substitute into the performance evaluation function of the design In the process, the corresponding evaluation value is calculated, and the parameters of the reward function are updated along the unloading path with the larger evaluation value; where, the evaluation function... for:

[0081] in, Indicates the first Momentary rewards As a discount factor, It is a constant, chosen based on experience; This represents the set of action spaces.

[0082] S43. When the set number of iterations is reached or the gradient of the reward function reaches the set threshold, training is stopped, and the offloading path, the number of base stations deployed, and the deployment parameters of photovoltaic modules that maximize both return on investment and carbon efficiency are obtained.

[0083] Compared to the existing TD3 algorithm, which directly uses the smaller of two Q values ​​as the evaluation value and may suffer from underestimation bias, the performance evaluation function designed in this invention... It takes into account the impact of two Q values ​​on the reward function, avoiding underestimation bias when the two Q values ​​differ significantly, resulting in high exploration efficiency and reducing the number of model iterations.

[0084] To further verify the ATD3 method of the present invention and compare it with the existing TD3 and DDPG methods, such as Figure 2 As shown, the DDPG method is prone to getting stuck in local optima and is difficult to obtain higher rewards. Its reward function fluctuates and is significantly lower than the other two methods. The TD3 method tends to underestimate the Q value. The ATD3 method of this invention is superior to the DDPG and TD3 methods in terms of learning efficiency.

[0085] like Figure 3 As shown, this paper presents the optimization results of the ATD3 method of this invention, compared with existing TD3 and DDPG methods, for the return on investment of a new energy hybrid power supply system under different base station coverage rates. From... Figure 3 As can be seen from the above, the ATD3 method of the present invention has the highest return on investment under different coverage probability conditions compared with the other two methods, with the improved return on investment ranging from 2% to 10%.

[0086] like Figure 4 As shown, this paper presents the optimization results of the ATD3 method of this invention, compared with existing TD3 and DDPG methods, for the carbon efficiency of a new energy hybrid power supply system under different base station coverage probabilities. From... Figure 4 As can be seen from the above, the ATD3 method of the present invention has the highest carbon efficiency under different coverage probability conditions compared with the other two methods, and the improved return on investment is between 5% and 20%.

[0087] According to another aspect of the present invention, a new energy hybrid power supply system based on energy offloading is provided, comprising: The model building module is used to build a new energy hybrid power supply system model, including multiple cells. Each cell includes a base station, multiple photovoltaic modules, and energy storage modules connected to the photovoltaic modules. The energy storage modules are used to offload excess energy to base stations in other cells when the power generation of the photovoltaic modules exceeds the demand of the base station in the current cell. The constraint construction module is used to construct the constraint conditions for the new energy hybrid power supply system model. It is used to constrain the capacity, charging and discharging state and charging and discharging power of the energy storage components, as well as the number of users served by each base station and the number of photovoltaic modules deployed at each base station. The objective function construction module is used to construct the objective function of the new energy hybrid power supply system model. The objective function is to maximize the return on investment and carbon efficiency. The model solving module is used to solve the model of the new energy hybrid power supply system, and obtain the energy unloading path, the number of base stations deployed, and the deployment parameters of photovoltaic modules at the current moment.

[0088] Each module is used to execute the corresponding steps in the new energy hybrid power supply method based on energy offloading in the above embodiments.

[0089] The present invention also provides a computer-readable storage medium including a stored computer program; when the computer program is executed by a processor, it controls the device where the computer-readable storage medium is located to perform each step of the new energy hybrid power supply method based on energy offloading in the above embodiments.

[0090] Those skilled in the art will readily understand that the above description is merely 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 within the scope of protection of the present invention.

Claims

1. A new energy hybrid power supply method based on energy offloading, characterized in that, include: A new energy hybrid power supply system model is constructed, including multiple cellular cells. Each cellular cell includes a base station, multiple photovoltaic modules, and an energy storage module connected to the photovoltaic modules. The energy storage module is used to offload excess energy to base stations in other cellular cells when the power generation of the photovoltaic modules exceeds the demand of the base station in the current cellular cell. The constraints for constructing the new energy hybrid power supply system model are used to constrain the capacity, charging and discharging state and charging and discharging power of the energy storage components, as well as the number of users served by each base station and the number of photovoltaic modules deployed at each base station. The objective function for constructing a new energy hybrid power supply system model is to maximize the return on investment and carbon efficiency. Solve the model of the new energy hybrid power supply system to obtain the offloading path of excess energy in the new energy hybrid power supply system at the current moment, the number of base stations deployed, and the deployment parameters of photovoltaic modules; Solving the model of a hybrid power supply system for new energy sources includes: S1. Constructing the state space of a new energy hybrid power supply system ,include: Return on investment at any time , Carbon efficiency at any time , The remaining capacity of the energy storage components at any time and The coverage rate of base stations at any given time; S2. Setting the operational space for the new energy hybrid power supply system, including: The unloading path at any given time, the number of base stations deployed, and the deployment parameters of the photovoltaic modules; S3. Based on the state corresponding to each action in the action space, construct the TD3 deep reinforcement learning network. Reward function at time step : in, , They represent Carbon efficiency at any time and Carbon efficiency at any given time This indicates the set threshold. This represents the average rate of return on investment over the previous t time points; S4. Optimize the reward function through reverse iteration to maximize the reward function.

2. The new energy hybrid power supply method based on energy offloading according to claim 1, characterized in that, The energy storage component is Remaining capacity at time step for: in, This indicates the self-discharge rate of the energy storage component when it is inactive. and These represent the energy exchange efficiency of the energy storage module during charging and discharging, respectively. This indicates the discharge power of the energy storage component; Indicates the first residential area The charging power of the energy storage component at any time is determined by the first... The power generated by the photovoltaic modules in the community and the discharge power transferred from the energy storage modules in other communities; Indicates a unit time interval.

3. The new energy hybrid power supply method based on energy offloading according to claim 2, characterized in that, The first residential area Charging power of the instantaneous energy storage component for: in, This indicates the total number of base stations deployed in the new energy hybrid power supply system. Indicates the first Community and the Distance between the residential areas The path loss factor represents energy transfer. The state function represents two states, 0 and 1. When, it indicates the first The community does not provide assistance to the first Energy transfer in the community, when When, it indicates the first The community to the first Energy transferred from the community; Indicates the first Base stations within the community Output power at any moment Indicates the first The power generation capacity of the photovoltaic modules in the community.

4. The new energy hybrid power supply method based on energy offloading according to claim 3, characterized in that, The objective function of the new energy hybrid power supply system model is: ROI represents the rate of return on investment. For carbon efficiency, the expressions are as follows: In the formula, This indicates the number of days the new energy hybrid power supply system operates each year. Indicates the number of hours the device runs per day. This indicates the system's service life in years; Power supplied to the municipal power grid; This represents the carbon efficiency factor of renewable energy power supply in the i-th community. This represents the carbon efficiency factor of the mains power supply in the i-th cell; The unit price of mains electricity; The unit price of the carbon tax; This indicates the number of photovoltaic modules equipped in each base station. , and These represent the net present value costs of the base station, photovoltaic modules, and energy storage modules within each cell, respectively. This indicates the system's communication capacity within one day.

5. The new energy hybrid power supply method based on energy offloading according to claim 1, characterized in that, S4 includes: S41. The reward function is trained using the temporal difference algorithm. During each iteration, the two time-series differences of the TD3 deep reinforcement learning network are calculated. Values ​​to predict the two values ​​at the next moment. Value, where each The value includes the current state and the corresponding uninstallation path. S42, predict the two next moments. Substitute the value into the performance evaluation function In the process, the corresponding evaluation value is calculated, and the parameters of the reward function are updated along the unloading path with the larger evaluation value; the evaluation function for: in, As a discount factor, It is a constant. This represents the set of actions in the space. , These represent the two predicted next time steps. value; S43. When the set number of iterations is reached or the gradient of the reward function reaches the set threshold, training is stopped, and the unloading path, the number of base stations deployed, and the deployment parameters of photovoltaic modules that maximize both return on investment and carbon efficiency at the current moment are obtained.

6. The new energy hybrid power supply method based on energy offloading according to claim 1, characterized in that, The capacity constraint of the energy storage component is: in, Indicates that the energy storage components are in Remaining capacity at any given time Indicates the rated capacity of the energy storage component; The charging and discharging state constraints are as follows: in, This indicates the current charge / discharge state of the energy storage components. This indicates the discharge power of the energy storage component. , These represent the minimum and maximum discharge states of the energy storage component, respectively. The charging and discharging power constraint is: in, Indicates the first residential area The charging power of the energy storage components at all times. This indicates the maximum rated power for charging the energy storage component.

7. The new energy hybrid power supply method based on energy offloading according to claim 6, characterized in that, The constraint on the number of users served by each base station is as follows: in, This represents the number of users in the i-th cell. This indicates the maximum number of users that each cell base station can serve. This indicates the total number of users in a new energy hybrid power supply system. This represents the total number of base stations deployed in the system; The constraint on the number of photovoltaic modules deployed in each base station is as follows: in, This indicates the number of photovoltaic modules deployed at each base station. This indicates the maximum number of photovoltaic modules deployed at each base station.

8. A new energy hybrid power supply system based on energy offloading, characterized in that, For executing the new energy hybrid power supply method based on energy offloading as described in any one of claims 1-7, the new energy hybrid power supply system based on energy offloading comprises: The model building module is used to build a new energy hybrid power supply system model, including multiple cells. Each cell includes a base station, multiple photovoltaic modules, and an energy storage module connected to the photovoltaic modules. The energy storage module is used to offload excess energy to base stations in other cells when the power generation of the photovoltaic modules exceeds the demand of the base station in the current cell. The constraint construction module is used to construct the constraint conditions of the new energy hybrid power supply system model, and to constrain the capacity, charging and discharging state and charging and discharging power of the energy storage components, as well as the number of users served by each base station and the number of photovoltaic modules deployed at each base station. The objective function construction module is used to construct the objective function of the new energy hybrid power supply system model, wherein the objective function is to maximize the return on investment and carbon efficiency. The model solving module is used to solve the model of the new energy hybrid power supply system, and obtain the unloading path of excess energy in the new energy hybrid power supply system at the current moment, the number of base stations deployed, and the deployment parameters of photovoltaic modules.

9. A computer-readable storage medium, characterized in that, Includes a stored computer program; when executed by a processor, the computer program controls the device containing the computer-readable storage medium to perform the functions of claim 1. The new energy hybrid power supply method based on energy offloading as described in any one of the 7 claims.

Images