Multi-energy complementary zero-carbon toll station construction process optimization method based on energy distribution

Abstract

The invention discloses a multi-energy complementary zero-carbon toll station construction process optimization method based on energy distribution, and relates to the field of toll station construction, and the method comprises the steps: predicting the power utilization demand in a toll station, obtaining the power supply distribution output of energy equipment based on the power utilization demand, and obtaining the power supply distribution output of the energy equipment; according to the power supply distribution output result and the power purchase and sale agreement, excess power sale transaction is carried out; the load demand in the toll station is predicted, the supply mode of the geothermal heat pump is adjusted based on the load demand, and the heat output of the geothermal heat pump is determined by combining the real-time temperature in the toll station and the frequency conversion control technology; and combining a power supply distribution output result, a surplus power selling transaction result and a heat output result of the geothermal heat pump as an operation process optimization strategy, and optimizing the multi-energy complementary zero-carbon construction process of the toll station. According to the invention, redundant power selling transaction is carried out based on the power purchase and sale agreement, so that redundant power can be reasonably sold to the power market when there is redundant power, and the economic value of power resources is maximized.

Description

Technical Field

[0001] The present invention relates to the construction of toll stations, and more particularly, to an optimization method for the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution. Background Art

[0002] As one of the main sources of carbon emissions, the transportation industry is gradually moving towards a low-carbon or even zero-carbon direction. The concept of "zero-carbon toll stations" has emerged based on this trend. As an important node in the transportation network, toll stations have a large energy demand, including lighting, heating, cooling, vehicle charging, and daily services. By integrating various new energy technologies, it is possible to achieve self-sufficiency in the energy supply of toll stations and upload excess electricity to the power grid, which can not only reduce carbon emissions but also lower energy costs and maximize economic benefits.

[0003] Traditional toll stations usually rely on fossil fuel power generation or conventional power grid power supply. This energy supply method not only leads to high energy consumption but also generates a large amount of carbon dioxide emissions, which cannot meet the requirements of current low-carbon development. Currently, most toll stations rely on a single energy form (such as power grid power supply), lacking the complementarity and balance of multiple energies. As a result, in the face of extreme weather or energy supply interruptions, they are unable to maintain normal operations. At the same time, the existing energy management systems of toll stations are usually relatively simple and difficult to conduct intelligent scheduling according to actual situations, resulting in low energy utilization efficiency and widespread energy waste. Moreover, when the power generation of renewable energies such as photovoltaic and wind exceeds the demand, the excess power of existing toll stations cannot be effectively processed, easily causing energy waste.

[0004] In response to the problems in the related art, no effective solutions have been proposed yet. Summary of the Invention

[0005] In response to the problems in the related art, the present invention proposes an optimization method for the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution to overcome the above-mentioned technical problems existing in the related art.

[0006] To this end, the specific technical solutions adopted by the present invention are as follows:

[0007] An optimization method for the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution, comprising:

[0008] Predicting the electricity demand inside the toll station, obtaining the power supply distribution output of energy equipment based on the electricity demand, and conducting excess power sales transactions according to the power supply distribution output result and the power purchase and sales agreement;

[0009] Predicting the load demand inside the toll station, adjusting the supply mode of the ground heat pump based on the load demand, and determining the heat output of the ground heat pump in combination with the real-time temperature inside the toll station and the variable frequency control technology;

[0010] Combine the power supply distribution output result, the surplus power sale transaction result and the heat output result of the ground heat pump as the operation process optimization strategy to optimize the multi-energy complementary zero-carbon construction process of the toll station.

[0011] Preferably, predict the electricity demand in the toll station, obtain the power supply distribution output of the energy equipment based on the electricity demand, and conduct the surplus power sale transaction according to the power supply distribution output result and the power purchase and sale agreement, including:

[0012] Obtain the historical charging data of the charging piles in the toll station and the historical power consumption data of the living service area, and combine the evidence theory and the convolutional neural network technology to predict the electricity demand in the toll station;

[0013] Construct a power supply distribution model based on the electricity demand and the dynamic characteristics of the energy equipment power supply, determine the power supply distribution output of the energy equipment, and judge the remaining energy storage of the energy equipment;

[0014] Combine the remaining energy storage and the energy generation rate to analyze the movable energy of the energy equipment at any time on the premise of meeting the electricity consumption of the charging piles and the living service area in the toll station;

[0015] Combine the movable energy with the power purchase and sale agreement to conduct the surplus energy sale transaction processing, and determine the sale time and sale quantity in the surplus energy sale transaction processing according to the real-time electricity price.

[0016] Preferably, obtain the historical charging data of the charging piles in the toll station and the historical power consumption data of the living service area, and combine the evidence theory and the convolutional neural network technology to predict the electricity demand in the toll station, including:

[0017] Establish a charging demand matrix based on the number of charging piles in the toll station, decompose the charging demand matrix into several types of charging modes according to the non-negative matrix factorization method, and judge the charging scale and charging scale of several types of charging modes in the toll station;

[0018] Generate a basis matrix and a coefficient matrix according to the charging scale and charging scale, and multiply the rows and columns of the corresponding charging modes in the basis matrix and the coefficient matrix to obtain a dimension matrix;

[0019] Obtain the historical charging data of the charging piles in the toll station, and substitute it together with the dimension matrix as the input into the graph convolutional neural network to capture the effective information between the charging mode and the charging data, and predict the electricity demand of the charging piles;

[0020] Obtain the historical power consumption data of the living service area in the toll station, and obtain the electricity demand in the living area according to the historical power consumption development trend and the evidence theory, and combine the electricity demand of the charging piles to obtain the total electricity demand of the toll station.

[0021] Preferably, obtaining the historical power consumption data of the living service area in the toll station, and obtaining the electricity demand in the living area according to the historical power consumption development trend and the evidence theory includes:

[0022] Obtaining the historical power consumption data at each sub-area in the living service area of the toll station, and generating a prediction set of electricity demand in each sub-area according to the historical power consumption data and the seasonal nature;

[0023] Predicting the basic probability assignment matrix according to the electricity demand prediction set and the nature of the sub-area electrical equipment, and synthesizing the electricity demand prediction set using the evidence theory;

[0024] Based on the synthesis result, performing evidence conflict analysis on the electricity demand prediction set, obtaining the belief function and plausibility function of the electricity demand prediction set, and converting the belief function into a probability distribution according to the probability conversion formula;

[0025] Obtaining the probability distribution of the electricity demand in each sub-area according to the distribution result, and combining the probability distribution with the basic probability assignment matrix and the electricity demand of the sub-area, and integrating the electricity demand of each sub-area to obtain the electricity demand in the living area.

[0026] Preferably, constructing a power supply allocation model based on the electricity demand and the dynamic characteristics of the energy equipment power supply, determining the power supply allocation output of the energy equipment, and judging the remaining energy storage of the energy equipment includes:

[0027] According to the power generation characteristics and energy storage characteristics of the energy equipment, analyzing the minimum expected frequency of the power supply allocation of any power supply equipment based on the energy storage characteristics, and defining the minimum power generation objective function of any power supply equipment using the power generation characteristics;

[0028] Based on the naive Bayes, making the maximum a posteriori hypothesis for the electricity demand of the toll station, and smoothing the maximum a posteriori hypothesis process, and calibrating the dynamic allocation parameters according to the processing result;

[0029] Combining the dynamic allocation parameters, the minimum expected frequency and the minimum power generation objective function to analyze the energy consumption of the energy equipment during the power supply process, and representing the objective function of the power supply allocation model in an octuple manner based on the naive Bayes;

[0030] Setting the boundary conditions of the power supply allocation model based on the dynamic allocation balance of the energy equipment, and combining with the objective function to obtain the constructed power supply allocation model;

[0031] Using the power supply allocation model to determine the power supply allocation output of any power supply equipment, and obtaining the remaining energy storage of the energy equipment according to the difference between the energy storage of the energy equipment and the power supply allocation output.

[0032] Preferably, the expression of the power supply allocation model is:

[0033] ;

[0034] Wherein, F(t) represents the power supply distribution of the t-th power supply device, e represents the energy consumption, α1 represents the probability that the power supply distribution output reaches high uniformity, α2 represents the probability that the power supply distribution output reaches low uniformity, υ represents the objective function of the power supply distribution model, β represents the minimum power generation target value of any power supply device, T min represents the minimum expected frequency, and ε represents the dynamic distribution parameter.

[0035] Preferably, combining the movable energy quantity with the power purchase and sale agreement to conduct the surplus energy sale transaction processing, and determining the sale time and sale quantity during the surplus energy sale transaction processing according to the real-time electricity price, including:

[0036] Combining the movable energy quantity with the power purchase and sale agreement to define the surplus energy sale rule, connecting with the power company's smart meter system, and automatically recording and uploading the output data of the surplus energy;

[0037] Obtaining the historical electricity price data, reconstructing the phase space of the historical electricity price data time series, analyzing the phase points at any time point of the historical electricity price, and the closest points of the phase points;

[0038] Obtaining the distance between the phase point and its closest point, traversing the distances between all phase points and their corresponding closest points in the phase space to obtain the dynamic evolution law, and analyzing the evolution time series of the electricity price based on the dynamic evolution law;

[0039] Predicting the real-time electricity price at any time based on the evolution time series, and determining the time and quantity of the surplus energy sale transaction according to the level of the real-time electricity price.

[0040] Preferably, predicting the load demand in the toll station, adjusting the supply mode of the ground heat pump based on the load demand, and determining the heat output of the ground heat pump in combination with the real-time temperature and frequency conversion control technology in the toll station, including:

[0041] Obtaining the historical heating and cooling data in the toll station, and predicting the heating or cooling load demand in the toll station at any time period in combination with seasonal factors and marginal distribution technology;

[0042] Determining the temperature demand in the toll station according to the load demand, determining the supply mode of the ground heat pump based on the temperature demand, and adjusting the supply mode of the ground heat pump through the determination result;

[0043] Adjusting the supply quantity of the ground heat pump according to the real-time temperature and load demand in the toll station, and calculating the optimal operating frequency of the ground heat pump in combination with the frequency conversion control technology and the characteristics of the ground heat pump;

[0044] Based on the heat output results of the ground source heat pump at the optimal operating frequency and supply volume, and adjust the optimal operating frequency according to the real-time feedback to optimize the operating efficiency of the ground source heat pump.

[0045] Preferably, obtain the historical heating and cooling data within the toll station, and combine seasonal factors and marginal distribution technology to predict the heating or cooling load demand of the toll station at any time period, including:

[0046] Obtain historical heating and cooling data based on the historical operation data of the toll station, construct an estimation sample containing two variables of heating and cooling according to the obtained results, and obtain the distribution function value of the estimation sample under seasonal factors through non-parametric estimation method;

[0047] Construct a conditional matrix according to the marginal distribution function value and the estimation sample, and use the conditional matrix to predict the probability of each data value appearing in the estimation sample and the corresponding marginal value;

[0048] Based on the upper and lower bounds of the preset load demand interval, and combine probability and marginal value to analyze the expected probability of the data value within the load demand interval, and obtain the marginal distribution matrix of the data value by using the expected probability;

[0049] Perform an inverse operation on the marginal distribution matrix to obtain the demand prediction interval for heating and cooling of the toll station, and determine the heating or cooling load demand of the toll station from the demand prediction interval according to the time period.

[0050] Preferably, the calculation formulas for the probability of the data value appearing and the marginal value are:

[0051] ;

[0052] In the formula, S i represents the probability that the i-th data value appears in the estimation sample, T i represents the distribution function value of the i-th data value, P represents the characteristic value of the estimation sample under seasonal factors, E i represents the marginal value of the i-th data value, H(i,m) represents the joint probability mass function of the data value i and the conditional matrix m, represents the load size of the data value i in the m-th conditional matrix.

[0053] The beneficial effects of the present invention are:

[0054] 1. An optimization method for the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution proposed by the present invention aims to achieve the efficient operation and zero-carbon goal of the toll station by comprehensively predicting and optimizing multiple energy elements such as power supply, load demand, and ground-source heat pump heating mode. By predicting the electricity demand of the toll station, the energy supply method can be planned in advance, and the distribution of different energies can be optimized according to the prediction results. Moreover, according to the load demand, the heating mode of the ground-source heat pump will be dynamically adjusted to accurately meet the heating and cooling load demands, avoid overheating or overcooling, improve energy efficiency. At the same time, through the sale transaction of excess electricity based on the power purchase and sale agreement, it is ensured that when there is excess electricity, it can be reasonably sold to the electricity market, which can not only effectively reduce the burden on the power grid but also bring additional income to the toll station, maximizing the economic value of power resources.

[0055] 2. The present invention can improve energy utilization efficiency, optimize power trading to obtain economic benefits, reduce carbon emissions, enhance system stability and reliability, and reduce the overall operating cost through accurate demand prediction, intelligent power supply distribution, energy storage management, power sale transaction, and maximum use of green energy. At the same time, it improves the intelligent level of the toll station, providing strong support for achieving the zero-carbon goal. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the following described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0057] Figure 1 is a principle block diagram of an optimization method for the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution according to an embodiment of the present invention;

[0058] Figure 2 is a principle block diagram of a multi-energy complementary zero-carbon toll station in an optimization method for the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution according to an embodiment of the present invention;

[0059] Figure 3 is a principle block diagram of a photovoltaic power generation module in an optimization method for the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution according to an embodiment of the present invention;

[0060] Figure 4 is a principle block diagram of a wind power generation module in an optimization method for the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution according to an embodiment of the present invention;

[0061] Figure 5It is a schematic block diagram of the geothermal heating and cooling module in an optimization method for the construction process of a multi - energy complementary zero - carbon toll station based on energy distribution according to an embodiment of the present invention. Detailed implementation manners

[0062] To further illustrate each embodiment, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. They are mainly used to illustrate the embodiments and can be combined with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, those of ordinary skill in the art should be able to understand other possible implementation manners and the advantages of the present invention.

[0063] According to an embodiment of the present invention, there is provided an optimization method for the construction process of a multi - energy complementary zero - carbon toll station based on energy distribution.

[0064] Now, the present invention will be further described in conjunction with the accompanying drawings and detailed implementation manners. As Figure 1 shown, the optimization method for the construction process of a multi - energy complementary zero - carbon toll station based on energy distribution according to an embodiment of the present invention includes:

[0065] Step S1, predicting the electricity demand inside the toll station, obtaining the power supply distribution output of energy devices based on the electricity demand, and conducting surplus power sale transactions according to the power supply distribution output result and the power purchase and sale agreement.

[0066] In one embodiment, predicting the electricity demand inside the toll station, obtaining the power supply distribution output of energy devices based on the electricity demand, and conducting surplus power sale transactions according to the power supply distribution output result and the power purchase and sale agreement includes:

[0067] Obtaining the historical charging data of charging piles and the historical power consumption data of the living service area inside the toll station, and combining the evidence theory and convolutional neural network technology to predict the electricity demand inside the toll station;

[0068] Constructing a power supply distribution model based on the electricity demand and the dynamic characteristics of energy device power supply, determining the power supply distribution output of energy devices, and judging the remaining energy storage of energy devices;

[0069] Combining the remaining energy storage and the energy generation rate to analyze the movable energy amount of energy devices at any time on the premise of meeting the electricity consumption of charging piles and the living service area inside the toll station;

[0070] Combining the movable energy amount with the power purchase and sale agreement to conduct surplus energy sale transaction processing, and determining the sale time and sale quantity during the surplus energy sale transaction processing according to the real - time electricity price.

[0071] Specifically, when obtaining the historical charging data of the charging piles in the toll station and the historical power consumption data of the living service area, and combining the evidence theory and convolutional neural network technology to predict the electricity demand in the toll station, a charging demand matrix can be established based on the number of charging piles in the toll station, and the charging demand matrix can be decomposed into several types of charging modes according to the non-negative matrix factorization method, and the charging scale and charging scale of several types of charging modes in the toll station can be judged; a basis matrix and a coefficient matrix can be generated according to the charging scale and charging scale, and the corresponding rows and columns of the charging modes in the basis matrix and the coefficient matrix are multiplied to obtain a dimension matrix; the historical charging data of the charging piles in the toll station is obtained, and together with the dimension matrix, it is used as an input and substituted into the graph convolutional neural network to capture the effective information between the charging mode and the charging data, and predict the electricity demand of the charging piles; the historical power consumption data of the living service area in the toll station is obtained, and the electricity demand in the living area is obtained according to the historical power consumption development trend and the evidence theory, and the total electricity demand of the toll station is obtained by combining the electricity demand of the charging piles.

[0072] It should be explained that when obtaining the total electricity demand of the toll station, the historical charging data of all the charging piles in the toll station can be collected. The data content includes the charging power, charging duration, charging frequency, charging amount, etc. of each charging pile at different time periods. According to the number of charging piles in the toll station, the charging modes of the charging piles (such as the charging behaviors of different electric vehicles) and their charging data at different time periods, a charging demand matrix is constructed. The rows of the matrix represent different time periods (such as hourly data of each day), and the columns represent the charging demands of different charging piles. Each row represents the charging demands of all the charging piles during a certain time period, and each column represents the charging demands of a certain charging pile during all time periods.

[0073] Non-negative matrix factorization (NMF) decomposes the charging demand matrix into two non-negative matrices, namely the basis matrix and the coefficient matrix. This decomposition helps to extract the potential charging modes in the charging demand matrix:

[0074] Basis matrix: It contains the potential charging modes and represents the charging demands of each charging pile under different charging modes; Coefficient matrix: It represents the charging intensity of each charging mode at different time periods.

[0075] The goal of NMF is to find the basis matrix (W) and the coefficient matrix (H) such that: X≈W.H, where X represents the charging demand matrix, W represents the basis matrix, and H represents the coefficient matrix. In this way, the charging demand can be decomposed into several types of charging modes, and the charging scale of different modes can be calculated.

[0076] By analyzing each element in the coefficient matrix, the charging scale of each charging mode is judged. The charging scale can be defined as the total charging amount of a charging mode within a specific time period, or the load size when multiple charging piles use this charging mode simultaneously. If the coefficient of a charging mode is large, it indicates that the charging demand of this mode within this time period is large, and more charging piles or higher power may be required. If the coefficient of the charging mode is small, it means that the charging demand of this mode is low, which may be an occasional demand for a specific charging pile.

[0077] The basis matrix obtained through NMF decomposition represents the characteristics of charging modes. Each row represents a charging mode, and each column represents the charging demand of a charging pile. The coefficient matrix represents the intensity of each charging mode at different time periods. Each row represents a time period, and each column represents the intensity of a charging mode (i.e., the demand of this mode at this moment).

[0078] Multiply the corresponding rows and columns of the basis matrix and the coefficient matrix to obtain a new matrix. This matrix reflects the relationship between different charging modes and charging piles, and further helps to capture the characteristics and trends of charging data. Each element of the matrix corresponds to a combination of a charging mode and a charging pile, reflecting the demand performance of a specific charging mode on a specific charging pile. Input the historical charging data (such as the historical charging power and duration of charging piles, etc.) and the matrix into the graph convolutional neural network (GCN). GCN can capture spatial or temporal dependencies in graph-structured data through convolutional operations.

[0079] During the construction of the graph structure, information such as the mutual relationship of charging piles and the change of charging modes is represented through the graph structure. In the graph convolutional neural network, nodes represent charging piles or charging modes, and edges represent the relationships between them (such as the similarity of charging behaviors). Through the graph convolutional layer, the GCN model can capture the effective information between charging modes and charging piles, and capture potential patterns and trends. For example, GCN can help identify which charging piles have similar charging behaviors in similar time periods.

[0080] The evidence theory (or Dempster - Shafer theory) can handle uncertainty. By combining evidence from different sources (such as historical electricity consumption data, environmental factors, etc.), the credibility of future electricity demand in the living area is obtained. Combine historical electricity consumption data with other influencing factors (such as weather, holidays, etc.), and use the evidence theory to reason and predict the electricity demand in the living area. The key of the evidence theory is to give weights to different data sources according to the relative importance of the evidence, and synthesize the final prediction result through the Dempster combination rule.

[0081] Specifically, when obtaining the historical power consumption data of the living service area in the toll station and obtaining the power consumption demand in the living area according to the historical power consumption development trend and the evidence theory, the historical power consumption data of each sub-area in the living service area of the toll station can be obtained, and the power consumption demand prediction set of each sub-area can be generated according to the historical power consumption data and the seasonal nature; according to the power consumption demand prediction set and the nature of the sub-area power consumption equipment, the basic probability assignment matrix can be predicted, and the evidence theory can be used to synthesize the power consumption demand prediction set; based on the synthesis result, the evidence conflict analysis and processing are carried out on the power consumption demand prediction set to obtain the belief function and the plausibility function of the power consumption demand prediction set, and the belief function is converted into a probability distribution according to the probability conversion formula; according to the distribution result, the probability distribution of the power consumption demand of each sub-area is obtained, and the probability distribution and the basic probability assignment matrix are combined with the power consumption demand of the sub-area, and the power consumption demands of each sub-area are integrated to obtain the power consumption demand in the living area.

[0082] It should be explained that in the process of obtaining the power consumption demand of the living area, it is necessary to collect the historical power consumption data of each sub-area in the living service area of the toll station. Each sub-area (such as a restaurant, a store, a rest area, etc.) has independent power consumption data, which may include the power consumption of each sub-area at different time periods (such as hours, days, months). These historical data are time series data. Therefore, the power consumption patterns of each sub-area can be analyzed based on time and seasonal trends. At the same time, the power consumption demand is greatly affected by seasonal factors. For example, the air-conditioning load will increase in summer, and the heating load will increase in winter. Based on the seasonal changes in the historical data, the power consumption demand trends of each sub-area in different seasons can be inferred. For example, the power consumption demand in summer may increase significantly due to the increased use frequency of air conditioners, while in winter it may increase due to the use of heating equipment.

[0083] Combining historical data and seasonal influencing factors, predicting the power consumption demand of each sub-area in different seasons and different time periods, generating a prediction set through time series analysis or other statistical methods (such as ARIMA, seasonal decomposition, etc.), including the estimated power consumption demands in different time periods and seasons. Different electrical equipment (such as air conditioners, lighting, electric equipment, etc.) has different power consumption characteristics. At this stage, it is necessary to generate a basic probability assignment matrix (BPA matrix) according to the nature of each electrical equipment in the sub-area (such as power consumption, usage frequency, etc.). The BPA matrix represents the initial probability distribution of the power consumption demand prediction of each electrical equipment under given conditions.

[0084] The set of electricity demand forecasts within each sub-region is combined with the basic probability assignment matrix to describe the probability distribution of different devices in each sub-region during future electricity consumption periods. This matrix reflects the possibility and reliability of different forecast results. The evidence theory is used to synthesize the electricity demand forecast sets of each sub-region. The evidence theory generates a more accurate electricity demand forecast by combining evidence from different sources (such as historical data, seasonal trends, device characteristics, etc.). Specifically, the Dempster-Shafer combination rule is used to combine different evidence sets (forecast results) to obtain a new combined result. The combined result is based on the combination of multiple different data sources and can effectively handle uncertainty and conflict. For each sub-region, the combination rule combines the known forecast set with its reliability (confidence level) to output a new, comprehensive forecast result.

[0085] During the synthesis process, conflicts may occur between different pieces of evidence. The goal of conflict analysis is to identify and handle the conflicts between different pieces of evidence to ensure the accuracy of the combined result. By calculating the conflict degree between different pieces of evidence, the relevance and reliability of each piece of evidence are judged. If the conflict is too large, it may be necessary to adjust the weight of the evidence or reprocess the data.

[0086] The belief function represents the reliability of each forecast result and reflects the credibility of the forecast result after synthesis. The higher the belief, the more reliable the forecast result. The belief function is usually deduced from the support degree in the evidence theory. The support degree reflects the degree of support of the evidence for a certain forecast result. The plausibility function represents the credibility of a certain forecast result (i.e., a measure of uncertainty) and is used to describe the possibility of a certain forecast result under the current evidence. The probability conversion formula is used to convert the belief function into a probability distribution. The belief function and the plausibility function, as tools for uncertainty assessment, can be converted into specific probability distributions through the probability conversion formula. These distributions describe different possibilities of electricity demand. The core idea of the conversion formula is to convert the belief function into a probability density function to obtain different probability distributions of electricity demand.

[0087] Through the probability conversion result, the probability distribution of the electricity demand of each sub-region is obtained. These probability distributions reflect the uncertainty and variation range of the future electricity demand in each sub-region. Key statistics are extracted from the probability distribution, such as the expected value (i.e., the most likely electricity demand), variance (i.e., uncertainty), confidence interval, etc. These statistics help to further optimize the electricity demand forecast and resource allocation. The probability distribution of the electricity demand of each sub-region is combined with the basic probability assignment matrix and integrated into the electricity demand forecast of the entire living service area according to the weighted average or combination rule. The electricity demands of each sub-region are synthesized, considering the weight of each sub-region (such as regional area, number of devices, importance, etc.), and finally the electricity demand forecast of the entire living area is obtained.

[0088] Specifically, when constructing a power supply distribution model based on electricity demand and the dynamic characteristics of energy device power supply, determining the power supply distribution output of energy devices, and judging the remaining energy storage of energy devices, according to the power generation characteristics and energy storage characteristics of energy devices, the minimum expected frequency of power supply distribution of any power supply device can be analyzed based on the energy storage characteristics, and the minimum power generation objective function of any power supply device can be defined using the power generation characteristics; perform a maximum a posteriori hypothesis on the electricity demand of the toll station based on Naive Bayes, smooth the maximum a posteriori hypothesis process, and calibrate the dynamic distribution parameters according to the processing results; combine the dynamic distribution parameters, the minimum expected frequency, and the minimum power generation objective function to analyze the energy consumption of energy devices during power supply, and represent the objective function of the power supply distribution model in an octuple manner based on Naive Bayes; set the boundary conditions of the power supply distribution model based on the dynamic distribution balance of energy devices, and combine with the objective function to obtain the completed power supply distribution model; use the power supply distribution model to determine the power supply distribution output of any power supply device, and obtain the remaining energy storage of the energy device according to the difference between the energy storage of the energy device and the power supply distribution output.

[0089] Preferably, the expression of the power supply distribution model is:

[0090] ;

[0091] In the formula, F(t) represents the power supply distribution of the t-th power supply device, e represents energy consumption, α1 represents the probability that the power supply distribution output reaches high uniformity, α2 represents the probability that the power supply distribution output reaches low uniformity, υ represents the objective function of the power supply distribution model, β represents the minimum power generation target value of any power supply device, T min represents the minimum expected frequency, and ε represents the dynamic distribution parameter.

[0092] It should be noted that during the process of constructing the power supply distribution model, according to the power generation characteristics and energy storage characteristics of energy devices, the minimum expected power supply frequency and target power generation that each device can provide within a given time need to be analyzed:

[0093] Power generation characteristics: Each power supply device (such as solar power generation, wind power generation, etc.) has different power generation capabilities under different environmental conditions, and this characteristic can be modeled through the relationship between the output power of the device and time;

[0094] Energy storage characteristics: Energy storage devices have limitations on the maximum storage capacity and charge-discharge efficiency. The discharge and charge processes of the energy storage system need to consider these characteristics to ensure that the maximum energy storage capacity is not exceeded.

[0095] Based on these characteristics, analyze the minimum expected frequency that any power supply device can provide under specific conditions, that is, how frequently the device needs to supply power within a given time.

[0096] To ensure that the energy equipment can provide a stable power supply without overloading, a target function can be established according to the characteristics of the power generation equipment to calculate the minimum power generation target and ensure that the power generation during power supply is not less than the equipment's demand:

[0097] ;

[0098] In the formula, P gen represents the power generation, E min represents the minimum power generation requirement, A represents the time range, and the minimization of the target function will help find the optimal power supply amount under the constraints of the equipment's power generation characteristics.

[0099] The Naive Bayes classifier can infer the power consumption demand of the toll station based on known historical data (such as historical power consumption demand, equipment status, weather conditions, etc.). Through the Maximum A Posteriori (MAP) method, it can infer the most likely power consumption demand based on the known prior information:

[0100] Based on historical data (power consumption demand) and equipment characteristics (power generation, energy storage, etc.), using Bayes' formula for inference, the power consumption demand inferred by the Maximum A Posteriori hypothesis may be affected by noise and fluctuations. Therefore, this process is smoothed, such as using a moving average or Kalman filtering technique, to eliminate noise, smooth the data, and improve the prediction accuracy. The result obtained after smoothing can be used as the basis for dynamically allocating parameters for the next step of power supply allocation.

[0101] Based on the smoothed Maximum A Posteriori hypothesis result, calibrate the dynamic allocation parameters. These parameters are used to represent the allocation rules during the power supply process to ensure that the load of the system and the power output of the power supply equipment can match the power consumption demand. Specifically:

[0102] Based on historical data and power consumption demand prediction, calibrate the power consumption demand for each time period and each sub-region; according to the power generation characteristics and energy storage characteristics of the equipment, dynamically adjust the power supply allocation parameters to meet real-time demands.

[0103] Based on the Naive Bayes classification result and the dynamic allocation parameters, the target function of the power supply allocation model can be represented in an eight-tuple manner. The eight-tuple contains the following elements:

[0104] Power generation, power consumption demand, remaining energy storage, time period, minimum expected frequency, power difference (such as the difference between the equipment's power generation and demand), maximum energy storage capacity, energy consumption cost. This eight-tuple describes the core objective of the power supply allocation model: how to reasonably allocate power generation to meet the power consumption demand in different time periods, while considering the power generation capacity of the equipment, energy storage capacity limitations, and costs.

[0105] When designing the objective function, it is necessary to ensure the dynamic balance of power supply distribution, that is, the difference between the power generation and the power consumption demand will not cause the equipment to operate overloaded or be overcharged, so as to maintain the stability and reliability of the energy equipment. To ensure the feasibility and stability of the model, appropriate boundary conditions need to be set, and these conditions usually include:

[0106] Equipment capacity limit: The maximum power generation capacity and energy storage capacity of each power supply equipment;

[0107] Demand limit: The power consumption demand cannot exceed the maximum power provided by the equipment;

[0108] Storage limit: The stored energy of the energy storage equipment cannot exceed its maximum capacity.

[0109] According to the objective function, boundary conditions and dynamic distribution parameters, a power supply distribution model is finally constructed, and with the constructed power supply distribution model, the power supply output of any power supply equipment can be determined.

[0110] Specifically, when analyzing the available energy of the energy equipment at any time by combining the remaining stored energy with the energy generation rate under the premise of meeting the electricity consumption of the toll station charging piles and the living service area, the energy generation rate refers to the amount of electricity generated by the energy equipment (such as solar energy, wind energy, etc.) per unit time. For example, the energy generation rate of a solar power generation system depends on the solar radiation intensity, weather conditions and the efficiency of the system; the energy generation rate of a wind power generation system depends on the wind speed and the efficiency of the generator. The calculation formula of the energy generation rate is p gen (r)=E produce (r) / Δr, p gen (r) represents the energy generation rate at time r, E produce (r) represents the amount of electricity generated per unit time, such as the amount of electricity generated per hour.

[0111] The available energy refers to the remaining energy that the energy equipment can provide under the premise of meeting the electricity consumption needs of the charging piles and the living service area. Its calculation process includes the following steps:

[0112] Sum up the electricity consumption demands of the charging piles and the living service area to obtain the overall electricity consumption demand. According to the total electricity consumption demand, calculate the amount of energy required within a specific time period (such as 1 hour, 1 day, etc.). If there are multiple energy equipment in the system (such as solar power generation, wind power generation, energy storage batteries, etc.), the amount of energy required needs to be calculated according to the output characteristics of each equipment.

[0113] The available energy is the result of the combination of the remaining stored energy and the energy generation rate. The formula is as follows:

[0114] ;

[0115] Wherein, E movable represents the movable energy quantity, and E remaining represents the remaining energy storage capacity. represents the total energy generated by the energy device within the time period T, and P totaldemand represents the energy quantity required to meet the electricity consumption demands of the charging piles and the living service areas.

[0116] If the movable energy quantity is greater than 0, it indicates that after meeting the electricity consumption demands, the energy device can provide the remaining movable energy for sale or storage; if the movable energy quantity is 0 or negative, it indicates that the energy storage capacity and generation rate of the energy device are insufficient to meet the electricity consumption demands, and it may be necessary to purchase electricity from the power grid or dispatch other standby energy devices.

[0117] Specifically, when combining the movable energy quantity with the power purchase and sale agreement to process the redundant energy sale transaction and determining the sale time and sale quantity during the redundant energy sale transaction process according to the real-time electricity price, the movable energy quantity can be combined with the power purchase and sale agreement to define the redundant energy sale rules, and connected to the power company's smart meter system to automatically record and upload the output data of the redundant energy; obtain the historical electricity price data, perform phase space reconstruction on the time series of the historical electricity price data, analyze the phase points at any time point of the historical electricity price, and the closest points to the phase points; obtain the distances between the phase points and their closest points, and traverse the distances between all phase points and the corresponding closest points within the phase space to obtain the dynamic evolution law, analyze the evolution time series of the electricity price based on the dynamic evolution law; predict the real-time electricity price at any time based on the evolution time series, and determine the time and sale quantity of the redundant energy sale transaction according to the level of the real-time electricity price.

[0118] It should be explained that during the process of determining the time and sale quantity of the redundant energy sale transaction, it is necessary to combine the movable energy quantity (such as the remaining electricity quantity of the energy storage system or the generated electricity quantity of renewable energy) and the power purchase and sale agreement to define the redundant energy sale rules:

[0119] Movable energy quantity: refers to the remaining energy that can be sold, such as the electric energy stored through energy storage devices, or the redundant electricity generated from renewable energy devices (such as solar energy, wind energy);

[0120] Power purchase and sale agreement: is a contract agreement between the power company and the user, which stipulates the basic conditions, unit price, transaction time, etc. of the power purchase and sale.

[0121] Based on these contents, define the redundant energy sale rules, including:

[0122] The maximum energy sale quantity for each transaction;

[0123] The electricity price and time window for sale (according to the regulations of the power company and the grid demand);

[0124] Trading restrictions on the electricity sold (such as whether it is allowed to sell within a certain time period, the upper limit of the electricity that can be sold, etc.);

[0125] These rules are connected to the power company's smart meter system, enabling the output data of excess energy to be automatically recorded and uploaded, ensuring the real-time and accuracy of the transactions.

[0126] In order to predict the future real-time electricity price, it is necessary to conduct an in-depth analysis of historical electricity price data. The specific steps are as follows:

[0127] By obtaining historical electricity price data from power companies or relevant data sources, this data is usually time series data, representing the changes in electricity prices over a past period (such as the electricity price per hour). Phase space reconstruction is a method of nonlinear time series analysis used to map time series data into a high-dimensional space in order to reveal the dynamic laws therein. Based on the historical electricity price data, phase space reconstruction is carried out, and the specific method is as follows:

[0128] Select the delay time and embedding dimension: First, select the delay time and embedding dimension for time series reconstruction. These two determine how to extract meaningful features from the time series. Using the delay time and embedding dimension, map the historical electricity price data into a high-dimensional phase space to obtain the corresponding "phase points". Each point in the phase space represents the state of the electricity price data at a specific time point.

[0129] The phase point is the position of the historical electricity price in the phase space, representing the dynamic characteristics of the electricity price change; the closest point is the point in the phase space that is closest to the current phase point, representing the occurrence of similar states in the electricity price time series.

[0130] By analyzing the historical electricity price data, the distance between any phase point and its closest point can be calculated, and a distance matrix can be obtained. This distance matrix reflects the similarity between electricity price states, helping to understand the change patterns and laws of electricity prices.

[0131] Through the reconstruction of the phase space and the calculation of the distances between phase points, the dynamic evolution law of historical electricity prices can be analyzed. Specifically, by traversing all the phase points and their closest points in the phase space, calculating the distances between the phase points, through calculating these distances, revealing the dynamic law of electricity price evolution, and at the same time, based on the distance matrix, analyzing the evolution time series of electricity prices, finding the change patterns and trends of electricity prices, such as: the frequency and amplitude of electricity price fluctuations; the periodic changes and trend changes of electricity prices; the sudden change points of electricity prices; these analyses can help predict the future trend of electricity prices.

[0132] Based on the evolutionary time series of electricity prices, time series prediction methods (such as ARIMA, LSTM neural networks, etc.) are used to predict the real-time electricity price at a future time point. The goal of this step is to predict the future trend of electricity prices according to the dynamic evolution law of historical electricity prices. Based on the predicted value of the real-time electricity price, a strategy for selling excess energy can be formulated:

[0133] If the real-time electricity price is high, it indicates that the demand in the electricity market is large. At this time, selling excess energy can obtain higher returns. Therefore, during high electricity price periods, priority should be given to selling excess energy; if the electricity price is low, the return from selling excess energy is low, and one can choose to delay the sale or decide whether to sell based on the energy storage capacity of the system.

[0134] Based on the real-time electricity price and the remaining energy of the energy storage device, determine the amount of electricity to be sold. For example, if the electricity price is at a peak, one can choose to sell to the maximum extent; if the electricity price is low, one can consider selling a part or not selling at all. Finally, based on the connection to the power company's smart meter system, automated transaction processing can be achieved.

[0135] To facilitate the understanding of the above technical solution of the present invention, the working principle or operation method of the present invention in the actual process will be described in detail below.

[0136] Step 1: Predict the electricity demand in the toll station;

[0137] (1) Historical charging data of charging piles:

[0138] Suppose there are 10 charging piles in the toll station, the charging power of each charging pile is 22 kW, the charging duration is 1 hour, the charging frequency is 2 times per hour, and the charging amount is 44 kWh (charging power × charging duration). The size of the daily charging demand matrix is 24 * 10, indicating the charging demand of each charging pile within 24 hours.

[0139] (2) Historical power consumption data of the living service area:

[0140] Suppose the living service area includes three sub-areas: a restaurant, a store, and a rest area. The historical power consumption data of each sub-area (for example: the daily power consumption of the restaurant is 300 kWh, the store is 150 kWh, and the rest area is 100 kWh). The collected power consumption data is a 24x3 matrix, indicating the power consumption demand of each sub-area within 24 hours.

[0141] (3) Prediction of charging demand based on charging pile data:

[0142] By recording the data of the past 30 days, a 30*24*10 charging demand matrix can be obtained, which contains the charging demands of each charging pile at different time periods. Decompose the charging demand matrix to obtain the basis matrix W (charging mode) and the coefficient matrix H (intensity of time periods).

[0143] Basis matrix W: Each column represents the demand intensity of a charging pile under a specific charging mode. For example, the first column may indicate that more power is required during the charging peak period.

[0144] Coefficient matrix H: Each row represents the charging mode intensity within each time period. For example, from 7 pm to 9 pm may be the peak electricity consumption period.

[0145] The dimensional matrix calculated from the basis matrix and the coefficient matrix can capture the relationship between the charging mode and the charging data, and be used as the input of the graph convolutional neural network (GCN).

[0146] (4) Predict the electricity demand of the living service area based on the evidence theory:

[0147] Analyze the seasonal electricity demand of the living area based on historical data. For example, the air-conditioning load of a restaurant may increase in summer, and the lighting load of a store changes with the seasons. By analyzing historical data and seasonal factors, and combining the evidence theory to fuse different prediction results, the electricity demand prediction of each sub-area can be obtained.

[0148] Combine the electricity demands of the charging piles and the living service area to obtain the total electricity demand of the toll station.

[0149] Charging demand: 10 charging piles, the charging mode and intensity are predicted by NMF. Assume the total charging demand is 500 kWh / day;

[0150] Electricity demand of the living service area: 300 kWh / day for the restaurant, 150 kWh / day for the store, and 100 kWh / day for the rest area, with a total demand of 550 kWh / day.

[0151] Total electricity demand: 500 kWh + 550 kWh = 1050 kWh / day.

[0152] Step 2: Construct a power supply distribution model;

[0153] (1) Analyze the characteristics of energy equipment:

[0154] Suppose two types of energy equipment are used:

[0155] Photovoltaic power generation system: The maximum power is 50 kW, and the average power generation is 400 kWh / day;

[0156] Wind power generation system: maximum power is 30kW, average power generation is 300kWh / day;

[0157] Energy storage device: maximum storage capacity is 500kWh, current stored energy is 200kWh.

[0158] (2) Power supply distribution model:

[0159] Through Naive Bayes classification and dynamic allocation parameter calibration, analyze the current electricity demand and power generation capacity. Assuming the time period is 24 hours, determine the hourly power generation and energy storage allocation according to the charging demand pattern predicted by NMF and the demand in the living service area. For example, during the charging peak period (7 - 9 pm), increase the output of solar and wind energy.

[0160] Step three: Calculation of available energy;

[0161] Assume that the remaining stored energy is 200kWh during a certain period; the energy generation rate is 20kW from solar energy and 15kW from wind power generation.

[0162] Charging demand: Assume the charging demand during this period is 300kWh, and the demand in the living service area is 150kWh.

[0163] According to the power generation of energy equipment: P gentotal =(20kW + 15kW)×1hr = 35kWh; The calculation of available energy is: E movable =200kWh (remaining stored energy) + 35kWh (generated energy) - 450kWh (total demand) = -215kWh. Since the available energy is negative, it means that the current energy storage and generated energy are not sufficient to meet the electricity demand.

[0164] Step four: Sale transaction of surplus energy;

[0165] Assume that the agreement stipulates that the maximum sale per hour is 100kWh, and the electricity price is 0.15 ¥ / kWh. Through phase space reconstruction and time series analysis of historical electricity price data, assume that the future electricity price is 0.20 ¥ / kWh. At this time, the price is relatively high, and surplus energy can be sold preferentially. Since the current available energy is negative, there is no available electricity for sale.

[0166] During the power supply distribution process, if the available energy of the energy equipment is greater than 0, the system will automatically initiate the transaction process for selling surplus electricity. If the available energy is less than 0, it may be necessary to purchase electricity from the grid or dispatch standby energy. The analysis and prediction of real-time electricity prices help to select the optimal selling time to maximize economic benefits.

[0167] Step S2: Predict the load demand inside the toll station, adjust the supply mode of the ground heat pump based on the load demand, and determine the heat output of the ground heat pump by combining the real-time temperature inside the toll station and the variable frequency control technology.

[0168] In one embodiment, predicting the load demand inside the toll station, adjusting the supply mode of the ground heat pump based on the load demand, and determining the heat output of the ground heat pump by combining the real-time temperature inside the toll station and the variable frequency control technology includes:

[0169] Obtain the historical heating and cooling data inside the toll station, and combine seasonal factors and marginal distribution technology to predict the heating or cooling load demand of the toll station at any time period;

[0170] Determine the temperature demand inside the toll station according to the load demand, and determine the supply mode of the ground heat pump based on the temperature demand, and adjust the supply mode of the ground heat pump through the determination result;

[0171] According to the real-time temperature and load demand inside the toll station, adjust the supply volume of the ground heat pump, and calculate the optimal operating frequency of the ground heat pump by combining the variable frequency control technology and the characteristics of the ground heat pump;

[0172] Output the heat output result of the ground heat pump based on the optimal operating frequency and supply volume, and adjust the optimal operating frequency according to the real-time feedback to optimize the operating efficiency of the ground heat pump.

[0173] Specifically, when obtaining the historical heating and cooling data inside the toll station and combining seasonal factors and marginal distribution technology to predict the heating or cooling load demand of the toll station at any time period, the historical heating and cooling data can be obtained based on the historical operation data of the toll station, and an estimation sample containing two variables of heating and cooling can be constructed according to the obtained result, and the distribution function value of the estimation sample under seasonal factors can be obtained through non-parametric estimation method; construct a conditional matrix according to the marginal distribution function value and the estimation sample, and use the conditional matrix to predict the probability of each data value appearing in the estimation sample and the corresponding marginal value; based on the upper and lower bounds of the preset load demand interval, and combine probability and marginal value to analyze the expected probability of the data value within the load demand interval, and use the expected probability to obtain the marginal distribution matrix of the data value; perform an inverse operation on the marginal distribution matrix to obtain the demand prediction interval for heating and cooling of the toll station, and determine the heating or cooling load demand of the toll station from the demand prediction interval according to the time period.

[0174] Among them, the calculation formulas for the probability of the data value appearing and the marginal value are:

[0175] ;

[0176] In the formula, S i represents the probability that the i-th data value appears in the estimation sample, T iDenote the distribution function value of the i-th data value, P represents the eigenvalue of the estimated sample under seasonal factors, E i Denote the marginal value of the i-th data value, H(i, m) represents the joint probability mass function of the data value i and the conditional matrix m, Denote the load magnitude of the data value i in the m-th conditional matrix.

[0177] It should be noted that when predicting the heating or cooling load demand in a toll station, historical heating and cooling data of the toll station need to be collected. These data usually come from the heating system and air conditioning system in the toll station and may include:

[0178] Heating load data: Denote the heating demand of the toll station in different time periods (such as hourly data per day); Cooling load data: Denote the cooling demand of the toll station in different time periods. These data contain demand fluctuations in different time periods under multiple seasons (such as winter and summer), and can reflect the impact of seasonal changes on heating and cooling loads.

[0179] Based on the historical heating and cooling data of the toll station, two variables of heating and cooling are constructed into an estimated sample, that is: each pair of data contains heating and cooling demands, and the heating demand and cooling demand in the same time period can be regarded as a bivariate data sample.

[0180] For example, on a certain day in winter, the heating demand is 30 kW and the cooling demand is 0 kW; on a certain day in summer, the heating demand is 0 kW and the cooling demand is 25 kW. In this way, a multi-dimensional sample set can be created, and these samples can be used for subsequent non-parametric estimation processes.

[0181] Use non-parametric estimation methods (such as kernel density estimation) to estimate the distribution function of sample data. In this method: it does not depend on the specific distribution assumption of the data, but estimates its probability density function according to the data itself. Suppose we hope to estimate the distribution of heating and cooling demands. The non-parametric estimation method will generate the marginal distribution functions of heating and cooling demands according to the distribution characteristics of historical data. This means that we can obtain the probability density estimates of heating and cooling demands in different time periods.

[0182] After obtaining the distribution function value of the estimated sample, the next step is to construct a conditional matrix according to the relationship between the data. This matrix describes the distribution of another variable given a certain variable. For example:

[0183] The heating demand is the known condition, and the cooling demand is the variable to be estimated. At this time, the conditional matrix reflects the probability distribution of the cooling demand value given the heating demand value. Through the relationship between the marginal distribution and the joint distribution, the probability of the cooling demand occurring under specific conditions (such as a specific heating demand) can be calculated.

[0184] Assume that the load demand for heating or cooling has a certain range. For example, the heating demand may be between 10 kW and 50 kW, while the cooling demand may be between 0 kW and 30 kW. Based on the known marginal distributions, by calculating the probability of each data value appearing within different load demand intervals (such as 10 kW to 50 kW), the expected probability is obtained, that is, the probability that the heating or cooling demand is most likely to occur within this interval. The expected probability helps to evaluate the probability that the heating or cooling demand falls within a specific load interval during a certain period. For example, in a colder season, the heating demand may be more inclined to the high-load interval.

[0185] Through the analysis of the expected probability and marginal values, the marginal distribution matrix can describe the probability distributions of heating and cooling demands at different time periods. For example, the marginal distribution of the heating demand may show different distribution characteristics of the heating demand at different time periods (such as morning, evening, or night); the marginal distribution of the cooling demand may reflect the demand patterns during the day and night in summer. The marginal distribution matrix provides a predicted distribution of the load demand at the toll station at different time periods through probability calculations.

[0186] The process of the inverse operation is to obtain the actual load demand prediction interval from the marginal distribution matrix. Assume that the expected probability of the heating demand within a specific time period is obtained. It can be transformed into an actual demand prediction interval through the inverse operation. This process is as follows:

[0187] According to the calculated marginal distribution, find the range within which the heating or cooling load demand is most likely to fall during a certain period. For example, assume that the predicted interval of the heating load demand for a certain period is from 15 kW to 40 kW. This interval provides a reference value indicating that there is a high probability that the actual load demand will fall within this interval during this time period.

[0188] Based on the combination of the demand prediction interval and the time period, the heating or cooling load demand of the toll station is finally obtained. For example: during a certain period in winter, the predicted interval of the heating load demand is from 15 kW to 40 kW. Combining the actual weather and historical data, the heating demand for a certain period may be finally determined to be 25 kW. Similarly, during a certain period in summer, the cooling load demand may be predicted to be 20 kW.

[0189] Therefore, based on the accurate prediction of the heating and cooling load demands of the toll station considering historical data and seasonal factors, the distribution of sample data is obtained through non-parametric estimation method, combined with the marginal distribution and conditional matrix, the expected probability in the demand interval is analyzed, and the demand prediction interval is obtained through the inverse operation, and finally the accurate heating or cooling load demand is determined. This method combines statistics, probability theory, and non-parametric estimation techniques, providing a scientific load demand prediction tool for the energy management of the toll station.

[0190] In the process of determining the heat output of the ground heat pump by combining the real-time temperature in the toll station with the variable frequency control technology, it is necessary to calculate the corresponding temperature demand according to the load demand of the toll station. The load demand usually represents the required energy load, which is specifically manifested as the heat or cold required by the heating (or cooling) system.

[0191] In the case of heating, the load demand is closely related to the temperature difference between the indoor and outdoor (the difference between the indoor temperature and the outdoor temperature). For example, when the external temperature is lower than the set indoor temperature, the heating system needs to provide more heat.

[0192] In the case of cooling, the load demand is related to the difference between the indoor temperature and the external temperature, as well as the indoor heat load (for example, the heat generated by lighting and equipment operation).

[0193] According to the load demand, the temperature demand can be estimated by the following formula: Q heat =c×(T inside -T outside ), where: Q heat represents the required heat (unit: kW or kWh); c represents the coefficient related to the heat insulation performance and heat conductivity of the toll station; T inside represents the target indoor temperature; T outside represents the external ambient temperature. Through the relationship, it is determined what the target temperature in the toll station is to meet the load demand.

[0194] When determining the supply mode of the ground heat pump according to the temperature demand, the ground heat pump is a device that uses the constant temperature characteristic of the underground soil for heat exchange. According to the target temperature demand, the supply mode of the ground heat pump can be determined. In the heating mode, the ground heat pump absorbs heat from the underground soil and transfers it to the building to maintain the indoor temperature. In the cooling mode, the ground heat pump absorbs heat from the building and releases it into the underground soil to lower the indoor temperature.

[0195] The supply mode of the ground heat pump can be matched with the temperature demand and the load capacity of the ground heat pump. For example, if the target temperature demand is high, the ground heat pump needs to provide a higher heat output in the heating mode. On the contrary, if the demand is low, the ground heat pump will operate at a lower load level. In the winter heating mode, the temperature control of the ground heat pump may need to adjust the water supply temperature to between 35°C - 45°C to adapt to different heating demands. For higher heating demands, it may be necessary to increase the output capacity of the pump. In the summer cooling mode, the output of the ground heat pump needs to be adjusted in reverse to maintain a lower temperature difference.

[0196] According to the real-time indoor temperature and load demand, the supply of the ground heat pump will be adjusted. To optimize energy usage efficiency and comfort, it is necessary to dynamically adjust the workload of the ground heat pump based on real-time temperature data and load demand. By installing temperature sensors inside the toll station, the indoor temperature data is monitored in real time. These data will be input into the control system for comparison with the target temperature to adjust the working state of the ground heat pump.

[0197] Winter heating: If the indoor temperature is lower than the set target temperature and the load demand increases, the ground heat pump will need to output more heat to quickly restore the indoor temperature. If the indoor temperature is close to the set value, the ground heat pump will reduce the output to maintain a stable indoor temperature.

[0198] Summer cooling: In the cooling mode, as the outside air temperature changes, the load demand will fluctuate, and the output of the ground heat pump will be adjusted in real time according to these changes.

[0199] Variable Frequency Drive (VFD) technology is a technology that adjusts the motor speed according to the load demand. In the ground heat pump system, the variable frequency control technology can adjust the working frequency of the pump according to the load demand, thereby adjusting its heat output. According to the difference between the indoor temperature and the target temperature, the control system will calculate a working frequency suitable for the current load demand. When the load demand is low, the working frequency of the ground heat pump is low, and less heat is output; when the load demand is high, the working frequency increases, and the heat output of the ground heat pump also increases. The relationship between the working frequency and the heat output: The heat output of the ground heat pump is approximately proportional to its working frequency. By optimizing the working frequency, the system can achieve a balance between energy conservation and load adaptability.

[0200] The heat output of the ground heat pump can be approximately estimated by the following relationship: Q output =K heat ×f pump Where: Q output represents the heat output of the ground heat pump; K heat represents a constant related to the characteristics of the pump; f pump represents the working frequency of the ground heat pump. As the frequency is adjusted, the system can flexibly adjust the heat supply according to the actual load demand.

[0201] By continuously monitoring the temperature, load demand, and system operation, and using real-time feedback to adjust the working frequency of the ground heat pump, its operating efficiency can be optimized. The real-time temperature and load data can be fed back to the ground heat pump control system through sensors and the control system. According to the feedback signal, the control system will dynamically adjust: increase or decrease the frequency of the ground heat pump; adjust the heat output to cope with temperature changes.

[0202] Based on real-time feedback, the operating frequency of the ground heat pump is optimized according to the following principles:

[0203] Reduce the frequency when the load demand is low to reduce energy consumption;

[0204] Increase the frequency when the load demand is high to ensure that the temperature demand is met.

[0205] This optimization strategy is achieved through variable frequency control technology, ensuring that the ground heat pump can operate with optimal efficiency under different operating conditions and minimizing energy waste.

[0206] Therefore, based on real-time temperature and load demand, combined with variable frequency control technology and the characteristics of the ground heat pump, flexibly adjust the supply mode and operating frequency of the ground heat pump to ensure the satisfaction of heating (or cooling) load while optimizing energy use efficiency.

[0207] Step S3, combine the power supply distribution output result, the surplus power sale transaction result and the heat output result of the ground heat pump as the operation process optimization strategy to optimize the multi-energy complementary zero-carbon construction process of the toll station.

[0208] In one embodiment, when combining the power supply distribution output result, the surplus power sale transaction result and the heat output result of the ground heat pump as the operation process optimization strategy to optimize the multi-energy complementary zero-carbon construction process of the toll station, the data of various energy systems can be integrated into a unified platform for real-time monitoring and management, including:

[0209] Power supply distribution output result: Data from renewable energy power generation systems such as photovoltaic and wind power, indicating the energy supply situation in each time period;

[0210] Surplus power sale transaction result: Calculate the trading volume and electricity price of the surplus power according to the difference between power supply and demand;

[0211] Heat output result of the ground heat pump: The changes in real-time temperature demand and load demand determine the working state and heat output of the ground heat pump.

[0212] Collect and process this information through intelligent sensors, data acquisition systems and real-time data analysis platforms to ensure that the system always maintains precise control over energy supply and demand.

[0213] The goal of the multi-energy complementary system is to improve energy utilization efficiency and reduce carbon emissions by reasonably configuring different energy systems (such as photovoltaic, wind energy, geothermal energy, etc.). The specific steps are as follows:

[0214] Predict the future load demand of each energy system through historical load demand data and weather prediction models, and optimize the matching relationship of different energy sources according to real-time power supply and load demand data. For example, if the electricity provided by solar and wind energy is sufficient, it can be preferentially used to meet the heating or cooling needs of buildings. The heat output of the heat pump cooperates with the power supply (especially the power from renewable energy sources) to ensure that the load demand of the ground heat pump is met and green power is utilized as much as possible.

[0215] In a multi-energy complementary system, ensuring the stability of power supply is crucial. At the same time, it is necessary to maximize the sale of surplus power to obtain economic benefits. Through means such as battery energy storage systems, store surplus renewable power to cope with insufficient power supply during peak demand periods or unstable weather. Combine power demand forecasting and real-time electricity prices to dynamically decide whether to sell surplus power. For example, when power demand is low and electricity prices are high, more power can be sold to maximize economic benefits. When the energy storage system stores surplus power, it is necessary to release power at an appropriate time to balance the load and avoid power system instability caused by excessive storage or sale.

[0216] The heat output of the ground heat pump needs to be adjusted according to demand and energy supply conditions to ensure comfort while maximizing energy utilization efficiency. Dynamically adjust the operating frequency of the ground heat pump according to real-time power supply and building load demand. If the power supply is relatively sufficient and the price is low, preferentially increase the output of the ground heat pump to meet heating or cooling needs; if power demand is high, the operating frequency of the ground heat pump can be appropriately reduced. Weather changes have a direct impact on the temperature demand of the building, so it is necessary to incorporate meteorological data into the load forecasting model in real time. For example, predict cold days in advance and adjust the ground heat pump in advance to ensure sufficient heating.

[0217] The core goal of the entire optimization process is to maximize the utilization of renewable energy by flexibly dispatching various energy systems, rationally configuring power supply and demand, and combining the heating and cooling load demands of the ground heat pump, and ultimately achieve multi-energy complementarity and zero-carbon goals for toll stations.

[0218] As Figure 2 As shown in the overall architecture diagram of the method for building a multi-energy complementary zero-carbon toll station, the control of the toll station includes a photovoltaic power generation module, a wind power generation module, an energy storage module, power consumption, and an intelligent control system. In the implementation of the photovoltaic power generation module in power supply applications, this embodiment uses high-efficiency monocrystalline silicon photovoltaic modules to ensure the maximum light energy conversion efficiency within a limited space. The power output range of the modules can be flexibly configured according to the actual needs of the toll station. The photovoltaic modules are installed on the roof of the toll station or the ceiling of the parking lot, and an adjustable-angle support structure is used to ensure that the inclination angle of the modules matches the local latitude to maximize light absorption.

[0219] AsFigure 3 And Figure 4 From the photovoltaic power generation module structure diagram and the wind power generation module structure diagram shown, the specific structure of the photovoltaic power generation module includes a photovoltaic module, an inverter, a DC / AC converter, an energy storage system, and a control system; the specific structure of the wind power generation module includes a wind turbine, a generator, a control system, and an energy storage module; the photovoltaic power generation system converts direct current into alternating current through an inverter and connects to the internal power grid of the toll station. The inverter is equipped with an intelligent grid connection control function, which can monitor the power generation status of the photovoltaic system in real time and automatically adjust the grid connection output to meet the power consumption needs of the toll station. At the same time, according to the wind conditions in the area where the toll station is located, a suitable vertical-axis or horizontal-axis wind turbine is selected. The blades of the generator are made of composite materials, with high strength and corrosion resistance, suitable for long-term outdoor use. The wind turbine is installed on a reinforced tower, and the tower foundation adopts the concrete cast-in-place pile technology to ensure the stability of the generator under adverse weather conditions. The tower height is optimized according to the local wind speed conditions to improve the wind capture efficiency. The wind power generation system is equipped with an intelligent control unit, which can automatically adjust the operating state of the generator according to the wind speed change to optimize the power generation efficiency and extend the equipment life. The control system is linked with the energy management module of the toll station to adjust the output power of the wind power generation in real time.

[0220] The energy storage conversion power supply module uses a lithium-ion battery pack as the main energy storage device. Lithium-ion batteries have become an ideal energy storage choice due to their high energy density, long cycle life, and stable performance. The capacity of the battery pack is designed according to the daily power consumption and peak-valley power demand of the toll station. The energy storage system is equipped with an advanced battery management system (BMS), which can monitor key parameters such as the voltage, current, and temperature of the battery in real time to ensure that the battery pack operates within a safe range. The BMS also has an equalization charging function to prevent overcharging or over-discharging of the battery pack, thereby extending the battery life. The energy storage system is connected to the internal power grid of the toll station through a bidirectional inverter. The bidirectional inverter can not only convert the direct current of the energy storage battery into alternating current for the toll station to use, but also store the excess electric energy in the battery when the photovoltaic or wind power generation is excessive. At the same time, it can also automatically dispatch the power output of the photovoltaic, wind power, and energy storage systems according to the real-time energy supply and demand situation to ensure that the power supply of the toll station is always in balance and give priority to using renewable energy to reduce the dependence on the power grid.

[0221] The toll station proposed in this embodiment is equipped with fast charging piles and ordinary charging piles to meet the charging needs of different types of new energy vehicles. The fast charging pile uses high-power DC charging technology, which can provide large-capacity charging for vehicles in a short time, while the ordinary charging pile is suitable for vehicles parked for a long time. The charging pile is connected to the internal power grid of the toll station through an independent power interface. The interface is equipped with a smart meter, which can monitor the power consumption during the charging process in real time and upload the data to the energy management module. The power interface also has a load management function to avoid overload of the power grid caused by large-scale charging.

[0222] The intelligent charging management system can dynamically allocate the output power of charging piles according to the real-time power load in the toll station, giving priority to ensuring the normal operation of other key equipment in the toll station. At the same time, the system supports remote control, and the administrator can view and adjust the operating status of each charging pile in real time through the mobile terminal. The charging pile is equipped with a touch screen user interface, supports multiple payment methods (such as mobile payment, IC card, etc.), and can display real-time charging progress and cost information. The system integrates big data analysis functions, which can predict and optimize based on historical charging data to improve user experience.

[0223] This embodiment integrates the living service areas (such as lounges, restaurants, shops, etc.) in the toll station with the energy management system to ensure that the power supply in these areas comes from renewable energy first. The power supply system has a built-in overload protection device to ensure stable power supply even during peak power demand periods. The lighting system in the living area uses LED energy-saving lamps, which are combined with an intelligent control system to automatically adjust the brightness according to the natural light intensity and personnel activities. Home appliances (such as air conditioners, refrigerators, etc.) are equipped with smart sockets that can automatically cut off the power when the equipment is idle, further reducing energy consumption.

[0224] Each type of equipment in the living service area is equipped with an independent energy consumption monitoring device to record electricity usage in real time. The energy consumption monitoring system is linked to the central energy management module, which can identify high-energy consumption equipment and provide energy-saving suggestions. The system continuously optimizes the energy usage pattern of the living service area through big data analysis and machine learning algorithms. Through the real-time feedback mechanism, the system can automatically adjust the working mode of certain equipment or remind the administrator to intervene manually to ensure that energy consumption is always at the lowest level.

[0225] The photovoltaic and wind power generation systems proposed in this embodiment are equipped with grid-connected inverters, which can convert excess electrical energy into alternating current that meets grid standards. The grid-connected inverters support a variety of grid-connected protocols and can seamlessly connect with the grid-connected requirements set by local power grid companies. The grid-connected inverters have built-in anti-islanding effect protection devices to ensure automatic disconnection in the event of a grid failure to avoid unnecessary impact and risk to the grid. The inverters are also equipped with multiple protection mechanisms such as overvoltage, undervoltage, overcurrent, and short circuit to ensure the safety of the grid-connected process.

[0226] This embodiment integrates a power trading management module. By connecting to the smart meter system of the local power company, it automatically records and uploads the output data of excess power. The system can automatically calculate and determine the time and quantity of power to be sold according to the real-time electricity price or the signed power purchase and sale agreement, maximizing economic benefits.

[0227] According to the actual operation of the toll station, the power sales contract can be dynamically adjusted. For example, the power sales ratio can be increased in seasons with large photovoltaic or wind power generation, and vice versa, the power sales can be reduced or the in-station electricity demand can be preferentially met when the power generation is insufficient. The system also supports negotiating with the power company to establish a customized power purchase and sale contract to ensure the long-term stability of the revenue. To ensure the power quality during the power sales process, the present invention has optimized the design of the power transmission line, adopted low-loss and high-efficiency cable materials, and set a reasonable line layout to reduce the energy loss during power transmission. The system is equipped with a power quality monitoring device that can monitor and adjust key parameters such as the frequency and voltage of the power in real time to ensure that the power output to the power grid meets the national or regional power quality standards. The monitoring data will be fed back to the central control system in real time. If an abnormal situation is detected, the system will automatically adjust the output or give an alarm prompt.

[0228] As Figure 5 It can be seen from the geothermal heating and cooling module shown that the specific structure of the shallow + medium-deep geothermal heating and cooling module includes an energy storage module, geothermal wells, ground heat pumps, and a control system. In this embodiment, shallow geothermal wells are arranged near the toll station, and shallow geothermal resources are collected through ground heat pump technology. The depth of shallow geothermal wells is usually within 100 meters, and the stable temperature below the ground surface is used to heat and cool the buildings of the toll station. In areas where conditions permit, the toll station can further utilize medium-deep geothermal resources. The depth of medium-deep geothermal wells can reach 1000 meters or deeper, which can provide a higher temperature difference for more efficient heating or cooling. Through the reinjection technology, the sustainable utilization of geothermal resources is ensured. Geothermal energy realizes heating and cooling through the ground heat pump system. The ground heat pump system intelligently switches between the heating and cooling modes according to the load demand of the building to provide a constant indoor environment. The system integrates variable frequency control technology and can adjust the heat output in real time according to the indoor temperature to improve the energy utilization efficiency. The geothermal heating and cooling system is linked with the energy management module of the toll station to adjust the intensity of heating and cooling according to seasonal changes and actual needs to ensure the comfort of the indoor environment and at the same time achieve the optimal utilization of energy.

[0229] Furthermore, by comprehensively applying multiple technologies such as photovoltaic power generation, wind power generation, energy storage conversion, geothermal heating and cooling, new energy vehicle charging, and selling excess electricity back to the grid, a self-sufficient, energy-cycling, and highly energy-efficient zero-carbon toll station has been constructed. Through the organic integration and intelligent management of various technologies, the toll station can not only achieve self-balance and independent supply of energy, but also sell excess renewable energy electricity, further improving economic benefits.

[0230] Photovoltaic and wind power generation technologies are adopted to provide the main power source for the toll station, and the energy storage system is used to regulate the power supply stability. Geothermal energy provides efficient heating and cooling for the toll station to ensure the temperature control requirements of the building. Through the energy management system, the production, storage, and use of various types of energy are monitored in real time and dynamically scheduled to achieve the optimal allocation of energy and maximize the energy use efficiency. At the same time, based on selling the excess electricity to the grid, the toll station can not only achieve self-sufficiency, but also create additional income and reduce operating costs. The intelligent management of the power selling system ensures the transparency and efficiency of power transactions, thereby achieving zero carbon emissions at the toll station. By making full use of renewable energy and intelligent management technologies, the dependence on traditional fossil energy is reduced, promoting the goals of environmental protection and sustainable development.

[0231] Through the above detailed technical solutions, this embodiment effectively achieves the invention purpose of constructing a zero-carbon toll station, not only realizing the comprehensive utilization and management of energy technically, but also ensuring efficient operation and the improvement of economic benefits through the intelligent system.

[0232] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An optimization method for the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution, characterized in that, Including: Predict the electricity demand inside the toll station, obtain the power supply allocation output of energy equipment based on the electricity demand, and conduct excess power sales transactions according to the power supply allocation output result and the power purchase and sale agreement; Predict the load demand inside the toll station, adjust the supply mode of the ground source heat pump based on the load demand, and determine the heat output of the ground source heat pump by combining the real-time temperature inside the toll station and the variable frequency control technology; Combine the power supply allocation output result, the excess power sales transaction result and the heat output result of the ground source heat pump as the operation process optimization strategy to optimize the multi-energy complementary zero-carbon construction process of the toll station.

2. The optimization method for the construction process of a multi - energy complementary zero - carbon toll station based on energy distribution according to claim 1, wherein, The predicting the electricity demand inside the toll station, obtaining the power supply allocation output of energy equipment based on the electricity demand, and conducting excess power sales transactions according to the power supply allocation output result and the power purchase and sale agreement includes: Obtain the historical charging data of charging piles inside the toll station and the historical power consumption data of the living service area, and combine the evidence theory and the convolutional neural network technology to predict the electricity demand inside the toll station; Construct a power supply allocation model based on the electricity demand and the dynamic characteristics of energy equipment power supply, determine the power supply allocation output of energy equipment, and judge the remaining energy storage of energy equipment; Combine the remaining energy storage and the energy generation rate to analyze the movable energy of energy equipment at any time on the premise of meeting the electricity consumption of charging piles and the living service area inside the toll station; Combine the movable energy with the power purchase and sale agreement to conduct excess energy sales transaction processing, and determine the sale time and sale quantity during the excess energy sales transaction processing according to the real-time electricity price.

3. An optimization method for the construction process of a multi - energy complementary zero - carbon toll station based on energy distribution according to claim 2, characterized in that, The obtaining the historical charging data of charging piles inside the toll station and the historical power consumption data of the living service area, and combining the evidence theory and the convolutional neural network technology to predict the electricity demand inside the toll station includes: Establish a charging demand matrix based on the number of charging piles inside the toll station, decompose the charging demand matrix into several types of charging modes according to the non-negative matrix factorization method, and judge the charging scale and charging scale of several types of charging modes inside the toll station; Generate a basis matrix and a coefficient matrix according to the charging scale and charging scale, and multiply the rows and columns of the corresponding charging modes in the basis matrix and the coefficient matrix to obtain a dimension matrix; Obtain the historical charging data of charging piles inside the toll station, and use the dimension matrix as an input and substitute it into the graph convolutional neural network to capture the effective information between the charging mode and the charging data, and predict the electricity demand of the charging piles; Obtain the historical power consumption data of the living service area inside the toll station, and obtain the electricity demand in the living area according to the historical power consumption development trend and the evidence theory, and combine the electricity demand of the charging piles to obtain the total electricity demand of the toll station.

4. An optimization method for the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution according to claim 3, characterized in that The obtaining the historical power consumption data of the living service area inside the toll station, and obtaining the electricity demand in the living area according to the historical power consumption development trend and the evidence theory includes: Obtain the historical power consumption data of each sub-area inside the living service area of the toll station, and generate a prediction set of electricity demand in each sub-area according to the historical power consumption data and the seasonal nature; Predict the basic probability assignment matrix according to the prediction set of electricity demand and the nature of sub-area electrical equipment, and use the evidence theory to synthesize the prediction set of electricity demand. Perform evidence conflict analysis and processing on the electricity demand prediction set based on the synthesis result to obtain the belief function and plausibility function of the electricity demand prediction set, and convert the belief function into a probability distribution according to the probability conversion formula; Obtain the probability distribution of the electricity demand in each sub-region according to the distribution result, and combine the probability distribution with the basic probability assignment matrix and the electricity demand of the sub-region, and integrate the electricity demand of each sub-region to obtain the electricity demand in the living area.

5. A method for optimizing the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution according to claim 4, characterized in that, The power supply distribution model is constructed based on the dynamic characteristics of electricity demand and energy equipment power supply, the power supply distribution output of the energy equipment is determined, and the remaining energy storage of the energy equipment is judged, including: According to the power generation characteristics and energy storage characteristics of the energy equipment, analyze the minimum expected frequency of the power supply distribution of any power supply equipment based on the energy storage characteristics, and define the minimum power generation objective function of any power supply equipment using the power generation characteristics; Perform maximum a posteriori hypothesis on the electricity demand of the toll station based on Naive Bayes, smooth the maximum a posteriori hypothesis process, and calibrate the dynamic allocation parameters according to the processing result; Combine the dynamic allocation parameters, the minimum expected frequency, and the minimum power generation objective function to analyze the energy consumption of the energy equipment during the power supply process, and represent the objective function of the power supply distribution model in an octuple manner based on Naive Bayes; Set the boundary conditions of the power supply distribution model based on the dynamic allocation balance of the energy equipment, and combine the objective function to obtain the constructed power supply distribution model; Use the power supply distribution model to determine the power supply distribution output of any power supply equipment, and obtain the remaining energy storage of the energy equipment according to the difference between the energy storage of the energy equipment and the power supply distribution output.

6. The optimization method for the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution according to claim 5, wherein, The expression of the power supply distribution model is: ； Wherein, F(t) represents the power supply distribution of the t-th power supply device, e represents the energy consumption, α1 represents the probability that the power supply distribution output reaches high uniformity, α2 represents the probability that the power supply distribution output reaches low uniformity, υ represents the objective function of the power supply distribution model, β represents the minimum power generation target value of any power supply device, and T min represents the minimum expected frequency, and ε represents the dynamic allocation parameter.

7. A method for optimizing the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution according to claim 6, characterized in that, The process of combining the movable energy quantity with the power purchase and sale agreement to conduct surplus energy sale transactions, and determining the sale time and sale quantity during the surplus energy sale transaction process according to the real-time electricity price, including: Combine the movable energy quantity with the power purchase and sale agreement to define the surplus energy sale rule, connect it with the power company's smart meter system, and automatically record and upload the output data of the surplus energy; Obtain the historical electricity price data, perform phase space reconstruction on the time series of the historical electricity price data, analyze the phase points at any time point of the historical electricity price, and the closest points to the phase points; Obtain the distance between the phase point and its closest point, traverse the distances between all phase points and their corresponding closest points in the phase space to obtain the dynamic evolution law, and analyze the evolution time series of the electricity price based on the dynamic evolution law; Predict the real-time electricity price at any time based on the evolution time series, and determine the sale time and sale quantity of the surplus energy sale transaction according to the level of the real-time electricity price.

8. A method for optimizing the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution according to claim 1, characterized in that Predict the load demand in the toll station, adjust the supply mode of the ground source heat pump based on the load demand, and determine the heat output of the ground source heat pump in combination with the real-time temperature and variable frequency control technology in the toll station, including: Obtain the historical heating and cooling data in the toll station, and predict the heating or cooling load demand of the toll station in any time period in combination with seasonal factors and marginal distribution technology; Determine the temperature demand in the toll station according to the load demand, determine the supply mode of the ground source heat pump based on the temperature demand, and adjust the supply mode of the ground source heat pump through the determination result; Adjust the supply of the ground heat pump according to the real-time temperature and load demand inside the toll station, and calculate the optimal operating frequency of the ground heat pump by combining the variable frequency control technology and the characteristics of the ground heat pump; Output the heat output result of the ground heat pump based on the optimal operating frequency and supply, and adjust the optimal operating frequency according to the real-time feedback to optimize the operating efficiency of the ground heat pump.

9. A method for optimizing the construction process of a multi - energy complementary zero - carbon toll station based on energy distribution according to claim 8, characterized in that, The obtaining of the historical heating and cooling data inside the toll station and the prediction of the heating or cooling load demand of the toll station at any time period by combining seasonal factors and marginal distribution technology include: Obtain the historical heating and cooling data based on the historical operation data of the toll station, construct an estimation sample containing two variables of heating and cooling according to the obtained results, and obtain the distribution function value of the estimation sample under seasonal factors by the non-parametric estimation method; Construct a conditional matrix according to the marginal distribution function value and the estimation sample, and use the conditional matrix to predict the probability of each data value appearing in the estimation sample and the corresponding marginal value; Based on the upper and lower bounds of the preset load demand interval, and combining probability and marginal value to analyze the expected probability of the data value within the load demand interval, and obtain the marginal distribution matrix of the data value by using the expected probability; Perform an inverse operation on the marginal distribution matrix to obtain the demand prediction interval for heating and cooling of the toll station, and determine the heating or cooling load demand of the toll station from the demand prediction interval according to the time period.

10. A method for optimizing the construction process of a multi-energy complementary zero-carbon toll station based on energy distribution according to claim 9, characterized in that, The calculation formulas for the probability of the data value appearing and the marginal value are: ； where S i represents the probability that the i-th data value appears in the estimation sample, T i represents the distribution function value of the i-th data value, P represents the characteristic value of the estimation sample under seasonal factors, E i represents the marginal value of the i-th data value, and H(i, m) represents the joint probability mass function of the data value i and the conditional matrix m, representing the load magnitude of the data value i in the m-th conditional matrix.

Images