Regional load prediction method and device

Abstract

The invention discloses a regional load prediction method and device. The method comprises the steps that historical load data of a land parcel is acquired, and the land parcel is obtained by dividing a region according to load characteristics; predicting the load of the land parcel based on historical load data to obtain a surface load time curve of the land parcel; predicting a total charging load curve of vehicles in the area based on a Monte Carlo algorithm; and superposing the surface load time curve and the total charging load curve by using a load superposition technology to generate a time curve of the regional load. According to the invention, the technical problem of low power load prediction accuracy is solved.

Description

Technical Field

[0001] The present invention relates to the field of power grids, and in particular to a method and device for regional load forecasting. Background Art

[0002] Power load forecasting is an important basis for power system planning and operation. Its purpose is to guide the optimal operation of power production and transmission and distribution systems by analyzing and predicting future power demand. Effective load forecasting results can guide the scheduling and adjustment of power generation resources, ensure the balance of power supply and demand, and avoid power surplus or shortage.

[0003] However, in the relevant schemes, the factors affecting load forecasting are not fully considered, resulting in low accuracy in power load forecasting.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a method and device for regional load forecasting, so as to at least solve the technical problem of low accuracy of power load forecasting.

[0006] According to one aspect of an embodiment of the present invention, a method for regional load prediction is provided, comprising: obtaining historical load data of a plot of land, wherein the plot of land is obtained by dividing a region according to load characteristics; predicting the load of the plot of land based on the historical load data to obtain a surface load time curve of the plot of land; predicting a total charging load curve of vehicles in the region based on a Monte Carlo algorithm; and superimposing the surface load time curve and the total charging load curve using a load superposition technique to generate a time curve of the regional load.

[0007] In one embodiment of the present application, the load of a plot of land is predicted based on historical load data to obtain a surface load time curve of the plot of land, including: determining the maximum load value of the plot of land in multiple time periods based on the historical load data; analyzing the historical load data using a time series analysis method to determine the point load time curve of the plot of land, wherein the point load time curve is used to characterize the load characteristics of the electricity consumption location in the plot of land; and superimposing the maximum load value and the point load time curve to generate a surface load time curve of the plot of land.

[0008] In one embodiment of the present application, a time series analysis method is used to analyze historical load data to determine a point load time curve of a plot of land, including: using a time series analysis method to identify features of historical load data to obtain load change features; using a mathematical function to fit the load change features to obtain a point load time curve.

[0009] In one embodiment of the present application, a time series analysis method is used to identify features of historical load data to obtain load change features, including: inputting historical load data into a prediction model, predicting the periodicity and trend characteristics of load changes through the prediction model, and obtaining load change features.

[0010] In one embodiment of the present application, a total charging load curve of vehicles in an area is predicted based on a Monte Carlo algorithm, including: predicting the charging start time and charging start power of the vehicle, wherein the charging start power is the battery power of the vehicle at the charging start time; determining the charging time of the vehicle based on the charging start time and the charging start power; and determining the total charging load curve of the vehicle according to the charging time.

[0011] In one embodiment of the present application, predicting the charging start time and charging start power of a vehicle includes: determining a first random number and a second random number based on a random number algorithm; predicting the charging start time of the vehicle based on the first random number and a charging start time probability density function; predicting the charging start power of the vehicle based on the second random number and a charging start power probability density function.

[0012] In one embodiment of the present application, the probability density function of the charging start time is obtained by fitting the discrete probability distribution of the historical charging start time by a first function, and the probability density function of the charging start power is obtained by fitting the discrete probability distribution of the historical charging start power by a second function, wherein the discrete probability distribution of the historical charging start time and the discrete probability distribution of the historical charging start power can be determined based on the historical charging records.

[0013] In one embodiment of the present application, the charging time of the vehicle is determined based on the charging start time and the charging start power, including: determining the charging power of the vehicle according to the charging mode; determining the charge state after charging based on the charging start time; determining the charging time of the vehicle based on the charge state after charging, the charging start power, and the charging power.

[0014] In one embodiment of the present application, determining a total charging load curve of a vehicle according to charging time includes: determining a charging load curve of the vehicle according to charging time and charging power; and integrating charging load curves of multiple vehicles to obtain a total charging load curve.

[0015] According to another aspect of an embodiment of the present invention, a regional load prediction device is also provided, including: an acquisition module for acquiring historical load data of a plot of land, wherein the plot of land is obtained by dividing the area according to load characteristics; a first prediction module for predicting the load of the plot of land based on the historical load data to obtain a surface load time curve of the plot of land; a second prediction module for predicting the total charging load curve of vehicles in the area based on a Monte Carlo algorithm; and a superposition module for superimposing the surface load time curve and the total charging load using a load superposition technique to generate a time curve of the regional load.

[0016] In an embodiment of the present invention, historical load data of a plot is obtained, wherein the plot is obtained by dividing the area according to load characteristics; the load of the plot is predicted based on the historical load data to obtain the surface load time curve of the plot; the total charging load curve of vehicles in the area is predicted based on the Monte Carlo algorithm; the surface load time curve and the total charging load curve are superimposed using the load superposition technology to generate a time curve of the regional load. It is easy to notice that in the process of predicting the regional load, while considering the surface load of the plot, the important influencing factor of the charging load of the vehicle is fully considered. By superimposing the surface load time curve of the plot and the total charging load curve of the vehicles in the area, the purpose of predicting the regional load is achieved, thereby achieving the technical effect of improving the accuracy of power load prediction, and then solving the technical problem of low accuracy of power load prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0018] Figure 1 is a flow chart of a regional load forecasting method according to an embodiment of the present invention;

[0019] Figure 2 is a schematic diagram of an optional regional load forecasting method according to an embodiment of the present invention;

[0020] Figure 3 is a flow chart of an optional Monte Carlo algorithm execution according to an embodiment of the present invention;

[0021] Figure 4 is a schematic diagram of a regional load forecasting device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0023] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0024] Example 1

[0025] According to an embodiment of the present invention, an embodiment of a regional load forecasting method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0026] Figure 1 is a flow chart of a regional load forecasting method according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0027] Step S102, obtaining historical load data of a plot, wherein the plot is obtained by dividing the area according to load characteristics.

[0028] The historical load data in the above steps is used to characterize the electricity consumption records at various locations or within various ranges in the plot, and may include electricity consumption records arranged in chronological order, such as hourly, daily or monthly electricity demand, etc., and may also include peak load data for each time period, such as daily, weekly, monthly peak load data, etc., but is not limited to this.

[0029] The load characteristics in the above steps can characterize the characteristics of the power load, including the size, type, time distribution, etc. of the load.

[0030] In an optional embodiment, the historical time period corresponding to the historical load data can be determined according to actual application requirements. For example, the historical time period can be the past year or the past few years, but is not limited thereto. Then, the historical load data corresponding to the historical time period can be retrieved from the database storing the load data. For example, the historical load data can include the daily, weekly, and monthly power consumption, peak load, etc. within the historical time period. The acquired historical load data is cleaned to remove abnormal values ​​and missing values ​​to ensure the integrity and accuracy of the data.

[0031] Step S104, predicting the load of the plot based on the historical load data to obtain a surface load time curve of the plot.

[0032] The surface load time curve in the above steps is a curve showing the change of the overall load of the plot over time. It is used to characterize the trend of the load change over time within the plot. It combines the load time curves of multiple electricity consumption locations within the plot, usually expressed as a daily load curve, weekly load curve or monthly load curve, which can reflect the dynamic changes of the plot load.

[0033] In an optional embodiment, the land parcel can be divided into multiple power consumption locations, and the historical load data can be analyzed using a time series analysis method to determine the load change patterns of multiple power consumption locations over time. Then, the load change patterns of multiple power consumption locations over time are integrated and summarized to obtain the surface load time curve of the land parcel.

[0034] In another optional embodiment, the load prediction model can be trained using the historical load data of the past few years to extract the regularity of the periodic changes of the plot load with the date, so that the load prediction model has the ability to predict the plot load according to the date. Then, the date for which the plot load needs to be predicted is input into the load prediction model, and the load prediction model is used to predict the plot load corresponding to multiple dates, and then the surface load time curve of the plot is constructed based on the plot load corresponding to multiple dates.

[0035] Step S106: predicting the total charging load curve of vehicles in the area based on the Monte Carlo algorithm.

[0036] The vehicle in the above steps may be an electric car that needs to be charged.

[0037] The Monte Carlo algorithm in the above steps is an algorithm that uses random sampling and statistical analysis to solve problems. It predicts the charging load of the vehicle and generates a total charging load curve by simulating the random process of uncertain factors in the load forecasting process, where the uncertain factors can be the charging behavior of the vehicle.

[0038] The total charging load curve in the above steps can represent the changing trend of the charging load of vehicles in the area over time.

[0039] In an optional embodiment, based on the charging record data of the electric vehicle, such as the charging start time, end time, charging power and vehicle ownership, the charging behavior of the vehicle is modeled to obtain a charging behavior model. Then, the charging start time and the battery power of the vehicle corresponding to the start time are randomly sampled multiple times. According to the results of the multiple random samplings and the charging behavior model, the charging behavior of different vehicles at different times is simulated to obtain multiple charging load curves. Multiple charging load curves are superimposed to obtain a total charging load curve.

[0040] Step S108, using load superposition technology to superimpose the surface load time curve and the total charging load curve to generate a time curve of the regional load.

[0041] The load superposition technology in the above steps is a technology for superimposing different types of loads in the time dimension. When superimposing the loads, it is necessary to consider the mutual influence between the loads and the alignment of time.

[0042] The regional load time curve in the above steps is the result of superimposing the surface load time curve of the plot and the total charging load curve of the vehicles in the region. It comprehensively shows the change of the total load in the region over time and provides key information for power system planning and scheduling.

[0043] In an optional embodiment, when the surface load time curve and the total charging load curve are superimposed, the surface load time curve and the total charging load curve need to be aligned in the time dimension, and can be synchronized at a time interval of hours or less. Then, the surface load time curve and the total charging load curve are added at each time point to obtain a preliminary regional total load time curve.

[0044] Then, considering that the surface load and the total charging load may not reach their peak values ​​at the same time, it is necessary to calculate the simultaneity rate, that is, the probability that the two loads reach their maximum values ​​at the same time, in order to more accurately predict the peak load. The preliminary regional total load time curve is adjusted based on the simultaneity rate to obtain the regional total load time curve. The regional total load time curve is then smoothed to eliminate abnormal points caused by data fluctuations and improve the prediction accuracy of the regional load.

[0045] In an embodiment of the present invention, historical load data of a plot is obtained, wherein the plot is obtained by dividing the area according to load characteristics; the load of the plot is predicted based on the historical load data to obtain the surface load time curve of the plot; the total charging load curve of vehicles in the area is predicted based on the Monte Carlo algorithm; the surface load time curve and the total charging load curve are superimposed using the load superposition technology to generate a time curve of the regional load. It is easy to notice that in the process of predicting the regional load, while considering the surface load of the plot, the important influencing factor of the charging load of the vehicle is fully considered. By superimposing the surface load time curve of the plot and the total charging load curve of the vehicles in the area, the purpose of predicting the regional load is achieved, thereby achieving the technical effect of improving the accuracy of power load prediction, and then solving the technical problem of low accuracy of power load prediction.

[0046] In one embodiment of the present application, the load of a plot of land is predicted based on historical load data to obtain a surface load time curve of the plot of land, including: determining the maximum load value of the plot of land in multiple time periods based on the historical load data; analyzing the historical load data using a time series analysis method to determine the point load time curve of the plot of land, wherein the point load time curve is used to characterize the load characteristics of the electricity consumption location in the plot of land; and superimposing the maximum load value and the point load time curve to generate a surface load time curve of the plot of land.

[0047] The maximum load values ​​of the plots in multiple time periods in the above steps are in the historical load data. For each plot, the highest electricity demand recorded in different time periods can reflect the peak load conditions of the plot in a specific time frame. For example, the time period can be hours, daytime, nighttime, weekdays, weekends, four seasons, etc.

[0048] The time series analysis method in the above steps is a statistical analysis method used to study the changing trend of historical load data. The time series analysis method can perform trend analysis, seasonal analysis, periodic analysis, random analysis, etc. on historical load data. The time series analysis method may include but is not limited to autoregressive model, moving average model, autoregressive moving average model, autoregressive integrated moving average model, etc.

[0049] The point load time curve in the above steps describes the changing trend of the power demand of a specific power consumption point in the plot over time, and can reflect the load fluctuation of the power consumption point within a day, a week or a month.

[0050] In an optional embodiment, the load peaks in each time period can be identified by statistically analyzing the historical load data. For example, for the maximum daily load value, the maximum load value of each day can be found from the historical load data to obtain 365 maximum daily load values; for the maximum seasonal load value, the maximum load values ​​in the four seasons of spring, summer, autumn and winter can be found from the historical load data to obtain the maximum seasonal load value.

[0051] When using time series analysis methods to analyze historical load data, the time series analysis methods that can be used may include but are not limited to ARIMA (AutoRegressive Integrated Moving Average), SARIMA (Seasonal AutoRegressive Integrated Moving Average) or Prophet (Prophet prediction model), etc. The time series analysis method is used to capture the time pattern of the load, such as daily periodicity, seasonal fluctuations, etc., and the load forecasting model is used to predict future load demand based on the time pattern, obtain the load value of each time period, and form a point load time curve.

[0052] When superimposing the maximum load value and the point load time curve, time alignment is required to ensure that the time axis of the maximum load value and the point load time curve are consistent, and the maximum load value is assigned to the time point when the maximum load value occurs. At each time point, the load value in the point load time curve is superimposed with the maximum load value at that time point. If the maximum load value is significantly higher than the point load prediction value, the interpolation method can be used to smooth the maximum value to the prediction value at the adjacent time point to generate a continuous and accurate surface load time curve.

[0053] It is also possible to consider the situation that the surface load and the total charging load may not reach the peak value at the same time. The probability that the two loads reach the maximum value at the same time, that is, the simultaneity rate, is calculated, and the surface load time curve is fine-tuned based on the simultaneity rate to make the surface load time curve more accurate.

[0054] The surface load time curve of the plot obtained through the above steps will cover the load characteristics and maximum load information of all power consumption points in the plot, which can provide a more detailed load forecast result for the power system and support more accurate power resource planning and scheduling.

[0055] In one embodiment of the present application, a time series analysis method is used to analyze historical load data to determine a point load time curve of a plot of land, including: using a time series analysis method to identify features of historical load data to obtain load change features; using a mathematical function to fit the load change features to obtain a point load time curve.

[0056] The load change characteristics in the above steps refer to a series of statistical and mathematical characteristics extracted from historical load data that reflect the time-varying patterns of electricity demand. These characteristics may include but are not limited to load periodicity, trend, seasonal fluctuations, differences between weekdays and weekends, holiday effects, etc.

[0057] The mathematical function in the above steps can convert the extracted load variation characteristics into a mathematical expression for generating a load time curve. The mathematical function can be a polynomial function, a Fourier series, or an exponential smoothing function, but is not limited thereto.

[0058] In an optional embodiment, when performing feature identification, trend feature identification can be performed on historical load data, and whether there is a linear or curvilinear trend in the historical load data can be identified by calculating the autocorrelation diagram and partial autocorrelation diagram of the first-order or high-order difference sequence, so as to obtain load change characteristics, wherein the autocorrelation diagram and partial autocorrelation diagram are tools in the time series analysis method. Periodic feature identification can also be performed on historical load data, and whether there is a short-term or long-term repetitive pattern can be identified by the autocorrelation diagram in the time series analysis method, so as to obtain load change characteristics. For example, the short period can be 24 hours, 48 ​​hours, etc., and the long period can be 7 days, 30 days, etc., but is not limited thereto.

[0059] Then, mathematical functions are used to fit the load change characteristics. Multiple mathematical functions are used to fit the load change characteristics. After repeated trials and adjustments, the function that best fits the load change characteristics is determined. Then the parameters of the data function are selected and optimized to minimize the prediction error and obtain the point load time curve.

[0060] In one embodiment of the present application, a time series analysis method is used to identify features of historical load data to obtain load change features, including: inputting historical load data into a prediction model, predicting the periodicity and trend characteristics of load changes through the prediction model, and obtaining load change features.

[0061] The prediction model in the above steps is used to predict future load change characteristics based on historical data, and can be a SARIMA (Seasonal Autoregressive Integrated Moving Average model) model or a neural network model based on LSTM (Long Short-Term Memory), but is not limited thereto.

[0062] In an optional embodiment, the historical load data may be converted into a time series format, including a timestamp and a load value corresponding to the timestamp. The historical load data in the time series format is then input into a prediction model, and the prediction model predicts the periodicity and trend characteristics of the load change to obtain the load change characteristics.

[0063] In one embodiment of the present application, a total charging load curve of vehicles in an area is predicted based on a Monte Carlo algorithm, including: predicting the charging start time and charging start power of the vehicle, wherein the charging start power is the battery power of the vehicle at the charging start time; determining the charging time of the vehicle based on the charging start time and the charging start power; and determining the total charging load curve of the vehicle according to the charging time.

[0064] The charging start time in the above steps refers to the specific time point when the electric vehicle starts charging.

[0065] The charging starting power in the above steps refers to the battery state of charge (SOC) of the electric vehicle when charging starts.

[0066] The charging time in the above steps refers to the time required from the start of charging to the end of charging of the electric vehicle. The charging time will be affected by the starting charge, target charge, charging power and charging efficiency. The charging time determines the distribution of the charging load on the time axis in the total charging load curve.

[0067] In an optional embodiment, the historical charging records of the vehicle may be collected, including data such as the charging start time, the starting power, and the charging duration, to identify the regularity and pattern of the charging behavior. The data in the historical charging records are then fitted to the statistical distribution to construct a probability distribution model of the charging start time and the charging start power, wherein the statistical distribution may be a normal distribution or an exponential distribution, but is not limited thereto. The probability distribution model is then used to randomly extract the charging start time and the charging start power through the Monte Carlo simulation method to simulate the actual charging behavior of the electric vehicle.

[0068] Then, the charging duration is determined based on the charging start time, charging start power, charging power, battery capacity, and charging efficiency. A load curve is created for each vehicle, and the time range of the load curve is determined based on the charging start time and charging duration. During each charging time period, the vehicle's charging power is applied to the load curve to form the charging load value at each time point. The charging load curves of all vehicles are superimposed to generate a total charging load curve.

[0069] In one embodiment of the present application, predicting the charging start time and charging start power of a vehicle includes: determining a first random number and a second random number based on a random number algorithm; predicting the charging start time of the vehicle based on the first random number and a charging start time probability density function; predicting the charging start power of the vehicle based on the second random number and a charging start power probability density function.

[0070] The random number algorithm in the above steps is a mathematical algorithm used to generate pseudo-random numbers. These pseudo-random numbers are statistically random. In practical applications, the repeatability of the results can be ensured by setting the seed. In Monte Carlo simulation, random numbers are used to simulate uncertain random events, such as the randomness of charging behavior.

[0071] The first random number and the second random number in the above steps are values ​​generated by a random number algorithm, and are used to predict the charging start time and the charging start power respectively.

[0072] The probability density function of the charging start time in the above steps describes the probability distribution of the time when the vehicle starts charging within a period of time. For example, the probability density function of the charging start time may show that the probability of charging the vehicle at night is higher, and the probability of charging during the day is lower. For example, the period of time can be one day or one week, but not limited to this. The probability density function of the charging start time is constructed based on the historical charging records. By analyzing the charging start time of a large amount of data in the historical charging records, a probability distribution model is fitted. The form of the probability density function of the charging start time can be a normal distribution, a Poisson distribution, or a multiple mixed distribution, etc., but not limited to this.

[0073] The charge start power probability density function in the above steps describes the probability distribution of the vehicle battery power state when charging starts. For example, some vehicles may charge when the SOC is low, while other vehicles may start charging when they are close to exhaustion or at a specific SOC. This function is also based on historical charging records. By counting the number of times charging starts at different SOCs, a probability distribution model is fitted. The charge start power probability density function can be in the form of a normal distribution, a Poisson distribution, or a multiple mixed distribution, but is not limited to this.

[0074] In an optional embodiment, a first random number and a second random number are generated using a random number algorithm, and then the first random number is input into a probability density function of the charging start time to obtain a simulation value of the charging start time, and the simulation value of the charging start time is used as the charging start time of the vehicle. Similarly, the second random number is input into a probability density function of the charging start power to obtain a simulation value of the charging start power, and the simulation value of the charging start power is used as the charging start power of the vehicle.

[0075] In one embodiment of the present application, the probability density function of the charging start time is obtained by fitting the discrete probability distribution of the historical charging start time by a first function, and the probability density function of the charging start power is obtained by fitting the discrete probability distribution of the historical charging start power by a second function, wherein the discrete probability distribution of the historical charging start time and the discrete probability distribution of the historical charging start power can be determined based on the historical charging records.

[0076] The first function in the above steps is a function used to fit the discrete probability distribution of the historical charging start time. A suitable mathematical function can be selected as the first function according to practical applications, and the specific content of the first function is not limited here.

[0077] The second function in the above steps is a function used to fit the discrete probability distribution of the historical charging start power. A suitable mathematical function can be selected as the second function according to practical applications, and the specific content of the second function is not limited here.

[0078] The discrete probability distribution in the above steps is to group the data according to certain intervals or categories, and then count the frequency or probability of the data in each group.

[0079] The historical charging records in the above steps refer to detailed information about the charging of electric vehicles at different charging piles over a period of time in the past, including the start and end time of charging, the power at the start and end of charging, the charging location, the charging power, etc. Among them, the historical charging start time is the charging start time in the historical charging records, and the historical charging start power is the charging start power in the historical charging records.

[0080] In an optional embodiment, when constructing the probability density function of the charging start time, the historical charging records of the vehicle may be collected, the historical charging start time in the historical charging records may be extracted, and the historical charging start time may be cleaned to remove outliers. A day is divided into multiple time periods. For example, each 15 minutes may be divided into a segment, with a total of 96 segments in a day, but the segmentation method may not be limited thereto. Then, based on the historical charging start time after data cleaning, the frequency of the number of charging starts in each time period is determined, and a discrete probability distribution of the historical charging start time is generated. Subsequently, the discrete probability distribution of the historical charging start time is fitted to a continuous probability density function using the first function to obtain a probability density function of the charging start time.

[0081] When constructing the probability density function of the charge start power, the historical charge start power can be extracted from the historical charging records, and the historical charge start power can be cleaned to remove outliers. The SOC is then divided into multiple segments, illustratively, one segment every 5%, but the segmentation method may not be limited to this. Then, based on the historical charge start power after data cleaning, the frequency of the historical charge start power distributed in each SOC segment is calculated to generate a discrete probability distribution of the historical charge start power. Subsequently, the discrete probability distribution of the historical charge start power is fitted to a continuous probability density function using a second function to obtain a charge start power probability density function.

[0082] In one embodiment of the present application, the charging time of the vehicle is determined based on the charging start time and the charging start power, including: determining the charging power of the vehicle according to the charging mode; determining the charge state after charging based on the charging start time; determining the charging time of the vehicle based on the charge state after charging, the charging start power, and the charging power.

[0083] The charging modes in the above steps can be slow charging, fast charging, super fast charging, wireless charging, etc., among which slow charging uses a lower power AC power supply for charging, and the charging time is longer, which is suitable for charging at night or in spare time; fast charging uses a high-power DC power supply for charging, which can quickly replenish the battery power, but the cost is high and the battery life may be affected; super fast charging is a faster charging method than fast charging, usually with higher power and shorter charging time; wireless charging is a charging method that does not require physical contact, which is suitable for specific occasions, but the efficiency and power may be lower.

[0084] The charging power in the above steps is the electric power obtained by the vehicle from the charging facility during the charging process, and the unit is usually kilowatts (kW). The charging power is determined by the charging mode. The charging power of the slow charging mode is usually lower, while the charging power of the fast charging mode is higher. The charging power directly affects the charging time, that is, the higher the charging power, the shorter the charging time.

[0085] The state of charge after charging in the above steps is the state of charge of the battery after the vehicle is fully charged, usually expressed as a percentage of the maximum capacity of the battery.

[0086] In an optional embodiment, when determining the charging power of the vehicle according to the charging mode, the corresponding charging power can be determined according to the charging mode. The determination method can be to find statistical information in the vehicle manual, industry standards or historical charging data, but the method for determining the charging power may not be limited to this. Since the charging start time determines the charging time window of the vehicle in a day, the charging start time will affect the SOC after charging. For example, if the charging start time is at night, the user usually hopes that the vehicle can be fully charged, and the state of charge after charging can be set to 100%. If the charging start time is during the day, then considering that the user may use the vehicle, the state of charge after charging can be set to less than 100%, such as 80%. Therefore, the state of charge after charging can be determined at the charging start time.

[0087] Then, the charging time is calculated as follows:

[0088]

[0089] Among them, T c is the charging time, α is the state of charge after charging, SOC is the starting charge, P c is the charging power, E h is the battery capacity, which is the maximum amount of electrical energy that the vehicle battery can store, usually measured in kilowatt-hours (kWh). σ is the charging efficiency, which is the ratio of the electrical energy actually charged into the battery during the charging process to the total electrical energy output by the charging facility.

[0090] In one embodiment of the present application, determining a total charging load curve of a vehicle according to charging time includes: determining a charging load curve of the vehicle according to charging time and charging power; and integrating charging load curves of multiple vehicles to obtain a total charging load curve.

[0091] The charging load curve in the above steps is a curve obtained by using the Monte Carlo algorithm model for each charging behavior of the vehicle. A single charging load curve can only represent one charging behavior of the vehicle. Therefore, the charging load curves of multiple vehicles should be integrated, and the total charging load curve obtained after integration should be used to represent the load generated by the charging behavior of all vehicles in the area.

[0092] In an optional embodiment, the charging time can be divided into multiple time intervals (such as 15 minutes or less), and in each time interval, the charging load of the vehicle is equal to the charging power. The charging load curve is constructed with time as the horizontal axis and charging load as the vertical axis. The time points from the start to the end of charging are determined according to the charging time and the start time of charging. At each time point from the start to the end of charging, the value of the load curve is the charging power at the current time point. By drawing points and connecting lines, the construction of the total charging load curve is completed.

[0093] When integrating the charging load curves of multiple vehicles to obtain the total charging load curve, it is necessary to initialize the total load curve and create a zero-value array on the time axis to store the total charging load curve. Then, for each vehicle's charging load curve, traverse each time point on the time axis and add the charging load value at that time point to the corresponding time point of the total load curve. After superimposing all vehicle load curves, the simultaneity rate needs to be considered. The curve obtained by superimposing all vehicle load curves can be adjusted according to the simultaneity rate to obtain the total charging load curve, which reflects the comprehensive performance of the charging load of all vehicles in the area.

[0094] A preferred embodiment is described below. Figure 2 is a schematic diagram of an optional regional load forecasting method according to an embodiment of the present invention. Figure 2 As shown, in the process of regional load forecasting, plot load forecasting can be performed to obtain a surface load time curve, and vehicle load forecasting based on the Monte Carlo algorithm can be performed at the same time. The typical daily load curve of the vehicle is obtained by predicting the vehicle charging behavior and the charging pile data. The typical daily load curve of the vehicle is also the total charging load curve of the vehicles in the region mentioned above. The surface load time curve and the typical daily load curve of the vehicle are superimposed and integrated to obtain the regional total load curve, where the regional total load curve is also the time curve of the regional load mentioned above.

[0095] Figure 3 is a flowchart of an optional Monte Carlo algorithm execution according to an embodiment of the present invention, such as Figure 3 As shown:

[0096] Step S302, setting the upper limit of the number of cycles and the number of vehicles in stock.

[0097] Step S304, initializing the current loop number to 1.

[0098] Step S306, initializing the current vehicle number to 0.

[0099] Step S308, adding 1 to the current vehicle number.

[0100] Step S310, extracting the probability density of the corresponding charging start time.

[0101] Step S312, randomly simulate the charging starting power and charging power, and calculate the charging time.

[0102] Step S314, calculating the charging probability and charging load at each moment to obtain a charging load curve.

[0103] Step S316, superimposing the charging load curve to obtain a total charging load curve.

[0104] That is, multiple charging load curves are superimposed to obtain a total charging load curve.

[0105] Step S318, determine whether the current vehicle number is greater than the number of vehicles in stock.

[0106] If the judgment result is yes, then continue to execute step S320; if the judgment result is no, then execute step S308.

[0107] In step S320, it is determined whether the total charging load curve has converged.

[0108] If the total charging load curve converges, step S326 is executed, that is, the curve is output, and the process ends; if the total charging load curve does not converge, step S322 is executed.

[0109] Step S322, determine whether the current number of cycles plus 1 is greater than the number of vehicles in stock.

[0110] If it is greater, execute step S326 to output the curve and end the process; if it is less, execute step S324.

[0111] Step S324, add 1 to the current loop count.

[0112] After executing step S324, steps S306 to S318 are re-executed, and then the next step to be executed is determined according to the judgment result of S318.

[0113] Example 2

[0114] According to an embodiment of the present invention, an embodiment of a regional load forecasting device is provided, which can execute the regional load forecasting method provided in the above-mentioned embodiment 1. The specific implementation method and preferred application scenario are the same as those of the above-mentioned embodiment 1 and will not be repeated here.

[0115] Figure 4 is a schematic diagram of a regional load forecasting device according to an embodiment of the present invention. Figure 4 As shown, the regional load forecasting device includes:

[0116] An acquisition module 40 is used to acquire historical load data of a plot, wherein the plot is obtained by dividing the area according to load characteristics;

[0117] A first prediction module 42 is used to predict the load of the plot based on historical load data to obtain a surface load time curve of the plot;

[0118] A second prediction module 44 is used to predict a total charging load curve of vehicles in a region based on a Monte Carlo algorithm;

[0119] The superposition module 46 is used to superimpose the surface load time curve and the total charging load using the load superposition technology to generate a time curve of the regional load.

[0120] Among them, the first prediction module includes: a first determination unit, which is used to determine the maximum load of the plot in multiple time periods based on historical load data; a second determination unit, which is used to analyze the historical load data using a time series analysis method to determine the point load time curve of the plot, wherein the point load time curve is used to characterize the load characteristics of the electricity consumption location in the plot; a generation unit, which is used to superimpose the maximum load value and the point load time curve to generate the surface load time curve of the plot.

[0121] The second determination unit is also used to perform feature recognition on the historical load data using a time series analysis method to obtain load change characteristics; and to fit the load change characteristics using a mathematical function to obtain a point load time curve.

[0122] The second determination unit is further used to input the historical load data into the prediction model, and predict the periodicity and trend characteristics of the load change through the prediction model to obtain the load change characteristics.

[0123] The second prediction module includes: a prediction unit, used to predict the charging start time and charging start power of the vehicle, wherein the charging start power is the battery power of the vehicle at the charging start time; a third determination unit, used to determine the charging time of the vehicle based on the charging start time and the charging start power; and a fourth determination unit, used to determine the total charging load curve of the vehicle according to the charging time.

[0124] The prediction unit is also used to determine a first random number and a second random number based on a random number algorithm; predict the charging start time of the vehicle based on the first random number and a probability density function of the charging start time; and predict the charging start power of the vehicle based on the second random number and a probability density function of the charging start power.

[0125] The probability density function of the charging start time in the prediction unit is obtained by fitting the discrete probability distribution of the historical charging start time through the first function, and the probability density function of the charging start power is obtained by fitting the discrete probability distribution of the historical charging start power through the second function, wherein the discrete probability distribution of the historical charging start time and the discrete probability distribution of the historical charging start power can be determined based on the historical charging records.

[0126] The third determination unit is also used to determine the charging power of the vehicle according to the charging mode; determine the state of charge after charging based on the charging start time; and determine the charging time of the vehicle based on the state of charge after charging, the charging start power, and the charging power.

[0127] The fourth determination unit is further used to determine a charging load curve of the vehicle according to the charging duration and the charging power; and integrate the charging load curves of multiple vehicles to obtain a total charging load curve.

[0128] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0129] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0130] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0131] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0132] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0133] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0134] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A regional load forecasting method, characterized in that: include: Obtaining historical load data of a plot, wherein the plot is obtained by dividing an area according to load characteristics; Predicting the load of the plot based on the historical load data to obtain a surface load time curve of the plot; Predicting a total charging load curve of vehicles in the area based on a Monte Carlo algorithm; The surface load time curve and the total charging load curve are superimposed using load superposition technology to generate a time curve of the regional load.

2. The regional load forecasting method according to claim 1, characterized in that: Predicting the load of the plot based on the historical load data to obtain a surface load time curve of the plot includes: Determine the maximum load of the plot in multiple time periods based on the historical load data; Analyzing the historical load data using a time series analysis method to determine a point load time curve of the plot, wherein the point load time curve is used to characterize the load characteristics of the electricity consumption location in the plot; The maximum load value and the point load time curve are superimposed to generate the surface load time curve of the plot.

3. The regional load forecasting method according to claim 2, characterized in that: The historical load data is analyzed using a time series analysis method to determine a point load time curve of the plot, including: Using the time series analysis method to perform feature recognition on the historical load data to obtain load change features; The load variation characteristics are fitted using a mathematical function to obtain the point load time curve.

4. The regional load forecasting method according to claim 3, characterized in that: The historical load data is characterized by using a time series analysis method to obtain load change characteristics, including: The historical load data is input into a prediction model, and the periodicity and trend characteristics of load changes are predicted by the prediction model to obtain the load change characteristics.

5. The regional load forecasting method according to claim 1, characterized in that: The total charging load curve of vehicles in the area is predicted based on the Monte Carlo algorithm, including: Predicting a charging start time and a charging start power of the vehicle, wherein the charging start power is the battery power of the vehicle at the charging start time; Determining a charging time of the vehicle based on the charging start time and the charging start power; A total charging load curve of the vehicle is determined according to the charging duration.

6. The regional load forecasting method according to claim 5, characterized in that: Predicting the charging start time and charging start power of the vehicle, including: Determine a first random number and a second random number based on a random number algorithm; Predicting a charging start time for the vehicle based on the first random number and a charging start time probability density function; The charging start power of the vehicle is predicted based on the second random number and a charging start power probability density function.

7. The regional load forecasting method according to claim 6, characterized in that: The probability density function of the charging start time is obtained by fitting the discrete probability distribution of the historical charging start time through a first function, and the probability density function of the charging start power is obtained by fitting the discrete probability distribution of the historical charging start power through a second function, wherein the discrete probability distribution of the historical charging start time and the discrete probability distribution of the historical charging start power can be determined based on historical charging records.

8. The regional load forecasting method according to claim 5, characterized in that: Determining the charging time of the vehicle based on the charging start time and the charging start power includes: Determining a charging power for the vehicle according to a charging mode; Determining a state of charge after charging based on the charging start time; The charging time of the vehicle is determined based on the state of charge after charging, the charging start power, and the charging power.

9. The regional load forecasting method according to claim 5, characterized in that: Determining a total charging load curve of the vehicle according to the charging duration includes: Determining a charging load curve of the vehicle according to the charging time and the charging power; The charging load curves of the plurality of vehicles are integrated to obtain the total charging load curve.

10. A regional load forecasting device, characterized in that: include: An acquisition module, used for acquiring historical load data of a plot, wherein the plot is obtained by dividing the area according to load characteristics; A first prediction module, used for predicting the load of the plot based on the historical load data to obtain a surface load time curve of the plot; A second prediction module, used for predicting a total charging load curve of vehicles in the area based on a Monte Carlo algorithm; The superposition module is used to superimpose the surface load time curve and the total charging load curve using load superposition technology to generate a time curve of the regional load.

Images